CZECH NATIONAL COMMITTEE FOR COOPERATION

WITH THE INTERNATIONAL INSTITUTE OF REFRIGERATION (IIR)

ACADEMY OF SCIENCES OF THE CZECH REPUBLIC
CZECH INDUSTRIAL GAS ASSOCIATION
FACULTY OF MECHANICAL ENGINEERING OF THE CZECH TECHNICAL UNIVERSITY
CZECH ASSOCIATION OF MATHEMATICIANS AND PHYSICISTS

10th CRYOGENICS 2008

IIR International Conference
Commissions A1, A2 and C1
Praha, Czech Republic
April 21 25, 2008

PROCEEDINGS

INSTITUT INTERNATIONAL DU FROID

INTERNATIONAL INSTITUTE OF REFRIGERATION

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system, or transmitted in any form or by any means, electronics,
mechanical, photocopying, recording or otherwise, without the prior permission of
the authors.

Published by
Icaris Ltd., Conference Management
Nm. Dr. Holho 8
180 00 Praha 8
Czech Republic
www.icaris.cz

ISBN 978-2-913149-62-5
ISSN 0151-1637

CONTENTS
COMMITTEES...................................................................................................................21
LIST OF EXHIBITORS.....................................................................................................23

OPENNING
LECTURE...................................................................................................25
CR08-63................................................................................................................................27
THE IIR, YESTERDAY, TODAY AND TOMORROW................................................27
Coulomb D. .................................................................................................................27

HELIUM
LIQUEFACTION.......................................................................................35
CR08-62................................................................................................................................37
THE CENTENARY OF THE FIRST LIQUEFACTION OF HELIUM.........................37
Scurlock R.G................................................................................................................37
CR08-56................................................................................................................................45
FIFTY YEARS FROM HELIUM LIQUEFACTION
IN CZECHOSLOVAKIA
AND A NEW TURBINE TECHNOLOGY.....................................................................45
Kaiser Z.1, Kouba M.2, Kundera R.3, Pruk J.4, afrata S.4, Schustr P.5, Chrz V.2,
.....................................................................................................................................45
CR08-17................................................................................................................................53
IMPROVEMENTS OF HELIUM LIQUEFACTION / REFRIGERATION PLANTS
AND APPLICATIONS....................................................................................................53
Muehlegger M., Berdais K.-H., Wilhelm H., Ungricht Th..........................................53
CR08-47................................................................................................................................59
EVOLUTION OF THE STANDARD HELIUM LIQUEFIER RANGE........................59
Caillaud A., Aigouy G., Crispel S., Delcayre F., Grabi V., Dauguet P.....................59
CR08-40................................................................................................................................67
LIQUID HELIUM IN LABORATORY USE PRACTICAL REMARKS...................67
Haberstroh Ch..............................................................................................................67

PULSE TUBES
AND
OTHER REFRIGERATORS....................................................................75
CR08-31................................................................................................................................77
EFFECT OF ALTERNATE TUBE CHARACTERISTICS ON HIGH CAPACITY
PULSE TUBE CRYOCOOLERS PERFORMANCE.....................................................77
Saidi M.H., Sarikhani N., Jafarian A., Hannani S.K...................................................77
CR08-58................................................................................................................................85
GREEN CRYOGENICS: THE USE OF NATURAL CONVECTION TO IMPROVE
THE EFFICIENCY OF CRYOGENS AND CRYOCOOLERS.....................................85
Scurlock R.G.1, Wang C.2..........................................................................................85
5

CR08-12................................................................................................................................91
THE DEVELOPMENT OF A VUILLEUMIER CRYOCOOLER FOR NEW
ZEALANDS HIGH TEMPERATURE SUPERCONDUCTOR INDUSTRY...............91
Gschwendtner M.A., Tucker A.S................................................................................91
CR08-20................................................................................................................................99
THE EFFICIENT MANAGEMENT OF LIQUID HELIUM AT SOUTH POLE
STATION DURING THE AUSTRAL WINTER ..........................................................99
Baker R.A., Sullivan P.................................................................................................99

HELIUM
TEMPERATURE
TECHNIQUES..........................................................................................107
CR08-04..............................................................................................................................109
CRYOGENIC SYSTEM
OF THE SWISS ULTRA-COLD NEUTRON SOURCE..............................................109
Anghel A.1, Blau B.1, Daum M.1, Kirch K.1, Grigoriev S.2...................................109
CR08-54..............................................................................................................................117
EXPERIMENTAL SET-UP OF HEAT TRANSFER MEASUREMENTS IN HE II...117
Chorowski M., Fydrych J., Strychalski M.................................................................117
CR08-39..............................................................................................................................125
S-N-S PHASE TRANSITIONS OF GEOMETRICALLY-METASTABLE
SUPERCONDUCTING THIN FILMS..........................................................................125
Ribeiro Gomes M......................................................................................................125
CR08-07..............................................................................................................................133
BLACK SURFACES FOR CRYOGENIC APPLICATIONS.......................................133
Krlk T., Hanzelka P., Musilov V., Srnka A..........................................................133

HIGH-TEMPERATURE
SUPERCONDUCTIVITY.......................................................................141
CR08-11..............................................................................................................................143
25 TESLA HTS MAGNET INSERT COIL IN ZERO BOIL OFF CRYOSTAT ........143
Good J., Bracanovic D...............................................................................................143

GAS SEPARATION
AND
LIQUEFACTION.....................................................................................151
CR08-59..............................................................................................................................153
LIQUID DISTRIBUTION FROM STRUCTURED PACKINGS AND
DISTRIBUTORS UNDER TILT AND MOTION RELEVANT TO FLOATING
CRYOGENIC AIR SEPARATION PLANTS..............................................................153
Kalbassi M.A.1, Waldie B.2, White V.1, Bell C.2....................................................153
CR08-43..............................................................................................................................161
COMPLEX SEPARATION OF MULTICOMPONENT FLOWS TO EXTRACT
INDUSTRIAL AND INERT GASES............................................................................161
6

Bondarenko V. L.1, Losyakov N. P.2, Simonenko O. Yu.2......................................161

CR08-41..............................................................................................................................167
SOLUBILITY OF PROPANE AND ETHANE IN LIQUID OXYGEN......................167
Houssin-Agbomson D.1, Arpentinier P.1, Delcorso F.1, Coquelet C.2, Richon D.2
...................................................................................................................................167
CR08-64..............................................................................................................................175
MODELING HEAT-MASS TRANSFER PROCESSES ON REGULAR PACKINGS
OF DISTILATION PLANTS.........................................................................................175
Arkharov I., Navasardyan E......................................................................................175

STORAGE
AND
TRANSPORT
OF INDUSTRIAL GASES.......................................................................183
CR08-42..............................................................................................................................185
OPERATION OF SMALL AND HIGH PRESSURE TANKS FOR LIQUEFIED AIR
GASES...........................................................................................................................185
Hnzdil T., Suma J., Kouba M., Chrz V. ..................................................................185
CR08-52..............................................................................................................................193
40 FOOT CRYOGENIC INTERMODAL ISO CONTAINERS...................................193
Mtl P., Lnsk M., Chrz V.......................................................................................193
CR08-53..............................................................................................................................201
CRYOGENIC LIQUID TRANSFER POSSIBILITIES FOCUS ON STATIC
VACUUM INSULATED PIPES...................................................................................201
Takcs D., Chrz V.....................................................................................................201

USE
OF LOW TEMPERATURES
IN INDUSTRY..........................................................................................211
CR08-57..............................................................................................................................213
THERMODYNAMIC STUDY OF THE SIMULTANEOUS PRODUCTION OF
ELECTRICAL AND COOLING POWER FROM LNG .............................................213
Parise J.A.R.1 , Esteves A.D.S.1...............................................................................213
CR08-51..............................................................................................................................223
GAS IMPURITIES FREEZING OUT TECHNOLOGIES...........................................223
Klepal J., Stoek P.....................................................................................................223
CR08-49..............................................................................................................................227
MULTISTAGE CRYOGENIC TREATMENT OF MATERIALS: PROCESS
FUNDAMENTALS AND EXAMPLES OF APPLICATION......................................227
Alava L.A..................................................................................................................227

CRYOSTORAGE
OF CELLS AND TISSUES......................................................................235
CR08-06..............................................................................................................................237
FROM THE TISSUE BANK TO THE TISSUE ESTABLISMENT............................237
Mika P., Strakov H., Horynov A.......................................................................237
CR08-32..............................................................................................................................245
VENTILATION OF CRYOSTORAGE FACILITIES OF TISSUE
ESTABLISHMENTS.....................................................................................................245
Lain M. 1, Mika P. 2, Dvok J. 2........................................................................245
CR08-25..............................................................................................................................249
SPECIAL EQUIPMENT FOR CRYOPRESERVATION OF TISSUE IN A
STANDARD FREEZING UNIT...................................................................................249
Sprl G.1, Klingner E.2, Quinger J.3........................................................................249

USE
OF LOW TEMPERATURES
IN CRYOTHERAPY................................................................................255
CR08-55..............................................................................................................................257
THE LIQUID AIR CRYOCHAMBERS FOR WHOLE-BODY CRYOTHERAPY....257
Strnad P.1 , Fortkov L.2, Brojek W.3,...................................................................257

POSTERS:
HELIUM
TEMPERATURE
TECHNIQUES..........................................................................................263
CR08-18..............................................................................................................................265
EXERGETIC ANALYSES USABLE FOR CONTROL THE OPERATION
PARAMETERS OF THE HELIUM PROCESSING PLANT FOR DIFFERENT
WORKING CONDITIONS...........................................................................................265
Gherghinescu S..........................................................................................................265
CR08-09..............................................................................................................................271
SPLIT PULSE TUBE CRYOCOOLER WITH INNOVATIVE DOUBLE-PISTON
LINEAR COMPRESSOR..............................................................................................271
Kaiser G., Albert S., Schmidt J., Heidrich R., Binneberg A., Klier J........................271
CR08-01..............................................................................................................................277
VERY-LOW TEMPERATURE THERMAL CONDUCTIVITY OF STRUCTURAL
MATERIALS FOR LARGE CRYOGENIC EXPERIMENTS.....................................277
Ventura G., Barucci M., Martelli V., Risegari L.......................................................277

POSTERS:
HIGH TEMPERATURE
SUPERCONDUCTIVITY.......................................................................283
CR08-14..............................................................................................................................285
DESIGN, FABRICATION AND TEST RESULTS
ON A CONDUCTION COOLED HTS MAGNET.......................................................285
Joonhan B., Seokho K., Kideok S., Myunghwan. S..................................................285
CR08-23..............................................................................................................................293
ANALYSIS ON THE QUENCH AT THE CONDUCTION-COOLED JOINT
BETWEEN HTS WIRE AND NORMAL CONDUCTOR...........................................293
Bae D.K.1, Bae J.H.2, Lee D.-Y.3, Lee S.-J.3, Park J.-S.3,......................................293
CR08-37..............................................................................................................................301
ANALYSIS OF THE MAGNETIC PROPERTIES OF HTC SUPERCONDUCTORS
AND APPLICATION THEM AS PERMANENT MAGNETS....................................301
Sosnowski J...............................................................................................................301

POSTERS:
NITROGEN TEMPERATURE TECHNOLOGY................................309
CR08-02..............................................................................................................................311
GAS FLOW THROUGH NARROW GAPS AT LOW PRESSURE IN SUPERINSULATION PACKAGES..........................................................................................311
Stipsitz J.1, Dobrozemsky R. 2, Hirschl C. 1, Laa C. 1.............................................311
CR08-46..............................................................................................................................317
NEW DEVELOPMENTS OF NON-METALLIC CRYOSTATS FOR HIGH
SENSITIVE ELECTRONIC DEVICES AND OTHER APPLICATIONS..................317
Klier J., Sprl G., Schumann B., Binneberg A., Herzog R.......................................317
CR08-03..............................................................................................................................323
CRYOGENIC DISTILLATION COLUMN BEHAVIOR
AT THE VARIATION OF AN EXTERNAL FACTOR...............................................323
Pearsica C.,Stefan L.,Preda A.,Vasut F.....................................................................323
CR08-44..............................................................................................................................329
ANALYSIS OF PERIODIC ADSORPTION PROCESSES,
USED IN NEON AND HELIUM PRODUCTION.......................................................329
Bondarenko V. L.1, Simonenko Yu. M.2..................................................................329
CR08-45..............................................................................................................................335
NEON LIQUEFIERS AND THEIR USAGE IN THE INSTALLATIONS FOR RARE
GASES EXTRACRION................................................................................................335
Bondarenko V.L.1, Diachenko .V.2, Diachenko O.V.2..........................................335
CR08-35..............................................................................................................................341
GAS-CHROMATOGRAPHIC ANALYSIS OF MIXTURES OF HYDROGEN
ISOTOPES USING DIFFERENT PARAMETERS......................................................341
Preda A., Bornea A., Pearsica C., Vasut F................................................................341
9

CR08-19..............................................................................................................................349
MATHEMATICAL MODELS CONCERNING THE DESIGN OF COLUMN FOR
ISOTOPIC EXCHANGE PROCESS IN THE PILOT PLANT FOR TRITIUM AND
DEUTERIUM SEPARATION......................................................................................349
Gherghinescu S.1, Popescu G.1.................................................................................349
CR08-15..............................................................................................................................355
THE CREATION OF VEHICLES FOR MULTIMODAL TRANSPORTATION OF
LIQUEFIED GASES.....................................................................................................355
Zashlyapin R.A., Cheremnych O.Ya.........................................................................355
CR08-28..............................................................................................................................363
THE INCREASE OF EFFICIENCY AND SAFETY OF LIQUID HYDROGEN
TRANSPORTATION....................................................................................................363
Cheremnych O.Ya.....................................................................................................363
CR08-29..............................................................................................................................371
THE CREATION OF VAPOR COOLING DEVICES FOR LIQUID OXYGEN IN
STATIONARY RESERVOIRS USING LIQUID NITROGEN AS A COOLING
REAGENT.....................................................................................................................371
Cheremnych O.Ya., Korneva I.I................................................................................371
AUTHOR INDEX.............................................................................................................377

10

11

13

15

17

19

COMMITTEES
The International Conference Committee
Rodney Allam, UK
Alexey M. Arkharov, Russia
Stan Augustynowicz, USA
John G. Baust, Pres. Comm. C1, IIR, USA
John Campbell, USA
Walter F. Castle, UK
Didier Coulomb, Dir. IIR, France
Vaclav Chrz, Pres. Com. A2, IIR, Czech Republic
Ralf Herzog, Pres. Comm. A1, IIR, Germany
Boris A. Ivanov, Russia
Milos Kadlec, Czech Republic
Zdenek Kaiser, Czech Republic
Mohammad Kalbassi, UK
Philippe Lebrun, Head Section A, IIR, France
Omar Maiwand, UK
Locksley Mc Gann, Canada
Hans Quack, Germany
Stanislav afrata, Czech Republic
Ralph Scurlock (chairman), UK
Gabriele Spoerl, Germany
The Local Organizing Committee
Vaclav Chrz, (chairman)
Zdenek Kaiser
Radoslav Kundera
Martin Lnsk (Org. Com. secretary)
Zdenek Machala
Pavel Mika
Vera Musilov
Pavel Urban
Josef Ota
Jiri Pazek
Stanislav afrata (vice-chairman)
Pavel Schustr (vice-chairman)
Martin Vin

21

The Editorial Committee

Vaclav Chrz
Petr Duda
Romana Koov
Milan Kouba
Tomas Krlk
Pavel Mika
Vra Musilov
Pavel Urban
Josef Ota
Stanislav afrata
Pavel Schustr
Logistics
ICARIS Ltd., Conference Management
Ladislav ervinka & Dalibor ervinka, Directors
Romana Koov, Project manager

22

LIST OF EXHIBITORS
Registered till April 5, 2008
ACD CRYO AG
Gutenbergstrasse 1, CH-4142 Muenchenstein
Switzerland
Phone: +41 61 413 0230, Fax: +41 61 413 0233
E-mail: patrick.ravinel@acdcryo.com

www.cryoind.com
AIR LIQUIDE S.A.
Division Matriel Cryognique
8 avenue Gutenberg, Parc Gustave-Eiffel,
F-77607 Bussy Saint-Georges, Marne la Valle
France
Phone: +33 1 6476 1537, Fax: +33 1 6476 1699
E-mail: bjoern.sindermann@airliquide.com
www.dmc.airliquide.com
AUSTRIAN AEROSPACE GmbH.
8 Stachengasse 16, A-1120 Vienna,
Austria
Phone: +43 1 80199 3070, Fax: +43 1 80199 3060
E-mail: johannes.stipsitz@space.at
www.space.at
CHART FEROX, a.s.
steck 30, 405 30 Dn 5
Czech Republic
Phone: +420 412 507 343, Fax: +420 412 510 200
E-mail: sales@chart-ind.com
www.chart-ferox.com
CRYOSTAR SAS
2 rue de lIndustrie ZI BP 48, 68220 Hsingue
France
Phone: +33 389 70 27 27, Fax: +33 389 70 27 77
E-mail: info@cryostar.com
www.cryostar.com

23

HEROSE GmbH., Armaturen und Metalle

Elly-Heuss-Knapp-Str. 12, D-23843 Bad Oldesloe
Germany
Phone: +49 4531 5090, Fax: +49 4531 509 120
E-mail: info@herose.de
www.herose.de
LINDE AG, Linde Enginnering Division, Schalchen Plant
Carl-von-Linde-Strasse 15, 83342 Tacherting
Germany
Phone: +49 89 7445 6291 Fax: +49 89 7445 6291
E-mail: Gerald.Hecht@Linde-LE.com
www.linde-engineering.com
NEXANS Deutschland Industries GmbH. and Co.KG
Kabelkamp 20, 30179 Hannover
Germany
Phone: +49 (0) 511 676 3250, Fax: +49 (0) 511 676 2134
E-mail: Klaus.Schippl@nexans.com
www.nexans.de
VRV Group
Via Burago 24, 20060 Ornago (MI)
Italy
Phone: +39 039 6025 1 Fax: +39 039 6025 499
E-mail: cryo@vrv.it
www.vrv-group.com

24

OPENNING
LECTURE

25

CR08-63
THE IIR, YESTERDAY, TODAY AND TOMORROW
Coulomb D.
International Institute of Refrigeration (IIR),
177 boulevard Malesherbes 75017 Paris, France
ABSTRACT
The second half of 19th century showed both a sharp increase in the demand for cold
storage, refrigerated transport, needs of various factories and in the development of
successful refrigerating machines. 1908 was the year of the creation of the International
Institute of Refrigeration. The IIR had then to adapt to new challenges such as the
protection of the environment, to new uses of refrigeration and scientific progress.
INTRODUCTION
The IIR is celebrating its hundredth anniversary. The first part of the text describes the
creation the IIR, with the history of artificial cold, the 1908 event and the first years of this
new international body. The second part explains what the IIR is today, the changes that
have appeared regarding the challenges, the organization and the actions.
I THE CREATION OF THE IIR
a) Introduction
At 3 p.m. on October 5, 1908, 3000 specialists in the field of artificial cold gathered in the
Grand Amphitheatre of the Sorbonne in Paris. They had come to attend the formal opening
of the First International Congress of Refrigeration. The Congress, which lasted a week and
attracted a total of 5000 participants from 40 countries, was a resounding success.
Representatives from the worlds of science, commerce, industry and government
exchanged views on low temperatures, refrigeration technology, food, applications of
artificial cold in trade and industry, and legislative issues. The Congress culminated in the
founding of the International Association of Refrigeration in January 1909. It became an
international organization, the International Institute of Refrigeration, in 1920.
While it was electrical engineering that had taken the world by storm in the last quarter of
the 19th century, the baton passed to the cold industry in the early years of the new century.
Buyers included breweries and ice-cream factories, cold storage and refrigerated transport
companies, hospitals (for the conservation of dead bodies), dairy, chocolate, rubber and
perfume factories, dyeworks and factories producing liquid carbonic acid, ammonia or air.
Those involved in building mineshafts and subway tunnels soon saw the potential of
artificial cold in their line of business: refrigerant pipes could be used to create a wall of
frozen ground, after which it became far easier to dig out the space inside. In short,
artificial cold was a growing market.
b) Artificial cold
The mid-19th century witnessed a sharp increase in the demand for natural ice during the
summer months in Europe, North America (where ice was soon being used by the middle
classes) and the colonies. The demand was perhaps greatest among breweries producing
lager, which ferments at 5-8C, unlike the 20-30C of many English beers. The advent of
railways and steamships boosted trade in natural ice from Scandinavia and Canada, but
suppliers could not keep pace with the growing demand. Furthermore, rising concern about
27

the sawing of blocks of ice from polluted rivers and lakes gave extra impetus to the
development of machines that could manufacture clean artificial ice. The producers of
natural ice lowered their prices in a fruitless effort to reverse the tide.
The first machine to produce a continuous output of ice was invented by the French
businessman Ferdinand Carr. His idea was to release ammonia from a water solution by
heating it, to condense the vapour under pressure until it was liquefied, and then to allow
this liquid to evaporate and expand in a sealed space. This would extract heat from an
adjoining space with water, which would immediately freeze. The vapour would be
absorbed by the aqua ammonia, after which the cycle would be repeated. A prototype
was placed in a brewery in Marseille in 1859. Carrs ice machine became rather famous
when it was displayed at the Paris World Exposition of 1867. He was already doing a brisk
trade before then: the Confederates had bought several machines from him during the
American Civil War (1861-1865). After some adjustments made by Mignon and Rouart in
Paris, the vapour absorption device was one of the best-selling refrigerators in the years
1870-1885, especially in France. After that it was superseded by the vapour-compression
refrigerator, which is based on a far simpler construction.
This system, which is still applied in household refrigerators, artificial ice rinks and
industrial plants today, was invented by the French engineer Charles Tellier, earning him
the title le pre du froid. It uses a closed cycle. A compressor is used to compress methyl
ether (which was later replaced by methyl chloride, sulphur dioxide, carbonic acid gas, and
above all ammonia); a water-cooled condenser turns this into liquid, which evaporates in
the space to be refrigerated (in a system of pipes the main difference with regard to
Carrs system) and thus extracts heat from it. Tellier built his first refrigerator in Paris in
1863. Four years later he installed an improved version, using methyl chloride as the
coolant, in an ice factory in Marseille, France.
Commercially speaking, the most successful machines were compression refrigerators
using ammonia, launched in 1875 after theoretical studies carried out by the scientifically
trained Carl von Linde. The Gesellschaft fr Lindes Eismachinen A.G., in Wiesbaden,
supplied its first machine to a brewery in Munich and was soon the market leader. By 1890
the German company had sold about a thousand machines, and around the turn of the
century the Wiesbaden factory was sending off one or two of its refrigerators every day[1].
A major innovation made possible by the new refrigerators was the export of frozen meat
from Australia, New Zealand and South America to Europe. Cooling the meat with ice
proved not to be an option; steamships were still slow in the 1870s, and clippers also took
over 100 days to cross the ocean. The problem had to be solved with machines. In 1876,
Tellier built a compression refrigerator on board the French ship Le Frigorifique. This
steam-powered three-master sailed from Marseille to Buenos Aires with a cargo of frozen
meat, to return to Le Havre a year later. Though not a commercial success, the voyage had
demonstrated that shipping frozen meat across the oceans was technically feasible.
Bulk transportation imposed more stringent demands, and the problem with Telliers
machine was that if built on a larger scale, it sometimes broke down. Besides this, the
toxicity of the coolants and the risk of explosion deterred ship owners from taking the
plunge. It was another type of refrigerator that made them change their minds: the air
expansion machine patented by the Scottish butchers Bell and Coleman in 1877. This
cooled the produce by the rapid expansion of compressed air, and in spite of poor efficiency
large steam engines were needed to compress the necessary quantities of air and
problems with frozen water vapour, the sailing vessel Strathleven transported 34 tons of
28

frozen meat from Australia to England safely in 1879 using one of these machines. Things
moved very fast after this. In 1907, Argentina exported 425 000 tons of frozen meat to
England alone.
Low-temperature science, too, progressed in leaps and bounds. The last quarter of the 19th
century witnessed the liquefaction of each of the permanent gases in turn. In 1877, the
Frenchman Louis Cailletet and his Swiss colleague Raoul Pictet liquefied air. In 1883, the
polish team Zymunt von Wroblewski and Karol Olszewski went a step further, by inducing
the blue liquid of oxygen to boil gently. James Dewar, working in the Royal Institution,
London, became the first to produce liquid hydrogen, in 1898, after which Heike
Kamerlingh Onnes won the race for liquid helium in Leiden on 10 July, 1908[2].
c) The Paris congress
In this atmosphere of up-and-coming artificial cold, of new, hitherto unsuspected
applications, of changing economies in countries such as Argentina, of a proliferation of
technical problems crying out for solutions, of a scientific quest for absolute zero, the idea
of holding a major international Congress of refrigeration, in Paris, emerged. A Congress of
this kind had been held in Vienna in 1873, to coincide with World Exposition there, but it
had focused on brewers and their need for natural ice. At the beginning of the 20th century,
artificial cold produced by refrigerators revolutionized agriculture all over the world and
offered an immense economic potential concerning national and international food trade by
using large-scale refrigerators.
The engineer J. de Loverdo was the prime mover of the Paris congress, and in May 1907 a
circular was distributed calling for participants. The initiative soon attracted a wide-ranging
and distinguished band of supporters including lInstitut de France, the French Parliament,
the Collge de France, lAcadmie de Mdecine, major transport companies. No-one
interested in cold could afford to miss the Premier Congrs International des Industries
Frigorifiques, which was finally held at the Sorbonne University on October 5-10, 1908
under the more appealing and inclusive name of Premier Congrs International du Froid, or
First International Congress of Refrigeration[3]. The broad aim was to exchange ideas and
discoveries in the field of cold technology.
To keep the Congress manageable, it was divided into six sections: low temperatures,
refrigeration installations, applications of cold to foodstuffs, applications in other
industries, applications in trade and transport, and a final section that would examine the
relevant legislation. The name of the Congress made it clear that it was not to be a one-off
initiative. Ideas for an international institute for cold and science, or for training courses in
refrigeration technology, to be founded in Paris, soon proved overambitious. Instead, the
preparatory committee offered to set up an International Association of Refrigeration. Its
remit would be to perform research on scientific, technological and industrial applications,
to set up a library covering all aspects of the field, to publish articles and inform its
members, to provide courses, set up excursions, and organize a biennial Congress on the
subject of cold, to be held in a different country each time. National committees were
formed to ensure that all went smoothly. The latter coordinated the submission of reports
for Paris, and once the Association of Refrigeration actually got off the ground, they were
to have seats on its Executive Committee.
During the opening session on Monday October 5, 1908, the French minister of
Agriculture, Joseph Ruau, emphasized that agriculture, being the dominant factor in the
economic growth in the second half of the 19th century, profited a lot from the science of
cold and its technical applications. After Ruaus speech, the national committee chairmen
29

were all invited to say a few words. Kamerlingh Onnes, who represented the Dutch
government, took the opportunity to define the mission of the International Association of
Refrigeration: to bring together all knowledge bearing on low temperature[4]. He also
emphasized that research on artificial cold and its applications was of importance to all
countries and all social classes. The congress on refrigeration, said Kamerlingh Onnes,
could help to expand international solidarity that precious treasure of humanity. In
conclusion, he emphasized the importance of studying the physical properties of matter at
extremely low temperatures. This would further clarify the relationship between matter and
electricity, thus preserving the dream of energy reservoirs of a size that passes
imagination. The French physicist Jacques-Arsne dArsonval, who spoke on behalf of the
scientific community during the opening ceremony, also emphasized the importance of
pure research. All your machines, he said, addressing the technicians around the hall,
rely on thermodynamic principles. The scientific community in turn derived great benefit
from experience gained in industry: a science-and-technology spiral avant la lettre.
During the closing session, he placed Kamerlingh Onnes in the limelight: his liquid helium
made him the star of the Congress.
In the avalanche of recommendations that the Congress adopted on its final day, applied
cold technology predominated, but there were also follow-up proposals to the goal that
Kamerlingh Onnes had formulated at the opening session. The most striking was: Given
the crucial interest attached to pursuing and coordinating scientific and practical work in
the field of low temperatures, the Congress emits the wish of the foundation of an
International Association for the promotion of scientific and other studies, with its head
office in Paris, which would pursue its study of the whole field of refrigeration and at the
same time continue to strengthen the already specialized work centres.
d) The International Association of Refrigeration
The International Association of Refrigeration duly materialized. It was founded on January
25, 1909 in the presence of delegates from 35 countries. Lebon was appointed President of
the new association, and De Loverdo became its director. The Paris Congress led to the
establishment of six international committees. Vice-president Kamerlingh Onnes was
chosen to chair the first committee, which was to focus on scientific matters, and which
also included Louis Cailleted, Charles-Edouard Guillaume (of the Bureau des Poids et
Mesures) and James Dewar.
While the Association started life with a few dozen members, by the time of the 2nd
Congress of Refrigeration, held on October 6-11 in Vienna, Austria, it had 1700. Argentina
contributed most, with 1000 members all because of its frozen meat and the United
States at 370, also had a strong contingent. For all Americas numerical preponderance,
however, 92% of the financial contributions came from Europe. The second congress
attracted over 3000 participants.
A proposal was adopted to set up a grants system enabling young physicists to perform
research relevant to cold technology in Leidens cryogenic laboratory. The 3rd
International Congress of Refrigeration was held in September 1913 in the dual venues of
Washington and Chicago[5].
e) Restructuring Association
After the Great War, the Association was restructured into the International Institute of
Refrigeration. This was triggered by the resignation of the president, Andr Lebon, on
December 12, 1918. Following this, the director of the Association convened a meeting of
the Executive Committee on February 6, 1919. The meeting in the Crdit Foncier
30

[mortage bank] dAlgrie et de Tunisie in Paris was attended by only six of the 28
members of the executive committee.
Discussions on restructuring designed to place the Association on a solid financial basis
were postponed until the end of the peace talks in Versailles. On June 21, 1920, the
Association was replaced by the International Institute of Refrigeration. This had a far more
tightly-knit organizational structure, based on that of the International Institute of
Agriculture in Rome: instead of individual members it had participating countries in six
categories, paying fixed contributions. And these rules are still valid in 2008.
The formal International Congress of Refrigeration, held in the premises of the Ministry of
Trade in Paris, was preceded by a meeting of the provisional Executive Committee, to
which Kamerlingh Onnes belonged. President Lebon, now back at his post, reviewed the
organisational structure: the Bulletin, the grants, and the international committees,
including the physics, chemistry and thermometry committee chaired by Kamerlingh
Onnes. Lebon opened the meeting by congratulating the Dutchman on his recently acquired
status of membre correspondant of the Acadmie.
In the Great Hall at Rue de Varenne, it fell to Kamerlingh Onnes, un grand savant
lavant-garde de la science, to respond on behalf of the 42 countries attending to the
welcome speech given by Ricard, the French Minister of Agriculture. Ricard emphasized
that the war had truly brought home the benefits of refrigerating food. The next step was to
make a concerted effort to improve the accessibility of the cold industry. A more farreaching and useful goal than improving the refrigeration could scarcely be imagined, the
minister concluded. In his response, Kamerlingh Onnes remarked, not without self-interest,
that as long as the Institute followed in the Associations footsteps, success was assured.
The science of refrigeration had a golden future and developments had time and again
exceeded their wildest expectations. What had begun with a little cloud of liquid air in a
Cailletet test tube had grown into a cryogenic industry producing billions of tonnes of
oxygen, nitrogen and argon annually both in liquid and gaseous state. This is a real
fulfilling of the prophetic words of Jacques-Arsne dArsonval, delivered at the occasion
of first liquefaction of air in 1877: Industrial liquefaction of air is not only a scientific
upheaval; it is also an economical and social upheaval. Preparation of oxygen and nitrogen
by liquefaction of air brings forth an upheaval in illumination, metallurgy, chemical
industry, health care and agriculture. There is not much to be added after more than 100
years.
II THE IIR TODAY
a) Fundamentals
The IIR is 100 years old. It has changed, but some principles are still alive:
The basic reasons for the creation of the IIR are still important: the role of refrigeration in
agriculture and food, the importance of science and technology in refrigeration and
especially cryogenics, and the need for scientists to share their research.
The IIR is still an intergovernmental organization with six member-country categories.
However, it also has, as in the beginning, private and corporate members who receive its
services.
The structure of the IIR, with congresses and with commissions or committees, is partially
the same. The Bulletin, which was the first IIR publication, has been maintained.

31

The roots are still presents, but new branches have appeared and the world has changed.
The IIR had to adapt to its environment and continues to adapt to new challenges.
b) The challenges
The most important challenges for humanity in the 21st century are health and the
environment.
Diseases and mortality are still too widespread in developing countries. The aim to live as
old as possible in good health is the goal of most people in developing countries.
Refrigeration is one of the answers: it is necessary to guarantee sufficient quantities of
food, available to everybody, and to preserve its quality, particularly in order to avoid
contamination.
Refrigeration is also necessary to enable the storage and transport of health products
(vaccines, certain drugs, diagnostic products), and to preserve organ and tissue in
hospitals, for cryobiology, surgery and medicine.
Refrigeration is at the core of two major threats to the environment: ozone depletion and
climate change, because of the use of certain refrigerants and because of the energy it
needs. We thus have to implement new refrigerants, to reduce electrical consumption and to
develop new environmentally friendly technologies. Refrigeration is also an answer to
global warming: air conditioning will be increasingly necessary in many cases and
refrigeration is needed in several leading-edge energy sources: liquefied natural gas,
liquefied hydrogen, thermonuclear fusion. Moreover, refrigeration will be needed for the
capture of CO2 in energy plants, the steel industry
The IIR is not only still necessary: it is more and more necessary.
c) Members
Our members have changed but not totally. The First and Second World Wars, and
decolonization had an important impact. However, the main countries present when the IIR
was set up are still there.
May I mention that Czechoslovakia, was one of the founding member countries (it joined in
1921) and both the successor countries, Czech Republic and Slovakia are still active in the
IIR.
Originally, there were about 40 member countries; they are now 61. We should have more
than 150 member countries according to our mission and global challenges.
The number of private and corporate members is also too low: there were about 500 such
members 50 years ago and there are currently almost 600. We certainly need to attract more
private and corporate members, and welcome suggestions.
The challenges we have are challenges for governments, but also for all public and private
sectors.
d) The committees, commissions and working parties
The IIR is more sophisticated than at the beginning, perhaps too sophisticated. However, it
reflects the various fields of actions and the broad mission we have. We have a General
Conference, an Executive Committee, a Management Committee, a Science and
Technology Council. The latter comprises ten commissions: each one has about 50
members from all paying IIR member countries. Three of them are involved in this
32

conference: Commissions A1 (Cryophysics, Cryoengineering), A2 (Liquefaction and

separation of gases) and C1 (Cryobiology, Cryomedicine).
We still have commissions dedicated to the cold chain, but also to issues that did not exist
at the outset, such as Air Conditioning. We also regularly create and sometimes disband
working parties, which do not have the same official status. However, they are very
necessary to actively work on precise subjects. There are several projects in the field of the
commissions involved in this conference. I hope we will be able to find here persons to
handle them. Your field is important for the future and we need, as for existing working
parties, conferences, workshops, publications and statements in your fields.
e) Publications
The first IIR publication was the Bulletin. It was first published in 1910; it still exists, but it
has of course changed: it is now an electronic Bulletin (e-Bulletin). We have an electronic
database, Fridoc, which now comprises more than 81 000 entries. It is the most important
database in refrigeration technologies. The Bulletin now essentially comprises the new
entries of abstracts of articles and documents published all over the world and is now
merging with Fridoc.
We launched two other publications:
The International Journal of Refrigeration was created 30 years ago. We needed a scientific
journal, with the same kind of selection and strict peer-review approach as the best ones.
We have succeeded and its impact factor is the best in the refrigeration sector: it is the 26th
out of 106 journals in mechanical engineering and 14th out of 42 journals in
thermodynamics.
The Newsletter was created 8 years ago. It comprises selected news from all over the world
in addition to IIR news and it is sent to our entire network (more than 3000 people).
The IIR progressively began to publish books and guides, thanks to its network of experts,
which is our main wealth. We publish technical books, brochures, diagrams, training
courses and recommendations. We publish reference documents that are used and
recognized all over the world. For example, we publish the International Dictionary of
Refrigeration in 11 languages.
Thanks to our intergovernmental status, we are invited to international events and are able
to deliver statements, to participate in meetings and side events, to prepare international
standards and recommendations for governments, the United Nations and its various
bodies, for decision-makers. For instance, we regularly deliver statements during the United
Nations Conferences on the ozone layer and on climate change.
f) Conferences and congresses
One of our main activities is now the holding of conferences. We organize about 3-4 IIR
conferences per year and we co-sponsor 8 conferences per year on the average. Most of
them are series of conferences, like this one, and it is important to have regular events on
key subjects for scientists and engineers. I am sure that this one will be successful as usual.
Abstracts of the papers presented will be inserted in our Fridoc database and some of them
can be presented to the International Journal of Refrigeration . The proceedings will be
inserted in our Catalogue of Publications and sold. We will do our best to raise the visibility
of your work.

33

We also organize The International Congress of Refrigeration every 4 years, which covers
all fields of refrigeration technologies, as was the case 100 years ago. It is a great pleasure
for me that the next one will take place in Prague in 2011. I hope you will all attend it.
CONCLUSION
In conclusion, the IIR is 100 years old. However, it is still young: refrigeration preserves
health and quality of life! We still have a lot of work to do throughout this century,
together, for a better health, for a better environment, in a way of sustainable development.
I hope the coming generation will continue to celebrate the IIR in 2108, with many more
people and many more countries and companies in healthier and safer world for everyone.
NOTA
The first part of this paper partially comes from a text of Dirk van Delft, which will be the
introduction of a brochure published for the IIR Centenary in 2008.
REFERENCES
1 Dienel, H. L., Linde: history of a technology corporation 1879-2004 (2004).
2 Van Delft, D., Freezing Physics. Heike Kamerlingh Onnes and the Quest for Cold
((2007).
3 Museum Boerhaave, archives of Heike Kamerlingh Onnes, inv. No. 190.
4 Bulletin Officiel du Premier Congrs International du Froid, Nos. 1 & 2 (1909)
5 Bulletin Mensuel de lAssociation Internationale du Froid (1910-1913)

34

HELIUM
LIQUEFACTION

35

CR08-62
THE CENTENARY OF THE FIRST LIQUEFACTION OF HELIUM
Scurlock R.G.
Kryos Technology, 22 Brookvale Road, Southampton, SO17 1QP, UK.
ABSTRACT
This paper marks 2008 as the Centenary Year since the first liquefaction of helium. On 10 th
July 1908, Professor Heike Kamerlingh Onnes and his team first liquefied helium at his
Cryogenic Laboratory, University of Leiden, The Netherlands.
Before describing this liquefaction event, the paper discusses the considerable difficulties
Kamerlingh Onnes had to overcome in achieving his success. In contrast to his rivals,
Dewar and Olszewski, he adopted the first ever big science approach to build up a large
laboratory at Leiden, with the extensive infra-structure and expertise needed for his attempt.
He also had a strong working relationship, through his experimental measurements on the
low temperature properties of gases, with theoretical physicist van der Waals at the
University of Amsterdam.
A brief chronology outlines how Kamerlingh Onness success in liquefying helium helped
to open the door from the classical physics of the 19 th century into the new scientific
world of macroscopic quantum physics of the 20th century.
1. INTRODUCTION
th

On 10 July 1908, Professor Heike Kamerlingh Onnes and his team first liquefied helium
in the Cryogenic Laboratory, University of Leiden, The Netherlands.

Figure 1. Building of the Cryogenic Laboratory, University of Leiden

Today, at Cryogenics 2008, we celebrate the centenary of this landmark achievement,

which opened the door from the 19 th century classical world of the physical sciences, into
the strange, new, quantum mechanical world of the 20 th century, with its many macroscopic
manifestations at low temperatures.

37

However, before we look at the consequences of his achievement, let us examine how
Onnes overcame many problems to achieve his success, when his main competitors, Sir
James Dewar, at The Royal Institution, London, and Dr Karol Olszewski, at The
Jagiellonian Unversity, Krakow, failed.
2. ONNES AND HIS COMPETITORS, DEWAR AND OLSZEWSKI
Dewar, having been the first to liquefy hydrogen in 1898 while working with a small team
at the Royal Institution, set out to liquefy helium in the same way with the same 2 technical
assistants and the same low level of funding.
Olszewski, working again with a small team, set out to liquefy helium in the same manner
as his earlier successful first liquefactions, but again with a low level of funding.
Onnes realised shortly after his 1882 appointment as Professor of Experimental Physics at
Leiden, at the age of 29 years, that the success of his new Cryogenic Laboratory required a
new approach ( probably the first big science approach ).
Over a continuous period of 26 years, he set about building
up the extensive infra-structure to support a large
laboratory and the cryogenic experience of his staff
through many research projects on the properties of matter
at low temperatures, before turning to
the liquefaction of helium [1]. He instituted a training
school for instrument makers and glass blowers, an open
house for visiting scientists, a new journal for the
publication of all research results at Leiden, and obtained
adequate government funding plus royal patronage. In
addition, he built up a strong collaborative relationship
with van der Waals, who was 16 years older than Onnes
and had been appointed in 1877 as Professor of Physics at
Amsterdam University.
Figure 2. Professor Heike Kamerlingh Onnes
3. THE RELATIONSHIP BETWEEN ONNES AND VAN DER WAALS
The influence of the friendship between Onnes and van der Waals cannot be
underestimated in the successful contributions they both made to physics. Van der Waals
was a powerful theoretician, making 2 major contributions: (i) pioneering the concept of a
general equation of state to relate the PVT behaviour of a real gas from high temperature
down to and below its critical temperature, by considering gas molecules to have (a) a
finite size, and (b) a finite sphere of influence or interaction between each molecule and its
near neighbours in the fluid state, (ii) hypothesising and testing a Law of Corresponding
States linking the PVT behaviours of different gases in terms of their critical P, V and T
parameters.

38

Over a period of years, Onnes carried out

measurements to test and confirm van der Waalss
theories, including the helium gas isotherms, so as
to predict the critical PVT parameters needed to
attain liquefaction with his available liquefier
equipment.
In fact, van der Waalss contributions were
recognised world-wide by his being awarded the
1910 Nobel Prize in Physics--- 3 years before
Kamerlingh Onness award.

Figure 3. Onnes with van der Waals at the helium

liquefaction stand

4. RESEARCH FUNDING AT THE END OF THE 19TH CENTURY

At the end of the 19 th century, in contrast with the Arts,
funding was sparse for supporting growth in the
Sciences, both as an academic activity and to meet the
demands of the new industries of the Industrial
Revolution. Only in Germany was the development
( with the royal sponsorship of the Kaiser ) of applied
sciences being supported by government funding,
through the introduction of the polytechnics. This policy
had led to the domination in Europe by the German
chemical industry, in the production of organic
compounds and aniline dyes, before the end of the
century.
Figure 4. Onnes with Flim and students

The Netherlands government adopted the same policy in order to compete (with the royal
sponsorship of Queen Wilhemina) and Onnes was one of the fortunate beneficiaries when
he was awarded continuous funding for his Cryogenic Laboratory from the beginning of his
appointment in 1882.
In comparison in the UK, Royal Commissions had been set up and, for example, had
advocated that education of science in universities should not be specialised and
recommended government funding of 4000 for the endowment of research throughout the
UK to be administered by the Royal Society. Dewar could expect little help from this
source, having made so many enemies.

39

In Krakow, there was little or no government

funding, and Olszewski and Wroblewski had to use
their small resources to carry out their research,
which had included the
first liquefaction of oxygen, carbon monoxide and
nitrogen. They also achieved the production of
liquid hydrogen in a transitory jet in 1884, some 14
years before Dewars achievement. However,
Wroblewski was seriously burned, and died soon
afterwards, from an experiment on the physical
properties of hydrogen in 1888.
Olszewski
continued on his own, to condense and solidify
argon. Onnes recognised him as the precursor of
cryogenics, but that with limited funding he could
not be a serious rival in the competition to liquefy
helium.
Figure 5. Stand of the first liquefaction of helium
5. GASES AND MATERIAL RESOURCES IN 1908
It is impossible today to imagine how difficult was every aspect of low temperature science
in 1908.
Gases
The gases, required in a pure state, for the study of their properties and for use as
refrigerants, needed to be produced and purified on a DIY basis as required in the
laboratory. They were not available commercially.
Hydrogen was made by the electrolysis of water, purified over heated catalysts and dried
before storing at low pressure in gas holders, or at higher pressures in cylinders. Helium
was a very rare gas in 1908, and the major recognised source was a rare mineral, monazite
sand, which Onnes obtained from North Carolina, USA. When heated, the sand could
produce about 1 litre of impure helium gas at 1 bar from each kg of sand. Elaborate
purification steps, taking many months, were then needed to remove every contaminant
including hydrogen, before the helium product was acceptable for a liquefaction attempt.
Onnes used about 400 litres of gas in continuous
circulation for his first successful liquefaction, producing a volume of liquid of a few tens
of millilitres.
The occurrence of helium in natural gases ( at 0.3 to 2.0 % ) was not publicised until 1907
by Cady and McFarland [2], while the first production did not start until late 1917 in
Hamilton, Ontario, Canada, and early 1918 in Fort Worth, Texas, USA.
Instrumentation
Instrumentation was crude, to say the least, in comparison with today..
A constant volume gas thermometer was the most reliable, employing a small copper bulb
filled with helium under pressure, with the absolute temperature varying linearly with
pressure down to 10K, about twice the critical temperature; but pressure readings at lower
temperatures were not reliable indicators, eg. for the NBP at 4.2K.
40

Platinum resistance thermometry was the alternative means of measuring temperature, but
again below 10K, was not reliable.
Constructional materials for cryogenic use
Metallic materials of construction in the form of thin-walled tubing and cylinders were
limited to low thermal conductivity nickel alloys like German silver, medium conductivity
copper alloys like brass, and high conductivity copper. No steels or aluminium alloys
existed. Joints were made with soft or hard solder--- oxy-acetylene welding had not yet
been invented.
The fall back material was glass tubing in the form of high borosilicate Pyrex glass, which
turned out to be permeable to helium at ambient temperature; it was later
replaced by low-borosilicate Monax glass, which is non-permeable. As a result, the major
parts of laboratory cryostats, dewars, liquefiers and associated pipework were made of
glass, with the aid of highly trained glass blowers.
This lack of materials and equipment resources for low temperature research did not change
until the mid 1950s. Until then, the idea of cryogenic applications using liquid helium
appeared to be impossible.
6. THE FIRST HELIUM LIQUEFACTION, ON 10TH JULY 1908
After 26 years of preparation, construction and trials, Kamerlingh Onnes and his team of
technicians, led by Gerrit Flim, were ready for the first attempt to liquefy helium.
On 9th July 1908, preparations began with the production of enough liquid air for the next
day, via the cascade liquefier of methyl chloride, ethylene and oxygen. Some of the liquid
air was then used as a precoolant for making sufficient liquid hydrogen for the liquefaction
attempt.
On 10th July 1908, work started at 0545, the first elaborate and tedious job being to remove
the last of the impurities from the helium gas down to the lowest possible level. The gas
was passed over copper oxide, after which oxygen and gases of similar volatility were
removed by freezing them out in liquid hydrogen. The helium was then compressed and
passed over charcoal, at successively liquid air and liquid hydrogen temperatures, for
several times until all impurities had ben removed as far as practicable. The purpose of
these purification stages was to prevent blockages caused by freezing out of residual
impurities in the fine passages of the high pressure pipework of the liquefier.
At 1620, the purified helium was admitted into the liquefier cycle, which involved
precooling with liquid air at 80K, liquid hydrogen at 20K, and pumped hydrogen at
14K, close to its triple point. At 14K, the compressed gas was well below its estimated JT
inversion temperature for helium, and was led into a Hampson recuperative cooling spiral
ending in a JT expansion valve.
By 1900, the gas thermometer had reached an apparent temperature of 5K and stopped
moving down, but no liquid could be seen. It was then realised that the thermometer was
behaving as though immersed in liquid.
Going beneath the cryostat, with an electric light, Onnes looked upwards into the glass
cryostat and clearly saw the liquid meniscus standing out sharply ( once seen ) like the
edge of a knife against the glass wall.
He had indeed liquefied helium [3].
41

His immediate regret was that his friend Professor van der Waals was in Amsterdam, and
would have to wait several days to see the liquid for himself at the second liquefaction run.
On the same day, with his first liquefaction run, Onnes was able to measure the normal
boiling point at 4.3K, and estimate the critical temperature at 5.0K . He also noted the
unexpectedly low density of the liquid at 125 g/litre (125kg/m3), and the extremely low
surface tension.
He also tried to reach the triple point, by reducing the vapour pressure. He reached 7mm Hg
(corresponding to 1.7K ) but the helium remained liquid.
For the next 13 years until 1921, Onnes would use larger and larger vapour pumps,
probably reaching a minimum temperature of 0.83K, but still the helium remained liquid. It
was beginning to appear that helium remained liquid to absolute zero.
The same day, Onnes sent a telegram to Dewar announcing his success. Dewars reply
shows his complicated feelings thus:
CONGRATULATIONS. GLAD MY ANTICIPATION OF THE POSSIBILITY OF THE
ACHIEVEMENT BY KNOWN METHODS CONFIRMED. MY HELIUM WORK
ARRESTED BY ILL HEALTH BUT HOPE TO CONTINUE LATER ON.
7. THROUGH THE DOOR TO MACROSCOPIC QUANTUM PHYSICS
Five years later in 1913, Kamerlingh Onnes was awarded the Nobel Prize in Physics for his
successful liquefaction of helium. His discovery of superconductivity in mercury in 1911
was not even part of the reasons for his Nobel award. However, he made certain in his
acceptance speech to mention superconductivity, expressing his wonder about the abrupt
loss of electrical resistance. In addition, he mentioned the extremely low density of liquid
helium, and suggested to the Nobel audience that explanations for these strange phenomena
could possibly be connected with the new quantum theory. How right he was!
8. SUBSEQUENT DEVELOPMENTS: A BRIEF CHRONOLOGY [4]
Following on the work of Onnes, in 1926 Keesom solidified helium at Leiden by the
application of pressures above 26 bar, and established the melting curve using an all-glass
cryostat.
In the 1930s, the phenomenon of superfluidity in liquid helium was identified and the
associated effects of -specific heat, fountain effect, wall-film flow etc. were measured and
formulated into various theories.
Before the mid 1950s, liquid helium temperature research involved cryostats which
included their own miniature helium liquefiers and glass dewars. Experimental work was
slow, time consuming and sometimes dangerous with high pressures inside glass systems.
Then the availability of 4-8 L/h Collins helium liquefiers, and the like, enabled the original
few, and many new, low temperature physics laboratories to have a centralised liquid
facility, or to have access to a commercial liquid helium facility and distribution system via
25, 50, 100 litre, or larger, portable metal dewars.
Superconductivity remained very much in the background (although it had been found in
many non-magnetic solids), until 1957 when the BCS theory of Type I superconductivity
via pairing of electron states was proposed and tested [5].

42

But it was not until 1961, when high current, high field, Type II superconductors were
discovered, that any application of superconductivity could begin to be considered seriously
[6].
A further 10 years of development led to the large scale manufacture of NbTi and Nb3Sn
conductors suitable for power engineering applications in motors, generators, cables, fault
current limiters and transformers. However, the power engineers have
hesitated since then, about their use of Type II superconductors, because of the large scale
liquid helium requirements.
The 1980s saw the new development of 5 T, liquid helium cooled, superconducting
magnets with a 1m diameter bore, leading directly to the invention of Magnetic
Resonance Imaging and its use as a medical diagnostic tool. Since then, most hospitals
around the world have invested in MRI, all using liquid helium, while higher fields, lower
stray fields, and improved computer techniques are leading to functional MRI (eg. for
studying brain functions ) when combined with liquid helium-cooled, Josephson-device,
encephalography.
A spin off during the 1990s has been the development of miniature closed cycle
cryocoolers to replace the use of liquid helium altogether. Thus, laboratory cryostats for all
kinds of application at liquid helium temperature can now operate in a liquid free state at
4K, just like a domestic refrigerator at 263K.
On the other hand, the large scale use of liquid helium has expanded rapidly since 1980 into
space science and particle physics, employing ever larger and ever more sophisticated
superconducting solenoid, dipole and quadrupole magnets.
The confidence gained during this period has led directly to the design, building and
commissioning in 2008 of the Large Hadron Collider LHC particle accelerator at CERN.
The LHC is in a tunnel 26 km in circumference, with liquid helium cooled 3 Tesla
superconducting magnets at 1.8K, around the whole length of the tunnel. The weight of the
superconducting magnets is a total of 36,000 tonnes. 36 MW of power is needed for the
refrigeration plant, in which the total helium inventory is 138 tonnes of liquid helium
(equivalent to 1.1 million litres of liquid; more than the total inventory for the rest of
Europe ).
The future of liquid helium is bright as more and more natural gas streams are being
stripped of their helium, before being exported as fuel gas or chemical feedstock. Even
larger applications than the LHC can be envisaged, particularly in the development of
fusion reactors with their requirement for very large volumes of high magnet fields, which
can only be created by liquid helium cooled superconductors at present.
Following their discovery in 1986 [7], the new era of ceramic superconductors, operated at
liquid nitrogen temperatures rather than liquid helium temperatures, now appear to be
particularly suitable for power engineering applications, with high current densities at
relatively low magnetic fields of 1 Tesla. Prototype cables, motors, generators, fault current
limiters and transformers are presently being built and tested.
However, where high magnetic fields of 5 10 Teslas, or higher, are required as in MRI,
liquid helium temperatures will continue to prevail for the present.

43

9. REFERENCES
1. Scurlock R.G, History and Origins of Cryogenics, (1992), Oxford University Press,
Oxford.
2. Cady H.P. and McFarland D.F., J.Am.Chem.Soc.,(1907) 29 1523.
3. Kamerlingh Onnes H., Commun.Phys.Lab., Leiden (1908) 108 ; Proc.R.Acad.,
Amsterdam (1908) 11, 168.
4. Timmerhaus K.D. and Reed R.P.(Editors), Cryogenic Engineering:Fifty Years of
Progress, (2007), Springer, New York.
5. Bardeen J., Cooper L.N. and Schrieffer J.R., Phys. Rev .,(1957) 108 1175.
6. Kunzler J.E., Buehler E., Hsu F.S.U. and Wernick J.H., Phys. Rev. Letters, (1961) 6 89.
7. Bednorz J.G. and Muller K.A., Z. Phys., (1986) B64, 189.

44

CR08-56
FIFTY YEARS FROM HELIUM LIQUEFACTION
IN CZECHOSLOVAKIA
AND A NEW TURBINE TECHNOLOGY
Kaiser Z.1, Kouba M.2, Kundera R.3, Pruk J.4, afrata S.4, Schustr P.5, Chrz V.2,
1

Ingersol Rand, Prague, Czech Republic (Ferox Dn formerly)

2
Chart Ferox, Dn, Czech Republic
3
PBS, Velk Bte, Czech Republic
4
Institute of Physics, Academy of Sciences of Czech Rep., Prague, Czech Rep.
5
ATEKO, Hradec Krlov, Czech Republic
ABSTRACT
When celebrating 100 years of helium liquefaction at the conference Cryogenics 2008 in
Prague, it is worth to look back onto the part of helium history in this part of world and in
the Czech Republic specifically. It is just fifty years from first steps for helium liquefaction
in the Institute of Nuclear Physics of the Czech Academy of Science. Technology of
Helium liquefiers has been developing in the Ferox company for thirty years, then. This
research required development of helium turbines, which was realized by ATEKO and PBS
companies. Successful development of highly efficient turbines with magnetic brakes using
dynamic gas bearings was a base of new production branch, which has been continuing up
today with series supplies to the most recognized European manufacturers of helium
liquefiers.
PIONEERING AFTER THE WORLD WAR II
Helium liquefaction was a long- time privilege of several
laboratories in the world. Still in the year 1946, there were
fifteen labs only in the world, who operated a helium liquefier
with a performance of several liters per hour. These were
results of several-year work of teams of physics and
technologists of particular
institutions.
Also
in
Czechoslovakia,
research in the branch of low
temperature nitrogen and
helium
physics
and
technology was developing
rapidly after the World War II.
After
founding
the
Department
of
Low
Figure 1. P.L. Kapitsa
Temperatures of the Institute
of Nuclear Physics of the
Czechoslovakian Academy of Sciences in Prague in the year
1956 the need of access to liquid helium became critical. It
was decided on building own Figure 2. First He liquefier from Dn, N2 and H2 cascade, in
liquefier using experience and the Institute of Nuclear Physics, e.
documentation of the Moscow

45

Institute of Problems of Physics with direct support of the Academy member P. L. Kapitsa
(Fig. 1) and his team.
Under supervision of the Department of LT, the liquefier was designed and manufactured
in the KSB works in Dn (Chart Ferox, a.s., now) , in the years 1957 to1958. It was a
cascade type using subcooling with liquid nitrogen and liquid hydrogen. The performance
was 8 liters of LHe or 15 liters of LH2 respectively. It was built in the new premises of the
LT Department in e near Prague, where in the night from 12 to 13 April 1960 helium
was liquefied first. As the second installation, Linde liquefier with performance of 3
liters/hour in the same year.
CUSTOMIZING OF KAPICAS LIQUEFIER
Increasing needs of other laboratories in Czechoslovakia and some other countries of the
Central and Eastern Europe faced difficulties with import from the USA or Germany,
where the helium liquefiers were already commercialized. Based on successful
manufacturing of the first liquefier, it was decided that commercial production would be
located at KSB, Dn.
This was again supported
by P. L. Kapitsa. His
Institute
provided
documentation as a base for
the new design, already
working as prototype, with
a piston expander instead
of hydrogen subcooling.
This type of liquefier was
suggested and first time
built by Kapitsa in 1934.
During the World War II,
liquefiers on this basis were
built by Collins in USA and
by Meissner in Germany.
The period of development
and manufacturing of these
liquefiers, namely ZH4 and
ZH9
(Fig.
3)
with
Figure 3. Helium liquefier ZH9 with the oscilloscope and the
production 4,5 and 9 liters
records of the indicator diagrams from top: a) normal, b) with
LHe/hour
respectively
premature... c) ...and delayed closure of the exhaust valve.
started in Dn in 1964.
Both the types were
equipped with expanders. Although manufacturing of apparatuses of copper and stainless
steel was on high level in the works that time, manufacturing of expander was a new
challenge for design and manufacturing technology.
The principal part of the liquefier, decisive for the correct function and especially for the
performance of this unit, is the piston expander. The expander must operate with a high
adiabatic efficiency and reliably. This is rather an exacting demand in view of the low
temperatures (28 to 12 K) of the moving parts such as the piston with cylinder and valves.
Since all known lubricants are unusable at the service temperature of the expander, gas
lubrication is effected here directly by the working medium, i.e. helium. There is a small
gap between the piston and cylinder, through which some of the expanded helium escapes
and forms a gas layer preventing contact of the surfaces of the piston and the cylinder. The
46

optimum width of this gap is given, on the one hand, by the admissible amount of the
escaping gas, which should be as small as possible, and, on the other hand, by the necessity
of preventing possible contact between piston and cylinder. The piston diameter amounts to
35 mm, clearance is 0,008 mm at the diameter at the working temperatures. Since the
material of the cylinder is stainless steel, while the piston is made of carbon steel with a
surface of plastic, both parts have different coefficients of thermal expansion and must
consequently be made with a wider clearance in such a way that only after cooling to the
working temperature is the mentioned value of 0,008 mm attained.
A part of the measuring system of the liquefier was an electric measuring system recording
a continuous indicator diagram of the piston expander during operation. This diagram
represents the dependence of the pressure in the working space of the expander on the
position of the piston in the cylinder. It provides the best information on the function of the
expander and it makes possible control of its operation on an oscilloscope (Fig. 3 a to c).
The first liquefier was finished in the Dn works (part of the trust Chepos that time) in
1965. Two years later, the liquefiers ZH4 and ZH9 were installed in several laboratories in
the Czechoslovakia, Poland and DDR (East Germany). Successively, other were built not
only in these countries but in Poland, Hungary and Bulgaria. Availability of liquefiers
enabled progress in basic research in physics, superconductivity and their applications in
countries of the eastern block of that time divided Europe. Encompassing of that time
difficult problems of manufacturing and operation of these liquefiers was only the first step,
which opened way to further development.
However unique were the first liquefiers with their
concept, the operation and maintenance required
continuous attention. Similarly it was with the
accessories. Oil lubricated piston compressors with
insufficient oil separation did not allow long time
operation. Helium, liquefied in a little separator
inside the liquefier, was periodically and with
thermal losses transferred to larger outside storage
Dewars, those of relatively small size with nitrogen
shielding. The expander worked on constant speed,
the only way of control of performance was the
filling of the cylinder by the timing of valves.
Modernization was the next step starting with 70s.
A new oil free piston compressor was developed
by CKD Prague (Fig. 4). The new liquefierrefrigerator enabled liquefaction directly into a
larger Dewar vessel with higher level of automatic
Figure 4. Oil free helium compressors
operation and lower maintenance.
Progress in applications of superconductive
magnets including first industrial applications
required large quantities of liquid helium. Two types new liquefiers-refrigerators with oil
free compressors and effective heat exchangers, as well as continuous modules of helium
purification were the answer to the challenge. They were still equipped with the Kapitsa
type expander, but of larger size and with electronic control of speed. Development of
control computers allowed continuous automatic operation in adjustable regimes.

47

The
largest
liquefier
designed in Dn, that time
already the state enterprise
Ferox, in the half of 70s. It
had two piston expanders,
built-in helium purification
and
fully
automated
operation. The performance

ZHR50

was

Figure 5.a, b. Ferox ZHR50 liquefier, 30 years in operation in the Slovak Academy of Sciences

was 30 to 100 liters of helium/hour or 150W at 4,5K in

refrigeration regime respectively. This liquefier is still in
operation at the Slovak Academy of Sciences in Bratislava.
(Fig. 5)
Another new type was a compact liquefier ZHR20 (Fig. 6),
with a single expander. Its performance was 25 liters of
liquid helium per hour. It was delivered for several
important projects on applications of superconductivity. It
was able to work in conditions of industrial surrounding in
months long operation.
HELIUM DEWARS, CRYOSTATS AND CRYOGENIC
SYSTEMS

Figure 6. ZRH 20 liquefier

with a helium Dewar
500 liter.

A He3- He4 diluents refrigerator was designed also in 70s

for continuous cooling of solid samples to temperatures in
the range of tens of mK ( Fig. 7 - with Stan Smr).
A range of superinsulated vessels for storage and transport of
helium, as well as purposely designed cryostats for the needs
of research in low temperature physics were developed and

delivered that time.

48

Figure 8. Cryostat for a superconductive quadrupole

Figure 7. He3- He4 refrigerator

Top
products

of that kind were

- the cryostat for a superconductive quadrupole installed at the Institute of Nuclear
Research in Dubna near Moscow (Fig. 8)
- cryostats for superconductive magnets of gyratrones for the Kurtchatov Institute of
High Energies in Moscow
- a super large helium cryostat 3,2 m in diameter and 10 m in length for tests of large
superconductive magnets for the IVTAN Moscow
- a cryostat for a NMR spectrometer developed by the Institute of Scientific
Instruments of ASCR was distinguished by a very low evaporation of liquid helium. It
was also manufactured at Ferox, then.
A very unique program was a rotating cryostat for a
superconductive generator koda with a power of 5
MVA with a superconductive rotor. The rotating
cryostat allowed cooling of winding at a speed of 3000
rpm with liquid helium delivered from the storage
vessel over a rotating joint.
For an experimental motor Skoda 55 kW with a
superconductive winding, a cryostat with minimum gap
between the superconductive stator and the rotor and
with ability to resist a high torque between the
superconductive winding and the outer vessel of the
cryostat was developed.
The most complex cryogenic system was developed,
built and operated for a project of a superconductive
magnetic separator of kaolin. By separation of oxides
of iron and manganese, high whiteness of kaolin was
Figure 9. SC magnet of the
achieved for production of top quality china.
separator of kaoline.
An industrial-scale magnetic separator for cleaning of In front: Demagnetization blanket of
kaolin clay was designed, installed and tested during
the separation matrix.
years 1983 1988 in a kaolin clay cleaning plant. The
superconducting system of the separator worked with
the magnetic field up to 5 T in the working space of the separator, whose active cavity
dimensions are 560 mm diameter and 1320 mm lenght (Fig. 9). The cryogenic system
comprises two helium liquefiers of different concept, one of them operating with a piston
type expander, the other one with an expansion turbine. This project was one of the first
real industrial applications of superconductive magnets in the world.
49

FROM THE PISTON EXPANDERS TO TURBINES

Further demand on industrial applications of superconductivity required highly reliable
liquefiers. Rotating machines were the answer of designers. Screw compressors instead of
the piston ones and radial turbines instead of piston expanders.
Development of expanders was started at the Research Institute of Food and
Refrigeration technology (ATEKO, a.s., in Hradec Kralove, today) in co-operation with
PBS, state enterprize, in
the middle of 80s. The
result was implementation of manufacturing
systems of two types of
turbines HEXT 0.5 with
the speed 216 000 and
237 000 rpm at the inlet
temperature 9K and 15K
respectively,
with
cooling power 91 W and
220 W (Fig. 10). This
design
concept
of
ATEKO is original and
unique in the world by
now.
Dynamic
gas
bearings lubricated by
the working helium and
Figure 10. a) Two ATEKO turbines on a liquefier b) Turbine cross
eddy current magnetic
section. c) Details of the shaft and the wheels.
brake
were
the
distinguishing
characteristics of the reliable and compact design. The other were small size, low weight,
high thermodynamic efficiency and a very simple speed control, which allows to maintain
optimum speed for achieving maximum thermodynamic and refrigeration efficiency
according to actual process parameters.
The first Ferox liquefier with two
turboexpanders ATEKO (Fig. 10 a) was the ZRH30T was supporting the above
described superconductive magnetic system of kaolin purification.
The last developed liquefier-refrigerator was again the two-turbine type ZRH3T, intended
for continuous refrigeration of helium cryostats, including those for magnetic tomography.
Oil lubricated screw compressor with highly efficient oil separator was used for helium
circulation. The development of this first Czechoslovak system of helium liquefaction
without any piston machines was finished by beginning of 90.
After political and economical changes by beginning of 90s the state supports of research
in superconductivity were drastically reduced and the existing liquefier markets collapsed.
This is why the design and manufacturing of helium liquefiers and cryostats was abandoned
in Ferox and the company, in frames of privatization by Air Products and Chart later,
concentrated fully onto the branch of storage and distribution equipment of liquefied air
gases and natural gas (LNG), which resulted in new progress and successful entering the
world market as one of the largest suppliers.
In total, Ferox delivered 45 helium liquefiers in the period 1964 to 1992.
Nevertheless, development and manufacturing of helium expansion turbines continued
successfully in PBS with deliveries to the main helium liquefier manufacturers in the world.
HELIUM TURBINES AS AN INDEPENDENT PRODUCTION BRANCH
50

First of all PBS deliveries of turbines started for Linde A.G., Munich. 258 turbines were
delivered for the liquefier LKKA for refrigeration of superconductive magnets Siemens
during the period 1988 to 1992. Turbines were delivered in two pieces for each liquefier.
The same it was with all the liquefiers described below except the liquefier Criotte Impianti
with four turbines PBS.
ATEKO developed new larger turbines HEXT 1 and HEXT 1.8 with cooling power
1000W and 1800 W respectively during 1988 to 1991. 14 of them were delivered to
liquefiers of Linde and Ferox. PBS, a.s. developed new type of 3D radial-axial wheels.
Measurements, done by ATEKO at the customer, indicated high turbine efficiency 78 to
80%.
Since 1991 PBS continued their development independently. Step by step, new types were
put over to other liquefiers.
Type HEXT 0.5 equipped with 3D wheels was delivered in 10 pieces for 5 liquefiers L5 of
Linde.
New type HEXT 1,5, cooling power 1500 kW has been delivered in 56 pieces since 1995
for the Linde liquefier, marked TCF 10.
Another achievement as acceptance of the next type HEXT2 with maximum power 2000 W
for the LAir Liquide liquefier HELIAL1000 since the year 2002 in the amount of 30
pieces.
Another type HEXT1.1 was delivered in four modifications for a university liquefier of the
company CRIOTEC Impianti in the year 2007.

Figure 11. The PBS HEXT turbine

a) (left): The turbine body.
b) The turbine 3D wheels 10 to 18,5 mm diam.

With systematic development and innovations, PBS

achieved considerable results in increasing of the thermodynamic efficiency of turbines
(Fig. 11) and reliability of their operation:
- The speed was increased up to 360 000 rpm, which enabled to design turbines with
maximum efficiency for all the input parameters.
- Circumferential velocity on the outer diameter of the wheel achieved 370 m/s.
- Implementation of 3D design and manufacturing of axial wheels on CNC machine tools.
(Fig. 11.b)
- Considerable increase of bearing capacity of the axial bearing. Circumferential velocity
on the outer diameter achieved 520 m/s.
- Radial bearings were improved
- Method of dynamic balancing of turbine rotors at operation speed was implemented.
Measurement of the amplitude of rotor vibrations proved considerable reduction at
commercial series production.
51

- CFD simulations of helium flow and consequently new guiding baffle profiles and
profiles of the channels of the low temperature part resulted in higher turbine efficiency.
- For reduction of parasitic heat flow into cold helium the low temperature part of the
turbines and the cold end of the shaft were redesigned.
- The control unit delivered with the turbines makes possible to operate turbines with
increased speed during the warm start and cooling down. This results in higher efficiency
and shorter startup time.
Total number of helium turbines manufactured and delivered up today is 376.
Development and manufacturing of turbines in PBS continues.
After customizing all this improvements
and achievements, PBS took active part at
development of baffle type cryogenic
compressors for helium refrigerating
systems for accelerators of elementary
particles for CERN, accelerator Elbe at
Rosendorf and others.
Flow systems were delivered for radial
compressors of Linde and Air Liquide
(Fig. 12).
Figure 12. Cryogenic compressor wheels
Portfolio of cryogenic products of PBS
60 to 250 mm diameter.
was enlarged by design and delivery of
flow parts and the low temperature part of
three circulators of supercritical helium for completion of the Linde delivery for the
accelerator Wendelstein 7 in Greifswald .
CONCLUSIONS
Development of helium liquefiers in Czechoslovakia was stimulating research of physics of
solid matters and superconductivity in the Central and Eastern Europe in the period 1964 to
1991. On this base, new branch of helium expansion turbines brought to a level of series
production and top efficiencies has been successfully progressed up today.
REFERENCES
1.
Kaiser Z.: Equipment for Helium Liquefaction, Czechoslovak Heavy Industry
(Journal), RAPID, Prague, July 1968
2. Kaiser Z., Fojtek J., Kouba M., Kotva J., uma J.: Magnetic Separator with a
superconducting magnet and a reciprocating matrix. Proceedings of the conference
CryoPrague 86, published by IIF-IIR at Ceuterix s.a., Leuwen 1986
3. Z.Kaiser, P.Vykydal, J.Fojtek,, S.Smr, M.Kouba, J.uma: Cryogenic Magnetic
Separator, Proceedings of the 9th Int. Conf. on Magnet Technology, Zurych, 1985
4. Tuek L., Kundera R.: Design, Manufacture and Operational Characteristics of helium
turbines used for helium liquefiers, Proceedings of the conference Cryogenics 94, House of
Technology, st nad Labem, Czech Rep. 1994
5. Kundera R., ervenka B., CFD application in cryogenic turbines design and testing,
Proceedings of the conference CryoPrague 2006, ICARIS, Prague, 2006

52

CR08-17
IMPROVEMENTS OF HELIUM LIQUEFACTION /
REFRIGERATION PLANTS AND APPLICATIONS
Muehlegger M., Berdais K.-H., Wilhelm H., Ungricht Th.
Linde Kryotechnik AG, Daettlikonerstrasse 5, CH-8422 Pfungen, Switzerland
ABSTRACT
Design features for a new range of helium liquefiers and refrigerators with capacities
ranging from 30 to 280 l/h of liquid helium (LHe) and 100 to 900 Watt, respectively. The
latest He cold box development shows an increased efficiency due to improved turbine and
heat exchanger design. Other benefits of the new design include short cool-down times and
a very compact design, which offers better flexibility and process control. The modularity
of the system was designed in order to cover a wide range of applications like sophisticated
shield cooling at different temperature levels or simultaneous operation modes for He
liquefaction and refrigeration purposes. The presentation will highlight the individual
improvements in the design.
During the presentation the influence of certain parameters like power requirement and cold
box inlet pressure in relation to the liquefaction and refrigeration capacity shall be shown
and discussed for the range of newly developed Helium liquefiers. In addition, the
presentation will cover the latest results of recently installed liquefiers in comparison with
the previous model.
Keywords: Liquefier, Refrigerator, L-Series, Turbine
INTRODUCTION
High reliability, availability, low operational costs and short delivery times have become
key requirements for small scale helium liquefiers. In addition, the demands concerning
product design and user friendliness have risen. By using standardised components, like
heat exchangers, expansion turbines and control software of highest quality, the L-Series
can be designed and manufactured according to customers requirements and specifications
maintaining short delivery times. This paper shows the impact of these optimised plant
components on liquefaction rate and specific power input.
LIQUEFACTION PROCESS
L-series liquefiers are designed for liquefaction rates at 4.4 K from 20 l/h up to 290 l/h. The
range is covered by three sizes of liquefiers - L70, L140 and L280, which are all based
upon a Claude cycle. FIGURE 1 shows the process flow diagram of an L-Series plant.
High pressure (HP) helium gas supplied by the compressor system enters the cold box. It is
cooled down in heat exchanger E3110 and E3120 by counter-flow low-pressure (LP)
helium gas. At the cold end of E3110 a liquid nitrogen (LN2)-evaporator is integrated so
that pre-cooling of the HP stream with LN2 becomes possible and the refrigeration or
liquefaction capacity of the plant is enlarged. The heat exchanger E3120 has two sections.
Between these two sections the high-pressure stream is split in two parts. The larger
fraction expands in turbine X3130. After a further cool-down in heat exchanger E3140 it
enters turbine X3150. It is expanded to low pressure and finally joins the returning JTstream. The smaller fraction, called Joule-Thomson stream, continues to be cooled down in
53

heat exchangers E3120 - E3160. After that it is throttled by the JT-control valve to dewar
pressure and gets partially liquefied in the dewar. The gaseous fraction is returned as a LP
stream to the cold box. It is warmed in the heat exchangers before returning to the intake of
the compressor.
Impure helium is fed to the integrated purifier. By cooling down the impure gas in counter
current with cold helium HP gas, impurities like nitrogen and traces of other gases
condensate and/or freeze out. The purified gas is fed into the cold-box HP-inlet side. By
warming up the purifier it will be regenerated and the impurities will be discharged.

Figure 1: Process Flow Diagram of an L-Series helium liquefier by Linde Kryotechnik AG

DESIGN FEATURES AND IMPROVEMENTS

Based on the well proven TGL turbine design and technology, the new TED (Turbo
Expander Dynamic) turbines have been developed. A TED 16 turbine is shown in FIGURE
2. Schoenfeld H., et al. [1] showed that efficiencies could be increased between 13 to 20%
per turbine. Improved bearing design allows higher thrust capacities and wider operation
range, leading to dramatically reduced cool-down times. The robustness of the TED turbine
is based on special bearing materials. The TED turbines are absolutely maintenance free.
Thus highest availability and reliability are ensured, providing an expected MTBF (Mean
Time Between Failure) of around 250,000 operating hours.
The L-series is equipped with a heat exchanger block, consisting of five counter-flow
brazed-aluminium plate-fin heat exchangers. All heat exchangers have the same block
length and identical plate-fin design concept. The heat-exchanger surface has been
increased compared to the TCF plants and pressure drops within the heat exchangers could
be reduced. The L-Series heat exchanger can be used for both liquefaction and refrigeration
plants. This concept ensures high operation flexibility in case of a combined liquefier and
refrigeration (mixed-mode) plant.

54

The outer piping has been minimised. The liquefaction and purification process has been
designed in such a way that no icing on outer pipe-work shall occur. The entire L-Series
(L70, L140 and L280) have the same design layout. Their external appearance only differs
by the cold box diameter. FIGURE 3 shows a L140 liquefier.
All L-Series plants use the same Siemens S7-300 control strategy and the same software
(PLC). In addition, a state-of-the-art visualisation is available for the L-Series. The control
panel is separated from the cold box, which provides additional flexibility for the
installation of the plant. Plant components are operated with a decentred control system and
are connected by PROFIBUS DA. Field wiring is minimised and maximum flexibility
concerning placement is ensured.

Figure 2: TED 16 turbine with cooler

Figure 3: State of the art helium liquefier L140

55

wo LIN
V_liq
[l/h]
153
115
101
70
59
46
34
28
21

TCF10 TCF20 TCF50

L280

Performance average
increase
increase
[%]
[%]
54
52
51
52
80
95
104
102
74
98
102
117

L140

with LIN
V_liq
[l/h]
190
150
130
78
57
45
39
27
19

L70

with LIN
V_liq
[l/h]
293
226
197
140
116
91
68
54
41

TCF10 TCF20 TCF50

L280
L140
L70

Compressor
massflow
[g/s]
79.5
61.6
54.2
40.8
34.3
27.5
21.1
17.7
14.5

wo LIN Performance average

V_liq
increase
increase
[l/h]
[%]
[%]
90
70
67
70
64
60
68
37
89
91
30
97
25
86
23
47
74
15
86
11
88

Table 1 Performance data of L-Series plants with and without LIN precooling compared to TCF
plants

COMPARISON OF PERFORMANCE DATA: L-SERIES VS. TCF

TABLE 1 shows performance data of the L-Series liquefier in comparison to the TCFSeries. Liquefaction rates of TED equipped L-Series plants are based on constant isentropic
efficiencies of 75% for turbine 1 and 80% for turbine 2, according to the measurements of
Schoenfeld et al. [1]. The cold box inlet state is 13.0 bar and 313K for L-Series and TCF
plants. Dewar pressure is 1.20 bar (4.4 K) for L-Series plants and 1.25 bar (4.45 K) for TCF
plants. The assumed heat in-leak into the dewar and transfer line are 10 Watts for both
types. Using the same mass flow and thus compression power, liquefaction capacities have
been increased between 47 and 97% for operation without LN2 pre-cooling. For operation
with LN2 pre-cooling, the increase of liquefaction capacity is between 51 and 117%. In [1],
a TCF20 was equipped with TED turbines. The increase of the liquefaction rate was 52%
compared to a TGL-equipped TCF20. The combination of TED turbines and improved heat
exchangers in L140 lead to an increase of up to 102% compared to a TCF with the same
compressor mass flow. Consequently, the specific compression power input per liquefied
helium could be significantly reduced. For liquefaction, specific compressor power is
defined as

liq [ kW l LHe @ 4.4 K ]

p = PComp m

(1)

where Pcomp is the shaft power of the helium recycle compressor. In TABLE 2 data for the
L-Series and TCF plants are presented. The required power for producing the LN2 is not
considered for the specific compression power.
For liquefaction of one litre liquid helium at 4.4 K without LN2, between 1.54 and 2.22 kW
of compression power is required for an L-Series plant. For TCF plants, between 2.62 and
4.06 kW are necessary (TABLE 2). This means that an L280 requires a third less
compression input power per produced litre LHe than a TCF50. TABLE 2 shows p for
liquefaction with LN2 pre-cooling as well. L-series plants require between 33% and 50%
less specific compression power for operation with LN2 pre-cooling.
RESULTS FROM THE FIRST OPERATING PLANTS
To date more than twenty L-series plants have been ordered and several have already been
commissioned. The first four plants are already in operation. The L280 at Ibaraki, Japan is
operating with 9.6 bar and 313K high-pressure inlet state.

56

wo LIN
p_spec
[kW/l Lhe]
1.54
1.78
1.70
1.85
1.81
1.95
2.04
2.01
2.22

TCF10 TCF20 TCF50

L280
L140

with LIN p_spec av. p_spec

p_spec decrease decrease
[kW/l Lhe]
[%]
[%]
1.24
35
34
1.36
34
1.32
34
1.66
44
49
1.87
51
2.01
51
1.77
43
49
2.08
50
2.42
54

L70

with LIN
p_spec
[kW/l Lhe]
0.81
0.90
0.87
0.92
0.92
0.99
1.02
1.03
1.11

TCF10 TCF20 TCF50

L280
L140
L70

Compressor
massflow
[g/s]
79.5
61.6
54.2
40.8
34.3
27.5
21.1
17.7
14.5

wo LIN
p_spec
[kW/l Lhe]
2.62
2.91
2.85
3.50
3.55
3.62
3.00
3.75
4.17

p_spec
decrease
[%]
41
39
40
47
49
46
32
46
47

av. p_spec
decrease
[%]
40

47

42

Table 2. Specific compression power for L-Series plants compared to TCF plants

Liquefaction capacities of 220l/h without purifier operation have been achieved. Cool down
time from ambient to operation temperature is less than 2.5 hours. To cool down a similar
TCF50, approximately 4 hours were necessary.
CONCLUSION
By using high-efficient TED turbines and optimised heat exchangers, liquefaction
capacities have been increased by 50% to 100% in comparison to the old liquefier
generation. The power input per litre LHe produced has been decreased by 33% to 50%.
The L-Series from Linde Kryotechnik AG established a new benchmark for small scale
helium liquefiers and refrigerators.
REFERENCES
1.

Schoenfeld H., Cretegny D. and Loehlein K., "Standard liquefier-test results with
improved turbines," Proceedings of ICEC 20, Beijing, China, 2004, pp. 119-122

57

CR08-47
EVOLUTION OF THE STANDARD HELIUM LIQUEFIER RANGE
Caillaud A., Aigouy G., Crispel S., Delcayre F., Grabi V., Dauguet P.
AIR LIQUIDE, Advanced Technologies Division, Rue de Clmencire,
B. P. 15, 38360 Sassenage, France
ABSTRACT
The standard helium liquefier and refrigerator range, called Helial and designed by Air
Liquide DTA, has recently been upgraded in order to improve the efficiency of these
machines. Indeed, over the multi-range markets requiring these cryogenic systems,
(international laboratories, aerospace applications, synchrotrons, HTS applications...), the
technological solution has to provide increasingly high performances. The new range,
equipped with very reliable DTA turbo-expanders, constitutes a highly efficient product for
this wide application field. The optimizations, adaptations and results of the Helial
Evolution series, doubling the performance for the same power consumption, will be
presented.
INTRODUCTION
The Helial was born in the 1980s. These machines were revolutionary in that they
constituted the first helium liquefiers to be fully automatic and therefore easily operable.
Nearly 30 years later, their refrigeration and liquefaction capacities have grown
enormously, but helium liquefiers-refrigerators still operate on the same principle. With a
30-year wealth of experience, Air Liquide decided to assess the situation, looking through
all the projects. This study led in 2007 to the launching of a new range named Helial
Evolution. Indeed, in the demanding high-tech markets, cryogenic systems such as standard
liquefiers and refrigerators must provide increasingly high performances with strong
reliability. This paper describes the capitalization of the last six years regarding small and
medium liquefaction. The results of this study are presented as well as the evolutions they
generated on the standard machines. Process and design improvements are detailed, and
finally benefits and adaptations of these new standard machines are described.
1. CAPITALIZATION OF THE PAST SIX YEARS
In 2001, Air Liquide upgraded its standard helium liquefiers range constituted by the Helial
7, Helial 20 and Helial 50. Three new liquefiers/refrigerators were created and called Helial
1000, Helial 2000 and Helial 3000. Their performances, which are presented in TABLE 1,
resulted from an evolution of the liquefaction market characterised by the arrival of third
generation synchrotrons requiring dedicated cryogenic systems with refrigeration and
mixed-mode operations. At the same time, small liquefiers were still required all over the
world, particularly in Asia with the premises of helium liquefaction in new developed
countries.

59

Helial 1000
40 L/h

Liquefaction capacity without LN2

Helial 2000
85 L/h

Helial 3000
180 L/h

Liquefaction capacity with LN2

80 L/h
175 L/h
Refrigeration capacity without LN2
130 W
415 W
Refrigeration capacity with LN2
160 W
500 W
Table 1. Helial 1000/2000/3000 performances

300 L/h
750 W
900 W

The design of the Helial 1000/2000/3000 range was therefore adapted to these specific
requirements, trying to find a good compromise between refrigeration and liquefaction, in
order to fulfil specifications requiring various modes of operation.

Cold neutron
source

Space
simulation
chamber

Applications

Operation modes

HTS
Liquefiers

Synchrotrons

Neutral Beam
Injector

Refrigerators

SMES
Liquefaction
centers

Mixed modes

Operating temperatures

Margin between guarantee and measurement

T<4.5K
T>10K

0% < Margin < 5%

5% < Margin < 15%
15% < Margin < 30%
Margin > 30%

T = 4.5K

Figure 1. Characterisation of past Helial projects.

After six years of experience with the Helial 1000/2000/3000 range, past projects were
analysed. First, Air Liquide examined the field of applications for which these systems
were installed. FIGURE 1 upper left corner shows the variety of customer applications
for which the Helial systems provided a solution. Two main fields appear: synchrotron
centres and liquefaction centres. They constitute about 75% of the needs in terms of helium
liquefaction and/or refrigeration. Nevertheless, the remaining 25% reveal new markets like
cold and ultra-cold neutron sources, HTS applications or neutral beam injectors for fusion
applications, which should develop in the near future for helium refrigeration. Therefore,
Helial machines should constitute a solution for cryogenic needs within these new fields.
On the upper right corner of FIGURE 1, the operation modes for past projects have been
distinguished highlighting a quasi-perfect repartition between liquefaction, refrigeration
and mixed modes. These two repartitions, i.e. final applications and operation mode
required, convey the difficulty to design a standard machine able to cover the whole range
of applications, satisfying the diversity of users.

60

Regarding the operating temperatures, FIGURE 1 bottom left corner shows that most of
the projects work at liquid helium temperature. Nevertheless, in recent years, more and
more projects at temperatures above 10 K were born in different fields such as HTS
applications, cold neutron sources and space chambers. The standard machines must remain
flexible to allow, with minor modifications, operation at this temperature range.
Finally, the last part of the analysis of past projects concerns the on site performance. The
graph on FIGURE 1 bottom right corner shows that more than 40% of the installations
in operation exceed the guaranteed performance by more than 30%. These conservative
results led to a better knowledge of the machines as well as of their components. This better
control of the process and component limits encouraged Air Liquide to propose a solution
closer to the customer needs.
Moreover, power consumption becoming a major topical concern, Air Liquide decided at
the same time to analyse the efficiencies of the Helial range. Each machine can be
characterised by a specific consumption, which is defined as the ratio between the power
dissipated to ambient at warm end and the power absorbed at the cold temperature. TABLE
2 shows that the specific consumption of each Helial corresponds, for a liquefaction mode
without liquid nitrogen (LN) pre-cooling, to about 8% of the percentage of the Carnot
efficiency. For machines characterised by an equivalent cold power between 150 W and
900 W, the percentage of the Carnot efficiency should be between 10 and 15% (FIGURE
2).

Equivalent power for liquefaction w/o LN2

Helial 1000

Helial 2000

Helial 3000

137 W

290 W

614 W

Specific consumption for liquefaction w/o LN2

967 W/W
862 W/W
% Carnot
7%
8%
Specific consumption for refrigeration w/o LN2
1015 W/W
602 W/W
% Carnot
6%
11%
Table 2. Helial 1000/2000/3000 efficiencies

732 W/W
9%
600 W/W
11%

This differences in performance can be explained by the multi-mode optimised design of

these machines, which was determined by the strategy decided upon by Air Liquide seven
years ago. This choice permitted to propose a single multi-purpose product in order to
cover a large market of applications, but had the drawback of not being optimised for pure
modes. The results of this capitalisation led Air Liquide to decide a new strategy
characterised by dedicated machines, i.e. refrigerators and liquefiers, in order to improve
the specific consumptions in a context of energy consumption reduction.
The process was then reviewed in order to dedicate and optimise a standard machine to one
pure operation mode, taking into account the acquired experience on components limits and
machines behaviour. This led to the upgrade of the Helial range.
2. HELIAL EVOLUTION RANGE
2.1 Process Optimisation
Helial machines can operate in a liquefaction mode or in a refrigeration mode. For both
modes, the basic cycle is the same, remaining a Claude cycle with two turbines installed in
series. Nevertheless, the components of the cold box do not have the same contribution

61

depending on the operating modes. The process studies performed were based on the
differences between refrigeration and liquefaction modes.

Figure 2. Efficiency for refrigerators and liquefiers compared to Carnot efficiency (Strobridge
[1])

In refrigeration mode (Figure 3 left side), the customers application is maintained at a

constant temperature with a liquid helium bath, where evaporation absorbs the power
injected into helium. Some gas is generated and recovered in the cold box heat exchangers.
The boil-off participates to the cool-down of the high-pressure (HP) gas to be liquefied by
counter-flow exchange. In this mode, the power extracted by the turbines compensates only
for the heat losses of the system, (i.e. non-reversibility of heat exchangers), heat-in leaks to
internal components of the cold box and the heat load transferred by the customers
application to helium. As shown by the left picture, the refrigerator is aimed at removing 18
J/g from helium gas (which corresponds to the latent heat) to condensate again liquid
helium in order to maintain the level.

62

14
13

1
N2

12

11

10

T1

15
16

T2

= 18 J/g

~30 J/g
(GHe @4.5K)

LHe

GHe 1570 J/g

14
13

~12 J/g (LHe @4.5K)

1
N2

12

11

10

T1

15

T2

= 1540 J/g

16

~30 J/g
6

7
LHe

= 18 J/g
~12 J/g

Figure 3 : Differences between refrigeration mode (left side) and liquefaction mode (right side).

In liquefaction mode (Figure 3 right side), no cold gas is returned through the heat
exchangers except for the flash generated by the Joule-Thomson expansion. Therefore the
turbines must balance not only the heat losses of the system but also must extract the power
to cool down the helium to be liquefied. The liquefier must remove about 1540 J/g in order
to cool the gas down to the liquefaction point, and then again 18 J/g in order to liquefy
helium.
In that way, one can easily understand that in a liquefaction mode, turbines have to extract
more power than in a refrigeration mode, whereas in a refrigeration mode, the heat
exchangers must provide the largest surface, particularly for the last heat exchanger, so as
to recover the maximum cold enthalpy of the cold gas generated by the application. When
the Helial was designed for a mixed mode, the efficiency in case of a pure liquefaction
mode or a pure refrigeration mode would not have been the maximum since turbines and
heat exchangers are operated in off-design modes.
63

Within the diversity of applications, the requirements for liquefaction and refrigeration
modes are very different in terms of process specificities, scope of supply, control system,
and system operating approach. The recent evolution of the Air Liquide Helial range relies
on the choice of the design mode in order to propose a standard system optimised for the
desired operation mode.
The range is therefore split into two series: liquefiers and refrigerators. This separation
permits one to obtain the maximum specific consumption with a given compressor for the
operation mode chosen by the customer. The new Helial Evolution range is composed of
three sizes of cold boxes : small, medium and large. For each size, the customers
specification leads to the design mode for which the performance will be optimised:
liquefaction or refrigeration. The new range now consists of three liquefiers: Helial SL,
Helial ML and Helial LL; and three refrigerators: Helial SF, Helial MF and Helial LF. Even
if the external parts of the cold boxes of Helial refrigerators and liquefiers of the same size
are identical, the heat exchangers and the turbines will be different being optimised for the
chosen range.
2.2 Turbines Improvements
Some studies have also been performed regarding the influence of turbines efficiencies. We
estimated the impact of increasing efficiency on the liquefaction rate. The results of the
calculation are given in Table 3. This information pushed Air Liquide to work on the
efficiency of the turbines, trying to improve these figures on their smallest machines that
only extract several hundreds of watts.
This has been managed thanks to the use of 3D-open wheels as it was already done on
larger turbines. About ten efficiency points have been gained and demonstrated on our
specific test bench in DTA, which permits testing the turbines under real cold conditions.
These good results, in addition to the optimisation resulting from the choice of operating
mode, also contributed to the improvement of the performances (as shown in Table 3).
Additional studies are still in progress with the goal of gaining further increases in
efficiency.
Turbine 1
Effect of one additional point upon the liquefaction rate (%)
+ 0.9
Table 3. Influence of turbines efficiencies on liquefaction rate

Turbine 2
+ 1.25

2.3 New Range of Performances

The liquefier optimisation performance results are summarised in Table 4.
As pointed out in the table, the specific consumptions of the new liquefiers when compared
to the former range are much lower. The specific consumption for liquefaction without LN
pre-cooling as a percentage of Carnot is now around 12%, which constitutes a good
position on the curve of Figure 2 and particularly considering the small size of these
machines. In terms of performances, the Helial ML anticipates a doubled capacity with the
same cycle compressor and size than the Helial 1000. The Helial LL offers 78% more
capacity than the Helial 2000 when both use the same compressor, and a small adapted
machine for low capacity requirements, the Helial SL, is born.
This improvement means lower operation costs thanks to a better adaptation to the
customers needs, process optimisation and better use of the main components.

64

HELIAL SL
25 L/h

Max. Liquefaction capacity without LN2

HELIAL ML
70 L/h

Max. Liquefaction capacity with LN2

50 L/h
150 L/h
Compressor electrical motor
55 kW
132 kW
Specific consumption for liquefaction w/o LN2
645 W/W
552 W/W
% Carnot
10%
12%
Table 4. HELIAL Evolution liquefiers performances

HELIAL LL
145 L/h
330 L/h
250 kW
505 W/W
13%

3. BENEFITS AND ADAPTATIONS OF THE STANDARD PRODUCT

The standard Helial units are dedicated to be used for a maximum range of applications,
from liquefaction centers to refrigeration applications. Hence, this product must fulfil a lot
of different requirements and should remain flexible.
3.1 Benefits of Standard Helial Range
First, this product must remain easy to operate, since the number of operators dedicated to
cryogenics has been decreasing in most centers. The Helial is then totally controlled by a
dedicated performing control system, in order to adapt automatically operations to the load
variations of the system. It is equipped with suitable sensors and automatic valves which
makes it easy to operate. The control system can also be exported through a supervisor with
friendly interfaces. This supervisor can offer different levels of control from simple
monitoring of parameters up to a total remote control with orders given through the
supervising system. Diagnosis and permanent check with recording is then made easier. Air
Liquide DTA also proposes a remote access to the PLC through a modem so that an office
expert can give direct assistance to each customer.
Helial product also constitutes a high reliability solution with reduced maintenance, resting
on the use of very robust components such as oil lubricated screw compressors, aluminium
plate fin heat exchangers and Air Liquide own static gas bearings turbines, characterised by
a calculated MTBF higher than 150,000 hours. These turbines are designed, manufactured
and tested in real conditions in the Air Liquide DTA workshops, which allows control of
the whole process, and qualification of the turbines before delivery. This secures the
delivery schedule and the performances of the machines. The reduced maintenance enables
one to propose a product adapted for a long and continuous operation time without shutdown, with a control system providing permanent monitoring of all the parameters,
permitting auto-diagnostic and safe-guarding the system from unanticipated stops. Thanks
to its worldwide experience, Air Liquide developed some partnerships with subcontractors
and suppliers, and can propose a product which complies with all the regulations and
norms.
This product is also designed in order to limit the operation costs. The new optimised Helial
Evolution range led to lower power and utility consumptions due to improved system
efficiencies. With the automated control system, a permanent operations teams is not
necessary, which also leads to the operation cost reduction.
Finally, Air Liquide can propose a turn-key system in order to take responsibility for the
entire cryogenic project. This includes the proposal of complete solutions including on-site
services such as the installation work, and maintenance and operation contracts which have
been developed. The delivery time has also been reduced by modifying the manufacturing
process so as to comply with customer requests to shorten the project construction phase.
65

CONCLUSION
The Helial Evolution constitutes a standard product line providing high performances with
high reliability and efficiency. Performances have been considerably increased for the new
Helial Evolution range. This product is the answer to a multi-range markets and remains
adaptable to specific requirements.
REFERENCES
Boissin, J.C., Gistau, G., Hbral, B., Pelloux-Gervais, P., Ravex, A. and Seyfert, P.,
Cryognie : Mise en oeuvre des basses tempratures, in Techniques de lIngnieur, B2
382, pp. 2-4.

66

CR08-40
LIQUID HELIUM IN LABORATORY USE
PRACTICAL REMARKS
Haberstroh Ch.
Lehrstuhl fuer Kaelte- und Kryotechnik, TU Dresden, Germany
ABSTRACT
Within the last decades an uninterrupted increase in the use of liquid helium for laboratory
applications was recorded. On the other hand increasing problems and excessive costs
caused by incorrect use or lack of knowledge about helium specifics are reported.
In this contribution a number of typical aspects are addressed, like commercial supply of
liquid helium vs. operation of an own liquefier, the use of helium dewar vessels and
balancing of helium amounts, usual causes for leakage and for contamination as well as
special features of helium recovery systems.
INTRODUCTION
Liquid Helium (LHe) is usually used for cooling tasks below liquid nitrogen temperature
level, mostly in the range between 2 and 5 K. The LHe demand especially in Germany can
be characterized as follows:
There is an increasing number of universities as well as national and semi-national
institutes with current utilization of LHe. Several of them had been founded within
the last decades.
The specific LHe consumption at the single points of use and thus the overall
consumption of the whole institutes has raised appreciably, compared e.g. with the
situation in the 70th or 80th.
Both aspects can be attributed at least partially to changes in the cryo technology within
the last decades: in the 70th or 80th typically small table-top flow cryostats or bath
cryostats with a capacity of a few liters only had been in use. Often these had been selfdeveloped and self-fabricated devices. For experiments these were cooled-down and used
for some hours at dedicated days only.
Meanwhile a wide range of bath, flow and magnet cryostats are commercially well
available and are often delivered as a turn-key system within a complex experimental setup. This does first broaden the subgroup of potential users: whereas in former times
experiments at LHe temperatures had been subject for dedicated cryo specialists only,
today as well chemists, material scientists or physicists find themselves among the LHe
users, mostly not involved in cryogenics else wise. Often the attention paid to the LHe
cooling is similar low as to cooling water of similar supplies, and the LHe handling is done
according to the manual of the system supplier, without further knowledge of any LHe
specifics.
Secondly, the dimensions of those commercial cryostats are appreciably bigger and more
complex compared with earlier self-made devices. Improvements in thermal insulation are
outbalanced by that by far. Moreover most of the commercial superconducting magnet,
NMR and ESR systems are customarily or necessarily kept at LHe temperatures
continuously. Thus a permanent LHe demand is generated all over the year, independent of
the periods of use.
67

1. COMMERCIAL LHE SUPPLY

As for Germany, meanwhile an excellent infrastructure can be stated concerning the supply
of LHe. LHe can be ordered from several gas companies as a standard product and is
delivered at short notice to almost all places. Transport dewar vessels with a capacity of
50 l mostly 100 l and up to 500 l are used. Prices are found between 6 and 12 per liter
LHe, including rental dewar vessel and transport fees. The helium gas released at the user is
either blown to the atmosphere or collected and compressed to high pressure cylinders,
in case a helium recovery system is installed in the respective institute.
For the institute administration such a commercial supply can be very attractive:
1.

onset of LHe utilization almost instantaneously,

2.

no or only minor invest cost in advance,

3.

no additional space or staff requirements.

Moreover a LHe provision from an external supplier fits well to the common trend of
outsourcing, what stands for reduction of all activities other than the institutes own
dedication.
Thus a LHe supply on commercial terms is practiced by many institutes, sometimes with
remarkable increase of volume upon successful implementation. Sometimes up to 100 000 l
LHe per year are obtained, i.e. a turn-over of 20 transport dewars per week and costs
beyond M per year.
Two aspects have led to a more critical estimation of a solely commercially-bases supply:
on the one hand delivery bottlenecks had to be registered in the past, concerning all gas
companies and the entire European market for periods of several weeks each time in the
last years - obvious with increasing frequency. E.g. in the summer 2000 as well as in the
autumn 2001, 2005 and 2006 or at the end of 2007 only parts of the necessary LHe
quantities could be delivered, partly had test series to be abandoned and MRI devices
allowed to warm-up.
Secondly: in institutes with large and regular LHe consumption there is generally a helium
recovery plant installed at least. Re-compressed helium gas at typically 200 bar is returned
to the gas company against an appropriate credit note. Thus the effective purchase price is
reduced significantly. The gas companies show up, however, increasingly less interest in
returned helium. The acceptance is rejected generally or only a poor refund of 1 ... 2 /m is
granted. Apart from high logistic expenditure the principal reason lies in the fact that there
is nearly no need for further gaseous helium (besides easily contaminated).
Both has its cause in the same point: Helium is won exclusively as by-product from the
natural gas extraction. Appropriate plants for this are installed at He-rich deposits, which
are predominantly situated overseas in the USA, in North Africa and in the Middle East.
For transport the helium is locally liquefied and transported in locked 11,000 gal overseas
containers to Western Europe. Due to the heat input during the transport the helium is
found in supercritical state at approx. 4 ... 5 barabs. finally. At the destination therefore only
about half of the contents can be filled up in liquid form and delivered to the LHe
customers. The remaining helium is merchandised as high-pressure helium gas, usually just
sufficient to meet the complete gaseous helium demand. Therefore there is normally no
reasonable use for further helium gas coming from the customer, moreover for a more or
less impure one.

68

2. HELIUM BALANCING
With the balance of helium conversion values apply in accordance with table 1. Special
attention should be paid to the high density of cold helium gas at 4.2 K or at slightly
elevated temperatures: in good approximation 10 % of the liquid density can be assumed.
Therefore still a tenth part of the LHe capacity is contained in an empty", but cold helium
dewar.
Table 1: Conversion numbers for helium [1]

liquid @ 4.2 K / 1013 mbar

1l

gaseous @ 4.2 K / 1013 mbar

7.2 l

gaseous @ 0C / 1013 mbar

702 l

gaseous @ 15C / 1 bar

751 l

The gaseous state at 0C /1013.25 mbar is often denoted as standard temperature and
pressure (STP) or Tnpn, in physics or science one mostly refers to that. At the gas industry
against it the so-called reference state (Bezugszustand) at 15C and 1 bar is used
exclusively [1]. This leads in practice frequently to wrong conversions and
misunderstandings.
For the use of commercially supplied LHe the scenario in an unfavorable case can be found
as follows:
Table 2: Balance example of poor utilization of a LHe dewar vessel

dewar filling level

order

100 l

filling ex factory and account

96 l

after transport

94 l

after 14 days stand-by

80 l

after cryostat refill

20 l

liquid return

20 l

In view of organizational and process uncertainties usually the LHe is ordered with some
buffer time. With a typical boil-off rate of approximately 1 l LHe/day the level already sank
first around 14 l. Secondly, against the wide-spread judgment, in this example only about
54 l instead of 60 l were transferred to the LHe cryostat. Approximately 10 % of the
quantity must be vaporized and remain in the transport dewar to replenish the void volume.
Remaining liquid is generally not credited with commercial supply. Actually in the
example shown above thus the real price per l LHe has nearly doubled. Against it, in an
optimized utilization down to the necessary 2 ... 3 l remaining liquid content and with
minimized evaporation losses, a substantially better balance can be effected. It should not
be forgotten however that a discrepancy of at least 20 % compared to the account balance
still remains.
69

3. LHE SUPPLY BY SELF-OPERATED LIQUEFIER

In Germany actually nearly 50 helium liquefiers are in use, several of them taken into
operation within the last years only. Liquefiers are operated mostly by universities or e.g.
by institutes of the Max Planck or Leibniz society. Those plants generally comprise the
following components:

cold box (liquefier) and cycle compressor

oil removal; line drier

gas management and gas analysis

stationary LHe vessel and dewar filling station.

For the recovery system in addition is needed:

feed-in ports and piping

a helium balloon (10 40 m)

a high pressure compressor for 200 300 bar

a high pressure storage volume at 200 300 bar.

Approximately 200 m of floor space are required, partly outside of the building. Additional
supply of cooling water and of LN 2 if used for pre-cooling must be taken into account. The
specific electric power consumption has dropped remarkable due to efficiency
improvements in the liquefier technology. 2 kWh per l LHe now can be calculated contrary
to 4.5 kWh per l LHe in the past. Besides that the additional power consumption of the high
pressure compressors must not be neglected: for an at least one-time re-compression of
each helium batch to e.g. 200 bar approx. 0.5 kWh/l LHe must be added. The total
investment sum for the components specified above is at present around 2 M. Two
operators must be engaged for operation and for the decant service. The installation of an
own liquefier plant therefore is a quite fundamental decision for the respective institute.
With an own plant the costs for mere liquefaction can be beat down to 1 / l LHe or even
appreciably lower (plant depreciation, building and personnel costs usually not directly
allocated to the single institutes within national institutions). The re-liquefied helium is
distributed next the places of use, the helium gas released from the experiment has to be
recovered as far as possible (a recovery rate of 90 95 % stands for good practice). Losses
must be compensated by purchased helium, actually at a price level equivalent to
5 10 / l LHe (regardless whether in liquid state of as high pressure gas). Beyond that,
for spring 2008 a further price increase of more than 40% was announced by the gas
companies. Thus the operation of an own liquefaction plant makes only sense if a high
recovery rate is sustainable. In case of losses up to 1/3 of the helium amount per cycle, as
observed particularly at new installations sometimes, the price advantage in relation to a
commercial supply shrinks dramatically or turns even into the opposite.
4. CAUSES FOR HELIUM LOSSES
Often the causes for helium losses in spite of an existing recovery system are unclear to the
responsible persons. In most cases quite trivial causes can be identified by on closer
inspection:

70

transport dewars not connected to the recovery system

bath cryostats cut off from the recovery system during refill (for insufficient
dimensioned lines or instruments)

LHe dewars are opened despite elevated pressure inside (due to its high density the
released gas volume equals to an appreciable helium amount)

excessive cool-down of the transfer lines before insertion into the cryostat

forgotten ball valves or helium paths at the cryostat warmed-up and opened after
completed series of measurement

loose line fittings (e.g. widened ends of PVC-hoses)

The helium recovery systems are usually held on some mbar positive pressure. By latter
point often unnoticed, slight leakages are caused, which can accumulate however to
substantial losses.
5. HELIUM CONTAMINATION
In certain way a reverse problem is additional gas i.e. impurities in the recovered
helium. Mostly this concerns ambient air (perhaps in somewhat changed composition), in
some cases also pure nitrogen (from pre-cooling). Practically all liquefiers therefore are
equipped with freeze-out purifiers, which can remove such impurities within certain limits.
Tolerable are generally concentrations up to some tenths percent by volume. Impurity
levels exceeding 2 ... 3 vol-% become problematic, since from this level the purifiers
normally are overtaxed.
As causes for infiltration of impurities are found mostly:
-

Faulty operation of the cryogenic equipment (often from unawareness, often in case
of acquired turn-key equipment or insufficiently understood cryostat technology,
often with insufficiently instructed experimenters with different specialization)

Leakages at pumped flow cryostats (despite the quiescent evaporation develops in the
LHe supply dewar with longer cryostat runs a negative pressure; smallest leakages
than result in sniffing in ambient air. For security the LHe dewars in such cases are to
be kept pressure-free over a compensating line.

Diffusion through hoses and balloon material (cf. 6.1)

Local negative pressure within the recuperation system (cf. 6.2).

5.1 Helium balloon diffusion problems

Usually in todays recovery plants a balloon is installed as temporary buffer storage before
re-compression. Mostly these are ball- or zeppelin-shaped with a capacity of 10 m to
30 m. As jacket material chloroprene rubber (CR, Neoprene) coated nettings are used
(thickness d 0.9 mm, 1.1 kg/m), newly as well PU-coated nettings (d 0.15
0.35 mm). All these materials have a more or less pronounced permeation for the different
kinds of gas. Generally each partial pressure for its own tries to equalize. So helium would
diffuse to the outside, the different atmospheric gases into the balloon. For a Neoprene coat
this way helium losses of about 1.1 l gaseous (at 15C/1 bar) per m and per 24 hours are
found. After all this sums up to about 16 m per year for a typical 20 m balloon with a total
surface of ~ 39 m.

71

With a PU-foil material these losses are reduced to about 0.4 l He gaseous / m 24 hours.
Meanwhile as well PU based material with additional diffusion barriers is available (nano
platelets, 100 x 100 x 5 nm in size). Helium losses are quoted here with 0.03 l He
gaseous / m 24 hours.
More problematic usually is the incorporation of atmospheric molecules into the helium
balloon. Unfortunately only vague diffusion parameters are known here. Generally these
should be lower by a factor 40 approx. for N 2, O2 or CO2. Against it for the polar H2Omolecule comparably high diffusion numbers are found.
5.2 Sub-atmospheric pressure in the recovery system
With a certain configuration of a helium recovery system it can come to a negative pressure
at certain feed points. This illustrates Fig. 1. As shown here, the recovery system extends
over a laboratory building with several floors. The helium balloon is installed here
according to general practice in the attic. The helium inside is held at ambient pressure
(1000 mbar supposed). The pressure level of the atmosphere for different height computes
itself from the barometric scale factor. Close to ground level in good approximation a linear
change of 0.12 mbar/m can be set for air (15C/1 bar; 1.21 kg/m). For helium at
15C/1 bar ( 0,167 kg/m) in the same way a pressure change of only 0.017 mbar/m can
be calculated. This results, as indicated in Fig. 1, in an air pressure change of approximately
0.5 mbar from floor to floor. Within the helium system, however, a pressure change of only
0.1 mbar per floor is effected. In result according to the example given in Fig. 1 in the
floors below the attic (without the check valve already drawn in) a negative pressure in the
helium lines between 0.4 mbar and 1.6 mbar is found.
This represents a perfectly unacceptable condition:
-

Small leakages or forgotten valves at the inlet ports lead immediately to immense
impurities (with positive pressure naturally on the other hand to appropriate helium
losses)
There is a severe risk to suck in ambient air into LHe dewars or cryostats connected to
the recovery system. Frozen air can plug neck tubes and blow-off lines, the latter
being an extremely dangerous situation.
Air

Helium
1000 m bar

attic

1000 mbar

3rd floor

+ 0,5 mbar

2nd floor

+ 1 mbar

1st floor

+ 1,5 mbar

check valve
(ca. 5 mbar)

- 0,4 mbar

+ 0,1 mbar

4.2 m
- 0,8 mbar

+ 0,2 mbar
- 1,2 mbar

+ 0,3 mbar

??
basem.

+ 2 mbar

- 1,6 mbar
+ 0,4 mbar

72

Figure 1: Scheme of a helium recovery system in a multi-storage laboratory building with sub
atmospheric pressure at the inlet ports.

The self-weight of the balloon jacket in this respect does not play any relevant role. In the
numerical example for a 20 m balloon from heavy neoprene material a jacket mass of
22.5 kg is distributed over a cross section of ~ 11 m. The resulting positive pressure p is
p = 22.5 kg g / 11 m = 0.2 mbar
For a reasonable overpressure of 4 mbar an additional weight of some 450 kg would be
necessary. The right solution is of course the introduction of an adequate check valve with
an opening pressure of some mbar, as already indicated in Fig. 1. Nevertheless a negative
pressure still remains at the suction side of the recovery compressor.
REFERENCES
1. 1 x 1 der Gase - Physikalische Daten fr Wissenschaft und Praxis (fluid handling and
data collection for science and practice), Air Liquide GmbH, 2005 (www.airliquide.de)

73

PULSE TUBES
AND
OTHER REFRIGERATORS

75

CR08-31
EFFECT OF ALTERNATE TUBE CHARACTERISTICS ON HIGH
CAPACITY PULSE TUBE CRYOCOOLERS PERFORMANCE
Saidi M.H., Sarikhani N., Jafarian A., Hannani S.K.
Center of Excellence in Energy Conversion, School of Mechanical Engineering,
Sharif University of Technology, Tehran, Iran
ABSTRACT
High capacity pulse tube cryocoolers offer the promise of the cooling capacity required for
operation of superconducting devices. The purpose of this paper is to investigate the
influence of the alternate tube characteristics on a high capacity pulse tube cryocooler
performance, intended to achieve 250 W at 80 K. In this respect the hydrodynamic and
thermal behavior of the cooler is explained by applying the mass and energy balance
equations to different parts of the cryocooler. Nodal analysis technique is employed to
simulate the tube section behavior numerically. Employing the proposed model the effect
of essential characteristics such as tube aspect ratio, frequency of oscillations, reservoir
volume and double inlet is considered. To determine the optimum operation conditions that
maximize the enthalpy flow and thus the coefficient of performance, net enthalpy flow
analysis is conducted. The results of this analysis are tested and validated by comparing
them with the experimental data.
INTRODUCTION
The use of very low temperatures for application in superconducting devices, space,
military and medical equipments requires the achievement of high performance cryogenic
systems. Pulse tube refrigerator as a reliable cryogenic cooler, contains no moving parts at
its cold head, and thus having considerable system advantages over the most other types of
cryogenic refrigerators in terms of reliability, life time and low vibration and cost.
However, the older types of cryocoolers like GM and Stirling coolers are going to be
replaced by pulse tube refrigerators in a wide variety of applications.
Gifford and Longsworth [1] were the first who introduced pulse tube refrigerator in 1966.
The modern pulse tube cryocooler, which was equipped with an orifice and a surge volume
on the warm end, was developed by Mikulin [2] in 1984. Due to this modification the
performance of the pulse tube increased and became comparable to the performance of
practical coolers such as Stirling cycle, Gifford-McMahon and Joule-Thomson cryocoolers.
In 1990 Zhu et al. [3] added a bypass to the device and introduced the double inlet pulse
tube refrigerator. The improvement of the double inlet pulse tube refrigerator was achieved
by implementing a bypass between the central zones of the Pulse tube and that of the
thermal regenerator. This improvement led to a new configuration named Multi Bypass
Pulse Tube Refrigerator [4]. By the end of the 1990s, temperatures below 2 K were
achieved with a three stage pulse tube [5]. In recent years two major models of pulse tube
refrigerators are currently under development. The first style, known as G-M type, is a
variant of the Gifford-McMahon cryocooler. The second type of pulse-tube refrigerators is
known as Stirling type.
In this paper we present the design and optimization of a high capacity Stirling type double
inlet pulse tube refrigerator (DIPTR), which is intended to produce 250 watts of
refrigeration at 80 K. Special precautions have to be considered in the design and
77

functioning optimization of the high capacity DIPTR systems. In fact, the energy
consumption of the high capacity cryocoolers is significantly greater that of the medium
capacity ones. Thus, the coefficient of performance that considers both the cooling power
and the rate of work transfer to the gas should be optimized to ensure the highest cooling
capacity and minimum work. Here, we study the influences of the cooler geometry and
operating parameters of the tube section on the cooling capacity and coefficient of
performance by using the developed model. Furthermore, in order to optimize the
performance of the pulse tube cryocooler, the effect of orifice valve and double inlet are
explored as well.
PHYSICAL MODEL AND GOVERNING EQUATIONS
A double inlet Stirling type pulse tube cooler contains several components such as, a piston
which generates pressure oscillations, the stainless steel regenerator which acts as a porous
medium, the tube through which the working fluid flows freely, heat exchangers which
make desirable thermodynamic contact with the gas, an orifice, one bypass line and a
reservoir. Figure 1 shows a schematic view of a single-stage double-inlet Stirling-type
pulse tube refrigerator.

Figure 1: Shematic view of double inlet pulse tube refrigerator

The operating process of the pulse tube refrigerator is complicated due to the nature of
unsteady, oscillating and compressible gas flow. To trace the process, we treat it as a one
dimensional, periodic, unsteady compressible flow. Following assumptions are introduced
in our model as well:
1.

The working fluid is an ideal gas

2.

Physical properties are functions of temperature

3.

The outer surface of the regenerator and tube wall are adiabatic

4.

Thermal behavior of the regenerator and heat exchangers are ideal

The hydrodynamic and thermal behavior of the cryocooler is predicted by classical

thermodynamic model. In this respect the mass and energy balance equations are applied to
six control volumes. Temperature and pressure are assumed to be spatially averaged in the
control volumes except in the tube section and regenerator. Nodal analysis technique is
employed to simulate the tube section behavior numerically. For the regenerator a linear
trend is acquired for the pressure and temperature distribution along the regenerator. The
implicit control volume method is used to perform the spatial discretization of conservation
of mass and energy in tube section. The nodal analysis is used to divide the whole tube
section into a series of control volumes. The tube section is divided into (n) control
volumes in longitudinal direction.
The temperature, pressure, density, mass, and gas properties of each control volume are
accounted in the center of that control volume. The node is supposed to be in the center of

78

each control volume. The second order up-wind is used to perform the spatial discretization
of conservation of mass and energy in the tube section.
Complete system of differential equations for DIPTR is as follows:

dPcp

dt

dt

dV cp

cp

dt

= T cp

Vt

[C

Pcp dt

Pj

V cp

1 dPcp

dT cp

Pt

V ac

cp

(1)

1 dV cp
V t T cp
.
C cp
( Pj Pcp )
V cp dt
T h PcpV cp

(2)

+ C cpT cp

Vt
Th

( Pj Pcp )

( Pj Pcp ) + (C d + C reg )( Pj Pt )

(3)

D reg L reg

Th
T
* V
* V
T c c +
(
ln h 1) T h ( h ) V t
Th
T c 4(T h T c ) T h T c T c

(C reg T c

Vt
Th

( Pt Pj ) + C d T h

Vt
Th

( Pt Pj ) + C O T h

Vt
Th

( Pt Pres )

D reg 2 L reg
Tc
T Pj
(1
ln h )
)
4(
T

T
)
T

T
T

t
h c
h
c
c

*
( 1)Q t T c

T i
t
Twi
t
Pres
t

P T i T i R
. +
( m i m i 1 )
t Pt Pt V i

(4)

C wV wi

= CO

Vt
V res

(5)

T
T
T
T
( k w Awi ( w ,i 1 w ,i ) + Q t + k w Awi ( w ,i +1 w ,i ))
x
x
( Pt Pres )

(6)

(7)

In the above system of equations:

The subscript (i) designates the number of control volumes in the tube section which is
varying from 1 to n.
The temperature at the boundaries is shown by superscript * and is defined as follows:
79

T cp = (T cp T h ) H ( Pcp P1 ) + T h

(8)

T c = (T t (0, t ) T c ) H ( m c ) + T c

(9)

*
T h = (T t ( L , t ) T h ) H ( m h ) + T h

(10)

The hydraulic conductance of the regenerator

C reg

is obtained from the Kozeny law

established for porous media [6].

2

C reg =

D reg K
4 L reg

Preg

(11)

RT reg

where, permeability K, is expressed in terms of the pressure drop coefficient

Reynolds number
K =

2 D h

Re h C f

Reh

Cf

and the

[6, 7].

where C f =

h
a
mD
+ b and Re h =
c
Re h
g A pass

(12)

The value of the coefficients a, b and c for different situations can be found in [8]. We have
chosen the empirical correlation for the oscillating flow which was proposed by Tanaka [9].
Heat transfer between the gas and tube wall in tube section is obtained from the Nusselt
number equation under oscillating flow, in complex form as described by Kornhouser [10].
RESULTS AND DISCUSSION
The geometry and operating parameters of the cryocooler, which has been considered in the
present paper, are presented in Table 1.
compressor swept volume(m3)
3

compressor dead volume(m )

Frequency (Hz)

1.75E-4
1.5E-5
50

average pressure (bar)

20
3

tube wall thickness (m)

1.00 E-3
3

hot heat exchanger volume(m )

3

cold heat exchanger volume(m )

3

reservoir volume (m )

1.25E-5
2.5E-5
2.25E-3

after cooler volume(m )

2.00E-5

cold end temperature (K)

80

regenerator diameter (m)

0.1

hot end temperature (K)

300

regenerator length (m)

0.08

compressor conductance

10

tube diameter (m)

0.04

orifice conductance

1.5

tube length (m)

0.2

double inlet conductance

Table 1: Geometery and operating parameters

Figure 2 shows the comparison of pressures along the pulse tube refrigerator. The
amplitude of pressure decreases as the flow passes through the cryocooler toward the

80

reservoir. The pressures are comparable in amplitude with an obvious phase shift, as the
above figure displays.

Figure 2: Comparison of pressures at different locations of the cryocooler

In figure 3 the instantaneous temperature along the alternate tube section has been
represented. In this figure the temperature has been plotted at different cycle times in
comparison with the cycle averaged temperature along the tube section.

Figure 3: Instantaneous temperature along the tube

Figure 4 shows the influence of the cryocooler orifice conductance on cooling capacity.
The effect of reservoir volume has been reported in this figure as well.
As expected optimize values are observed on the cooling capacity of the cryocooler, which
depend on the reservoir volume. Cooling capacity of the cryocooler increases almost 10%
by increasing the relative reservoir volume from 9 to 18. No significant increase is
observed in the cooling capacity when the relative reservoir increases more. The significant
effect of orifice conductance is noticeable in this figure. The trade off between the phase
shift angle, between pressure and velocity at the cold end and pressure amplitude in the
tube section results is as an optimum as figure 4 displays.

81

Figure 4: Cooling capacity of the cryocooler vs. orifice conductance

Figure 5 represents the variation of cooling capacity vs. frequency of oscillations at

different average pressures of the cryocooler. A significant increase is observed in cooling
capacity as the average pressure of the cryocooler increases from 10 to 25 bar. This figure
shows the influence of the frequency of oscillations on the cryocooler capacity as well.

Figure 5: Effect of average pressure and frequency on cyocooler capacity

Figure 6 shows the effect of double inlet valve conductance on cryocooler coefficient of
performance for different magnitudes of alternate tube aspect ratios. By increasing the
double inlet valve conductance cooling capacity of the cryocooler decreases due to the
decrease of the mass flow rate at the cryocooler cold head. However, by increasing the
double inlet conductance, the rate of work transfer to the working fluid decreases as well.
Thus, an optimum is observed on cryocooler coefficient of performance due to the trade off
between the cooling capacity and transferred work to the gas.
To verify the results of the present model, the cooling capacity and the transferred power to
the working fluid have been compared with experiment at the same operating parameters
and geometry [11]. The computational model predicts 126 W cooling capacity at 80 K cold

82

Figure 6: Effect of double inlet valve and alternate tube section aspect ratio on
cryocooler coefficient of performance

end temperature with 1.8 kW net power transfer to the gas. Results show that 2.4%
deviation is found between the cooling capacity, predicted by the model and that of the
experiment. The experimental model reports the cooling capacity of 123 W at 80 K with
compressor power of 3.1 kW. The reasons of the difference between the transferred power
to the gas predicted by our model and the experiment include the compressor efficiency and
lost work due to the entropy generation, which have not been considered in the present
study.
CONCLUSION
The hydrodynamic and thermal behavior of a high capacity Stirling PTR was predicted by
control volume analysis model. Nodal analysis technique was employed to simulate the
tube section behavior numerically. The influences of the tube section geometry and
operating parameters on the cooling capacity and coefficient of performance have been
studied. In order to optimize the performance, the effect of orifice valve and double inlet
has been explored as well. The cooling capacity of 250 W at 80 K with 3.9 kW net power
transfer to the gas has been predicted by the computational model. Results have been
compared with experiments and a good agreement was observed.
NOMENCLATURE
A (m )

area

Subscripts &Superscripts

C (J / kgK )

specific heat capacity

ac

after cooler

Cf

inertia coefficient

cold end

Cc

compressor conductance

chx

cold heat exchanger

Co

orifice conductance

cp

Compressor

Cd

double inlet conductance

Gas

D (m )

diameter

hot end

h (W / m 2K )

heat transfer coefficient

Hydraulic

83

Heaviside function

hhx

hot heat exchanger

K (W / m .K )

thermal conductivity

node number

K (m 2 )

permeability

double inlet junction

L (m )

length

Tube

m (kg / s )

mass flow rate

constant pressure

number of nodes

Wall

Nu

Nusselt number

reg

Regenerator

P (bar )

pressure

pass

Passage

Pe

Pecklet number

boundary value

r (m )

radius

Greeks

R (J / kgK )

gas constant

Porosity

Re

Reynolds number

(kg / m .s )

dynamic velocity

t (s )

time

(1/ s )

angular velocity

T (K )

temperature

W (w )

rate of work transfer

V (m 3 )

volume

x (m )

element length

REFERENCES
[1] Popescu G., Radcenco V., Gragalian E. and Ramany Bala P., A Critical Review of Pulse Tube
Cryogenerator Research, Int. J. Refrig. Vol. 24, 2001, 230-237.
[2] Mikulin EI., Tarasov AA. and Shkrebyonok MP., Low Temperature Expansion Tubes, Adv.
Cryog. Eng., 1984, 629-637
[3] Zhu SW., Wu PY. and Chen ZQ., A Single Stage Double Inlet Refrigerator Capable of Reaching
42 K, ICEC 13 Proc. Cryogenics, 1990, 257-261.
[4] Cai JH., Wang JJ., Zhu WX. and Zhou Y., Experimental Analysis of Double Inlet Principle in
Pulse Tube Refrigerator, Cryogenics, 1993 522-525.
[5] Xu M. Y., A. de Waele T. A. M., and Ju Y. L., A Pulse Tube Refrigerator Below 2 K, Cryogenics
39(6), 1999, 865-869.
[6] Nika Ph., and Bailly Y., Comparison of Two Models of a Double inlet Miniature Pulse Tube
Refrigerator: Part B, Electrical Analogy, Cryogenics (42), 2002, 593-603.
[7] Nika Ph., and Bailly Y., Comparison of Two Models of a Double inlet Miniature Pulse Tube
Refrigerator: Part A, Thermodynamics, Cryogenics (42), 2002, 605-615.
[8] Nika Ph., Bailly Y., Jeanot J.C. and Labachelerie M. D., An Integrated Pulse Tube Refrigeration
with Micro Exchangers: Design and experiment, International Journal of Thermal Science, Vol. 42,
2003, 1029-1045.
[9] Tanaka M., Yamashita I., Chisaka F., Flow and Heat Transfer Characteristics of the Stirling
Engine Regenerator in an Oscillating Flow, JSME Internat. J. Ser. (2) 33 (2), 1990, 283289.
[10] Kornhauser A.A., and Smith J.L.Jr., Application of a Complex Nusselt Number to Heat Transfer
during Compression and Expansion., Journal of Heat Transfer 116, 1994, 536-542.
[11] Imura J., Shinoki S., Sato T., Iwata N., Yamamoto H., Yasohama K., Ohashi Y., Nomachi H.,
Okumura N., Nagaya S., Tamada T., Hirano N., Development of High Capacity Stirling Type Pulse
Tube Cryocooler, Physica C 463465, 2007, 13691371.

84

CR08-58
GREEN CRYOGENICS: THE USE OF NATURAL CONVECTION
TO IMPROVE THE EFFICIENCY OF CRYOGENS AND
CRYOCOOLERS
Scurlock R.G.1, Wang C.2
1

Kryos Technology, 22 Brookvale Road, Southampton, SO17 1QP, UK.

2
Cryomech. Inc., 113 Falso Drive, Syracuse, NY 13211, USA.
ABSTRACT

Cryogenics and Cryogenic Engineering are major users of energy, by their nature in
providing or using refrigeration or cold at low temperatures.
The magnitude of this energy is inversely proportional to the lowest temperature and is far
larger than the enthalpy absorbed by the cryogen produced, or by a cryocooler/condenser.
Economy and green practice is becoming increasingly important in Cryogenics to:
1. reduce energy related emissions,
2. reduce cost of making cryogens, and the cost of losses in storage and use of them,
3. reduce losses in handling and transfer.
Natural convection can be used as a major factor for reducing radiative, conductive, and
convective heat in-leaks, in the design of low loss dewars, and in minimising transfer
losses.
Cryocoolers are becoming increasingly important as cryogen-free replacements for cooling
large industrial, laboratory and medical systems. In many cases the cryocooler is used as a
condenser of boil-off vapour in, for example, LHe cooled superconductivity magnet
systems.
Recent work by Chao Wang, Cryomech Inc. Syracuse, NY , USA, has revealed how the
helium liquefaction performance of cryocooler/ condensers can be significantly improved.
Simple modifications to the cold head of a Cryomech PT410 pulse tube
cryocooler/condenser has led to a 67% increase in helium liquefaction rate, at the same
compressor power. These modifications are the first experimental results from using natural
convection to enhance the precooling of the helium boil-off vapour prior to condensation.
Further improvements may be possible.
Additional attention to matching cryocooler head to the dewar neck geometry may yield
even further beneficial results.
INTRODUCTION
In the current thinking on climate change, it is important to note that Cryogenics is a major
user of energy, in providing and using refrigeration, or cold, at low temperatures. Even
with a Carnot efficiency of 100%, the magnitude of this energy consumption is
considerable, being inversely proportional to the lowest temperature; and is far larger than
the enthalpy absorbed by the cryogen produced, or by a cryocooler/condenser. We therefore
need to develop a green attitude, and examine ways of reducing energy consumption and
losses to reduce carbon footprints.

85

This paper is, perhaps, a beginning, but it does illustrate how the use of natural convection
can offer some green credentials leading to improved efficiency.
The starting point is the fact that, with decreasing temperature, natural convection heat
transfer coefficients increase by 10 100 fold in the vapour near the liquid boiling point. In
addition, the local temperature gradient, or stratification, in the vapour can give rise to a
further enhancement of the heat transfer by a factor of 10 or more, in, for example, helium
vapour at 10K. Together, these heat transfer enhancements approach 1000 times the
ambient convection heat transfer coefficients, so that, for example, helium vapour at 10K
can transfer heat to, or from, solid surfaces as effectively as the subcooled liquid at 4K.
These very large enhancements can be used advantageously in several ways.
This paper reports on two sequential developments, namely
(a) low loss dewars and
(b) cryocooler/condensers.
1. NATURAL CONVECTION IN THE DESIGN OF LOW LOSS DEWARS
The boil-off vapour from a liquid bath does not flow uniformly up the neck. Some 20 years
work at Southampton [1] and elsewhere, has demonstrated that
(a) the vapour flow is confined to a thin boundary layer flow at the neck wall,
(b) heat flux from the wall drives the convection flow up the wall,
(c) the magnitude of this upward flow is many times larger than the boil-off mass flow and
forms part of a thermosyphon recirculation,
(d) the downward flow part of the thermosyphon occurs at the centre of the neck.
Studies have shown that some, or all, heat fluxes down the neck, whether radiative,
convective or conductive, can be absorbed by the cold vapour rather than by the liquid.
This simple concept can be applied directly to the design of low loss dewars, containers and
tanks.
2. CRYOCOOLERS AS CONDENSERS
Placing a cryocooler cold head in the neck of a dewar can achieve two bonuses at the same
time.
(a) The cooling produced by the cold head, of the central vapour core of the thermosyphon
flow, will increase the mass flow of the recirculation and hence the upward mass flow at
the neck wall. The heat influxes down the neck will be further absorbed, thereby yielding a
significant reduction in the boil-off rate.
(b) The distributed cooling of the downward flow over the cold head can be used as a
simple precooling function to increase the rate of condensation. This bonus in condensation
rate cannot be achieved if the cryocooler head is mounted in a vacuum with spot cooling
heat exchangers attached to the dewar vapour vent line.

86

3. ENHANCEMENT OF CONDENSATION RATE

A number of convection features need to be considered to maximise condensation.
3.1 Central position in cryostat neck
Since the thermo-syphon flow is cylindrically symmetrical, the cryocooler head should be
placed on the axis in the centre of the down flow. No part of the cold head should project
into the upward boundary layer flow at the neck wall.
3.2 Use of regenerator/recuperator for distributed cooling
While the regenerator is primarily for exchanging heat cyclically between warm and cold
flows, the mean temperature of the outer wall has a non-linear temperature gradient
between upper and lower ends. Thus the regenerator (or recuperator) wall can be used,
accidentally or deliberately, as a continuous heat exchange surface for cooling the
downward boundary layer flow of condensable gas.
Theoretical analysis and experimental tests by Wang in 2001, [2] and 2005, [3] and those
by Zhu et al [4] and Ravex et al [5], have indicated that significant cooling effects are
available.
One limit on the cooling effect may be the plain surface area of the regenerator. It is
therefore postulated that improved heat transfer may be obtained by increasing the heat
transfer area with a set of low thermal conductivity vertical stainless steel fins, or high
thermal conductivity horizontal copper pins or rings, perhaps 1-2 mm high, along the
vertical length of the regenerator (or recuperator).
3.3 Use of pulse/expander tube for distributed cooling
The primary source of cooling is provided by expanded gas at the lower end of the pulse
tube/expander tube. After each expansion, the cold gas passes back via the cold end heat
exchanger into the regenerator. It is interesting to consider whether some of the cold in
the expanded gas can also be taken from the surface of the pulse tube or expander tube by
natural convection; this would further increase the usable cold available for precooling.
Since the pulse tube/expander tube also has a non-linear temperature gradient between top
and bottom, the cooling effect would be distributed, and again could be used to provide
significant precooling prior to condensation.
Intermediate precooling from the pulse tube wall is suggested by Wang [6].
3.4 Integration with spot cooling condensing performance
Clearly, a balance must be made between the distributed cold taken from the
regenerator/recuperator and pulse tube/expander tubes, and the consequent reduction in
spot cooling power of the cold head. Caution is therefore needed in not modifying the tubes
with too much additional heat transfer surface by adding too many vertical fins, or
horizontal pins or rings.
Reducing convective heat transfer can be achieved by applying layers of MLI
superinsulation to the tube surfaces. This reduction is demonstrated by the difference in
liquefaction rates between Tests 2 and 1 described below.
3.5 Radiation cooling
As mentioned earlier, the cryocooler/condenser increases the mass flow in the vapour
thermosyphon loop. As a result, thermally floating horizontal radiation shields, (no colder
than 70K) and not thermally attached to the cryocooler head, will be adequately cooled by
the vapour flow so as to absorb all 300K radiation down the cryostat neck. Alternatively, if
the radiation shields are thermally attached to the bottom of the first stage, then they may
87

also assist the precooling of the downward boundary layer flow of condensable gas at the
higher temperatures, as well as absorbing 300K radiation down the neck. These alternatives
will be tested later on.
3.6 Matching cryocooler and cryostat neck geometries
For maximum precooling, the overall temperature gradients in the cryocooler cold head and
cryostat neck should be closely similar. It follows that the length of the cold head should
match the length of the cryostat neck, or be slightly shorter.
4. PRELIMINARY EXPERIMENTAL RESULTS
So far, 4 preliminary test results have been obtained [7], using modifications to the
Cryomech PT410 helium liquefier/cryocooler described in Reference [3], as follows:
Test 1. Result achieved with modifications to cold head described in Reference [3], in 60L
dewar with 180mm diameter neck.
Test 2. As Test 1 but with removal of G10 sleeve and superinsulation around cold head,
also condensing coil, as shown schematically in Figure 1.

Figure 1. Helium liquefaction with pulse tube cold head inside dewar neck
1. dewar neck; 2. portion of down stream flow of helium gas to be liquefied;
3. a thermosyphon loop; 4. pulse head cold head; 5. 1st stage cooling station; 6. 2nd stage
cooling station/condenser.
Test 3. Reduction of diameter of dewar neck from 180 mm to 150 mm in new 60L dewar,
and addition of horizontal copper fins to 2nd stage pulse tube and regenerator.

88

Test 4. As Test 3 but with addition of horizontal radiation shields/fins at the lower end of
the 1st stage regenerator, as shown schematically in Figure 2.

Figure 2. Latest helium liquefaction system with pulse tube cryocooler.

1. helium dewar; 2. dewar neck; 3. fins on the 2nd stage regenerator; 4. 1st stage cooling
station; 5. radiation shields/fins on the 1st stage regenerator; 6. pulse tube cold head;
7. flexible lines; 8. fins on the 2nd stage pulse tube; 9. 2nd stage cooling/condenser;
10. compressor.
The performance improvements, in terms of liquefaction rates per day, are listed in Table 1.

Liquefaction rate

Test 1
12.8 L/day

Test 2
15.0 L/day

Test 3
19.5 L/day

Test 4
21.4 L/day

Table 1: Test results

It can be seen how the modifications led to significant improvements in the liquefaction
rate, from 12.8 L/day to 21.4 L/day, an increase of 67 %.
It should be added that a newly made cold head with copper fins on the 2 nd stage
regenerator and pulse tube was used for Tests 3 and 4.
Further tests should enable more improvements in condensing performance to be achieved
using natural convective heat transfer, towards the theoretical limit, for example, of
34 L/day for a 1W/4.2K cryocooler.

89

CONCLUSIONS
Simple modifications are possible to improve the low loss boil-off performance and the
condensing performance of cryocoolers, using the natural convection in the neck of a
cryostat.
Natural convective heat transfer coefficients are very high in dewar necks and can be used
advantageously both for cooling the dewar neck and for extracting precooling from both the
pulse tube and regenerator, without major modification to the cryocooler head.
Clearly, there remains a great deal of optimisation to attain the maximum condensation rate
with a given cryocooler.
REFERENCES
1. Scurlock R.G., Low Loss Storage and Handling of Cryogenic Liquids: The Application
of Cryogenic Fluid Dynamics. Kryos Publications, 2006.
2. Wang C., Helium liquefaction with a 4K pulse tube cryocooler. Cryogenics 2001, 41,
491-496.
3. Wang C., Efficient helium recondensing using a 4K pulse tube cryocooler. Cryogenics
2005, 45, 719-724.
4. Zhu S.W., et al., 4K pulse tube refrigerator and excess cooling power. Adv. Cryo. Eng.
2002, 47A, 633-640.
5. Ravex A., Trollier T., Tanchon J., and Prouv T., Increase in Performance of 4K pulse
tube cryocooler. Proc.CryoPrague 2006/ICEC 21, 2006, 515-524.
6. Wang C., Intermediate cooling from regenerator and pulse tube in a 4K pulse tube
cryocooler. Cryogenics, 2008, 48, to be published.
7. Wang C., and Scurlock R.G., Improvement in performance of cryocoolers as condensers.
Cryogenics, 2008, 48, accepted for publication.

90

CR08-12
THE DEVELOPMENT OF A VUILLEUMIER CRYOCOOLER FOR
NEW ZEALANDS HIGH TEMPERATURE SUPERCONDUCTOR
INDUSTRY
Gschwendtner M.A., Tucker A.S.
University of Canterbury, New Zealand
ABSTRACT
A Vuilleumier cryocooler with a projected cooling capacity of 100 W at 77 K has been
developed at the University of Canterbury in New Zealand. While most cryocoolers use
mechanical compressors with associated vibration, leakage and lubrication problems, the
Vuilleumier concept employs a thermal compressor with only a marginal mechanical work
input requirement. This paper highlights its significant design advantages and presents
results from a computer model that justifies this approach, despite the apparent
disadvantage of the Carnot limitation arising twice.
INTRODUCTION
New Zealand has invested in high temperature superconductor technology (HTS) and is
now in a significant position globally. Various local companies have already spun off this
technology such as HTS-110 with laboratory magnets and Canterbury TX with
transformers. It is anticipated that more companies will follow. The sheer number of
possible applications of HTS technology necessitates the development of affordable,
reliable and mobile cooling systems. This project originated from the need to provide an
already existing and fast growing industry with a locally developed cooling system that can
be specifically adapted and produced at lower cost.
1. BACKGROUND
1.1 Theoretical background
Any cold-producing devices such as Gifford-McMahon, Stirling or pulse-tube cryocoolers
require a variation in the pressure of the working gas in order to achieve expansion and
compression. This is usually done by a mechanical compressor in the form of a piston that
reciprocates in a cylinder. However, a reliable seal design is required in order to avoid gas
leakage past the piston and, unless precise and expensive clearance seals are employed,
lubrication is involved. Lubrication, on the other hand, does not go well with machines that
incorporate a regenerator such as Stirling and pulse-tube refrigerators as there is the risk of
contaminating the fine pores of the regenerator matrix with the lubricant and thus
significantly reduce the performance of those machines. Vibration and noise are further
potential problems of mechanical compressors [1, 2, 3].
An even greater issue arises from the current demand of higher cooling capacities. The
scaling-up of existing equipment to meet the required higher cooling capacities by simply
increasing the physical size of these machines is often very difficult and typically
disproportionately more expensive. One of the limiting factors in this exercise is the
mechanical compressor that is usually driven by linear motors. It is common knowledge
that the cost of linear motors increase disproportionately with their size and they become
exorbitantly expensive. In addition to that, large mechanical compressors tend to be very

91

noisy and heavy quite often a decisive criterion in some cryocooler applications. A higher
moving mass of a larger mechanical compressor also leads to more severe balancing issues.
A promising alternative is to use a thermal compressor that creates a pressure variation by
a change in the temperature of the working gas instead of a change in volume as in a
reciprocating piston machine. This principle is based on Rudolph Vuilleumiers invention
from 1917 [4] and basically consists of a Stirling heat engine that is coupled to a Stirling
refrigerator while sharing the same working gas. While it can be argued that a considerable
disadvantage of using a heat engine as a compressor lies in the fact that the Carnot penalty
has to be paid twice, significant design advantages may still justify the implementation of
the Vuilleumier concept. This is the case even more so if a design is chosen that promises
to be as nearly as efficient as a mechanically driven as will be shown below.
1.2 Working principle
A Vuilleumier refrigerator operates between three temperature levels hot, ambient and
cold. Its working principle can be best understood by first looking at how a mechanical
compressor-type Stirling refrigerator works. Figure 1 shows a gamma-type Stirling
configuration.

Figure 1: Gamma-type Stirling refrigerator.

A power piston reciprocates in a cylinder and creates a variation in pressure by changing

the volume of the working gas. Separate, but connected to the compression and expansion
cylinder by a port, is the refrigerator part. Here a displacer shifts the gas back and forth
between two heat exchangers a heat rejector at ambient temperature and the absorber at
the cold end temperature. The motion of the displacer is out of phase by 90 with the power
piston, such that the gas in the refrigerator is exposed to the cold end during the expansion
phase and is exposed to the heat rejector during the compression phase. In between the two
heat exchangers the regenerator is located, serving two main purposes: Firstly, it acts as a
thermal barrier between the two temperature levels in order to avoid short-circuiting of
thermal energy. Secondly, due to the porous structure of its matrix through which the
working gas oscillates, the regenerator acts as a temporary heat store for the gas which
picks up heat when moving away from the cold end and rejecting heat when moving away
from the heat rejector heat exchanger. A Vuilleumier refrigerator has a similar working

92

principle; however, the mechanical compressor on the right hand side is replaced by a heat
engine while the remaining components in the refrigerator are identical (Figure 2).

Figure 2: Vuilleumier refrigerator.

Here, the pressure variations are created by a change in temperature of the working gas. An
additional displacer moves the gas between two heat exchangers and alternately exposes
most of the gas to a hot temperature and to a cold temperature. Since the overall volume of
the system remains constant at all times the gas pressure varies with the temperature
according to the Equation of State. Again, compressor and refrigerator parts are connected
via a port. Figure 3 shows a comparison between the Carnot Coefficients of Performance
(COP) for a Vuilleumier machine and a Stirling refrigerator at given cold end temperatures.
Ambient temperature is assumed to be 300 K whereas the hot temperature of the heat
engine in the Vuilleumier configuration is 1000 K.
0.45
Stirling (100% efficient)

0.40

Stirling (75% efficient)

0.35

Stirling (65% efficient)

COP [-]

0.30

Vuilleumier

0.25
0.20
0.15
0.10
0.05
0.00
0

10

20

30
40
50
60
70
Cold space temperature [K]

80

90

100

Figure 3: Comparison of the Carnot-Coefficient of Performance (COP) for a Vuilleumier and a

Stirling refrigerator with three different mechanical efficiencies.

93

The diagram illustrates two things: the Carnot-COP of a Stirling is superior to a

Vuilleumier where the Carnot efficiency of the thermal compressor has to be included as
well, but the situation looks different if the efficiency of the Stirlings mechanical
compressor is also taken into account. The diagram shows two efficiencies (75% and 65%)
that straddle the COP-curve of the Vuilleumier and the assumed temperatures. The point to
note is that mechanical compressors also have a limited efficiency and, as long as
reasonably high temperatures are realised in the thermal compressor of a Vuilleumier
machine, the COPs of both configurations are in the same range. However, there are
significant design advantages for the Vuilleumier configuration as will be shown below.
2. Design considerations
The fact that the overall volume of the working gas remains constant in a Vuilleumier
machine means that sliding seals which have to maintain a large pressure differential
between the inside and the outside of the system can be abandoned. Instead, the gas
envelope can be hermetically sealed by either static seals or, if practical, even be a fully
welded enclosure.
A further advantage is the fact that no pistons are required to operate between a large
pressure difference (and small T). Instead it is possible to use displacers which do not
have to compress the gas, rather, move it from one side to the other, only having to
overcome gas friction (small p, large T). This comparatively small work requirement
means that smaller and cheaper linear motors are sufficient to drive the displacers. Since
displacers have only a small pressure difference across them, seal problems are much less
severe and the manufacturing tolerances are less tight which reduces the cost to a great
extent.
The absence of problematic sliding seals and the ease with which the whole system can be
sealed results in another advantage for thermal compressors: the average pressure of the
working gas can be increased without the requirement for stronger mechanical
compressors. Since the cooling performance is an almost linear function of the average
system pressure, the machines can either be built smaller or, alternatively, the cycle
frequency can be reduced. This last possibility is beneficial for the efficiency of the cycle as
more time is available for heat transfer processes and fluid friction losses can be reduced.
Lubrication with all its associated issues can be avoided by the use of flexure bearings
which are common practice in todays cryocooler technology. Flexure bearings are circular
discs made of steel with slots in the form of a spiral. This allows the discs to flex normal to
their faces while, at the same time, providing a high stiffness radially. A shaft that is
supported by stacks of these discs is more or less centrally located and constrained, but is
able to reciprocate along its axis. If the flexure bearings are designed such that occurring
stresses remain well below the fatigue limit these components are not only maintenancefree but also exhibit a long lifetime.
Furthermore, the choice of flexure bearings and the need for displacers only instead of
pistons facilitates the use of clearance seals. These are basically tight-fit displacers in
cylinder liners that provide a gas seal through a very small clearance. Since displacers have
to seal against a relatively small pressure difference across them, the machining tolerances
are not as tight as in the case of pistons operating between a large pressure difference. This
way of sealing makes redundant the use of sliding seals that are prone to wear and also
eliminates the risk of seal material debris floating around in the gas cycle. Again, this
contact-free gas seal in conjunction with lubricant-free flexure bearings are a very good
match that promises reliability and longevity.
94

Using heat as the main energy input allows a high energy transfer rate per unit volume and
per unit mass which is of particular importance for the demand of higher cooling capacities.
Heat can be applied to the system without any noise or vibrations and the fact that different
heat sources can be used depending on the requirements or the environment of the
application may be an additional significant benefit. While, at first sight, it may seem
irresponsible from a thermodynamic viewpoint to use electric heating, the situation looks
different if the efficiency of mechanical compressors is taken into account (see diagram in
Figure 3). Electric heating is very convenient and readily available and, after all,
mechanical compressors also use electricity.
Thanks to recent design developments such as flexure bearings and linear motors, the
Vuilleumier concept is now more attractive than it may have been several years ago. Also,
advances in materials have pushed the boundaries in the use of light-weight components
and high temperature insulation materials. Finally, and not least, a new territory has been
opened up by digital electronic controls of linear motors. Instead of being constrained by a
kinematic drive system, the position-versus-time relationships for the displacers can now be
accommodated such that their motion follows more naturally the gas cycle in Stirling
machines.
3. MODELLING
3.1 Background
For modelling of the gas cycle in a Vuilleumier cryocooler it is not necessary to represent
the above mentioned design aspects in great detail. For the analysis and optimisation David
Gedeons Stirling simulation software package SAGE was used [5]. This allows the user to
plug various modules together and connect them through relevant variables. For instance,
the face of a piston is connected to the adjacent gas space through a volume variable. Or, a
heat exchanger is linked to a connecting tube with a mass flow variable, whereas the heat
exchanger wall is in thermal contact with a heat sink or source through a heat flow variable.
All fundamental equations are solved for each module separately until the program
converges to a final solution. SAGE has a mapping function which allows the specification
of various parameters that are to be varied in given increments, as well as an optimisation
function. All modules/components in SAGE may have two or more sub-levels where
further input variables can be specified. For example, one sub-level of a heat exchanger
could be the geometry of internal fins. Depending on how detailed the user wishes to model
a system, SAGE takes heat conduction paths along solid and gaseous components into
account, calculates fluid friction and pressure drops, and even shuttle and seal losses in
piston/cylinder arrangements can be modelled.
Figure 4 shows how above mentioned Vuilleumier cryocooler configuration was modelled
in SAGE. Both the refrigerator part in the lower half and the thermal compressor in the top
half comprise the same components. Both double-acting displacers are neighboured by their
adjacent gas spaces to the left and the right in the diagram, followed by connecting ports to
the respective heat exchangers (upwards in the diagram). Both the refrigerator regenerator
and the heat engine regenerator are sandwiched by the adjacent heat exchangers. The
connecting duct module at the far left at the bottom of the diagram connects the gas
spaces of the refrigerator and the thermal compressor. Finally, the three temperature
symbols at the bottom of the diagram represent the three temperature levels to which a
number of heat flow variables are connected to.

95

Figure 4: SAGE-model of a Vuilleumier cryocooler configuration.

3.2 Procedure of analysis

From the approximately 150 input variables 25 could be indentified as most critical. In
order to find an optimum configuration, however, if one assumed a variation of each
variable in three increments and a computing time of 10 seconds for each solution, it would
take some 250,000 years to test all combinations. The dreaded search for the elusive needle
in the haystack turned out to be relatively simple, though. It was found that if each
component was optimised separately , one at a time, then repeated after all components had
undergone this process, the optimum performance of the system had almost converged after
one loop only.
The optimisation of each component was achieved through mapping by which typically
three, four or five critical parameters were varied in up to ten increments each. SAGE
records the user-specified output values for each combination which can then be displayed
graphically in a spreadsheet. The optimum combination is easily found by sight, although a
compromise between efficiency (COP) and cooling capacity had to be sought almost all the
time.
Figure 5 shows a typical result of a mapping process for the regenerator. The achieved
cooling capacity of the Vuilleumier cryocooler and its percentage of Carnot efficiency are
plotted over the length of the regenerator for various outer housing diameters. For design
constraints the inner diameter of the concentric tubular shape of the regenerator housing
was held constant. Also the wire diameter and the porosity of the regenerator matrix were
fixed after a material had been chosen. The arrows indicate the increase of the regenerators
outer diameter for both the cooling capacity plots and the Carnot-efficiency plots. Here it
was relatively straight-forward to pick the smallest possible outside diameter of the
regenerator since an increase in diameter had a detrimental effect on both the cooling
capacity and the Carnot efficiency. The choice of the regenerators length, however, was
less clear as an increased length resulted in an ambiguous performance. A compromise had
96

to be found where the Carnot efficiency was reasonably high at a still-acceptable cooling
performance. Quite often the selection process was influenced by design considerations
and/or constraints.
100

30

increasing outer diameter

90

25

70

20

% Carnot [-]

Cooling capacity [W]

80

60
50

15

40

10

30

increasing outer diameter

20

Cooling capacity

10

%Carnot

0.01

0.03

0.05

0.07

0.09

0.11

0.13

0.15

Regenerator length [m]

Figure 5: Mapping result of regenerator analysis.

4. Analytical results
Once all components were optimised and the performance of the chosen configuration
could not be increased further, a performance chart could be plotted (Figure 6). This was
obtained in varying the cold space temperature in the SAGE model and calculating the
respective cooling capacity and Carnot efficiency. It can be seen that the cooling capacity is
an almost linear function of the cold space temperature, with a value of somewhat more
than 100 W at the design point of 77 K, reaching down to around 30 K with almost zero
cooling capacity.
120

60
Cooling capacity

50

%Carnot
80

40

60

30

40

20

20

10

% Carnot [-]

Cooling capacity [W]

100

0
0

20

40

60

80

100

Cold space temperature [K]

Figure 6: Predicted performance chart of a Vuilleumier cryocooler configuration by SAGE.

97

The predicted Carnot efficiency also drops with a decreasing cold space temperature but in
a non-linear fashion. At the design point the Carnot efficiency of 25% is almost twice as
high as commercially available pulse-tube or Gifford-McMahon cryocoolers with similar
cooling capacities. It should be noted, however, that this is a computer model only that is
partially based on ideal assumptions despite the fact that it takes a number of realityeffects into account such as heat conduction paths, shuttle losses and irreversible heat
transfer processes.
CONCLUSION
In view of an increasing demand of higher cooling capacities in todays cryocooler
applications an alternative to mechanical compressor-type coolers is proposed. The use of
thermal compressors as in Vuilleumier refrigerators offers a variety of design advantages
which may result in quieter and more reliable operation, lower manufacturing cost and, as a
computer model predicts, at least comparable efficiencies. The model predicts a cooling
capacity of somewhat more than 100 W at a cold space temperature of 77 K and at a Carnot
efficiency of 25%. The simulation results of the proposed Vuilleumier cryocooler
configuration not only compare well with existing Stirling, Gifford-McMahon and pulsetube cryocoolers, but also seem to justify the use of a thermal compressor as opposed to a
mechanical one. While it should be noted that above presented results are only simulated
and reality effects will take their toll, the outlook is very promising. The Stirling Group at
Canterbury University has gained some confidence in SAGEs predictions from earlier
modelling and subsequent experimental performance testing of non-cryogenic Stirling
refrigerators [6, 7]. A prototype of the described Vuilleumier cryocooler is being
manufactured at the University of Canterbury according to the results of the optimisation
procedure with SAGE. Extensive testing is planned for the first half of 2008 with the
implementation of smaller modifications in the second half of this year. Results will be
published elsewhere.
REFERENCES
1. Walker, G., Cryocoolers, Part 1: Fundamentals, Plenum Press, New York and London
(1983)
2. Walker, G., Cryocoolers, Part 2: Applications, Plenum Press, New York and London
(1983)
3. Walker, G. and Bingham, E. R., Low-Capacity Cryogenic Refrigeration, Clarendon
Press, Oxford (1994)
4. Vuilleumier, R., Method and Apparatus for Inducing Heat Changes, U.S. patent
1,275,507 (1918)
5. Gedeon Associates, 16922 South Canaan Road, Athens, OH 45701, U.S.A.
6. Haywood, D., Investigation of Stirling-type Heat-pump and Refrigerator Systems Using
Air as the Refrigerant, Doctoral Thesis, University of Canterbury, Christchurch, New
Zealand (2004)
7. Haywood, D., Raine, J. K., Gschwendtner M. A., Investigation of seal performance in a
4-a double-acting Stirling cycle heat-pump/refrigerator, Proceedings of the 10th
International Stirling Engine Conference, Osnabrck, Germany (2001)
ACKNOWLEDGMENT
The New Zealand Foundation for Research, Science and Technology funded this project
and contributed to its success in a manner which was very supportive but, at the same time,
quite unbureaucratic.

98

CR08-20
THE EFFICIENT MANAGEMENT OF LIQUID HELIUM AT
SOUTH POLE STATION DURING THE AUSTRAL WINTER
Baker R.A., Sullivan P.
Raytheon Polar Services Company, 7400 South Tucson Way,
Centennial, CO 80112, USA
ABSTRACT
Liquid helium (LHe) is critical for the operation of various astrophysical projects at the
South Pole. Getting large quantities of liquid helium to the South Pole is logistically very
difficult and costly. In order to supply these experiments with a combined 40 liters per day
of liquid helium South Pole Station has historically had to begin the austral winter viewing
season with 34,000 liters. Of this amount 34% is supplied to the projects while 66% is
boiled away. By capturing the boiled off gas and using pulse tube technology 65% of the
liquid helium can now be delivered to the projects and only 35% is unrecoverable. Now,
34,000 liters can supply projects with more than 80 liters per day combined total. More
projects that require liquid helium can be funded, or the logistical load on the South Pole
can be reduced. South Pole Station has recently procured three Cryomech PT410 pulse tube
cryo-refrigeration systems custom fitted to three 4,000 liter Wessington liquid helium
storage dewars. These units are tuned to re-liquefy the boil-off inherent in the Wessington
dewars. This technology does not come without cost. It takes power to operate the pulse
tubes and at the South Pole that translates into fuel supplied by flights. The net trade-off
was beneficial to the NSF, RPSC, and science projects.
KEYWORDS: South Pole, liquid helium, pulse tube, cryorefrigerator
INTRODUCTION AND HISTORY
For logistical purposes there are two seasons at Amundsen/Scott South Pole station in
Antarctica, the austral summer and austral winter. The summer season begins around 1
November of a given year when the ambient temperatures warm up enough for aircraft to
safely fly in and out of the station, and ends on 15 February each year when the ambient
temperatures fall below the safe operating temperatures of the aircraft. The station closes
to all incoming and outgoing traffic on this date. A small crew of station operations
personnel and scientists remain behind and work throughout the austral winter. All
logistics take place within the 3 months while the station is open during the austral
summer. All material, scientific equipment, food, fuel, personnel, etc. must be in place
when the station closes on 15 February.
There has been a continuous scientific presence at Amundsen/Scott South Pole Station
since 1957. It was originally a meteorological station but over the years has evolved into a
station that supports cutting edge science projects, some of which can be performed at no
other location on the planet. In 1988 liquid helium was used for the first time to cool
astrophysical receivers that are looking at extremely faint energy signatures from the
Cosmic Microwave Background (CMB) radiation and the origins of the universe. The low
relative humidity (RH) of the atmosphere, no diurnal cycle, and the fact that a telescope can
be pointed at a specific location in the celestial dome for long integration periods makes the
South Pole an ideal place for astrophysical work. Since then liquid helium has been used

99

increasingly for a variety of projects over the years, peaking in 2006 and 2007.
Figure 1.

See

Figure 1 LHe requirements at South Pole have increased yearly to the maximum capacity of the
station. New technology is being incorporated into the design of new telescopes and
cryorefrigeration is replacing cryostats.

There are logistical challenges involved with getting liquid helium to the South Pole and
maintaining a supply there. Large volumes, usually in increments of 13,140 liter Gardner
LHe dewars, must travel by ship or be flown approximately 10,000 miles via commercial
and military aircraft from the continental USA (CONUS), through New Zealand, to the
coast of Antarctica, and on down to the South Pole. This size transport dewar just fits in
the cargo hold of the LC-130 Hercules military cargo airplane which flies from McMurdo
Station on the coast of Antarctica to the South Pole. This logistics chain can only be
accomplished during the supply window of the austral summer of a given year, because
temperatures outside this time window prevent travel of any kind to and from the South
Pole by these aircraft. See Figure 2
When the South Pole closes for the winter the quantity of liquid helium on station must be
sufficient to provide a regular, uninterrupted supply to the astrophysical projects for 270
days. The harsh conditions of the South Pole winter present challenges that make this
difficult. The air is dry (<1% RH), the elevation is high (~2835 meters), the atmospheric
pressure is low (~68.3 kilopascals) and the median ambient temperature in the winter is <65 C. A poor understanding of how these conditions affect instruments and equipment
resulted in some early failures of the South Pole Cryogenics Program. There were seasons
when the winter supply of liquid helium was exhausted or lost before the viewing season
was over. See Figure 3.

100

Figure 2 The logistics chain from CONUS to South Pole.

LHe Tracking
10,000

9,000

8,000

7,000

Liters

6,000

3K
W1
W2
W3

5,000

4,000

3,000

2,000

1,000

liquid empty 11 Sept

11/8/2003

11/1/2003

10/25/2003

10/18/2003

10/4/2003

10/11/2003

9/27/2003

9/20/2003

9/6/2003

9/13/2003

8/30/2003

8/23/2003

8/9/2003

8/16/2003

8/2/2003

7/26/2003

7/19/2003

7/5/2003

7/12/2003

6/28/2003

6/21/2003

6/7/2003

6/14/2003

5/31/2003

5/24/2003

5/17/2003

5/3/2003

5/10/2003

4/26/2003

4/19/2003

4/5/2003

4/12/2003

3/29/2003

3/22/2003

3/8/2003

3/15/2003

3/1/2003

2/22/2003

2/15/2003

Date

Figure 3 Early in the cryogenics history of South pole the station would run out of LHe before the
austral winter observing season was over. 3K refers to the 3,000 gallon transport dewar and W1,
W2, W3 refer to the Wessington dewars.
101

The South Pole Astrophysical Program grew faster than the South Pole Cryogenics
Program. Telescopes operated outside and liquid helium was stored in large transport
dewars in the same inhospitable environment. Seals froze and failed, vacuum insulations
were lost, and large quantities of LHe were rapidly evaporated. The United States
Antarctic Program (USAP) and the NSF quickly learned that heated shelters for key pieces
of equipment had to be built, but the planning and execution of such endeavors is a multiyear process. Ad hoc and temporary structures were built and used
until a comprehensive cryogenics facility could be designed, engineered, and constructed at
the South Pole.
Resources at the South Pole are limited and finite, particularly in terms of power and fuel.
Large helium liquefiers required more power than was available at the South Pole. As a
result the station employed a passive approach to storing and supplying liquid helium to the
astrophysical projects. It stored large volumes, enough to supply the needs of the
experiments plus the natural boil-off inherent in the storage dewars, transfer losses, and
some contingency for equipment failure. This is inefficient and wasteful but historically
less costly for the station than employing an active method of supply, i.e., re-liquefying the
boiled off helium gas.
THE ISSUE
The LHe requirements of the astrophysical projects at the South Pole grew in successive
years. The South Pole had to find a way to supply as much as 60 liters per day total to the
various projects over the 270 days of the austral winter. The supply plan also had to take
into account the amount of helium boiled off from each of the storage dewars, transfer
losses, cryostats, and a contingency amount to mitigate unforeseen problems or catastrophic
failures, usually on the order of 20%. This translated to beginning the austral winter
viewing season with 34,000 liters of liquid helium, the maximum capacity of South Pole
Station. This left no room for further growth.
Passive storage of liquid helium outside is not a viable method for supplying cutting edge
astrophysical projects with cryogens. Supplying liquid helium to the astrophysical projects
at the South Pole is similar in scope and complexity to supplying liquid helium to
experiments in spacecraft, but with a far smaller budget.
THE SOLUTION
A comprehensive South Pole Cryogenics Program was established and a three-pronged
approach was taken. First, a working group, the Liquid Helium Working Group (LHeWG)
consisting of grantees, the National Science Foundation (NSF), and Raytheon Polar
Services Company (RPSC), the support contractor, was established. Considerations of
budget, existing South Pole infrastructure, resource limitations, logistics, and emerging
technology were made.
Already underway were plans to modernize the station in terms of size, infrastructure,
logistics, and resources. The second prong was to include the construction of a facility
specific for housing, tracking, and working with cryogens and equipment in a warm
environment. Third, a search was initiated for technology that would reduce liquid helium
waste from dewar boil-off but not bankrupt the station fuel and power supply.
The South Pole Station Modernization Project
The South Pole Station Modernization (SPSM) project was a multi-year project managed
by the NSF that increased the size of the station, fuel capacity to 450,000 gallons, power
102

output to 800 kilowatts, population, logistics, and generally upgraded the infrastructure.
With these improvements it was possible to look at modern solutions to the South Pole
Cryogenics Program.
The New Cryogenics Facility
SPSM made possible the construction of a new 2900 square foot, three-module cryogenics
facility. All cryogenic equipment, instruments, and storage were moved out of the harsh
South Pole environment into a spacious and heated structure. The station could have some
assurance of the integrity of storing dewars and other cryogenic equipment during the
austral winter. Seals would no longer fail due to the cold and vacuum insulations would
not be compromised.
There is now room for two 13,140 liter Gardner liquid helium transport dewars and three
4,000 liter Wessington liquid helium storage dewars. There is also room for smaller dewars
(250 liter and 100 liter) used to transport liquid helium from the storage facility to the
telescopes. Permanent hard-pipe transfer lines were installed in the facility to make liquid
helium transfers simpler and more efficient with less transfer loss. Space was allocated for
maintenance, monitoring, and tracking. The new cryogenics facility also has the capacity
for housing refrigeration units to reliquefy helium gas. For the first time the South Pole
could explore the possibility of taking an active approach to liquid helium storage.
The first of three modules of the new South Pole cryogenics facility was brought online in
February 2006. This module housed two Gardner 13,140 liter liquid helium transport
dewars for the austral winter. Prior to this construction the large transport dewars wintered
outside. The Wessington liquid helium storage dewars were housed in a temporary heated
structure for the austral winter. In February, 2007 all three modules of the new cryogenics
facility were brought online and all equipment related to cryogenics was for the first time
brought inside under one roof.
Pulse Tube Cryorefrigeration
With the construction of the new cryogenics facility and the increase in overall station
power it was now possible to investigate helium re-liquefaction at the South Pole. Several
approaches were studied.
The facility was designed to house two 13,140 liter Gardner liquid helium transport dewars
and three 4,000 liter Wessington liquid helium storage dewars. The measured static boiloff rate of the large transport dewars was 35 liters per day and the measured static boil-off
rate of the Wessingtons was 15 liters per day for a combined loss of liquid helium of 115
liters per day, the biggest single user on station.
One option was to capture the boiled off helium gas into a bag and run it through a large
liquefaction unit. A unit of sufficient size to liquefy this much helium gas would require in
excess of 90 kilowatts of power to operate. This would be a very large burden to the
overall station power budget.
Another option was to mate pulse tube cryorefrigeration units to the Wessington storage
dewars. In order to match the evaporation rate of 15 liters per day (0.02 grams per second)
for a Wessington dewar the cryorefrigerator would be required to provide an enthalpy
change of 1563 Joules per gram from 300K to 4K and 21 Joules per gram for the latent heat
of condensation.

103

The cooling power required to match this boil-off rate would be:
dm/dt * Hfg + dm/dt * Cp * T

(1)

0.02 * 20 + 0.02 * 5 * 300-4 = 30 Watts

(2)

where:
dm/dt = 0.02 grams per second
Hfg = 20 Joules per gram
Cp = 5 Joules per gram Kelvin
This takes the temperature down to 4K and then the re-condensing would take place.
Approximately 0.4 Joules per second or 30 Watts of heat must be removed from the helium
gas to match the boil off rate of the Wessingtons and it must be done within the power
budget of South Pole Station.
Cryomech, Inc manufactures a 4K pulse tube cryorefrigerator, the PT410 that could be
custom fitted to the neck tube of a Wessington dewar, reliquefy the boiled off gas and drain
the condensate back into the dewar, all at a cost of 8 kilowatts of input power. Another 4
kilowatts of power per unit would be required to cool the associated compressor for a total
of 12 kilowatts per unit or 36 kilowatts for systems fitted to all three Wessingtons. This is a
substantial power savings over larger all-encompassing units and is a reasonable
compromise.
In February, 2005 a Cryomech PT410 cryorefrigeration unit was installed on one of the
Wessington liquid helium storage dewars in the temporary cryogenics facility built in 2001,
and operated throughout the austral winter as a proof of concept. The liquefaction rate of
this unit successfully matched the evaporation rate of the Wessington it was fitted to.
In fact, the cooling power was slightly greater than the Wessington boil-off and gas from a
second storage dewar was plumbed into the cryorefrigeration system and the total
liquefaction rate was measured to be 17.5 liters per day, 2.5 liters per day more than the
boil off rate of the Wessington dewar. These results were not totally unexpected since the
inlet temperature of the gas into the system was less than 300K resulting in less heat
removed from the vapor to be liquefied.
For the austral Winter of 2006 two more Cryomech PT 410 cryorefrigeration units were
procured and mounted on the remaining two Wessington storage dewars. As indicated by
Figure 4 all three units performed better than expected and some gas from the two 13,140
liter Gardner transport dewars was plumbed into the cryorefrigeration systems and that gas
was re-liquefied along with the Wessington boil-off.

104

Wessington Volumes
5000
4500
4000
3500

Liters

3000

W1
W2
W3

2500
2000
1500
1000
500

2/
15
/2

00
7
3/
1/
20
0
7
3/
15
/2
00
7
3/
29
/2
00
7
4/
12
/2
00
7
4/
26
/2
00
7
5/
10
/2
00
7
5/
24
/2
00
7
6/
7/
20
07
6/
21
/2
00
7
7/
5/
20
0
7
7/
19
/2
00
7
8/
2/
20
0
7
8/
16
/2
00
7
8/
30
/2
00
7
9/
13
/2
00
7
9/
27
/2
00
10
7
/1
1/
20
07
10
/2
5/
20
07

Date

Figure 4 In 2007 3 cryorefrigeration units were fully operational. Helium gas from the transport
dewars was plumbed into the system resulting in all 3 Wessington dewars increasing in volume over
the course of the season.

The cumulative total of gas re-liquefied amounted to 52.5 liters of liquid helium per day, a
total savings of product that, up until this year, had been lost to the atmosphere.
A manifold was designed, constructed, and installed that combined the boil-off of all five
dewars and distributed it to the three cryorefrigeration systems so they could all liquefy
helium gas at maximum capacity. Gas in excess of what the cryorefrigeration systems
could liquefy was then recovered and compressed into gas cylinders for use by the station
Meteorologists and other science projects. The difference for the 2006 austral winter was,
in previous years a cumulative total of 115 liters of liquid helium per day was being lost as
boil-off while in 2006 only 62.5 liters of LHe per day was being not being recondensed, a
product savings of 54%. These savings were repeated in 2007.
Additional fuel and power savings were made in 2007 when the new cryogenics facility
was modified to use the ambient temperature of the South Pole to cool the associated
compressors and heat the building. A glycol cooling loop and heat exchangers were
installed in the facility that removed waste heat from the compressors and distributed it
through the three modules of the cryogenics facility. The mechanical chillers were taken
offline and a savings of 4 kilowatts per system was realized for a total savings of 12
kilowatts. Less fuel is also used to heat the building for even more power and fuel savings.
2008 and the Future
The South Pole cryogenics infrastructure is now established. The capability to provide as
much as 80 liters per day of liquid helium to astrophysical projects now exists. This is a
105

100% improvement over seasons without the new innovations. The program also has the
capability to be flexible and to grow with developing technology. The South Pole
cryogenics program as well as the astrophysical projects it supports has grown from infancy
to maturity. The telescopes are now starting to incorporate pulse tube technology into their
design and using cryorefrigerators instead of cryostats to keep their detectors cold. By
keeping abreast of the latest technology, upgrading equipment and instruments as needed,
careful monitoring, and being innovative it is likely that increased improvements will be
realized among all stakeholders without a substantial increase in cost to the station.
CONCLUSION
Several factors were involved with the successful turn around of the South Pole Cryogenics
program. First, all member of the Liquid Helium Working Group had a stake in the success
of the program. All had expertise in areas that were relevant to the program. Conferences
were attended and manufacturers were consulted to explore options and new technology.
Second, budgets were established for the construction of a comprehensive cryogenics
facility. It was determined that the cost of lost scientific data far outweighed the cost of
constructing the new facility and the purchase of helium liquefiers. By bringing the
cryogenic equipment and instruments indoors and providing a warm space for the
cryogenics technician to work the dewars, transfer lines, vacuum pumps and other
equipment could be protected from the harsh South Pole environment and properly
maintained.
Cryorefrigerators had been looked at in the past but their cost, complexity, and lack of
reliability made them a questionable choice for use at the South Pole. Pulse tube
technology is more reliable because of the lack of moving parts in the cold head and
associated with that reliability is less downtime and less maintenance. They are less
expensive to purchase, less expensive to operate, produce less vibration, and generally have
longer lifetimes than traditional cryocoolers such as GM or Stirling cycles. They are more
suited to the challenges presented to the Cryogenics Program at the South Pole.
ACKNOWLEDGMENTS
The authors would like to thank Jesse J. Alcorta at Raytheon Polar Services for his
invaluable input to the South Pole Cryogenics program for over 15 years. Thanks also to
Chao Wang, Peter Gifford, and Brent Zerkle at Cryomech, Inc. for the design, installation,
and testing of the cryorefrigeration systems at the South Pole, Eddie Rowe at Wessington,
and to the NSF for funding this venture.
REFERENCES
1. Wang, C., Efficient Helium Recondensing Using a 4K Pulse Tube Cryocooler,
Cryogenics (2005) 45 719-724
2. Handbook of Chemistry and Physics, 69th Edition, CRC Press, Inc, Boca Raton, FL
(1989)
3. Van Sciver, S. W., Helium Cryogenics, Plenum Press, New York, New York (1986)

106

HELIUM
TEMPERATURE
TECHNIQUES

107

CR08-04
CRYOGENIC SYSTEM
OF THE SWISS ULTRA-COLD NEUTRON SOURCE
Anghel A.1, Blau B.1, Daum M.1, Kirch K.1, Grigoriev S.2
1

Paul Scherrer Institute, CH-5232 Villigen-PSI, Switzerland,

2
Efremov Institute, St. Petersburg, Russia
ABSTRACT

The ultra cold neutron source under construction at the Paul Scherrer Institute is a new
facility dedicated to the production of UCN. An essential element of this source is a 5 K
solid ortho-deuterium moderator. The cryogenic system of the UCN source is presented
emphasizing on the thermal process, design criteria, and evaluation of heat loads in the D 2
condenser, para-ortho D2 converter and moderator.
INTRODUCTION
Ultra cold neutrons (UCN) are neutrons with very low kinetic energies (less than ~300neV)
that can be trapped in evacuated storage volumes. They are mostly used for studying
fundamental properties of the neutron and its interactions. A new facility is under
construction at the Paul Scherrer Institut (PSI), Switzerland, dedicated to the production of
UCN and designed to produce almost two orders of magnitude larger intensities and
densities of UCN than presently available [1]. The production of UCN takes place in steps
starting with a several seconds long pulse of the PSI proton beam (590 MeV, 2 mA) hitting
a Pb target [2] and generating about 10 spallation neutrons per proton. The neutrons are
thermalized in a 3 m3 D2O moderator tank surrounding the proton target. A cold moderator
of 30 dm3 solid D2 (Fig.1, left panel) further cools the neutron spectrum and generates
UCN. Sufficiently cold neutrons can excite a phonon in the solid D 2 and, by that, loose
almost all kinetic energy to become ultra cold. The reverse process of gaining energy due to
collisions of UCN with phonons is suppressed by cooling the moderator at a temperature of
about 5K. In order for UCN to leave the solid D 2 into a vacuum storage system and to
experiments, the D2 quality is crucial: besides the low temperature also a low hydrogen
contamination (<0.1%) and a low (<1-2%) para-D 2 concentration is required [3]. The
cryogenic system is not only used to cool the D 2 moderator to 5 K, but also to prepare the
appropriate quality of D2. Starting from 30m3 normal D2 gas at 1bar in (warm) storage
tanks, the D2 is liquefied and solidified in a condenser vessel and transferred as liquid to a
para-to-ortho converter (Fig.1, right panel). At a liquid temperature close to the triple point
temperature of deuterium (18.7K), ortho D2 is produced in a catalytic conversion process
using OXISORB [4]. The ortho D2 is then transferred to the moderator, either as liquid
(~20 K) or vapor. The solid D 2 is formed in the moderator by slowly cooling from the
liquid through the triple point to 5K or by directly solidifying from the gas phase,
respectively. The cryogenic system is also used to operate a cryopump (5K) and the thermal
radiation shield (60K) of the large UCN storage tank of the source. Therefore, a dedicated
helium cryogenic system, built around an existing refrigerator, has been designed to deliver
cooling at the three required temperature levels, i.e. at 5K, 20K, and 60K.

109

Condenser

sD2 Moderator

Converter
Phase-separator

Figure 1: The layout of the ultra-cold neutron source. The solid D2 moderator (left panel) and the
combined helium-deuterium cold-box (right panel) are shown.

1. COOLING CONCEPT
1.1 System components and heat loads
The main part of the cooling system is installed in a cold box (Cryobox) which includes
the cold helium system and the preparatory part of the D 2 system. The Cryobox, apart from
valves, piping and sensors, consists mainly of three vessels provided with external
(condenser, converter) or internal (phase separator) heat exchangers as shown in the right
panel of Fig.1. The vessel at the top is the deuterium condenser. It is cooled down to about
5K by a 4.5K supercritical helium loop in order to condense and freeze the deuterium from
the storage tanks. The reason we chose the freezing process instead of condensation is the
relatively high vapor pressure of the liquid deuterium down to the triple point (194.6mbar
at 18.71K). At this pressure a considerable amount of deuterium gas will remain in the
30m3 storage tanks (almost 1kg) and therefore lost for the process. At the end of the
condensation process, the condenser is separated from the D 2 storage tanks by closing a
valve. The solid deuterium is then melted by increasing the temperature of the condenser to
about 20K. Liquid deuterium is transferred by gravity to a second vessel, the converter,
already at ~20K. Here, with the help of a special catalyst material (OXISORB ), a chromic
oxide (Cr2O3) impregnated silica gel, the para-to-ortho conversion of deuterium takes place.
Once the desired ortho-D2 concentration ~98-99% has been reached, the user has two
options: 1) ortho-D2 crystal growth in the moderator vessel directly from the liquid phase or
2) ortho-D2 crystal growth in the moderator vessel by resublimation.
In the first option a valve placed at the bottom of the converter is opened and liquid
deuterium is transferred by gravity to the D 2 moderator vessel. The moderator vessel has
already been cooled down before (i.e. during the para-ortho conversion phase) to about 20K
and prepared to receive the liquid ortho-D2. After the transfer by gravity, it is further cooled
down slowly to 5K and solid deuterium is produced from the liquid phase. This cooling
process is continued until all deuterium in the moderator vessel is frozen. The moderator
vessel temperature is then maintained at the lowest possible temperature (~5K).
In the second option the ortho-D 2 crystal is grown by resublimation i.e. solidification from
the gas phase. For this purpose, the moderator vessel is cooled down to 5K and then a valve
is opened such that the saturated ortho-D2 vapor from the converter can condense on the
cold walls of the moderator vessel. The converter is kept at about 20K during this process.
110

A third vessel, the helium phase separator, will be placed in the Cryobox as well. It serves
as an interface to the helium refrigerator. This vessel contains 5 helium heat-exchangers

Figure 2. PID of the UCN cooling system. The heat exchangers HX2-HX5 are actually placed in the
phase separator in contact with the liquid helium bath.

(HX1-HX5 in Fig.2) in contact with the liquid helium bath where the warm helium coming
from the process is re-cooled before being throttled in the Joule-Thomson valves and
returned as cold gas to the refrigerator. The special heater EH1 in the phase separator is
installed in order to simulate the full heat load and will consume the unused cooling power.
Other important components of the cryogenic system are the solid D 2 moderator vessel, the
cryopump and the thermal shield of the UCN storage tank. The D 2 moderator vessel and the
cryopump are located inside the UCN tank several meters below the Cryobox.
The complete Process and Instruments Diagram (PID) scheme of the cryogenic system for
UCN is shown in Fig.2. The temperature, pressure, liquid level and mass-flow sensors used
for the process control are indicated. A 10m long transfer line with a loss of about 0.5W/m
connects the 4.5K and the 60K refrigerator ports with the Cryobox.
1.2 Solid D2 moderator vessel
The main heat load on the solid deuterium moderator vessel occurs during the ultra cold
neutron production at initially 5K when the proton beam hits the spallation target. In order
to estimate the thermal load a thermal transient FEM model was built based on the
moderator vessel geometry and the data for power deposition. The moderator consists of a
vertical cylinder of solid-D2 at 5K of diameter 474mm and height 157mm (V=27.7dm 3,
M=5.51kg). The bottom surface of the D2 is 356.5mm above the target axis. The D 2 is
surrounded by a wall of AlMg4.5 alloy with a thickness of 1.5mm on the outer and bottom

111

side and 1mm on inner side. On the top surface a radiation boundary condition is applied
and on the inner, outer and the bottom surface convective boundary conditions are

Figure 3: The time dependence of the moderator heat load and maximum (a) and minimum (b)
temperature in the solid deuterium during the neutron pulse for h=100W/m 2K .

applied corresponding to the cooling by 20g/s supercritical helium with an inlet temperature
of 4.5K. Two scenarios were investigated: 1) a good, but conservatively low, heat transfer
coefficient (h=100W/m2K) on the convective boundaries and 2) an adiabatically insulated
vessel, corresponding to a poorly-cooled moderator. The power deposition in solid-D 2 was
taken from [5]. Interpolated data tables for the temperature dependence of the material
properties like specific heat, density, thermal conductivity of solid deuterium, AlMg4.5
alloy and supercritical helium at 3bar have been used in the calculation. The applied pulse
is rectangular with a flat top of 8s and the temporal development is followed for 120s.
The time dependence of the temperature in the solid deuterium is shown in Fig.3. As can be
seen the maximum temperature is below 13K i.e. the solid deuterium does not melt during
the pulse. The total heat load on the refrigerator has a peak of 250W at the end of the pulse.
The average load is 80W, well below the available cooling power of the refrigerator, 300W.
1.3 The condenser
The purpose of the condenser is to pump/condense as much gaseous D 2 from the storage
tanks as possible. Due to the rather large volume of the storage tanks (30m 3, ~5kg nD2) and
the relatively high vapor pressure of liquid deuterium (~194.6mbar at triple point) the
condensation of deuterium even at the triple point temperature (18.71K) is not enough to
extract the whole D2 mass from the storage tanks. Deuterium solidification is therefore
required. A heat exchanger helix around the condenser is used to cool it down to 4.5K
during the condensation process. In this phase the condenser works practically as a
cryogenic pump. In order to improve the condensation efficiency, the condenser is
equipped with 6 internal fins to extend the condensing surface and improve the heat
transfer. The condenser volume is around 40dm3 and is made of copper.
The heat load on the condenser during the 5K operation was estimated using a simple film
condensation model. The equilibrium wall temperature was calculated by equating the heat
deposited by the condensation process to the heat transferred to helium in the cooling coil
wound around the condenser vessel
h ( Tw ) Sc ( Tsat Tw ) = hScoil Tm ,log
112

where Sc is the condenser area including the fins, h the heat transfer coefficient from wall
to helium and Scoil the coil area. The heat-transfer coefficient from D 2 vapor to wall was
calculated using the Nusselt correlation for the film condensation [6]
g ( v ) kl3 hvl
h ( Tw ) = 0.943 l l

l ( Tsat Tw ) Lc
which depends on the liquid and vapor density l , v , the liquid viscosity l and the latent
heat of vaporization hvl of D2. Tsat is the saturation temperature, Tw the wall temperature
and Lc the height of the fins. Tm ,log is the helium logarithmic mean temperature difference.
If on the helium side we assume a heat transfer coefficient of h = 500W/m2K and a
temperature difference at the outlet of 1K we obtain a wall temperature of 23K and a heat
load of 767W. Under these conditions the deuterium condensation speed will be 2.4g/s and
the ~5kg deuterium will condense in about 0.6 hours. The calculated helium mass flow
necessary to sustain this process is around 7g/s. Unfortunately, under this condition the heat
load is higher than 300W, the maximum available cooling power of the refrigerator. The
solution is to throttle the deuterium flow using a valve placed before the condenser such
that during the condensation process the load on the refrigerator does not exceed the
available cooling power. The drawback is a longer, but not prohibitive, condensation time.
0.25

1.4 The converter

The purpose of the converter is the catalytic para-ortho conversion of deuterium. The
catalyst (OXISORB) is placed inside a cylinder with a volume of about 50dm 3. About one
third of the converter is filled with catalyst. It is not possible to fill the converter completely
with catalyst because it will adsorb a big volume of D 2 and this volume will be lost. The
converter vessel is provided with a demountable flange which allows the change of the
catalyst for regeneration purposes. A cooling pipe, controlling the process temperature, is
soldered on the outer wall of the converter vessel.
Because the generation of UCN is proportional to the amount of ortho deuterium and in
order to estimate the heat load generated by the catalytic conversion a mathematical model
has been developed. The underlying physics is an exoenergetic catalytic conversion
reaction coupled to a non-homogeneous diffusion-convection process and a thermal
conduction process with a volume heat generation problem. Initially, at 300K, deuterium is
in the so called normal deuterium phase (nD 2) with the equilibrium composition: 66.67%
ortho-D2 and 33.33% para-D2. After cooling to 5K and reheating to 20K the composition
does not change appreciably with time because the natural conversion rate of para-D 2 to
ortho-D2 is very small (~0.06%/hr, time constant ~70days). The initial condition is
o
o
therefore D2 ( t = 0 ) = D2,i = 66.67% . The catalyst and the forced convection in the
liquid accelerate dramatically the conversion. Measurements performed at PSI [4] show
that the time constant of the conversion is in the range = k 1 = 3 3.5hr . The initial
temperature is assumed to be 20K, the temperature of the heat sink at the converter wall.
o
The conversion equation is a first order reaction for the ortho-D2 concentration D2
D2o
t

= k D2,o eq D2o

113

o
where D2, eq = 1 exp ( 85.0 T ) is the equilibrium concentration of ortho-D2 at temperature T (K). At 20K the ortho-D2 equilibrium concentration is 98.6%. Newly created
ortho-D2 diffuses out from the catalyst, a process described by a standard diffusion equation
D2o
= D44 ( T ) 2 D2o
t
8
where D44 ( T ) = 5.70 10 exp ( 54.0 T ) is the D2 bulk self-diffusion constant [7]. Inside
the catalyst the effective diffusion constant is probably smaller. We assume it to be three
times smaller. The para-to-ortho conversion generates heat at a rate given by
D2o
Q = p o LD2
t
where
N E
J
p o = A J =1 = 180.91103 ; EJ =1 = 7.5 meV
M D2
kg

is the latent heat of conversion and LD2 is the density of liquid D2.
Finally, the thermal conduction process is described by the equation of heat conduction in
liquid D2 with a volumetric heat source Q, the heat generated by the para-ortho conversion
T
LD2 C p
= ( T ) + Q
t
where and C p are the thermal conductivity and the specific heat of liquid deuterium.
The simulation result presented in Fig. 4 indicates that in the absence of forced convection
the converter performance is poor even after 24 hours. Two thirds of the D 2 volume not in
contact with the catalyst will practically not convert while the other one third will convert
almost completely resulting in an average ortho-D 2 concentration of ~77.4% much lower
than the required 98%. Therefore a gas-lift pump driven by a small heater (EH4 in Fig.2) to
circulate the liquid D2 trough the converter is used. The D 2 vapor generated by the gas-lift
pump is recondensed in the condenser at 19K. With 50W of heating power, 0.15g/s D 2 can
be circulated. The recycling time for the 5kg of D2 contained in the converter is ~9hr.
1.5 The thermal shield
Due to the high level of neutron radiation we can not use super-insulation for the thermal
insulation of the UCN storage tank and moderator vessel. We resort therefore to the
classical solution of actively cooled thermal shield i.e. an aluminum screen cooled by a pipe
wound around it. The optimum distance between the turns is calculated using the equation
Dopt =

4km wTmax
q

where km is the thermal conductivity of the shield material, w the thickness, Tmax the
maximum allowed temperature difference and q the heat load from thermal radiation in
W/m2 . A typical calculation for an Al shield of 3mm thickness shows that for a helium inlet
temperature of 65K and a maximum temperature difference of 1K, the optimal distance
between the turns is 0.26m which results in having ~10 turns with an overall length of 42m.

114

Figure 4: The elevation of the equilibrium (a) and non-equilibrium (b) ortho-D 2 concentration (right
panel) after 24 hours along the vertical diameter of the converter. The time dependence of the
maximum D2 temperature (a) and the temperature in the center of the converter (b) (left panel).

The calculated radiation heat load is 550W and the total helium pressure drop is 120mbar.
If the helium inlet is at the bottom of the shield then the bottom shield temperature is 65K
and the top shield temperature will be 70.5K. The helium outlet temperature will be 70.3K.
The Cryobox, which is placed 6m above the moderator vessel, has a low level of neutron
radiation and super-insulation can be used without any problem of degradation.
2. REFRIGERATOR
2.1 Specifications
The UCN source needs cooling at three temperature levels: 1) ~5K used to condense and
keep the deuterium in the solid state in the condenser and moderator, 2) ~20K to liquefy the
solid-D2 in the condenser and to keep the liquid D 2 temperature in the converter during the
para-to-ortho conversion process and 3) ~60-80K for cooling the thermal shields of the
UCN storage volume. The cold source available is an existing helium refrigerator, which
can deliver cooling power at two temperature levels, 4.5K and ~60-80K. For the 5K
operation, the plant will be operated according to the regular specifications i.e. 300W at
4.5K with 32g/s. Typical for this operation is that the helium pressure before the last heat
exchanger stage will be reduced by a special valve from 10-12bar to any other pressure
down to 3bar. With this method we avoid an additional pressure reducing valve in the
Cryobox and assure a moderate operating pressure in the moderator vessel. Operation in the
20K temperature range is performed using heaters (EH2, EH3, EH5 in Fig.2) placed before
the cooling coils of the condenser, converter and moderator vessels. The warm gas coming
out of these coils is cooled back to 4.5K in the phase separators heat-exchangers HX3,
HX4 and HX5 and then throttled to low-pressure by valves as shown in the PID scheme of
Fig.2. The calculation for operation at 23K is illustrated in the T-S diagram of Fig.5. We
see that this operation can be performed with at most 5.5g/s while still ensuring that the
refrigerator can operate in steady-state conditions. Q heater is the power necessary to increase
the helium temperature to 23K. Q last and QHX1 are the heat transferred to the low pressure
side of the last refrigerator heat exchanger and to the liquid helium bath of the phase
separator respectively. Finally, Qres is the rest cooling power which covers other thermal
losses like those in the transfer line, heat conduction and thermal radiation.

115

m = 0.042 [kg/s]

m heater = 0.005522 [kg/s]

m = 0.042 [kg/s]

T5 = 23 [K]
h5 = -1.430E+06 [J/kg]
P5 = 3 [bar]

P1 = 12 [bar]
T1 = 5.38 [K]

P2 = 3.1 [bar]

h1 = -1.545E+06 [J/kg]

T2 = 5.439 [K]
h2 = -1.545E+06 [J/kg]

last heat exchanger of refrigerator

Qlast = 205 [W]

QHX1 = 80.76 [W]

HX1

Qtot = 772 [W]

P3 = 3.1 [bar]
T3 = 4.541 [K]
h3 = -1.552E+06 [J/kg]

T6 = 4.541 [K]
h4 = -1.552E+06 [J/kg]

h6 = -1.533E+06 [J/kg]

xL = 99.54 [%]
Qres = 20 [W]

Qheater = 671.2 [W]

P6 = 1.35 [bar]

Figure 5: The T-S diagram of the proposed 20K operation. Maximum 5.5g/s can be used to
cool the condenser/moderator at maximum 23K. This mode requests 42g/s and operation at
1.35bar in the phase separator.
CONCLUSIONS
A helium cooling system for a new ultra cold neutron source readapting an existing
refrigerator has been designed and optimized. The main processes responsible for the heat
load: the neutron and gamma irradiation during a proton pulse, the D 2 condensation and the
para-ortho conversion have been modeled and corresponding heat loads and time scales
estimated. Special attention was given to the unconventional operation in the 20 K mode.
Numerical simulations were essential to understand the different processes.
REFERENCES
1. Atchison, F. et al., The UCN Source at PSI, Proc. of the Int. Collab. On Advanced
Neutron Sources, ICANS-XVIII, Guandong 2007
2. Wohlmuther, M. and Heidenreich, G., The spallation target of the ultra-cold neutron
source at PSI, Nucl. Instr. Meth. A564, 51 (2006)
3. Morris, C.L. et al., Measurement of Ultracold-Neutron Lifetimes in Solid Deuterium,
Phys. Rev. Lett. 89, 272501 (2002).
4. Bodek, K. van den Brandt, B. et al., An apparatus for the investigation of solid D 2 with
respect to ultra-cold neutron sources, Nucl. Instr. Methods, A 533, 491 (2004)
5. Atchison, F., Calculated values for heating, particle fluxes and activation in components
of the UCN source, Table IV, pag. 9, PSI Internal report TM-14-02-02 (2002)
6. Incropera, F. P. and DeWitt, D. P., Fundamentals of Heat and Mass Transfer, 4th edition,
John Wiley & Sons, Inc.
7. Souers, P.C., Hydrogen properties for Fusion Energy, University of California, Berkeley
1986

116

CR08-54
EXPERIMENTAL SET-UP OF HEAT TRANSFER
MEASUREMENTS IN HE II
Chorowski M., Fydrych J., Strychalski M.
Wrocaw University of Technology, Faculty of Mechanical and Power Engineering,
Wybrzee Wyspiaskiego 27, 50-370 Wrocaw, Poland
ABSTRACT
Superconducting magnets and cavities can be cryostated with superfluid helium under
saturation or elevated pressure. To optimize heat transfer processes through electrical
Rutherford superconducting cable insulation, a dedicated helium II cryostat has been
designed, manufactured and commissioned. The cryostat enables reproduction of vast range
of the helium thermodynamic states, including pressurization up to 6 bar of the superfluid
helium. The cryostat construction, instrumentation and external equipment including
vacuum pumping system are described. The paper also presents experimental results
obtained during the cool-down of the cryostat.
INTRODUCTION
Currently built superconducting magnets, with the coils wound of low temperature
superconductors, in most cases use NbTi alloy with its critical temperature equal to 9.6 K.
There are attempts to replace the NbTi alloy with inter-metallic compound Nb 3Sn
characterized by the critical temperature of 18.1 K, however the technological difficulties
encountered in case of Nb 3Sn mechanical and thermal processing are still in favour
of NbTi [1]. Hence in case of accelerator magnets, that usually have to provide stable
magnetic field at the level of 10 T, maximum acceptable temperatures can be as low as
1.8 K. The temperature can be practically achieved only with the use of superfluid helium.
In high-energy particle accelerators the superconducting magnet coils are exposed to the
particles resulting from the beam losses and interacting with the magnets what leads to the
energy dissipation in the coils and the helium. Additional amount of heat can be dissipated
in the magnet structure due to small inter-movements of the cable wires caused by thermal
stresses; moreover there are always residual heat fluxes from the environment through
unavoidable thermal bridges. All these heat fluxes must be uninterruptedly transferred to
the cooling helium within the allowable temperature margin which in case of superfluid
helium can be of the order of 0.1 K.
In practice the accelerator magnets are usually immersed in liquid helium which penetrates
and fills all voids in the magnet structure like for example: spaces in-between magnet yoke
layers, longitudinal channels designed for magnetic field quality improvement, a distance
between beam tube and magnet coil. As the NbTi superconducting coils do not need any
resin impregnation, the helium can penetrate also through superconducting cables due to
their inter-wire porosity, as well as through intentionally porous electrical insulation.
As mentioned above, all the integrated heat fluxes must be finally transferred to the helium.
The efficiency of this process is determined by the heat transfer intensity between the
magnet elements and the helium that can be in different thermodynamic states like twophase saturated helium at normal pressure (He I), supercritical helium, superfluid helium at

117

saturated pressure (saturated He IIs) or superfluid helium at about atmospheric pressure

(pressurized He IIp).
The high energy physics accelerators with the superconducting magnets (Hera at DESY,
Tevatron at FERMILAB, RHIC at Brookhaven National Laboratory and finally LHC at
CERN) have relatively small heat influxes to the helium (not exceeding few Watts per
meter). Sufficient heat transfer efficiency was guaranteed by magnet construction resulting
from the field requirements. Therefore during the magnet design phase the heat transfer
capability has not been the critical issue. However, because of planned LHC upgrade which
will be focused on the improvement of the beam luminosity, it is predicted that the
longitude density of the heat dissipated in the coils of some magnets can be as high as about
100 W per meter. To transfer this amount of heat to the helium requires reconsideration of
the construction of superconducting coils with respect to the heat transfer intensity. It is
very probable that the heat transfer can become a limiting factor in the increase of the
energy and luminosity in upgraded and future accelerators.
In spite of the widespread use of liquid helium for cooling of superconducting magnets, RF
cavities, cryogenic vacuum pumps and other devices and instruments, there is a lack
of sufficient knowledge that can be used for accurate calculation of heat transfer between
magnet structure and helium remaining in different thermodynamic states. There is also
a lack of experimental data which allow reliable estimation of the heat fluxes from the
Rutherford-type cables to supercritical or superfluid helium.
To perform heat transfer measurements a dedicated helium II cryostat was designed,
manufactured and commissioned at the Faculty of Mechanical and Power Engineering of
Wrocaw University of Technology. The paper presents the cryostat construction and
chosen reception test results.
1. EXPERIMENTAL SET-UP DESCRIPTION
The experimental set-up for performing heat transfer measurements in superfluid helium is
composed of three major parts (the scheme of the set-up is presented in Figure 1). The main
part is a Claudet-type cryostat which enables to pressurise superfluid helium inside the
measurement chamber [2]. During the cool-down and the measurements, the cryostat needs
to be regularly supplied with liquid helium. For this purpose the experimental set-up
is equipped with a liquid helium supply system.

Figure 1: Scheme of the experimental set-up

118

The supply system is composed of a 100-litre liquid helium dewar and vacuum insulated
helium transfer line. To force the liquid helium to flow into the cryostat, the dewar is
pressurized with the gaseous helium supplied from the helium cylinder. The third part of
the set-up is vacuum system which is composed of a high capacity vacuum pump, vacuum
line and vacuum control valve. The vacuum system allows lowering temperature by
decreasing the pressure of saturated helium in a dedicated volume in the cryostat.
1.1 Cryostat construction
The cryostat construction is based on the Claudet principle and its scheme is shown in
Figure 2. Inside the cryostat there are three separated helium vessels that allow keeping
helium simultaneously in three thermodynamic states: at normal boiling point temperature
(He I), at superfluid helium saturated (He IIs) and superfluid pressurised (He IIp) states.
Helium from the external dewar is regularly supplied to the He I vessel. The lambda plate
separates thermally the He IIp volume from the He I vessel, while the lambda valve allows
filling the He IIp vessel with helium and its replenishment during the subcooling phase.

Figure 2: The cryostat scheme and the helium thermodynamic state in the vessels

Liquid helium from the He I vessel can flow into the He IIs volume through a recuperative
heat-exchanger followed by Joule-Thomson valve. During the subcooling phase the helium
vapour from the He IIs volume is pumped away by a vacuum system and the helium vapour
pressure is decreased to the level of a few millibars. The evacuated subcooled helium
vapour preliminarily cools down the liquid helium streaming in the heat exchanger coil.
Due to low pressure and isenthalpic expansion in the Joule-Thomson valve, the temperature
of the helium is decreased below the lambda point temperature (2.17 K) and in the vessel
He IIs a certain amount of helium is constantly kept in a superfluid saturated state.
The wall of the He IIp vessel, which is made of oxygen-free copper, does not cause
a significant heat resistance and the helium that is closed in the He IIp volume is smoothly
cooled-down by the helium in the external He IIs vessel. As a result, the helium
temperature in the He IIp volume also decreases below 2.17 K, but the pressure in this
vessel can be kept at the stable level that is much higher than the equivalent saturated
helium pressure.

119

1.2 Lambda plate and high pressure He IIp vessel

A lambda plate is a partition that separates the He IIp volume from the He I vessel. To
decrease the heat transfer from liquid helium He I to superfluid helium a lambda plate is
usually made of high performance insulator material, as for example composite G10. Then
the pressure in the He IIp volume cannot differ significantly from the pressure in the He I
vessel, which is usually a little bit higher than atmospheric pressure.
The lambda plate in the described cryostat, as well as the He IIp vessel have been designed
to enable the measurements in superfluid helium which can be pressurised significantly
above normal pressure level. The wall of the vessel is thick enough to withstand pressure
difference up to 6 bars, whereas the lambda plate is composed of two rigid parallel metal
sheets. Figure 3 shows a cross-section view of the high pressure He IIp vessel together with
the lambda plate. The space between two metal sheets of the lambda plate is open to the He
IIs volume where temperature is a little bit lower than in the He IIp vessel and pressure is of
about few millibars. This solution helps with reducing heat flux from the He I vessel to
the He IIp volume. Heat is transfer only by thermal conduction through the lambda valve
seat pipe and through a thin metal collar that connects two vessels in their circumferences.
a

Figure 3: Photo (a) and cross-section view (b) of the lambda plate and high pressure He IIp vessel

1.3 Vacuum pumping system

Vacuum pumping system is composed of a 4-meter vacuum pipe DN50, a vacuum control
valve, a helium vapour heater and two vacuum pumps SOGEVAC SV 100 and SV 300
with the nominal pumping speeds equal to 97.5 m3/h and 280 m3/h, respectively. The
system is presented schematically in Figure 1.
The vacuum control valve allows adjustment of the pressure in the He IIs volume and in
this way the change and automatic control of the superfluid helium temperature. The
helium vapour heater is made of a coil-shaped copper pipe with an electrical heater wound
around it. It is installed in order to protect the vacuum pumps from damages that could be
caused by cold helium vapour inflow.
To design correctly the vacuum system and especially to select a vacuum pump with
a sufficient pumping speed a dedicated mathematical model of the cryostat subcooling
process has been applied.
1.4 Indispensable pumping capacity calculation
In the cryostat the subcooling process is based on the helium temperature lowering below
4.2 K in both He II vessels by decreasing the pressure in the He IIs volume. In the
presented calculations we assumed a constant subcooling rate at the level of 1 mK/s. This
120

0,8

0,8

0,6
0,4

0,6
T

0,4
p

0,2

0,2

3,2 3,4 3,6 3,8

4,2

Temperature T, K

3600

3000

2,2 2,4 2,6 2,8

2400

1800

1,8

1200

0
600

Pressure p, bar

Temperature T, K

Pressure p, bar

rate should lead to reaching 1.8 K at about 2500 s. Depending on the presented in Figure 4
relationship between the saturated helium pressure and its temperature, and taking into
account the assumed subcooling rate, we have obtained the evolution of the pressure in He
IIs volume. Figure 5 shows the dynamic characteristics of helium temperature and pressure
during the analysed subcooling process.

Time t, s

Figure 4: Helium saturated pressure versus

temperature

Figure 5: Evolution of temperature and pressure

in the He IIs vessel

We have also assumed that the content of helium in both He II volumes is V = 8 dm3 and
total heat leaks Qst to the helium through the vacuum insulation and thermal conduction
bridges is equal to 3 W. Making use of the relationship between helium density and its
temperature and pressure = (T, p), as well as between helium specific heat and its
temperature and pressure cp = cp (T, p), we derived a formula for an instant heat flux that
must be given up by the helium to guarantee assumed temperature drop rate:

Q=

V c p T
t

+ Qst .

(1)

Figure 6 shows the evolution of the heat flux Q during a whole subcooling period. At the
beginning of the process the heat flux is equal to 7.5 W only, and after decreasing gradually
to 5.5 W during the first 1800 s, it rises rapidly to above 14 W. This rapid change is caused
by an intense growth of the helium specific heat near the lambda line.
During the cryostat subcooling an instant heat flux Q is taken over by a certain amount
of evaporating helium m in a time t. This process can be described by the following
equation:

Q=

m
rHe ,
t

(2)

where rHe stands for heat of vaporization.

Due to the fact that the heat of vaporization for helium does not change significantly in
analysed temperature range (1.8 K 4.2 K), we have treated it as a constant and equal to 23
kJ/kg.
Combining the equations (1) and (2) yields a formula for the helium mass flow rate qm:

qm =

m
1 V c p T

=
+ Qst .
t
rHe
t

121

(3)

The variations of the calculated helium mass flow rate qm during the cryostat subcooling
process are also presented in Figure 6. At the beginning of the process the mass flow rate
is lower than 0.4 g/s and decreases to 0.27 g/s within the first 1800 s. Then, because
of an intense growth of the specific heat near lambda line, to keep the assumed rate
of temperature drop, it has to rise rapidly to the value above 0.7 g/s. As soon as the final
temperature (1.8 K) is reached, the mass flow rate decreases to 0.16 g/s, which is necessary
to keep a stable temperature in both superfluid helium vessels.
0,8

140

0,2

40
20
0
0

3600

3000

2400

1800

1200

600

0
0

60

Time t, s

3600

80

3000

0,4

2400

100

1800

10

120

1200

0,6

Volume flow rate qv, dm3/s

12

Mass flow rate qm, g/s

Heat flow rate Q, W

14

600

16

Time t, s

Figure7: Evolution of the volume flow rate of

helium vapour warmed up to 310 K.

Figure 6: Evolution of heat flux and helium

mass flow rate

Estimation of both the helium mass flow rate (Fig. 6) and helium pressure evolution
(Fig. 5) was very useful for determining the volume flow rate of the helium vapour through
the vacuum pumping system. Because the applied vacuum pumps can work only with warm
gases it is necessary to warm helium vapour up to 310 K before the pump inlets. For this
purpose we installed a coil with an electrical heater (see Fig. 1). This solution helps
to protect the pumps but it leads to the increase in an indispensable pumping speed.
The calculated evolution of volume flow rate of the helium vapour in the cross section just
before the vacuum pumps is presented in Figure 7. In the analysed case the volumetric flow
is affected not only by helium specific heat growth close to lambda line, but also by
continuous pressure decrease. Therefore the volume flow rate, which is equal to a few liters
per second at the beginning, rises to above 130 dm 3/s at the end of the subcooling process.
When the final temperature (1.8 K) is obtained the volume flow rate drops to 75 dm3/s.
The presented analysis proves that the indispensable pumping speed of the vacuum pumps
applied for subcooling of the cryostat should be higher then 75 dm 3/s. And of course the
higher pumping speed the faster the final temperature is reached. Therefore we have
decided to apply two pumps with the total nominal pumping speed of about 380 m 3/h,
which is equal to 105 dm3/s.
1.5 Instrumentation and data acquisition system
The cryostat described in the paper is equipped with two pressure sensors connected to the
He IIp vessel (shown in Figure 3) and with six cryogenic temperature sensors - Cernox.
All the temperature sensors for the cryostat reception test were installed as shown
schematically in Figure 8. Before the test the Cernoxes were calibrated in the temperature
range 1.64 290 K and during the test they were connected to a specially designed 1A
DC source.
He I

Figure 8: Schematical location of temperature sensors in He IIp and IIs vessels

T3 T4

T1

He IIp

122
T5 T6

He IIs

T2

To acquire measurement results we applied a dedicated LabVIEW code. The code

controlled KEITHLEY 2000 multimeter and read 6 resistances every second and converted
them on-line into corresponding temperature values.
2. THE CRYOSTAT RECEPTION TEST RESULTS
The analysed cryostat is dedicated for performing heat transfer measurement in superfluid
helium environment. After construction and checking its overall behaviour first in nitrogen
and then in helium temperature levels the cryostat was put to the final reception test.
The test was focused on subcooling helium and obtaining superfluid condition as well
as on decreasing temperature below 1.7 K. Measured temperature developments are shown
in Figure 9. The subcooling phase from the temperature 4.2 K to 2.17 K lasted 30 minutes.
During this period there were noticeable differences among measured temperatures, even
up to 0.4 K. Furthermore the helium inside the He IIp vessel was strongly stratified until the
superfluid helium temperature was reached. The difference between temperatures measured
in the top and bottom of the vessel exceeded 0.2 K. It was caused by a continuous small
leak of liquid helium through the lambda valve that was driven by thermal compressibility
of helium. This temperature difference disappeared when the helium in the He IIs vessel
crossed lambda point compare Figure 9.
4,3
Temperature T, K

3,9
3,5
3,1

He I

2,7

T1 & T2

T5 & T6
T3 & T4

2,3
1,9

He II

150

140

130

120

110

100

90

80

70

60

50

40

30

20

10

1,5
Time t, min
Figure 9: Measured temperature courses during subcooling of the cryostat

Figure 10 presents temperature evolutions during the superfluid condition period with a few
remarks concerning some important actions. The duration of this period was of about
60 minutes while the lowest temperature was equal to 1.6 K. The spread of measured
values was rather stable along whole period. The difference between temperatures
measured by coupled temperature sensors (T1& T2, T3& T4 and T5& T6) was lower than
25 mK.

123

Cool down
Disassembly
Cool down without
with helium vapour
of helium
helium vapour heater
heater
vapour heater

2,5

Helium in He IIs
vessel is over

Temperature, K

2,3
2,1

T5 & T6

1,9
T3 & T4

T1 & T2

1,7

140

130

120

110

100

90

80

70

60

1,5
Time t, min

Figure 10: Temperature changes during the superfluid helium state period
of the cryostat subcooling
The cool-down rate, expressed by the slope of the temperature evolution curve, indicates
that it is feasible for the cryostat to reach a lower temperature level than measured during
the commissioning. The subcooling phase was stopped due to the full use of the helium in
the supply dewar.
CONCLUSIONS
The heat transfer phenomena may be the critical issue in future high-field accelerator
magnet design and operation. There is a need of new experimental data enabling
verification of new ideas of enhancing the heat transfer between the magnet elements,
including superconducting cable and the cooling helium. A Claudet type cryostat oriented
at the heat transfer measurements in superfluid helium at elevated pressure has been
designed, manufactured and commissioned. A special attention has been paid to a high
capacity helium pumping system and a simple analytical model of the required pumping
process has been proposed.
REFERENCES
[1] Devred A. et al., High filed accelerator magnet R&D in Europe, IEEE Trans. Appl.
Supercond (2004) 14 2 339-344
[2] Claudet G., Aymar R., Tore Supra and He II cooling of large high field magnets, Adv.
in Cryogenic Eng. (1990) 35 55-67
ACKNOWLEDGEMENTS
The work has been partly supported within the cooperation agreement K944 signed
between CERN, Geneva, Switzerland and Wrocaw University of Technology, Poland.

124

CR08-39
S-N-S PHASE TRANSITIONS OF GEOMETRICALLYMETASTABLE SUPERCONDUCTING THIN FILMS
Ribeiro Gomes M.
Centro de Fsica Nuclear da Universidade de Lisboa, Av. Prof. Gama Pinto, 2,
1649-003 Lisboa, Portugal
ABSTRACT
Phase transitions of superconducting films continue to be a frontier research area, not only
as fundamental solid state issue but also due to the current potential applications of these
devices as radiation detectors. A description of the S-N transition is provided in terms of
flux penetration regimes and experimental characteristics fields; the N-S transition, being a
continuous process, is associated with the different regimes of superconductivity
nucleation.
INTRODUCTION
A complete theoretical description of Superconducting-Normal-Superconducting (S-N-S)
phase transitions in laminar pure type-I/II superconductors is still. A key contribution to the
better understanding of the S-N-S phase transition is achieved, using an unconventional
experimental technique of real time flux motion detection, complementary to the
conventional SQUID measurements which provide static magnetization results integrated
over time scales ~3 orders of magnitude larger; or to static images, and more recently realtime high-resolution movies, with time scales ~0.1 s and spatial resolution of a few m.
The combination of these three techniques provides a promising laboratory for further
investigation of the S-N-S phase transitions driven by an applied magnetic field, transport
current, or radiation induced. Highly-sensitive energy detectors are continuously a
challenging R&D topic, and cryogenic microcalorimeters using superconducting films are
at the present time the best detector candidates for X-rays, electrons, nuclear recoils, etc, in
many different fields of physics [1], such as mass spectroscopy of biomolecules, X-ray
chemical analysis for industrial applications, cosmological dark matter searches, solar
neutrino studies, X-rays astrophysics, double beta-decay and direct neutrino mass
determination experiments. Superconducting microcalorimeters provide a large number of
advantages over the other type of detectors. They can be relatively large and still be
sensitive to small amounts of deposited energy, since the energy is sensed after it has been
converted to heat (even interactions that produce little or no ionization can be detected).
Being a thermal detectors, they do not depend on the charge transport properties of the
absorber and a variety of superconducting materials can be used, in contrast to more
conventional ionization detectors. While the microcalorimeter operating principle is simple,
the full detector performance and optimization is not trivial. A full understanding behind
the detection application rests on the solid state behavior of the superconducting films.
1. EXPERIMENTAL TECHNIQUE
The measurements were performed in a single-shot He-3 refrigerator at temperatures 300
mK<T<4.2 K, with an overall experimental uncertainty of better than 0.5%. The magnetic
field was applied perpendicularly to the sample by a coil external to the refrigerator, with
an homogeneity of 1% over the sample area and relative precision of better than 2x10 -4. The
applied magnetic field (Ha) sweep rate was varied from 0.5 G/s to 250 G/s, and it is PC125

remote controlled allowing a continuous increase or decrease of H a and its reversal to zero.
The experimental technique is a real-time fast-pulse detection, described in detail elsewhere
[2, 3]. The detection principle is based on Faradays law: the motion of a flux bundle within
the sample, surrounded by a sensor loop, induces an electromotive force, which is
proportional to the amount of flux quanta associated with the applied magnetic field
strength and contained in a moving bundle of a given surface area. The flux motion induces
a voltage-signal in the Cu pickup loop that is transformed-coupled to a fast amplifier. A 10
kHz frequency cutoff on the detection bandwidth allows a discrimination between applied
magnetic field increments and flux bundle motions within the sample bordered by the
sensitive loop. The acquisition system permits the synchronization of the magnetic field
step-rise with the recording gate during which pulses are detected. The samples were
mounted in a multi-layered design mechanically fastened to the cold plate of the He-3
refrigerator. The samples were cut from pure, annealed, pinhole-free 12.5-125 micron thick
(Ly) polycrystalline metal foils provided by different commercially available suppliers. The
purity was, in general, better than 99.99%. All samples were observed under different
optical microscopes, with a < 1 m spatial resolution, for surface dimension measurements
and geometry imperfections. Any sample showing significant geometrical imperfections
and non-parallel edges (maximum tolerance of 10% on the sample width) was not used in
experiments. Prior to sample mounting, all samples were cleaned to eliminate any surface
contamination by hand handling. The sample width (Lx) and length (Lz) were chosen in the
majority of the cases to fit within the outer width of the Cu-loop (width of each Cu layer
150 m and length of 3.2 cm; inner width of the loop 550 m and outer 850 m). Different
type-I/II materials were studied: Re, Sn, Al, In, Pb, Ta, Zn,V and Nb, even if the majority
of the studies were performed on Re and Sn.
2. SUPERCONDUCTING-TO-NORMAL PHASE TRANSITION
As is commonly known, the S-N transition is characterised by three distinct regimes of flux
penetration demarcated by two fields, denoted H fdp and Hldp for first and last detected
penetration event, respectively. Typical integral and differential flux penetration curves are
presented in Fig. 1(a) for a Sn and 1(b) for a Re film, respectively, where the number of
events is plotted as a function of the applied magnetic field, H a, which was varied from 0 G
up to Ha>Hc(T).
Re 25mx760mx30mm - 340 mK, 3 G/s, 13 mV, 3 cycles
3

1.2 10

Sn 50mx800mx25mm - 320 mK, 15 mV, 3 G/s, 4 cycles

4

4.5 10

1 10
4

3.6 10

8 10
ldp

events

differential events

2.7 104

1.8 104

9 10

6 10

4 10

fdp

fdp

ldp

2 10

0
0

50

100

H (G)

150

200

250

50

100

150

200

H (G)
a

126

250

300

350

Figure 1: (a) Integral S-N curves for 50-m thick Sn film; (b) Differential S-N penetration curves for
a 25-m thick Re strip.

In the first regime, delimited by 0<H a<Hfdp, no flux is admitted in the sample volume and a
diamagnetic band is created along the sample volume. At a magnetic field given by
H a = (1 D ) H c ( T ) , where D = Lx Lx + L y is the demagnetization factor for a
rectangular cross section, the superconducting state becomes unstable at the sample edges
and flux is allowed to penetrate in the sample edges, up to an extent of the order of a few
coherence lengths, (T), creating therefore an edge intermediate state, well described by
Landau [4]. This process is reversible and undetectable with the fast-pulse readout
technique. The second regime, defined by HfdpHaHldp, is characterized by the migration of
flux bundles across the diamagnetic band, containing a finite number of flux quanta, which
diameter is determined by the curtain-like structure generated at the sample edges. Each
flux bundle, circled by a superconducting current, moves to the center of the sample. This
process is irreversible and detectable with the fast-pulse technique. As the magnetic field is
further increased, the central region occupied by flux bundles expands and the diamagnetic
band shrinks. The onset of this regime is geometry dependent and the end corresponds to
the complete destruction of the diamagnetic band. The third and last regime is again
reversible and delimited by Hldp<Ha<Hc(T). In this regime, the normal central region has
reached the curtain-structure at the edges, creating normal corrugated channels that expand
reversibly with the magnetic field increase until the sample is in a complete normal state.
No events are detectable.

In contrast with earlier wisdom [5, 6] this first flux penetration field severely depends on
the trapped flux inside the sample volume after a first magnetic field scan. The so-called
Hfdp in virgin (trapped flux free) penetration curves shows a good agreement with the
predictions from Benkraouda [7] proposed for type-II or type-I superconducting flat strips.
In the case of non-virgin curves the first entry field is defined by:

H fdp = H C (T )

1
1+ 1

where g = L y Lx is the aspect ratio and 1 and depends on the amount of trapped
flux; for the virgin case =1 , and within the same measurement but different magnetic
field cycle is constant. The end of the irreversible regime of penetration, initiated by H fdp,
is defined by the magnetic field H ldp where the last event is detected; no more events are
detected for a further increase of the magnetic field. This field is generally smaller than the
critical field Hc(T). However, when extrapolated to zero-threshold, and consequently
eliminating the noise cut-off, Hldp equals the critical field, in contradiction with the common
knowledge but in agreement with published work [5]. Consequently, the H ldp studies
indicate that this observable field is coupled with the annihilation of the diamagnetic band
surrounding the strip edges, and therefore has the same geometry dependence as H fdp. The
various experimental results yielded:
H ldp = H fdp

where depends on the superconducting material and threshold setting: for Sn

=1.4 0.4 independent of the temperature and for Re = 4.1 0.4 . The
interpretation underlying these two constants is under investigation for a systematic study
conducted in other materials undergoing a complete phase transition. For H a>Hldp, a
127

continuous flux penetration driven by Ha occurs until Ha=Hc(T) and the strip is fully
normal. In this regime no events are detected since they are outside the bandwidth of the
pulse amplifier.
3. NORMAL-TO-SUPERCONDUCTING PHASE TRANSITION
The N-S phase transition, on which fewer theoretical predictions exist and the disagreement
among researches as to the existence of a barrier to flux expulsion is still open [8]. The
measurements indicate that, although the N-S transition is a continuous process [9]
characteristic fields defining the different regimes of flux expulsion are observable and
depend on intrinsic superconducting properties of the material. In contrast with other
experimental techniques and previous published work [3], it was possible to identify for the
first time the two major characteristic expulsion fields: first expulsion field, H fe, associated
with the spontaneous nucleation field of superconductivity on a perimeter surface sheath,
Hc3, suggested by Saint-James et al.[10] but never observed previously with this technique;
and second expulsion field, Hse, the onset field of the spinodal regime, H c2. Both fields were
measured in different materials with different geometries.
In our experiments it was observed that the N-S phase transition is typically characterized
by four regimes, demarcated by three characteristic fields: H fde, and Hfe and Hse, as shown in
Fig. 2. The integral curve corresponds to the expulsion part of the hysteresis cycle,
following a complete field ramping to well above Hc(T).

events (A.U.)

regime 4

regime 2

regime 3

regime 1

rd

3 expulsion
field
nd

expulsion field

st

1 expulsion field

H (A.U.)
a

Figure 2.:Typical expulsion curve with regimes and characteristic fields identification.

In the first regime for H fde<Ha no events are detected, this regime may be extended to H fe
with the disappearance of Hfde; this field may be lower or higher than H c(T) depending on
the superconducting material and sample properties. The regime 2 of expulsion is defined
by Hfe<Ha<Hfde may begin at fields ~H c(T) depending on the superconducting properties of
the material. The third and more perceptible then regime 2 occurs for H se<Ha<Hfe, where a
variable number of events may be detected. The majority of events are detected in regime
4; frequently this is the only detectable regime, and occurs for 0<H a<Hse. The identification
of the characteristic expulsion fields is also shown in Fig. 3, which represents typical a
differential expulsion curves with four visible regimes of expulsion present. This
measurement was conducted in a 50-m Sn sample at 350 mK.

128

Sn 50mx800mx25mm - 350 mK, 3 G/s, 14 mV, 5 cycles

100

differential events

80

60

H
40

se

fde

fe

20

0
0

50

100

150

200

250

300

Ha (G)

Figure 3: Typical differential expulsion curve with characteristic fields identification.

The flux expulsion onset occurs at Hfde and is characterized by a narrow signal. A further H a
decrease induced no additional expulsion events until H a=Hfe where a small event rate is
observed. The field Hse defines the beginning of a rapid event rate which persists
throughout the remainder of the field decrease. H c3 has a weak signature corresponding to
~0.05% of the total detected signal, consistent with a surface sheath of width 2 at the strip
perimeter. This field is often not observed due to the noise level associated with that
measurement. Hc2, in contrast, sets the beginning of the last regime of flux expulsion where
~99% of the signal is detected. While H se does not depend on the sample geometry nor
surface quality, providing a convincing agreement with Hc2(T), as can be seen in Fig. 4, and
the extracted Landau-Ginzburg parameter, as indicated in Fig. 5 for the case of Re; H fe
deviates from the theoretical predicted Hc3(T) or Hfe(T)/Hse(T) depending on the sample
surface roughness.
Material survey
0.7

0.6

Pb

Sn

Ta

In

Re

h calculated
c2

0.5

se

0.4

0.3

0.2

0.1

0
0

0.05

0.1

0.15

1/ H

Figure 4: Variation of hse with Hc/2 for various materials.

129

0.2

Re at 350 mK
0.55

kfrom Hc2
kfrom Hc3

0.45

experimental

0.5

0.4

0.35
0

20

40

60

80

100

120

140

L (m)
y

Figure 5: Experimental values as a function of sample thickness for various Re samples.

A third expulsion field, Hfde, is frequently observed at higher magnetic fields and in the
vicinity of Hc(T) which is associated with heterogeneous nucleation resulting from
metallurgy and/or impurities that stimulate nucleation in the vicinity of the defect, also
observed with other experimental techniques [11].
CONCLUSIONS
In this work, a complementary approach of the S-N-S thin film phase transition is provided
in terms of the magnetic flux motion within an intermediate state evolution process,
achieved with an unconcentional real-time fast-pulse readout technique. Different
superconducting materials were studied under various experimental conditions, and for
different geometrical aspect ratios. The magnetic fields characterizing the energy barrier to
flux penetration, of geometrical original, were clearly identified and compared with the
existing theoretical models for the first flux penetration field. The N-S phase transition,
being a continuous process, was shown to be geometry independent and directly related
with the superconducting properties of the material. The two characteristic expulsion fields,
were observed, for the first time with this technique, corresponding to the spontaneous
nucleation field of superconductivity on a perimeter surface sheath and the onset of the
spinodal regime field. A third expulsion field, associated with heterogeneous nucleation
resulting from metallurgy and/or impurities stimulating nucleation near the defect, together
the other two fields were proven to be an excellent tool to extract fundamental
superconducting parameters, such as the Landau-Ginzburg of a material. These are key
features when superconducting systems are designed for radiation detectors [12, 13], in
general. The achievements in the -decay 187Re spectrum for Hp<Hfdp for direct electron
antineutrino mass determination yielded promising expectations [14]. On the other hand,
superconducting cryogenic detectors are currently the best performing detectors, in
particular for neutrino mass measurements [15], where thin superconducting films are used
as thermal sensors. Specifically, transition edge sensors (TES) have demonstrated a
sensitivity of a few eV. The theoretical prediction for its optimum performance has not yet
been experimentally reached and the modeling of such a behavior is necessary. The
intrinsic vortex dynamics and intermediate state evolution, within the superconducting
volume, are fundamental features to clarify for pushing an experimental detector sensitivity
to the sub-eV threshold.
130

ACKNOWLEDGEMENTS
This work was supported by the grants PRAXIS/10033/1998 and SFRH/BPD/36293/2007
of the Foundation for Science & Technology of Portugal. Part of this work was
accomplished during the author PhD programme at the Center of Nuclear Physics of the
University of Lisbon (CFNUL), Portugal and the experimental work supported by the
scientific projects funds of the Advanced Detectors Group of the CFNUL.
REFERENCES
th

1. Proceedings of the 11 International Workshop on Low Temperature Detectors, Nucl.

Instr. And Meth. A (2006) 559
2. Jeudy, V., Limagne, D. and Waysand, G., Irreversible flux entry in tin superconducting
strips: Geometrical Metastability, Europhys. Lett. (1991) 16 491-496
3. Jung, G., Girard, T.A.,Valko, P., Gomes, M.R., Jeudy, V., Limagne, D. and Waysand,
G., Expulsion of magnetic flux in a type-I superconducting strip, Physica C (2002) 377
121-129
4. Landau, L.D., Zh Eskp. Teor. Fiz (1937) 7 371 and Jour. Phys. (1943) VII(3) 99
5. Jeudy, V., Jung, G., Limagne, D. and Waysand, G., Irreversible flux penetration regimes
in type-I superconducting strips, Physica. C (1994) 225 331-336
6. Gomes, M.R., Destruction of the geometrical energy barrier in type-I superconducting
rhenium, Master Thesis, Faculty of Science/University of Lisbon, Portugal (1998)
7. Benkraouda, M. and Clem, J. R., Magnetic hysteresis from the geometrical barrier in
type-II superconducting strips, Phys. Rev. B (1996) 53 5716-5726
8. Valko, P. Gomes, M.R:, and Girad, T.A., Nucleation of superconductivity in thin type-I
foils, Phys. Rev. B (2007) 75 140504-1-4
9. Huebener, R. P., Magnetic flux structures in superconductors, (1979), SpringerVerlag, Berlin Heidelberg, 2nd ed., (2001)
10. Saint-James, D., and Gennes, P.G., Onset of superconductivity in decreasing fields,
Phys. Lett. (1963) 7 306-308
11. Shoenberg, D., Superconducting cylinders, Proc. Camb. Phil. Soc. (1937) 33 260-276
12. Gomes, M.R., Study of phase transitions in geometrically-metastable superconducting
-laminar systems, PhD Thesis, Faculty of Science/University of Lisbon, Portugal
(2005)
13. Jeudy, V., Collar, J.I., Girard, T.A. Limagne, D. and Waysand, G., S-35 beta irradiation
of a tin strip in a state of superconducting geometrical metastability, Nucl. Instr. And
Meth. A (1996) 373 65-67
14. Gomes, M.R., Girard, T.A., Oliveira, C., Jeudy, V. and Limagne, D., A
superconducting measurement of the 187-Re beta decay spectrum, Nucl. Instr. And
Meth. A (2000) 444 84-87
15. Nucciotti, A. for the MARE collaboration, The MARE Project, J. Low Temp. Phys.
DOI 10.1007/s10909-008-9718-5

131

CR08-07
BLACK SURFACES FOR CRYOGENIC APPLICATIONS
Krlk T., Hanzelka P., Musilov V., Srnka A.
Institute of Scientific Instruments of the ASCR, v.v.i
Academy of Sciences of the Czech Republic, Krlovopolsk 147, 612 64 Brno
ABSTRACT
Surfaces with high absorption or emission of thermal radiation are often needed in
cryogenic and space applications. But it can be more difficult to realize a sufficiently black
surface applicable in cryogenic systems than to make highly reflective surfaces.
Experimental results on black surfaces concerning the thermal radiative properties in the
temperature range from 20 to 300 K of the source of the thermal radiation are presented.
INTRODUCTION
Black surfaces which absorb or emit thermal radiation at low temperatures have found their
typical application in space-borne devices [1] and cryopumps [2, 3]. They are realized as
coatings on surfaces of metals that ensure the heat transfer towards or from the black
surface. Especially in this region of applications it is not a trivial task to prepare sufficiently
black surfaces. In addition to the blackness in infrared and far infrared, the properties of
the surface like outgassing rates (vacuum compatibility), mechanical properties or possibly
directional residual reflectivity (diffuse or specular) are of great importance. Primary
motivation of this work was to find a suitable coating for baffles of a cryopump [4] and to
develop a reference black surface in apparatus for emissivity and absorptivity
measurements [5, 6]. For these reasons we have measured and compared low-temperature
emissivity and absorptivity of various coatings on copper and aluminium.
1. METHOD AND APPARATUS
We have performed measurements of radiative heat flow between the opposite surfaces of
two interchangeable concentric parallel discs (samples) of the same diameter, the radiator
and the absorber [5, 6].
From the measured radiative heat flow QR, the mutual emissivity (emissivity factor) RA of
radiator and absorber surfaces was evaluated:

RA =

QR
A (TR T A )
4

(1)
In this relation, TR and TA (TR>TA) are the temperatures of the radiator and absorber,
whereas A and represent the area of the measured sample surface and the StefanBoltzmann constant, respectively.
Let R represent the emissivity of the radiator and A the absorptivity of the absorber,
respectively. Then an approximate relation can be written:

RA

(2)

This relation is exactly valid for spectral values of RA, R and A and when parallel surfaces
are infinite. Thus for total hemispherical values R(TR) and A(TA, TR), the formula (2)
133

represents a good approximation when properties of the surfaces depend weakly on

wavelength of the radiation (grey surfaces), i.e. when A(TA, TR) = A(TA) = A(TA). Here
A(TA, TR) is the total hemispherical absorptivity of the absorber surface at the temperature
TA for incident radiation of blackbody at the temperature TR. For finite area we can use the
formula (2) when the diameter of radiator and absorber is much larger than the distance
between them (view factor approaches unity).
The approximate values of absorptivity A or emissivity R of a sample can be calculated
from (2) for the known values R or A of the surface opposite the sample (reference
surface). When the emissivity or absorptivity of the reference surface is much higher than
the absorptivity or emissivity of the sample, the measured RA may be considered, with a
small error, to be the absorptivity or emissivity of the sample, respectively.
From the measurement of heat transfer between two identical samples, the emissivity of
one surface can be evaluated if we assume that absorptivity does not depend on the
temperature of the material or this dependence is weak, i.e. when A(TA, TR) A(TR). Then
the emissivity of the radiator equals the absorptivity of the absorber and thus the emissivity
R(TR) = A(TR) can be evaluated from the measured value RA using the approximate
relation (2).
In our apparatus, the heat exchanged between radiator and absorber QR is conducted from
the absorber through the thermal resistor into the LHe bath and its value is derived from the
temperature drop on the thermal resistor in steady state. The range of measurable heat QR
can be changed by adding of a thermal shunt parallel to the thermal resistor. A bronze tape
or copper wire was soldered in parallel to the thermal resistor, which is made from a thin
stainless steel tube. Consequently, for the same QR the temperature TA of the absorber is
different with different shunt, which we also used for testing of the influence of the
absorber temperature on the result of measurement.
The accuracy of QR measurement is within 3 % of the measured values. In addition, a
systematic error, reducing the measured values, corresponds to the value 0.96-0.97 of view
factor for radiator and absorber configuration. This error reduces by about 3-4 % the
measured value of RA for nearly black surfaces.
2. MATERIALS
Sample substrates are made of a Cu sheet (99.5%) and they have the form of discs with the
diameter of 40 mm and thickness of approx. 1 mm. The side of the substrate on which
coating is applied was usually treated by grinding with wet corundum sand or with steel
wool.
Three types of the epoxies were used. The first of our samples was covered with epoxy
Araldite LY 5210 filled with soot. Later we applied epoxy resin Spolimer with hardener
HT1 and Epoxy Chs520 with hardener Telalit 600. Both epoxies are produced by the Czech
firm Spolchemie, a.s [7]. In some coatings polyester screen printing fabric Sefar PET
1000 77-48Y PW (abbrev. SPF) was used as filling and reinforcement. This net-like fabric
is 80 m thick and its projected area is about 65%.
2.1 Samples
Epoxy Araldite LY 5210 + 1% soot - Samples 1, 3
The roughened substrate was brush-painted with the epoxy Araldite LY 5210 with
hardener Aradur HY 2954 (prod.: Ciba A.R.L. Ltd) filled with the 1 wt% soot
134

Chezacarb (prod.: Chemopetrol a.s, Czech Rep.). The thickness of the coating is on average
of 70 m and it has glossy black colour.
Epoxy Spolimer with one layer of the SPF - Samples 32, 33
One layer of the SPF impregnated with the Epoxy Spolimer was placed on the substrate.
The excess of the epoxy was wiped away. The epoxy was hardened 1 h at 100C, 1 h at
140C and 2 h at 180C. The total thickness of the coating is 90 m. The coating has red
brown appearance with visible texture of the SPF and it is visually semitransparent. This
coating was also used as an absorptive coating on chevrons in a small helium bath
cryopump [4].
Epoxy Chs520 with one layer SPF covered with the Mylar foil - Samples 49, 50
The SPF impregnated with the Epoxy Chs520 was applied on the substrate. The SPF was
covered with the Kapton foil coated with Teflon and a force 10 N was applied. After the
successive epoxy curing for 1 h at each of the temperatures 60C, 80C, 100C and 120C,
the Kapton foil was peeled off and the Mylar foil without aluminization (th. 12 m) was
glued on the epoxy surface with the adhesive Loctite 480. The total thickness of the layer
is 120-130 m (Nr. 49) and 140-160 m (Nr. 50). The coating has spotted black appearance
due to black colour of the adhesive.
Epoxy Chs520 380 m (four layers of the SPF) - Samples 57, 58
Four layers of the SPF impregnated with the epoxy Chs520 were placed on the Kapton foil
coated with Teflon and than the substrate was pressed on this layer. The curing proceeded
under a force of 10 N and at the same temperatures as by the samples 49 and 50. After
curing the epoxy, the Kapton foil was peeled off. The samples have yellowish appearance.
The total thickness of the coating on both samples is on average 380 m.
Chemglaze Z306 on aluminized Kapton - Samples 39, 40
Double side aluminized Kapton coated with the paint Chemglaze Z306 on one side was
applied on these samples. The Kapton foil has thickness 25 m and the total thickness of
the coated foil is 80 m. The paint is based on a polyurethane binder filled with fumed
silica and carbon [1]. This foil was glued with the Loctite 480 to the substrate covered
with the similar layer as that on the sample Nr. 32 and 33. The foil was provided by the
firm Austrian Aerospace GmbH, Austria [8]. The paint Chemglaze Z306 (Aeroglaze) is
used for thermal control in space-borne instruments and it is also often used as a well-tried
black surface for cryogenic applications.
Black paints Samples L1, L2 and L3, L4
Commercially available paints produced by the German firm Warnecke&Bhm GmbH
(WB-Lacke) were applied on substrates. The samples L1 and L2 were coated with visually
matt-black epoxy baking paint. The binder of the paint is acrylic epoxy resin. The samples
L3 and L4 have the same appearance but the type of the paint binder is amino-alkyd. The
thickness of the coating on samples L1 and L2 is approx. 20 m and 25 m for the samples
L3 and L4. Both paints are one-component, they were applied by spraying and hardened by
baking at 150C.
Diamond Like Carbon (DLC) on copper substrate - Sample 65
The DLC coating was deposited by combined PVD/PACVD (Physical Vapour
Deposition/Plasma Assisted Chemical Vapour Deposition) method in industrial coating
equipment of the firm HVM Plasma s.r.o [9]. The base layer Cr/WC:H was deposited by
135

magnetron sputtering to maintain good adhesion of the DLC, whereas the top DLC layer
was deposited by pulsed plasma from acetylene. The total coating thickness is 3.3 m. The
thickness of the top DLC layer is 2.2 m. This coating is commercially used in mechanical
engineering for reduction of friction and wear. Maximum working temperature of this
coating is 300C.
Fractal Black on Al foil - Samples 60, 61
Fractal Black is a coating very black for visible and infra-red radiation. It was provided
by the firm Acktar Advanced Coatings Ltd. (Israel) [10]. The coating was created by
magnetron sputtering on a thick Al foil. The foil was glued with Loctite 480 to the
substrate. Fractal Black is completely inorganic layer and its thickness is 10-12 m.
Nanocomposite coatings on copper substrate - Samples 67, 68
MARWIN SI (sample 68) is a nanocomposite AlTiSiN system which is commercially
used as a super-hard coating of the machinery cutting tools. The coating is created by using
PVD in special coating facility in the firm SHM s.r.o. (Czech Rep., umperk) [11]. The
micro hardness of the coating is 43 GPa, the thickness is 2-3 m and the thermal stability is
better than 1000C. The colour of the coating is black.
LUBRIK SI (sample 67) consists of the gradient film TiAlN as the base layer and of the
end lubricating layer TiAlCO. The total thickness of all layers is 3.8 m. The coating has
shiny black colour. This coating is used for working of nonferrous metals and for pressing
and forming tools and it was also created by the firm SHM s.r.o.

ME_P5
ME03
ME06
ME09
ME11

1
32
49
57
58

3
33
50
58
57

Coating and Configuration

Radiator Absorber
Epoxy Araldite HY 5210 + 1% soot
Epoxy Spolimer + 1 SPF
Epoxy Chs520 + 1 SPF + Mylar foil
Epoxy Chs520-380 m
Epoxy Chs520-380 m

ME05
E28
MEL1
MEL2
A62
A63
E25
E29
E30
E32

40
40
L1
L3
58
58
65
61
68
67

39
58
L2
L4
65
65
58
58
58
58

Chemglaze Z306 on Kapton

Chemglaze Z306 Epoxy Chs520-380 m
Acrylic epoxy paint, WB-Lacke
Amino-alkyd paint, WB-Lacke
Epoxy Chs520-380 m DLC
Epoxy Chs520-380 m DLC
DLC Epoxy Chs520-380 m
Fractal Black Epoxy Chs520-380 m
MARWIN SI Epoxy Chs520-380 m
LUBRIK SI Epoxy Chs520-380 m

Meas. Radiator Absorber

Thermal
shunt
Shunt 0
Shunt 0
Shunt 0
Shunt 1
Shunt 1
Shunt 0
Shunt 2
Shunt 2
Shunt 2
Shunt 2
Shunt 1
Shunt 2
Shunt 2
Shunt 2
Shunt 2

Table 1 : Description of individual measurements and their configuration.

3. MEASUREMENT RESULTS AND DISCUSSION

Table 1 lists the measurements. The letters ME refer to the measurements in which the
same surfaces are both on the radiator and absorber. Letter E refers to the emissivity
measurement, i.e. when the studied sample is placed on the radiator and for the absorber is
used a reference surface as black as possible (samples 57 and 58). Letter A refers to the
136

absorptivity measurement with the sample mounted on the absorber opposite to black
radiator. Measurement without shunt (Shunt 0) means that the basic thermal resistor made
of a stainless steel tube is used. In most measurements, a bronze (Shunt 1) or copper shunt
(Shunt 2) was soldered in parallel to the basic thermal resistor.
In following figures, the open symbols and dashed lines represent the emissivity R or the
absorptivity A of the samples, evaluated from the measured values RA (full symbols) by
using the relation (2) for known A or R of the opposite surface or using relation A = R.
3.1 Epoxies
No cracking or peeling of coatings after repetitive cool down from room temperature to 5 K
was observed. Outgassing, very probably of water, was observed during several
measurements of very low values of absorptivity of metals when the epoxy on the radiator
achieved temperatures over 250 K. This effect was registered as a very slow increase in
measured metal absorptivity, probably due to formation of a condensate on the metal.
Identical samples 57 and 58 were used as the radiator and absorber in the measurements
ME09 and ME11. In each of these measurements, we used different thermal shunts and
thus the influence of the absorber temperature TA (Fig.1 right hand side axis) on the
measured mutual emissivity could be checked. Although the temperatures of material on
the absorber differ at the same TR in the experiments ME09 and ME11, the measured values
of RA are identical within the accuracy of measurement. It is in agreement with expectation
of weak temperature dependence of optical properties of dielectrics in infrared and farinfrared. From the measurement we can deduce that at least within the interval of material
temperatures TA from 5 to 60 K and radiation temperatures TR from 40 to 160 K, the
absorptivity of the samples 57 and 58 depends very weakly on the temperature of the
coating.
100

200

90

180
160
ME09, ME11 - Chs520 380 m, R
ME09 - Chs520 380 m, RA, Shunt 1
ME11 - Chs520 380 m RA, Shunt 2
ME06 - Chs520 + 1 SPF + Mylar foil 120 m, R
ME03 - Spolimer + 1 SPF 90 m, R
ME06 - Chs520 + 1 SPF + Mylar foil 120 m, RA
ME_P5 - Araldite LY 5210 + 1% soot 70 m, RA
ME03 - Spolimer + 1 SPF 90 m, RA

RA, R [%]

70
60
50
40

140
120
100
80

30

60

20

40

10
0

ME09 TA - Shunt 1
ME11 TA - Shunt 2

20

40

60

80

TA [K]

80

20

0
100 120 140 160 180 200 220 240 260 280 300

TR [K]

Figure 1 : Epoxy coatings. Measured RA and emissivity R evaluated using relation (2) under
condition R=A. Gray curves TA(TR ) are temperatures of the absorber in measurements
ME09 and ME11.

137

3.2 Paints
The emissivity of the Chemglaze Z306 at TR=300 K, evaluated from our measurement, is
85 % (Fig. 2). This is a very good result in comparison with epoxies if we take into account
that the thickness 55 m of the Chemglaze Z306 coating is nearly one order smaller than
the thickness of the epoxy layer with emissivity 89 %. The data found in literature show
higher values of the Chemglaze Z306 emissivity [12, 13]. This may be caused by the
higher thickness of the brush-painted coating (100 m) in the case of the Jamottons data
[12]. Nevertheless the values in [12] are unusually high in comparison with our results for
the temperatures TR<100 K. The data published by Fabron [13] are more consistent with
our results.
For both paints from WB-Lacke and Chemglaze Z306, we obtained nearly the same value
of emissivity at 300 K. For the paint with amino-alkyd binder, also the low temperature
emissivities approach the values obtained for Chemglaze Z306. The thickness of this
paint is half of the Chemglaze Z306 paint thickness.
Z306, Jamotton 1997, Centre Spatial de Liege, Belgium, 100 micron
Z306, Fabron [13], Alcatel Space Industries
Z306, Fabron [13], citation of Paul Scherrer Institute, Switzerland
Z306, Fabron [13], citation of ISRO Satelite Centre, India

100
90
80

RA , R [%]

70
60
50
40

ME05 - Chemglaze Z306 55 m, R

E28 - Chemglaze Z306 55 m, R
ME05 - Chemglaze Z306, 55 m, RA
MEL2 - Amino-alkyd paint 25 m, WB-Lacke, R
E28 - Chemglaze Z306, 55 m Chs 520 380 m, RA
MEL1 - Acrylic epoxy paint 20 m, WB-Lacke, R
MEL2 - Amino-alkyd paint, WB-Lacke, 25 m, RA
MEL1 - Acrylic epoxy paint, WB-Lacke, 20 m, RA

30
20
10
0

20

40

60

80 100 120 140 160 180 200 220 240 260 280 300

TR [K]

Figure 2 : Paint coatings. Measured RA, emissivity R evaluated from (2) and published data on
emissivity of Chemglaze Z306.

3.3 Thin film coatings

The highest values of emissivity were achieved with the coating Fractal Black (Fig. 3).
The total hemispherical emissivity R estimated from the relation (2) for the temperature
region 230300 K is approximately 90 %. For comparison, the spectral hemispherical
absorptivity of the Fractal Black, according to the producers data [10], decreases from
99.5 % at wavelength 10 m (300 K black body temperature) to 96 % at 13.5 m (230 K).
The measured absorptivity of the Diamond Like Carbon is of about 60 % in the range
TR = 150-300 K (Fig.3). Similarly like for thick epoxy layers (see sec. 3.1), we checked the
influence of the temperature of DLC layer on its absorptivity. In two measurements with
different shunts, the different temperatures TA of DLC layer at the same temperature TR
(Fig. 3) were achieved. The measured absorptivity did not change with temperature of the
138

layer. We can conclude that in the region of temperatures TA = 575 K and TR = 60190 K
the absorptivity of the DLC layer does not depend on the material temperature. The
independence on material temperature was confirmed by emissivity measurement (not
presented in Fig. 3). The values of mutual emissivity RA obtained in absorptivity and
emissivity measurements at TR = 30300 K are equal within a relative accuracy of 2 %.
The films MARWIN SI and LUBRIK SI have lower emissivity above 40 K than the DLC
(Fig. 3).
E29 - Fractal Black 10 m, RA
E29 - Fractal Black 10 m Chs520 380 m, RA
A62, A63 - DLC 3.3 m, A
A62 - Chs520 380 m x DLC 3.3 m, RA, Shunt 1
A63 - Chs520 380 m x DLC 3.3 m, RA, Shunt 2
E30 - MARWIN SI 2-3 m
E30 - MARWIN SI 2-3 m x Chs520 380 m
E32 - LUBRIK SI 3.8 m

RA, A, R [%]

90

200
180

80

160

70

140

60

120

50

100

40

80

30

A62 TA - Shunt 1
A63 TA - Shunt 2

20
10
0

TA [K]

100

60
40
20

20

40

60

80

100 120 140 160 180 200 220 240 260 280 300

TR [K]

Figure 3 : Thin films. Measured RA and from (2) evaluated emissivity R and absorptivity A. TA(TR)
are temperatures of absorber in measurements A62 and A63.

CONCLUSIONS
We have measured total hemispherical emissivity and absorptivity of various types of
coatings below room temperature up to 15 K. Coatings based on epoxy resins (thickness
70-380 m), polyurethane and alkyd paint (thickness 20-55 m) and also inorganic thin
films (2-11 m) were applied on Cu or Al substrate. All coatings withstood cryogenic and
vacuum conditions. Also the films that are commercially used as super-hard coatings of
machinery cutting tools lasted out the cool down to 5 K in spite the fact that the coatings
were applied on Cu whereas the technology is originally developed for cutting tools
coatings. The nanocomposite coatings and DLC film may be applicable in cryogenics if
also high temperatures are needed.
For two coatings, thick epoxy layer and DLC layer, we have tested the dependence of
radiative properties on the temperature of the coating in the region of material temperatures
5-80 K (temperatures of the source of radiation were 20-160 K). We did not observe any
dependence of measured radiative properties on the temperature of material. The
independence on material temperature was confirmed for DLC layer by measurement of
both absorptivity and emissivity at temperatures from 20 to 300 K.

139

The increase of transparency for far infra-red electromagnetic radiation with increasing
wavelength is the property expected for dielectric materials. This behaviour results in the
decrease of emissivity of dielectric layers on metals with decreasing temperature. For all
coatings we observed the decrease in measured emissivities at low temperatures, i.e. the
surfaces are not grey at low temperatures. At higher temperatures, weak dependence on the
temperature was obtained. The non-greyness at low temperatures could influence the values
of low temperature emissivities and absorptivities that were evaluated under the assumption
of grey surfaces.
The highest values of emissivity were achieved for thick epoxy composite layer (380 m)
and Fractal Black (11 m). The epoxy coating is filled with polyester net (filling more than
50 %). At temperatures over 60 K its emissivity approaches 90 % and about 80 % at 30 K.
For Fractal Black coating, we have measured values 85-88 % above 120 K and about 60 %
at 30 K. In spite of very low coating thickness (approximately 3 m), relatively high
emissivity values were observed for DLC layer.
ACKNOWLEDGEMENT
This work was supported by the Academy of Sciences of the Czech Republic (Project Nr.
AV0 Z20650511). The authors thank the firms HVM Plasma s.r.o. (Czech Rep.), SHM
s.r.o. (Czech Rep.), Spolchemie a.s. (Czech Rep.), Acktar Advanced Coatings Ltd. (Israel),
and Austrian Aerospace GmbH (Austria) for providing the samples.
REFERENCES
1. Persky, M. J., Review of black surfaces for space-borne infrared systems, Rev. Sci.
Instrum. (1999) 70 2193-2217
2. Benvenuti, C.J., Characteristics, advantages, and possible applications of condensation
cryopumping, J. of Vacuum Science and Technology (1974) 11 591-599
3. Haefer, R.A., Kryo-Vakuumtechnik, Springer-Verlag, Heidelberg (1981)
4. Musilova, V., Dupak, J., Hanzelka, P., Kralik, T. and Urban, P., Economical helium bath
cryopump: design and testing, Vacuum (2004) 74 77-83
5. Kralik, T., Hanzelka, P., Musilova, V., Srnka, A., Device for measurement of thermal
emissivity at cryogenics temperatures, 8th Cryogenics 2004 IIR international
conference, Icaris, Praha (2004) 23-29
6. Musilova, V., Hanzelka, P. Kralik, T., Srnka, A., Low temperature radiative properties of
materials used in cryogenics, Cryogenics (2005) 45 529536
7. Spolchemie a.s. Czech Rep. [online]. [2008] <http://www.spolchemie.cz>
8. Austrian Aerospace GmbH. [online]. [2008] <http://www.space.at/htmldocs/2.html>
9. HVM Plasma s.r.o. Czech Rep. [online]. [2008] <http://www.hvm.cz>
10. Acktar Advanced Coatings Ltd. [online]. [2008]
<http://www.acktar.com/category/FractalBlack>
11. SHM s.r.o. Czech Rep. [online]. [2008] <http://www.shm-cz.cz/en/products/pvdcoatings>
12. Jamotton, P. et al., Measurement of the total hemispherical emittance of different
surfaces at temperature from 4.4 to 200 K, Sixth European Symposium on Space
Environmental Control Systems, Noordwijk, The Netherlands, ESA SP-400 (1997) 543547
13. Fabron, Ch., Meurat, A., Measurement of total hemispheric emissivity at low
temperatures / designing a cryogenic test bench, Fourth International Symposium
Environmental Testing for Space Programmes, Lige, Belgium, ESA SP-467 (2001) 5159
140

HIGH-TEMPERATURE
SUPERCONDUCTIVITY

141

CR08-11
25 TESLA HTS MAGNET INSERT COIL IN ZERO BOIL OFF
CRYOSTAT
Good J., Bracanovic D.
Cryogenic Ltd, 30 Acton Park Industrial Estate, London W3 7QE, UK
ABSTRACT
There is increasing interest in magnetic fields for NMR at above 1 GHz (23.48 Tesla) but
these fields are not available with commercial Low Temperature Superconductors (LTS) at
either 4.2 or 2.2 K or reduced temperature.
Cryogenic Ltd has manufactured a coil which is designed to demonstrate the feasibility of a
magnetic field of 25 Tesla in a working bore of 50mm. The magnet uses HTS conductors
combined with LTS and is suited to solid state research for NMR and ESR. It is cooled by
closed cycle cryostat and runs at 4.2K.
The application of High Temperature Superconductor (HTS) is additionally attractive
because the magnet can run at 4.2K rather than being pumped to 2.2K as the critical field of
HTS is much higher than 25 Tesla.
The magnet consists of 5 coils; the outer coil is of NbTi section, and the next of Nb3Sn
section. Inside are 3 coils of a high-temperature superconductor HTS BiSrCaCuo-2223
tape.
The outer two coils have a 140mm bore and provide 15 Tesla at 4.2K. The target for the
HTS is to provide up to 9 Tesla at 4K.
The performance of the magnet both LTS and HTS section is discussed together with the
operating characteristics of the closed cycle cryostat.
1. INTRODUCTION
There is increasing interest in magnetic fields for NMR at above 1 GHz (23.48 Tesla) but
these fields are not available with commercial Low Temperature Superconductors (LTS) at
either 4.2 K at a reduced temperature of 2 K.
Cryogenic Ltd has manufactured a coil which is designed to demonstrate the feasibility of a
magnetic field of 25 Tesla in a working bore of 50mm at 4.2K.
At present the only way fields of 25 Tesla can be generated continuously is by a bybrid or
resistive magnet. However, hybrid coils require 10 megawatts of DC power to operate and
the field generated is not sufficiently stable [1].
The applications for extra high field magnet systems are many and varied. Some examples
are:o

Characterization of new material and semiconductors including nitrides and zinc

oxides, by Hall effect resistivity, specific heat, magnetic moment, De Haas-van
Alphen measuremnts.

High field spectrometry using NMR at 1 GHz for research into structures of bio
macro-molecules which is important for the development of new drugs. Both

143

sensitivity and resolution can be improved by moving to higher frequencies of 1

GHz and beyond.
o

Measurements of other Physical properties such as quantum fluids and plasma

phyiscs.
2. THE PROJECT

In this project the ultimate goal is to provide a superconducting magnet capable of

generating a continuous magnetic field of 25 Tesla. The magnet was built under the
HIGINS project and funded by the EU under the 6th framework for Research.
The magnet with a first generatioin HTS set of insert coils has been delivered to the
Engineering Department of Cambridge University. Since the BISCO [2] tapes used in the
HTS insert coils were relatively low current density after insulation, the field provided is
modest so the magnet only provides 19 Tesla at 4.2K. New coils are now being build using
second generation HTS tapes that have much higher currents and with these the magnet
should achieve the expected 25 Tesla field performance.
The magnet had to meet strict requirements for stray field with a 5 Gauss line not more than
4 metres from the field centre and this motivates a compact design since it is the diameter
of the outer windings which give use to a large diopole moment and stray field.
It was also required that the magnet should have low or zero helium consumption.
Accordingly, the magnet was designed to run in liquid helium with a zero loss cryostat.
The cryostat has its own 4K cryocooler [3] built in to recycle and recondence helium.
2.1.The Magnet
The outer magnet is designed to be as compact as possible but to provide 15 Tesla at 4.2K
in a 140mm bore. Inside this diameter are three coils each made of HTS conductor, two
coils have single lengths and one required a single joint within the winding. These coils
ideally would have provided 9 Tesla at 4.2K giving a total field of 24 Tesla. The outer
magnet is made with two winding, one of NbTi and one of NbSn.
Filamentary NbTi conductor is used for the outer winding and provides a field of 8.5 T at
4.2K in a bore of 220mm.
The inner coil is wound using bronze route Nb 3Sn. This conductor is widely used and very
reliable. It can withstand high tensile stress and is well suited for fields up to the 17 Tesla.
Above that the critical current decreases making alternative conductors more useful. To
obtain good high field performance, the Nb 3Sn is alloyed with small quantities of tantalum
and titanium, which both increases Jc and Bc.
The three inner sections are made of a high performance, high strength BiSrCaCuO tape in
a silver alloy matrix. The project generates up to 9 Tesla in a background field of 15 Tesla.
Research on silver alloys to provie a high strength matrix proceeded in Poland [4] with
additional tests in Dresden [5] under the Higins Project. In the event, however, the best
conductor available to wind the coils was purchased from American Superconductor. It
was insulated by winding a thin Kapton tape as a lap around the BISCO tape at Trithor [6].
The specification of the conductor is shown in Table 1 below. The dimensions of the
conductor winding are given below in Table 2.

144

Table 1: Specification of HTS tape

Average thickness:

0.4 mm

Minimum width:

3.9 mm

Maximum width:

4.3 mm

Minimum bend diameter:

50 mm

Maximum tensile stress:

65 Mpa

Critical current at 77K

100 A

Critical current at 4K

300 A

Table 2: Dimensions of the Conductor Winding

Inner
Radius
(mm)

Outer
radius
(mm)

Magnet
length
(mm)

No. of
turns

Wire
length
(km)

Inductance
(H)

Nb3SnN

72

108.9

320

NbTi

114

168.7

360

12232

6.93

29.4

23699

21.052

91.3

HTS 1

25

43

100

590

0.112

8-3

HTS 2

43

49

120

532

0.125

10-3

HTS 3

575

65

140

576

0.221

10-2

Each of the three windings is made as conventional layer winding and not as a series of
pancake windings as this reduces the number of joints and power dissipation. Connections
are made to each end of the HTS tape by NbTi conductors using soft solder. These joints
are at a low field point at one end of the magnet. Figure 1 shows one of the coils being
wound. After winding each coil is impregnated with resin.

Figure 1: HTS tape winding

145

2.2 The cryostat

The important requirement for the cryostat was to have the helium boil-off from the magnet
system kept to a minimum and preferably zero. The cryostat is designed to be
recondensing and uses a Gifford-McMahon refrigerator with a base tempertaure of below
4 K. The first stage of the cryocooler is used to cool a single radiation shield surrounding
the helium bath at about 50K. The 2 nd stage re-condenses gas evolving from the helium
reservoir using a special design of heat exchange.
The cryocooler and compressor require 6.5kW for operation. When operating normally
the cryostat has no boil-off and will in fact condense gas from room temperature so that the
helium reservoir slowly fills over time. The magnet is supported inside the cryostat with a
standard support structure from the top plate. The whole assembly can be lifted out of the
cryostat so that the magnet can be inspected or modified. Four sets of current leads are
provided so that the inner HTS and outer LTS coils can be energised seperately. A lambda
plate has been built into the cryostat to allow the magnet to be operated at 2.2K but it has
not been used. A drawing is shown in Fig 2.
When running the magnet power supplies an additional 3kW is drawn, giving a total power
consumption of less than 10 kW. This compares extremely favourably with the 10 mW
power requirements of conventional resistive high field magnets.

146

Figure 2: Drawing of zero-boil off

cryostat housing the magnet.

147

3. TEST RESULTS
The NbSn / NbTi magnet was built and tested first together with the zero boil off cryostat.
The cryostat, which has been installed at Cambridge [7], has proved very successful and
without current in the magnet leads, it slowly condensed helium from room temperature at
a rate of nearly 100cc per hour.
The outer LTS magnet has been tested at 4.2K to 14 Tesla at which point stress induced
training quenches were observed.
The inner HTS coils were first seperately tested in liquid helium at 4K. The coils show
good performance with low resistive losses up to about 300 Amps. All three coils have
some losses which are believed to be due to the resistance of the joints between sections or
to the NbTi connecting cables. The effective resistance appears to be about 1.5 Ohm on
each coil as can be seen from the test results show of Figure 3.
400

Inner HTS coil, (63.29 G)

tested to 1.51T (240A)

350
300

V(

V)

250
200
150
100
50

Voltage
Linear Fit of Data1_B

0
0

50

100

150

200

250

I(A)

Figure 3: Inner HTS coil test.

The rather low engineering current density achieved in these HTS windings limits the field
available from the combined magnet to 18 to 19 Tesla.
4. CONCLUSIONS
While the field achieved today by this magnet is not higher than can be achieved with LTS
conductors at 4.2K it does demonstrate the practicality of the approach. Furthermore new
2nd generation YBCO conductors with up to four times the current density are now
becoming available and with these conductors a field of 25 Tesla appears practical for this
magnet provided the windings can be designed in such a way as to support the forces
involved.
A magnet of 25 Tesla which is all superconducting will have unique advantages compared
to the hybrid alternative. Firstly, both capital and running costs are very much lower.
Secondly, the field is more stable and quieter. It will also be possible to make the magnet
of high homogeneity more easily. Ideally, a true persistant coil would be made however
this may prove difficult due to flux creep in the HTS material as well as the difficulty of
making superconducting joints to HTS materials.

148

5. REFERENCES
1] Zhehong Gan, Hyung-Tae Kwak, Mark Bird, Timoth Cross, Peter Gorkov., wiiliam
Brey, Kiran Shetty. High Field NMR using resistive and hybrid magnets, High Field NMR
resistive and hybrid magnets
2]. American Superconductors Corportation, 64 Jackson Road Devens MA 01434
3] Sumitomo Heavy Industries Ltd, 2-1-1, Yato-cho,Nichitokyo-city, Tokyo, 188-8585
4] Maciej Chorowski, TTA Techtra Sp. z o.o.ul. Muchoborska 18, PL 54-424 Wroclaw,
Poland
5] Wolfgang Hassler, IFW, Institut fr Festkrper- und Werkstofforschung Dresden e. V.,
Helmholtzstr. 20, 01069 Dresden, Germany
6] Jan Wiezoreck, Trithor GmbH Heisenbergstr. 16, D - 53359 Rheinbach,
7] Archie Campbell, Tim Coombs, Univerysity of Cambridge From, Dept. of Engineering,
Trumpington St., Cambridge CB2 1PZ, UK

149

GAS SEPARATION
AND
LIQUEFACTION

151

CR08-59
LIQUID DISTRIBUTION FROM STRUCTURED PACKINGS AND
DISTRIBUTORS UNDER TILT AND MOTION RELEVANT TO
FLOATING CRYOGENIC AIR SEPARATION PLANTS
Kalbassi M.A.1, Waldie B.2, White V.1, Bell C.2
1

Air Products PLC, Surrey, UK

Offshore Processing Research Group, Herriot-Watt University, Edinburgh, UK

ABSTRACT
Adaptation of land-based gas-to-liquid processes to floating production plants is being
proposed as a means of recovering large offshore reserves of stranded gas. The
cryogenic air separation plants which supply oxygen to the process need to cope with the
tilt and motion conditions experienced on large production ships or barges. Results for
liquid distribution under tilt conditions are presented and this experimental data is used to
estimate separation efficiency and product oxygen composition under static, tilt and motion
conditions.
INTRODUCTION
There is considerable activity worldwide on the development of offshore floating
production systems for conversion of natural gas into liquid hydrocarbons. Some large
reserves of gas, so called stranded gas, are too far from land for a pipeline to be
economic. In that situation chemical conversion to liquids would reduce drastically the
volume of hydrocarbon to be moved and allow use of shuttle tankers. Large-scale plants
for such conversion are being used on land to produce high value liquid products including
clean diesel fuels for which there is increasing demand. Fischer Tropsch processes for
conversion of the intermediate syngas to liquids are a key part of these plants. The
cryogenic air separation plants which supply oxygen to the gasification step of the Fischer
Tropsch reactors need to be adapted to cope with the tilt and motion conditions experienced
on ships and barges proposed for offshore plants. The studies reported here are part of an
ongoing investigation into how tilt and motion affect liquid distribution and consequently
mass transfer performance in the packed columns used in cryogenic oxygen production and
other separation processes.
Previous reports on the effects of tilt and motion have been mostly on columns of up to 0.5
m diameter using water or aqueous solutions of similar surface tension. For example,
columns of 0.22 and 0.40 m diameter were used by Tanner et al [1] in a comparison of the
mass transfer performance of different packings subjected to tilt and motion. Results from
detailed studies of liquid distribution in the 0.40 metre column were applied [2] in a parallel
column technique to model the observed mass transfer performance. Columns for oxygen
production need to be larger in diameter, around 4 metres, and provide many more
theoretical stages than those used offshore for water de-aeration or regeneration of glycol.
Maldistribution has an increasingly severe effect on separation performance as the number
of theoretical stages increases [3]. In addition the extent of deflection or motion increases
with column height. Another factor in modelling cryogenic separation columns is the low
surface tension of the liquids, typically only 20% of that for water. Present studies are
therefore based on a larger, 1 m diameter, column and use of a lower surface tension liquid

153

as well as water. The results from these tests provide input for parallel column studies to
predict the effect of the measured distribution on distillation performance.
PACKING STUDIES: EXPERIMENTAL TECHNIQUES
Experiments were done on the 1 metre diameter by 5 metre high transparent column shown
in Figure 1. Metal structured packings were studied, these being used in cryogenic air
separation to minimise pressure drop. This paper reports experiments on 4 m (20 layers) of
Sulzer 500Y. Consecutive layers were rotated by 90. Metal foil wall wipers were fitted at
each layer. The column is mounted on pivots to allow tilting or near sinusoidal motion by a
mechanical drive with variable speed and amplitude. In the usual vertical orientation the
column was within 0.2 of true vertical. A pressurized distributor was used to avoid tilt or
motion affecting the initial liquid distribution. This was a ladder type giving 150 holes/m 2.
Detailed data on the distribution of liquid from the column were obtained with a multi cell
collection and online cell flow measuring system. Outgoing liquid passed first through 580
cells, mainly 35mm by 35mm in size, arranged in a 27 by 27 row array located immediately
under the packing. Liquid streams were then led through flexible PVC tubes (visible in
Figure 1) to a set of flow measurement cells. These contained wire electrodes for
measurement of rate of fill by conductance. Groups of sixteen collection cells are
connected in sequence to the sixteen measuring cells fitted with rapid acting fill and drain
valves. Rates of fill are measured via a fast response conductivity meter, multiplexer and
PC. Further details of a smaller version are available [2]. Mains water and a surfactant
solution with defoamer were used to study the effect of surface tension. Surface tension of
the solution, 34 mN/m, was significantly lower than that of water though still higher than
that of liquid oxygen and nitrogen. Safety, materials incompatibility and cost precluded
other lower surface tension liquids. Foam suppressant was essential to avoid foam effects
in the column and measuring system. Mean liquid fluxes were 2 and 4 l/m.s, typical of the
low liquid fluxes in some parts of an oxygen column.
RESULTS: VERTICAL AND TILT
The influence of tilt on liquid distribution can be shown most clearly by grouping the cells
into 27 rows running at 90o to the plane of tilt (Figure 2). This is also useful for subsequent
parallel column modelling. In Figures 3 and 4 mean flux per row is plotted against row
position across the column.
For the vertical column with 4 metres of 500Y, flux distribution across the column is more
even with the lower surface tension solution than with water Figure 3. With tilt of 4 the
surface tension effect is reversed, the flux distribution being worse for the low surface
tension liquid than for water (Figure 4). Both liquids suffer considerable maldistribution
with the uppermost wall region devoid of liquid. Near the lower wall, fluxes over twice the
column mean occur with water and surfactant.

154

R ow 27

R ow 1

R ow 14
T ilt andM o toi n P lane

Figure 1:
Motion

1 metre Diameter Column in

Figure 2:
of Plan Area for
F gi u re 2 .CChordal
ho rd a lSubd iv is oi n o fPSubdivision
lan A rea
fo rP a ra lleModelling
lC o ulm n M ode llni g
Parallel Column

3.75

4.75
4.5

3.5

4.25

3.25

3.75
3.5

2.75

3.25
FLUX l/m .sec

FLUX l/m .sec

2.5
2.25
2
1.75
1.5

3
2.75
2.5
2.25
2
1.75

1.25

1.5

1.25
Water Vertical

0.75

Surfactant Vertical

0.5

Water 4deg tilt 1

0.75

Water 4deg tilt 2

0.5

0.25

Surfactant 4deg tilt

0.25

0
0

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

ROW No.

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

ROW No.

Figure 3: Flux distributions for water and

surfactant from 4m of 500Y in vertical
orientation

Figure 4:
Flux distributions for water and
surfactant from 4m of 500Y at 4 tilt

The low surface tension of cryogenic liquids and some hydrocarbons is well recognised as
an important factor in packed column performance but there are few published reports on
the effect of surface tension on distributions from actual columns and none apparently at
the present scale. More data is now available on fundamental aspects of liquid flow on
single sheets of packing materials including the influence of surface tension and contact
angle but there remains the substantial task of applying that to predicting the distribution
from an actual column.
Another, more concise, way of comparing distributions is the fractional standard deviation
parameter, FSD. This though it is not of use in subsequent modelling. FSD was applied by
Reiss [4] to co-current flow distributions in packed columns and by Waldie [5] to compare
packings under tilt.

155

FSD =
n

[ ]

v 1 2 a
A
V

where,
v = flux in given chordal slab [kmol/m2]
V = mean flux over whole column [kmol/m2]

a = area of slab [m2],

A = total column area [m2]
n = number of slabs

FSD = 0 for a perfectly even distribution

The influences of surface tension and column orientation on distribution are summarised in
terms of FSD in Table 1. The trends shown by the graphs are confirmed quantitatively.
Table 1 also shows how FSD is affected by the degree of subdivision of the distribution
data. The finer the degree of subdivision the greater is the FSD.
Column
conditions*

500Y
3x9 rows

500Y
9x3 rows

500Y
27x1 row

4m/2/W/V

0.0164

0.0239

0.0314

4m/2/SA/V

0.00519

0.0103

0.0148

4m/2/W/4

0.1269

0.1450

0.1583

4m/2/SA/4

0.3164

0.3441

0.3555

Table 1: Liquid distributions in terms of flow distribution parameter FSD

*Code example: 4m/2/W/V means 4m packed height/ 2 l/ms / Water/ Vertical

RESULTS: MOTION
Fluxes to selected cells were measured continuously either in 16 cells connected to the
measuring cells through flexible tubes or in a single cell from which liquid fell freely into a
measuring cell which moved with the column. The latter gave the best resolution but could
only be applied to one collection cell at a time.
Volume/time plots for 16 cells spaced along a line near the middle of the base parallel to
the motion plane are shown in Figure 5. There is some evidence of cyclical variations in
slope, hence flowrate, in cells nearer the wall. The possibility that flexing of the tubes
contributed to these variations cannot be ruled out. Definite confirmation of visual
evidence of variations in flowrate is given in Figures 6 and 7 from the single cell
measurements. Fluctuations are more pronounced at the higher feed rate (4 l/s m mean
flux), probably due to a lower proportion of the liquid remaining attached to the surface of
the moving packing.

156

550

CUMULATIVE VOLUME COLLECTED IN CELL (ml)

500

CELL 1
CELL 4
CELL 7
CELL 10
CELL 13
CELL 16

450

CELL 2
CELL 5
CELL 8
CELL 11
CELL 14

CELL 3
CELL 6
CELL 9
CELL 12
CELL 15

400

350

300

250

200

150
0

10

15

20

25

30

35

40

TIME (secs)

Figure 5: Volume/Time Plots for 16 Cells over Single Cycle of 3 at 35 sec motion period for 500Y
with Surfactant

Figure 6: Time Dependant Flowrate in Cell B

over Single Cycle of 3/35 secs , 2 l/s m

Figure 7: Time Dependant Flowrate in Cell B

over Single Cycle of 3/35 secs , 4 l/s m

DISTRIBUTOR STUDIES
In the above packing study a pressurised distributor was used to ensure a constant initial
distribution pattern independent of tilt or motion. On an actual plant a gravity distributor
would avoid the need for a pump and thus be preferable if it was not affected too much by
tilt and motion. Conventional and proprietary designs have been studied on both the 1 m
column and on a larger computer controlled motion simulator in the Heriot Watt Offshore
Processing pilot plant. A thin slice model technique has been developed which allows
elements of gravity distributors up to 4.5 m diameter to be studied at accelerations similar
to those at the top of a 40-50 m column subjected to motions expected on a large
production ship. This has provided data for example on the relative significance of
hydrostatic and dynamic heads during motion and the development of proprietary designs
of gravity distributors [6].
INTERPRETATION OF PACKING DISTRIBUTION ON COLUMN
PERFORMANCE
The results of the tilt and motion studies can be used to determine the performance of a
cryogenic air separation distillation column system under these conditions. Here we
157

studied specifically the bottom section of packing of the low pressure (LP) column and the
top section of packing of the high pressure (HP) column. The results show the deterioration
in performance of the packing section due to the liquid maldistribution in the packing
caused by the motion. This was carried out using Aspen Plus to simulate the distillation.
First, the nominal cases were simulated. Two sections were to be studied. The first, the
bottom of the LP column, is shown in Figure 8. The L/V in the packing is 1.4 and the feed
composition is 0.95 O2 and 0.05 Ar, product is pure Oxygen. The second section to be
studied, the top of the HP column, is shown in Figure 9. Here the L/V is 0.6 and the feed,
stream 3, is air and the product is pure Nitrogen. In both cases 20 layers of Sulzer 500Y
were used.
To study the effects of motion, this flowsheet was adapted by creating 8 parallel columns.
The feeds to the column, liquid and vapour, were split between these 8 columns, the split
fractions depending upon the results of the experimental motion studies. In the case of the
LP column, the liquid from the bottom of the packing was then mixed into the reboiler. For
the HP column, the vapour is mixed into the condenser. The 8 parallel column arrangement
is shown for the LP column in the Aspen flowsheet in Figure 10.
.

Figure 8: Aspen Flowsheet of LP Packing

Simulation

Figure 9: Aspen Flowsheet of HP Packing

Simulation

158

Figure 10: Parallel Columns used for motion simulation

Only the results from experiments in which surfactant was used are analysed using parallel
column analysis since these experiments more closely match the expected behaviour of a
cryogenic liquid.
For the stationary column cases, where data is available for all 27 rows, shown in Figures 3
and 4, neighbouring cells were grouped and averaged to give 8 average cells across the
diameter that were used to determine the liquid flows into the eight parallel columns. The
vapour splits were then determined by the areas of the chords over which the averaging had
taken place.
For the results with the column in motion 16 cells were used to collect the data in the
experiment and these cells were almost across a diameter, Figure 5. Therefore, the eight
flows were taken by averaging Cell No. 1 & 16, Cell No. 2 & 15, etc. This gives 8 flows
and dividing each by the total gives the liquid split fractions required.
Using the above calculated split fractions, the simulation was performed to determine the
performance. In order to determine the effective efficiency of the packing under these
conditions, we return to the Aspen flowsheet with the single column and performed a
simulation to determine the section efficiency required to obtain the separation achieved
with the motion results applied to the parallel columns. These results can be found in
Table 2.
The section efficiency of the stationary packing is also surprisingly low. Note that the two
column sections investigated are not pinched. The efficiency of the LP column simulations
works out greater than that of the HP columns. Note that stationary data shows packing
inherent tendency to maldistribute and thus reduced packing efficiencies shown in Table 2.
Packing
500 Y
500 Y

Column
LP
HP

Ideal
100%
100%

Stationary
71.6 %
45.4 %

Motion
22.1 %
19.1 %

4 degree tilt
12.4 %
11.1 %

Table 2: Separation Efficiency

CONCLUSIONS
Reducing liquid surface tension to about half that of water reduces maldistribution from
500Y structured packings in a vertical column but increases it when the column is tilted by
4.
Tilt of 4 causes significant maldistribution from 500Y structured packings at a mean
column flux of 2 l/sm. Other studies though on the same column have shown that these
packings suffer less maldistribution than random packings under tilt.
On the very large ships now proposed for gas to liquids schemes tilt is not expected to
exceed 1 to perhaps 2. Present data therefore indicates that it will still have to be taken
into account in column design especially as there is a further decrease in surface tension
with cryogenic liquid.
With the column moving 3 at 35sec period, cyclical variations in local flux occur, at least
near the wall. These are more pronounced at a higher mean column flux.
Even though 20 layers of Sulzer 500Y has been shown to give low performance in a
shipboard scenario, the techniques reported in this paper have been demonstrated to provide
a method to compare the performance of different proprietary types of packing [7] or
159

distributors in a shipboard scenario and relate this to the expected performance of the
distillation column in which the packing is employed.
REFERENCES
1.
2.
3.
4.
5.
6.
7.

Tanner, R.K., Baker, S.A. and Waldie, B., 1992, Proc.Distillation & Absorption
92. I.Chem.E. Symp. Series No 128, B.111- B.118.
Tanner, R.K., Baker, S.A., Millar, M.K. and Waldie, B., 1996,
Trans.I.Chem.E.74A.177-182
Billingham, J.F. and Lockett, M.J., Trans I ChemE 2002, 80A , 373-382
Reiss, L.P., 1967, Ind.Eng.Chem.Proc.Des.Dev. 6, 486
Waldie, B., 2002, Proc.Gas Proc.Assn Europe Annual Conference, Rome
Kalbassi M.A and Zone,I.R. U S Patent 6,907,751 B2 ,June 21 2005
Armstrong P.A, Kalbassi, M.A, Miller D, US Patent 5, 984,282, Nov 16 1999

160

CR08-43
COMPLEX SEPARATION OF MULTICOMPONENT FLOWS TO
EXTRACT INDUSTRIAL AND INERT GASES
Bondarenko V. L.1, Losyakov N. P.2, Simonenko O. Yu.2
1

Moscow Bauman State Technical University, 5, 2-nd Baumanskaya Str.,

107005, Moscow, Russia
2
Iceblick, Ltd., 29, Pastera Str., 65026, Odessa, Ukraine
ABSTRACT

The paper analyzes the composition of the waste flows, which appear in ammonia
production. The ways of the inflow of the inert gases into ammonia synthesis circuit have
been shown. The potential volumes of helium, neon, argon, krypton and xenon, which can
be extracted in chemical industry, have been calculated. The preferable sequence of the
multicomponent mixtures processing has been reasoned. The conditions resulting in
acceptable degrees of extraction and the given quality or rare gases and accompaniments
have been detected.
INTRODUCTION
The rare gases volumes and applications are continually increasing. A significant part of
them is extracted from the waste products of the oxygen production at the steel mills. In
spite of the developed metallurgical industry in Ukraine and Russia, the raw sources of rare
gases have been practically exhausted. Therefore, there is an acute task of finding
alternative sources of raw mixtures. The chemical industry plants can be considered as
such. In the technological cycle of ammonia production there are considerable quantities of
waste gas containing all the range of inert gases. Separating these flows into individual
components is possible by means of sequential separation at cryogenic temperatures.
COMPLEX TECHNOLOGY OF WASTE FLOW SEPARATION
To get hydrogen, needed in ammonia production the process of natural gas conversion is
used. This reaction requires huge volumes of oxygen, which is the basis of atmospheric air.
Together with the air flow considerable amounts of inert admixtures get into the synthesis
device. Because they do not take part in the reaction, they are accumulated in the NH3
synthesis circuit and they are discharged into the atmosphere as a waste flow. This mixture,
besides the rare gases, contains valuable products: methane and components of singas
nitrogen and hydrogen, as well as the product of synthesis ammonia, which all, as by
products of the separation of rare gases can be returned back to the synthesis process.
Typical composition of by-product gas mixtures is shown in Table 1 on the example of two
biggest enterprises of Ukraine. The presence of several substances in the waste flow allows,
prima facie, many variants of separation. However, considering the physical properties of
separate components and a number of technological limitations, a certain sequence of
separation processes can be observed in the industry. Preferable variants of waste flow
separation are shown in Figure 1. The width of lines, characterizing separate products on
the diagram, corresponds to their content in the mixture. All the complex of technologies
can conventionally be split into several processes: I purification from ammonia; II
preliminary cryogenic separation of the mixture; III recovery of high-boiling products;
IV hydrogen-helium mixture processing.

161

ENTERPRISE

NH3

Azot enterprise,
Severodonetsk, Ukraine
Priportovy Zavod,
Odessa, Ukraine

Xe+Kr

1,7

CH4

Ar

N2

Ne

13,0

5,3

20,7

8,6

5,6

19

<0,001
2

0,0
1
0,0
1

H2

He

59

0,3

64,4

0,4

Table 1: Volume content of the components in the waste flows, %

N itr o g e n

A rg o n
M e th a n e
A m m o n ia

He
H

CH

Stage
of preliminary
separation (SPS)

H yd ro g e n

Contour
of extraction NH3

H e liu m

He

IV -
PURIFYING
from N2

II

B lo c k
o f c o n v e r s io n
o f h y d ro g e n
fr a c tio n

H E L IU M
CO NCENTRATE

IV -b
HYDRO G EN

N IT R O G E N
N IT R O G E N
N

III

r
A M M O N IA

B lo c k
o f h e liu m
p u r ific a tio n

SEPARATION

IV -c

H IG H P U R IT Y
H E L IU M

CH

B lo c k o f e x tr a c tio n o f
h ig h b o ilin g p r o d u c ts
4

M ETHANE
(K r+ X e )

ARG O N

Figure 1: The preferable sequence of complex waste flow processing

The necessity of the preliminary mixture separation on stage II into light and heavy
fractions allows relieving circuit III from high hydrogen consumption, making 2/3 from the
end product volume. Besides helium, hydrogen also contains traces of neon. The need of
recovery of the trace content of Ne and He in the source flow on stage II makes it necessary
to use the rectification column. Using simpler means as preliminary separators (phase or
reflux condensers) is unacceptable. This leads to the light inert gases dissolving in the
liquid nitrogen-methane fraction, which is equivalent to their loss. The other, not less
important function of column II, is the partial decrease of the concentration of high-boiling
admixtures (N2) in its gas fraction, consisting mainly of nitrogen. As follows from Figure 2a, the effective means of hydrogen-helium flow enrichment is lowering the temperature in
the column II condenser (Figure 3). It is rational to additionally cool the mixture ( 2-N2He) and lower the nitrogen concentration in it in a separate reflux condenser RC1. The
autonomy of this device allows applying alternative variants of cooling. As such, nitrogen
boiling at a reduced pressure ( = 0,02 MPa, = 66 ) or the upstream of gaseous
hydrogen ( = 24 ) after separation in the circuit (VI-b, Figure 1) can be used.
As it is shown in the diagram (Figure 2-a), due to reducing the phase equilibrium
temperature in RC1 from 83 to 64 K it is possible to condense and return to the column up
to 70% of high-boiling components (mainly nitrogen). This leads to considerable reduction
of loading on the final purification stage IV-a. Further cooling and enrichment of the flow
during the continuous operation is difficult because of the danger of freezing of N 2,
contained by the hydrogen fraction at 63,15 . Total purification of hydrogen-helium
162

mixture is achieved in the devices with periodically working process. This can be realized
by means of adsorption at = 64 or freezing at = 40 . The research showed that the
second variant is preferable because it allows achieving the same result at 45% less energy
consumption. Reduction of operating costs of the cryogenic purification of the H 2-He
mixture from nitrogen is conditioned by the narrow corridor of working temperatures in
the freezer. Unlike the adsorbers, in this device, it is enough to raise the temperature from
40 to 65 for the admixtures disposal.

Figure 2: Phase equilibrium isotherms of the systems 2-N2 () and -2 (b) in the vapor phase.
1V-2V-3 - enrichment and purification of 2-fraction
in the additional reflux condenser RC1 and the freezer Fr (Figure 3);
6V-12-12 - helium concentration process under different phase equilibrium conditions

Creating the circuit IV-b for H 2-He separation was partially based on the same physical
principles and technological methods that are inherent to the process of hydrogen fraction
purification from N2 considered above (circuit IV-, Figure 1). In spite of different
temperature levels (64 for the mixture 2-N2 and 1516 for -2), these problems
have a common solution creating favorable phase equilibrium conditions outside the
respective rectification columns. This fact is illustrated by the phase equilibrium isotherms
on Figure 2-b and it is proved by the closeness of schematic solutions of the circuits IV-a
and IV-b (Figure 3).
Raw helium flow in point 12 is two order smaller than at the entry into HC (point 3).
This allows processing of efficient enrichment and final helium purification by use of
helium refrigerators.
The specifics of the circuit III, in which pure argon and Kr-Xe concentrate from the
methane basis are obtained, is introducing the additional column KC into the typical
scheme (Figure 4). As the preliminary research showed, krypton enrichment in the cube of
column MC cannot achieved within this step even in the case of liquid phase extraction on
the L line. It is explained by the fact that the 4-Kr system has a rather low relative
volatility <1,6. Such coefficient is about 20 times smaller than the one of the extensivelly
studied Kr-2 mixture, from which krypton-xenon concentrate is extracted in the airseparation plants.
The preliminary enrichment of krypton concentrate can also be achieved by the sorption
method. This method, besides the concentration, also allows substituting methane for
nitrogen. The reduction of 4 content from 9599% to several percent simplifies the
further nitrogen-krypton flow processing and obtaining Kr and Xe in pure form.

163

Figure 3: The scheme of H2-He concentrate extraction and the flows parameters in characteristic
points: SPS - system of preliminary separation; RC1 - N 2 reflux condenser; F - N 2 freezer (one of the
two sections is shown); HC - hydrogen column; RC2 - H2 reflux condenser; HC - helium column; HR
- helium refrigerator; NA - neon adsorber (the circuits markings correspond to Figure )
IV - C
IV - b

IV - a

12
9

RC2

H E L IU M
CO N CENTR ATE

HR

Fr
6L
10
2L

2V

H YDR O G EN

N
1L

II

NA

RC1

11

6V

HC
2

1V

4
IN IT IA L
M IX T U R E

SPS

,
MPa

1
1L
2V
2L
3
4
5
6V
6L
7
8
9
10
11
12

3,5
3,5
3,5
3,5
3,5
3,5
1,0
1,0
1,0
1,0
0,25
0,25
0,25
0,25
0,25

,
K
83
83
64
64
40
33
31
26
26
31
24
24
60
78
23

Volume content, %
He
0,4
0,44
0,45
0,45
0,45
45
1,7
0,001
0,001
0,001
0,001
0,001
70

164

H2
88,6
10
97,7
9
99,55
99,55
99,55
55
98,3
99,999
99,999
99,999
99,999
99,999
30

N2
11,0
90
1,9
91
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001
<0,0001

 2+

II

IV - a

N IT R O G E N

III

M ETH AN E
N

M C

I n itia l

m ix tu r e

AC

SPS
KC

ARG O N
III

M E T H A N E + (K r+ X e )

Figure 4: The preferable sequence of waste flow complex processing.

SPS - system of preliminary separation; MC - methane column; AC - argon column; KC - krypton
column

CONCLUSION
1.

The potential of chemical industry producing helium and argon are

commensurable with the productivity of metallurgic industry oxygen plants.

2.

The waste flows complex processing is possible only on basis of cryogenic

separation methods.

3.

There is a rather definite technological sequence of standard multi-component

mixture processing. It includes ammonia redemption, gaseous fraction 2- separation
in a column; purification of this fraction from nitrogen and further separation into
hydrogen and helium.

4.

Liquid fraction processing on the preliminary separation stage (II) takes place at
least in two subsequent columns and is accompanied by the output of argon, N2 and 4
product flows.

5.
6.

The methane fraction is the raw product for krypton and xenon extraction.
It is rational to extract rare gases in pure form in separate devices, not connected
with the waste flow separation complex.
REFERENCES

1. Bondarenko V. L., Simonenko Yu. M., The method of xenon separation, (Variants) and
installation for its realization. Patent of Russia 2134387. B.I.-22 (1999).
2. Arkharov A. M., Bondarenko V. L., Losyakov N. P. and all, A unit for the extraction of
the krypton-xenon mixture from off-gases at the ammonia production. Proc. 6 Int. Conf.
Cryogenics2000, Praha (2000) 122-125.

165

CR08-41
SOLUBILITY OF PROPANE AND ETHANE IN LIQUID OXYGEN
Houssin-Agbomson D.1, Arpentinier P.1, Delcorso F.1, Coquelet C.2, Richon D.2
1

Centre de Recherche Claude-Delorme Air Liquide, Jouy-en-Josas, France

2
Laboratoire CEP/TEP Mines Paris, Fontainebleau, France
ABSTRACT

Industry is large consumer of air gases for many and varied applications. Among the
various processes of separation of air components, the most employed remains fractional
distillation at low temperatures. The presence of pollutants like hydrocarbons in the
feed atmospheric air of air distillation units (ASU) can be at the origin of drastic
dysfunctions. That is the reason why a more accurate knowledge of the solubility of
hydrocarbons in liquid oxygen and of the thermodynamic behaviour of these flammable
systems under process operating conditions (from 93 to 153 K) would improve both
evaluation and control of the risks specific to ASU and their performances.
INTRODUCTION
Air gases, mainly oxygen, nitrogen, argon are essential for metallurgy, chemistry,
petrochemistry, refining, energy, electronics, health The fractional distillation which
operates at cryogenic conditions is the most employed process. The feed atmospheric air
must be cleaned, before liquefaction, by removing all components being potentially
obstructive at low temperatures. In particular it is necessary to take special care of carbon
dioxide and water, and of secondary pollutants that are either natural or produced by the
various anthropic activities (industry, heating, road traffic), like hydrocarbons (ethane,
propane or ethylene), or ozone and nitrogen protoxide. The presence of pollutants in the
feed of air distillation units can be at the origin of their drastic dysfunctions. Specifically,
hydrocarbons can form highly flammable mixtures with oxygen [1]. The risk is controlled
today through several means, which allow operating air distillation units in effective and
safe way. However these aspects can be improved by a better knowledge of physical
properties of air pollutants. Concerning hydrocarbon-oxygen binary systems scientific
literature presents only few data, probably because of the danger with handling of such
mixtures in laboratories. In order to be able to build the required database, Air Liquide and
CEP/TEP Laboratory have designed, built and set up new experimental equipment allowing
to work under safe conditions. The study of the solubility of propane in liquid oxygen was
the really interesting first subject of investigation using new installation. Nevertheless, to
validate equipment and procedure before working on propane-oxygen system,
investigations began with the study of the non hazardous mixture: propane-nitrogen. And to
improve our knowledge on hydrocarbon-oxygen systems behaviour, the experiments
have been extended with the study of ethane-oxygen binary system.
1. INDUSTRIAL CONTEXT
Main steps of air distillation process are presented in the Figure 1 (A): compression of air,
elimination of pollutants, generation and transfer of cold, compression of the products [2].
The main element of the process is the double fractionating column (see Figure 1 (B)),
composed by the medium pressure column (MP-column: 0.5 to 0.6 MPa) and the low
pressure column (LP-column: about 0.15 MPa). Compressed air is fed into the medium
pressure column.
167

The two columns are connected by a reboiler-condenser which ensures at the same time the
heating of the LP-column, the vaporization of oxygen and the backward flow of each
column by condensation of nitrogen. In this reboiler-condenser, heated and vaporized liquid
oxygen (by nitrogen, which condenses) contains dissolved impurities: nitrogen protoxide
(N2O), carbon dioxide (CO2), hydrocarbons (C2, C3).

(A)

(B)

Figure 1: (A) Unit operations involved in air distillation process.

(B) Conventional double column apparatus for air distillation [1].

However, taking into account the extent of the air flows, impurities, even if they are present
only at very low amount in the air feed, can under particular conditions, in spite of the
careful air cleaning steps, accumulate over long period of time in liquid oxygen to reach
considerable contents. According to the operating conditions and technologies of
vaporization used, they will settle then in a solid state or will form a second liquid phase (in
addition to the oxygen rich phase). For industry, it is of primary importance to control the
formation of these phases, solid or liquid, rich in impurities in order to maintain, on one
hand, transfer efficiency by avoiding the clogging of the reboiler-condenser and, on the
other hand a satisfactory safety level by minimizing ignition risk of hydrocarbons with
oxygen. That is why, air processors have developed equipment (liquid oxygen filters),
materials (new adsorbents for the air cleaning step) and strategies and procedures for check,
analysis and follow-up of the impurities behaviour from feed air to the reboiler-condenser.
The principal quantities whose knowledge is fundamental to control the formation of
undesirable phases are: the solubility of the impurity in liquid oxygen, the partial pressure
of the impurity in gaseous oxygen and the molar fraction of vaporized liquid oxygen.
Thus the solubility of hydrocarbons in liquid oxygen appears as a key variable. In fact its
knowledge as a function of pressure and temperature will allow to predict the formation (or
not) of a hydrocarbon rich second liquid or solid phase. Among hydrocarbons present in
atmospheric air, propane is potentially one of the most critical to control taken into account
the efficiency of the front-hand purification step. Unfortunately, up to now, its solubility
was badly-known under industrial operating conditions.
2. STATE OF THE ART
The study of the literature concerning the solubility of hydrocarbons in liquid oxygen
shows that only few references are available on the subject: Karwat (1958) [3] and
McKinley and Wang (1960) [4] for propane; Tsin (1940) [5], Cox and de Vries (1950) [6],
McKinley and Wang [4], Amamchian et al. (1973) [7] and Bulanin (1973) [8] for ethylene;
Cox and de Vries [6], Karwat [3] and McKinley and Wang [4] for ethane. Furthermore,
168

experimental procedures are not sufficiently detailed in these references to make it possible
to evaluate uncertainties on the measured values. In particular, few values of solubility in
liquid oxygen are published at 90 K: 10 000 ppm [3] and 50 000 ppm [4] for propane, and
78 000 ppm [6], 128 000 ppm [3] and 215 000 ppm [4] for ethane. Moreover, temperature
ranges are relatively limited (from 77 to 90 K) and do not cover the exploitation field of air
distillation units. Experimental techniques used for the determination of the thermodynamic
equilibria properties, are generally classified according to the method of equilibrium
creation (static methods, dynamic methods) and according to the technique of
determination of the compositions (synthetic methods, analytic methods) [9].
Measurements of solubility are based on static-synthetic or static-analytic methods.
3. EXPERIMENTAL ASPECTS
3.1 Experimental method
The technique retained for this study is a static-analytic method with sampling of phases
by ROLSI samplers (Rapid On-Line Sampler-Injector) [10], followed by gas
chromatography analyses. This technique is based on a method described by Laugier and
Richon [11]. The solute is introduced into the equilibrium cell and then it is diluted with the
solvent. Once the compounds loaded, the cell is comparable to a batch reactor. Equilibrium
is reached in a static way.
3.2 Description of the equipment
A 12-cm3 Hastelloy C276 cell is fixed inside a cryostat partially filled with liquid nitrogen.
The cryostat used on this apparatus is a 55-dm 3 double envelope vessel (LAir Liquide
GT55 model). The stability of the temperature, 0.05 K, is achieved using a heating
resistance cable rolled up around a brass housing containing the equilibrium cell and
connected to a proportional-integral-differential (PID) thermal regulator (West Mini 6100
model). The rounded heating resistance acts as the hot source and the vapour of liquid
nitrogen as the cold one. Two ROLSI pneumatic samplers fitted on the top of the
equilibrium cell (one for the liquid phase, the other for the vapour phase) and connected to
a gas chromatograph (Varian 3800 model), allow direct injection of liquid and vapour
phase samples into the carrier gas circuit of the gas chromatograph. Two pressure
transducers measure the total pressure inside the equilibrium cell: a 0-1 MPa transducer
(Drck PTX610) for low pressures and a 0-10 MPa transducer (Drck PTX611) for higher
ones. After calibration, accuracies of pressure measurements are estimated to be better than
0.07 kPa for the low pressure transducer and 0.18 kPa for the other. Temperatures are
measured by two four-wire 100- platinum probes introduced inside wells managed in the
walls of the equilibrium cell body (one at the top, another one at the bottom). Uncertainties
on temperatures are lower than 0.02 K on the (100-140) K range. Once the system has
reached equilibrium (i.e. P and T are constant), the analysis of the samples taken by the
ROLSI samplers is carried out thanks to a Varian gas chromatograph (model 3800)
equipped with two types of detectors fitted in series: a thermal conductivity detector (TCD)
and a flame ionisation detector (FID). The TCD is used for the detection of oxygen,
nitrogen and high quantities of propane, while the FID, more sensitive, allows detecting the
very small quantities of propane. TCD and FID were repeatedly calibrated by introducing
known amounts of each pure compound through a syringe into the injector of the gas
chromatograph. Chromatograph calibrations realized for this type of mixtures lead to
relative uncertainties around 1 % for each component.

169

3.3 Equipment and procedure validation

In order to validate our experimental equipment, nitrogen-oxygen system was studied at
110 K and the experimental results obtained have been compared to those of Baba-Ahmed
et al. [12]. Baba-Ahmed used a - thermodynamic approach and adjusted the binary
interaction parameter kij of the Soave-Redlich-Kwong equation of state [13, 14] on its
experimental values (temperature range: 100 to 123 K) measured with similar equipment
and procedure, but in an equilibrium cell of a volume 3.5 times larger than ours. The
standard mixing rules [14], and the Mathias-Copeman alpha function [15] were used for
this calculation. This adjustment (kij = -1.58.10-2), using an objective function based on
the total pressure and the nitrogen vapour phase composition, led to relative average
deviations on these variables of respectively 0.6 % and 1.5 %. Average deviations
obtained between our measured values and Baba-Ahmed computed values [12] concerning
the vapour phase composition and the total pressure are about 0.5 %. These deviations,
lower than the uncertainty of Baba-Ahmed adjustment, validate the experimental protocol
and show that our equilibrium cell, in spite of its low volume, is still well adapted to the
study of thermodynamic equilibria under cryogenic conditions.
4. RESULTS AND DISCUSSION
In this section, validation of the procedure on a similar but not flammable system is
demonstrated by measurements performed on liquid compositions of the propane-nitrogen
system under various experimental conditions (propane low and high composition areas).
Then the study of propane solubility in liquid oxygen is presented with that of the ethaneoxygen binary system.
4.1 Propane-nitrogen binary system
Experimental values obtained between 109.98 and 125.63 K for the propane-nitrogen
system at thermodynamic equilibrium allowed, according to a - approach, to adjust
parameters of a thermodynamic model based on the Peng-Robinson equation of state (PREoS) [16] using the Mathias-Copeman alpha function and the Huron-Vidal mixing rules
[17] coupled with NRTL activity coefficient model [18].

1.6

L I + L II

L II

LI

1.2

L II + V
V

0.8

LI + V

0.4
0.0
0.00

1.470
Total pressure (MPa)

Total pressure (MPa)

2.0

1.465
L II

1.460

L I + L II

1.455

L II + V
V

1.450
LI + V

1.445
0.991

0.20

0.40

0.60

0.80

1.00

0.993

0.995

0.997

0.999

x N2 , y N2 (molar fraction)

x N2 , y N2 (molar fraction)

(A)

(B)

Figure 2: (A) Pressure-composition diagram for propane-nitrogen system at 109.98 K. (B) Zoom on
diagram close to nitrogen vapour pressure. (): experimental data from this work; solid lines:
calculated VLE on experimental data from this work and LLE prediction with PR-EoS.

170

Thanks to this work, four (P, x)-isothermal curves have been studied and modelling results
have been compared to scientific literature (Cheung and Wang (1964) [19], Poon and Lu
(1974) [20], Kremer and Knapp (1983) [21], Llave et al. (1985) [22]): good agreement is
observed with Kremer and Knapp data while sets of data present systematic deviations with
Poon and Lu data and Llave et al. ones that are consequently judged as suspicious [23].
Calculated complete phase diagrams (see Figure 2) show that they are of type III according
to the van Konynenburg and Scott classification [24]: this system presents vapour-liquidliquid phase separation.
4.2 Propane-oxygen binary system
Before determining solubility values with the apparatus previously described, another
experimental device (fairly simple) has been set up in order to visualize propane-oxygen
system at thermodynamic equilibrium (see Figure 3) and have more information on its
behaviour in cryogenic conditions.

Vapour phase

Propane rich liquid

phase

Oxygen rich liquid

phase

Figure 3: Photography of the vapour-liquid-liquid equilibrium

for propane-oxygen system in cryogenic conditions at atmospheric pressure.

Thanks to this first experiment, some assumptions have been made and finally experimental
protocol was modified in several times in order to better control the propane loading and
succeed to determine propane solubility. These modifications, so simple in fact, have
significantly contributed to improve the operability of the apparatus, and both the reliability
and the repeatability of measurements.
0.06

0.0570
0.0500

xC3 (molar fraction)

0.05
0.04
0.03

0.0224

0.02
0.0098

0.01
0.00
85

90

95

100

105

110

Temperature (K)

115

120

125

Figure 4: Propane solubility in liquid oxygen as a function of temperature.

() this work, () Karwat [3], () McKinley and Wang [4].

171

Data measurements have been performed in propane high and low composition conditions
and in the vapour-liquid-liquid area allowing the determination of the propane solubility
in the oxygen rich phase. Consequently, solubilities of propane in liquid oxygen have been
determined at 110.22 and 120.13 K. These experimental values are presented in Figure 4,
for comparison to literature data. At 110.22 K, the value of propane solubility in liquid
oxygen is xC3 = 0.0224 0.009 for a total pressure of 0.536 0.001 MPa. At 120.13 K, the
solubility is higher and reaches a xC3 value of 0.0570 0.009 for a total pressure of
0.979 0.001 MPa. Literature data presented in Figure 4 (Karwat [3], McKinley and Wang
[4]) are the only available ones for the propane-oxygen binary system up to now. Figure 4
shows that the solubility value measured by McKinley at 90 K is not consistent with values
of this work obtained at 110.22 and 120.13 K and not with that of Karwat at 90 K. Data
from scientific literature were obtained with experimental techniques different from that
developed in CEP/TEP Laboratory. McKinley and Wang [4], in particular, used a synthetic
method coupled with a visual criterion of solubility detection. Unfortunately experimental
technique and detection criterion of Karwat [3] are not clearly described. A detailed reading
of these publications does not make it possible to conclude as for the repeatability and the
accuracy of their corresponding measurements. It seems that the precautions taken in this
work allow approaching solubility values with a more acceptable level of accuracy.

L +L

Total pressure

L +V

II

II

II

L +V

x O2 , y O2

Figure 5: Pressure-composition diagram for propane-oxygen system at 110.22 and 120.13 K.

Solid lines: results of the modelling with adjustment on our experimental data;
LI, LII: respectively propane rich and lean liquid phases; V: vapour phase.

Based on our experimental data a modelling work was performed and complete phase
diagrams have been calculated at two temperatures, as presented in Figure 5, with a
particular adjusted thermodynamic model. These diagrams, as for propane-nitrogen system,
are of type III according to the van Konynenburg and Scott classification [24].
4.3 Ethane-oxygen binary system
Data have been carried out at several temperatures and phase diagrams of ethane-oxygen
system were calculated thanks to modelling after adjustment on these isothermal data. Only
results obtained with ethane at 112.1 K are presented herein. At 112.1 K, ethane is
completely soluble in liquid oxygen as shown on the pressure-composition diagram
calculated displayed on Figure 6: no vapour-liquid-liquid equilibrium is observed,
whatever liquid phase composition. Thus the phase diagram of ethane-oxygen system
seems of type I according to the van Konynenburg and Scott classification [24].
172

Total pressure

L
V

L+V

x O2 , y O2

Figure 6: Pressure-composition diagram for ethane-oxygen system at 112.1 K. Experimental data

() and result of the modelling with adjustment on our experimental data (solid lines);
L, V: respectively liquid and vapour phases.

Experiments and modelling are found in good agreement concerning the thermodynamic
behaviour of ethane-oxygen system. Now it is important to emphasize that solubility values
presented in scientific literature by authors at 90 K and at lower temperature, certainly are
solubilities of solid ethane in liquid oxygen (note that melting temperature of pure ethane is
about 90.4 K).
CONCLUSIONS
A new equipment was developed to study hydrocarbon-oxygen flammable systems and
in particular propane-oxygen system. Preliminary studies, which were carried out on non
dangerous systems for which solubility data are known or partially known (nitrogenoxygen and propane-nitrogen systems), allowed to validate equipment and procedure.
Initially, composition measurements of the liquid phase, realized in propane low
composition area, were dispersed according to propane loading rate. This was probably due
to the properties of the two coexisting liquid phases (density, viscosity, interfacial tension,
wettability) which are more influent with low quantities loaded in the equilibrium cell.
Thus a great part of the study consisted in developing a loading protocol making possible
the control of the quantity of propane injected into the cell in order to measure the propane
solubility in the liquid phase with an acceptable level of accuracy. Propane solubilities in
oxygen (mole fractions) determined during this study are: 0.0224 at 110.22 K and 0.0570 at
120.13 K. In comparison, the solubility value measured by McKinley and Wang at 90 K [4]
is not consistent with values obtained in this work; this could be explained by the difference
of experimental techniques or by the state of propane either a liquid or a solid close to pure
propane melting point. However concerning Karwat value at 90 K [3], the agreement with
our values can be considered as acceptable. Concerning ethane-oxygen system, experiments
and modelling showed that ethane is completely soluble in liquid oxygen at 112.1 K.
Other data have been performed on few hydrocarbon-oxygen binary systems and will be
presented in further communications (to be published). This work brings new information
on thermodynamic behaviour of such systems (properties and phase envelopes), which will
be used for accumulation phenomena interpretation highlighted in high capacity air
distillation units.
173

REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.

Arpentinier, P., Cavani, F. and Trifir, F., The Technology of Catalytic Oxidations, Tome 2:
Safety Aspects, Editions Technip, Paris, France (2001) 471-670
Foerg, W., Refrigeration Science and Technology Proceedings, Munich, Germany (1996) 1-12
Karwat, E., Some Aspects of Hydrocarbons in Air Separation Plants, Chem. Eng. Prog. (1958)
54(10) 96-101
McKinley, C. and Wang, E. S. J., Hydrocarbon-Oxygen Systems Solubility, Adv. Cryog.
Progress. (1960) 53 11-25
Tsin, N. M., Solubility of Ethylene and Propylene in Liquid Nitrogen and Liquid Oxygen, Zh.
Fiz. Khim. (1940) 14(3) 418-421
Cox, A. L. and de Vries, T., The Solubility of Solid Ethane, Ethylene, and Propylene in Liquid
Nitrogen and Oxygen, J. Phys. & Colloid. Chem. (1950) 54 665-670
Amamchian, R. G., Bertsev, V. V. and Bulanin, M. O., Infrared Spectral Analysis of Cryogenic
Solutions, Zavodskaya Lab 4 (1973) 39(4) 432-434
Bulanin, M. O., Infrared Spectroscopy in Liquified Gases, J. Mol. Struct. (1973) 19 59-79
Coquelet, C., Etude des Fluides Frigorignes. Mesures et Modlisations, PhD Thesis ENSMP,
France (2003)
Guilbot, P., Valtz, A., Legendre, H. and Richon, D., Rapid On-Line Sampler-Injector. A Reliable
Tool for HT-HP Sampling and On-Line Analysis, Analusis (2000) 28 426-431
Laugier, S. and Richon, D., New Apparatus to Perform Fast Determinations of Mixture VaporLiquid Equilibria up to 10 MPa and 423 K, Rev. Sci. Instrum. (1986) 57 469-472
Baba-Ahmed, A., Guilbot, P. and Richon, D., New Equipment Using a Static Analytic Method
for the Study of Vapour-Liquid Equilibria at Temperatures down to 77 K, Fluid Phase Equilib.
(1999) 166(2) 225-236
Redlich, O. and Kwong, J. N. S., On the Thermodynamics of Solutions. V. An Equation of State.
Fugacities of Gaseous Solutions, Chem. Rev. (1949) 44 233-244
Soave, G., Equilibrium Constants for Modified Redlich-Kwong Equation of State, Chem. Eng.
Sci. (1972) 4 1197-1203
Mathias, P. M. and Copeman, T. W., Extension of the Peng-Robinson Equation of State to
Complex Mixtures: Evaluation of the Various Forms of the Local Composition Concept, Fluid
Phase Equilib. (1983) 13 91-108
Peng, D. Y. and Robinson, D. B., A New Two-Constant Equation of State, Ind. Eng. Chem. Sci.
(1976) 15 59-64
Huron, M. J. and Vidal, J., New Mixing Rules in Simple Equations of State for Representing
Vapour-Liquid Equilibria of Strongly Non Ideal Mixtures, Fluid Phase Equilib. (1979) 3 255271
Renon, H. and Prausnitz, J. M., Local Composition in Thermodynamic Excess Function for
Liquid Mixtures, AIChE J. (1968) 14 135-144
Cheung, H. and Wang, D. I. J., Solubility of Volatile Gases in Hydrocarbon Solvents at
Cryogenic Temperatures, Ind. Eng. Chem. Fundam. (1964) 3 355-361
Poon, D. P. L. and Lu, B. C. Y., Phase Equilibria for Systems Containing Nitrogen, Methane,
and Propane, Adv. Cryog. Eng. (1974) 19 292-299
Kremer, H. and Knapp, H., Three-Phase Conditions during Gas Processing are Predictable,
Hydrocarb. Process. (1983) 62 79-83
Llave, F. M., Luks, K. D. and Kohn, J. P., Three-Phase Liquid-Liquid-Vapor Equilibria in the
Binary Systems Nitrogen + Ethane and Nitrogen + Propane, J. Chem. Eng. Data (1985) 30(4)
435-438
Houssin-Agbomson, D., Coquelet, C., Richon, D., Delcorso, F. and Arpentinier, P., Solubility of
Hydrocarbons in Liquid Oxygen, 2007 AIChE annual meeting, Salt-Lake-City, Utah, USA
(2007) 384c
Van Konynenburg, P. H. and Scott, R. L., Critical Lines and Phase Equilibria in Binary van der
Waals Mixtures, Philos. Trans. R. Soc. (1980) 298 495-594

174

CR08-64
MODELING HEAT-MASS TRANSFER PROCESSES ON REGULAR
PACKINGS OF DISTILATION PLANTS
Arkharov I., Navasardyan E.
Bauman Moscow State Technical University, Moscow, Russian Federation
ABSTRACT
The paper presents the results of mathematical modeling heat-mass transfer processes on
regular packing of two profiles rectangular and parabolic types. The paper gives graphic
illustrations of distribution of pressure, temperature and concentration in one cell for every
type of packing. Modeling geometrical structure of the packing and heat-mass transfer
processes is plotted in the program complex STAR-CD.
INTRODUCTION
Heat-mass transfer processes realized in column apparatuses are widely used in chemical,
food, cryogenic, oil-processing and other industries. They are the base of many
technological processes of separation, distillation, rectification of multicomponent
mixtures, solutions and emulsions. So, their effectiveness is directly linked with technical
and economic indexes of plants. It is evident that the intensification of heat-mass transfer
processes in separation plants is one of the principle conditions of increasing efficiency and
profitability of air separation plants (ASPs) as a whole. Creation of efficient heat-mass
transfer apparatuses makes possible to increase thermodynamic efficiency of the ASPs, to
decrease capital costs for their production as well as power inputs at exploitation, as a result
to decrease cost price of oxygen, nitrogen, argon.
1. INCREASE OF AIR SEPARATION PLANT COLUMNS EFFICIENCY
Now the columns of 0.2 12,0 m diameter and 3,0 70,0 m high at mass flowrate of raw
material of 0.05 6,500 kg/s are widely applied in industrial plants. Before the trays of
different type and model were used as a packing, many of which are used up to the present
time. The rectification process in the tray column is stepped, it is characterized by high
hydroresistance of a step, and the improvement of the process is limited. At present the
performance of some types of trays is up to 80%, but power inputs indices of the ASPs are
not decreased. New developments in this way are considerably curtailed; development of
packing of other types is actively stimulated. Industrial packed columns for deuterium and
tritium at 20K appeared in 70s of the last century. A continuous or film regime of
rectification on packing is used in them. The packing is a contact surface formed by
irregular fill of contact elements of different form or by a specially ordered structure of
elements layers.
In the middle of 80s of the last century the Swiss company Sulzer Chemtech Ltd. has
generalized successfully its own experience and know-how in the field of the packing in the
chemical production, it elaborated and put on sale the packing Mellapack for air separation
that is being used till now. Today the volume of world production of ASPs with packed
columns reaches 45% of the total volume of ASPs production, and the tendency to
increasing has been noticed.

175

The packed columns work at rather large range of pressure of 0.1 1.0 MPa and liquid
loads of 5 20 m3/m2 per hour, they have low hydro resistance (approximately 5-7 times
less than a tray packing).
The upper columns (nitrogenous, low pressure), columns of raw and pure argon, columns
of extraction and concentration of kryptonxenon mixture are packing apparatuses of
modern ASP. Little by little some ASPs are equipped with lower (oxygen) packing
columns. Today world producers of cryogenic heat-mass transfer equipment for
apparatuses of average and high efficiency use a structure packet gofer packing. The
creation of new type packing, perfection and renovation of production technology as well
as determination of heat and mass transfer and exploitation characteristics inevitably is
accompanied by a big volume of construction and technological developments. The
experiment is still a decisive factor to solve this optimization task and select a type of
packing. The interest to mathematical modeling heat-mass exchange processes in the layer
of complex geometrical configuration of the packing is constantly growing.
Modeling and investigation of heat and mass exchange processes is done in several stages.
At the first stage the problem of distribution of fluid and gas flows along the packing is
resolved, dynamic retention is determined as well as hydrodynamic resistance of the
packing. At the second stage on the base of the obtained result the process of mass transfer
in each element of the packing taking into account inequalities of phases flow in it is
modeling. Then the fields of a components concentration and their transfer in the volume of
the packing are defined.
All the volume of the structure packing may be presented as some ordered in fixed manner
combination of elementary cells or calculated elements. Such an approach allows applying
a classic mathematical model of heat and mass transfer to the cells determined
geometrically, and finding the solution of the system of hydrodynamic equations
concerning the concrete cell. After calculation of the processes in a separate cell a layer of
these cells is formed according to real dimensions of the separation section of the
apparatus. Specifying initial values of concentration and desirable low of their distribution
along the height we can define the necessary quantity of layers or the height of the packing.
2. MODELING GEOMETRY OF THE PACKING
AND HEAT-MASS TRANSFER IN IT
Let us examine a plate and parallel packing Mellpack 250 type without perforation. Its
working surface is formed by gofer sheets at an angle of 450 that lay in vertical position, for
all this, even layers are turned at 1800 relative to odd layers on a vertical axis. Crossed
sheets form a combination of cavities that are calculated elements or cells (Figure 1). A
three-dimensional cell formed by adjacent plates of the packing is shown in Figure 2 (a),
(b).

176

y
z

Cell

Figure 2. A three-dimensional cell:

) cell; b) finite-difference grid

Figure 1. Crossed sheets form

W
wall

b)

film
liquid; W1liquid

Boundary i; W1vapor.i

N2
Uliquid, y

U vapor, x

Uliquid, x

Zone of liquid and vapor

vapor; W1 vapor

U vapor, y
Zone of the main vapor
vapor.; W1vapor.; P=const

Figure 3. Scheme of symmetric flow of the liquid film in a

counterflow with vapor

Coordinate axes coincide with the sides of the cell y and z in horizontal plane, x in
vertical plane; the origin of coordinates is at the top of the cell angle. A gravitational force
is acting in the direction of the axis x.
When plotting a mathematical model let us assume the following assumptions:
-regime of flow of two phases is laminar and stationary;
-process of impulse, heat and component transfer is considered as steady;
-interaction of phases is one-sided;
-coefficients ,,

D are homogenous and constant in the field of transfer;

177

-physical properties of two phases, mass flowrate and value of disturbing force along the
axis x of the channel are adopted as constant;
2W
in energetic balance is neglect. At the boundary of the section
y 2
temperatures Twall; Ti, Tvapor.; mass concentrations - W1vapor.i, W1vapor. ; velocities of vapor
and liquid Uvapor.i, Uliquid.i, Uvapor. dont depend on the coordinate x; here i is an index
designating the parameters on the boundary of the film at y = .
-component

The obtained systems of equations are solved by finite-difference methods. For example, by
the method of finite elements. A finite-difference network consists of cubic or other type
calculated cells (Figure 2, b). The type of calculated cells is chosen based on geometry of
the cell. A recommended relation of network dimensions of the cell is 16x9x9, though
decrease of the relation doesnt influence much the precision of calculation.
Initial and boundary conditions:
1.

Initial conditions
- for vapour
t = 0; 0x b cos
;

0yb; 0z; (along y)

0y; 0za; (along z)

T = T0,
- for liquid

Ux = Uy = Uz = 0.

yb - ; z - ;

t = 0; 0x b cos ;

T = T0,

Ux = Uy = Uz = 0.

2.Boundary conditions (conditions of adhesion, solid impermeable walls):

x = 0, b cos

t 0;

y = 0, b;

z =0, a;

Ux = Uy = Uz = 0 for liquid and vapor

3.Heat boundary conditions: (adiabatic walls)
T
x

x =0, b cos

=0;

T
y

=0;
y =0, b

T
z

=0;
z =0, a

A convective movement is initiated by a constant gradient of temperature across a liquid

layer:
T=T1, at x=0
where T1 is warm side temperature, K;
T=T2, at x= b cos

, at t 0

where T2 is cold side temperature,K

y

y=0

y=

Uliquid, = 0

Uvapor, y = Uvapor.

a) Uvapor. i = Uliquid. i = Ui

Uliquid, y = 0

Tvapor = Tvapor.

b) Tliquid. i = Tvapor. i = Ti

Tliquid = Twall

W1 vapor = W1vapor. ;

c) W1 vapor = W1 vapor. i

Table 1: Boundary Coditions for vapour and liquid

178

Components

2W
2W
and
y 2
z 2

are neglect in energetic balance. The components of

velocity are defined on the edges of calculated cells; the temperature and the pressure are
determined in the center of the cells. To keep corresponding transferable characteristic it
is recommended to use the form of finite-difference presentation of adequate members on
the cell-donor. To apply necessary boundary conditions we assume that the liquid is
surrounded by one layer of dummy cells. The difference scheme is explicit. The velocity
distribution is calculated using the equations of momentum applying the values of the
antecedent time pitch. Then a new distribution of velocity changes by circuit to perform the
equation of mass conservation by means of changing pressures in calculated cells. After
calculation of a new distribution of velocity it is used for calculation of a new distribution
of the temperature applying the equation of energy. Then these new distributions of
velocity, pressure, and temperature are used to define concentration, and as initial values
they are used for calculation cycle at the next time pitch.
To illustrate this method the calculations of non-stationary heat-mass exchange processes in
cells of several types using software STAR-CD and subprogram Star-design (to create
a three-dimensional model of an elementary cell and a subprogram Pro-amm for
generation of a 3-D finite-difference network from a modeled cell) were carried out.
The initial conditions are as follows: vapor air (y N2=0.79; yO2=0.21); liquid air
(XN2=0.79, XO2=0.21), counterflow Pinit=0.13 MPa, vn=0.5 m/sec, Tinit=83K. The stationary
condition is modification of parameters at inlet no more than 0.1%. Figures 4 -7 show the
distribution of velocity, pressure and concentration of O2 in cells in steady process.

b)

Figure 4. Field of vapor velocities in the cell of rounds:

) general view; b) middle section of the cell

179

b)

Figure 5. Field of velocities and pressure of vapor in the cell of rounds:

) general view; b) middle section of the cell

Relative change of
concentration

Unit measure, m
Local maximum and
minimum

O2

Figure 6. Distribution of oxygen concentration

180

yO2 = yO2 2

in the cell of rounds

Relative pressure change

Relative velocity of vapor

Unity measure, m/s

Unity measure, P

b)

Figure 7. Distribution in the cell of triangular shape:

) pressure change (minus corresponds to a zone of exhaustion);
b) vapor velocities relative to the liquid film

NOMENCLATURE
T- temperature [K]; - thickness of a liquid film [mm]; - thickness of a boundary vapor
layer [mm]; - kinematic viscosity [m2/s]; - density [kg/m3]; K- heat exchange
coefficient [W/m.K]; D diffusion coefficient [m2/s]; - vapor region remote of the zone
of the contact with a liquid film; W - mass concentration of the component; 1,2
component.
CONCLUSIONS
It is clear from the given illustration that the field of velocities is more irregular. So, in the
central region of the cells (Figure 4 b) a zone of slowdown of vapor is well seen, and a zone
of acceleration is seen in nodal points. The zones of low and high pressure are distributed
correspondently. But in laminar regime the field of concentration of one of the components
(oxygen, Figure 6) practically has no features and its configuration is near to the field
corresponding to the field in lengthy channel.
It is to note that the cited results of calculation according to the mathematical model
obtained with software STAR-CD are our fist steps to apply modern software for
modeling heat-mass transfer processes inside space structures at complex geometry.
REFERENCES
1.
Mallinson G.D., De Vahl Davis G., Three-Dimensional Natural Convection in a
Box: a Numerical Study, J.Fluid Mech. (1977) 83 1-31
2.
Del Giudice S., Strada M., Comini G., Three-Dimensional Laminar Flow in Ducts,
Numerical Heat Transfer (1981) 4 215-228
3.
Chjan A.M.S., Benergi S., Numerical Simulation of Three-Dimensional cell Vortex
in closed Cavities with Solid Impermeable Walls, Teploperedacha (1979) 101 2 52-57

181

STORAGE
AND
TRANSPORT
OF INDUSTRIAL GASES

183

CR08-42
OPERATION OF SMALL AND HIGH PRESSURE TANKS FOR
LIQUEFIED AIR GASES
Hnzdil T., Suma J., Kouba M., Chrz V.
Chart Ferox, a.s., steck 30, Dn 5, 405 30, Czech Republic
ABSTRACT
Small vacuum storage tanks (up to 11 m3 capacity) are generally used for applications,
where, for logistics optimization, the re-filling periods are longer, typically two-three
weeks. With long retention time the bulk of liquid becomes warmer and doesnt accept so
much heat leak, which is transformed more into vaporization, then. In high pressure tanks
with relief pressure set to 37 bar and operating pressure close to the critical pressure of
nitrogen at 33 bar, as encountered at laser applications, e.g., the heat of vaporization is very
low. Any heat leak causes large vaporization and holding times are very short.
Minimization of heat leak is very important, then. Holding time and nominal evaporation
rate was improved by using of multi-layer insulation at the new multi-layer insulated EVT
tanks in sizes 3, 6 and 11 m3,. Measurement of temperature distribution in a cryogenic
pallet tank 600 liters at high operating pressure proves warming up of the bulk of liquid.
Recommendations for prolongation of the holding time of high pressure tanks are
presented.
INTRODUCTION
Storage tanks used for storing of cryogenic liquefied gases are double-wall vacuum
insulated vessels. Annular space has to be filled by special insulation material for to reduce
rest gas convection and thermal radiation. This way, heat leak is minimized and optimum
holding time achieved.
The outer jacket is only a container for the inner vessel and its thermal insulation. Its main
purpose is to ensure a high vacuum in the insulation space between both the inner vessel
and the outer jacket to provide good thermal insulation of the inner vessel.
Despite the good quality of insulation of cryogenic tanks, there is still a continuous heat
flow (heat leak) from the environment to the fluid in the inner vessel, which increases the
temperature of the fluid, causes partial evaporation and increases the pressure in the inner
vessel. The withdrawal of fluid reduces the pressure in the inner vessel as a result of
expansion of the vapor space. At best, thermal energy intake and fluid withdrawal rates are
in balance. If the liquid withdrawal is too low so that it cannot compensate for the
increasing quantity of evaporated gas, the pressure is growing until the pressure relief
system set pressure is achieved; the relief system opens and releases gas from the tank from
time to time to avoid over-pressurizing. In the majority of tanks the liquid withdrawal is
larger and heat leak doesnt compensate for the expansion of the vapor space. To prevent
the pressure from falling below the required operation value, a pressure build up vaporizer
(PBU) is installed at each tank. The pressure build-up regulator allows liquid from the
bottom of the tank to flow into the PBU if the operating pressure falls below a set point.
Evaporated gas is returned into the upper part of the tank, which results in a pressure
increase, until the upper pressure set point is achieved. The lower and upper set points of
the regulator are close to each other with the pressure difference 0,5 bar and far below the
set relief pressure. This makes possible accumulation of vapor during interruption of liquid
185

withdrawal.

Figure 1. Flow diagram of cryogenic storage tanks for air gases

The number of ordered and delivered HP tanks has greatly increased in the last few years. It
is closely related to increased number of laser application requiring nitrogen at higher
operation pressures around 30 bar. A few years ago the laser cutting equipment was very
expensive, which was driving continuous operation of these tanks with no time for increase
of pressure. Today, more and more lasers work at small companies under part-time
operation. For laser cutting operation the tank operating pressure should be between 30 and
33 barg. In such cases, minimization of heat leak is very important. Regular shut-down
Friday evening to Monday morning (app. 60 hrs) can cause pressurization to the relief
valve set pressure 37 bar. In case of shut-down throughout the weekend, these tanks use to
pressurize up to either set pressure of the tank relief valves or set pressure of the boil-of
regulator conected to the consumption net . When the quick pressure increase occurs, the
typical first reaction of the end-user is that he is losing gas and the first impression is that
the tank is faulty, because other tanks of the same type are working well.
1. SPECIFICS OF SMALL HIGH-PRESSURE TANKS
Holding time of a cryogenic tank is a time from a defined initial state of liquid up to
achieving the relief valve set pressure due to heat leak and opening the relief valve of the
tank. This situation is not desirable because of losses of the gas. At high pressure (HP)
tanks with relief pressure set to 37 bar (3.7 MPa) and under operating pressure close to
critical pressure of nitrogen at 33.97 bar (abs) = 33 bar (g=gauge) approx., the heat of
vaporization is very low (Table 1), any heat leak causes large vaporization and holding
times are very short. In the range of super-critical pressures 33 to 37 barg, the vapor and

186

pressure
bar(g)
0
10
20
30
32
32,8
33
Table 1.

equilibrium
evaporation rate
temperature
r
(if entire liquid saturated)
K
kJ/kg
%/day
kg/day
77,2
199
0,49
23
105,1
148
0,66
31
116,3
110
0,89
42
124,1
61
1,62
75
125,5
42
2,33
109
125,9
28
3,54
165
126,08
0 critical point = non-defined
Heat of vaporization of nitrogen
and evaporation rate of the tank VT6 (6 m3)

liquid in the upper layers become above-critical fluid, which expands according to its rising
temperature. It should be accounted, that the heat leak of small tanks is relatively high
because of large surface to volume ratio.
Forecast of pressure increase is especially difficult in case of small high-pressure tanks,
because of following aspects:
a)

Unknown average thermal state of liquid, depending on different temperature in

various layers of liquid, with the lower layers being colder than the upper ones.
The distribution and the average saturation of the liquid depends on the thermal
history of the liquid, as it was transported, mixed with the rest of liquid in the tank,
stored, the liquid withdrawn, etc. In consecutive steps of filling-emptying of the
tank there is stratification of the bulk. Always the coldest liquid is being
withdrawn and warmer liquid remains in the tank. When filling the tank with fresh
cold liquid, the vapors are condensed and possible excessive pressure reduced. The
mixed liquid gets colder. As the period between individual fillings is longer, the
average liquid gets warmer.
b) Low heat of vaporization at high pressures, rapidly reducing with growing
pressure
c) Uncertainty of the real physical level of liquid, which depends on the liquid
density, which further depends on average saturation of the liquid. Consequently,
there is large uncertainty on the real volume of the vapor space above the liquid.
Smaller vapor space results in quicker pressure rise. Real level of liquid in the tank
is higher than shown on the scale of the DP gauge. The level increase is rapid as
the pressure is getting close to the critical point.
Common causes of rapid pressure increase:
1.
When there is no withdrawal from full (95%) tank, liquid expands and fills in the
ullage space (the vapor space above the liquid level). As liquid is almost
incompressible this results in really fast pressurization (tens of bar per day)
2.
Too high temperature of the liquid.
If the volume of the tank is oversized in relation to the average liquid consumption
and if there is a long period between each two fillings. the liquid stored for a long
time is gradually heated up close to its boiling point. Such a heated liquid has no
capacity to absorb the heat leak as it had during warming up. Majority of the heat
transferring through the insulation causes vaporization of liquid and consequently
the tank is pressurized more quickly.

187

We can reach significantly different pressure increase in case of a single monitored high
pressure tank by measurement, caused by minor changes of operating parameters.
(Long-term monitoring of VT6/37 in laser application, operated at 32 barg has shown
increase by 1 to 3 bar/day, e.g.).
When the tank is equipped with the economizer line (Fig. 1, line J), minimum required
product withdrawal (Fig. 1, line D) for compensation of the increase of pressure in the
vapor space is equal to the vaporization which makes possible venting of the vapor space to
the production line (Table 2.). Without any economizer line, the product withdrawal would
have to be larger in the ratio of liquid to vapor density, which at saturated liquid at 30 barg
is 2.7, but it can be much larger when some colder liquid is still at the bottom of the tank.
Tank volume

Type

liter
3400
6100
8700
11200

VT3
VT6
VT9
VT11

Evaporation rate
at 30 barg
kg/day
56
75
97
110

Type

EVT3
EVT6
EVT9
EVT11

Evaporation rate at
30 barg
kg/day
39
53
68
77

Table 2. Minimum withdrawal rates of the product for compensation of evaporation at

constant pressure 30 bar in small high-pressure tanks, equipped with an economizer line.
(Compare to Table 1.)

When liquid nitrogen tanks are operated close to the critical point of nitrogen (33 barg) and
the liquid is warmed up close to saturation pressure because of longer storage time, the
evaporation rate is high due to low heat of vaporization. It is typically observed as quick
pressurization when the tank is on stand by at a high pressure.
The situation with palletized tank in the range of sizes from 600 liter to 2000 liters and the
relief pressure 37 is very similar to that, stated above on standard bulk storage tanks with
the difference, that multi-layer insulation has been used for these tanks already for thirty
years. The pressure and temperature changes are much quicker at these tanks because of
their smaller volume and larger NER.
2. EXPERIMENTAL PART
We modeled the situation on a small 600 liter tank (Euro-Cyl 600/37). For to follow up
temperatures in various layers of liquid and vapor space, two thermometers were at the
bottom head, one was in the middle of the liquid bulk, the third was at the top head. The
temperature on the surface of the liquid was calculated from the pressure in the vessel as
vapor-liquid equilibrium. The tank was filled with liquid nitrogen and saturated to 1 bar
abs. boiling point by venting the excessive gas to the atmosphere. (Measurement of
temperature changes at constant surface temperature of the liquid was the aim of this
experiment. As it can be seen from the course of the temperature curves, both liquid and
vapor stratification was observed at beginning.
When the liquid saturated at a low pressure and corresponding boiling temperature (1 bar
and 77 K e.g.) is exposed to higher pressure (28 bar, e.g.), condensation of vapors takes
place at the liquid surface, the surface becomes warm and the heat penetrates down by
thermal conductivity of liquid.

188

Temperature (K)

Measuring of temperature on EC600/37

130
125
120
115
110
105
100
95
90
85
80
75

T1

50

100

T2

150

T3

T4

Tp

200

250

300

Time (hours)

Figure 2. Temperature changes in a tank, exposed to periodical pressurization to 28 bar.

Liquid level 84%.

Figure 2. shows how the pressure in the tank collapses as result of condensation of vapor on
the surface of the cold liquid even, when impulse pressurization is repeated.
Collapsing of pressure is seen by decreasing of the equilibrium temperature in the top of the
tank (Tp). The heat transfer into the bulk of the liquid is seen by increasing the liquid
temperature (T1 center of the bulk 50% height, T2 25%, T3, T4 0% -bottom of the
tank). After 250 hours the entire volume of the bulk and the tank walls are on a temperature
around 105 K corresponding to saturation at 10 barg, while the pressure above the liquid
level was still 20 barg. This experiment proves warming up of the liquid due to the pressure
in the tank and simultaneous collapsing of the pressure due to condensation on the cold
liquid. .
Another experiment was done with the functioning pressure build-up (PBU) regulator
function (Fig. 3). The regulator set pressure 28 bar is stabilized at the liquid surface.
Measuring of temperature on EC600/37
165
155
Temperature (K)

145
135
125
115
105
95

T1

T2

T3

T4

Tp

85
75
0

20

40

60

80

100

Time (hours)

120

140

160

180

Figure 3. Temperature changes in a tank, exposed to continuous pressure control at 28 bar.

Liquid level 40%. The 50% thermometer measures the temperature of the vapor.

Than the tank was sealed and PBU, set to 31,7 barg (equilibrium temperature 125 K). The
experiment was started by full opening of the pressure build-up regulator and increasing the
pressure in the tank to the regulator set pressure. After achieving the set pressure, further
inlet of vapor from evaporated liquid from PBU compensated for condensation of vapor on
the surface of the liquid. Maintaining constant high temperature on the surface results in
189

continuous warming of the bulk of the liquid. At warmer liquid the rate of condensation
was reduced and tendency of pressure growth even at closed regulator was compensated by
manual venting to the atmosphere. This way, constant pressure and, consequently constant
temperature on the liquid level was maintained during the entire experiment time. All the
liquid temperatures are equalizing to the temperature of equilibrium. The same happens in
the vapor space. The original stratification of vapor is equalized. We presume that more
important, than the conductivity of vapor and its possible circulation, thermal conductivity
of the thick walls and head of the high pressure tank play an important role, when all the
metal equalizes at the temperature of equilibrium.
When the bulk of liquid is colder than equilibrium, large part of heat leak can be absorbed
by liquid on account of increasing its temperature and only a small part is converted into
vaporization of part of the liquid, which causes increase of the pressure. But after the
temperature of the liquid is close to equilibrium, all the heat leak is converted into the
vaporization and the pressure is growing quickly. This cannot be seen in this diagram,
because the pressure was maintained strictly constant by opening the PBU in the first
period of the experiment during 95 hours, but by manual opening of the vent valve after
this time. The purpose was to measure heating of the bulk of liquid at the constant surface
temperature.
3. OPTIMUM OPERATION FOR REDUCTION OF PRESSURE INCREASE
3.1. Two important aspects to be considered, when operating high-pressure tanks:
a) The pressure increase is significantly influenced by the level of the liquid in the tank.
Differential pressure gauge does not show this real level of the liquid, because the
temperature and density of the liquid should be close to equilibrium. As standard, the scale
of the level gauge is calculated for the density of liquid, saturated at the atmospheric
pressure, as the only state, which can be definitively reproduced in a real tank. But a typical
operation state is a non-defined and continuously changing saturation to a temperature,
which corresponds to medium thermal saturation of the liquid between the original cold
state and the saturation at the tank pressure. This means that the liquid is always warmer, its
density always lower and the real liquid level proportionally higher that what can be seen
from the level gauge. The vapor space can be much lower than indicated from the nominal
level, which results in quick vaporization.
b) The pressure increase is also directly influenced by the temperature and its stratification.
If the temperature of the liquid is low then the major part of heat input is consumed for the
heating-up of the bulk.
3.2. Following recommendations can be applied to operation of high-pressure tanks
A) Maintaining the liquid temperature as low as possible.
Lower average temperature of the liquid results in larger absorption of heat leak into the
liquid and reduction of vaporization. Quantity of vapor, which has to be accommodated in
the vapor space, is lower, then.
-

First rule comes out directly from the above described experiments. One part of liquid
heating is caused by condensation of vapor, which is higher, the higher the pressure in
the tank. The liquid level surface always has a temperature, which is the equilibrium
one to the pressure. Heat is transferred downwards into the colder layers of the bulk of
liquid. If the temperature gradient is large shortly after the tank filling, the tank
pressure would drop as result of vapor condensation on the level. PBU comes into
function then and maintains the pressure by vaporization of liquid. This should be
190

prevented during stand-by periods like nights or weekends wherever are conditions for
such a sophisticated control. When a tank is filled with fresh liquid and left unused
over weekend, e.g., the PBU circuit should be shut and the pressure allowed to
drop for the period not in use. This would reduce the warming up of the liquid
and increase the non-vent holding time during this and possible next stand-by
periods.
-

Another of sources of heat is mixing of the fresh cold liquid filled in the tank with the
rest of the previous liquid, which is warm after long time storage. The resulting
temperature is lower, when the quantity of the rest of liquid is lower. The average
temperature is lower, when the volume of the rest of liquid is small. This results in
another useful rule: The rest of liquid before the filling of the tank should be
minimized, with respect to specific conditions, by optimized logistics of liquid refilling.

B) Maintaining the vapor space (ullage) sufficiently large.

Two aspects have to be considered:
-

real vapor space is always smaller than what could be read from the scale of the level
gauge, because the density of liquid is lower than the maximum one, considered for the
scale. The real liquid height is inversely proportional to the liquid density,
approximately. Average density of liquid is always somewhere between that one of
mixed liquid after the tank filling and the equilibrium one corresponding to the tank
pressure. More detail analysis find in [1].

larger vapor space (in the range of liquid filling 80 to 95%) reduces the pressure raise
rate. (At smaller liquid filling the pressurization rate is larger because of lower heat
absorption in the liquid.)

C) Maintaining the optimum filling regime

Conclusion of the items A and B: It is better to fill the tank from 15 % to 50% rather than
from 30 % to 65 %, e.g. We evaluated similar cases a the tan VT6/37:
- filling from 49,3% to 63 % - average pressure increase was 2,55 bar /day
(average from measured 2,80 and 2,39 and 2,47 bar increase per 24 hours).
- filling from 18,5% to 49 % - average pressure increase was 1,66 bar /day
(average from measured 0,99; 1,81; 1,61; 2,0 and 1,82 bar increase per 24 hours).
4. LOW HEAT LEAK TANKS
As it was reasoned and documented above, the general problem of rapid pressurization at
HP tanks, especially the small ones, is caused by thermodynamics of the storage processes,
For to reduce this problem, Chart Ferox, a.s. developed new range of multi-layer insulated
small tanks of the range 3 to 11 m3 (marked EVT tanks), based on 20 years experience of
Chart Distribution & Storage Division.
The heat leak of these tanks is lower than of those, insulated by perlite, by 20%
approximately.
With respect to importance of the minimization of the heat leak at the high-pressure tanks,
the heat leak is further reduced by increasing the thickness of the insulation by higher
number of layers. The tank EVT6/37 (multi-layer insulated tank) with 50 % layers addition
represents NER decrease by 38 % compared to the perlite insulated tank VT6/37.
191

Measurement of holding time of liquid, saturated initially at atmospheric pressure doesnt

represent the typical real operation conditions, but it is a precise high-reproducibility
measurement, which, besides of NER measurement, proves the quality of insulation. The
pressure increase in 400 hrs in case of EVT6/37 at 42 % of filling from 0 bar was 2,5 bar,
while it was 6,5 bar in case of the VT6/37.
Substitution of perlite insulated tanks by multi-layer insulated EVT tanks with the addition
of layers for high pressure tanks would eliminate majority of users problems with pressure
increase at high pressure tanks.
Pressurizing of VT6 (EVT6) at filling 42%

12

VT6

Pressure (bar)

10
8

EVT6

6
4
2
0
0

200

400
Time (hours)

600

800

Figure 4. Comparison of pressure rise in VT and EVT tanks

CONCLUSIONS
Specifics of small high pressure cryogenic tanks results in relatively short holding times
between operation pressure and relief system set pressure, which causes loss of media
during several-day stand by. Following measures can be recommended for prolongation of
holding time.
Refilling of the tanks at minimum possible level of the rest liquid.
Not filling the tanks up to the trycock, but to lower level, when stand-by period is shortly
after the refilling.
Shutting the PBU circuit during the stand-by period.
Range of special EVT tanks with multi-layer insulation was developed for maintaining low
heat leak, which results in longer holding times.
REFERENCES
1. Chrz V., Suma J. : Dynamics of tank pressure during storage of cryogenic liquids.
Proceedings of the 22 Congress of Refrigeration, Beijing (CD), International Institute of
Refrigeration, Paris, 2007

192

CR08-52
40 FOOT CRYOGENIC INTERMODAL ISO CONTAINERS
Mtl P., Lnsk M., Chrz V.
Chart-Ferox, a.s., steck 30, Dn 5, 405 30, Czech Republic
ABSTRACT
As LNG has become more and more important energy source, demand for cost effective
transport has arisen. The word effective relates to several features. First of all, it is a high
payload, further capability to store and deliver liquefied natural gas without any losses by
venting by providing long holding time and finally, very important, compatibility with
requirements of intermodal transport. ISO containers make possible continuous rail, road,
river, sea shore, over sea and ocean transport. Transport flexibility creates new 40 feet
container opportunities. For instance, they can be used as a base of virtual pipelines that
represent a viable alternative where capital and operational costs of fixed natural gas
pipelines make the project too expensive.
Transport of LNG is not the only use of ISO containers. They can also be used as
temporary source of LNG during pipeline maintenance or instead of stationary storage
tanks at satellite LNG stations. Advantage of 40 foot containers is that they have large
volume which is also attractive for transport of other low density cryogenic liquids,
ethylene or ethane.
In order to meet all customers requirements Chart Ferox has successfully developed and
built a new 10 bar 40 foot LNG ISO container with currently the largest volume in the
world, 43,5 m3.
Chart-Ferox made the best of its experience of proven 20 foot cryogenic container design.
It is essential that final element method of strength calculation is used to develop light
container and retain its reliability and durability. Container approval meets ADR, IMDG,
RID, EN 13530 and CSC requirements, for which, among other, the container was
subjected complex static and dynamic load tests.
Currently, first series of the product was delivered to a customer in Pakistan, where it will
be used in frames of a virtual pipeline project.
40 FOOT CONTAINER DEVELOPMENT
Firstly, usage of the container has to be defined. If the container is intended for intermodal
transport, it must comply with specific codes and norms for road transport (ADR, DOT or
equivalent national codes), for railway transport (RID) and for maritime transport (IMDG).
Not all countries accept all design codes. For instance, the combination of ADR and EN 13
530 could not be used for ISO containers that are to be freely used in the USA because
local regulations require that such equipment must have DOT approval when transported
over roads. JB 150 coded containers are required in China. Each code has significant
influence on container design and prototype testing.
Major weight difference is obvious on inner vessel. In order to keep same MAWP of the
container ASME/DOT inner vessel is approximately 1.3 times thicker than ADR/EN coded
containers. Other option is to keep same thicknesses of inner vessel but, due to the
difference between the two codes, the maximum allowable working pressure of an ISO

193

container manufactured according to ASME is lower compared to the same ISO container
manufactured according to EN13530.
The container meets TPED and can be registered and operated in each country of the
European Union, or countries, who accept this code. Other codes are applicable case by
case.
The name ISO container comes from compliance with ISO 668 and ISO 1496-3, which
clearly describes ISO containers dimensions, gross weights and testing. Meeting IMDG
code, the container can be used for overseas or ocean transport. It is important to know that
for cryogenics liquids like LNG, ethylene or liquefied gases container has to be designed
according to the tank instruction T75 of IMDG and, consequently, the chapter 6.7.4 for the
design of the container. Such containers have a specific name portable tank. Besides
portable tanks we can see cryogenics containers called tank container on the market,
which is according to ADR but not IMDG. Difference between portable tank and tank
container is that the tank container can be designed without meeting requirements of the
dynamic test.
Chart Ferox has developed and built a new 40 foot LNG 10 bar ISO container, marked
TVS-43-PB-10, which meets requirements of the category of portable tank.
CHART FEROX CONTAINER FEATURES
ISO container is characterized by following parameters:
Parameter

Data

Product name

TVS-43-PB-10

Type of the container according to

IMDG/ADR
Tank instruction for portable tank
Design codes
Approvals

UN Portable tank

Gross capacity
Tare weight
Maximum gross weight

T 75
EN 13530, ISO 1496, ADR, RID, IMDG,
CSC, TPED, Gost-R, Rozreshenie
Rostekhnadzora
43 500 Liters
12,000 kg
30,480 kg

Max. payload
MAWP

18,480 kg
10 bar

NER

0.2% LNG per 24 hours

Max. Holding time

60 days at 81% LNG filling

Table 1 Chat-Ferox container features

Because of low LNG density it makes possible using the entire available size of the tank,
still in compliance with the maximum allowable gross weight of 30 480 kg. Possibly, this
type is the most effective with the ratio of the payload 18480 kg to tare weight 12000 kg
with 43.5 m3 volume .

194

The cryogenic ISO container is a vacuum insulated double walled vessel fitted into a fortyfoot container frame. Highly efficient insulation and low heat leak of supports are critical
for all cryogenics ISO containers. Container supports hold the inner vessel in its position
inside the outer vessel. Non-metallic materials with low thermal conductivity are suitable
for this type of supports. The most effective thermal insulation is a multilayer super
insulation combined with high-level vacuum in the interspace. It consists of reflective
aluminium foil and low conductivity glass-fibre-paper. These two components are laid in
multiple layers providing thermal protection against heat leak.
Thermal performance is characterized by NER index, which shows how much liquid
evaporates during 24 hours. The lower the NER index the better thermal performance the
container has.
Other important indicator is the Holding Time. It specifies how long it takes the container
to pressurize from atmospheric pressure up to the safety relief valve set pressure. As the
relief valves should not vent during the whole transport time, the holding time determines
the maximum transport time of a container without any product loss. Calculation method
and measurement procedure is described in EN 12213 Methods for performance
evaluation of thermal insulation.
Design pressure of the ISO container is another technical aspect to be taken into account
when selecting the right product for certain applications. The higher the operating pressure
of the container, the longer the holding time. Therefore, it may seem to be better to select a
container with the highest possible operating pressure. Higher operating pressure of the
vessel results in higher tare weight, which may negatively affect the payload. Therefore
design pressure should be set in order to ensure holding time and certain tare weight limit.
The table below shows holding times for 10 bar and 7 bar container with different filling
conditions but both with NER rate 0.2% LNG/day. 0 bar filling saturation pressure can be
expected from marine terminals, e.g.. Then holding time is sufficient for both the design
pressures. Most common case is that the inner vessel of the container arrives to filling
station not with LNG temperature which increases saturation pressure of filled liquid.
Moreover LNG from liquefiers or from intermediate pressure storage is often saturated to
higher boiling pressure, 3 bar e.g.. We can consider from practical experience 5 bar
saturation pressure. Holding time for 7 bar container would be probably too short, then.
Holding time of
7 bar MAWP
container
0 bar saturated LNG after filling

day
50
s
5 bar saturated LNG after filling
day
10
s
Table 2 Holding time comparison

195

Holding time of
10 bar MAWP
container
60
22

PRODUCT DESIGN AND TESTING

As mentioned above, this ISO container is of the category of portable tank according to
IMDG. Quality and reliability of such a product is a fundamental requirement of every
operator of cryogenic containers. This is why we paid maximum attention to the design
concept. Its design arose from proven 20 foot version, but increased length brought new
challenges. Finite element analysis (FEM), which was previously used during
improvements of 20 foot container, made possible to compare stress in all the critical points
and compare with those, which occur on the many years operated designs. Gained
experience established qualification for design and optimization of new-longer container
using FEM.
During design, it was necessary to optimize between contradictory requirements:
- Minimum weight
- Low heat losses, which mean suitable support system
- Stiff and strong construction able to withstand all operational loads
- Fatigue resistant construction.
All the design and testing states were calculated. It included static load of 1G in vertical
upward direction and 2G in all other directions. Static tests included stacking, lifting from
the four top corner fittings, lifting from the four bottom corner fittings, external restraint
(longitudinal), internal restraint (longitudinal), internal restraint (lateral), rigidity
(transverse), and rigidity (longitudinal). The most difficult state was the dynamic test;from
the view of the assembly strength and from the view of evaluation as well.
According to ADR 2007 and IMDG each new portable tank has to meet requirements on
dynamic test according to Manual of Test and Criteria, which is based on code Transport
Canada. The container is fixed by all 4 corner fittings to a staying wagon. Another wagon
strikes it several times until the impact acceleration meets the required criteria. Measured
acceleration is analysed and Shock Response Spectrum (SRS) is calculated. The required
impact is characterized by a defined SRS curve. The older method required maximum
measured acceleration of 4g. The peak value of the acceleration record according to the
new method during the test of the container prototype was 18g. Nevertheless the container
was able to withstand that impact and dynamic test was recognized as successful.
One of the test cases simulated by FEA is shown in Figure 1. Linear elastic analysis was
used which means that no deformation affecting model were considered. As deformations
are negligible in comparison with structure dimensions they were multiplied by 150 in
order to make them visible.
Picture 2 shows container during physical Bottom lift static test. Container is held by 4
arms whereas the angle between vertical line and arm axis is 45 which simulate lifting by
ropes.

196

Figure 1 FEM stress view

Deformed shape of ISO Container; Transverse stiffness test, ISO 1496-3. displacement
enlarged 150 times, max displacement of 4.3 mm.

Figure 2 Bottom lift static test according to EN 1496-3

197

40 FOOT CONTAINER USAGE

An advantage of ISO containers is the possibility of continuous transport over road, rail,
river, sea or ocean directly to the end user without any liquid transfer. Another advantage is
an easy implementation. The container of course undergoes all related regulations and
therefore it doesnt face possible problems with crossing frontiers. Road regulations, which
relate just to the carrying lorries of general use for the container transport, are usually
available at logistics companies. The third important advantage is the possibility of using
the container not only as a means of transport, but also as a temporary satellite station or
back up system in case of pipelines and storage tank maintenance.
If transport of LNG is regular and its defined by an exact amount (tons/day) we call it
virtual pipeline. It has the same purpose as gas pipelines but with extra benefits. Gas
pipelines have to be designed for expected flow rate but number of ISO containers operated
to certain destination can be easily adjusted to the gas demand. At the region of destination,
the containers can be directed to particular customers and unloaded to satellite stations or
used as temporary installation by exchange full for empty. Owners of large logistics or gas
companies can flexibly decide where containers are needed worldwide.
There are many typical examples, where it is effective to use LNG transport instead of gas
pipeline.
If there are difficult geographic conditions for laying down pipelines (sea, mountains)
The existing pipeline is not sufficiently sized for growing demand of gas at the end site.
The gas consumption is irregular with high peaks caused by a batch process technology,
weather, contract campaigns etc. The pipeline size may not be sufficient for covering the
peaks, or the user is penalized for high or low consumption.
Isolated users, such as hotels in the country, factories, small towns, villages, where the
energy consumption is of interesting size, but not sufficient for justifying a longer distance
pipeline
Each new gas delivery project should be evaluated by feasibility study where there is
comparison of all related costs.
CONCLUSION
40 foot ISO containers are specific products for LNG, liquid ethane and liquid ethylene,
where the large volume can be used for the light liquid within the maximum container gross
weight. It is an effective means of transport for complex logistics conditions when
combining various paths like road, railway, river, sea and ocean without any transfer of
product. This is based on meeting requirements of ADR, RID and of portable tank
according to IMDG. It is also suitable for pilot projects, because the container may serve as
temporary storage tank at satellite stations.

198

Figure 3 Routine liquid nitrogen testing

Figure 4 40 FT ISO Containers

REFERENCES
1. Chrz V., Bures I.: ISO containers modern means of cryogenic liquid transport.
Proceedings of the 4th International conference on cryogenic technology and equipment,
MirExpo, Moscow, 2007
2. Miroslav Cerny, Ivan Bures, Karel Mudroch: Design of cryogenics ISO containers
Cryogenics 2004

199

CR08-53
CRYOGENIC LIQUID TRANSFER POSSIBILITIES FOCUS ON
STATIC VACUUM INSULATED PIPES
Takcs D., Chrz V.
Chart-Ferox, a.s., steck 30, Dn 5, 405 30, Czech Republic
ABSTRACT
Liquefied gases can be transported in several piping systems. The most common systems
are: naked pipe, mechanically insulated pipe, dynamic vacuum insulated and static vacuum
insulated pipe. Each system is justifiable under certain conditions. Naked or mechanically
insulated pipes can be used for periodical service with large flowrates, as filling pipelines
of storage installations. Maximum insulation performance is required for pipelines under
continuous operation with low flow rates, just to mention two extreme examples.
Chart has developed efficient static vacuum insulated pipe with patented features as
bayonet-type connections for tighter seals using Invar alloy in the design. Several predesigned solutions of joints of the pre-fabricated sections and several systems of thermal
expansion compensation are presented with their particular advantages and suitability for
different projects as well as impact on the project economy.
Several special accessories were developed as vapor separators. Experience from
installation, testing of the final systems and follow-up service will be also presented.
INTRODUCTION - CRYOGENIC LIQUID TRANSFER POSSIBILLITIES
There are several possibilities to transfer liquefied gases at cryogenic temperatures such as:
- dewars
- piping systems
a) bare pipe
b) mechanically insulated pipe
c) dynamic vacuum insulated pipe
d) static vacuum insulated pipe
Dewars are suitable for small gas quantities. The filling losses are up to 40% (flash, cool
down). There are handling labour costs and risk of injuries during handling
Bare pipes can be used for a larger bore, short length piping with high flowrates (tank
filling, space craft LOX filling). They have high liquid losses. Frost and condensation can
occur on the surface of the piping system. Condensation of air can take place on the pipe
outer surface when transporting low pressure nitrogen. Enrichment of the condensate with
oxygen is high, which could lead to an explosion.
Mechanical insulation is an economical option for short runs (< 5 m). A foam could
degrade over the time and require periodical replacement. This cause operational cost. The
insulation is bulky and it has 15x higher heat leak than static VI line.
The dynamically pumped system enables fast delivery and easy installation but the vacuum
pump electricity consumption and maintenance cost can be critical. There is no radiation
shield installed and in case of any failure the insulation of the system is lost. The

201

dynamically pumped piping system has 3x higher heat leak than a static vacuum insulated
line.
1. WHY MULTI LAYER VACUUM INSULATION?
Heat transfer is a transfer of mechanical energy between colliding molecules expressed as
apparent mean thermal conductivity in [W/m_K] between boundary temperatures 300 K
(ambient temperature) and 80 K (liquid nitrogen @ 0 barg)
air
solid insulations (cork, balsa wood)
polyurethane foam
unevacuated powder (perlite)
evacuated perlite
straight vacuum
superinsulation

- 0.015 W/m_K (conductivity only)

- 0.04 W/m_K
- 0.02 W/m_K
- 0.02 W/m_K
- 0.001 W/m_K
- 0.005 W/m_K
- 0.0002 W/m_K

Case Study for liquid nitrogen transfer system (DN15):

Nitrogen for freezer filling
Tank pressure: 2 bar
Freezer pressure: 0 bar (flash loss 11.1%)
Distance: 15 m
End point quantity: 1000 kg
Filling: 1x week
Project parameters see in the table 1.
Operation Loss in rigid pipe systems
Flow rate
[kg/h]

Static VIP
973

Mechanical Ins.
956

"naked" pipe
809

Heat leak
Loss due to Heat Leak

[W]
[kg/h]

15
0,3

225
4,6

2250
45

Time
Operation Loss

[hr]
[kg]

1,2
125

1,2
131

1,5
199

Operation Loss

[%]

11,1

11,6

16,6

Table 1: Operation Loss in rigid pipe systems

2. CHARACTERISTICS AND DESCRIPTION OF VIP

The vacuum insulated pipeline is composed of vacuum insulated sections. The sections are
fully manufactured, evacuated and sealed in the factory. The individual sections are
connected preferably my means of bayonet coupling or, alternatively, by evacuated field
joints.
The vacuum insulated piping comprises a tube-in-tube conduit for transfer of mainly
cryogenic fluids. The inner tube holds liquid, the outer tube keeps the vacuum insulation
and bears the external loads. The annular space between the tubes is provided with
insulation. The thermal contraction of the inner tube is compensated by inner line bellows.
Vacuum insulated pipes have very low heat in-leak because of the static vacuum insulation
and the radiation shielding. Under normal circumstances (based on relative humidity and
outside temperature) the VIP will be free of condensation of atmospheric water.
Standard piping sizes are listed in the Table 2 according to nominal DN sizes.

202

Unit
Inner tube
dimensions
Inner tube ID
(approx.)
Outer tube
dimensions

mm

Line nominal dimension

DN25
DN15
DN40
D33.4x1.65
D21.3x1.65
D48.3x1.65

mm

18

mm

D60.3x1.65

30

45

D88.9x2.11

DN50

DN80

D60.3x1.65

D88.9x2.11

57

85

D101.6x2.11 D101.6x2.11 D141.3x2.77

Table 2. VIP line nominal dimensions

Design parameters of the vacuum insulated transfer systems are designed in accordance
with the data in the table 3.
Parameter
Units
Design pressure
Bar (psi)
10.3 (150) *
Design temperature
K
73 to 338
Design code
ANSI Section B31.3, PED
Inspection authority
By code
* Other pressure ratings available upon request
Table 3. Design data of vacuum jacketed pipelines.

The inner and the outer pipe material is stainless steel AISI 304L or equivalent. The tube is
of a longitudinally welded type. The compensating bellows are incorporated within the
inner tube. Therefore the thermal dilatations of the process tube are not transferred to the
outer tube.
The concentricity of the inner tube within the outer is ensured by spacers. The spacers also
guide the compensators.
The line insulation comprises a combination of the laminar radiation shield with vacuum.
Also a system of rest gas capturing by getters and the evacuation port is considered to
pertain to the insulation.
The laminar radiation shield consists of 24-26 layers of spirally wound aluminum foil
interleaved with glass paper.
The sections are fully manufactured, evacuated and sealed in the factory. There is no
evacuation on site expected. The shipping vacuum level is < 1 Pa, the operating vacuum is
expected to be of 100-1000x lower magnitude (0.01 - 0.001 Pa).
The rest gas capturing system comprises molecular sieve for absorption of most of the
vacuum rest gas, which is mostly water and air, as well as the gases released to vacuum by
construction materials or penetrating there from the ambient. Further, it comprises
palladium oxide as a hydrogen converter.
Doses or these getters are installed in each section.
Each section can also be optionally provided with a Hastings DV-6 vacuum sensor (range
0.133 to 133 Pa) for quick vacuum checks.

203

3. PIPE JOINTS
The individual sections are connected preferably by means of bayonet couplings (Fig.1).

B
A

For DN15 and DN25 lines the bimetallic bayonets are proposed. The tip, or nose, of the
male bayonet half is made from Invar 36, the female counterpart is from stainless steel.
When cooled down to the cryogenic temperature, the stainless steel tip shrinks against Invar
material, which stays nearly unchanged. The bayonet seal is thus achieved at its nose, the
O-ring at the flange only prevents moisture from penetrating into the inner bayonet space.
Both bayonet halves are fixed with a clamp.

a)
b)
Figure 1., Bayonet joints female (a) and male (b)
The larger bore piping will use close tolerance bayonets (Johnston style). The fluid
pressure is kept by the O-ring that is protected from the cold by the gas cushion. The gas
layer in between both bayonet parts shall be very thin to prevent thermal oscillation
(pulsation of the liquid/gas in the gap).
Both the above described types of joints minimize the installation work on site and the
project lead time. On-site joints (Fig. 6. and 7) with locally perlite-insulated chambers
offer the lowest cost solution, but the construction work lasts longer and the work exposed
to weather may result in lower quality. The chambers are made from pipes of larger
diameter than the outer pipe of the pipeline. After the overhanging inner pipes are welded
together, the chamber pipe is set on and welded to vacuum tight chamber, filled by perlite
and evacuated. Some on-site joints are mostly required even with bayonet joint system for
compensation of tolerances in distances and for easier assembly of the system.
4. MODULAR VACUUM INSULATED PIPE
Modular Vacuum Insulated Pipe (MVIP) is a prefabricated interlocking system for
cryogenic liquid service. Piping modules are connected with vacuum insulated bayonets for
a simple and flexible system installation.
The following modules are available:
a) Straight lengths
b) Flexible module lengths
c) Vacuum Insulated Valve module
d) Cryovent module
e) Elbow module
f) Tee modules (male branch or female branch)
g) Drop modules with internal liquid trap
h) Bayonet adapters to adapt to Chart Vacuum Insulated Pipe design

204

Figure 2. Modular pipeline

5. PLANNING FOR VACUUM INSULATED PIPELINE SYSTEMS
The vacuum insulated pipeline system is assembled from sections that are fully factory
manufactured. It means there is very little flexibility in this type of assembled lines. The
piping routing and joint positions have to be carefully planned. Some degree of freedom
can be obtained by using flexible hose sections, or field welded and evacuated joints.
From the process point of view, the VIP transports saturated liquid or even two-phase flow.
The piping routing and dimensioning shall be planned for with this in mind (e.g. avoiding
gas traps, vents use and distribution, etc.)
6. INSTALLATION
The VIP installation is very straightforward. Normally available supports and hangers can
be used. Bayonet couplings require no welding on site and normally are preferred over the
field joints. These are more laborious to make, however they could be of lower cost (say at
diameters of DN50 and more), and provide additional flexibility for mismeasurements in
planning and ease of installation.
7. APPLICATIONS OF SYSTEMS
Turn-Key packages of vacuum insulated pipelines are delivered for typical process
installations:
Storage of cells and tissues
Long term storage of cells and tissues for artificial insemination, for transplants and other
genetic and biological applications require liquid nitrogen temperature. Liquid nitrogen is
distributed from the liquid nitrogen storage tanks to individual cryo-bio storage tanks.
Electronics industry
Inert atmospheres are required at some manufacturing procedures in microelectronic
production. Continuous purging of nitrogen in small quantity is needed at individual
manufacturing stands. For example, Chart Ferox delivered a turn key installation of 80
meters with phase separators and cryogenic vacuum insulated valves to a microchip factory
at Roznov, CR.

205

Nitrogen Injection:
Liquid Nitrogen Injectors deliver
competitive advantages to food
and beverages manufacturers.
Nitrogen gas in and around the
product in a container displaces
the oxygen normally in the
atmosphere. Products packaged
without hot process sanitation
will degrade in the presence of
oxygen.
Nitrogen
inert
atmosphere prevents it. By
dropping liquid nitrogen into the
drink, oxygen is displaced and a
protective
atmosphere
of
nitrogen prevents oxidation and
consequential degradation of the
Figure 3. Typical lay-out of a vacuum insultated
drink.
pipeline application in an industrial installation

Oxygen Exclusion Applications

in
solid products:
Granular, round and irregular shape products such as: candy, nuts, potato chips, pills, and
coffee, are saturated by nitrogen during processing for displacement of oxygen.
Pressurization Applications:
Beverages including water, juices, sweet drinks, low carbonation drinks. Usually bottle or
can rigidity is necessary for bottle handling, distribution, and vending. A drop of liquid
nitrogen is injected into the drink shortly before sealing, which makes overpressure by its
rapid evaporation.
Typical accessories, used in the complete installation systems are:
Vacuum jacketed valves. Original Chart product is currently manufactured in cooperation by a subvendor.
Phase separators for removal of vapor from the two phase stream after reduction of
liquefied gas pressure, when the liquid is transferred at its boiling point and
throttled by the pipeline pressure drop. The APPS 160 (Adjustable Pressure Phase
Separator) is used to lower the saturation point of liquid nitrogen. The APPS 160
will take liquid from a bulk tank, where the liquid is stored at a higher pressure
(8,3 bar (120 psig) for example) and recondition it to the lower pressure, required
for example, for test chambers (3,4 bar (50 psig) for example). This is achieved by
separation of vapor from liquid and venting the vapor to the low pressure space.
Liquid nitrogen injectors, which are dosing small quantities of liquid or gaseous
phase for local cooling or inerting
Flexible hoses are used for compensation of distances or connection of
components, which can move during the use of the system.

206

Figure 4. Example of an industrial installation

8. LNG APPLICATIONS
Larger DN pipelines are often used for unloading or transfer of LNG at LNG terminals and
satellite stations. 1 km long VIP was built by Chart at Trinidad for transfer of LNG from
liquefier storage tanks into ocean-going LNG carriers.
Chart Ferox delivered an LNG re-fueling station for LNG fuelled ferries in Norway. A 100
m long DN 150 pipeline had to be built under a road to a sea shore. The transfer capacity is
1700 liters per minute. With respect to corrosion and impossibility of maintenance in an
underground channel, the Vacuum insulated option was chosen for the project. The pipes
are connected with on site joints with respect to the method of assembly inside the channel
(Fig. 5 to 9)

207

Fig. 5. Twelve 10 m long

sections were delivered to
the site.

Fig. 6.
VIP DN 150 in the channel

Figure 8. From LNG storage

tanks to the undergroung
channel entry. Transition form
bare and polyurethane insulated
pipes to vacuum insulated pipe.

Fig. 7 On-site joint with

vacuum port

Figure 9. During transfer of LNG:

Black part: The polyurethane foam insultated pipe in
waterproof jacket winding.
White part: Bare pipe, frosted from atmospheric humidity
Grey part: Outer jacket of the vacuum insulated pipeline,
No frost nor condensation.

CONCLUSIONS

208

Liquefied gases can be transported in several piping systems. Each system is justifiable
under certain conditions.
Delivery of liquefied gases with minimum losses can be most efficiently done by using
static vacuum insulated pipelines. This justifies VIP pipeline suitability for continuous or
periodical transfer of smaller quantities, as it is typical at cryogenic storage, for
refrigeration of continuous processes, etc. Another important advantage is resistance to
weather and other ambient corrosion conditions. Bayonet joints make possible quick on site
assemblies of the complete systems.

209

USE
OF LOW TEMPERATURES
IN INDUSTRY

211

CR08-57
THERMODYNAMIC STUDY OF THE SIMULTANEOUS
PRODUCTION OF ELECTRICAL AND COOLING POWER FROM
LNG
Parise J.A.R.1 , Esteves A.D.S.1
1

Pontifcia Universidade Catlica do Rio de Janeiro, 22453-900, Rio de Janeiro, Brazil

ABSTRACT

The paper studies the use of LNG for electrical power production. Typical regasification
systems, ORV (open rack vaporizers) and SCV (submerged combustion vaporizers), do not
envisage the possibility of recovering the cryogenic energy of the Liquefied Natural Gas
(LNG) for producing refrigeration power, which may be needed down the electrical power
utilization chain. A simple thermodynamic analysis compares the energy utilization
efficiency of a LNG electrical and cooling power cogeneration arrangement with
conventional thermo-electrical power systems making use of LNG as fuel. Energy
conservation fundamental principles are applied in the analysis to provide a description of
the thermal behavior of the system, in terms of the electrical and cooling power demands.
The energy efficiency of the cogeneration scheme is compared to those of traditional
regasification systems, in terms of the electrical power to cooling load ratio. Typical system
configurations, capable of matching cooling and electricity demands in a realistic situation,
were devised for this preliminary analysis.
1. INTRODUCTION
Usage of natural gas as a primary energy source is growing in importance, due to large
proven reserves and to its relatively lower air pollution and emission of greenhouse gases,
if compared to other fossil fuels. It is expected that natural gas will account, by 2020, for
30% of the total electricity production. Liquefaction of natural gas (by condensation to a
cryogenic temperature of about -161oC) makes it possible to convey this energy source
from remotely located reserves to the consuming markets. The maritime transportation of
LNG, in double-hulled carriers, competes with long distance pipelines, now contributing to
nearly a quarter of worldwide gas exports [1]. At the import terminal LNG is regasified,
where several types of LNG vaporizers are commonly used. The following five types have
either been used or demonstrated in LNG receiving terminals [2]: (i) Open Rack Vaporizers
(ORV), (ii) Submerged Combustion Vaporizers (SCV), (iii) Shell and Tube type
Vaporizers (STV) including modified designs such as the Reli-Vap type vaporizer, (iv)
Combined Heat and Power unit with Submerged Combustion Vaporizer (CHP-SCV), and
(v) other types, such as Ambient Air-Heated Vaporizers [3, 4]. LNG receiving terminals
commonly use one of two types of LNG vaporizers: the ORV and the SCV. In general, the
ORV system uses seawater as the heating medium. It has a lower operating cost than the
SCV, but normally a higher capital cost due to the vaporizer equipment, the added seawater
intake/outfall system, the large diameter seawater pipes, and the seawater pumping and
treating systems The SCV requires fuel for the LNG vaporization, and the fuel consumption
is about 1.5% of the send-out [2]. A common characteristic of these methods is that they
have room for significant improvement on energy utilization, since no recovery is made of
the cryogenic energy of LNG. Moreover, environmental impacts are expected from the
combustion of the fuel gas for vaporization (SCV) and from the reduction of sea water
temperature (ORV).
213

Two papers, among others, can be found in the literature reporting on studies of
arrangements that recover part of the energy consumed in the liquefaction process of
natural gas. Kaneko et al [5] make use of a combined cycle, comprised by a conventional
gas turbine, working as the topping cycle, and an inverted Brayton cycle, for bottoming
purposes. The authors claim a superiority of this system, if compared to traditional ORV
systems, in terms of thermal efficiency and specific output. Deng et al [6] propose a new
cogeneration power system, with two energy outputs rates (electricity and refrigeration),
making use of both chemical and cryogenic energies of LNG.
This paper presents a basic study on the production of refrigeration power from the
regasification process of LNG-fuelled thermal power plants for electricity production. The
analysis is based on energy conservation fundamental principles and on the fact that, within
the range of consumption of the produced electrical power, there is potential for
concentrated refrigeration power demand, like from large-capacity refrigerated storage
spaces.
2. SYSTEM DESCRIPTION
Three systems are considered for the production, from LNG, of electrical and refrigeration
power. The first two systems, figures 1 and 2, are based on traditional thermal power
plants, meeting electrical power demand as well as the electrical power consumption of a
vapour compression chiller. Natural gas, produced from traditional regasification methods,
ORV (figure 1) or SCV (figure 2), fuels the thermal power plants. The third system studied,
figure 3, is based on cogeneration, where the refrigeration power is produced directly from
the regasification process, thus diminishing the power required from the chiller. Neither
heat recovery nor heat demand are considered in this study. Main components, or plants, of
the system are: the thermal power plant (TP), the vapour compression chiller (VC) and the
regasification plant (RG). The thermal power is characterized by an overall thermal
efficiency, t , and the vapour compression chiller, by the refrigerating coefficient of
performance, COP . Two demands are to be met: the electrical and the refrigeration power
loads, E load and Q ref , respectively. The LNG stream is represented by an energy rate
equivalent of LNG consumption, H
, which, after regasification, converts to the energy
LNG

rate equivalent of natural gas consumption, H NG . Processes at the thermal power plant,
chiller and regasification, involve unrecoverable rates of heat gain or loss, denoted by Q .
l

214

Thermal
Power
Plant

NG

LNG

Vapor
Compression
Chiller

Electricity

Regasification
Scheme
(ORV)

Refrigeration

Figure 1: Simplified energy flow diagram of an electricity and refrigeration power production plant
from LNG, using an ORV regasification scheme

Thermal
Power
Plant

NG

Vapor
Compression
Chiller

Electricity

Regasification
Scheme
(SCV)

LNG

Refrigeration

Figure 2: Simplified energy flow diagram of an electricity and refrigeration power production plant
from LNG, using an SCV regasification scheme

215

Thermal
Power
Plant

NG

Regasification
Scheme
(Cogeneration)

LNG

Vapor
Compression
Chiller

Electricity

Refrigeration

Figure 3: Simplified energy flow diagram of an electricity and refrigeration power production plant
from LNG, using a cogeneration regasification scheme

The compressor power consumption of the chiller is W CV . Specifically, the ORV

regasification scheme, figure 1, has an overall seawater-to-LNG heat transfer effectiveness,
RG . By its turn, the SCV regasification scheme, figure 2, is characterized by the ratio of
imported gas used as fuel gas for LNG vaporization, RG . Finally, the cogeneration scheme
of figure 3 allows for two possibilities: (a) the heat transfer rate for the vaporization of
LNG, Q RG , is greater then refrigeration power demand, Q ref , which means that
refrigeration power demand is met but heat transfer rate from an additional heat source,
Q s , RG , is required to fully vaporize the LNG stream; otherwise, (b) a vapour compression
chiller, of smaller capacity, if compared to non-cogeneration schemes, complements the
LNG vaporization load, Q RG , with the evaporator power capacity, Q VC , in order to meet
the refrigeration load, Q .
ref

3. THERMODYNAMIC MODEL
A set of parameters is taken from previous studies on heat recovery [7] and cogeneration
systems [8]. The energy conversion ratio, ECR , is here defined as the total energy delivery
rate (electrical and refrigeration power) divided by energy input rate from LNG. In other
words, ECR is the ratio between the sum of all energy products and the total energy
consumption.
ECR =

Q ref + E load
H
LNG

Two other non-dimensional ratios, RCE and RCV , compare the magnitudes of the cooling to
electricity loads and of cooling load to LNG vaporization, respectively.

216

RCE =

Q ref
E

load

RCV =

Q ref
Q

RG

Dividing equation by E load and introducing equation , one obtains:

Q ref
+1
E
ECR = load
=
H LNG
E
load

RCE + 1
H LNG
E load

The ratio between LNG energy equivalent consumption rate and electrical load will depend
on the characteristics of the regasification system employed. Each regasification scheme,
figures 1 to 3, will be dealt with separately. First, the energy balances applied for the
thermal power plant and the vapor compression chiller, common to all three schemes, are,
respectively:
W CV + E load = H NG t
Q ref = COP W VC
3.1 ORV Scheme
Applying an energy balance applied to the ORV regasification plant control volume,
depicted in figure 1, provides:
H NG = H LNG
Substituting equations and into and dividing it by Q ref , the energy conversion ratio is
obtained in terms of RCE and system characteristics, COP and t , respectively:
ECRORV =

( 1 + RCE ) t
1+

RCE
COP

3.2 SCV Scheme

The use of imported gas as fuel for the vaporization of LNG reduces the flow of energy to
the thermal power station by:
H NG = ( 1 RG ) H LNG
With the substitutions indicated in section 3.1, the energy conversion ratio becomes:
ECRSCV =

( 1 + RCE ) ( 1 RG ) t
1+

217

RCE
COP

3.3 Cogeneration Scheme

Two situations are to be considered, regarding how the required vaporization heat rate,
Q RG , compares with the refrigeration power demand, Q ref , equation . If RCV 1 , i.e.,
Q RG Q ref , it is assumed that the additional heat supply rate for LNG vaporization, Q s , RG ,
is provided by an ORV regasification system and, of course, a supplemental chiller is not
required, i.e., W CV = 0 . On the other hand, if RCV > 1 , i.e., Q RG < Q ref , the supplemental
chiller will come into effect. For both cases, equation applies. Development of the energy
conversion ratio equation, for both cases, follows:
a) Refrigeration load greater than LNG vaporization load, RCV > 1 :
Equation applies. An energy balance applied to the vapour compression chiller control
volumes provides:
Q VC = COP W VC
Taking equation into and dividing it by E load , yields:

H LNG 1 1 Q VC
=
+ 1
E

t COP Eload
load

In the cogeneration scheme, the chiller only accounts for the refrigeration power that is not
met by LNG vaporization:
Q VC = Q ref Q RG
Taking equations and into equation and
H LNG 1 1
Q
=
RCE RG + 1

E load t COP
E load
Writing Q RG E load in terms of the load ratios, equations and , and taking equation into , the
energy conversion ratio equation becomes:

( 1 + RCE ) t

ECRCOG ( RCV > 1) =

1+

RCE
1
1

COP RCV

b) LNG vaporization load greater than refrigeration load, RCV 1 :

The chiller does not operate,
W CV = 0
so that all the electrical power produced by the thermal power plant goes to the electrical
power load:
E load = H NG t
Taking equation into :
218

ECRCOG ( RCV 1) = ( 1 + RCE ) t

4. RESULTS
Equations , , and were applied to provide an overview of the first-law performance in
terms of the system characteristics and of the energy demands (refrigeration and
electricity). Typical values for the system characteristics were chosen as follows:
RG = 0.50 , RG = 0.015 [2], t = 0.40 (Rankine cycle) and COP = 3
Figure 4 shows how the energy conversion ratio varies with the two load ratios. One
observes the superiority of the cogeneration scheme. The least energy efficient
regasification arrangement is, of course, the SCV, because of the consumption of a fraction
of LNG stream to convey vaporization, thus reducing ECR , in relation to ORV, by a
factor of ( 1 RG ) , equation . The greater the cooling to electrical load ratio, RCE , the
greater the energy conversion ratio. Oppositely, greater energy conversion ratios are
obtained when the LNG vaporization load surpasses the refrigeration load, RCV 1 , as no
need of is made the electricity consuming chiller. Figure 5 shows the same results plotted in
a different manner, emphasizing the reduction of the energy conversion ratio of the
cogeneration scheme when the chiller comes into operation, in order to meet the
refrigeration load. The limit of the curves, for very large values of RCV , is the energy
conversion ratios of the traditional regasification processes, in the present simulation, the
ORV scheme. Such limit case would be the situation when the refrigeration load would be
far greater than the vaporization load.
1,4

ECRCOG

ECR

1,2

RCV 1

RCV = 1,5
RCV = 2

0,8

ECRORV

0,6

ECRSCV

0,4
0,2

0,4

0,6

0,8

1,2

1,4

1,6

1,8

2,2

2,4

RCE
Figure 4: Variation of the energy conversion ratio for different regasification schemes, with varying
cooling to electricity load and cooling load to LNG vaporization ratios, RCE and RCV , respectively.

219

1,3
1,2

RCE=2

ECR

1,1
1

RCE=1,5

0,9
0,8

RCE=1

0,7
0,6
0,5
0,2

RCE=0,5

0,4

0,6

0,8

1,2

1,4

1,6

1,8

RCV
Figure 5: Variation of the energy conversion ratio for different regasification schemes, with varying
cooling to electricity load and cooling load to LNG vaporization ratios, RCE and RCV , respectively.

5. CONCLUDING REMARKS
The present analysis has proven, as expected, that the recovery of the cryogenic energy of
LNG, in order to match a certain refrigeration load, can provide reasonable improvement
on the energy utilization of LNG, here measured by a proposed non-dimensional parameter,
the energy conversion ratio. Moreover, environmental impacts of the traditional schemes,
like seawater temperature reduction (ORV) or additional CO 2 (SCV) emission, could be
lessened with the application of a cogeneration scheme. However, attention should be
brought to the fact that the present analysis did not take into consideration certain
thermodynamic aspects, such as temperature requirements in the recovery arrangements, as
well as technical, economical, cost-effectiveness and eventual operational limitations that
may arise from the implementation of such scheme. For instance, a LNG receiving terminal
with a C2+ or C3+ separation facility can import LNG feeds with varying compositions, in
order to meet stringent calorific value export gas specifications and to decrease capital and
operating costs. Studies found in the literature shows that process schemes for extracting C 2
or C3 from rich imported LNG are feasible, effective, and economical [9]. The presence of
such facilities may, then, impact the overall balances as proposed above. Furthermore,
environmental changes due to global warming have caused various kinds of serious
problems on a global scale. From this standpoint, the LNG industry must reduce harmful
effects on the environment with the LNG related working chain, which includes
exploitation, liquefaction, transportation of a natural gas and re-gasification issues [10].
Further study on the subject is recommended.
ACKNOWLEDGEMENTS
The authors are indebted to CNPq (from the Brazilian Ministry of Science and Technology)
and to FAPERJ (State of Rio de Janeiro Research Funding Agency) for the financial
contribution to this study.

220

REFERENCES
1. Chrz, V., Liquefied Natural Gas: Current Expansion and Perspectives, 19th Informatory
Note on Refrigerating Technologies, International Institute of Refrigeration, Paris,
France, November (2006)
2. Yang, C.C., Huang, Z., Lower Emission LNG Vaporization, LNG Journal (2004)
November /December 24-26
3. Hubbard, B.S., Floating Storage and Regasification Concepts Using the LNG Smart Air
Vaporization Technology, AIChem 2007 Spring National Meeting, Houston, USA
(2007)
4. Davis, J.F., Understanding Ambient LNG Vaporizers, AIChem 2007 Spring National
Meeting, Houston, USA (2007)
5. Kaneko, K., Ohtani, K., Tsujikawa, Y., Fujii, S., Utlization of the Cryogenic Exergy of
LNG by a Mirror Gas-Turbine, Applied Energy (2004) 79 355-369
6. Deng, S., Jin, H., Cai, R., Lin, R., Novel Cogeneration Power System with Liquefied
natural Gas (LNG) Cryogenic Exergy Utilization, Energy (2004) 29 497-512
7. Parise, J.A.R. and Cartwright, W.G., Experimental Analysis of a Diesel Engine Driven
Water-to-Water Heat Pump, Journal of Heat Recovery Systems and CHP, (1988)8 75-85
8. Herbas, T.B., Dalvi, E.A., Parise, J.A.R., Heat Recovery From Refrigeration Plants
Meeting Load and Temperature Requirements, International Journal of Refrigeration,
(1990) 13 264-269
9. Yang, C.C., Kaplan, A., Huang, Z., Cost-effective design reduces C2 and C3 at LNG
receiving terminals, Oil & Gas Journal, (2003) 101 issue 21
10. Kajitani, M., Efforts to minimize the environmental load at LNG receiving terminals,
23rd World Gas Conference (2006) Amsterdam

221

CR08-51
GAS IMPURITIES FREEZING OUT TECHNOLOGIES.
Klepal J., Stoek P.
ATEKO Hradec Kralove Czech Republic
ABSTRACT
An experimental testing study verified the influence of small amount of epichlorhydrine
vapour on character of the ice accretion formation during toluene freezing out from the
nitrogen / toluene / water vapour mixture.
A toluene freezing out spiral wound heat exchanger with glass shell was installed into the
testing line. Forms of the ice accretions were observed optically.
Epichlorhydrine impurities changed character and density of the ice accretion.
INTRODUCTION
One of methods used for separating impurities from off gases from chemical production
plants consists in freezing out the impurities. However, the effectiveness of this process is
dependent on the one hand on the concentration of an undesirable substance in a gas and on
the other hand on the fact whether or not are present at the same time other impurities or
other components (water vapour). In particular water vapour presence influences markedly
the form of obtained substance frozen out from a off gas. Mostly they are so called hydrates
formed by some hydrocarbons in the presence of water at a low temperature and at a high
pressure eventually.
Particularly the form similar to hydrates formed at freezing out toluene vapours from off
gases from the production line was the subject of the experimental measuring as described
hereinafter.
1. EXPERIMENTAL APPARATUS AND PROCEDURE
A schematic diagram of the apparatus used in this work is given in Figure 1. The main
component of the apparatus consists of a spiral wound vertical heat exchanger. The
exchange is made by upward spiral pipe with a diameter of 6 x 1 mm wound with a
diameter of 41 mm with a pitch of 20 mm.
The air with impurities was fed to the bottom end of the heat exchanger through a stainless
steel tube with diameter of 17 mm fitted in the centre of the winding. For visual observation
of the degree of formation of ice accretion on the wound heat transfer surface the heat
exchanger was built in a glass tube with a diameter of 54 mm with the bottom end closed.
The glass tube with the wound tube exchanger was placed into a visual glass Dewar vessel
for eliminating heat losses.
In the upper part the Dewar vessel was closed as a protection against penetration of the
ambient atmosphere. The inter space of the heater was cooled with freezed ethanol. The
circulation of ethanol is ensured with the help of a circulation pump and by LIN freezed
bath filled by ethanol. The temperature of the circulating coolant was maintained at a value
of 25 C with the help of liquid N2. The off gas for freezing out was fed into the bottom

223

L
egend: 1. Gas INPUT 2. Pressure reducing valve, 3 Gas flow meter, 4. Water saturation, 5. Toluene
saturation, 6. Demister, 7. Glass Dewar vessel, 8. Experimental heat-exchanger, 9. Freezed ethanol
bath

part of the shell side of the heat exchanger through the tube fitted in the axis of the heat
exchanger winding. As a carrying gaseous fluid it was used compressed air treated in a
reduction station to a lower pressure. The gas was led through a gas meter and two vessels
designed to serve for saturating the air with water and toluene. In this circuit it was installed
also a separator with drop and overrun catcher.
In the course of the measuring they were read temperatures of the freezing out fluid and
temperatures of the leaving air together with air rate of flow.
The proper experimental verification of the freezing out process was divided in two
experiments. In the course of the first experimental verification (Fig. 2) it was verified the
course of freezing out of the pure toluene from
the air and in the other experimental
verification (Fig. 3) the course of freezing out
of toluene from the air in the presence of
impurities. In this case they were trace
amounts of epichlorhydrine (1-chloro-2,3epoxypropane) and 2,3-dichloro-1-propen.
2. RESULTS AND DISCUSSION
Already from the beginning of the measuring
in both the cases a growth of ice accretion
occurred in the form of needles on the proper
winding of the heat exchanger tubes and then
also on distance of wind.
A part of toluene presented in the air
condensed in the feeding central tube and in
the form of a condensate flown down to the
bottom of the glass cylinder (Fig. 1).

224

Fig. 1

The other part was then absorbed by the arisen ice accretion formed by hydrates on the heat
exchanger winding (Fig. 2, 3).
Both the experiments ware carried out up to a complete freezing and clogging of the heat
exchanger winding. Then the temperature of the circulating cooling liquid was increased
and in the course of about 20 minutes all ice accretion was removed from the heat
exchanger winding.
In the course of the experiments they were
caught gradually liquid portions from single
phases in the course of the freezing out and
also in the course of thawing of the ice
accretion resulted. After a stabilization and
separation of single liquid components of the
two-phase system (water + toluene) the
measuring of volumes caught was carried out.

Fig. 2

Fig. 3

By carrying out a balance of inlets to the experimental equipment and outlets from the
experimental equipment it has been found the effectiveness of the process of separating
toluene from the air, in the first case (pure toluene) about 24 % and in the other case
(contaminated toluene) about 43,5 %. A value of the theoretical efficiency of 48% has been
calculated on the basis of a balance computation made with the help computing programme
CHEMCAD with the use of a model for ideal gases.
3. CONCLUSIONS
Experimental results have shown that the formation of freezed out impurities (hydrates) in
the course of freezing out influences mainly the character of ice accretion arising on the
heat exchanger winding. The structure of the ice accretion arisen has not been characterized
by a compact ice layer but rather as a layer of a wet snow. The influence of impurities has
manifested itself as a richer form of ice accretion with the ability to catch more the liquid
toluene in its structure in the same time period when compared to the freezing out of pure
225

toluene (compare Fig. 2 and 3). The resulting needle-like ice accretion (consisting mainly
of water) is able to catch the condensing toluene and in this way it is increased its specific
mass but also thermal conductivity.
In the course of the final comparison of measured values with values calculated on the basis
of simulation programme CHEMCAD it has been attained a very good conformity of single
parameters observed.
The results have been then used for optimising industrial plant in the line for liquidating off
gases from resin production plant.

226

CR08-49
MULTISTAGE CRYOGENIC TREATMENT OF MATERIALS:
PROCESS FUNDAMENTALS AND EXAMPLES OF APPLICATION
Alava L.A.
Cryobest International, S.L. Vitoria, Spain
ABSTRACT
The multistage cryogenic process is an evolution from the conventional cryogenic
treatments of materials. It needs shorter process time achieving the same or even better
results.
This paper introduces some basic fundamentals of these treatments, their effects and
applications, the equipment, etc. Some examples with different materials and from different
industrial sectors are also presented as well as some brief comments about R&D and future
trends of the technology.
INTRODUCTION
People usually relates heat treatments with high temperatures, but thermal treatments can
also involve cooling. Although it has been traditionally considered that deep cold
temperatures have no permanent effect on the materials, it is not true at all.
Heat treatments were already known and used centuries ago but the access to really low
temperatures was only possible in relatively recent days. Although the first experiences
took place at the beginning of last century, it is not possible to properly speak about
industrial cryogenic treatments of materials until the 70s when the liquefied gases became
more affordable and the treating equipment had more accurate process control systems.
During the 80s and 90s the use of this technology increased and some treating facilities
were opened, mainly in the US. Nowadays it is possible to find cryogenic processing
companies in many countries all over the world.
Although still hardly known and used in Europe, the cryogenic treatment of materials is a
technology that is slowly getting acceptance in industry. It basically consists in submitting
the materials to deep low temperatures for increasing some of their performance
characteristics like wear resistance or fatigue life.
1. THE PROCESS
Cryogenic treatments basically consist in submitting the materials to deep cryogenic
temperatures (below 120K) following predetermined time-temperature curves in order to
enhance some of their physical or structural properties.
According to the previous definition, it must be noticed that the subzero processes (at about
-80 C) that are used in many traditional heat treating facilities to reduce the austenite
content in some tool steels cannot be considered cryogenic treatments.
1.1 Conventional cryogenic treatments
There is no a standard process for cryogenic treatments but most of them are quite similar.
In conventional cryogenic treatments the materials are slowly cooled down to a temperature
around -180 C and maintained for a period of time that lasts from eight hours to two or
even more days. After the soak, the materials are slowly heated up to ambient temperature.
227

Sometimes the treatment is completed with a soft tempering. The entire process typically
needs two to three days to be completed.
There is a conventional process sub-category called wet process where the soak is made
by submersion in liquid nitrogen. Anyway, the previously described dry process (no
liquid nitrogen in the chamber) is more widely used.
The cryogenic treatments are performed in chambers designed for this purpose. The
material is usually cooled using liquid nitrogen that is introduced in the processor through
solenoid valves controlled by computer. Most of the modern chambers have heaters that
also allow to control the temperature during the heating phases of the process.
1.2 Multistage cryogenic process
The multistage cryogenic treatment is a more advanced process that has been developed as
an evolution from the conventional ones. In this treatment the isothermal soak at cryogenic
temperature is substituted by several cryogenic cooling/heating phases. This process is
more effective but its main advantage is that it is much faster (an average of fifteen hours
for the whole process) than the conventional ones.
The cryogenic chambers that are used to apply a multistage cryogenic treatment are
specially designed to perform this type of process.

Figure 1: Multistage cryogenic processor designed and manufactured by Cryobest International, S.L.

2. THEORIES AND FUNDAMENTALS

The cryogenic treatments and their applications have been developed mainly in an empiric
way. This technology is becoming widely accepted and used in industry but there are still
certain controversy and some mystery concerning the effects of deep cryogenic
temperatures on the material microstructure. Some studies have been performed in recent
years in this field and more are currently in progress all over the world.
Metallurgists know that, when a steel is quenched, usually the higher the carbon content the
lower the temperature (Mf) at which the transformation of austenite into martensite finishes.
In high carbon steels, cold temperatures lead to higher contents of martensite and therefore
to harder and more stable structures (more desirable in most of the applications). More
recently it has also been confirmed that cryogenic temperatures promote the precipitation of
fine -carbide particles in the steel matrix contributing to improve the material
characteristics.
These facts could explain some of the improvements in material performance due to its
submission to a cryogenic treatment but, basically, they are only valid for steel. Although
228

many grades of steel (alloyed, cold working, hot working, HSS, stainless) can be
cryogenically treated, the cryogenic processes have also evident effects in other materials
like casting, cemented carbide, cooper alloys, aluminium alloys, titanium, some ceramics
and even certain polymers. That means that there must be something else, more general,
that explains the changes in the properties of the materials.
Recent theories point to stress relieving in the microstructural level and more stable,
continuous and homogeneous lattices as a key factor to achieve improved behaviours in the
service performance of cryogenically processed materials. This point of view is becoming
quite accepted nowadays but it must be more deeply investigated.
3. EFFECTS AND APPLICATIONS
3.1 Cryogenic treatment effects
We have seen that there is a wide range of materials that can be cryogenically treated with
good results. The effects of these treatments depend on the material but, in general, it is
possible to obtain several of the following improvements:
- better wear resistance
- improved fatigue life
- stress relieving and dimensional stability
- increased conductivity
- improved machinability
- slight increase of hardness
- better corrosion resistance
Apart from the material, the results depend on the considered application. Of course, the
cryogenic treatments are not a cure-all but there are innumerable situations where the
improvements are significant or even impressive. Some examples will be commented later.
3.2 Industrial applications
It is possible to find applications in practically every industrial sector: machining, casting,
injection moulding, forging, welding, automotive, aerospace, electronics, steel, timber
industries, mining, agriculture, motorsports, etc. Some examples of parts that can improve
their performance and increase their lives are: knives, cutting tools (drill bits, carbide
inserts, mills, hobs, broaches), saws, punches, dies, rolls, moulds, electrodes, gears,
shafts, bearings, springs, cables When a wear or fatigue problem occurs or more life is
needed there is usually a good chance for using cryogenic treatments.
Cryogenic processes do not substitute conventional heat treatments although sometimes
could be considered as an extension of them. One important characteristic is that they
permanently affect the whole mass of the components, not just the surface. If, for example,
a cutting blade is cryogenically treated, it can be sharpen as many times as desired without
loosing its improved performance. Another point to be taken into account is that cryogenic
treatments are fully compatible with most of the surface treatments and coatings (nitriding,
PVD, CVD, etc.) that are commonly used in industry to enhance the working performance
of tools and components.
It is important to remark that cryogenic treatments are environmentally friendly. Absolutely
no waste or residues are produced during the process. And even more, the use of this
technology allows for significant reductions in energy and materials consumption.

229

4. EXAMPLES OF APPLICATION
The number of applications of this technology is practically unlimited. Many of the most
typical applications of cryogenic treatments are in the field of perishable tools. Machining
tools are usually made of different grades of HSS or cemented carbide (and very often are
PVD or CVD coated). These materials usually react very well to the application of a
multistage cryogenic treatment. Furthermore, machining tools are often sharpened (and
sometimes also coated) several times during their life. As this processes only need to be
performed once this is a key advantage of this technology in many applications.
Cryogenic treatment is a good way to achieve important cost reductions and improved
productivities in gear making processes. A well-known automotive supplier company can
certify it. In one of the plants they manufacture steering systems for cars and trucks. The
truck steering racks manufactured in this facility are machined with specially designed
Maag type cutters (Fig. 3). These cutters are made of ASP2030 (powder-metallurgical
HSS) and are coated with TiN every time they are sharpened.

Figure 2: Maag type cutters (TRW, Spain)

In this application the use of the multistage cryogenic treatment has made a clear
difference. Without the treatment, an average of 60 parts were machined between
sharpening. Nowadays all the new cutters are treated and the average production is 160
parts between sharpening. Not only this; as the wear is more homogeneous, it is only
necessary to remove 0.2 0.3 mm. of material to sharpen the cutters (0.3 0.4 mm. without
the treatment) what means about 30% extra uses of the tool. In this case the savings due to
the use of the cryogenic processes are really important.
Another typical application of this technology is the treatment of gear cutting tools like
hobs. A company specialized in the manufacturing of flywheel starter ring gears for the
automotive industry uses inserted blade hobs. These hobs are made of HSS (M35) and
coated with TiN. This kind of tools cannot be coated every time they are sharpened because
the coating temperatures are too high for the resins that accurately fix the blades in its
position. Nevertheless, the multistage cryogenic treatment avoids this problem and, in this
case, the cryogenically treated hobs cut between 50 % and 100 % more ring gears than the
untreated ones.
A well known aircraft manufacturer uses thousands of drill bits to make holes in difficult to
machine materials like stainless steel, titanium or nickel alloys. After some months doing
tests with HSS and solid carbide tools they could check that the life of the cryogenically
treated drill bits was, as an average, three times longer than the untreated ones (in some
cases even five times longer). Obviously, nowadays this company is achieving important
tool cost savings thanks to the use of the multistage cryogenic treatment.
230

Figure 3: Insert blade hobs after a multistage cryogenic treatment.

Timber and wood industries as well as pulp and paper industry can also benefit from the
use of multistage cryogenic treatments. Just as an example, the big knives that are used to
convert the wood logs in small chips suffer severe wear. In most pulp mills these knives
have to be grinded (after being previously removed) once and even twice a day. These
knives are usually made of HSS and increase their life between two and three times when
cryogenically treated. Taking into account that the treatment is applied only once, it is easy
to understand that a growing number of companies is using this process to reduce downtime
and tooling costs in their chipping facilities.

Figure 4: 760 mm. long wood chipping knife.

A forging plant is using the multistage cryogenic treatment for the hot forging dies. These
tools are made of hot working steel X40CrMov51 (H13). The dies are nitrided to achieve a
higher hardness value in the surface. As nitriding is a surface treatment it has to be repeated
after any die regrinding or restoration.
This company made tests using the multistage cryogenic treatment and the results were
clear: the untreated dies made an average of 1430 parts while the cryogenically treated ones
were able to make an average of 3550 (the nitrided dies have a similar performance). The
next step would be to test a combination of nitriding + cryogenic but, meanwhile, the
forging dies are cryogenically treated instead of nitrided.
A well known manufacturer of bearings uses steel rolls to laminate and calibrate the outer
rings of the bearings. These rolls can be used to calibrate an average of 3000 units before
231

needing grinding. The cryogenically treated rolls can make as much as 17000! units before
grinding. Of course, all the rolls are now submitted to the multistage cryogenic treatment.
But not only steel or carbide tools can be bettered through the use of cryogenic treatments.
Copper alloys also respond very well to the process. The resistance welding electrodes are
usually made of alloys like CuCoBe or CuCrZr that have a good balance between strength
and conductivity.
The influence of the multistage cryogenic treatment in the performance of resistance
welding electrodes has been tested in several applications finding performance increases up
to 500% in some cases. That is why nowadays there are some companies that are already
using this technology to reduce their welding electrodes consumption and to increase their
welding quality and productivity.
Other copper alloys like brass are also suitable for the multistage cryogenic treatment. A
manufacturer of rolled profiles uses forming punches and wear parts made of brass. After
two years of tests, now all the brass parts that are used in the factory are submitted to the
cryogenic process. The life of those components has increased from two to three times
compared with the untreated ones.
5. TECHNOLOGY USE AND EVOLUTION
After some decades seeking legitimacy, it seems that the cryogenic treatments will
probably become accepted and used in industry. Nowadays it is possible to find cryogenic
treatment facilities in many countries all over the world, but the growth in the use of these
technologies is slower than expected (the present situation of these technologies in Europe
is clear example).
5.1 Research needs
Cryogenic treatment fundamentals are not totally understood and this fact can explain the
delay in the complete acceptance of these processes. Fortunately this situation is already
changing due to the evident results of the treatments. Anyway, there is a clear need of
improving the research activity in this field in order to achieve a more complete
understanding of the mechanisms that are involved in the microstructural enhancement of
the deep cold treated materials and also in a search of new applications.
There has been few research activity in this field of and, traditionally, most of it has been
made in North America apart from some studies made in Europe. But during the last few
years this situation has changed and, nowadays, there are more R&D projects concerning
cryogenic treatments of materials. And, nowadays, most of the research activity takes place
in Asia.
The European commission has approved the first big cooperative project in the field of
cryogenic treatments. Its title is Improvement of automotive tools and components trough
the application of deep cryogenic treatments. This three year project was launched in
October, 2007 and the participants are research centers, universities and industrial
companies from Austria, Italy, Germany and Spain. Hopefully, more new similar project
proposals will also be presented in the near future in Europe.
Anyway, the empirical development of the technology will probably prevail in the
development of new applications of the cryogenic treatments during the coming days.
5.2 Cryogenic treatments application forecast
Until now the use of this technology in industry has been mainly focused on the cryogenic
treatment of all kind of tools and consumables: knives, saws, mills, inserts, dies, punches,
232

welding tips, moulds... This is probably the most evident use for the technology because the
increase of tooling life has a direct positive consequence in productivity and cost,
something that is always interesting for the users.
The increase in wear resistance is the prevailing treatment effect when treating tools. But
there are other treatment properties that will probably get more relevance in the coming
years and one of them has special interest: the increase of fatigue life.
Probably, during the coming years there will be more and more applications where the
cryogenic treatments will be used to increase the fatigue life of all kind of components.
Apart from the tools the process will be applied to the manufactured components in order to
increase their service life and their reliability. If this forecast is right, there will be
necessary to treat much more quantities of materials. This would be a situation where the
multistage cryogenic treatment has a clear advantage compared with the conventional ones:
the process time is much shorter and, consequently, the treating facilities are much more
productive.
CONCLUSIONS
The treatment of materials at cryogenic temperatures is a promising and cost effective
technology that, although is not new, is still hardly known and used. Only a small part of its
potential has been developed.
The metallurgical fundamentals are not fully understood and the cryogenic treatments have
been, and still are, developed in an empirical way. During the last years the research
activity in this field has significantly increased and nowadays it is also more global than in
the past.
The multistage cryogenic treatment is an evolution from the conventional ones. It is more
efficient and its process time is much shorter.
The wear resistance improvement and the increase in fatigue life are just two of the effects
of the cryogenic treatments. A wide range of materials can be cryogenically treated and the
number of applications is unlimited. They can be found in every industrial sector.
There are two main fields of application of cryogenic treatments:
- the treatment of tools is nowadays the more common application. It is a cost effective
way to increase their productivity allowing less downtime and reduced costs.
- cryogenically treating components (bearings, gears, shafts, springs) it is possible to
greatly improve their performance and reliability and also to reduce their weight and
size. This field of application will probably have a big increase in the future.
The multistage cryogenic treatment is an environmentally friendly technology that helps the
materials to perform better. No doubt cryogenic treatments are an exciting technology that
holds much future promise.
REFERENCES
1. Meng, F., Tagashira, K., Azuma, R. and Sohma, H., Role of Eta-carbide Precipitations
in the Wear Resistance Improvements of Fe 12Cr-Mo-V-1.4C Tool Steel by Cryogenic
Treatment, ISIJ International (1994) vol. 34, No 2 205-210.
2. Diekman, R., Cryogenics in the Thermoset Molding Industry, Thermosettings (2001)
vol. 29.
3. Schiradelly, R. and Diekman, F.J., The Racers Edge, Heat Treating Progress (2001)
nov. 43-50.
233

4. Molinari, A., Pellizzari, M., Gialanella, S. Straffelini, G. and Stiasny, K.H., Effect of
deep cryogenic treatment on the mechanical properties of tool steels, Journal of
Materials Processing Technology (2001) 118 350-355.
5. Huang, J.Y, Zhu, Y.T., Liao, I.J., Bourke, M.A. and Mitchel, T.E., Microstructure of
cryogenic treated M2 tool steel, Material Science Engineering (2003) A339 241-244.
6. Zhiseng, W., Ping, S., Jinrui, L. and Shengsun, H., Effect of deep cryogenic treatment
on electrode life and microstructure for spot welding hot dip galvanized steel, Materials
and Design (2003) 24 687-692.
7. Huang, M.C., Gao, C.H., and Huang, L.G. Study on cryogenic phase change & wear
characteristic of high speed steel, Acta Metallurgica Sinica, (2003) vol. 16, No. 6 524530.
8. Manoj, V., Gopinath, K. and Muthuveerappan, G., Rolling contact fatigue studies on
case carburized and cryogenic treated En 353 gear material, International Sympossium
on Material Science and Engineering, Chennai (2004).
9. Singh, P.J., Mannan, S.L., Jayakumar, T. and Achar, D.R.G., Fatigue life extension of
notches in AISI 304L weldments using deep cryogenic treatment, Engineering Failure
Analisys (2005) 12 263-271.
10. Yong, A.Y.L., Seah, K.H.W., Rahman, M. Performance evaluation of cryogenically
treated tungsten carbide cutting tool in turning, International Journal of Machine
Tools& Manufacture (2006) 46 2051-2056.
11. Latas, Z., Ciski, A. and Suchmann, P., Cryogenic Treatment and Combination of
Nitriding and Cryogenic Treatment of Hot Forging Tools, Proceedings of the 4th
WSEAS International Conference (2006) 133-139.
12. Zhirafar, S., Rezaeian, A., Pugh, M., Effect of cryogenic treatment on the mechanical
properties of 4340 steel, Journal of material processing Technology (2007) 186 298-3

234

CRYOSTORAGE
OF CELLS AND TISSUES

235

CR08-06
FROM THE TISSUE BANK TO THE TISSUE ESTABLISMENT
Mika P., Strakov H., Horynov A.
Tissue Bank, University Hospital, Hradec Krlov, Czech Republic
ABSTRACT
A review of changes in a role of a tissue bank in assuring clinical cell and tissue
transplantation is presented. At the beginning the tissue bank was regarded a place to which
the tissue collected by a surgeon was put until its use, mostly by the same physician. The
role of a tissue bank was only to extend as much as possible the shelf life of the preserved
tissue. Later the tissue banks started to introduce methods modifying the properties of the
original tissue with the aim to lower its immunogenicity as well as to enhance or to prevent
tissue rebuilding after transplantation. The chance to overcome the tissue rejection was
enlarged after introducing methods combining the autologous cultured cells, e.g. epithelial
keratinocytes with allogeneic or biosynthetic matrices. For a long period of time the
activities of cell and tissue banks were not regulated by law, only the standards of voluntary
organizations of tissue bankers have been available. The term tissue establishment was
introduced by the Directive of the European Parliament and Council issued in 2004, that
put high requirements on the safety and quality of the cell and tissue processing and
banking procedures. The authors demonstrate the results of their effort to meeting these
requirements at their workplace, Tissue Bank of the University Hospital, Hradec Krlov.
1. INTRODUCTION
At the beginning of cell and tissue banking there was no regulation of these activities
neither by voluntary standards nor by state health care authorities. The voluntary standards
were first formed by the American Association of Tissue Banks in the 70s, regulation by
state authorities by the issue of the FDA interim regulations in 1993. In Europe the process
was started later by standards of the EATB in 1993, issue of the Convention of
Biomedicine in 1997, by recommendation of the European Council in 2002 and last but not
least by the issue of the Directive of the European Parliament and Council in 2004 and of
the European Commission Directives in 2006. Simultaneously the scope of tissue banking
activities has been changing. Originally the tissue bank was regarded a place, where the
collected tissue was stored for some time until its clinical use. The role of the tissue bank
was to extend as long as possible the shelf life of the stored tissue. In living cells and
tissues extending their shelf life was not possible without cryopreservation that has been
available since the discovery of the cryoprotective action of glycerol in 1949. Later the
tissue banks started to introduce methods modifying the properties of the original tissue
with the aim to lower its immunogenicity and to enhance or prevent tissue rebuilding in the
organism of the host. Different kinds of tissue grafts started to be prepared by tissue banks
in large quantities, covering the needs of a particular region, or of the whole state. Because
of these new approaches as well as of increased requirements on safety of the product the
tissue banks needed to cover more activities than storage of tissues itself and developed into
establishments with precisely defined requirements on premises, equipment, personell,
control laboratories and their cell and tissue collection centres.

237

2. ORIGINAL PROJECT OF THE TISSUE BANK, ITS DEFINITION AND

EARLY HISTORY
The Tissue Bank of the University Hospital Hradec Krlov was established in 1952 by
prof. R. Klen on basis of his own concept presented at the occasion of the 14th Congress of
the Czechoslovak Society for Orthopaedic Surgery and Traumatology in 1951 and
published one year later (Klen 1952). Prof. Klen, who was in the position of the head of the
department till 1984, defined the tissue bank as a facility specialized to collection,
processing, preservation, storage and distribution of different kinds of tissues collected both
in living and deceased donors and used for clinical or experimental purposes (Klen 1952).
This concept was different from that of existing monobanks oriented to preservation of one
tissue and being regularly parts of clinical departments. At the time of establishment the
bank covered the needs of the departments of ophthalmology and orthopaedic surgery that
performed cornea (Klen 1954) and bone transplantations (Lny, 1954). Cornea and/or the
eye bulb was preserved in a moist chamber at +4C. In bone tissue the low temperature
preservation at 20C was preferred to chemical preservation in merthiolate used
predominantly in that time. Later the use of solid tissue grafts was started in neurosurgery
( Kro and Klen 1960, burn medicine (Klen et al. 1967) , stomatosurgery and face surgery.
After moving of the bank to new premises at the building of Institute of Pathology in 1967
the technological background was enlarged, and freezing of tissues to 50C and freezedrying followed later by radiation sterilization were introduced(Klen et al 1977). Freezedried grafts were distributed to many clinical departments on the whole territory of
Czechoslovakia and sent also abroad . Since the beginning of the 70s cryopreservation
methods bave been elaborated for storage of cells and tissues in a viable state. In the 80s
and 90s regular programme of autologous and allogeneic transplantation of haematopoietic
progenitor cells was introduced (Mika et al. 1991, Blha et al.1990). Simultaneously the
methods of pretreatment of solid tissue grafts prior to freeze-drying leading to modifying of
the process of their rebuilding in the organism of the host were introduced, such as
glutaraldehyde stabilization in xenogeneic freeze-dried, radiation sterilized pericardium
used in reconstruction of dura mater in neurosurgery (Mika et al., 1986, Pazek et
al.1989) and demineralisation of allogeneic freeze-dried cancellous bone for use in
stomatosurgery and dental implantology (imnek et al. 1997). The burn surgeons used
besides allogeneic skin grafts a big quantity of pigskin grafts as a temporary skin cover
(Klein et al.,1995 Mika et al.1995). In the beginning of the 90s the culture of
autologous epidermal keratinocytes as a permanent skin cover for use in burns or chronic
skin defects was introduced under the financial support of the grants of Ministry of Health
and of the Ministry of Defence (Klein et al.1997, Mika et al. 1996-1997). In the second
half of the 90s this programme was replaced by culture of autologous chondrocytes for
reconstruction of articular surface (Folvarsk et al. 2002, Pavlata et al. 2003).
3. THE LEGISLATIVE REQUIREMENTS ON TISSUE ESTABLISHMENTS SET
BY THE EUROPEAN UNION AND RESULTS OF THEIR IMPLEMENTATION
INTO THE NATIONAL LEGISLATION OF THE CZECH REPUBLIC
In the 90s of the last century there was a rapid development of tissue banking and
engineering technologies, there was also a rising criticism, however, focused on the
following issues:
1. Respect of the cell and tissue donor rights.
2. Safety of the cell and tissue transplantation.
The solution for the first problem was included in the Convention of Medicine and its
Additional Protocol.
238

The new high requirements on quality and safety of cell and tissue transplants were set
recently by the European Parliament and Council Directive 2004/23/EC (European Union,
2004) as well as by the European Commission Directives 2006/17/EC and 2006/86/EC . All
directives are to be implemented into the national legislation within 2 years after their issue.
The National Transplantation Act approved in 2002 (Ministry of Health of the Czech
Republic, 2002) included already the principles of voluntary and unpaid donation settled by
the Directive 2004/23/EC. This act introduced a strong protection of the autonomy of the
living or deceased donor as required by a Convention on Biomedicine (European Treaty
Series, 1997). In living donors informed consent was introduced in all cases including
harvest of surgical residues, e.g. femoral heads removed during hip-joint plasties. In
deceased donors the presumed consent was preserved, the national register of persons
rejecting post-mortem donation was established, however, and information of the next-ofkin became obligatory. This Act also defined the duties of tissue banks. The safety aspects
were oriented to defining the contraindications of the cell, tissue and organ donation and
traceability assurance. The new Act on Cells and Tissues of the Human Origin that sets the
quality and safety norms for use of banked human cells and tissues in the clinical practice is
expected to come into force in 2008. The main tool for achieving rapid implementation of
the new quality and safety requirements into daily practice of tissue banks using the
existing national legislation was starting the licensing process in 2004. The requirements
set by the Ministry of Health in 2004 included compliance of the tissue banks system of
work with the Directive 2004/23/EC and with the National Transplantation Act. In the
years 2006 and 2007 when the licences were renewed, the compliance with the Directive
2006/17/EC and 2006/86/EC was required.
4. METHODS OF MEETING THE REQUIREMENTS IN THE TISSUE BANK
The similarities between tissue banking processing distribution and quality control practice
and practice of manufacturing and control of sterile drugs were recognised by authors
already in the 80s ( Mika 1983, Mika et al. 1990). In that time it was obvious that
meeting of these criteria was not possible within the tissue bank premises that were at the
disposal of the authors. As the result of analysis of trends of world and European legislation
the author proposed a radical rebuilding of the Tissue Bank of the University Hospital
Hradec Krlov in accordance with standards of the International Society for
Pharmaceutical Engineering (The Society for Pharmaceutical and Medical Devices
Professionals, 1999). A new concept of the cell and tissue bank as combination of
cryogenic and clean-room technology was proposed and the bank was designed and built
with the financial support of the Ministry of Health of the Czech Republic and of the
University Hospital Hradec Krlov in the years 1998-2002 (Mika 2006,2007). The
cryopreservation and storage facility with the aseptic processing rooms started operation in
2003, the new freeze-drying facility in 2007. Simultaneously the internal and external cell
and tissue collection centres were established and their duties were defined. The
collaboration with control laboratories and companies authorized to perform validation of
the equipment and of clean rooms both at rest and operation conditions was settled as well.
The bank was granted the multifunctional licence by the Ministry of Health in 2004. In the
years 2006 and 2007 this broadest type of licence was regranted. In the year 2007 the
model inspection of the State Institute for Drug Control, Prague was performed in the bank
within the European Union twining project of this institute with AFSSAPS, France and
MHRA, Great Britain.

239

5. CURRENT ORGANISATION OF THE TISSUE ESTABLISHMENT AND

RESULTS OF ITS ACTIVITY
The Tissue Establishment of the University Hospital Hradec Krlov consists of internal
and external cell and tissue collection centres, clean rooms serving for aseptic processing
before cryopreservation or freeze-drying of tissues, cryostorage and freeze-drying facility
and control laboratories. The number of living cell and tissue donors is presented in the
Tables I and II. The number of deceased solid tissue donors is presented in the Table III.
The number of solid tissue grafts delivered for clinical application is presented in the Table
IV.
Table I: Number of living cell and solid tissue donors

Year

Living cell donors

Autologous
use
95
117
80
66

2004
2005
2006
2007

Living solid tissue donors

Allogeneic use

Autologous use

Allogeneic use

14
18
10
11

15
6
10
13

84
110
140
174

Table II: Number of living cell donors

Year

Autologous
haematopoietic
progenitor cells

Allogeneic
haematopoietic
progenitor cells

Cultured
autologous
chondrocytes

Sperm

2004
2005
2006
2007

57
66
34
22

15
14
8
6

17
26
21
22

21
25
25
22

Table III: Number of deceased solid tissue donors

Year

2004

2005

2006

2007

Number

16

19

17

20

Table IV: Number of preserved solid tissue grafts delivered for clinical transplantation

Year

Musculoskeletal tissue

Fascia lata

2004
2005
2006
2007

231
249
288
245

165
126
138
156

240

6. DISCUSSION
Despite of the fact that the definition of the tissue establishment included in the Directive
23/2004EC does not differ substantially from the original definition of the tissue bank
(Klen 1952), the legislative changes that were initiated by the issue of the Directives
represented the most radical change in the position of tissue banks in the Czech Republic in
their more than 50 years old history. The Transplantation Act made it possible to start and
to run the tissue bank licensing process immediately after the issue of the Directives.
Implementation of the principles of voluntary and unpaid cell and tissue donation led to
considerable changes in the number of tissue donation (Tables I, II, III). From the tables it
is clear that there was a considerable increase of the living solid tissue donations while the
number of post-mortem donations remained relatively low. To secure satisfactory sources
of the bone tissue collected in living donors it was necessary to establish external collection
centres on a contract basis in four county hospitals. In post-mortem tissue donations a basis
of the hospital donor management system was introduced in cooperation with the Regional
kidney transplantation centre. On this basis all necessary measures are performed including
contacting the donors family.
As the Transplantation Act was accompanied by a decree defining the conditions for health
suitability of the donor, donor screening and serological testing (Ministry of Health, 2004),
the implementation of the Directive 2006/17/EC did not represent a major obstacle for the
majority of banks including our establishment.
Similarly the strict requirement of keeping traceability of the way from the donor to the
host of the tissue graft did not represent any change of existing practice as this approach
has been applied since the very beginning of the activity of the Tissue Bank of the
University Hospital Hradec Krlov (Klen 1954). The same case is a continuous
monitoring of clinical results (Klen 1957, Mika 1983) including registration of adverse
reactions or graft failures.
On the contrary meeting of the requirements of the Directive 86/2006/EC setting high
demands on the technical background of the tissue bank was not possible without radical
rebuilding of premises of existing cell and tissue banks. It was proved on the example of
our establishment that a concept of a tissue establishment based on combination of
cryogenic and clean-room technology introduced as a priority in the Czech Republic
represents an efficient tool how to meet the requirements of this directive. By strict keeping
aseptic conditions during cell and tissue collection and processing of cell and tissue grafts a
real alternative was established to the use of terminally sterilised grafts, which is important
in the instances where terminal sterilisation impairs their mechanical and/or biological
properties. Establishing of the system of internal and external quality control is an
additional new feature of the practice of tissue banks as in the past the quality control was
made predominantly by the tissue bank itself. The system is based on separating of personal
responsibility for production and for the control, that is performed in majority of cases by
laboratories not being a part of the tissue bank. The results of all control tests made by these
laboratories are evaluated by a person responsible for release of the grafts for clinical
application. This person, who must be an employee of the tissue bank, is nominated by the
Ministry of Health and his/her name is included in the text of the licence. To assure higher
involvement of the state in the graft safety and quality regular inspections of tissue banks
are planned to be performed in the near future by the State Institute of Drug Control.

241

7. CONCLUSION
1. Despite of the fact that the definition of the tissue establishment included in the Directive
23/2004/EC does not differ substantially from the original definition of the tissue bank
published in the 50s, the legislative changes that were initiated by the issue of the
European Union Directives represented the most radical change in the position of tissue
banks in the Czech Republic in their more than 50 years old history.
2. Implementation of the principles of voluntary and unpaid donation into the practice of
the tissue bank led to increased use of solid tissues collected in living donors.
3. The concept of a tissue bank based on combination of cryogenic and clean-room
technology represents the useful tool to meeting the requirements of the Directive
2006/86/EC. This concept introduced by the authors as a priority in the Czech Republic is
now being followed by other tissue establishments in the country.
8.REFERENCES
1.

Blha M, Maisnar V, Jebav L, Mika P, Rondiak J, ulc K, Vasek J. jr. 1990,

Transplantation of blood progenitor cells collected from peripheral blood - first
experience, Voj. Zdrav. Listy, LIX:218-223.
2. Blha M, Mika P, k P, tpnov V, Vvra L, Mal J, Touovsk K. 2003,
The Risk of Infection Transmission from Blood Progenitor Cell Concentrates, J.
Hemotherapy and Stem Cell Res., 12:161-164.
3. Convention for Protection of Human Rights and Dignity of the Human Being with
Regard to the Application of Biology and Medicine: Convention on Human Rights
and Biomedicine Oviedo, 4.IV.1997 European Treaty Series/164..
4. European Union, 2004, The European Parliament and the Council: Directive
2004/23/EC of the European Parliament and of the Council on Setting Standards
of the Quality and Safety for the Donation, Procurement, Testing, Processing,
Preservation, Storage and Distribution of Human Tissues and Cells, Strasbourg,
2004, 41 p.
5. Folvarsk, J., Ddek,T., Dobe, D., Frank, M., Adler, J.:Treatment of deep
cartilage defects in the knee by mosaicplasty combined with autologous cultivated
chondrocytes in Tissucol, 2002. European Journal of Trauma, 28, Suppl. P. 43
6. Klen, R. 1952, Tissue Centre (TC), s. nemocnice, 20:120-121 in Czech.
7. Klen,R.: Collection and storage of grafts for transplantation,1954. In.Group of
authors: Tissue Transplantation. SZdN, Prague, 13-21 (In Czech)
8. Klen, R. et al.: Experimental study of factors influencing take and survival of
human skin homografts, 1968. Plast. Reconstr. Surg. 41,p.471-476
9. Klen, R., Metelka,M., Pazek, J., 1977 : Freeze-dried homogeneous grafts of
fascia lata in neurosurgery. J. Neurosurg. Sci 21/4, p.247-250
10. Klein, L., Mika, P., Preis, J.,1995:Clinical experience with skin xenografts in
burned patients. In:Maselis,M.,Gunn,S.W.A. (eds.) Proceedings of the Second
International Conference on Burns and Fire Disasters:Perspectives 2000.-I.st ed.,
Palermo, Kluwer Academic Publishers, p.343-345
11. Klein, L., Mika, P., Strakov, H., Jebav, L., Noikov, M., Blha, M.,
Talbov, Z., Hoek, F.:Biological skin covers in treatment of two cases of the
Lyell syndrome 1997. Annals of Transplantation, 2, p.45-48
12. Kro, M. and Klen,R.:Our experience with the dura mater replacement with frozen
fascia and dura grafts, 1960. Lk.Zpr. LF UK v Hradci Krlov, p. 80-85. In
Czech
242

13. Lny, J.: Use of bone grafts in orthopaedic surgery,1954. In. Group of Authors:
Tissue Transplantation. SZdN, Prague, p. 32-37 (In Czech).
14. Mika, P.: Quality Control of Freeze-dried Tissue Grafts.In.: Proc. 16 th Int.
Congr. Refrig., Paris 1983, Fr.3,1984, p.161-165
15. Mika P, Vvra L, Huek Z, Strakov H, paek J. Pazek J, Klein L, Levnsk
M.1990, Low Temperature Preservation of Tissues for Clinical Use. In: Bose A,
Sengupta P (eds.) Advances in Cryogenics, Proceedings of the International
Conference on Cryogenics, Calcutta, India, December 6-10,1988 (INCONCRYO),
VII.Cryogenics in Life Sciences and Medicine, Surgery, Agriculture, McMillan,
India Ltd., New Delhi, p.625-644.
16. Mika P, Schustr P, Vin M, Dudek A, Vvra L, ervinka M, Rondiak J. 1991,
Containers for Freezing and Storage of Bone-Marrow Stem-Cells, Sb. Vd. Prac
Lk. Fak. Karlovy Univ. Hradec Krlov, 34:367-387.
17. Mika, P., Klein, L., Preis, J., Ettlerov, E. 1995:The role of the tissue bank in
disaster planning. In:Maselis, M., Gunn, S.W.A. (eds.) Proceedings of the Second
International Conference on Burns and Fire Disasters:Perspectives 2000.-I.st ed.,
Palermo, Kluwer Academic Publishers, p.75-83
18. Mika, P., Strakov, H., Klein, L., ubrtov, D: Practical aspects of establishing
an allogeneic human keratinocyte bank 1996-1997.Roczniki oparzen, Annals of
burns 7/8, p.105-109
19. Mika P. 2000, Brief History of the Tissue Bank, Charles University Hospital,
Hradec Krlov, Czech Republic, Cell and Tissue Banking, 1,p.17-25.
20. Mika P, Blha M, Vvra L, tpnov V. 2003, Our System of Crosscontamination Prevention During Storage of Haematopoietic Progenitor-Cells paper ICR 0302, In: Grof, G., Menzer, M. (eds): 21st IIR International Congress of
Refrigeration, Washington 2003.
21. Mika P, Strakov H, ermk P, tpnov V, Hradeck Z, Drahoov M. 2002,
New Safety-Assurance for Biological Skin Covers, Acta Chir. Plast., 44:23-29.
22. Mika P, Vvra L, Vin M, Schustr P. 2004, The Importance of Oxygen-Level
Monitoring in the Cryostorage-Facilities In: Chrz V (ed.), The Eighth Cryogenics
2004, IIR International Conference: Refrigeration Science and Technology.
International Institute of Refrigeration, Paris, Paper C 04-08.
23. Mika P, Strakov H, Vvra L, Schustr P, Vin M, Popil J, Postupa J, Plasov
B. 2004, The Cell- and Tissue- bank as a Combination of Cryogenic and Cleanroom Technology. In: Programme and Abstracts International Congress of the
European Association of Tissue Banks, October 13-16, 2004, p. 69.
24. Mika P. 2006, Contribution to Safety Assurance in the Cryopreservation of
Cells and Tissues used for Clinical Transplantation. Doctoral (Ph.D.) Thesis
Summary, Nucleus HK, 34 p.
25. Mika P., Blha, M., Strakov, H., Navrtil, P., Grfov, M., Poepcov, I.,
Folvarsk, J., Trlica, J., vk, I., mejkal, K., Ddek,T. 2007: New Safety
Assurance for Processing and Cryopreservation of Cell and Tissue Grafts used for
Clinical Transplantation. Transplant International 20 Suppl.2, p.325.
26. Ministry of Public Health of Czech Republic, Decree number 437/2002 Code of
law, from October 3rd, 2002, which establishes in detail conditions for health
suitability and extent of screening and testing of living or deceased tissues or
organ donor for the use of transplantations. (Decree about health suitability of
tissue and organ donors for the use of transplantations), Code of law number
437/2002 Part 153, pp. 82218223 (in Czech).

243

27. Ministry of Public Health of Czech Republic, Act number 285/2002 Code of law
from May 30th, 2002 about donation, recovery and tissue and organ transplants and
about change of other Acts (Transplantation Act). Code of law of the Czech
Republic No. 285/2002 Part 103, p. 6050 (in Czech).
28. Pazek, J. Mika, P., paek, J.,Nmeek,S., Eli,P.,ercl,M.,1989:
Xenogeneic pericardium as a dural substitute in reconstruction of suboccipital dura
mater in children. J. Neurosurg. 70,p. 905-909
29. Pavlata, J., Urban, K., Karpa, K.,Mika, P., Strakov, H., Brtkov, J 2003:.
Reconstruction of the joint surface. Acta Medica (Hradec Krlov), 46, p.57
30. The Society for Pharmaceutical and Medical Device Professionals - ISPE:
Baseline. Pharmaceutical Engineering Guide. Pharmaceutical Engineering Guides
for New and Renovated Facilities. Volume 3. Sterile Manufacturing Facilities,
First Edition/January 1999. ISPE Headquaters, ISPE European Branch Office,
1999, 162 p.
31. imnek, A.: Author own experience with using of freeze-dried demineralised
bone in dental implantology 1997. In: Mika, P., Strakov, H., Stacey, G (eds.):
Cryoprotectants in Medical Practice. Long Abstracts. International Institute of
Refrigeration, Paris p.47

244

CR08-32
VENTILATION OF CRYOSTORAGE FACILITIES OF TISSUE
ESTABLISHMENTS
Lain M. 1, Mika P. 2, Dvok J. 2
1

Technical University in Prague, Mech. Eng. Faculty, Prague, Czech Republic

2
Tissue Bank, University Hospital, Hradec Krlov, Czech Republic
ABSTRACT

The liquid nitrogen is commonly used for long-term storage of viable cells and tissues in
the cryostorage facilities of tissue establishments. As frequently large biological containers
are used the continuous evaporation of nitrogen during storage as well as evaporation
during refilling containers can lead to decrease of the air oxygen level below acceptable
limits (20% for women, 18% for men - according to the Czech law and recommendation of
producers). For this reason sufficient ventilation of such facilities is essential for safety of
the staff. Ventilation should also eliminate the possible leak of liquid nitrogen in
extraordinary situations. The character of nitrogen production is non-stationary due to
above mentioned situations. The authors present the basic model based on non-stationary
pollutant production. The nitrogen production in the model is calibrated according to on site
measurements, and can be used for prediction of specific operation and emergency
scenarios. The principles and recommendations for cryogenic storage rooms ventilation
systems, its control and environment monitoring equipment is also presented.
INTRODUCTION
One of the main reasons for ventilation of spaces is maintaining the pollutants in the room
within permissible concentration. In the cryostorage facilities of tissue establishments large
quantities of liquid nitrogen are frequently used for long-term storage of viable cells and
tissues. The continuous evaporation of nitrogen during storage as well as evaporation
during refilling containers is the main source of the contaminant. The increasing
concentration of nitrogen is not dangerous due to nitrogen itself, but due to decrease of the
air oxygen level below acceptable limits.
The influence of oxygen level on mans health can be classified in 4 stages:
1st stage - oxygen level 20-14 % - changes in concentration, accelerated pulse.
2nd stage - oxygen level 14-10 % - people stay conscious, loss of discernment.
3rd stage - oxygen level 10-6 % - belly-ache, loss of movement control, swoon.
4th stage - oxygen level below 6 %, spontaneous breathing stops, death in few minutes .
Up to 3rd stage, it is difficult for unexperienced people to recognize any problem, there is
not any smell or other warning. It can be very dangerous in nitrogen polluted places. In the
hygienic standards usually 18 % is the bottom limit for oxygen level in working places,
accordinbg to the Czech standard 288/2003. The bottom limit for oxygen level 20 % is set
for women and teenagers.

245

1. THE MODEL
1.1. Contaminat concentration model
The contaminant concentration model is based on the contaminant balance of the ventilated
space under presumption of complete mixing
of the air in the room (Fig.1).

M d +V p C p d V p Cd =VR dC
(1)
M
Vp
VR
C

contaminant production [g/s]

volume flow rate [m3/s]
space volume [m3]
contaminant concentration [g/ m3]
time

C 2 = C1e

V p
VR

V p

M

+
+ C p 1 e VR
V p

Figure 1: Contaminant concentration balance

There are two possible solutions for

(2) non steady contaminant production or

ventilation flow rates. The operation

time can be divided into periods with

constant productions and ventilation

and commonly known analytical solution of equation (1) can be applied. For finite time
intervals equation (1) can be directly applied. The presented model is based on equation (2)
for nitrogen concentration. The oxygen level is calculated from nitrogen concentration.

1.2. Nitrogen production

There are two sources of contaminant in the storage room: The continuous evaporation of
nitrogen from large biological containers
21
and evaporation of nitrogen during
refilling. In the Tissue Bank in University
Hospital Hradec Krlov, there were
20
following containers at the time of our
measurements:
19
2 x MVE XLC 1200,
Meassurements
1 x Cryocyl 230LP,
Model
2 x MVE XLC 230,
18
0
600
1200
1800
2400
3000
3600
4200
4800
1 x MVE Cryosystem 2000,
Time [s]
2 x Cryometal KL32,
21
1 x Eurocyl 230LP,
.
20.5
According the data of the producers of
containers the total nitrogen evaporation is
22,5 m3/day.
20
1.3 Ventilation system
The existing ventilation system supplied
constant amount of 1720 m3/h
of
conditioned fresh air into the storage room
which was equal to the air exchange rate 8
/h (8 ACH). The supply of the fresh air
should be close to the breathing zone of

Oxygen level [%]

Oxygen level [%]

Prbh koncentrace kyslku, pi polovinm objemu mstnosti

Meassured
Model
19.5
0

600

1200

1800

2400

3000

3600

4200

4800

Time [s]

Figure 2: Model calibration and verification

results

246

people, and the exhaust should be close to floor and at the largest nitrogen source (refilling
valve).
1.4 Model calibration
The model vas calibrated (Fig. 2a) according to measured oxygen level in the room. Results
of calibration are the nitrogen productions during the refilling and steady state. Finally
according to calibration the 49,4 l/day of liquid nitrogen is evaporated and 1,54 l/min is
evaporated during refilling. The calibrated model vas verified (Fig. 2b).
2. TESTED CASES
2.1 Controlled ventilation
The various scenarios were tested on the model. To save energy the two stage control of the
fresh air flow rate was recommended. During refilling and after it full flow rate (8 ACH)
and during common operation reduced flow rate (1 ACH). The calibrated model was
applied to check the oxygen level for such an operation (Fig.3).
21.00

Oxygen level [%]

Oxygen level [%]

21.00

20.50

20.50

20.00

Model

Model

20.00

19.50

500

1000

1500

2000

2500

3000

3500

4000

200

400

600

800 1000 1200 1400 1600 1800 2000 2200

Time [s]

Time [s]

Figure 3: Oxygen level for 2 stage flow rate:

just storage - 1 ACH (left),
during refilling 8 ACH (right).

2.2 Emergency scenarios

The model was applied to test various emergency scenarios as well. The oxygen level was
modelled for a long-time maximum nitrogen production, that usually happens just during
refilling period. The time of the modelled emergency situation was much longer than the
normal duration of the re-filling period, which typically is 15 min. (Fig.4).
21
18

20.50

Oxygen level [%]

Oxygen level [%]

21.00

20.00
19.50
19.00
18.50

15
12
9
6

Model

Model

18.00
0

500

1000

1500

2000

2500

3000

3500

4000

500

1000

1500

2000

2500

3000

Time [s]

Time [s]

Figure 4: Oxygen level for emergency situation 8ACH - left, 1 ACH - right.

247

3500

4000

CONCLUSIONS
The proper ventilation flow rate is very important for rooms with the cryostorage facilities
of tissue establishments. The detailed model of the oxygen level can be very useful for
system design and operation. In generally, two steps control of fresh air flow rate can be
recommended. The low flow rate for just the storage period and the high flow rate for the
refilling period. In the emergency situation the high ventilation is important as well.
REFERENCES
1. Dvok, J.: Vtrn prostor se zdroji dusku, diploma work, CTU in Prague, Mech. Eng.
Faculty, Department of Environmental Engineering, 2006.
2. Mika, P., Vvra, L., Vin, M., Schustr,P..: The importance of oxygen level monitoring
in the cryostorage facilities C04-08). In: The Eight Cryogenics 2004 IIR International
Conference .1st edition. Refrigeration Science and Technology Proceedings, Paris,
International Institute of Refrigeration, 2004, p.242-247
This research was supported by research plan MSM6840770011.

248

CR08-25
SPECIAL EQUIPMENT FOR CRYOPRESERVATION OF TISSUE
IN A STANDARD FREEZING UNIT
Sprl G.1, Klingner E.2, Quinger J.3
1

Institute for Air Conditioning and Refrigeration, Dept. of Applied New Technologies,
Bertholt-Brecht-Allee 20, D-01309 Dresden, Germany
2
edecto GmbH, Erlweinstrae 9, 01069 Dresden, Germany
3
Ingenieurbro und Plastverarbeitung Quinger GmbH, Schwarzer Weg 7, 09557 Flha
ABSTRACT
With forthcoming efforts in tissue engineering tissue banks will be established. For freezing
tissues with different sizes standard freezing units will be taken up. Hence the effectiveness
of such a unit was investigated. Results: The temperature homogenity was not sufficient for
developing cryoprotocols for tissues. Commonly used vials and Petri dishes are not suited
for storage of tissues during the cryopreservation procedure. Special hardware was
developed to overcome this deficit.
INTRODUCTION
Cryopreservation is a well established technique for long time storage of living biological
materials, especially of cell suspensions. According to the-state-of-the-art of science and
technology these suspensions are transferred to cryovials or bags (e. g. blood). For these
cryopreservation can be done in two major ways. First, for small amounts of vials
Qualifreeze vessels from VWR or Mr. Frosty vessels from Nalgene are often used.
They guarantee a cooling rate of about 1C per minute when filled with isopropanol and
placed in a freezer (-80C). [1] Second, controlled rate freezing can be carried out in the
gaseous phase of an automatic Standard freezing unit. In this case the vials or bags are
placed in special racks which allow controlled freezing and additional seeding if necessary.
[2] For long time storage the vials and bags are transferred into a liquid nitrogen storage
vessel.
Cryopreservation of living tissues is under way not only because of the forthcoming efforts
in the field of tissue engineering. All in all regenerative medicine is in progress. In contrast
to cell suspensions living tissues, especially tissue engineered constructs consist of a
scaffold, seeded with different cell types, and are well defined in terms of the upper and
lower surface. Therefore turning around of the tissue must be prevented. Furthermore the
tissue area can be larger than 1 cm. [3] All these aspects refer to lacking applicability of
vials for tissue cryopreservation. For that reason special equipment was developed for
usage in a standard freezing unit.
1. BASIC INVESTIGATIONS ON A FREEZING STAGE
1.1 Investigations of thin carrier for the supervised development of a cryoprotocol
At the beginning of our research, we have observed the behaviour of cells within a scaffold
during cooling down to 120 K at a freezing stage (LINKAM). Therefore the constructs are
placed on a glass cover slip which was mounted on the surface of the silver block of the
freezing stage. In order to qualify the visual results thermocouples were placed at several
points on the sample and on the stage (see figure 1, right). In respect to further application
of the cryoprotocol glass cannot be used in practice. So other materials for cover slips were
249

tested. The thickness of the glass cover slips ranges from 0,1 to 1 mm. Plastic cover slips
were 1,1 mm thick. All measurements were done with the same cryoprotocol: 1. cooling
with 5 K/min down to 273 K; 2. equilibration for 20 minutes;3. further cooling with 1
K/min down to 268 K; 3. equilibration again for 5 minutes; 4. continue cooling with 1
K/min down to 233 K; 5. final equilibrate for 5 minutes and 6. rewarming with different
rates to 300 K. The appearance of spontaneous ice nucleation was used as a measure for the
heat transfer from the silver block to the upper side of the tissue. Further cooling down to
120 K didnt gave much more information concerning the first ice formation.
Tissue on cover slip
Silver block
Thermal couples

Figure 1: Freezing stage on the microscopic stage (left)

freezing stage interior, sample preparation (right)

1.2 Results
At first very thin cover slips were used. The results are shown in figure 2. The temperature
gradient between the silver block and the upper side of the tissue ranges from 5 to 7 K.
30

TE 2: middle of the tissue

TE 3: Cryoprotectant
TE 4: edge of the tissue
TE 5: on the silver block

20

Temperature in C

10

-10

-20

-30

-40

-50
0

50

100

150

200

250

300

350

400

450

500

Time in (sek/10)

Figure 2: Temperature profiles of thermal couples, placed at different points (see fig. 1),
using 0,1 mm thick glass cover slips

250

The spontaneous nucleation causes a very small anomaly because the latent heat can be
dissipated quite well through the very thin cover slip. Thicker sample holders caused worse
results. The worst values we got when using the plastic cover slips (see figure 3).
40

TE 2: middle of the tissue

TE 3: Cryoprotectant
TE 4: edge of the tissue
TE 5: on the silver block

30

20

Temperature in C

10

-10

-20

-30

-40

-50
0

50

100

150

200

250
Time in (sek/10)

300

350

400

450

500

Figure 3: Temperature profiles of thermal couples, placed at different points (see fig. 1)
using 1,1 mm thick plastic cover slips
Here, the temperature gradient between the silver block and the upper side of the tissue
varies between 8 and 15 K. The spontaneous nucleation causes a very broad anomaly
because the latent heat cannot be dissipated well through the thick plastic cover slip.
In order to manage the spontaneous ice formation, additional cold (about 80 K) was applied
by cold forceps. The latent heat could be reduced or diminished at a large extent. But in
case of tissue larger than 1 cm the centre of it was still destroyed. The way out was found
by a device, which brings cold inside the tissue hedgehog). A satisfying cryoprotocol
could be developed as the proof of principle.
Cryopreservation of biological materials requires high cell respectively product viability
and integrity at point of use. For the integration of this freezing process in a tissue bank,
first the cryoprotocol has to be transferred to an automatic standard freezing unit. These
units were developed for a controlled rate freezing with high accuracy and repeatability.
As mentioned before the temperature distribution within the tissue must be extensively
homogeneous for a successful cryopreservation. For that reason the temperature distribution in the cooling chamber of the freezing unit should be well known.
2. BASIC INVESTIGATION ON A STANDARD FREEZING UNIT (FREEZER)
2.1 Investigation of the temperature distribution in the cooling chamber
Cell type dependent cryoprotocols take care for an optimal cooling rate to avoid cell
damaging during freezing. In case of controlled rate freezing the vials are placed in a rack
inside the cooling chamber. If controlled seeding is necessary a special rack for autoseeding
can be used. For both applications gaseous nitrogen cools the cryovials slowly. In case of
251

seeding additional liquid nitrogen flows through the autoseeding rack at the distinct
temperature. Figure 4 shows a standard freezing unit and an autoseeding rack for cryovials
inside the cooling chamber.
Thermometers (Pt 100)

Gas inlet

Figure 4: ICE CUBE 15M from Sylab overview (left)

and autoseeding rack inside the cooling chamber (right)

The freezing unit is connected to a storage vessel for liquid nitrogen (LIN). The required
size of that vessel depends on the flow rates of the freezing processes.
These freezers are often used for cryopreserving vials and bags, containing cell suspensions
or blood. For that reason an arrangement of vials in an autoseeding rack was chosen in
order to obtain the temperature distribution inside the cooling chamber under real
conditions. Thirty cryovials of 2 ml volume have been placed in the rack, each filled with
1,5 ml of a cryoprotectant. Five characteristic positions were evaluated (see in figure 4)
with thermocouples. One precision thermocouple was centred in each vial. An additional
vial contained the sample thermometer (Pt 100: ) of the freezing unit. All temperatures
were monitored during running a well established cryoprotocol without autoseeding. But
this protocol initiated a small spontaneous ice formation. Here, the appearance of
spontaneous ice nucleation was also used as a measure for the heat transfer from the
chamber via the rack into the vials.
In order to eliminate errors of measurements the thermocouples were cyclic permuted.
Furthermore all sensors were placed and fixed in one vial in all five positions for two runs.
2.2 Results
First, comparing the results from more than 20 measurements significant deviations could
be found between the five positions. At the point of spontaneous ice formation temperature
differences up to 10 K were measured (see insert in figure 5). Second, the sample
thermometer of the freezer didnt represent the real freezing process within the cell
suspension (see figure 5).
Looking at cell suspensions, the cells are round and in an equilibrium state. Suspended cells
freeze well and without visible deviation between different locations in the rack using cell
specific cryoprotocols. In conclusion, the cell suspensions seem to tolerate the observed
temperature differences.

252

Looking at tissues, the cells are adherent and spread and got in contact. When cryo preserving several tissue pieces larger than 1 cm at the same time, the inhomogenous
temperature distribution inside the cooling chamber can course undefined freezing and
disturbtion of the tissue. This we have measured in the temperature supervision at the
different positions as well as in an unfavourable effect on the tissue and cell integrity. This
situation is not acceptable for cryopreservation of tissues and prohibit the transfer into
tissue banks.
Thermocouples

5K
5 min

Onset of the rapid cooling

Left bottom
Right bottom
Middle
Left up
Right up

Pt 100 sample
thermometer
and
Pt 100 chamber
thermometer of the
freezer, equal to the

Figure 5: Representative temperature run at the place of the sample Pt 100 of the freezer
Insert: Temperatures of the thermocouples at the five measuring positions
3. CONSEQUENCES FOR SPECIAL EQUIPMENT
Petri dishes and cryovials are not suited to store tissue during the cryopreservation process.
The tissue shouldnt be wrapped and turned. Thats why special tissue cryogenic vials with
screw closure (can) similar to a small Petri dish were developed. The material for the tissue
vials must tolerate low temperatures, be thin enough for a good heat transfer and be tight
during the whole preservation process. In order to ensure a sufficient heat exchange
between tissue and the surrounding two additional hardware components were designed.
First, the device called Hedgehog takes care for an optimal and reproducible heat transfer
through the tissue. Inside the tissue cryogenic vials some pins are placed as means for
fixing tissue and for conducting cold into that tissue. This gives the opportunity to
cryopreserve large-area tissue of more than 1 cm. Furthermore two kinds of cans were
developed. The metallic one is comparable to a Petri dish with = 35 mm and ensures a
good thermal contact to the surrounding area inside the freezer but its only for laboratory
use. The plastic one also fulfils the thermal requirements for cryopreservation and will be
qualified for registration as a medical device in next future.
253

Second, the device called Rack ensures the heat exchange between cooling chamber or
thawing unit, can and tissue. It allows efficient, defined and reproducible heat transfer to
the special vials during cooling and rewarming. It guarantees a rapid thawing. Beside this it
organises that some tissue pieces can be handled and stored in more than one can under
equal conditions.
CONCLUSIONS
All developed parts for the cryopreservation procedure were tested also with tissue
engineered constructs. Though tissue engineered mucosa of the mouth was cryopreserved
and after thawing transplanted in an animal model [4] with good success.
The new equipment overcomes the disadvantages if inhomogeneous temperature
distribution in freezer and allows in the same way an more precise rewarming of tissue with
high temperature rates. In summary cryopreservation of the engineered mucosa becomes
more successful with better results in biochemical tests and higher viability of the cells.
Both special equipment - can with hedgehog and rack - are prepared for patent
registration.
REFERENCES
1 Lehle, K., Hoenicka, M. et al., Cryopreservation of human endothelial cells for vascular
tissue engineering, Cryobiology (2005) 50 154-161
2 Sputtek, A., Mingers B., Langzeitkonservierung von menschlichen roten Blutkrperchen,
KI (1994) 30/9 441-443
3 Applegate, D.R., Liu, K., Mansbridge J., Practical considerations for large-scale cryopreservationof a tissue engineered human dermal replacement, Adv. In Heat and Mass
Transfer in Biology (1999) HTD-Vol. 363/ BED-Vol. 44 77-91
4 Sprl, G., Eckelt, U., Lauer, G. and Klingner, E., Cryopreservation of Tissue Engineered
Mucosa, Proceedings of ICMC06 + 9th Cryogenics (2006) 359-363

254

USE
OF LOW TEMPERATURES
IN CRYOTHERAPY

255

CR08-55
THE LIQUID AIR CRYOCHAMBERS FOR WHOLE-BODY
CRYOTHERAPY
Strnad P.1 , Fortkov L.2, Brojek W.3,
1
DN FORMED Brno s.r.o., Czech Republic
Masaryk University - Faculty of Medicine, Department of Biophysics,Brno, Czech
Republic
3
METRUM CryoFlex , Sp. z o.o., Blizne Laszczyskiego, Poland

ABSTRACT
The paper describes the liquid air cryochambers used for whole-body cryotherapy, namely
the cryochambers exploiting the principle of the retention of cold. The cryochambers have
already been installed in the Czech Republic and further studies leading to the fructification
of the efficiency of this method with regard to diagnosing the patients have been started.
INTRODUCTION
The whole-body cryotherapy is one of the physical therapy methods of the application
which is based on body reaction to the stimulation of the temperature lower than 100 C.
At present, the whole-body cryotherapy ranks the modern methods used in the
rehabilitation and in physical medicine.
The introduction in 1978 of the whole-body cryotherapy to medical prophylaxis is
attributed to Professor Toshiro Yamauchi and his team (liquid nitrogen cryochamber). [1],
[3].
Having been inspired by the positive treatment results of Japanese scientists, professor
Reinhardt Fricke and his team from Germany transferred the method in all its applied
forms to Europe [2],[3].
Significant role in this field is played by the Polish specialists. The Polish cryotherapy
originated in 1983 (Professor Zdzisaw Zagrobelny). In the year 1989, the first
cryochamber (liquid nitrogen cryochamber) was constructed in Poland (the second in
Europe, the third in the world) [3].
In 2002 Wieslaw Brojek and Wlodzimierz Szmurlo first on the world used the liquid air
in whole- body cryotherapy and built the first cryochamber with retention of cold on the
world [4].
In 2006, the first cryochamber with retention of cold using the liquid air in Czech Republic
was built and this cryochamber was the first cryochamber of this model which has been
built outside of Poland. Evidently, the development of the cryochambers was influenced by
the development of scientific knowledge and technology.
CRYOCHAMBER WITH RETENTION OF THE COLD
The cryochamber can be built in numerous designs. The selection of the appropriate design
depends on local conditions and room available. Typical design of the cryochamber is
presented in the Figure 1.

257

The operation principle consists in the application of the cold laying low in a trough and in
using the cold to keep low temperature inside the treatment chamber. Liquid air is used in
order not to apply heat exchangers as applied in traditional chambers with liquid nitrogen.
The elimination of losses on heat exchangers allowed to increase the cryogenic liquid
capacity by 50 %. The low temperature results from spraying the cryogenic liquid from
a system of jets inside of the chamber.
Collectors and jets
Louvers

Floor level

Window
Chamber

Handrail
Recess

Stairs

Thermal insulation

Figure 1. The cryochamber with retention of cold cut-away view

The operation principle consists in the application of the cold laying low in a trough and in
using the cold to keep low temperature inside the treatment chamber. Liquid air is used in
order not to apply heat exchangers as applied in
traditional chambers with liquid nitrogen. The
elimination of losses on heat exchangers
allowed to increase the cryogenic liquid
capacity by 50 %. The low temperature results
from spraying the cryogenic liquid from
a system of jets inside of the chamber.
Actuated valves are being closed one by one in

Figure 2. Stairs into the cryochamber with

retention of cold

Figure 3. Ground floor room

258

sequence as the temperature approaches the temperature pre-set at the controller, and the
liquid air is being decreased till the main valve is closed. The main valve is the last one to
cut off the liquid air flow into the cryochamber. Thermocouples inside of the chamber
measure temperatures on different heights. The controller calculates the average
thermocouple temperature indication, compares it against the temperature pre-set for the
treatment carried out at the moment, and controls the operation of actuated valves. Via the
controller, the software provides the correct temperature inside the chamber. Besides, the
software monitors the oxygen concentration inside the chamber.
An acoustic alarm is generated as soon as the oxygen concentration reaches the hazard
level. For patients safety, the chamber is equipped with two independent oxygen
concentration meters.
The chamber for cryogenic treatment can be built in versions for two, three or four persons.
The chamber is thermally insulated with glass foam panels of approx. 15 cm thickness
fastened to a load bearing structure of wood. The inner surfaces of the chamber are finished
with soft wood. Inside the chamber, between wooden louvers and chamber walls is a nozzle
system for spraying the cryogenic liquid. The chamber is equipped with double leaf
transparent door of organic glass. The chamber is also equipped with
a transparent roof (Fig. 2) to enable the operator (physiotherapeutist) watch the patients.
The electrical installation inside of the cryochamber is supplied from a safe voltage source
(12 V AC, 24 V AC).

Figure 4. The cryochamber with retention of cold rear view

The cryochamber is connected via a cryogenic pipeline with a tank which is the cryogenic
liquid source. As for dimensions of the tank, the tanks for 6 or 11 tons of liquid air have
been used usually. The cryogenic liquid is synthetic liquid air (mixed of pure nitrogen and
oxygen). Liquid air contains 21% 2% oxygen. Each delivery is provided with a
manufacturer certificate of the liquid air composition. The liquid air manufacturer is
responsible for the composition, purity and for the procedure of the tank filling with the
cryogenic liquid.
At present, liquid air is manufactured in Poland and it is delivered by two companies to the
cryochambers in Czech Republic.

259

The cryogenic pipeline is either vacuum-insulated, or it is equipped with a traditional

thermal insulation. The consumption of the liquid air is from 90kg/per hour to 120 kg per
hour in dependence on the size of the cryochamber. A microprocessor control computer
with appropriate software is part of the cryochamber system (Fig. 4). It provides the
cryochamber control during treatments. Besides, the cryochamber is outfitted with an ozone
generator and a blower; and it is also connected to the controller. The cryochamber control
functions, drying procedures and sterilisation procedures have been divided into two
independent circuits within the controller. An UPS is a component of the equipment, to
enable the operation (treatment) to be safely finished in case of voltage failure.
DISCUSSION
An original solution, using a phenomenon of cold retention and direct injection of liquid air
into the cryochamber, notified in the Europe Patent Office, provides an opportunity for
widening the range of interventions offered so far in well-known cryo-chambers solutions.
A novelty in the construction of cryotherapeutic cabin is the location in the hollow below
the level of the operative floor, which allows the use of the cold retention effect and
advantage of direct liquid air injection into the cryochamber. The slope of the stairs is mild
and the steps are wide which enables the patients who are less fit to enter the cabin without
problems and can be used as an adaptation area (Fig. 2).
The staff is able to observe the patient permanently thanks to the fact that the cabin is well
lit and that the ceiling is transparent. This makes the cryochamber very comfortable
especially for patients who tend to claustrophobia.
APPLICATIONS OF THE WHOLE-BODY CRYOTHERAPY
Generally in all models or types of cryochambers:
The cryochamber is applied for therapy of the following diseases:
Acute and chronic arthropathy and diseases of articular cartilage, e.g. rheumatoid
arthritis, Bekhterovs deformans-ankylosing spondylitis, Reiters syndrome, lupus
erythematosus, gout, collagen disease, soft tissue rheumatism, infectional arthritis,
chronic arthropathy as e.g. degenerative arthropathy, secondary inflammation in
degenerative arthropathy, periarticular inflammation of shoulder,
Back pain, eg. pains in the lumbar region (after eliminating of kidney disorders),
sciatic neuralgia, condition after the operation of nucleus pulposus,
Soft tissues rheumatism as e.g. tendinitis, or peritendinous inflammation;
Consequences of accidents and other injuries, f.e. sprain, dislocation, muscle
strain, burn, oedema after injuries, Sudecks atrophy disease- post-traumatic
osteoporosis;
Surgical diseases, e.g. contracture in joints, oedema after the surgery of bresast,
hand, jaw, pain in the scars, bone abscesses, fistula, local infections,
haemorrhoids, pruritus of the anus, acute superficial inflammation of the veins,
epistaxis, acute inflammatory conditions within abdominal cavity,
Neurological diseases like spastic hemiparesis and spastic paraparesis,
disseminated inflammation of brain and cord, myasthenia, Parkinsons disease,
different pains, inflammation of anterior horns of the spinal cord, acute neuritis;

260

It should be emphasised that cryotherapy can be used in case of paroxysmal tachycardia,

diabetes varices of lower limbs, systemic lupus erythematosus, systemic scleroderma.
Because of strong stimulation of immunological and endocrine systems, cryotherapy of the
whole body is more and more often applied in:

Biological renewal, and in building of immunological immunity in patients, who

often suffer from infectious diseases (angina, flu, catarrh, inflammation of
respiratory tract) and it may be a perfect noninvasive alternative to different kinds
of vaccines;
In case of intensive exercise, e.g. sportsmen both before and after the competition;
In case of reducing obesity and cellulite;
Treatment of mood depression of patients with depression;
CONCLUSIONS

At present, four cryochambers with retention of cold, using liquid air, were installed in the
Czech Republic. The interest in building the new ones is even increasing for to be used not
only for rehabilitation centers but for medical applications and research too.
REFERENCES
1. Yamauchi, T.,Nogami, S.,Miura, K.: Various aplication of extreme cryotherapy and
strenous excercice program focusing on chronic reumatoid arthritis. Physiotherapy and
rehabilitation 5. 1981.
2. Fricke R.: Ganzkrperkltherap in einer Klter-Kammer mit Temperaturen 110
Celsia. Z., Phys.Med.Bahns.Klin., 18 (1989) 43.
3. Zagrobelny Z. a kol.: Krioterapia miejscowa i oglnoustrojowa, Wydawnictwo
Medyczne Urban & Partner, Wroclaw 2003.
4. CryoFlex Poland Sp. Z o.o.: The Usage Description of a Cryogenic Cabin ( Chamber
for Cryotherapy -110 C to -150 C).
5. Brojek, W: Kryoterapia uvagi oglne. Balneologia Polska, tom XLVIII, nr.1, 2006
6. Strnad,P.,Fortkov,L.: Terapeutick aplikace nzkch teplot.In:Sbornk konference
XXVIII. Dny lkask biofyziky,Valtice,kvten 2005.
7. Strnad,P.,Fortkov,L,: The Whole-Body Cryotherapy. In: Bratislava Medical
Journal,2006, 107(4), XXIXth Days of Medical Biophysics,Bratislava, May 2006.
8. Steinerov,A.,Korotvika,M.,Racek,J.,Zeman,V.,Strnad,P.,Bajgar,M.: Posouzen vlivu
celotlov kryoterapie na lidsk organismus.In:Sbornk XIV.sjezdu Spolenosti
rehabilitan a fyzikln medicny, Luhaovice,duben 2007

261

POSTERS:
HELIUM
TEMPERATURE
TECHNIQUES

263

CR08-18
EXERGETIC ANALYSES USABLE FOR CONTROL THE
OPERATION PARAMETERS OF THE HELIUM PROCESSING
PLANT FOR DIFFERENT WORKING CONDITIONS.
Gherghinescu S.
National Institute of R&D for Cryogenics and Isotopes Technologies (ICIT), Rm. Valcea,
code 240050 Uzinei 4, CP10, Valcea, Romania
ABSTRACT
The present work proposes several analyses and control models of a helium liquefaction
plant.
The monitored functional parameters of a cryogenic plant are temperature and pressure;
based on these two, enthalpy and entropy can be then easily derived. These will provide the
feedback for the control loop which will control entire plant.
The LABWIEW software is used for data acquisition and has great stability and safety data
acquisition.
Based on the experimental data, we can calculate the enthalpy and entropy for obtaining the
variation of exergy in different working conditions.
The main objective of this paper is to develop a method to optimize the cryogenic cycle.
The optimization would determine the working regimes (temperature, pressure, flow,
power consumption, etc.).
The contribution of the present work consists in the fact that the soft is easy to use, provides
real time response and convenient accessibility. This can be done by calculating the
enthalpy and entropy by a simplified method based on two-variable function.
Key-Words: exergetic analyze, monitored functional parameters.
THEORETICAL MODEL FOR THE CALCULATION OF THE NONREVERSIBLE LOSSES OF EXPANSION AND COMPRESSION
Losses ir_exp = c p T 0

exp =

h1 h2
h1 h2 s

Losses ir_comp = c p

1 k

p1 k

ln 1 exp 1
p2

comp =

h2s h1
h2 h1 k 1

p2 k

1
p

1
T 0 ln 1 +

comp

265

k 1

p 1 k

p 2

1k
1

p2 k

p1

T2

T2s

T2

P
1

T2s

22

T2

T1
1

2s
S

expansion

compression

Fig. 1 Non-reversible process of expansion and compression in T-S coordinates

MODEL OF CALCULATION THE NON-REVERSIBLE LOSSES TROUGH HEAT
TRANSFER

E 1` E 1`` = E 2`` E 2 ` + Losses ir _ T

dE 1 = E 1` E 1``
dE 2 = E 2 `` E 2 `

dE 1 = m 1 [ ( h1' h1' ' T 0 (s1' s1' ' ) ]

dE 2 = m 2 [ ( h2 ' h2' ' T 0 (s 2' ' s 2' ) ]

E _ T =

dE 2
dE 1

Losses ir _ T = dE 1 dE 2

E1
E2

E1

dQ

2
Fig.2 Variation of exergy in heat transfer process
266

E2

dE 1 , dE 2
1
m

2
m

- exergy of the heat exchanged between the two fluids

- flow rate of cool and heat fluid

MODEL OF CALCULATION THE NON-REVERSIBLE LOSSES TROUGH

THROTTLING (J-T EFFECT)
The J-T effect is the main source of irreversible losses. In our installation we hawe only one
J-T type proces (the transformation 10-11).

Losses ir_trottle = T 0 (s11 s12 )

T
T0

293K

1
5

14

70K

13

64K
4
16K
7K

Ts

4.2K

12

10

11

Fig. 3 Thermodynamic transformations of cryogenic cycle in T-S coordinates.

267

Kaese
r

10 g/s

15

290 K

E201A
2
14

143 K A A

E201B
7 g/s

70 K
3
13
4

Turboexpander 1

3 g/s

E202 ,
5

E203A

34 K

16 K

12

E203B ,

E204

7 g/s
6
9

Turboexpander 2

10 g/s

C205
7K

4.2 K

10

11

He liquid

Fig. 4 The cryogenic cycle of helium.

Considering that the throttle process takes place between certain pressure limits we may
write a polynomial equation which will help us to determine the temperature at the end of
the throttle process.

268

10

11
12

Y%

He liquid

Fig. 5. Fases separation in dewar of helium.

Knowing the input temperature in the Dewar vase with liquid helium we can determine the
liquid percentage (y):
2

y = 0.0357 T 11 0.6974 T 11 + 3.3141

y (%) at P11=1.3bara
LABWIEW PANEL FOR DATA ACQUISITION

CONTROL PANEL FOR DATA ACQUISITION AND SYSTEM MANAGEMENT

The above presented control panel was realised in LABWIEW software and has two main
functions:
269

data acquisition from the system

work process simulation on respect of the above equations

Fig. 7 Simulation of working process in Labwiew software based on experimental data.

CONCLUSIONS
The exergetic analyses allow a global evaluation of cold losses with impact on cryogenic
installation performances. It also ensures a stationary regime by a on-line control of the
working parameters using the LABWIEW software.
The mathematical model used to compute different parameters gives us results in good
agreement with the experimental data. The exergetics analyses may be extended over
complex cryogenic systems, where a precise parameters control is required to reaches the
desired working parameters. This is necessary for all systems where very short setting-up
time is required.
BIBLIOGRAPHY
1.
2.

3.

Adrian Bejan, George Tsatsaronis, Michael Moran Thermal Design &

Optimization ISBN 0471584673 Wiley 1996.
Adrian Bejan Entropy Generation Minimization Wiley 1996.
Philip Thomas Simulation of Industrial Processes for Control Engineers
Butterworth-Heinemann 1999.

270

CR08-09
SPLIT PULSE TUBE CRYOCOOLER WITH INNOVATIVE
DOUBLE-PISTON LINEAR COMPRESSOR
Kaiser G., Albert S., Schmidt J., Heidrich R., Binneberg A., Klier J.
Institut fr Luft- und Kltetechnik gemeinntzige GmbH, Dresden, Germany
ABSTRACT
A split pulse tube cryocooler with double-piston linear compressor is developed. The
compressor is driven via an innovative electro-dynamic linear motor operating by use of a
moving coil and core principle. This new development offers several advantages compared
to moving coil or moving magnet systems. It is possible to generate higher forces, power
and efficiency like in the case of moving coil systems. The reluctance force of the magnetic
system can be used as a substitute for mechanical springs. In this way the average position
of the mover is assured as for moving magnet systems. After an introduction into the new
driving principle, we present the theoretical and experimental results.
INTRODUCTION
There exist a market request for highly reliable medium power range cryocoolers for the
application in the field of cryo-conservation (cryogenic storage and carrier vessels), cooling
of high-Tc superconductor magnetic bearings for centrifugal machines and flywheel energy
storages and also cryogenic test chambers. To comply with this request a project for the
development of a 10 20 W @ 80 K split pulse tube cryocooler was elaborated.
During the project, the development of an innovative electrodynamic linear motor drive
was emphasized. The idea behind was to combine the advantages of the moving-coil and
the moving-magnet system and to eliminate their disadvantages. By use of moving-coil
motors it is possible to generate large electrodynamic Lorentz-forces. The disadvantage is
that there are additional means like mechanical springs, required in order to set the average
position of the linear motor mover. Using a moving-magnet system this problem is possible
to overcome. The reluctance force can be used to set the average mover position. The
forces of a moving-magnet system are limited by the mass of the moving magnet.
An innovative moving-coil and -core system was therefore developed and tested in a
double-piston linear motor driven inertance pulse tube cryocooler. The details of the drive
and a simulation model of the cryocooler are described. The experimental results are
presented and discussed.
1. THE LINEAR MOTOR
1.1 Principle
The principle design of the moving-coil and -core (MC&C) motor is shown in Figure 1.
The stator consists of a axially magnetized permanent magnet ring with two pole rings at
both sides. The moving part consists of the central magnetic core and the coil for the drive.

271

p e rm a n e n t m a g n e t p o le r in g
c o il

m a g n e t ic c o r e

Figure 1: Principle design of the MC&C motor

Without driving current, the mover is in its average position. Each shifting in axial direction
leads to a force generated by the magnetic field, which tends to minimize the field energy
of the magnetic system (reluctance force).
If a current flows through the magnetic coil, the windings are directed in such a way that a
Lorentz-force is generated against the reluctance force of the magnetic system leading to
axial movement. The combination of the mover mass and the reluctance spring lead to a
resonant system. Inside the assembled cryocooler an additional pneumatic spring occurs
leading to an increase of the resonant frequency.
1.2 Motor model and calculation
In order to calculate the motor performance a simulation model (EXCEL sheet) was
developed. Two different versions of the motor were calculated in order to operate with an
AC power supply of 12 V or 24 V respectively. The input data are presented in the
following table:
Parameter

Value

Mover stroke

9 mm

Electric load

250 W
Figure 2: Motor input data

A NdFeB permanent magnet was given with the following parameters:

Parameter

Value

Remanence

1.20 T

Coercitive field strength

905 kA m-1
Figure 3: Permanent magnet data

272

The calculation gives the following parameters for the motor:

Parameter

Value

Magnetic induction

438 mT

Number of windings, wire diameter

136, 1.4 mm (12 V) / 275, 1.0 mm (24 V)

DC resistance

0.206 Ohm (12 V) / 0.818 Ohm (24 V)

Force coefficient

8.23 N A-1 (12 V) / 16.7 N A-1 (24 V)

Mechanical power

166 W

Coefficient of performance

66.4 % (12 V) / 67.1 % (24 V)

Figure 4: Motor data

1.3 Motor measurements

Figure 5 shows the assembled linear drive and the motor mover. In order to get information
about the actual performance of the manufactured motors the magnetic induction was
measured locally inside the gap by use of a Model 8202 Honeywell Hall magnetometer.
The results for the two motors are given in Figure 6.

Figure 5: Linear motor and MC&C mover

The magnetic induction inside the main part of the motor gap is just 10 % less than
calculated. Considering the scattering field, which was not included in the calculation, this
deviation represents a good result. Load operation at 500 W of both motors inside the
cryocooler have shown a voltage range between 26.4 V and 28.0 V and currents between
22.8 A and 23.1 A in the frequency range between 45 Hz and 55 Hz. These current and
voltage observations correspond with the observation of a 10 % less force coefficient.

273

0.40

Magnetic induction [T]

0.35

0.30

0.25

0.20

0.15

10

20

30

Position [mm]

Figure 6: Magnetic induction inside the motor gap for two different motors

2. THE CRYOCOOLER
2.1 Cryocooler model and calculations
The cryocooler was modelled by use of SAGE (Gedeon Assoc.). The calculation was
performed under the following input conditions:
Parameter

Value

Swept volume

12.5 cm per cylinder / 2 cylinders

Average pressure

25 bar

Motor parameters

12 V (24 V) / 50 Hz, 500 W el. load

Figure 7: SAGE model input parameters

Under these given conditions one should obtain the following experimental results:
Parameter

Value

Cooling power @ 80 K

21 W

Minimum temperature

60 K
Figure 8: SAGE model results of calculation

2.2 Cryocooler testing and results

Figure 9 shows the MC&C pulse tube cryocooler integration in the test rig. The test setup
consists of a frequency- and voltage-variable power supply for the linear motor operation, a
vacuum system for the thermal insulation, a measuring system for temperature and pressure
wave recordings during operation. A cooling water supply was used to remove heat from
the motor housings and from the hot end heat exchanger and precooler.

274

Figure 9: MC&C pulse tube cryocooler test setup

Figure 10 shows the pressure variation, P, inside the compressor at an electric load of
500W. The peek-to-peek value of the pressure variation inside the compressor is 4.1 bar.
The simulation shows a pressure variation of 4.5 bar. This is in good agreement with the
experimental results, taking into account additional void volume which is not considered in
the simulation model.
3

Pressure variation [bar]

-1
30 bar
25 bar
20 bar

-2

-3

10

20

30

40

50

Time [ms]

Figure 10: Pressure variation at 500 W electric load for different average pressures and frequencies

First cool down measurements, at P = 2.2 bar, have shown a minimum temperature of
138 K (see Figure 11). Next cool down runs with the higher pressure variation of 4.1 bar
should lead to much lower temperatures. Further cold head optimization like the integration
of flow-straightening means and the optimization of the inertance tube are currently under
progress.

275

160

Minimum temperature [K]

155

150

145

IT ID = 2 mm, L = 300 mm
IT ID = 2 mm, L = 400 mm
IT ID = 2 mm, L = 500 mm
IT ID = 4 mm, L =2300 mm

140

135
20

22

24

26

28

30

Average pressure [bar]

Figure 11: Average pressure dependence of the minimum temperature for different inertance tubes

CONCLUSIONS
A new type of moving coil and -core electrodynamic linear motor was developed. This
motor is used as a drive of a double piston linear motor compressor for an intertance pulse
tube cryocooler, designed and tested for the medium cooling power range. The
experimental results for the magnetic and electric motor measurements show good
agreement with the calculations. A deviation of 10 % less force coefficient was observed.
The calculation for the inertance pulse tube cryocooler shows the possibility to generate
21 W cooling power at 80 K for an electric load of 500 W at the linear motors. The pressure
variation inside the compressor cylinder is expected to be 4.5 bar.
Experimental observations during the test of the cryocooler show a maximum pressure
variation of 4.1 bar which is in good agreement with the simulation results. During cool
down experiments, at lower pressure variations, a minimum temperature of 138 K was
reached. For higher pressure variations a lower temperature is expected. Improvement of
the cooling performance data by use of flow-straighteners and the optimization of the phase
shifter is currently under progress.
ACKNOWLEDGEMENT
This work was financially supported by the German BMWI under contract no. IW041392.

276

CR08-01
VERY-LOW TEMPERATURE THERMAL CONDUCTIVITY OF
STRUCTURAL MATERIALS FOR LARGE CRYOGENIC
EXPERIMENTS
Ventura G., Barucci M., Martelli V., Risegari L.
Department of Physics, University of Florence, Italy
ABSTRACT
In large cryogenic experiments like CUORE (Cryogenic Underground Observatory for
Rare Events) a mass of several tons is sustained by rods which must possess outstanding
mechanical parameters together with a very low thermal conductivity. We have measured
the thermal conductivity of some structural materials below 1K. A comparison is done
among data for metallic alloys, polymers and reinforced materials.
INTRODUCTION
CUORE (Cryogenic Underground Observatory for Rare Events) [1] experiment consists of
a large array of detectors for the search of (0) decay, to be installed in 2010 at the
underground National Laboratory of Gran Sasso (LNGS). A packed array of 988 TeO 2
detectors will be cooled down to ~10 mK. The experiment is housed in a large cryogen-free
cryostat cooled by five pulse tubes and one high-power specially designed dilution
refrigerator [2].
The cryostat is ~3 m high and has a diameter of ~1.6 m. About 510 3 kg of lead shielding
are to be cooled to below 1 K and a mass of 1.510 3 kg must be cooled to 10 mK. Several
tie-rods sustain the different parts of the experiment. One end of each rod is at low
temperature (10 mK for the detector frame, 50 mK for the coldest radiation shield and lead
shield, 700 mK for the shield linked to the still), the other end being, in some cases, at room
temperature. A thermalisation of the rods at the temperature of the first and the second
stage of the pulse tubes will be realized. Hence also the value of the thermal conductivity of
the material up to room temperature is important.
At the lowest temperatures, the thermal conductivity has great influence in establishing the
thermal load on the dilution refrigerator. The thermal conductivity of the structural
materials candidates for such tie-rods is usually known down to 4 K.
Here we present data of thermal conductivity below 1 K of PERMAGLAS ME771 [3], a
new oriented glass epoxy laminate material. A comparison is also done with other
materials, such as Torlon, Kevlar and metallic alloys, candidates for the realization of the
CUORE tie-rods.
1. EXPERIMENTAL SET UP AND MEASUREMENTS
The thermal conductivity of PERMAGLAS ME771 was measured along the direction of
the reinforcing fibres by the longitudinal steady heat flow method.
The experimental set up is shown in Fig. 1. The cylindrical sample has a length
L=(89.60.2)mm and a sectional area A=(501)mm2. At room temperature, the form factor
is g=A/L=(0.5610.016)mm.

277

Figure 1: Set up of the experiment (shorter sample)

The thermal contacts at the ends of the sample have been realized by means of two copper
blocks and two copper screws 4 mm in diameter. Since the thermal contraction of
PERMAGLAS ME771 is lower than that of copper [4,5], the thermal contact between the
blocks and the two ends of the sample becomes better on cooling. A SMD (Surface Mount
Device) NiCr heater and a RuO2 thermometer were glued onto the two copper blocks at the
ends of the sample.
The electrical connections to the heater and to the thermometer were made with NbTi wires
(= 25 m). The NbTi wires were electrically connected by tiny crimped CuNi tubes. At
the ends of the NbTi wires a four lead connection was adopted. The bottom copper block
was screwed onto a copper support in thermal contact with the mixing chamber of a
dilution refrigerator. Another RuO2 calibrated thermometer was used for the measurement
of the mixing chamber temperature. Both thermometers were calibrated by means of a SRD
1000 (Superconductive Reference Device) [6,7] and a NBS-SRM 767a fixed point device
[8]. A copper shield, in thermal contact with the mixing chamber of the dilution
refrigerator, surrounded the experiment.
A known power P was supplied to the upper end of the sample to establish a difference of
temperature T1 T0 between the ends of the sample. A LR-700 a.c. resistance bridge for the
upper thermometer (measuring T1) and a Picowatt for the lower thermometer (measuring
T0) were used. A four wire IV source delivered the power P to the heater.

278

By derivation of the integrated power P(T1) at constant T0:

T

1
A 1
P ( T1 ) = k ( T ) dT = g k ( T ) dT
L T0
T0

(1)

the thermal conductivity k(T) can be obtained.

A second set of measurements was carried out on a sample of g=(1.100.02)mm to cover
the lowest temperature range. The second run on the shorter sample gave the same value of
k in overlapping temperature range. This result guarantees that the effect of the contact
thermal resistances do not influence the value of thermal conductivity.
2. RESULTS AND CONCLUSIONS
The measured thermal conductivity of PERMAGLAS ME771 in the 100mK-1K
temperature range is shown in Fig. 2. Data of k(T) can be fitted by the formula:
k(T) = aT n = (9,000.01)10-5 T (1.8800.002)

(2)

Figure 2: Thermal conductivity of PERMAGLAS ME771 in the 100 mK- 1 K temperature range

The maximum estimated error on k(T) is 4%.

279

Figure 3: Thermal conductivity of PERMAGLAS ME771 compared with other candidate materials
for the CUORE supports

In Fig. 3 the thermal conductivity of PERMAGLAS ME771 is compared with that of other
candidate materials for the CUORE supports [9-12]. From a purely thermal point of view,
the PERMAGLAS ME771 is a good candidate for the realization of the supports. Of
course mechanical data will play a decisive role in the choice of the material.
REFERENCES
1. Ardito, R., et al., CUORE: a cryogenic underground observatory for rare events,
http://arxiv.org/abs/hep-ex/0501010 (2005)
2. Nucciotti, A., et al., Design of the cryogen-free cryogenic system for the CUORE
experiment, J. Low Temp. Phys. (2008) 151
3. http://www.permali.com/english/pdf_files/Permaglas.pdf
4. Roechling, private communication
5. Pobell, F., Matter and Methods at Low Temperatures, Springer-Verlag, Berlin (1995)
6. Bosch, W.A., et al., Status Report on the Development of a Superconductive Reference
Device for Precision Thermometry below 1 K, in Proceedings of the 8th International
Symposium on Temperature and Thermal Measurements in Industry and Science
TEMPMEKO 2001, VDE Verlag, Berlin (2001) 397-401
7. Schottl, S., et al., Evaluation of SRD1000 Superconductive Reference Devices, J. Low
Temp. Phys. (2005)138 941-946
8. Schooley, J. F., Soulen, R. J., Jr., Evans, G. A., Jr., Preparation and Use of
Superconductive Fixed Point Devices, SRM 767, NBS Special Publication 260-44,
Washington, DC, National Bureau of Standards (1972) 1-35
9. Risegari, L., Barucci, M., Lolli, L., Ventura, G., Low temperature thermal conductivity
of Ti6Al4V alloy, J. Low Temp. Phys. (2008) 151

280

10. Barucci, M., Lolli, L., Risegari, L., Ventura, G., Measurement of thermal conductivity
of the supports of CUORE cryostat, submitted to Cryogenics (2007)
11. Ventura, G., Bianchini, G., Gottardi, E., Peroni, I., Peruzzi, A., Thermal expansion and
thermal conductivity of Torlon at low temperatures, Cryogenics (1999) 39 481-484
12. Ventura, G., Barucci, M., Gottardi, E., Peroni, I., Low temperature thermal conductivity
of Kevlar, Cryogenics (2000) 40 489-491

281

POSTERS:
HIGH TEMPERATURE
SUPERCONDUCTIVITY

283

CR08-14
DESIGN, FABRICATION AND TEST RESULTS
ON A CONDUCTION COOLED HTS MAGNET
Joonhan B., Seokho K., Kideok S., Myunghwan. S.
Korea Electrotechnology Research Institute, Changwon, Korea
ABSTRACT
This paper describes design, fabrication, and testing of the conduction cooled HTS magnet.
The magnet is composed of 22 double pancake coils. The magnet was conductively cooled
down to 5.6K with two stage GM cryocoolers. The temperatures of the HTS magnet were
measured in the charging and discharging process. The successful operation of the magnet
illustrates that the technology of cooling HTS magnet with GM cryocoolers is fully
established.
INTRODUCTION
The superconductor can carry larger current without electricity loss because of having no
resistance. The superconducting magnets using theses merits are utilized in
superconducting equipments such as MRI, NMR, SMES, magnetic separator,
superconducting generator and motor. Because the conventional superconducting magnets
are cooled with coolants such as liquid helium or nitrogen to keep their superconducting
properties, there are many disadvantages in this magnet type. The cryostats used to store the
liquid cryogens and keep the magnet cold are rather complex. When a quench occurs in the
magnet, sudden evaporation of a large amount of liquid coolants might be dangerous. Also,
the liquid cryogen to run the magnet is very costly and there are annoying to refill coolants
periodically. Recently, rapid progress of refrigerators and superconducting wires allows the
superconducting magnets to be operated without cryogen use. Since Hoenig demonstrated a
thermal design of the conduction cooled superconducting magnet combined with Gifford
McMahon (GM) cryocooler in 1983, many researches on the refrigerator-cooled magnets
have actively been performed world widely [1]. The most outstanding value of cryogen free
magnet lies in easy and safe handling, low running costs, and compact system. Also, the
cost for the electricity and the maintenance of cryocoolers would be less than 1/10 of that
for the cryogens [2].
This paper describes the design, fabrication and test results on the conduction cooled high
temperature superconducting (HTS) magnet for superconducting applications. The optimal
design of the magnet was conducted using the objective function of minimizing the total
required amount of the HTS conductor. The electromagnetic and mechanical behaviors of
the magnet are analyzed. The prototype magnet composed of 22 double pancake coils was
fabricated on the basis of design and analysis results. Finally, the performance of the
magnet was evaluated during charging and discharging operation.
1. OPTIMAL DESIGN OF THE HTS MAGNET
Generally, there are several types of HTS magnets such as solenoid, multiple solenoid,
toroid, and so on. Solenoid type coils are simple to design and easy to fabricate, but they
can not prohibit or confine stray field. Multiple solenoid coils show very good
characteristics on stray field, but it has very poor energy density. Toroid coil can be a
compromise proposal. A perfect toroid coil makes no stray field. The field is confined
285

inside coil, but it is very hard to realize such kind of coil. Instead, coils wound in pancake
or stacked pancake coils can be configured to simulate similar effect. Meanwhile, toroid
coils require more conductors than solenoid coils but fewer wires than multiple solenoid
coils [3]. In this study, the modular single pole double pancake coil (DPC) was selected in
consideration of its installation environment.
The 4-ply HTS conductor is used to design the HTS magnet. The conductor is composed of
two AMSC Bi-2223 tapes and two brass tapes which are soldered at each side of Bi-2223
tapes for the mechanical reinforcement. To minimize the minimum bending radius, the
thickness of the brass tape and solder were controlled precisely. Finally, it was wrapped
with kapton tape for electrical insulation. Table 1 shows the specifications of the 4-ply HTS
conductor. It is well known that the critical current of the HTS tape depends on the
direction and amplitude of the external magnetic field, so we confined the maximum
operating current of the magnet within 70% of the critical current in the 4-ply HTS
conductor. The main objective of the design is to find optimal dimensions of the magnet
with minimum conductor length. There are several constraints to be considered for the
design process, such as critical magnetic field, total HTS conductor length, and geometrical
constraints for supporting and cooling equipments. Among them, the total HTS conductor
length was selected as the main objective function.
Table 1: Specifications of 4-ply HTS conductor
Composition
Average width of the conductor

2 BSCCO-2223 tapes and 2 brass tapes

wrapped with 2 kapton tapes
4.5 mm

Average thickness of the conductor

0.77 mm

Critical tensile stress

150 MPa at room temperature

Critical bend diameter

150 mm at > 95% Ic retention

Critical current

480 A at 20 K, 3 T

Al bobbin

DPC 1

HTS coil

DPC 2

FRP spacer

DPC 11

DPC 12

DPC 21
DPC 22
500 mm
691 mm

Figure 1: The across section of the HTS magnet

286

Table 2: Design Results of the HTS magnet

Operating temperature [K]

20

Inner diameter [mm]

500

Outer diameter [mm]

691

Number of turns per DPC

262

Number of DPCs

22

Gap between DPCs [mm]

Operating current [A]

275

Height [mm]

330

Maximum parallel field [T]

3.92

Maximum perpendicular field [T]

2.49

Central field [T]

2.98

Stored energy [kJ]

605

Inductance [H]

16

Length of HTS conductor [km]

10.8

The across section of the HTS magnet is shown in Figure 1. The aluminum bobbin to
support DPC is coated with ceramic powder for insulation and plays a role of the cooling
plate to cool down the HTS magnet. FRP (fiberglass reinforced plastics) spacer of 2 mm
thickness is located in the middle of DPC for the winding arrangement. The design
operating temperature of the HTS magnet is 20K, which is attained by conduction cooling
with two GM cryocoolers. Table 2 shows the design results of the HTS magnet achieved by
Auto-Tuning Niching Genetic Algorithm [4].
2. ELECTROMAGNETIC AND MECHANICAL ANALYSIS
HTS magnets experience the electromagnetic force, which cause the instability of the
superconducting magnet and deformation of the conductors. The problem is more serious in
large scale HTS magnet. Therefore, it is important to consider the mechanical forces on
conductors of superconducting magnets caused by this electromagnetic force.
When electromagnetic forces and stress are balanced in a nonmagnetic material, the
following equations are valid.
B = 0 J

(1)

B = 0

(2)

J B + S = 0

(3)

Where, J is current density, B is magnetic flux density and S is a stress tensor. When the
operating current circulates in the solenoid HTS magnet, the magnetic flux density
distributions in the magnet are obtained from expression (1) and (2). Also, Lorentz force
interaction between the operating current and the magnetic field results in stress within the
magnet, which tend to burst the windings radially outward and crush it axially and it could
be computed by using equation (3).

287

Figure 2: The magnetic field distribution in the HTS magnet

Figure 3: The radial and hoop stress distributions in the DPC 11 of the magnet at current of 275 A

Figure 2 shows the magnetic field distribution in the HTS magnet. The strongest axial field
in the middle of the magnet was produced and it caused the maximum hoop stress on the
HTS conductor in outmost layer of the magnet. The evaluation of the maximum stress
within the windings during charging the magnet is important in order to keep the magnet
from the quench due to the movement of the HTS conductor. Assuming that the
microscopic stress distribution due to the shielding current in the each HTS conductor can
be neglected, the radial and hoop stress distributions in the solenoid HTS magnet are
achieved through the axis-symmetric two dimensional numerical analysis using the finite
element method. Poisson ratio of 0.35 and equivalent Youngs modulus were adopted to
calculate the stress within the magnet because 4-ply HTS conductors are composite
material [5]

288

The radial and hoop stress distributions in the DPC 11 of the magnet at the current of 275 A
are shown in Figure 3. The maximum radial and hoop stress are produced in the DPC 11
because it is placed in the middle of the HTS magnet. The 3.61 MPa of the maximum radial
stress was calculated at the center of the DPC 11. Also, 35.6 MPa of the maximum hoop
stress was computed at the innermost radius of DPC 11.
3. FABRICATION AND ASSEMBLY OF THE HTS MAGNET
3.1 Cooling system
The cryostat for the conduction cooled HTS magnet has the outer diameter of 1330 mm, the
height of 924 mm, and the weight of 1100 kg. The two stage GM cryocoolers (Sumitomo,
model 415D) were adjacently mounted on one side of the cryostat. The current leads and
radiation shield are cooled by the first stage of the cryocooler and the HTS magnet is
cooled down by the second stage of the cryocooler.
For improving cooling capacity, the HTS magnet was cooled down by using high purity
cooper braid for a flexible thermal link and the thermal stress relaxation at the low
temperature, which connected the magnet with the second stage of the refrigerator. 15
layers of MLI (multi-layer insulator) were wound on the HTS magnet to minimize radiation
heat from the room temperature. A pair of brass current leads was anchored between the
room temperature and the first stage of the cryocooler. Also, pairs of the HTS current leads
composed of AMSC Cryoblock wires were inserted between brass current leads and the
HTS magnet terminals for blocking conduction heat invasion through the brass current
leads. In order to reduce the thermal contact resistance at the many mechanical boundary of
the thermal path, each boundary surface was precisely controlled by mechanical machining
and indium foil or thermal grease (Lakeshore, Apiezon N) was used at the boundary for
providing good thermal contact.
3.2 HTS magnet
The 4-ply conductor was arranged on the aluminum alloy bobbin, which extract the heat
generation from the magnet and was wound from the midpoint of the whole length of the
conductor. The 4-ply conductor with length of 490 m was wound into a DPC of 262 turns.
The HTS magnet is composed of 22 DPCs and they are cooled by two stage GM
cryocoolers through conduction. The aluminum bobbins were anodized for the additional
electrical insulation and the dielectric strength was measure as 2 kV at the flat surface and
500 V at the edge. To avoid the conductor displacement under electromagnetic force or
detachment between the coil and the bobbin, a cryogenic resin was impregnated in vacuum
after winding and they are hardened at room temperature for 24 hours. The configuration of
the assembled HTS magnet with the top flange of the conduction cooled cryostat is shown
in Figure4.
3.3 Monitoring the temperature and quench of the magnet
Monitoring the temperature and voltage of the conduction cooled HTS magnet is important
issue to provide safe operation during the charge and discharge of the magnet. The total 24
Cernox temperature sensors and 21 E-type thermo-couples were installed along the thermal
path to find out the poor thermal contact point. 22 DPC voltages, 22 joint voltages and
current lead voltage were also measured through isolation amplifiers to detect the quench
signal in the magnet in the energizing and deenergizing process.

289

Figure 4: The assembled HTS magnet with the top flange of the cryostat
4. TEST AND DISCUSSION
4.1 Cool down
To reduce the initial cool down time, we circulated liquid nitrogen through the heat
exchanger, which acted as thermal sink connected to the second stage of the cryocoolers.
After the temperature of the heat exchanger reached around 170 K, supplying the liquid
nitrogen was stopped and the HTS magnet was cooled down only using the two
cryocoolers.
Figure 5 shows the cool down history for the major measuring points. It took 90 hours to
cool down the HTS magnet to the saturation temperature. By minimizing the heat
penetration through the support, the current leads and radiation, the final temperature was
5.6 K lower than the operation temperature in design. Before installing the cryocoolers, the
cooling capacity of the cryocoolers was measured and the heat penetration could be
estimated. According to the measured cryocooler temperature, the heat penetration by
conduction and radiation to the HTS magnet is estimated to be about 6 W considering the
two GM cryocoolers.
4.2 The temperature variations in the HTS magnet at different currnet charging rates
Figure 6 shows the temperature variations during charging the current to 275 A with ramp
rate of 2 A/s and discharging with a 1 dump resistor. The hot spot temperature at the
HTS magnet was 12.9 K during discharging operation. Fig 7 shows the hot spot
temperature in the HTS magnet when the current charging rate was varied from 0.2 to 2.5
A/s. In Figure 7, it was observed that the maximum temperature of the magnet rose as the
current charging rate increased. This can be thought that the removed heat loss by
cryocooler decreases as the ramp rate increases because the cooling time of cryocooler is
longer than charging time of the magnet

290

Figure 5: The cool down history for the HTS magnet

Figure 6: The temperature variations during charging and discharging the HTS magnet

Figure 7: Hot spot temperatures in the HTS magnet at different current charging rates
.
291

CONCLUSIONS
The optimal design of the HTS magnet for superconducting applications was carried out
with several constrain conditions and the prototype HTS magnet was fabricated on the basis
of design results. The thermal and electromagnetic behaviors of the magnet were performed
during charging and discharging operation. The results are as follows.
a)

Maximum perpendicular and parallel field density in the designed HTS magnet
are 2.49 T and 3.92 T, respectively.

b) On the basis of the field analysis results, The 3.61 MPa of the maximum radial
stress was calculated at the center of the magnet and the maximum hoop stress
was 35.6 MPa at the innermost radius of the magnet.
c)

The temperature of the HTS magnet was cooled down to 5.6 K after 90 hours
using two GM cryocoolers and the total heat penetration by conduction and
radiation to the HTS magnet is calculated to be about 6 W using the cooling load
map and the temperature of the cryocoolers

d) The hot spot temperature at the HTS magnet was 12.9 K in discharging with a 1
dump resistor after charging current to 275A with ramp rate of 2A/s.
e)

The maximum temperature of the magnet rise as the current charging rate increase
because the cooling time of cryocooler is longer than charging time of the magnet.

The results through this research will be utilized in the optimal design and stability
evaluation of the conduction cooled HTS magnet for superconducting applications.
ACKNOWLEDGEMENT
This work was supported by Electric Power Industry Technology Evaluation & Planning,
Republic of Korea.
REFERENCES
1. Heonig, M. O., Design concepts for a mechanically refrigerated 13K superconducting
magnet system, IEEE trans. on Magn. (1983) MAG-19 3 880-883
2. Katano, S., Minakawwa, N., Hasebe, T. and Sakuraba, J., New cryocooler-cooled
superconducting magnet: A 13.5T high-field split-pair coil magnet for neutron
scattering, Physica B (2006) 385-386 1300-1302
3. Schonwetter, G., SMES solenoids with reduced stray field, IEEE trans. on Magn (1994)
30 2636-2639
4. Wooseok, K., Design of HTS magnets for a 600 kJ SMES, IEEE trans. Appl. Supercond.
(1994) 16 2 620-623
5. Crandall, Dahl and Lardner, An introduction to the mechanics of solid-2nd ed., McGrawHill Inc., New York, USA (1983)

292

CR08-23
ANALYSIS ON THE QUENCH AT THE CONDUCTION-COOLED
JOINT BETWEEN HTS WIRE AND NORMAL CONDUCTOR
Bae D.K.1, Bae J.H.2, Lee D.-Y.3, Lee S.-J.3, Park J.-S.3,
1

Department of Safety Engineering, Chungju National University, Chungju, Korea

2
Korea Electrotechnology Research Institute, ChangWon, Korea
3
Division of Energy & Electrical Engineering, Uiduk University, Kyongju, Korea
ABSTRACT

The heat generated in the high-Tc superconducting (HTS) devices is related with the cost
efficiency and safe factor of HTS devices. This paper deals with the quench at the
conduction-cooled joint between the HTS wire and normal conductor. The 3-D numerical
simulation of this phenomenon was implemented and compared with the experimental
results. The experiment was implemented with the HTS wire mounted on the copper blocks
cooled with a Gifford McMahon (GM) cryocooler.
INTRODUCTION
With the successful commercialization of the Bi-2223 powder-in-tube type HTS wire (1 G
wire) and the YBCO coated conductor type HTS wire (2 G wire), the HTS magnet for high
magnetic field and electric power applications has been researched & developed actively
[1], [2]. To name a few, there are superconducting magnetic energy storage (SMES),
magnetic resonance imaging (MRI), HTS fault current limiter, HTS motor, HTS cable and
HTS transformer. Especially, the conduction-cooled HTS magnets are widely developed
actively [3], [4]. The conduction-cooling system has many advantages such as easy
operation, compact size, wide range of operating temperature and so on. The operating
temperature range of recently developed conduction-cooled HTS magnets was from 4.2 K
to 50 K [1]-[4]. But the thermal runaway occurred easily in many conduction-cooled HTS
magnet systems. The thermal runaway could degrade or destroy the HTS coil. Although the
possibility of the thermal runaway of the HTS magnet is small, because the most of the
conduction-cooled HTS magnets are DC magnet and the operating current is usually much
less than the critical current of the magnet, it should be noticed that AC loss, pulse current,
larger operating current to reduce the cost, and electric power interruption may cause the
thermal runway. The possibility of the thermal runaway should be considered when
designing the conduction-cooled HTS magnets.
This paper deals with the quench at the conduction-cooled joint between the HTS wire and
normal conductors, as an initial step in the analysis on the cause of the thermal runaway
phenomenon in the conduction-cooled HTS magnet. The normal conducting copper block
is usually used as the current lead in the HTS magnet system. The copper current lead may
be heated by the AC loss of the HTS wire, pulse current and electric power interruption, at
which point the thermal runaway phenomenon initiates. The minimum quench energy
(MQE) of HTS wire is several orders higher than that of the LTS wire. However, because
the normal zone propagation (NZP) velocity of the HTS wire is several orders slower than
that of LTS wire and even very slow, the initiated thermal runaway causes damage to the
HTS wire and finally could cause the burnout of the HTS wire. So it is important to
evaluate the thermal runaway condition of the conduction-cooled HTS magnet.

293

In this study an YBCO coated conductor (CC) was mounted on the copper current leads.
The mounted CC and copper leads were cooled down with a Gifford McMahon (GM)
cryocooler. The temperature difference between the CC and the leads was made
deliberately in the system so that the experiment of thermal runaway could be implemented.
1. SIMULATION OF JOINTING PART
1.1 Connecting methods between HTS wire and normal conducting lead
Figure 1 shows two connecting methods between HTS wire and normal conducting copper
current lead. In method (a), both sides of HTS wire are in contact with the copper blocks. In
method (b) one side of HTS wire is in contact with the copper block. The bottom side of the
copper block is connected to the cold head of cryocooler. The distance between HTS wire
and the bottom of the copper block of method (a) is closer than that of (b). So, method (a) is
more efficient in cooling connecting parts than method (b) as a conduction-cooled
connecting method.
1.2 Numerical simulation of jointing part
Because the numerical analysis of method (b) in Figure 1 can also simulate method (a) in
Figure 1, method (b) was considered as the 3-D analysis model as shown in Figure 2. 3-D

(a)

(b)
Figure 1: Connecting methods between HTS wire and normal conducting lead, (a) HTS wire
mounted between copper blocks, (b) HTS wire mounted on copper block

294

Finite element method (FEM) was used in the analysis. The boundary conditions of this
analysis can be summarized as follows:
1) Perfect insulator boundary condition of whole outside surfaces except the bottom
surface of the copper block.
2) Specified temperature boundary condition (connected to cryocooler) of the bottom
surface of the copper block.
3) No thermal resistance between the HTS wire and the copper block.
The thermal conductivity of used copper was like Figure 3 [5]. The temperature of the
cryocooler was set to 75 K. In this calculation, AC loss of HTS wire was not included.

HTS wire

Copper block

Thermal conductivity (W/mK)

Figure 2: 3-D analytic model with mesh

10

10

10
Temperature (K)

Figure 3: Thermal conductivity of copper [5]

295

Figure 4: Distribution of temperature in connecting part (75 K, 300 A-200 Hz)

HTS wire

HTS wire

Copper block

HTS wire

HTS wire

Copper block

Copper block
Copper block

Figure 5: Distribution of temperature in connecting part (75 K, 300 A-600 Hz)

Figure 4 shows the distribution of the temperature in the connecting part with the transport
current of 300 A-200 Hz. The temperature of the bottom side was specified to 75 K. The
temperature of the HTS wire was about 76.5 K, and then the difference between HTS wire
and the bottom of copper block was about 1.5 K. Figure 5 shows the distribution of the
temperature in the connecting part with 300 A-600 Hz. As the frequency of the current
increases, the difference of the temperature became slightly bigger than that of 200 Hz. The
temperature difference between the HTS wire and the bottom of the copper block was about
1.9 K. The results of this simulation mean that AC current and/or pulse current may cause
the difference of the temperature between main HTS coil and normal jointing part, which
will be the source of thermal runaway.
2. EXPERIMENT
2.1 Experimental setup
Figure 6 shows the experimental setup for the deliberate thermal runaway test. Three
thermal sensors were mounted on the central part of HTS wire and two connecting parts,
respectively. Three voltage taps were also mounted on the HTS wire. One was on the center
of the wire (central voltage tap) and another was on the connecting part (Cu lead voltage
tap) and the third was on the whole wire (whole wire voltage tap) Left and right copper
blocks were connected to the cryocooler via thin glass fiber reinforced plastic (GFRP) plate
to make deliberate temperature difference. The thickness of the GFRP was 1 mm. The
copper blocks at the center of the system were connected directly to the cryocooler. The
heater was installed on the cold head of the cryocooler to control the temperature.
The HTS wire used for this study was CC from AMSC Inc. The specifications of the used
CC are shown in Table 1.

296

To make deliberate temperature difference between center point and connecting part, the
whole system was cooled down to 10 K and warmed up to 80 K, and then re-cooled down
the testing system to 75 K. Due to the GFRP plate between copper block and cold head,
there was some delay in cooling the connecting parts. When the temperature of the center
part reached 75 K, transport current began ramping-up.

Center voltage tap

and center temp. sensor
Block2 temp. sensor

Cu lead voltage tap

and block1 temp. sensor

Figure 6: Experimental setup for deliberate thermal runaway test

Average thickness

0.18-0.22 mm

Minimum width

4.27 mm

Maximum width

4.55 mm

Minimum double bend diameter (Room Temperature)

25 mm

Maximum rated tensile stress (RT)

250 MPa

Maximum rated wire tension (RT)

19.3 kg

Maximum rated tensile strain (77 K)

0.3 %

Minimum Ic (77 K, self-field 1 V/cm)

80 A

Table 1: Specifications of used CC

3. EXPERIMENTAL RESULTS AND DISCUSSIONS

3.1 Measurement of DC critical current
Figure 7 and 8 show the voltage and current characteristics of used CC at 75 K and 80 K
respectively. The measured Ic at 75 K was 90 A and that at 80 K was 47 A. Four-probe

297

method and 1 V/cm criterion were used in the measurement. The whole voltage tap was
used in these measurements. The Ic at 75 K was about two times larger than that at 80 K.
3.2 Experiment of deliberate thermal runaway
Figure 9 shows the temperature and transport current profile during the experiment. The
delay of the temperature falling at the copper blocks was shown in Figure 9. When the
transport current began ramping-up, the temperature at the connecting part was about
78.5 K and that at center was about 75 K. Block temperature was cooled down to 76.4 K at
the ending of the experiment. The effective temperature difference between center and
block was about 1.4 K.
10
9
8
Voltage ( V)

7
6
5

Ic=90A

4
3
2
1
0
0

20

40

60

80

100

Current (A)

Figure 7: Voltage and current characteristics of used CC at 75 K

10
9
8

Voltage ( V)

7
6
5
Ic=47A

4
3
2
1
0
0

10

20

30

40

50

Current (A)

Figure 8: Voltage and current characteristics of used CC at 80 K

298

82

100

81
80

79
60

78
77

40

76
75

Center temp.
Block1 temp.
Block2 temp.

74
73

Current (A)

Temperature (K)

80

20

72
0

100

200

300

400

500

Time (sec)
Figure 9: Temperature and current profile

Figure 10 shows the experimental results of the deliberate thermal runaway. The length of
center and Cu lead voltage tap was 4 cm and that of whole wire was 100 cm. Because the
temperature of the Cu lead was of 1.4 K higher than that of center, the voltage on the Cu
lead tap rose firstly. When the ramping-up transport current reached to 95 A, CC was
burned out.
The purpose of the experiment was to verify the possibility of the burning out of the HTS
conductor at the joint between HTS wire and normal conducting lead in the conductioncooled HTS magnet system. High energy is concentrated into a small zone when quench
occurs in the HTS wire cooled by cryocooler. The high energy concentrated into a small
zone gives severe damage to HTS magnet. The temperature difference between HTS wire
part and Cu lead part due to AC loss of HTS material, pulse current, power interruption and
so on should be considered when design the conduction-cooled HTS magnets.

299

5
Center
Whole wire
Cu lead

Voltage (mV)

3
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
0

20

40

60

80

100

Current (A)
Figure 10: Deliberate thermal runaway

CONCLUSIONS
In this paper, temperature of connecting part between HTS wire and normal conducting
copper lead was numerically calculated by using 3-D finite element method and the
deliberate thermal runaway experiment using conduction-cooled CC was implemented. The
temperature of normal conducting lead may be higher than that of HTS winding part so the
critical current of the HTS conductor on the normal conductor decrease. The difference of
temperature between HTS winding part and normal conducting part causes the degradation
of HTS wire. The possible temperature difference between HTS wire part and Cu lead part
should be considered in the design of conduction-cooled HTS magnets.
ACKNOLEDGEMENT
This work has been supported by KESRI (R-2005-7-068), which is funded by MOCIE
(Ministry of commerce, industry and energy).
REFERENCES
1.
2.
3.
4.
5.

Waynert, Joseph A., Boenig, Heinrich J., Mielke, Charles H., Willis, Jeffrey O., and
Burley, Burt L. Restoration and testing of an HTS fault current controller, IEEE Trans.
on Applied Superconductivity (2003) 13, 1984-1987
Kalsi, S. S., Aized, D., Connor, B., Snitchler, G., Campbell, J., Schwall, R. E.,
Stephanblome, Th., Tromm, A., and Kellers, J., HTS SMES Magnet Design and Test
Results, IEEE Trans. on Applied Superconductivity (1997) 7, 971-976
Seong, K.C., Kim, H.J., Kim, S.H., Park, S.J., Woo, M.H., Hahn, S.H., Research of a
600 kJ HTS-SMES system, Physica C (2007), 463-465, 1240-1246
Obana, T., Tasaki, K., Kuriyama, T., Okamura, T., Thermal stability analysis of
conduction-cooled HTS coil, Cryogenics (2003) 43, 603-606
Iwasa, Y. Case Studies in Superconducting Magnets, Plenum Press, New York, USA
(1994) 385

300

CR08-37
ANALYSIS OF THE MAGNETIC PROPERTIES OF HTC
SUPERCONDUCTORS AND APPLICATION THEM AS
PERMANENT MAGNETS
Sosnowski J.
Electrotechnical Institute, 04-703 Warsaw, Poaryskiego 28, Poland
ABSTRACT
Among various present or near future applications of HTc superconductors one of more
promising is utilizing these materials as permanent magnets. It is a very actual topic,
especially because quite recently it has been revealed the giant magnetic trapped flux in
these materials of the range of 17 T. This value is much higher than received for
conventional permanent magnets, even based on rare earths, for which magnetic remanence
does not exceed a few Teslas. In this paper the subject of the trapped flux will be
considered taking into account the granular structure of the ceramic materials. The model
describing the potential barrier height capturing the vortices, which determines the critical
current density, magnetic induction distribution and trapped flux will be presented for HTc
ceramic superconductors. The influence of ceramic material parameters on the trapped flux
magnitude will be considered at an aim of indicating meaning these parameters, which
should be useful for optimizing trapped flux value. Trapped flux has important meaning
from the point of view of application of HTc superconductors in the magnetic levitation
process, as superconducting bearings, which will be also briefly discussed in the paper.
Keywords: HTc superconductors, magnetic levitation, permanent magnets
1. INTRODUCTION
High temperature oxide superconductors have been discovered more than twenty years ago.
Now it is time therefore for applications of these materials in electric devices. Exceptional
electromagnetic phenomena appearing in these materials should be taken into account to
this aim. In the paper mechanism of the trapped flux in HTc superconducting materials is
considered, which giant value allows to treat these superconductors as promising permanent
magnets, very useful in magnetic levitation process, for construction of magnetic bearings,
future motors of new generation etc. The present work is partly stimulated by the recent
experimental data, showing that in single macro-grain of YBaCuO trapped magnetic flux
reaches giant value of 17 T at low temperature of 29 K (according to the results of Dowa
Mining, Japan). It is just actual world record of the remanent magnetic induction among all
materials presently known, including permanent magnets based on the rare earths. In the
paper is performed theoretical analysis of these unique, from technical point of view
magnetic properties of HTc ceramic superconductors, basing on modeling of the pinning
interaction. The critical current and influence of it on trapped flux magnitude is considered.
2. OUTLINE OF THE PINNING INTERACTION IN HTC SUPERCONDUCTORS
Unique magnetic properties of HTc superconductors are essentially related to the pinning
effects arising in these materials. Pinning phenomena describe the capturing of the
magnetic vortices in the superconducting materials and determine therefore critical current
density and trapped in vortices magnetic flux. Layered structure of the high temperature
superconducting materials influences specific shape of the pancake type vortex in HTc
301

superconductors, which is presented in Fig. 1. The core of that vortex retains the normal
state, which is characterized by higher energy than superconducting one. As follows from
theoretical analysis, especially based on the Ginzburg-Landau theory, the superconducting
state is characterized by the order parameter, which denotes that this state is energetically
more favourable than the normal one. It means further that the increase in the volume of the
normal phase in system enhances its energy and therefore the amount of normal phase
should be minimized [1]. This effect is considered just in the proposed pinning interaction
model. The shift of the captured on the nano-sized pinning centre vortex possessing the
normal core of the radius , equal to the coherence length, enhances normal state energy
of the superconductor. On the other hand the Lorentz forces, acting on the pinned vortices
during current flow, as well as elasticity forces tear off the vortices from the pinning
centres, which leads to their movement and dissipation effects. Potential barrier appears, to
be a function of the material parameters, such as pinning centres kind and dimensions,
elasticity shear modulus of vortex lattice, transport current density and as usually magnetic
field and temperature. Critical current density for the flux creep process is reached, if the
potential barrier height U against tearing the vortices off vanishes, while the current
satisfying the voltage criteria is achieved.

Figure 1: Scheme of the pancake type vortex in layered HTc superconductor

The model presented briefly above leads in its final form to the relation describing the
potential barrier for flux creep process U as the function of reduced current density i =
j/jC, where jC is defined as critical current density for magnetic vortices creep process, when
potential barrier disappears [2]:

U ( i ) =

o H c2 l 2
z + 2 1 i 2 1 i 2 2
2

(1)

Parameter describing the elasticity forces of vortex lattice is defined in Eq. 3, while z
appearing in Eq. 1 is determined according to the relation:

302

d
d
d
z = arcsin
+
1

i 1 i 2 arcsin i

2 2
2

(2)

Hc is the critical thermodynamic magnetic field, d pinning centre width, l pinning centre
thickness. In the paper the flat shape of the pinning centre has been considered.

T=1K, B=1T

Jc(10 A/cm )

2T

3
4T
6T

0
0

0,5

d/2 1,5

Figure 2: The influence of the dimensions of the nano-sized pinning centres of width d, normalized
to the coherence length , on the critical current density versus applied magnetic field

As it follows from relation 1 the potential barrier height and therefore the critical current
density are influenced significantly by the elasticity energy of the vortex lattice. Capturing
of the vortices by the nano-sized pinning centres causes the deflection of the vortex from
its equilibrium position in the regular vortex array, thus leading to an enhancement in the
elasticity energy of the magnetic structure of the vortex lattice. This effect is the function of
the shift of the single vortex from its equilibrium position in the lattice, described by
parameter x, denoting the distance from the origin of the vortex core to the pinning centre
edge. The elasticity energy increase is mathematically given by Eq. 3, assuming that the
increase of the vortex elasticity energy is proportional to the square of the length of the
vortex deflection from equilibrium position in lattice, with coefficient of proportionality
expressed by the value of the parameter .

2cs 2 ( x )
U el =
= ( x) 2
la
2

(3)

Parameter cs in equation 3 is the corresponding elasticity shear modulus, while la l denotes

the length on which the magnetic flux of vortices is distorted. We insert then expressions 13 into the constitutive relation describing generated electric field E in the flux creep process
[2], just in the function of the potential barrier height. It allows us already to predict the
form of the current - voltage characteristics and to determine then critical current density
applying the appropriate electric field criterion. In Fig. 2 an example is shown of the
computer calculations according to the elaborated model of the influence of the flat pinning
centres width on the critical current density. This result indicates the meaning of the
pinning centres parameters for analysis of the critical current density and then magnetic
properties of the superconducting materials.
303

3. THEORETICAL ANALYSIS OF POSSIBILITY APPLYING HTC

SUPERCONDUCTORS AS PERMANENT MAGNETS
The model presented in the previous section describes dependence of the critical current
density of HTc superconductors on pinning centres parameters. The received results
determine also the magnetic induction distribution in superconductors and describe
therefore magnetic properties of the superconductors, which have very significant technical
meaning, since ideal diamagnetism and trapped induction in superconductors is utilized
already in magnetic levitation process, in superconducting bearings, trains and magnetic
shields. This effect is utilized just in construction of the magnetic bearing built usually from
a superconducting cylinder levitating around permanent magnets, as presents schematically
Fig. 3, which shows six oppositely oriented permanent magnets in superconducting tube.

Figure 3: Schematic view of the cross-section of the magnetic bearing composed from 6 permanent
magnet disks and levitating above them superconducting cylinder, with marked calculated
by finite element method (FEM) magnetic field lines

The ideal diamagnetism of superconducting materials leads then to the expulsion of the
magnetic field lines, as indicate the graphical results of calculations performed here using
FEM numerical code. Fig. 4 presents results of the calculations using FEM method of the
levitation force acting on the superconducting cylinder from the Fig. 3, as the function of
the distance between the edge of the permanent normal magnets and the superconducting
element. Very strong dependence of the levitating force on the distance is observed here.

300
F(N)
200

100
b(cm)
0
0

Figure 4: Calculated levitation force versus distance b between superconducting cylinder and
permanent magnets for magnetic bearing of the geometry shown in Fig. 3.

304

Figure 5. Calculated using FEM method of the magnetic induction distribution for the model of
levitating magnetically train, built from the HTc superconductor placed above the
permanent magnets arranged in the n-s-n magnetic configuration.

Another example of utilizing ideal diamagnetism of HTc superconductors appearing in

weak magnetic field is shown in Fig. 5, which presents magnetic model of the track of
levitating train, built from three oppositely arranged permanent magnets and the
superconductor levitating above them. The expulsion of the calculated magnetic field lines
for the model of this device seen in this figure results from an ideal diamagnetism of the
superconducting element. It should be noticed in that point that although in present
calculations ideal diamagnetism of superconducting materials has been utilized, in
levitation process also remanent moment of superconductor can be applied in that aim what

Figure 6. Magnetic irreversible curve of sintered YBa2Cu3O7-x sample. The arrows indicate the
direction of magnetic field variation and finally magnitude of the trapped flux
305

brings even higher repulsion force. It follows directly from the giant remanent moment of
HTc superconductor observed experimentally, as it was mentioned previously. Essential
property of superconductor, which is the persistent current flow causes then that frozen
magnetic flux theoretically should not change, while neglect creep process and leads further
to the return of superconductor to its initial position in magnetic field, if shifted from it.

Figure 7: Magnetic induction profile in the HTc superconducting slab of the thickness 2x m for
magnetic induction cycle 0 B m 0. The influence on induction distribution of the
granular structure of the ceramic superconductors is presented in this picture.

The trapped magnetic induction magnitude discussed here is marked in Fig. 6 with arrow,
as the value of remanent magnetization of measured according to presented in [3] method,
loop of the magnetic hysteresis curve of sintered us YBa 2Cu3O7-x sample [3], in external
magnetic induction cycle 0-Bm-0. Considered in theoretical model magnetic induction
profile in this cycle of external magnetic induction, taking into account the granular
structure of HTc ceramics is shown in Fig. 7, for the slab geometry of the superconducting
sample exposed to the parallel, varying magnetic field. Basing on this figure we determine
now mathematically flux trapped value normalized to the unit cross-section of the
superconductor, it is average remanent magnetic moment, in function of amplitude of an
external induction applied to the surface of superconducting slab Bm= Be -Bc1 - .

(Ftr(T))1/2

R=50 nm
30 nm

10 nm

3 Bm(T) 4

Figure 8: Dependence of the square root of the average trapped magnetic induction in HTc
superconducting ceramic on the maximal magnetic induction in the magnetic induction
cycle 0 Bm 0, versus grains radius.
306

1,5

Sqrt(Ftr)
0.3

1
Bc1=0.1 T

0.5

0,5
Bm(T)

0
0

Figure 9: Dependence of the square root of the average trapped magnetic induction in HTc
superconducting ceramic on the maximal magnetic induction in the magnetic induction
cycle 0 Bm 0, versus first critical magnetic field of superconducting matrix.

Bm is defined here as the difference between external magnetic induction Be, first critical
field Bc1/0, (0vacuum permeability) and generally magnetic surface barrier , appearing
frequently on the surface of homogeneous superconductors. In present calculations
however the surface barrier effects have been neglected. For Bm<0 trapped flux disappears:

Ftr == 0

(4)

For the next range of the magnetic induction increase, when B p > Bm > 0 trapped flux
normalized to the sample cross-section is given as:

Ftr ==

Bm2
4B p

(5)

where Bp = 0jcxm - Bc1 is value of the first magnetic induction penetrating totally inside
the sample. For next range of the magnetic field amplitudes described by the condition:
2Bp > Bm > Bp trapped flux is described by the following relation:

Ftr == Bm

Bp
Bm2
B

+ nB g ( m 1)
4B p
2
Bp

(6)

For higher values of the magnetic induction, it is satisfying the condition Bm > 2Bp the
trapped flux saturates and is described then by the following formula:

Ftr ==

Bp
2

+ nBg

(7)

Parameter n in the above equations describes the fulfillment of the ceramic material with
superconducting grains, while Bg is the mean frozen magnet induction in individual grain:

Bg = Bc1g +

g
3

Bc1g denotes the first critical field in the grain of ceramic superconductor, while g is the
product of the superconducting grain radius Rg and jcg critical current density inside the
grains, defined according to relation:
307

(8)

g = 0 jcg Rg

(9)

Results of computer calculations of the influence of the ceramic material parameters on the
trapped flux magnitude are shown in Figs. 8 - 10 and indicate the influence of these
parameters on the trapped flux, what has meaning from the point of view of the
optimization trapped flux value. As it is seen trapped flux is the function of applied
magnetic induction and for higher magnitudes finally saturates. Figure 8 presents the
influence of the grains radius on the trapped flux value, while Figure 9 the dependence of
the first penetration magnetic induction B c1 of the superconducting matrix on the trapped
flux. In Figure 10 is given the dependence of the trapped flux on the grains concentration.
Performed analysis allows therefore to predict relevance of material parameters for
receiving optimal flux trapping, essential parameter from the point of view of the
application HTc superconductors as permanent magnets.

1,2
(Ftr(T))1/2

0,8
3

0,4

Bm(T)

0
0

Figure 10: Dependence of the square root of the average trapped magnetic induction in HTc
superconducting ceramic on the maximal magnetic induction in the magnetic induction
cycle 0 Bm 0, versus grains concentration in cross-section of superconductor:
(1) 3*10 5 cm-2 , (2) 105 cm-2 , (3) - 10 3 cm-2.

CONCLUSISONS
Technically important magnetic properties of the HTc superconductors have been described
using the proposed new model based on pinning interaction, taking into account the
granular structure of these unique ceramic materials. According to this has been considered
the influence of the parameters of superconducting grains and surrounding them matrix on
trapped magnetic flux. The importance of the interaction pancake type vortex - pinning
centre for determining critical current and then trapped magnetic flux in HTc
superconductors has been considered and relevance of this interaction emphasized.
REFERENCES
1.
2.
3.

Sosnowski, J., Superconductivity and applications, Book Publisher of

Electrotechnical Institute, Warsaw, Poland (2003) 1-200 in Polish.
Sosnowski J., Vortex pinning in HTc superconductors, Studies of High
Temperature Superconductors, Ed. A.Narlikar, Nova Science Publishers, Inc. New
York (2006) 389-122.
Sosnowski J., Gajda D., Analysis of the flux trapped in HTc superconductors,
Electrotechnical Institute Works, PL (2007) 231 126-134 - in Polish.
308

POSTERS:
NITROGEN TEMPERATURE
TECHNOLOGY

309

CR08-02
GAS FLOW THROUGH NARROW GAPS AT LOW PRESSURE IN
SUPER-INSULATION PACKAGES
Stipsitz J.1, Dobrozemsky R. 2, Hirschl C. 1, Laa C. 1
2

1
Austrian Aerospace GmbH, Stachegasse 16, A-1120 Vienna, Austria
Vienna University of Technology, Wiedner Hauptstrasse 8-10, A-1040 Vienna, Austria

ABSTRACT
Super-insulation is the most effective thermal insulation for cryogenic applications and is
employed in the vacuum space between the cold surfaces and the outer vacuum vessel.
Satisfactory insulation performance can only be maintained if gas conduction is suppressed.
Conductance measurements have been performed by means of an ultra-high vacuum
system. Based on these conductances and outgassing rate data from own measurements and
literature, the pressure drop between the center of the Super-insulation and the chamber
wall was calculated.
INTRODUCTION
Super-insulation (SI) is composed of alternate layers of reflector foils and spacer material.
The reflector (aluminum or aluminized polyester foil) minimizes radiative heat transfer.
The spacer prevents direct contact of, and minimizes solid conduction between adjacent
layers. Satisfactory insulation performance can only be maintained if gas conduction is
suppressed. Gas conduction becomes significant only for pressures higher than 1E-4 mbar
(1E-2 Pa) and therefore the interstitial volume must be pumped and kept below this level,
see Figure 1 and [1].

Heat Flow [W/m]

2,0

He
N2

1,0

0,0
1E-7

1E-6

1E-5

1E-4

1E-3

1E-2

Residual Pressure [mbar]

Figure 1: Typical SI performance with gas conduction [2]

The pumping of SI interstitial volume has been analyzed and measured by several authors
[3-7]. Due to the low conductance of the narrow gaps in molecular flow, the gas pressure
within the SI may be higher than the pressure measured at the chamber wall. Pumping of
the SI internal volume depends on the outgassing of the used SI materials and the
conductance of the narrow gaps between the foils for molecular gas flow. Knowing these
two parameters, the pressure inside the SI can be estimated.

311

1. MEASUREMENT
1.1 Test Setup
The conductance measurements have been performed by means of an ultra-high vacuum
(UHV) system equipped with Bayard-Alpert gauge (BAG) and quadrupole mass
spectrometer (QMS). Flow rates Q of pure gases were admitted to the upstream side by a
variable leak valve (VLV). Upstream pressures P up were adjusted by means of a capacity
diaphragm gauge (CDG) and the downstream pressures P dwn were measured by the BAG
with the gas composition monitored by the QMS. The experimental setup is illustrated in
Figure 2.

Figure 2: Setup for conductance measurement

A cylindrical test specimen was prepared by winding a SI strip of length B m = 2400 mm

and width Lm = 50 mm around a 10 mm diameter stainless steel rod. This cylindrical
specimen fits into the 32 mm inside diameter stainless steel test chamber in a way that the
distance between layers is comparable to the actual situation in an extended SI blanket with
no gap between the outermost layer and the wall of the test chamber (layer density: 3.3
layers/mm, distance between adjacent layers Hm = 0.3 mm), see Figure 3.

Figure 3: Test chamber with specimen

312

1.2 Measurements
The pressure reading instruments as well as the effective pumping speed S eff on the
downstream side were calibrated with respect to a secondary standard.
On pages 67-68 of [8] the throughput Q is defined as the product of the pumping speed S p
and the inlet pressure P, i.e.

Q = P S p = P (dV / dt ) ,

(1)

and the effective pumping speed S eff obtained in a chamber connected by a conductance C
to a pump having a pumping speed Sp is given by

1 / S eff =1 / S p +1 / C .

(2)

For equilibrium conditions the mass flow must be same for both the upstream and the
downstream sides of the specimen. The conductance C of the SI test specimen can now be
calculated from

Q=

Pup C Seff
= Pdwn Seff .
C + Seff

(3)

For all gases employed, equilibrium conditions were achieved within rather short periods of
time. This means that gas flow rates Q as well as contributions due to outgassing and
adsorption are constant during measurement periods. The upstream pressures were adjusted
in the 1E-3 and 1E-2 mbar (1E-1 and 1 Pa) range to maintain molecular flow conditions
with respect to the narrow spacing of the SI-layers, and the BAG-readings were evaluated.
With the test specimen at 21C (294 K) the following conductances of the SI package were
derived (Table 1).
C
[liter/s]
H2O

0.11

Ar

0.19

N2

0.23

H2

0.90

Table 1: Measured conductances for Bm=2400 mm and Lm=50 mm

For ideal gases in molecular flow, the conductance of an aperture is proportional to (1/M) 1/2,
with M the molecular weight of the gas see pages 80-81 of [8]. There is a good fit of the
measurement results for Ar, N 2, and H2 with this theoretical relation (3%). Water vapor
shows a lower value, because it is no ideal gas. It can be adsorbed to the MLI surfaces and
diffuses more slowly through the specimen [9].
2. ANALYSIS
Based on these conductances and outgassing rate data from own measurements and
literature (see Table 2), the pressure drop between the center of the SI (P up) and the chamber
wall (Pdwn) was calculated.

313

Outgassing rate
[mbar.liter/
(m.s)]

Main outgassing
products

Glass fiber spacer

1E-9

H2

Polyester spacer

2E-9

H2

Aluminium [10]

1E-10

H2

Aluminized polyester [11]

3E-7

H2O

Table 2: Measured outgassing rates for spacer materials and literature values for foils

The unit area for the outgassing rates of the spacer materials is defined as the one-sided
geometrical area. The unit area for the outgassing rates of the foil materials is the exposed
surface area. These considerations depend essentially on pumping cycles, including baking
and purging.
For the calculation it was assumed that the total outgassing of one foil plus one spacer layer
occurs in the center plane of a SI blanket of 1 m (see Figure 4), resulting into flow rates Q 0
from the center to either side.

Figure 4: Model for calculation of pressure drop in SI blanket of 1 m, having length 2L and width B

The pressure differences listed in Table 2 were calculated by means of

Pup Pdwn =

Q0
.
C SI

(4)

For a slot with H<<B, the conductance is proportional to BH/L see page 84 of [8].
Therefore the SI conductance CSI for a SI blanket of 1 m can be calculated from the
measured conductance C, which relates to the specimen geometry (B m, Lm), using the
formula

C SI =

B Lm C
.
Bm L

(5)

The calculated pressure drop caused by the outgassing of different SI materials is given in
Table 3.
314

Main
outgassing
products

CSI
[liter/s]

Pup Pdwn
[mbar]

Pup at
Pdwn=1E-6 mbar

Glass fiber spacer

H2

0.0375

1.7E-8

1.0E-6

Polyester spacer

H2

0.0375

2.3E-8

1.0E-6

Aluminium [10]

H2

0.0375

5.3E-9

1.0E-6

Aluminized polyester [11]

H 2O

0.0046

5.9E-5

6.0E-5

Table 3: Calculated typical pressure differences for SI blanket of 1 m

CONCLUSIONS
The suggested method for the measurement of conductances of super-insulation in
molecular flow yields consistent results.
Assuming an outside pressure Pdwn of 1E-6 mbar (1E-4 Pa) and the use of polyester foils
without perforation, the pressure within the SI will come close to the level of 1E-4 mbar
(1E-2 Pa), where gas conduction becomes important. In this case the influence of the spacer
outgassing can be neglected.
For aluminum foils the spacer will contribute most to the outgassing and there is no
significant pressure drop between the center of the SI and the chamber wall.
ACKNOWLEDGEMENTS
The work was carried out in a cooperation of the Vienna University of Technology and the
Thermal Systems department of Austrian Aerospace (AAE), the largest supplier of space
products and related ground support equipment in Austria, focusing on electronics,
mechanisms and thermal insulation. The company of 130 employees is owned by the Saab
Space Group. The business mission of AAE is to market, develop and manufacture satellite
equipment for the global space industry and cryogenic insulation for terrestrial applications
that strengthen and support the competitiveness of its customers. AAE produces high
quality Multi Layer Insulation (MLI) since 1991, and is the leading supplier of MLI for
spacecraft of the European Space Agency and a number of other European programs. AAE
designs, manufactures and integrates thermal hardware products in-house or at the
customer's site. As a product diversification, AAE has adopted space MLI for cryogenic
applications, covered by the Coolcat line of insulation products. Coolcat is used for the
insulation of superconducting magnets (MRI, NMR, and others); cryostats; transfer lines;
and automotive liquid hydrogen tanks. AAE offers engineering consultancy and thermal
analysis for the customer systems. In its state-of-the-art manufacturing facility for
insulation, AAE is equipped with automated cutting machines and all processes for
production on an industrial scale.
Prof. Dr. Rudolf Dobrozemsky is guest scientist at Institut fuer Allgemeine Physik (IAP) of
Vienna University of Technology. Besides its program of introductory physics courses, IAP
is well established in the fields of plasma, surface, nanotechnology, and ultrasonics
research with a broad experience in the respective experimental and analytical techniques.
The project was funded by the Austrian Federal Ministry of Transport, Innovation and
Technology (Austrian Research Promotion Agency), and Austrian Aerospace.

315

REFERENCES
1. Augustynowicz, S.D., and Fesmire, J.E., Cryogenic Insulation System for Soft Vacuum,
Advances in Cryogenic Engineering, (2000) 45 1691-1698
2. Laa, C., Hirschl, C., and Stipsitz, J., Heat Flow Measurement and Analysis of Thermal
Vacuum Insulation, presented at CEC/ICMC 2007
3. Keller, C.W., Cunnington, G.R., and Glassford, A.P., Thermal Performance of
Multilayer Insulation, Final Report of NASA Contract NAS 3-14377 (1974) Section 5
4. Kaganer, M.G., Thermal Insulation in Cryogenic Engineering, Israel Program for
Scientific Translations, Jerusalem (1969) 170-174
5. Bapat, S.L., Narayankhedkar, K.G., and Lukose, T.P., Performance prediction of
multilayer insulation, Cryogenics (1990), 30 700-710
6. Bapat, S.L., Narayankhedkar, K.G., and Lukose, T.P., Experimental investigation of
multilayer insulation, Cryogenics (1990), 30 711-719
7. Reiss, H., A coupled numerical analysis of shield temperatures, heat losses and residual
gas pressures in an evacuated super- insulation using thermal and fluid networks. Part 1:
Stationary conditions, Cryogenics (2004) 44 259-271
8. Roth, A., Vacuum Technology, Elsevier Science B.V., Amsterdam, The Netherlands, 3rd
Ed. (1990)
9. Dobrozemsky, R., Menhart, S., and Buchtela, K., Residence times of water molecules on
stainless steel and aluminum surfaces in vacuum and atmosphere, J. Vac. Sci. Technol.
A 25(3) (2007), 551-556
10. Ishimaru, H., Ultimate pressure of the order of 10-13 Torr in an Al alloy chamber, J. Vac.
Sci. Technol. A7 (1989) 2439
11. Glassford, A.P.M. et al, Effect of temperature and preconditioning on the outgassing
rate of double aluminized mylar and Dacron net, J. Vac. Sci. Technol. A2 (3), (1984)

316

CR08-46
NEW DEVELOPMENTS OF NON-METALLIC CRYOSTATS FOR
HIGH SENSITIVE ELECTRONIC DEVICES AND OTHER
APPLICATIONS
Klier J., Sprl G., Schumann B., Binneberg A., Herzog R.
Institut fr Luft- und Kltetechnik gemeinntzige GmbH,
Bertolt-Brecht-Allee 20,01309 Dresden, Germany
ABSTRACT
For a number of high-precision devices working at low temperatures magnetic stray fields
can drastically reduce their sensitivity. This problem can be avoided by the use of nonmetallic cryostats. For most applications of high-Tc-SQUID devices in material
investigations, geological exploration and medicine low noise cryogenic cooling systems
are required. Under certain conditions, however, there is the demand of special geometries
between sensor and testing area, and furthermore position independence and mobility.
Examples are measurements of heart and brain biomagnetic signals or non-destructive
evaluation on parts of air planes and reinforced concrete constructions. A series of special
cryostats has been developed to meet these demands. The construction of the inner vessel
allows reliable cooling of a sensor in vacuum referring to liquid level and stability of
temperature and when the cryostat is turned around its transverse axis. For very small
cryostats (< 150 mm) the sensor can be placed in vacuum on a sapphire rod mounted on a
heat transfer area cooled by cryogenic liquids.
INTRODUCTION
Extensive work in the field of superconducting electronics has lead to the development of
superconducting interference devices, known as SQUID magnetometers. These highly
sensitive devices are able to measure very small magnetic fields down to 10 14 Tesla. Their
sensitivity, however, depends on the level of thermal noise, which can be minimized
through the use of non-metallic materials for the cryostats providing the necessary low
temperatures. SQUIDs made from conventional superconductor usually operate at liquid
helium temperatures. Due to the need of liquid helium as cooling medium and the high
costs of low-Tc SQUIDs there are some limits in their application. This situation changed
after the discovery of the high-Tc superconductors in 1986 and the subsequent development
of high-Tc-SQUIDs. Especially high-Tc superconductors like yttrium barium copper oxide
(YBCO) with a transition temperature above 90 K allow the use of liquid nitrogen as
cooling agent for high-Tc-SQUIDs.
For various applications of high-Tc-SQUIDs we have developed a variety of non-metallic
cryostats. In this paper we present some of our most exceptional designs and latest
developments.
GENERAL PROPERTIES OF NON-METALLIC CRYOSTATS
Usually standard laboratory cryostats are double-walled vessels, made from glass or metal,
which are evacuated. In the case of glass the walls of the vessels are silver-plated in order
to reduce thermal radiation into the cryostat. In the case of metal the spacing between these
walls usually contain multilayer insulation. Most common the insulation material consists

317

of thin polyester foils (superinsulation) coated with aluminium on one side. Such cryostats
can be used when their magnetic influences do not affect the experimental results.
Changes of magnetic fields, however, induce eddy currents in metals. There is further
thermo-magnetic noise in metals which can compromise the sensitivity, interaction with the
measuring signal or reduction in the signal to noise ratio. Therefore cryostats free of any
metal near the sensor are essential for applications with high magnetic field sensitivity. This
requirement is fulfilled when glass fibre reinforced epoxy materials are used for the
cryostat fabrication.
Foils of superinsulation can act as shields for higher frequency signals or induce additional
noise. To avoid these problems we use polyester foils with partially demetallized
aluminium face. The use of these modified foils does not affect the vacuum condition and
evaporation rate of the cryostats. For some methods of geological exploration, however, the
insulation must be replaced by completely non-metallic polyester foils.
DESIGNS AND APPLICATIONS OF NON-METALLIC CRYOSTATS
For cooling highly sensitive measuring devices (e.g. to detect magnetic fields down to 10
14
T) we have developed and manufactured special liquid nitrogen cryostats. The high
sensitivity of the devices require effective means against outer electromagnetic
disturbances. To meet this requirement the cryostat must be free of any metallic parts in the
vicinity of the sensor. Composite materials such as glass fibre reinforced epoxy resin are
used instead of metals. The design of thermal insulation and integrated shielding were
adopted to the special requirements, such as transparency for high-frequency radiation, very
low intrinsic noise, unusual geometrical conditions, long operating time, suitable for field
tests, and low fabrication and maintenance costs.
Typical fields of applications are:
medicine
e.g. measurements of currents of heart and brain (i.e., biomagnetic
measurements with SQUIDs)
material tests e.g. detection of micro defects in materials by non-destructive
evaluation
geology
e.g. surveying of the earth in the vicinity of the surface, specification
of building ground and ground water
science
e.g. biophysical and standard investigation at and with SQUIDs,
scanning tunnelling microscopy, etc.
1. Variable distance and position independent cryostats
The demand of using SQUID systems for the non-destructive evaluation on parts of air
planes and reinforced concrete constructions initiated our reserach activities to develop
variable distance and/or position independent cryostats. Having a cryostat in which the gap
between SQUID and bottom of the tail of the vacuum vessel can be adjusted, opens the way
for high resolution magnetometry of biological specimens. So it is aimed to optimize the
local resolution of the measuring method by changing the distance between device and
detector element outside the cryostat.
The SQUIDs are located in vacuum on a sapphire rod (outside of the liquid nitrogen vessel)
mounted on a copper face cooled by the liquid nitrogen, see figure 1 and 3 (left). The
smallest possible distance between sensor and object of measurement can be only 4 mm.
For universal use of the cryostats special inserts are developed for filling and refilling,
made of different non-metallic materials. The SQUIDs can be changed without warming-up
the cryostat. Continuous measurements are possible by use of automatic filling stations.
318

Figure 1: Variable distance cryostat in z-direction.

The distance between the device and the outer wall of the vessel can be changed by spring
bellows mounted on top of the cryostat. In this way the whole liquid nitrogen vessel is
moved relative to the outer vacuum vessel, see figure 1. The bellows are pressed or
streched by several timing bolts or precision bolts. The distance between SQUID and
bottom of the outer vessel can be varied between 1 mm and several 10 mm. The isolation
vacuum usually holds more than 6 months and the evaporation rate is less than 7% per day
of LN2.
For some applications it is necessary to vary the distance between sensor and outer wall of
the cryostat not only in the z-direction but also in the horizontal plane, i.e., xy-direction.
The realisation of such a demand is shown in figure 2.

Figure 2: (left) Variable distance cryostat: The position between sensor and cryostat itself can be
varied both in horizontal and vertical direction (xyz-plane). (right) Sketch of the adjustment facility in
horizontal direction (xy- direction).

319

The position independent cryostats are designed in such way, that they can be turned
around their axis by 360, see figure 3 (right). Hereby the construction of the inner vessel
allows reliable cooling of the sensor with respect to liquid level and stability of
temperature. Changes of temperature are less than 0.2 K. By turning the cryostat there are
no significant changes of the evaporation rate and maximum working time.

Figure 3: (left) Drawing of a position independent cryostat, i.e., the cryostat can be turned by 360.
(right) Picture of a position independent cryostat which can be turned by 360. In addition the
distance and plane between SQUID and vacuum vessel is variable and can be inclined.

2. Very small mobile cryostats

SQUIDs can be used for magnetic non-destructive material tests in order to find very small
failures (less than 1 mm) in thick material layers or hidden metals (e.g. steel in concrete).
For this application the cryostat should be as small as possible and rotatable. A series of
cryostat has been developed to meet these demands. The smallest of this type is shown in
figure 4. The diameter is about 60 mm and the height about 140 mm. The weight of the
cryostat is less than 500 g. The operational time is three hours in thermal equilibrium, when
filled with about 36 ml of liquid nitrogen.
The construction of the inner vessel allows a reliable cooling of a sensor in vacuum
according level and stability of temperature, when the cryostat is turned around its axis. The
working temperature depends on the position and varies from 77.5 K to 78 K. The stability
of the temperature in one position is about 0.1 K.

320

Figure 4: Drawing (left) and picture (right) of a very small mobile cryostat.

3. Lift cryostat for superconductor-based magnetic levitation lift

The use of magnetic bearings has become of increasing importance. Although
electromagnets permit the contactless transmission of power and, thus, rotation or motion,
the generation of the required guiding forces in the magnetic bearings require costly
measuring and control techniques. This problem could be overcome by the use of
superconducting magnetic bearings.
For the realisation of a superconductor-based magnetic levitation lift the core of the
superconducting magnetic bearings comprises a magnet rail made of conventional
permanent magnets and superconducting blocks cooled to liquid nitrogen temperature.
Arranged at a defined distance from the magnet rail, the superconductors freeze the
magnetic field of the permanent magnets. This puts the superconducting magnets in a
position to maintain by themselves a certain distance from a magnet rail. Linear drives
ensure synchronous, contactless lifting and lowering of the superconductor-based magnetic
levitation lift.
The lift cryostat, especially developed for this project, is made from glass fibre reinforced
epoxy material and cooled by liquid nitrogen. A predetermined space is kept between the
superconductors and the surface of the magnet rails while the superconductors are cooled to
below the transition temperature of about 92 K, see figure 5 (left). As soon as this
temperature is undershot, the cryostat is non-positively and contactless anchored in the
magnetic field of the magnet rails (at a distance of 1 mm). Since the configuration of the
magnetic field does not change along the rail, the cryostat can move only along the rail at a
given distance.
The cryostat holds 12 litre of liquid nitrogen. This quantity is sufficient for operating the
magnetic levitation lift, see figure 5 (right), for approximately 10 hours without
replenishing the nitrogen. Thereafter, some 5 litres of evaporated nitrogen must be
replenished. Possible applications of such a system, superconductor-based magnetic
levitation lift, are in clean rooms and pharmaceutics.

321

Figure 5: (left) Picture of lift cryostat and superconducting bearings. (right) First working model of
a superconductor-based magnetic levitation lift.

4. Liquid Neon cryostat cryo-collector

For climate and environmental research the analyses of trace gas within the atmosphere, up
to heights of 40 km, is essential. The focus of such measurements are the halogenated trace
gases, like Chlorofluorocarbon, which are responsible for the hole in the ozone layer. In
these heights of the atmosphere the air pressure is only a few hPa. However, for gas
chromatography certain amounts of gas samples are necessary. To overcome this problem
one uses so-called cryo-collectors, which are transported to the appropriate height via large
stratospheric balloons. The working principle of such cryo-collectors is to freeze the air in
special sample containers. Due to the experimental conditions of such balloon flights, the
time for cryogenic gas collection is quite short, sometimes only a few minutes. Therefore
one needs a cryogenic fluid with low boiling point and high heat capacity. This makes
liquid neon quite suitable as refrigerant. Figure 6 shows such a LNe cryo-colector.

Figure 6: Liquid Neon cryostat: A cryo-collector for air samples in the atmosphere.

CONCLUSIONS
We designed and fabricated a number of non-metallic cryostats, mainly using liquid
nitrogen as cooling agent. The field of applications were high-precision devices working at
low temperatures. Usually magnetic stray fields can drastically reduce the sensitivity of
such devices. This can be avoided by the use of non-metallic materials, in our case glass
fibre reinforced epoxy materials, for the fabrication of the cryostats. For some special
applications of high-Tc-SQUIDs there is the demand of special geometries between sensor
and testing area, and furthermore position independence and mobility. Therefore a series of
special cryostats were developed.

322

CR08-03
CRYOGENIC DISTILLATION COLUMN BEHAVIOR
AT THE VARIATION OF AN EXTERNAL FACTOR
Pearsica C.,Stefan L.,Preda A.,Vasut F.
Institute of Isotopic and Cryogenic Technologies - ICIT Rm.Valcea,Romania
ABSTRACT
Behavior of a cryogenic distillation column for a case of tritium-deuterium separation under
periodic instability of the distillation process was studied by mathematical modelling.
Influence of an external factor, analyzing a non steady state with the fluctuation of the
hydrogen level in the column condenser was subject of investigation. The mathematical
model is based on balance equations, column operated at total reflux and sinusoidal
variation of the hydrogen level. There is studied the non steady state evolution in the
distillation equipment. For several situations, results were presented in specific diagrams
and plots.
INTRODUCTION
In order to realize correct operation of the distillation column from a liquid hydrogen
isotopic plant it is necessary to maintain all the parameters constant, which corresponds to
steady state. Practically, this cannot be realized and the operating column would suffer
some perturbations, the regime would become non-steady. The effect of the non steady
state is the decrease of the separation performance of the column [1].
We followed the way the variation of the liquid level in the condenser influences the
operation of the cryogenic distillation column. Such a perturbation determines an
immediate modification of the liquid flow in time and in length, but also modification in
time of the condenser holdup and on column packing.
The system (the distillation column) was defined for the mathematical analysis of the
distillation process of a tritium deuterium mixture.
The modell is analyzing the following situation: after attained normal operation steady
state, there is applied to the system a sinusoidal perturbation by the variation of the liquid
level from the column condenser. This moment represents the beginning of the study. The
concentrations existing in the column at this moment of time there are be called in the paper
as initial concentration.
A sinusoidal function that describe the liquid variation from the column condenser is
defined by two variable parameters: the size (amplitude) and the frequency (period). This
perturbation has a finite action, the intention being to analysis the influence of this kind of
operation on the distillation process.
There are calculated the gas concentrations for every time step, on each plate. The results
can give information about the process quality and the time for the entrance in the steady
state (defined like the time between the perturbation initiations, till attaining the constant
values corresponding to steady state).
The graphics for the entrance in steady state represent the moment from when the
concentration has constant values.

323

1. TRANSFER UNIT HEIGHT

A previous study analyzed the non-steady state for a case of a cryogenic distillation column
condenser cooling circuit. The non-steady state determines the variation of the liquid level
in the column condenser. This one induces the variation of the holdup at the top of the
column, which determines the variation of the liquid down-flow.
The quality of the transfer element can be expressed using the Transfer Unit Height, IUT,

IUT =

H col
where Hcol is the column height and N is the number of plates corresponding
N

to given separation [2].

The non-steady state involves the time variation of the Transfer Unit Height, so this
parameter is calculated at every time interval and represents an indicator of the distillation
process quality.
In order to show the non stationary character of this parameter, it was named Apparent
Transfer Unit Height, IUTcalc =
Unit Height: IUTref =

H col
. This factor is related to the Reference Transfer
NF

H col
where NT is the number of theoretical plates calculated for
NT

the steady state column operation. Input data for the hydrogen isotopic separation plant are:
NT = 30, Hcol = 2m which means IUTref = 0.067m, for a tritium-deuterium mixture.
2. MATHEMATIC MODELL FOR STUDYING THE NON-STEADY STATE OF A
SEPARATION COLUMN WORKING AT TOTAL REFLUX WITH
PERTURBATION IN CONDENSER
Due to the specific characteristic of the isotopic
distillation column, big number of theoretical plates, so
a big amount of package, a hypothesis was done, that
the gas flow along the column is constant and the flow
variation is all taken by the liquid flow variation.
We make the simplifying theory that the pressure
variation from the column does not influence the
separation factor [3].
In a situation like this the distillation column is the one
represented in Figure 1.
The holdup from the distillation column filling is
uniformly distributed all the way. The equilibrium
equation for the column:

H c
H n
+N
=0
t
t

(1)

where Hc and Hn are the quantities of liquid in the

condenser and on the respective plate. In fact, the
plate means a part of the column, corresponding to a

324

single IUTref. Total vaporization is considered in the boiling vessel, while the column is
operated at total reflux and total boil-off, x1=y0.
Balance of a stage n:

H n
= Ln +1 Ln
t

(2)

( H n y n )
= G ( yn 1 yn ) + Ln +1 xn +1 Ln xn
t
H c
For condenser:
= G Lc
t
( H c yc )
= Gy N Lc xc
t

(3)
(4)
(5)

Sinusoidal variation of the liquid hold-up is considered in the condenser, according to the
equation, with amplitude A and the period w :
H c = H c 0 + A sin t
(6)
A
cos t
After the calculation it is obtained: Ln = Ln +1 +
(7)
N
For the last plate it can be written:
N 1
LN = G
A cos t
(8)
N

Ln = G

n 1
A cos t
N

For the stage n

(9)

Making the next calculation the concentrations throughout the column can be obtained.
Applying the method of finite time-differences Dt, the gas phase concentration throughout
the column, on each respective stage:

L t
L t
A t
G t
yn, m +1 = 1 +
cos t y n +
y n 1 y n + n +1 x n +1 n xn
N Hn
Hn
Hn
Hn

(10)
The liquid phase concentration results from the relation of separation factor:

xn, m +1 =

yn, m +1
1 + yn, m +1 ( 1)

(11)
In a similar way, for the condenser:

yc ,m+1 =

G t
At cos t
(G A cos t )t
y N + (1
) yc
xc
Hc
Hc
Hc

(12)

xc, m +1 =

yc, m+1
1 + yc, M +1( 1)

(13)

The equation system that describes the process, which takes place in the column at the
liquid level variation in the condenser, is formed of the equations (9)-(13), which together
with the functions that describe the variation of the holdup in the condenser and on each
plate will be solved throughout the column. A program, solving the equations system above
was developed according to a logical diagram [4].
325

RESULTS AND DISCUSSIONS

For the analysis of the perturbation influence owed to the variation of the liquid level in
distillation column condenser operated at total reflux, we considered the following cases
[5]:
- initial concentrations of 25% and 50% tritium in the tritium-deuterium mixture
- variable sinusoidal amplitude function between 0 and 40% change of Hco
- sinusoidal function period between 0 and 30 minutes.
Some results are presented in the diagrams. Such as, in Figure 2 there are the concentration
profiles along the column at the entrance in steady state for the case, when the initial
concentration is 50% T/T+D.
yc
Hcol [m]

0.26

2.0

0.24

1.8

Initial concentration 50%

Amplitude 10%
T=10min
T=15min
T=30min

1.6

1.4

1.2

0.22
0.20
0.18

1.0

0.16

0.8

0.14
0.6

A=40%

0.12

A=15%
A=10%

0.4

0.10

0.2

0.08

0.0
0.0

0.1

0.2

0.3

0.4

0.5

Figure 2

0.6

0.7

0.8

0.9

1.0

20

40

60

80

100 120 140 160 180 200 220 240

time[min]

Figure 3

There was dealt the situation when amplitude was kept constant at a value of 10% and takes
place a variation of the perturbation frequency: 10, 15 and 30 minutes. The variation in
time of the gas concentration in the top of the column, in the case of 50% initial
concentration and constant period is shown in Figure 3.
A similar situation is dealt in the case when initial concentration is 25%. There are
represented the gas concentrations at the entrance in steady state in the two cases: when is
kept the amplitude constant (Figure 4) or the period constant (Figure 5).
For the latter case (Figure 5), there is observed that together with the growth of the
amplitude value, the profile changes in the way of decreasing the separation performance,
as a consequence, as the amplitude value grows, the more detrimental the situation is.
Noticing with attention the profiles drawn in figure 2, inversion of the concentration curve
profile was observed. Considering that this result gets out of normality it is recommended
to avoid to achieve or to overcome the value of 30% amplitude function and also it is
recommended a period more than 15 minutes.

326

For the case when the amplitude is maintained constant but perturbation frequency varies
(fig. 2 and 4 respectively), from the calculations we get to the conclusion that for
acceptable values of the perturbation frequency, I mean a frequency of 2-4 times an hour,
the concentration profile at the entrance in steady state is almost the same, while once
overcome the frequency of 4 times an hour takes to a deep lack of balance, case when the
distillation column manipulation becomes difficult and the time to achieve the steady state
is longer.
The liquid level variation in the distillation column condenser causes decrease of the
separation performance of the column, this one being lower as the perturbation is stronger.
The influence of the perturbation on the operation column can be followed through the
IUTcalc parameter. There were considered for plotting the IUT values at the entrance in
steady state. This parameter is represented as a function of amplitude at different values of
IUT [m]

IUT [m]
0.26

0.30

0.24

T=10 min

T=10minute
T=15minute
T=30minute

0.22

T=15min

T=10min
0.25

0.20

T=15min

0.20

0.18
0.16

0.15
T=30min

0.14
0.12

T=30min
0.10

0.10
0.05

0.08
0.06
0

10

15

20

25

30

0.00
0
40
45
Amplitude [%]

35

Figure 6

10

15

20

25

30

35
40
Amplitude [%]

Figure 7

frequency in the analyzed cases. In Figure 6 (the case of initial concentration 25%T/T+D)
and 7 (the case of initial concentration 50%T/T+D), IUTcalc increases with the amplitude
value and perturbation frequency. From calculations results that IUTcalc can increase 3.5
times towards IUTref value which means an increase of the column height with the same
value if it is wanted the same separation.
The model allows the representation of this profile at any time. The file in format *.dat
containing the necessary data for this kind of representations can be used if we want to
Hcol[m]
2.0

Hcol [m]
2.0

1.8

1.8

1.6

1.6
T=30min
T=15min
T=10min

1.4
1.2

1.2

1.0

1.0

0.8

0.8

0.6

0.6

0.4

0.4

0.2

0.2

0.0
0.0

0.1

0.2

0.3

0.4
0.5
Figure 4

0.6

0.7

A=40%
A=15%
A=10%

1.4

0.8

0.9

0.0

1.0

0.0

327

0.1

0.2

0.3

0.4

0.5

Figure 5

0.6

0.7

0.8

0.9

1.0
y

analyze the evolution of the concentrations along the column in time. The profiles
presented in the previous figures represent the moment of entrance in steady state.
CONCLUSIONS
Graphic representations can give information about the operation of a distillation column
under conditions of non-steady state, resulted as a consequence of perturbation in
sinusoidal shape of the liquid level in condenser. This way, concentrations profiles along
the column at its evolution in time can be traced, but most of all, at the entrance in steady
state, as is defined in the introduction, different values as S separation, Fenske number N F
or the height of the transfer unit IUT can be determined. Every case, by itself, may offer
information upon the behavior of the distillation column operated at total reflux when the
perturbation from condenser is described by sinusoidal function.
The models of calculation realized can be used to study more aspects of the evolution of the
non-steady state in the distillation plant.
REFERENCES
1. PECULEA M., Instalatii criogenice, Ed. Conphys (1997)
2. CONSTANTINESCU D.M., DIMULESCU A., PECULEA M., URSU I., Influenta
instabilittaii vaporizatorului asupra functionarii coloanelor de distilare izotopica, St.
Cerc.Fiz., Tom 31, nr.2, p 119-127, Bucuresti (1979)
3. STRATULA C., MARINOIU V., SORESCU GH., Metode si programe de calcul al
proceselor de distilare, fractionare si absorbtie, Ed. Tehnica, Bucuresti (1976)
4. TOMA M., ODAGESCU I., Metode numerice si subrutine, Ed. Tehnica, Bucuresti
(1980)
5. CLAUDIA PEARSICA, Contributions on the Non-steady in a Liquid Hydrogen Isotopic
Distillation Plant, Ph.D. Thesis (2007)

328

CR08-44
ANALYSIS OF PERIODIC ADSORPTION PROCESSES,
USED IN NEON AND HELIUM PRODUCTION
Bondarenko V. L.1, Simonenko Yu. M.2
1

Moscow Bauman State Technical University, 5, 2-nd Baumanskaya Str.,

107005, Moscow, Russia
2
Iceblick, Ltd., 29, Pastera Str., 65026, Odessa, Ukraine
ABSTRACT

The analysis of inert gases purification in a single adsorber has been made. The criteria for
the comparison of adsorption devices with different canal geometry have been offered. The
dependence of the sorbent dynamic capacity on the operational and constructive factors has
been studied. The complex of processes characteristic of the working phase and the
regeneration period has been considered. The obtained information permitted to reduce the
time of unproductive fragments of the cycle and increase the efficiency of the periodical
purification in adsorbers.
INTRODUCTION
The sorption technologies are widely used in the technologies of rare gases extraction.
There is a number of special requirements to the adsorbers used in this branch. The desire
to keep the low content of residuals often increases the length of the working cycle and, in
some cases, is accompanied with the increasing of the number of adsorbers necessary for
ensuring the continuous purification.
The said specific phenomena are different in the adsorbers of different shape. It can be
assumed that for each set of operational parameters there is a definite correlation of the
adsorber canal length and its diameter, at which each kilogram of the sorbent will perform
its functions most effectively. To find this optimum a sequence of process characteristics of
a single adsorber work should be considered.
THE ANALYSIS OF MIXTURE SEPARATION PROCESSES IN THE ADSORBER
Figure 1 illustrates the temperature change in the adsorber during one cycle. The sorbent
R e g e n e r a t io n
,
2= 3 0 0

H e a t in g
R = (21 )

Vacuum .

V = (32 )

W o r k in g W = ( 1 0 )

C o o lin g

C = ( '0 3 )

1= 8 0

= ( '0 0 ) = W + R + V + C
Figure 1: The temperature change during one cycle of the adsorber work

329

,
'0

regeneration on the stage H is achieved by the heat feeding through the walls of the device
or by means of supplying the heated circulate flow through the sorbent layer. The shown
operations sequence (or most of its stages) is characteristic for the neon and helium
separation devices, the neon-helium mixture purification and a number of other
technologies of production of rare gases and their isotopes.
An important utilitarian function of the adsorber is the admixture holding (1) and receiving
the pure product on the outlet (2) during the time W. First of all, this property is
characterized by the value of the sorbent adsorbing capacity adm [norm.m3/kg] with respect
to the admixture. The value adm is equal to the volume of admixture-substance, captured by
the sorbent mass unit. During the gas mixtures separation, their components mutually
influence the individual values of adsorbing capacities.
For example, according to the Langmuir theory, the following correlation is acceptable for
the estimation of the adsorption capacity in the binary mixture
m
aadm = aadm

badm P yadm
,
1 + badm P yadm + bcl P ycl

(1)

where: aadm is the sorbent adsorbing capacity against the pure component of admixture
under the conditions of total saturation;

yadm and ycl are the inclusion volume fractions of the components of the mixture;
is the working pressure;

badm and bcl are constants, depending on the properties of the sorbent and the mixture
components.
Reference sources give the sorbents characteristics for the cases of the sorbent static
saturation. These absorption levels are usually rather higher than the characteristics
obtained under real conditions. The measure of discrepancy between the adsorbing
capacities of the dynamic adm and static 1 depends of the canal shape, flow rate and a
number of other operating and design factors. For the interpretation of the sorbent
saturation we introduce the value =Aadm/aadm called the sorbent usage degree and the
degree of the saturation of the working layer. Notwithstanding the indisputable reliability of
adm, it characterizes only one (even though very important) cycle fragment. The dynamic
adsorption capacity unambiguously determines the duration of the working phase ( W), but
cannot totally reflect the whole complex of the adsorption separation processes. Indeed, the
task of getting the maximum of the saturation degree admadm (or 1) is easy to realize.
It is achieved, for example, by the multiple increase of the time of the working phase W.
Though this regime is not attractive from the customers point of view as it is realized at
negligibly small product recoveries, we believe that the inconsistency of as the
optimization object is compensated by the introduction of the time factor.

adm
aadm
=

W + R + V +

330

(2)

where: () is the duration of the whole period which includes the working time ( W), the
duration of heating (R) and pumping (V) during the regeneration and also the cooling time
before putting into operation (C).
The value is equal to the volume of the admixture adsorbed from the flow (yadm), falling
on the sorbent mass unit in a time unit. In terms of physics, in a freer interpretation, this
coefficient can be interpreted as the mean performance of the layer work. The reduction
of the criterion to the mass unit (via specific adsorbing capacity adm, m3/kg) is not
accidental. It is the quantity of the sorbent under the conditions of the periodical operation
that in many respects determines the separation operating expenses. But one and the same
mass of sorbent can be placed in the canals of different shape. The influence of the way of
packing the sorbent in the device is evident if we compare the adsorbers with the unequal
correlation of the diameter and length of the canal. In long devices (with small radial layer
extent) the ideal conditions for the heat transmission through the wall will be created. This
will result in the reduction of the time of the heating R and cooling C. At the same time, in
such adsorbers, the processes, determined by the speed parameters will last longer ( W,
V). In short canals the reverse situation will be observed: it is the processes of heat transfer
through the wall of the adsorber that will become overextended ( R C). Besides, for the
second case, a small degree of the sorbent usage will be typical ( 0).
To calculate the cycle fragments duration a block of calculation models has been designed,
which reflect the sequence of the single adsorber work. A complex of experimental stages
has been created, which allowed obtaining the data, characterizing individual adsorption
processes, in particular, the values of the static sorption capacity of silica gels and activated
carbon during the pure nitrogen, neon and helium adsorption. The sorption heat of the
mentioned gases has been estimated. The influence of the operating and design parameters
on the sorbent dynamic capacity has been studied (these data are interpreted by the
factor). The heat phenomena in the sorbent layer during the regeneration and cooling have
been investigated. The data, characterizing the pumping process of the adsorber, having
additional admixtures, have been accumulated.
The obtained information allowed making resulting estimations of the characteristics of the
devices with different size correlations. An adsorber for purifying the neon-helium mixture
from nitrogen has been taken as an example object of optimization. Operation parameters,
taken as the source data are characteristic of the typical technology of light inert gases (Ne
and He) extraction. Two ways of heat feeding during the regeneration have been studied:
through the wall of the device and by means of the heating gas supply (Figure 1). A series
of devices of the same volume (v = 0,1 m3) have been analyzed. The adsorbers had different
diameters and, naturally, different canal lengths L. For the indicated diameters the length
ranged from L = 51 m t L = 1,4 m correspondingly. The influence of the adsorber
geometry on the duration of certain cycle stages is shown in Figure 2.
The dependence of the optimization criterion for different device geometry is shown on
Figure 3. The value allows forecasting the utilitarian factors of the adsorbers with the
given capacity. The generalization of the obtained information showed that in the studied
range and at the stipulated device volume there exists a definite correlation of L/D
dimensions at which the coefficient possesses the maximum value. The value of the said
optimum is influenced, besides the operational and design factors, by the way of heat
transfer in the process of regeneration. For the considered example, in the case of heat
feeding through the wall of the device, the L/D optimal values are in the range from 50 to
331

100. When the sorbent is heated by means of the flow circuit, these values decrease to 20
50.
The information, shown on the graphs 2 and 3 is true for the following conditions: sorbent
mass m = 45 kg; imaginary flow speed w = 0,04 m/s; static adsorption capacity of the
admixture (N2) adm= 0,377 norm.m3/kg; sorbent packed density (coal SKT-4)
= 430 kg/m3; admixture content (nitrogen in neon-helium mixture) yadm= 0,1; working
pressure = 1,0 MPa; working temperature = 84 .

Figure 2: The duration of certain phases of the adsorber working period

a - during the regeneration by means of heat transfer through the wall of the device;
b - during the regeneration by the heat flow

Figure 3: The influence of the canal diameter and the regeneration method
on the optimization factor (upper curve not taking into consideration the adsorber heating
time)

332

CONCLUSION
The realized research proves the correctness and consistency of the optimization criterion
. It allows revealing the most successful packing shape of the given sorbent mass which
will result in each kilogram of the sorbent performing its functions most efficiently.
REFERENCES
1. Arkharov A. M., Bondarenko V. L., Savinov M. Yu. et al. System of neon concentrate
fine purification. Vestnik MGTU. Special issue Refrigeration, cryogenic technology,
systems of air-conditioning and life provision (2005) 24-32.

333

CR08-45
NEON LIQUEFIERS AND THEIR USAGE IN THE
INSTALLATIONS FOR RARE GASES EXTRACRION
Bondarenko V.L.1, Diachenko .V.2, Diachenko O.V.2
1

Moscow Bauman State Technical University, 5, 2-nd Baumanskaya Str.,

107005, Moscow, Russia
2
Iceblick, Ltd., 29, Pastera Str., 65026, Odessa, Ukraine
ABSTRACT

We have shown throttle cycles for neon liquefaction with preliminary cooling at the
temperature level T = 6678 K in comparison. The classic Lindes cycle and the variants
of installations with the intermediate pressure working agent have been considered. The
influence of the pressures on liquefaction coefficient, specific nitrogen flow and power
consumption have been researched. The usage of diaphragm compressors used in rare gases
extraction technologies has been justified.
INTRODUCTION
During the last stages of neon-helium mixture separation and extraction of Ne isotopes by
rectification the temperatures of about T = 30K are used. Under these conditions it is
preferable to use neon as a working body as an effective and safe refrigerant. Besides the
task of cryogenic support of separation, the liquefaction of neon reduces the freight and
warehouse costs. Liquid neon can be extracted by the direct recovery of the product from
the cube of the neon-helium rectification column [1] or with the help of a separate
installation [2]. The second alternative is more preferable since it allows efficiently
producing a given volume of liquid Ne without hindering the operation of the installation
for the neon-helium mixture separation.
THROTTLE LIQUEFIERS ON THE BASIS OF TWO PRESSURE CYCLE
In comparison with the classic Lindes cycle, the throttle cycle with the working agent
intermediate flow is less power-consuming. Consider two types of such installations
(Figure 1, 2), different only in the way of the compressors connection. In scheme (I) the
compressor of the type 1,6DC-10/12,5 is used as the machine of medium pressure (Table
1). For providing high pressure of the working medium one-stage boosters 1,6DC-16/12,5200 or 4,0DC-60/12,5-200 with increased suction pressure (1 = 1,2 MPa) are suitable. The
last unit has higher capacity and along with the standard high-pressure compressor 4,0DC20/200 is used as C2 in the installation (II).
Overpressure, MPa
Initial

Final

Engine
power, kW

9,8 (2,3)

0,02

20

5,4

1,6DC-10/12,5

11 (2,5)

0,02

1,25

2,0

1,6DC-16/12,5-200

21 (4,8)

1,25

20

6,7

4,0DC-20/200

20 (4,6)

0,02

22

11,4

4,0DC-60/12,5-200

70 (16,1)

1,25

20

15,0

Compressor type

Capacity, nm3/h (g/s)

1,6DC-8/200

Table 1: Diaphragm compressors produced by Ural

335

Compressor Plant OJSC

C1

G
3

C2
3N

C1
17

1N

16

13
HE

G*m

CC

11

T r2

10

HE
9

CC

10

T r2

12

CC

11

13

15

T r1

16
HE

8
HE

NB
5

12

CC

2N

14

15

T r1

17
HE

NB

3N

2N

C2

G*m

14
HE

II

Figure 1: Two pressure cycles with series (I) and parallel (II) connection of the compressors
T
T

2 1

17

14

5
N

6
9 8

T
16

7 15

0 10

11

13

12

Figure 2: T-s diagram of two-pressure cycle

With respect to the G consumption on the output of the second compressor the liquefaction
coefficient is x = m z . In this correlation m is the content of liquid in the condensate
collector C1, and z is the liquefaction coefficient with respect to the last throttle stage HE 3Th2

336

z=

i12 i8 q4
i12 i0

(1)

where: q4 is the specific heat leakage to the throttle stage, kJ/kg, i0, i8, i12 enthalpy in the
corresponding cycle points, kJ/kg.
The intermediate flow consumption is found from the circuit balance HE2-Th1-C1

m=

i16

i16 i 5 q 3
.
i8 (1 z ) (i13 i12 )

(2)

The specific nitrogen consumption, fed into the bath NB, is equal to

G N i3 i5 (1 m) (i17 i16 ) (m x) (i14 i13 ) + q1 + q 2

=
,
G
i3 N i1N
(3)
where: G and GN are the working medium expense on the outcome of the second
compressor and the outer refrigerating medium (nitrogen), correspondingly, kg/s; q1, q2, q3
heat leakages to the corresponding cycle stages.
Specific power expenses of the installation with the intermediate pressure IN according to
the scheme (Figure 1, I) is

l=

R TA
L

1000 x C

P
P G
ln 2 + m ln IN + N l N + lVAC ,
P1 G
PIN

(4)

where: L= 1206 kg/m3 liquid neon density; TA temperature of ambient air; R = 411 J/
(kg) gas constant of neon; P2, P1 pressures of the direct and reverse flow, MPa,
correspondingly; C isothermal coefficient of efficiency of the compressor; lN = 4,3 MJ/kg
specific energy expenses for obtaining liquid nitrogen. The value lVAC includes energy
consumption for the vacuum pump driving gear during nitrogen boiling in the bath NB at
reduced pressure (N < 77,4 K). Depending on the type of the evacuation system and the
temperature level lVAC can reach 25% of the compressor capacity.
In case of parallel connection (Figure 1, II) specific energy expenses are

l=

R TA

L
P
P G

(1 m) ln 2 + m ln IN + N l N + lVAC .
1000 x C
PIN
P1 G

(5)

The calculations have been made for the mode: P1 = 0,15 MPa; P2 = 20 MPa; TN2 = 66 .
Analysis shows, that other things being equal, the intermediate flow pressure increase IN
leads to the increase of the content of the liquid m in the collector C1. At m1 the
intermediate flow expense is (1 - m)0 and the cycle degenerates into Lindes throttle
cycle.
The information shown in the graphs (Figure 3) proves the obvious energy advantages of
the two pressure cycle. At lower expenses for compression, the liquefaction coefficient of

337

the cycle under study is in the same range as the Lindes cycle. The extent of energy
consumption depends on the compressors connection scheme. At the series connection
specific energy consumption is on average 1013% lower than for the Lindes cycle.
Parallel connection of the compressors reduces energy consumption 1520% more.
Diaphragm compressors of Ural Compressor Plant OJSC (Yekaterinburg, Russia) are
produced on two bases: 1,6DC 4,0DC (Table 1). The outward appearance of several units
is shown on Figure 4. Table 2 contains design parameters of real liquefiers with heat
leakages and under recuperation taken into consideration. The basic criterion of the choice
of the design is specific expenses per 1 dm3 of liquid neon. This coefficient included the
given equipment and operational costs.

Figure 3: a, b: two pressures cycle (for the temperature of neon

after the nitrogen bath 6 = 66)
4,0DC-60/12,5-200

4,0DC-20/200
1,6DC-8/200

Figure 4: Diaphragm compressors of low (a), medium (b) and high (c) productivity
Table 2: The description of two pressures throttle liquefiers on basis of diaphragm compressors
( = 20 MPa; PIN = 1,25 MPa)
Compressor type

Productivity, l/h

Power
inputs,
kWh/l l.Ne

Liquid N2
consumption,
kg/h

At the series compressors connection (Figure 1, I)

1,6DC-10/12,5

1,6DC-16/12,5-200

3,8

3,8

5,0

4,0DC-20/200 (2)

4,0DC-60/12,5-200 +
+ 1,6DC-16/12,5-200

16,4

3,3

22,0

338

At the parallel compressors connection (Figure 1, II)

1,6DC-8/200

1,6DC-16/12,5-200

3,2

3,9

4,3

4,0DC-20/200 (2)

4,0DC-60/12,5-200

12,6

3,8

16,8

High pressure throttle cycles with intermediate cooling on liquid nitrogen level are used for
light inert gases separation [1] and getting neon isotopes [2]. The first installation
(Figure 5, a) needs an additional neon refrigeration cycle with nitrogen vapour pumping
( 66), as the throttle effect of the neon-helium mixture under separation is not enough
to compensate the heat leakages to the cryogenic unit. In a relatively compact device for
neon separation into isotopes 20Ne and 22Ne (Figure 5, b) a simpler alternative of the cycle is
used (without the nitrogen vapour pumping). High pressures 2 = 1618 MPa are
practiced into start-up period lasting 1620 h. In the steady-state regime lasting from 5 to
20 days, the compression pressure decreases to 1012 P, and the required refrigerating
capacity of the cycle does not exceed 30 W.
The considered cycles do not include all the alternatives of the installations we have
researched and introduced. Particularly promising are the medium pressure designs using
outer low-temperature refrigeration systems. As such it is possible to use two-stage gas
cryogenic machines, CGM-100/20, helium refrigerators of the type KGS-600 and also the
stages using the cold flow exergy in liquid helium evaporation devices.
VP

77% N e+23% He

18% N e+82% H

20N

e + 22N e
20N

H E1
N B

N B

H E2
N

Rs

D C

9 9 ,9 9 9 % N e

D C

H E2

R C
2

20N

E
R C

PS

C n

C n

H E3

22N

Figure 5: Throttle neon cycles: a in the unit of neon-helium mixture rectification separation;
in the installation for getting neon isotopes (refrigerated cube compressor is used for pumping
productional isotope 20Ne). DC - diaphragm compressor; VP - vacuum pump; N - nitrogen bath;
1-3 - heat exchanger; PS - helium phase separation ; RC - rectification column;
Cn and E - condenser and neon cycle evaporator;
Rs - working agent receiver (20Ne)

339

CONCLUSION
For cryostatting the objects within the temperature range of = 2840 it is preferable to
use neon as the working agent as an effective and safe refrigerant. It is rational to create
low-expense liquefiers on the basis of traditional high pressure throttle cycle with
intermediate cooling. For the units with the productivity more that 15 l/h the transition to
more economical schemes with using two pressures cycles is more preferable.
Using diaphragm compressors in neon cycles simplifies the system of working agent
preparation. Such a solution ensures high quality of the products in saleable neon
liquefaction units and isotopes separation systems 20Ne 22Ne.
REFERENCES
1. Bondarenko V. L., Arkharov A. M., Golubev A. A. et al. Pilot-Commercial Plant for
High Purity Neon Production. Preprints of the XX International Congress of
Refrigeration, Sydney, Australia (1999) 14.
2. Arkharov A. M., Arkharov I. A. Bondarenko V. L. et al. Production of neon isotopes by
rectification method at 28K Proc. 9 Int. Conf. Cryogenics 2006, Praha (2006) 247250.

340

CR08-35
GAS-CHROMATOGRAPHIC ANALYSIS OF MIXTURES OF
HYDROGEN ISOTOPES USING DIFFERENT PARAMETERS
Preda A., Bornea A., Pearsica C., Vasut F.
Institute of Isotopic and Cryogenic Technologies ICIT Rm.Valcea,Romania
ABSTRACT
Gas-chromatography is considered to be the most appropriate of many analytical
techniques used to determine the composition or purity of gases.
The method for separating the isotopic species of hydrogen on a moderately large scale
using low temperature is gas chromatography.
In this paper, for the analysis of gas mixtures containing hydrogen isotopes, we presented
the analysis of mixtures of hydrogen isotopes using different parameters as: different
temperature of the column oven and different sample loops. It is realized a comparative
study to develop or improve existing methods for the qualitative and quantitative
determination of the composition of gas mixtures of hydrogen isotopes.
As results, there are presented chromatograms for different H2, HD, D2 mixtures and
different operated parameters.
INTRODUCTION
Chromatography is now an extremely versatile technique; it can be separate gases, and
volatile substances by Gas-chromatography.
By classical definition, chromatography is a separation process that is achieved by
distributing the substances to be separated between a moving phase and a stationary phase.
Those substances distributed preferentially in the moving phase pass through the
chromatographic system faster than those that are distributed preferentially in the stationary
phase. As a consequence, the substances are eluted from the column in inverse order of
their distribution coefficients with respect to the stationary phase [1].
Gas chromatography separation of hydrogen isotopes have been reported in the literature
dating from the late 1950s. Basically, three approaches have been employed to effect
separations on an analytical scale, and these approaches may be distinguished on the basis
of the column packing material used.
Three different column packing materials have been reported for the gas chromatographic
separation of hydrogen isotopes. These are molecular sieves used for size exclusion
chromatography, alumina used for gas/solid adsorption chromatography, and palladium
such as palladium dispersed on alumina used for catalytic adsorption chromatography.
The species isotopic of hydrogen are: H2, D2, T2, HD, HT, DT, ortho-H2, para-H2, orthoD2, or para-D2, where D stands for 2H and T for 3H [2].
One of the objective of our laboratory is the enhancement and/or development the gaschromatographic method for the separation of hydrogen isotope mixtures.

341

This paper discusses the specific design of the gas chromatograph using at our analysis.
Details of flow inside the column, about the column, the utilized detector and
chromatograms measured for a specific gas mixture at different parameters are presented.
EXPERIMENTAL
The gas chromatograph employed in this work was a type 3800 from Varian Analytical
Instrument.
The Varian 3800 gas-chromatograph is equipped with a capillary molecular sieve 5A
column with following characteristics:
The length of the GC column is: 50 m;
The inside diameter of the GC column is: 0,32 mm;
The film thickness of the GC column is: 30 m.
As detector, we used a Pulsed Discharge Helium Ionization Detector (PDHID).
The temperature of the filament of detector was fixed at 2000C.
The carrier gas used was: Helium (99,999% purity). It is recommended that a quality grade
of helium 5.0 (99,999 % pure or better) be used at all times.
The sample loops used for these analysis was: 5 L and 10 L.
The operating temperature of the GC column for these experiments was -99 oC and -75 oC.
The temperature of the oven of the GC column was maintained in the range of 0 C to -99
C, by spraying liquid nitrogen into the oven. A temperature controller to control the liquid
nitrogen flow and the heater was used [3].
RESULTS AND DISCUSSIONS
In this paper, for the analysis of gas mixtures containing hydrogen isotopes, we present the
analysis of two sample of mixtures of hydrogen isotopes with different concentration, using
different parameters as: different temperature of the oven column at -75 0C and -990C and
different sample loops : 5L and 10 L.
The first sample has 10% concentration of deuterium in the hydrogen isotopes mixture and
the second sample has 50% concentration of deuterium in the hydrogen isotopes mixture.
We present a comparative study between gas chromatograms obtained at these parameters.
The gas chromatograph column was conditioned before to use this for the separate of the
hydrogen isotopes mixtures.
The method used was calibrated with standard gas of protium and deuterium by external
standard calibration type.
In the Figure 1 and Figure 2 we analyzed the sample number one.
The method described in this was based on using a capillary molecular sieve 5A column
which has been operated at -750C.

342

Figure 1: Chromatogram of sample 1 at -750C temperature of GC column and 5 L sample loop

In Figure 1, the sample loop used for this analysis was 5L and in the Figure 2, the sample
loop used was 10 L.

Figure 2: Chromatogram of sample 1 at -750C temperature of GC column and 10 L sample loop

In the both cases, the carrier flow rate was 3,0 mL/minute, the linear velocity was 30,2
cm/second and the pressure was 10 psi.
The retention times were relatively short, about 8-9 minutes, and the result is a bad
separation of H2 and D2. In the Figure 1 when we used 5L sample loop can observe a
separation of mixture of hydrogen and deuterium but it is bad and in the Figure 2 we used
10L sample loop the separation between hydrogen isotopes dont exist.

343

Figure 3: Chromatogram of sample 1 at -990C temperature of GC column and 5 L sample loop

In the Figure 3 and Figure 4, we analyzed the same sample as in first two chromatograms,
sample number one.
The method described in this was based on using a capillary molecular sieve 5A column
which has been operated at -990C.
In Figure 3, the sample loop used for this analysis was 5L sample loop and in the Figure
4, the sample loop used was 10 L sample loop.

Figure 4: Chromatogram of sample 1 at -990C temperature of GC column and 10 L sample loop

In the both cases, the carrier flow rate was 3,7 mL/minute, the linear velocity was 32,8
cm/second and the pressure was 10 psi.
The retention times were about 8-9 minutes. In the both chromatograms we can observe a
good separation of hydrogen isotopes mixture but better in Figure 3 when we used 5L
sample loop than in the Figure 4 when we used 10 L sample loop.
For the next four chromatograms we used for analyzing the sample number two with a 50%
deuterium concentration in the hydrogen isotopes mixture.
344

Figure 5: Chromatogram of sample 2 at -750C temperature of GC column and 5 L sample loop

In Figure 5, the sample loop used for this analysis was 5L sample loop and in the Figure
6, the sample loop used was 10 L sample loop. The temperature of column oven was
-750C in the both cases. Also, other parameters of method had the next values: the carrier
flow rate was 3,0 mL/minute, the linear velocity was 30,2 cm/second and the pressure was
10 psi.

Figure 6: Chromatogram of sample 2 at -750C temperature of GC column and 10 L sample loop

In these chromatograms, we can observe a bad separation of hydrogen isotopes at this
operating of column temperature.
In the next two figures we analyzed the sample number two and the method described in this was

based on using a capillary molecular sieve 5A column which has been operated at -990C.

345

Figure 7: Chromatogram of sample 2 at -990C temperature of GC column and 5 L sample loop

In Figure 7, the sample loop used for this analysis was 5L sample loop and in the Figure
8, the sample loop used was 10 L sample loop.
In the both cases, the carrier flow rate was 3,7 mL/minute, the linear velocity was 32,8
cm/second and the pressure was 10 psi and the retention times were about 8-9 minutes.
In these chromatograms, we can observe a very good separation of the isotopic species of
the hydrogen: para-hydrogen, ortho-hydrogen, HD and deuterium.
Molecular hydrogen occurs in two isomeric forms, namely with its two proton spins aligned
either parallel (orthohydrogen) or antiparallel (parahydrogen). In the state of thermal
equilibrium at room temperature dihydrogen contains 25 % of parahydrogen (nuclear
singlet state) and 75 % of orthohydrogen (nuclear triplet state).
Orth- and para-hydrogen may easily be separated on column of molecular sieve. In all
cases, the para-form appears before the ortho-form. Ortho- and para-deuterium are not so
easily separated. Normal deuterium contains 66% ortho-deuterium and 33% paradeuterium.

346

Figure 8: Chromatogram of sample 2 at -990C temperature of GC column and 10 L sample loop

Chromatograms from Figure 7 and 8 are presented in Figure 9 to can see an evident
difference between two sample loops: 5 L sample loop is for the small chromatogram and
10 L sample loop is for the high chromatogram.

Figure 9: Chromatogram of sample 2 at -990C temperature of GC column and 5 and 10 L

sample loops
CONCLUSIONS
In this paper we could observe an unsatisfying separation of isotopic species of hydrogen at
-750C temperature of column and a good separation of isotopic species of hydrogen at
-990C temperature of column.
Also, we could observe a good separation of the two isomeric forms of hydrogen,
orthohydrogen and parahydrogen.
At 750C temperature of column, the carrier flow rate was 3,0 mL/minute, the linear
velocity was 30,2 cm/second and the pressure was 10 psi, and at -99 0C temperature of
column, the carrier flow rate was 3,7 mL/minute, the linear velocity was 32,8 cm/second
and the pressure was 10 psi.
Between two sample loops used, a good separation was obtained using 5 L sample loop.
Conclusively, the good method is the method based on using a capillary molecular sieve 5A
column which has operated at -990C and using 5 L sample loop.
REFERENCES
1.

SCOTT, R.P.W., Techniques and Practice of Chromatography, Marcel Dekker,

INC, New York, USA (1995)10-15

2.

LITTLEWOOD, A.B., Gas Chromatography, Second ed., Academic Press: New

York and London (1970) 427-430

3.

LASSE, R., GLUGLA, M., GUNTHER, K., PENZHOM, R.D., WENDEL, J.,
Fusion Science and Technology (2002), 41-43.

347

CR08-19
MATHEMATICAL MODELS CONCERNING THE DESIGN OF
COLUMN FOR ISOTOPIC EXCHANGE PROCESS IN THE PILOT
PLANT FOR TRITIUM AND DEUTERIUM SEPARATION
Gherghinescu S.1, Popescu G.1
National Institute of R&D for Cryogenics and Isotopes Technologies (ICIT), ROMANIA,
ABSTRACT
The present work has the purpose to determine the flow behavior of both phases, gaseous
and liquid, of the hydrogen isotopes in order to obtain a better separation factor between
hydrogen and water, D/T, in the D2-DTO large scale isotopic exchange column.
Seeing that direct determination of the fractions of the hydrogen isotopic species is difficult
and rather imprecise, the numerical estimation method with high precision is strongly
required.
Therefore we shall use series of isotope exchange process characteristic equations to obtain
the precise results needed in the process design for the achievement of the operating data
target.
A special attention will be paid to the two first weight reactions in the isotopic exchange
process:

D 2 ( g ) + DTO (v ) DT ( g ) + D 2O (v )
DTO (l ) + D 2O (v ) DTO (v ) + D 2O (l )
KEY-WORDS: separation factor, hydrogen isotopic species, isotopic exchange column.
INTRODUCTION
In the D2-DTO isotopic exchange column, the tritium transfer occurs through three phases
of hydrogen gas, water vapor and liquid water. Along the column the concentration of
tritium will decrease from the column top to the bottom. Therefore, for the optimum plant
design, it is necessary to determinate the D/T-isotopic separation factor and the dependence
with the tritium concentration.
The object of this study is to establish a simple calculation method, in which the changes of
overall D/T -isotopic separation factor on the operation temperature and the concentration
of tritium in liquid water could be evaluated accurately.
A special attention will have the two first weight reactions in the isotopic exchange process
between different phases (gas/vapor and vapor/liquid):
DT(g)+D2O(v)D2(g)+DTO(v)
DTO(v)+D2O(l) D2O(v) +DTO(l)

349

DTO
D2

Purification

CD

DT or T2

LPCE

Storage

MODERATOR

D2
D2O

Fig. 1

For these exchange reactions the separation factor g and l are defined as follows:

l =

[ DTO( l) ][ D2O( v ) ] = [ x / (1 x ) ]
[ D2O( l) ][ DTO( v ) ] [ z / (1 z ) ]

for the tritium transfer between hydrogen gas and water vapor

g =

[ DTO( v ) ][ D2 ( g ) ] = [ z / (1 z ) ]
[ D2O( v ) ][ DT ( g ) ] [ y / (1 y ) ]

and for the tritium transfer between water vapor and water liquid, where x, y and z shows
the atom fractions of tritium in liquid water, hydrogen gas and water vapor. In the end, the
totally separation factor D/T is determined as following:

D/T= gl
In the D2-DTO isotopic exchange process, the transfer of tritium among hydrogen gas,
water vapor and liquid water is described by following exchange reactions:

D2(g)+T2(g) 2DT(g)
K1
(R1)
DT(g)+D2O(v)D2(g)+DTO(v)
`
K2
(R2)
D2O(v)+T2O(v)2DTO(v)
K3
(R3)
DTO(v)+D2O(l)DTO(l)+D2O(v)
K4
(R4)
D2O(l)+ T2O(l)2DTO(l)
K5
(R5)
350

where K1, K2, K3, K4 and K5 are the correspondent equilibrium constants, temperatureresponsive, for the isotope exchange reactions, (R1) to (R5), respectively.

[ DT ]2g
K1 =
[ D2 ] g [T2 ] g
K2 =

= f1 ( T )

[ D2 ] g [ DTO]
[ DT ] g [ D2O]

= f 2 (T )

2
[
DTO ]
K3 =
[ D2O] [T2O]

K4 =

= f3 (T )

[ D2O] [ DTO] l = f ( T )
[ DTO] [ D2O] l 4

2
[
DTO ] l
K5 =
= f (T )
[ D2O] l[T2O] l 5

x) The expression of the atom fraction of tritium in liquid water,x is

x =

2lT 2O + l DTO

2 lT 2O + l DTO + l D 2O

where lD2O, lDTO and lT2O represent the mole fractions of D2O, DTO and T2O in liquid water.
Hence, lD2O+lDTO+lT2O=1 and

x = lT 2O +

1
1

l DTO = l DTO R l +
2
2

where Rl is the molar fraction ratio of T2O to DTO

Rl =

lT 2O
l DTO

therefore

l DTO =
And K5 became

x
Rl +

1
2

lT2O =

Rl x
l
1 D2O
Rl +
2

351

Rl + ( Rl + 1) x
2
=
1
Rl +
2

K5 =

x2
( Rl x ) Rl + 1 ( Rl + 1) x
2

By solving the above second order equation, Rl can be expressed as function of x and K5

( 2 x 1) +
Rl =

1 + 4 x(1 x )
1
K5

4(1 x )

and the atom fraction of tritium in liquid water x is given as

x=

K 5 Rl ( 2 Rl + 1)
2[1 + K 5 Rl ( Rl + 1) ]

Then, the abundance ratio of tritium in liquid water is derived as

K R ( 2 Rl + 1)
x
= 5 l
1 x
2 + K 5 Rl
y) The expression of the atom fraction of tritium in hydrogen gas y is

y=

2hT 2 + hDT
1

= ... = hDT Rg +
2( hT 2 + hDT + hD 2 )
2

Where hT2, hDT and hD2 represent the molar fractions of T2, DT and D2 in hydrogen gas, and
Rg is the molar fraction ratio of T 2 to DT. Using the same calculation pattern we obtain the
equilibrium constant for the gaseous reaction K1 as

[ DT ] 2g
K1 =
[ D2 ] g [T2 ] g

y2
( y Rg ) Rg + 1 ( Rg + 1) y
2

and the atomic fraction of tritium in hydrogen gas y as

y=

K1 Rg ( 2 Rg + 1)

2 1 + K1 Rg ( Rg + 1)

Then, the abundance ratio of tritium in hydrogen gas is derived as

K R ( 2 Rg +1)
y
= 1 g
1 y
2 + K1 Rg

z) The expression of the atom fraction of tritium in water vapor z is

z=

2vT 2O + vDTO
1

= vDTO Rv +
2( vT 2O + vDTO + vD 2O )
2

Where vDTO represent the molar fraction of DTO in water vapor and R v is the molar fraction
of T2O to DTO. The molar fractions of DTO(vDTO), T2O(vT2O) and D2O(vD2O) will be

352

calculated as functions of z and Rv and the equilibrium constant of the vapor phase reaction
K3 became

K3 =

[ DTO]2
[ D2O] [T2O]

z2
( Rv z ) Rv + 1 ( Rv + 1) z
2

and from here z is expressed by

z=

K 3 Rv ( 2 Rv + 1)
2[1 + K 3 Rv ( Rv + 1) ]

where the molar ratio Rv is represented as

Rv =

[T2O] v
[ DTO] v

1 [ DTO] v

K 3 [ D2O ] v

By substituting the equilibrium constants K4 and then K5 we get the relation for z

z=

Rl K 5 ( 2 Rl K 5 + K 3 K 4 )
2
2 K 3 K 4 + Rl K 5 ( Rl K 5 + K 3 K 4 )

Then, the abundance ratio of tritium in water vapor is derived as

z
R K ( 2 Rl K 5 + K 3 K 4 )
= l 5
1 z
K 3 K 4 ( 2 K 4 + Rl K 5 )
CONCLUSIONS
Now that we have the description of x, y and z, which shows the atom fractions of tritium
in liquid water, hydrogen gas and water vapor, the equilibrium constants K 1 to K5 and the
molar ratios Rg, Rv and Rl we can evaluate the separation factors g, l and D/T

[ x / (1 x ) ] = K 3 K 4 ( 2 Rl + 1)( 2 K 4 + Rl K 5 )
[ z / (1 z ) ] ( 2 + Rl K 5 )( K 3 K 4 + 2 Rl K 5 )
[ z /(1 z ) ] = K1K 2 ( 2 Rl K 5 + K 3 K 4 )( 2 K 2 K 4 + Rl K 5 )
g =
[ y /(1 y ) ] K 3 ( 2 K 4 + Rl K 5 )( 2 Rl K 5 + K1K 2 K 4 )
l =

Finally, the overall separation factor will be

D / T = l g =

K1K 2 K 4 ( 2 K 2 K 4 + Rl K 5 )( 2 Rl + 1)
( 2 + Rl K 5 )( 2 Rl K5 + K1K 2 K 4 )

where

353

( 2 x 1) +
Rl =

1 + 4 x(1 x )
1
K5

4(1 x )

With the temperature-responsive equilibrium constants K1, K2, K3, K4 and K5 and
knowing the atom fraction of tritium in liquid water, x, it will be possible to determinate the
theoretical shape of the overall separation factor
according with the operation
temperature and the concentration of tritium in liquid water. In the real operating conditions
it will be necessary to determinate the efficiency of the catalyst which will determinate the
D /T

D / T . By direct comparation with the

D /T
theoretical shape of the overall separation factor
between different operation
real shape of the overall separation factor

temperatures and concentrations of tritium in liquid water it will be possible to determinate

the optimal operating conditions for the isotopic exchange column.
BIBLIOGRAPHY
(1) J.H.Rolston, J.dem Hartog and J.P.Butler: The Deuterium Isotope Separation
Factor between Hydrogen and Liquid Water, J.Phys. Chem., 80(10), 1064-1067
(1976)
(2) Masami Shimizu, Kenji Takeshita "Simplified calculation method of deuterium
separation factor between hydrogen and water (H/D) depending on D-atom
fraction of liquid water" FIFTH CONFERENCE ON ISOTOPIC AND
MOLECULAR PROCESSES PIM 2007
(3) J.H.Rolston and K.L.Gale: Deuterium-protium Isotopic Fractionation between
Liquid Water and Gaseous Hydrogen, J.Phys.Chem., 86(13), 2494-2498 (1982)

354

CR08-15
THE CREATION OF VEHICLES FOR MULTIMODAL
TRANSPORTATION OF LIQUEFIED GASES
Zashlyapin R.A., Cheremnych O.Ya.
JSC UralCryoMash, Russian Federation, Sverdlovsk region, Nizhny Tagil
ABSTRACT
In this report there given information about the enterprise activities concerning the creation
of effective means of transportation, tank-containers for multimodal transportation of LPG,
(LH2). Application analysis of different types of TC heat isolation is also given. For safety
transportation of LNG and LH2 special drainage and locking-safety devices are used.
Delivery experience of the first LNG consignment (using tank-container) from Russian
Federation to European country is analyzed.
INTRODUCTION
JSC UralCryoMash is established as a development contractor and producer of vehicles
for transportation and storage of liquefied low-temperature gases (nitrogen, oxygen, argon,
ethylene, propane-butane mixtures, carbon dioxide, hydrogen, LNG) for different industrial
branches [1, 2]. Enterprise has a great experience (gained in 1967-1990) in creating and
safety operation of railway and motor vehicles for large-scale transportations of liquid
nitrogen, oxygen, argon (8513, 15-558 ), liquid hydrogen (LHT-100, LHT-100 M),
carbon dioxide (15-559), ethylene (15-147). This experience is used today at creation of
new generation vehicles - tank-containers for multimodal liquefied gases transportation for
the sake of space, energetic, oil-gas industrial branches.
Nowadays multimodal transportation (mixed) of liquefied gases (methane, hydrogen,
oxygen, nitrogen, argon, ethylene, etc.) becomes more and more effective. Such
transportation has got the following advantages:
-

the ability of tank-container transportation by railway, motorway and seaway

direct delivery from manufacture to consumer
no liquid loss, arising from its being pouring from transportation vessels to stationary ones
no necessity to use expensive terminals for filling/discharge
high reliability and safety
comparatively cheap transportation
When creating tank-containers the producer has the following objects:
Its a constructive solution that allows tank-container operating on present filling trestle
bridges at liquefied gases filling works and operation on discharge trestle bridges at
consumers site, particularly at Europeans;
Nomenclature widening of transported liquefied gas in the given type of tank-container or
its modifications;
Increasing of the transported mass of liquefied gas due to creation of 40 or 45 foot
containers in accordance to international standards, which determine basic design and
operating regulations at different ways of transportation [3,4,5];
Generality and individuality of constructive solution from the point of view of similar to
thermo physical and chemical properties of transported liquefied gases (propane-butanepropylene, nitrogen-oxygen-argon, LNG-ethylene, hydrogen-helium);
355

1. TANK-CONTAINERS FOR NONCOOLED LIQUEFIED GASES

TRANSPORTATION (LPG)
Liquefied noncooled gases (propane, butane, propylene and others) are widely used in
industry and in private life. The enterprise has developed and mastered the production of
different types of tank-containers for these gases: type size 1CC-TC-25/2,0, TC-25/2,2, TC25/1,8; they have both combined discharge (upper-bottom) and upper or bottom;
type size 1 AA TC- 52/1,8 with combined joint filling-discharge and joint of gas
drainage (picture 1).
Technical characteristics of tank-containers are given in table 1.
Tank-containers are gathered with own safety locking valves or with valves of Fort Valve
firm.
Tank-containers filling control is done with the help of level control block (in contrast to
railway tanks) which is connected by the cable to transmitter set on a vessel; this system
ensures the control of percentage filling of tank-container. When filling tank-containers
with noncooled gases IMDG code is followed; By IMDG the maximum degree of TC
filling with propane-0,42kg/l, butane 0,51 kg/l, propylene 0,43 kg/l.

TC-25/2,0
1CC
9,6
24
2,0
B
B
TC-25/2,8 HC
1CC
96,
24
2,0
B
H
B
TC-25/2,0 HC-01
1CC
9,6
24
2,0
H
B
TC -25/2,2
1CC
9,6
24
2,2
B
B
TC -25/2,2 HC
1CC
9,6
24
2,2
B
H
B
TC-25/1,8
1CC
6,4
24
1,8
B
H
B
H
TC-52/1,8
1AA
11
34
1,8
B
H
B
H
Table 1. Technical characteristics of tank-containers for noncooled liquefied gases.

Total capacity, m3

Upper

Bottom

Drainage
method

Bottom

Filling/
discharge
method
Upper

MPaWorking pressure in vessel,

Tank-container
model

Maximum gross weight, t

According to table 1 at empty weight of TC-25/1,8 6,4 t and TC-52/1,8 11t, maximum
allowable mass of transported liquefied gas is 17,6 t and 23 t. At normal filling with
propane 0, 42 kg/l its mass in TC-25/1,8 comes to 10,8 and 21,53. Thus, from the point of
view of steel intensity 40-foot container is 1620% more effective than two 20-foot
containers model TC-25/1, 8.
Empty weight

To ensure the maximum period of nondrainage transportation or control retainer time (time
between the first filling condition at the pressure 0,130,15 MPa and pressure raising as a
result of heat leakage, i.e. pressure opening of safety valves);
Observance of requirements for ecological and fire safety regulations in case of emergency
(heat isolating cavity decapsulation or vessel MAWP exceeding)

Designation ISO

25
25
25
25
25
25
52

Tank-containers are certified by Russian Maritime Register and correspond to international

standards IMDG;ADR;RID;CSC;CCC;ISO1496; ISO 668:1995; ISO 6346:1995; Russian
national standards: GOST 14249-89; GOST25290-82; GOST 26291-94.
356

2. TANK-CONTAINER FOR LNG TRANSPORTATION

Creating tank-container TCM -35/0,6 constructors took into account requirements to its
overall mass characteristics, determined by ISO 6346;1995. Regulations and instructions
for big-volume heavy loads transportation by motor way in Russian federation, CIS, Baltic
and European countries were also considered.
The analysis of allowable overall dimensions, full mass and axial loading of carriers in RF,
CIS, Baltic and European countries at 40 foot TC transportation showed that allowable
carrier mass including hauler with semi-trailer and 40 foot TC ISO can variate from 38 t
(RF, Belorussia, Kazakhstan) to 42:44t (Ukraine, Baltic, Finland, Sweden, Spain and
others).
If proceeding from normative requirements and taking into account that the mass of a five
tractive unit is 11-12t and allowable gross mass up to ISO standarts -30,48 t, so TC mass
for LNG for its exporting to European countries from RF mustnt exceed 30,48 t.
These directions became the basis for creation of TCM 30/0,6 with capacity 35 m3, tare
mass 14,95 and maximum mass of transported product 15,53 t (picture 2).
Technical characteristics of TC are given in table 2.
Parameters
Values
Type size ISO 14961995
1AA
Maximum gross weight, t
30,48
Tare mass, t
14,95
Total capacity, m3
35,36
Maximum allowable working pressure in
0,6
vessel, MPa
Loss of evaporation per day, l/day
0,39
Filling degree, kg/l not more
0,355
Period of nondrainage storage at pressure,
35
in days, not less
Type of heat isolation
Fiber-vacuum
Reservoir material:
12x18H10T
- vessel
092-14
- casing
Discharge method
Reservoir blowing from the side source
Table 2. Technical characteristics of TC for liquefied natural gas (LNG) model TCM 35/0,6

The enterprise has developed and uses different types of isolation for vehicles transported
liquefied cryogenic gases: powder-vacuum (using of basalt fiber as heatisolating mats);
multilayer screen-vacuum (using metalized film and glasscloth). For LNG tank-container
both types of isolation can be used.
Powder-vacuum isolation on basis of pearlite has got some disadvantages such as:
unavoidable pearlite shrinkage in cavity at transportation loadings and considerable
vacuuming time of cavity filled with pearlite.
Screen-vacuum heatisolation is more effective than powder-vacuum one, but its cost is
considerably higher than the latter.
Time of nondrainage transportation or the control time of product retaining is an important
characteristic of cryogenic tank-container and comes out from the heatisolation
effectiveness and maximum allowable value of working pressure in the vessel of TC.
An important moment must be taken into consideration when transporting LNG
requirements keeping for ecological and fire safety regulations.
357

Tank-container model TCM-35/0,6 is equipped with safety drainage device, that ensures
safe LNG vapor dumping from the vessel due to gasdynamic destabilization of drainaged to
the atmosphere gas burning at reaching the maximum allowable pressure in the process of
transportation [6].
In case of an emergency at LNG transportation and vacuum loss in heatisolating cavity,
safety devices protect the vessel from destruction and provide safety vapor dumping of
LNG to the atmosphere.
For safety LNG discharge to the reservoir tank-container model TCM 35/0,6 is equipped
with the following safety devices:
- Safety-locking device intended for discharge pipeline by means of its automatic cutoff at a
casual movement of tank-container during discharge.
- Safety-locking device together with fireproof device ensuring the discharge pipeline cutoff
in case of fire
- High-speed valve, used for discharge pipeline cutoff in case of decapsulation of discharge
system at consumers site.
For some technological questions working out, from the liquefied gas manufacturer plant
(AGNKS, St. Petersburg, RF) to European consumer (Lingcheping, Sweden) LMG
transportation in TC model TCM-35/0,6 on automobile container carrier (hauler with semitrailer) was done: St.Petersburg Helsinki (Finland) sea ship Stockholm Lingcheping
(Sweden); total way of transportation 1100 km, way time 48 hours, average speed 6070 km/per hour.
At arriving LNG discharge from TC to stationary vessel-reservoir was done (capacity 53m,
working pressure 1,5 MPa, picture 4).
For LNG exporting tank-containers model 35/0,6 (developed by UralCryoMash) can be
used.
3. TANK-CONTAINERS FOR LIQUED HYDROGEN TRANSPORTATION.
At present time liquid hydrogen is widely used in rocket- space techniques, aviation,
energetic [2]. For LH transportation to consumer, railway transportation tank-containers
model RHT 100M, road tank model 17228 are created [8].
Parameters
Values
Type size ISO 1496-31995
1BB
Gross weight, t
12,47
Tare mass, t
11,2
Mass of transported hydrogen, t
1,27
Total capacity, m3
20,0
Maximum allowable working pressure in vessel, MPa
1,2
Loss of evaporation per day, l/day
56
Period of non drainage storage at pressure increasing
53,6
from 0,11MPa to 1,2 MPa, in days
Type of heat isolation
Vacuum-screen
Reservoir material: -Vessel
12x18H10T
- Casing
092-14
Discharge method
Reservoir blowing from the side source
Frame contacting with LH2
and GH2
Table 3. Technical characteristics of liquid hydrogen transportation tank

358

General view of LHT can be seen in picture 5, technical characteristic is given in table 3.
REFERENCES
1.

2.

3.
4.
5.
6.
7.

8.

Zashlyapin R.A., Cheremnych O.Ya., The creation of vehicles and stationeries for
liquefied gases storage and transportation, Technical gases - 1, Odessa, Ukraine
(2006) 20-126
Zashlyapin R.A., Cheremnych O.Ya.,Pavlenko S.T., The creation of vehicles and
stationeries for lunar orbital complex filling with liquid hydrogen, Technical gases
- 4, Odessa, Ukraine (2007) 15-19
IMDG
CODE
(International
Maritime
Dangerous
Goods
Code),
St.Petersburg,ZNIIMF (2007) 512
IMDG CODE (International Maritime Dangerous Goods Code) (2004)
RID (International regulations for dangerous goods transportation by railway)
(2005)
ADR (European agreement on international road transportation of dangerous
goods) (2006)
Zashlyapin R.A., Cheremnych O.Ya., Organization and development of effective
production means for LNG multimodal and railway transportation, Technical
gases - 3, Odessa, Ukraine (2006), 32-36
Cheremnych O.Ya., Analysis of transportation peculiarities for LNG export in
tank-containers and technologies of its discharge to reservoir, Technical gases 6
Odessa, Ukraine (2007) 65-68

Picture 1. Tank-container model TC-25/2,0

359

Picture 2. Tank-container model TC-52/1,8

Picture 3. Road tank carrier with TCM 35/0,6, at cryogenic complex LNG, Lingcheping, Sweden

360

Picture 4. LNG discharge from TCM -35/0,6 vessel to stationary reservoir.

Picture 5. General view of tank-container for liquid hydrogen.

361

CR08-28
THE INCREASE OF EFFICIENCY AND SAFETY OF LIQUID
HYDROGEN TRANSPORTATION.
Cheremnych O.Ya.
JSC UralCryoMash, Russian Federation, Sverdlovsk region, Nizhny Tagil
ABSTRACT
Efficiency and safety criteria of vehicles for liquid hydrogen transportation are analyzed:
rate of evaporation of cryogenic product, emergency situation in the process of
transportation decapsulation of heatisolating cavity of a tankage with liquid hydrogen.
INTRODUCTION
Liquid hydrogen has been widely used in different brunches of industry: aircraft-space
techniques, aviation, power engineering [1, 2]. Liquid hydrogen delivery from cryogenic
manufacture to consumer is carried by tank wagons, truck tanks and tank-containers for
multimodal transportations [3].
More demands for liquid hydrogen transportation are made by Regulations for dangerous
goods transportation . As there exist a constant risk of explosion at hydrogen vapor
dumping from the vessel, and risk of emergency situation at tanker breakdown at humps of
a consist.
Thats why efficiency and safety criteria of liquid hydrogen transportation are considered
by the example of a tank wagon model HTC-100M.
1.

TANK WAGON FOR LIQUID HYDROGEN TRANSPORTATION.

Overview and schematic circuit of a tank wagon for LH transportation are given in picture
1; technical characteristic is given in table 1.
Parameters
Values
Geometrical volume, m
119
Hydrogen mass in the vessel, kg
7350
Working pressure in the vessel at 0,25
transportation, MPa
Empty weight, t
77
Overall dimensions GOST 9238-83
1T
Length at coupler pulling faces, mm
25730
Axle loading, t
21,2
Heatisolation
Screen-powder-vacuum
Loss of liquid at evaporation, % per day Not more than 0,8
Liquid hydrogen discharge
Side boosting
Time of nondrainage transportation, 12
days
Table 1. Technical characteristics of tank wagon model HTC-100M

The peculiarity of tank container construction is in its fixing devices which must bear
striking dynamic loads up to 12 g; this is achieved by fixing the vessel on sectors supports.

363

The main criteria of effectiveness of LHT is the rate of hydrogen (loss) evaporation in the
process of transportation. It is this criteria that determines: transportation time without
hydrogen gas release from the vessel and cryogenic product quality assurance.
In order to decrease the rate of evaporation of liquid hydrogen and also the components
contamination by admixtures (nitrogen, oxygen) characteristics of different types of
isolation were analyzed (layer-vacuum, powder-vacuum, layer-powder-vacuum);
experiments were carried out on samples of the tank model LHT-100M.
In layer-vacuum isolation metallized film and glass fiber were used as gasket material. The
quantity of screens in heatisolating place was 30, 20, 10; the number of layers of layervacuum isolation 60, 90, 120.
Aerogel was used as heatisolating powder ( in comparison with perlite it doesnt shrink).
The quantity of aerogel was changing from 5,5 t.- 100% filling, 2,75 t 50 % filling and
1,5 t aerogel filling only strength elements parts (bearings, shafts, chains)
The results of analysis are given in table 2.
Type of heatisolation

Vacuum value in the vessel

(mm merc. column)

Day loss, % per

day

Powder-vacuum

510-3

1,5

Layer-vacuum (90 screens)

510-4

1,4

510-3

1,2

110-3

1,01,1

Layer-powder-vacuum, at
aerogel mass:
-

5,5 t (100%)

2,75 t (50%)

1,5 t (shaft and

strength elements
level)

810-5

Table 2. Rate of LH evaporation in LHT-100M depending on type of isolation.

Tank container testing in powder-vacuum isolation determined the rate of hydrogen

evaporation at level 1,5% per day . It is explained by insufficient width of walls-between
clearance , filled with heatisolating powder, increasing of which can be done only by
decreasing of vessel diameter and this in its turn leads to transported hydrogen mass
decreasing.
The results of testing with layer-vacuum isolation showed the rate of hydrogen evaporation
at level 1,4 % per day. That is explained by low effectiveness of layer-vacuum isolation in
supporting area and chains in lower part of the vessel. Analysis allowed to determine the
optimal number of layers of layer-vacuum isolation (equals to 60), level of aerogel filling

364

by mass 1,5t, vacuum value in cavity 810-5 mm merc column, that ensured hydrogen rate of
evaporation at the level 0,8% per day.
RAILWAY TANK TESTING IN EMERGENCY CONDITIONS (LHT-100M)
LHT overview and its airhydraulics scheme are given in photo 1. In tank scheme safety
devices are shown.
In operating conditions of HTC-100M real danger of casing crippling could arise (for
example, because of its breakdown at railway consist reforming). Liquid hydrogen
transportation was carried out by five cistern cars with attendant personnel. Total amount of
transported liquid hydrogen 36,7t. At such mass consequences of vapor loss in heat
isolating space of tankage could be unpredictable.
Thats why it was decided to carry out an experiment in the testing area (photo 2). The
experiment was carried out on a real tank LHT-100M with screen-vacuum heatisolation vacuum loss imitation through a vacuum valve set on tankage casing (photo 3).
The main testing object was to determine operational reliability of safety devices in case of
vacuum loss in heatisolation cavity because of casing leakage. While testing it was
necessary to determine:
- Risk value at emergency situation as a result of decapsulation
Rate of pressure increasing in the vessel with liquid hydrogen
- Propriety of safety valve throat calculation
For imitation of emergency situation caused by mechanical failure of casing, pneumatic
valve 8 was open (photo 1, a). After the valve opening, air from the atmosphere started
intensively coming to the heatisolation cavity. The noise from absorbing air could be heard
during three hours at a distance of 300m. Moreover some cracks, strong and weak strokes
could be heard. This noise was because of separate tank parts destruction. The inferior limit
of operating temperature of casing (steel 092) was 70C, and the lowest casing
temperature during the experiment was 74 K.
The first acute sound was heard after 22 minutes, and the strongest noise could be heard in
the period 2h 2h 34 m after the beginning of testing. Pressure increasing in the vessel
caused the disk bursting and safety valve response (photo 4).
Full evaporation of liquid hydrogen evoked by decapsulation of heatisolation cavity
occurred during 6h 10 m. In an hour after the experiment beginning the casing surface (in
bearing area) started frosting, then moisture from the atmosphere also started frosting on it
(picture 5). In 2 hours 30 minutes moisture frosting could be seen along the full length of
the casing.
At tank control examination after the experiment completion the following was revealed:
-

along the full length of the casing cracks of different patterns and length were
formed (0,2 -1,5 m);
on a joint weld (head sticking to sidewall at the bottom part) there appeared a
crack on a length 0,025 m;
in the middle part of a center sill at the upper part of the H platform there also
appeared a crack.

365

All that arouse because of liquid air accumulation in the bottom part of heatisolating cavity.
It was here that cracks appeared. Some condensate flew through the cracks to the center
sill, hence its destruction occurred.
During the experiment safety valve ensured the full release of increasing pressure in the
vessel. At safety valve opening there was an acute clap with hydrogen vapor jet appearance
(length about 8 meters). During the experiment fire risk situation did not occur. After the
experiment some recovery works were done, so the tank could go to UralCryoMash at its
own pace.
CONCLUSION
The results of experiments showed that even at emergency situation safety conditions of
tank container LHT-100M can be ensured as well as to prevent negative influence on
environment. That allowed to use the results, obtained during the unique experiment, in the
process of new generation transport units creating.
REFERENCES.
1.
2.
3.

JSC UralCryoMash Little land of Vagonka, SV 96, Ekaterinburg, Russia

(2004) 208
Shuttle space system Energy-Buran, SMF OVM-LUCH, Moscow, Russia
(2004) 356
Zashlyapin R.A., Cheremnych O.Ya., Pavlenko S.T., The increase of efficiency
and safety of liquid hydrogen transportation at railway and multimodal
transportation, Technical gases - 6, Odessa, Ukraine (2007) 57-60
APPENDIX

a)

366

b)

Picture 1. Railway container for liquid hydrogen transportation, model LHT-100M: a general view; b schematic circuit: 1 vessel; 2 casing; 3 vacuum valve; 4
membranous safety device; 5 filling-discharge line; 6 side boost line; 7 gas release; 8
safety valve; 9 gage board; 10 control pane; 11- preheater; 12 vapor drainage line
in transit; 13 end elements blowing by nitrogen; 14 fire fighting system; 15 balloons
for nitrogen.

Picture 2. Railway container LHT 100M on experimental stand of testing area.

367

Picture 3.Valve block for sudden vacuum loss imitation.

Picture 4. Hydrogen vapor dumping to the atmosphere through the safety devices at vacuum loss in
heat isolating space of the tankage.

368

Picture 5. Casing frosting in the process of vacuum loss in heat isolating space of the
tankage

369

CR08-29
THE CREATION OF VAPOR COOLING DEVICES FOR LIQUID
OXYGEN IN STATIONARY RESERVOIRS USING LIQUID
NITROGEN AS A COOLING REAGENT.
Cheremnych O.Ya., Korneva I.I.
Uralcryomash, Sverdlovsk region, Nizhny Tagil, Russia
ABSTRACT
Cooled liquid oxygen (non boiling) is used as a cryogenic component of fuel in space
apparatus. Its got lower temperature than its boiling temperature at atmospheric pressure,
hence its density is high. Thus it allows to increase liquid oxygen reserves in a space
apparatus tank. Thats to say that the equal capacity tank is filled with more product mass
and product losses at storage and fillings are substantially decreased.
Liquid oxygen cooling in stationary conditions is carried out in both ways: stationary
reservoirs and special devices (heatexchanging apparatus, cooling reservoirs).
The results of analysis of liquid oxygen cooling conditions and storage are given. In this
case liquid nitrogen is used as a cooling reagent. Nitrogen is cooled by means of vapor
space vacuuming on liquid surface with the help of ejectors.
The constructive solution of tank-reservoirs for cooled liquid oxygen is given.
INTRODUCTION
Low density is considered to be a grave disadvantage of some fuel cryogenic components
(such as liquid hydrogen, LNG) as it leads to tank capacity increasing. Another
disadvantage is their low heat per unit of volume. Moreover there is a necessity in
compensation of losses as a result of component evaporation, maintenance of the given
product quantity in rocket tanks, exclusion of nonequilibrium processes in surface filling
stations.
The characteristics of rocket-carriers, acceleration blocks, and space ships can be improved
by tanks filling with cooled (nonboiling) cryogenic products, which have lower temperature
than their boiling temperature at atmospheric pressure and hence higher density. Thus it
allows to increase fuel components recourses on board the rocket, i.e. the tank with equal
capacity is filled with more product mass and its losses at storage and filling are
considerably decreased. Moreover the use of cooled product ensures a single-phase liquid
flow, some size decreasing, so the mass of pipelines and fittings that is particularly vital for
aircrafts. Cooled liquid oxygen is more applied in space techniques.
Cryogenic liquids cooling in a start complex can be carried out both in filling systems
reservoirs and in the process of rocket filling, acceleration block and a space ship in special
facilities (heat exchanging apparatus, reservoirs-coolers), installed in a filling system [1].
For products cooling in tank-reservoirs cooling devices with relatively little productivity
can be created as cooling is carried out in nontechnological time; it can be extended in time
and not connected with technological preparation of rocket for start.

371

When cryogenic liquid is being cooled in the process of filling the productivity of cooling
devices is determined by the rate of filling; in this case the equipment is some more
complicated.
1.1 COOLING METHODS
Basic cooling methods.
With the help of cooling machine devices. The process is effective from the point of view
of thermodynamics. Any temperature level can be reached. But the use of complicated
machinery is its disadvantage.
With lower temperature products assisted. In this case refrigerating fluids are:
1.Cryogenic components, having boiling temperature at atmospheric pressure lower than
that of necessary for fuel components cooling.
2.Cryogenic components, whose temperature in the reservoir with heatexchanger is carried
to the required one by vacuuming of vapor space with the help of different devices.
The distinctive feature of these cooling methods is the loss of refrigerating fluid,
evaporating in the process of fuel cooling. For collecting of refrigerating fluid vapors and
their back condensation complicated additional equipment is required; in conditions of a
start position it is considered to be inexpediently.
Due to liquid evaporation - by vacuuming of vessels vapor space or by barbotage through
the liquid of low soluble noncondensable gas (usually helium). These methods allow
reaching the products temperature up to the triple point. Moreover with their help an icelike condition of some products can be got.
When products are cooled by barbotage practically helium is only used because of its
safety, little solubility and low condensation temperature. This method is comparatively
expensive. Under its realization at an open-ended scheme large quantity of expensive and
deficit helium is required; when realizing at a closed scheme complicated system of helium
refinement from cooled liquid vapors is required [2].
Cooling method by vapor space vacuuming above liquid surface is more applied. Cooling
by vacuuming is a universal method of cryogenic products cooling; it doesnt require
complicated equipment, big capital and operating expenses.
1.2 THE USE OF NITROGEN AS REFRIGERATING FLUID IN THE PROCESS
OF OXYGEN COOLING
Having analysed all advantages and disadvantages of different cooling methods in
conditions of a start position such method as vapor space vacuuming above liquid surface
of a side refrigerating fluid (nitrogen) was chosen.

372

Figure 1: Scheme of oxygen cooling by vapor space vacuuming above liquid surface of a side
refrigerating fluid (nitrogen).

1 liquid oxygen, 2- liquid nitrogen, 3 ejectors block

1.3 COOLING OPERATING FACTORS DETERMINATION
1.Pressure determination at a pumping block absorbing (by nonlinear equation)

Pn1(T) = Gn(Pn) -

((PN2 (T) 10 6 ) 2 ( Pn 10 6 ) 2 ) 2 N2 (T ) 0.616

PN2 (T) 10 6 1.5 (A1 + A2)

Where
Pn1 (T) pressure at a pumping block absorbing, MPa;
Gn(Pn) - pumping consumption mass depending on pressure, kg/s;
PN2(T) nitrogen saturated vapors pressure in the vessel depending on temperature, MPa;
Pn pressure at a pumping block, MPa;

N2 (T ) density of gaseous nitrogen depending on temperature, kg/m3;

A1 local resistance equal to pumping line;
A2 - local resistance equal to pumping block;

373

Figure 2: Diagram of pumping consumption mass depending on temperature

2. Determination of pumping block heat rate

Q(T) = Gn(Pn) .rN2(T)
Where
Q(T) - pumping block heat rate depending on temperature, J/s;
Gn(Pn1) pumping consumption mass depending on pressure at pumping block absorbing,
kg/s;
rN2(T) - latent heat of nitrogen vaporization depending on temperature kJ/kg;

Figure 3: Diagram of pumping block heat rate depending on temperature

3. Heat leakage determination from environment

Qinp(T) =

Qinp (303 T )
303 70

Where
Qinp(T) heat leakage to a product depending on temperature, kJ;
374

Qinp - rated heat leakage to a product, K;

4. Product cooling time determination
Solution of differential equations

Qinp(T) Q(T)
dT
=
d

MwCp1N2 (T) 103 + Mo Cp1O2 (T) 103 + Mm Cp m (T) 103

dMw = Gn(T)
d
Where
T- current product temperature, K;
- current time of product storage, sec;
Q(T) - pumping block heat rate depending on temperature, J/s;
Qinp(T) heat leakage to a product depending on product temperature, kJ;
Mw current nitrogen mass, kg;
Gn(T) mass consumption of nitrogen depending on product temperature, kg/s;
Cp1N2(T) liquid nitrogen heat depending on temperature, J/kg*K;
CpO2(T) - liquid oxygen heat depending on temperature, J/kg*K;
MO oxygen mass;
Mm metal mass (vessel mass in casing);

Figure 4: Diagram of product temperature depending on cooling time.

375

5. Plotting of product T temperature dependence on cooling time.

Figure 5: Diagram of nitrogen mass depending on cooling time.

CONCLUSION
This type of cooling allows getting oxygen temperature (at a given rating of electors block)
equal to 70K. The main advantages of such cooling method are: the ability of preparation of
oxygen beforehand at nontechnological time, no filling fuel components losses, simplicity
of equipment, efficiency.
REFERENCES
1. Filin N.V., Liquid cryogenic systems, Mechanical engineering, Leningrad, Russia (1985)
247-249
2. Arkharov A.M., Kunis I.D., Filling cryogenic systems of
MSTU by Bauman, Moscow, Russia (2006) 252-260

rocket-space complexes,

3. Sychyov V.V., Vasserman A.A., Thermodynamic properties of nitrogen, Standards

publishing house, Moscow, Russia (1977)
4. Sychyov V.V., Vasserman A.A., Thermodynamic properties of oxygen, Standards
publishing house, Moscow, Russia (1977)

376

AUTHOR INDEX
Aigouy G.
59
Alava L.A.
227
Albert S.
271
Anghel A.
109
Arkharov I.
175
Arpentinier P.
167
Bae D.K.
293
Bae J.H.
293
Baker R.A.
99
Barucci M.
277
Bell C.
153
Berdais K.-H.
53
Binneberg A.
271, 317
Blau B.
109
Bondarenko V. L.
161, 329
Bornea A.
341
Bracanovic D.
143
Brojek W.
257
Caillaud A.
59
Cheremnych O.Ya. 355, 363, 371
Chorowski M.
117
Chrz V.
45, 185, 193, 201
Coquelet C.
167
Coulomb D.
27
Crispel S.
59
Dauguet P.
59
Daum M.
109
Delcayre F.
59
Delcorso F.
167
Diachenko O.V.
335
Diachenko .V.
335
Dobrozemsky R.
311
Dvok J.
245
Esteves A.D.S.
213
Fortkov L.
257
Fydrych J.
117
Gherghinescu S.
265, 349
Good J.
143
Grabi V.
59
Grigoriev S.
109

Gschwendtner M.A.
Haberstroh Ch.
Hannani S.K.
Hanzelka P.
Heidrich R.
Herzog R.
Hirschl C.
Hnzdil T.
Horynov A.
Houssin-Agbomson D.
Jafarian A.
Joonhan B.
Kaiser G.
Kaiser Z.
Kalbassi M.A.
Kideok S.
Kirch K.
Klepal J.
Klier J.
Klingner E.
Korneva I.I.
Kouba M.
Krlk T.
Kundera R.
Laa C.
Lain M.
Lnsk M.
Lee D.-Y.
Lee S.-J.
Losyakov N. P.
Martelli V.
Mtl P.
Mika P.
Muehlegger M.
Musilov V.
Myunghwan. S.
Navasardyan E.
Parise J.A.R.
Park J.-S.
Pearsica C.
377

91
67
77
133
271
317
311
185
237
167
77
285
271
45
153
285
109
223
271, 317
249
371
45, 185
133
45
311
245
193
293
293
161
277
193
237, 245
53
133
285
175
213
293
323, 341

Popescu G.
Preda A.
Pruk J.
Ribeiro Gomes M.
Richon D.
Risegari L.
Saidi M.H.
Sarikhani N.
Schmidt J.
Schumann B.
Schustr P.
Scurlock R.G.
Seokho K.
Simonenko O. Yu.
Simonenko Yu. M.
Sosnowski J.
Sprl G.
Srnka A.
Stefan L.
Stipsitz J.
Stoek P.
Strakov H.
Strnad P.
Strychalski M.
Sullivan P.
Suma J.
Takcs D.
Tucker A.S.
Ungricht Th.
Vasut F.
Ventura G.
Waldie B.
Wang C.
White V.
Wilhelm H.
Zashlyapin R.A.
afrata S.

349
323, 341
45
125
167
277
77
77
271
317
45
37, 85
285
161
329
301
249, 317
133
323
311
223
237
257
117
99
185
201
91
53
323, 341
277
153
85
153
53
355
45

378