Journal of Energy

and Power Engineering

Volume 12, Number 1, January 2018 (Serial Number 122)

David Publishing

David Publishing Company

www.davidpublisher.com
Publication Information:
Journal of Energy and Power Engineering is published monthly in hard copy (ISSN 1934-8975) and online (ISSN
1934-8983) by David Publishing Company located at 616 Corporate Way, Suite 2-4876, Valley Cottage, NY 10989, USA.

Aims and Scope:

Journal of Energy and Power Engineering, a monthly professional academic journal, covers all sorts of researches on
Thermal Science, Fluid Mechanics, Energy and Environment, Power System and Automation, Power Electronic, High
Voltage and Pulse Power, Sustainable Energy as well as other energy issues.

Editorial Board Members:

Prof. Ramesh K. Agarwal (USA), Prof. Hussain H. Al-Kayiem (Malaysia), Prof. Zohrab Melikyan (Armenia), Prof.
Pinakeswar Mahanta (India), Prof. Carlos J. Renedo Estébane (Spain), Prof. Mohamed Ahmed Hassan El-Sayed (Trinidad
and Tobago), Prof. Carlos Redondo Gil (Spain), Prof. Roberto Cesar Betini (Brazil), Prof. Rosário Calado (Portugal), Prof.
Dr. Ali Hamzeh (Germany), Prof. Noor-E-Alam Ahmed (Australia), Prof. E. Ubong (USA), Prof. Shripad T. Revankar
(USA), Prof. Almoataz Youssef Abdelaziz (Egypt), Prof. Guihua Tang (China), Prof. Mohammad Rasul (Australia), Prof.
Rene Wamkeue (Canada), Prof. Ya-Ling He (China), Prof. Filippo de Monte (Italy), Prof. Masoud Rokni (Denmark), Prof.
Hosni I. Abu-Mulaweh (USA), Prof. Quan Zhang (China), Prof. Peng-Sheng Wei (Taiwan), Prof. Vinod Kumar (India), Prof.
Yuan-Kang Wu (Taiwan), Dr. Kaige Wang (USA), Dr. Fude Liu (Hong Kong), Prof. Isa Salman Hasan Qamber (Bahrain),
Dr. GM Shafiullah (Australia), Dr. Kelum Gamage (UK), Dr Isam Janajreh (UAE), Prof. Marco Casini (Italy), Prof.
Meicheng Li (China).

Manuscripts and correspondence are invited for publication. You can submit your papers via Web Submission, or E-mail to
energy@davidpublishing.com or info@davidpublishing.com. Submission guidelines and Web Submission system are
available at www.davidpublisher.com.

Copyright©2018 by David Publishing Company and individual contributors. All rights reserved. David Publishing
Company holds the exclusive copyright of all the contents of this journal. In accordance with the international convention,
no part of this journal may be reproduced or transmitted by any media or publishing organs (including various websites)
without the written permission of the copyright holder. Otherwise, any conduct would be considered as the violation of the
copyright. The contents of this journal are available for any citation. However, all the citations should be clearly indicated
with the title of this journal, serial number and the name of the author.

Abstracted / Indexed in:

Database of Cambridge Science Abstracts (CSA), USA
Chinese Database of CEPS, American Federal Computer Library Center (OCLC), USA
Chinese Scientific Journals Database, VIP Corporation, Chongqing, China
CNKI(China National Knowledge Infrastructure), China
Ulrich’s Periodicals Directory
Summon Serials Solutions
Chemical Abstracts Service (CAS)
Google Scholar
Subscription Information:
Price: Print $640 (per year), Online $600 (per year), Print and Online $880 (per year).

Editorial Office:
616 Corporate Way, Suite 2-4876, Valley Cottage, NY 10989, USA.
Tel: 1-323-984-7526, 001-323-410-1082; Fax: 1-323-984-7374; 323-908-0457
E-mail: energy@davidpublishing.com; energy@davidpublishing.org; energy-power@hotmail.com;
energy@davidpublishing.net; order@davidpublishing.com
Digital Cooperative Company: www.bookan.com.cn

D DAVID PUBLISHING

David Publishing Company

www.davidpublisher.com
 Journal of Energy and
Power Engineering
Volume 12, Number 1, January 2018 (Serial Number 122)

Contents
Clean and Sustainable Energy

1 Wave Power Convereter Pendulor with Hybrid H.S.T.

Tomiji Watabe and S. D. G. S. P. Gunawardane

11 An Investigation to Assess Storage Stability of Pomelo Seed Oil Biodiesel

Madhurjya Saikia, Dilip K. Bora and Kalyan Kalita

16 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor

Sergio Beneitez, Juana L. Gervasoni and Juan Carlos Furnari

Power and Electronic System

26 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions
Shoji Kawasaki and Masahiro Ise

35 Evaluation of Height of Tsunami Induced by Submarine Landslide

Yoshikane Murakami, Takemi Shikata and Koji Tonomo

44 Characterization of Photovoltaic Modules under Arid Environments

Abubaker Younis, Esam Elsarrag, Yousef Alhorr and Mahmoud Onsa
Journal of Energy and Power Engineering 12 (2018) 1-10
doi: 10.17265/1934-8975/2018.01.001
D DAVID PUBLISHING

Wave Power Converter Pendulor with Hybrid H.S.T.

Tomiji Watabe1 and S. D. G. S. P. Gunawardane2

1. T-Wave Consultant, 5-23-3, Misono, Noboribetsu 059-003, Japan
2. Dept. of Mechanical Engineering, Faculty of Engineering, University of Peradeniya, Peradeniya 20400, Sri Lanka

Abstract: This Pendulor, wave power converter, was invented to be robust towards storms. The key is Hybrid H.S.T. for generator
driving by a higher the speed and a smaller the torque of the piston pump which can do over 360 degrees rotation free of the
pendulum motion. This idea will bring the non-shock operation to the moving body type wave power conversion and when the
Pendulor applies with the antenna principle, it will have a possibility to convert safe and cheap electricity from the ocean by this
technology.

Key words: Ocean wave power, Pendulor, H.S.T., hybrid, large plant.

1. Introduction the surface so that there appears periodical moving of

the water (ripples); birth of the waves. The growth of
Muroran Institute of Technology Japan developed
the sea waves depends on the wind power, lasting the
the wave energy converter Pendulor 39 years ago. The
blowing time and along to a place follows to the wind
research advanced well at the beginning but faced a
pass away from beginning and lasting to the final
fundamental problem on the survivability towards
place. It tells that the waves are phenomenon altered
storms. Since the original Pendulor depends on
from the wind energy, and it propagates from the birth
moving body type energy conversion, the system must
place, as it to be travelling waves. They are classified
have a stopper for the pendulum to keep the stroke
themselves within near sized waves together while
within a limit, in any type of operation of energy
travelling through such a long distance. Therefore,
conversion [1, 2]. The stopper should be a sockless for
utilization of the wave energy, applied with the antenna
the stopping of the massive Pendulum. If we cannot
principle, it can be such a long distance (Muroran
pass this situation, the dream to make electricity from
Institute Technology took the principle from the
the ocean waves would be impossible. This new
beginning stage [4]). For development of the large
Pendulor has changed the system completely safe with
Pendulor, driving it with energy rich waves was
no stopper [3]. The Pendulum can rotate over 360o
studied from a basic view point. The reason is that the
both directions freely by applying with the new hybrid
Pendulor must do resonant operation with the incident
H.S.T. for generator driving. Plural number of piston
waves and the generator load condition must be adjusted
pumps rotates increased speed by a step-up gear to
in coincide with the impedance of the Pendulor device
deliver much oil with all. The system shall be able to
for the optimal driving. Therefore, the Pendulor study
be applied for a large Pendulor to improve the cost
cannot do anything without understanding on the
saving (three patents pending).
incident waves for optimization of the wave power
2. Birth of the Ocean Waves and Use of Them conversion. Fig. 1 [5, 6] is the first floating type
Pendulor which has been studied by Korea after
Blowing wind on the ocean excites the sea water at
Muroran I. T. closed the study in 2000. The device of
floater has applied new idea. (1) Floater is stable with
Corresponding author: Tomiji Watabe, Dr., consulting
engineer, former professor, research field: ocean wave energy. a damping action by wave force in three directions.
2 Wave Power Converter Pendulor with Hybrid H.S.T.
Fig. 1 Floating Pendulor 300 kW, developed by, KRISO & YOOWON, Korea.
(2) Optimization control of the wave power
conversion. (3) Giant Rotary Vane pump direct
coupled with the Pendulum. (4) Oil seal fits to the
large pump. Japan created the soft of the System and
Korea built the Power Plant of Pendulor [7].
3. Wave Power Density of the World
Fig. 2 shows the wave energy distribution,
investigated by US Dept. of ENERGY by kW/m. The
department estimated of the world wave energy
potential; that is at 2 or 3 million MW. In favorable
location, the wave energy can average 40 MW/km (=
40 kW/m of coastline) [8]. Comparing with the power
density of Japan and the world, Japanese one is only
1/3 of the world. Nevertheless, the giant level typhoon
frequently attacks Japan with strong power of storm of
10 times or more of the normal wave power. The
Pendulor must survive towards them with not such the
strongest machine but a clever and inexpensive one.
The machine lets the dangerous waves go back to the
sea as they do nothing harmful to the Pendulor.
4. Wave Power for Large Pendulor
Considering the magnitude of the wave energy
potential on the Earth, we must pay attention to the
big energy converter driven by high energy density
waves. Here introduces an interesting study on the
waves which are grown by several wind conditions.
Since the glowing wave height and its period can be
calculated when the wind speed and the fetch length
are given, applying with the data shown in the chart
(Fig. 3) [9]. The authors tried some investigations to
find preferable waves which are good for driving the
large Pendulor. Figs. 3 and 4 show the wave height:
H1/3 (m) and its period: T1/3 (s) at given wind speed
(m/s), blowing duration (hour) and distance: (Fetch
length, km). In case of 10-hour wind blow with
several wind speeds, glowing wave heights estimated
are shown in Figs. 4 and 5 [9]. The wave height: H1/3
increases as wind speed being higher. For the wave
period: T1/3, it becomes longer (by slower wave motion)
excited with the higher blowing wind. It means the
wave motion changes slower as the blowing wind to
be stronger. Therefore, when a large Pendulor is
driven by energy rich waves, the richer the energy
density, the operating condition of the Pendulor
becomes the slower speed and the bigger torque
operation. This result tells us an important direction
 Wave Power Converter Pendulor with Hybrid H.S.T. 3

Fig. 2 Wave power distributions in the world with the unit kW/m (US Dept. of ENERGY).

Fig. 3 Wind-wave forecasting curves of deep water.

4 Wave Power Converter Pendulor with Hybrid H.S.T.

Fig. 4 Period change by growth of wave.

Fig. 5 Wave characteristics relation between blowing wind on the sea and glowing waves.
 Wave Power Converter Pendulor with Hybrid H.S.T.
to challenge to the wave energy utilization with a large
converter, the system must overcome the slower speed
and the greater torque load operation. The authors
decided to attach to this study with new ideas: shown
here. New ideas for the large Pendulor: (1) Survival to
the Storm by no-stopper ± 360° free rotation, (2)
Invention of, hybrid H.S.T. for Generator driving, and
(3) piston pump can work at the condition, well.
5. Experience on System Survivability
Since the wave energy conversion depends on the
reciprocal wave motion, the action parts by waves
must be limited even when the device encounters
storms. The wave power of storm would be over 10
times greater level than the normal, the shock by
impingement to the stopper often offers a damage
situation to the system. During sea operation of a
Pendulor, we experienced three times of accidental
breaking of the flapper legs by impingement between
the stopper and the legs [10]. The happening gives us
a hint to avoid the trouble; by not improving the legs
stronger but change the condition safer; by no use of
the stopper. We selected the hybrid H.S.T. system for
the Pendulor which requires no stepper as shown in
Fig. 6 [11]. In Fig. 6, a single Pendulum drives four
sets of geared piston pumps of the HST which drives a
generator. Since the gear is to be speed up use, it
drives the four pumps with increased speed with
divided power of 1/4 each, so, the displacement of the
pump unit becomes the value shown in Eq. (1).
/ (1)
where, Do: displacement of the pump when it is driven
directly. In a case of 4 sets of pumps with teeth
number of the pinion: zp = 7 and teeth number of the
gear: zg = 70, Dp takes the value shown by the Eq. (2),
that is the case of Fig. 6.
1/4 7/70 0.025 (2)
This result means that the system of Fig. 6 can
reduce the pump capacity required drastically,
comparing with the former system of one set pump.
5
The hybrid H.S.T. of Fig. 6 has no stopper because of
the system permits over 360° rotation of the pendulum
in either direction. The pendulum motion amplitude:
θa is shown by Eq. (3) as a function of the incident
wave height: H [1].
(3)
4 sinh 2
sinh2 2 (4)
sinh cosh 1 (5)
where, : wave number, H: wave height and h: water
depth, l: distance between the center of pendulum
shaft and the water surface.
Piston pump of Fig. 6 is non rotating type. Its
feature is much simpler than the rotary type piston
pump and being strong towards the cavitation and
dynamic load because of the tiny moment of inertia.
The pump can be inherently fit to the reciprocal
rotation as well.
6. Improvement of the Survivability of the
Pendulor
The cause of damage on the wave energy converter
observed was by shock load most, not only on the
Pendulor but also the Pelamis [11] (Fig. 7) and, the
Oyster [12] (Fig. 8). The shock load worked likely
both places with the stoppers and the mechanical
power-transmission where small gaps existed between
the parts faced with two of them which cannot be
combined into one part. The gaps make shocking
impingement. In case of impingement at the stopper,
there relates often with big force by collision.
Therefore, some components of the stopper are
damaged frequently. For instance, the tightening bolts
of the stopper to be loosened at even primary
condition then the all bolts were broken away to the
final stage. The incident lets us make the mind, to
improve the power take off components to be
improved for being reliable much more. The decision
is, to exchange the design principle with no use of the
stopper. This idea has been realized by the hybrid
H.S.T. developed shown in Fig. 6.
6 Wave Power Converter Pendulor with Hybrid H.S.T.

Fig. 6 Principle of the pendulum combined with the Hybrid H.S.T. (Patent app. 2017-01 5068).

Fig. 7 Typical W. E. converter: Pelamis UK.

7. Design of the Hybrid H.S.T. In 16 pumps of the displacement Dp each is shown by

Fig. 9 is a trial design with 16 pumps to make a set
of geared pumps for the hybrid H.S.T. This idea is for
an MW class study on the large Pendulor with 16
pumps incorporated under the speed ratio: zg/zp =
120/9 = 13.33. It seems one of the balanced features
from the engineering and the economic stand point.
The case shown in Fig. 9, the pump speed is increased
in 13.33 times higher than the case of no gear driving.
Eq. (6).
1/16 6/120 1/320 (6)
Therefore, the case of Fig. 9, the 16 pumps have the
320 times small displacement though, it is enough for
the pumps of the Hybrid H.S.T. Despite of such as 16
pumps drive in parallel, each the load is divided into
each the pump exactly by the error absorption effect
into the teeth pitch error of the pumps. For design on
the 7 teeth pinion of the Hybrid H.S.T, there is no fear
 Wave Power Converter Pendulor with Hybrid H.S.T. 7

Fig. 8 Inverse pendulum converter Oyster UK uses water power transmission.

(a)

(b)
Fig. 9 (a) Geared piston pump set for 1 MW class (Patent App. No. 2014-173194, 2014, T-Wave); (b) Geared pump set
shown inside.
8 Wave Power Converter Pendulor with Hybrid H.S.T.
of appearance of an under cutting trouble by the
interference towards the 7 teeth form pinion. That is
an allowable limit condition with the sharpened teeth
tips (the larger the speed ratio: zg/zp, the hybrid H.S.T.
can be made smaller mass and dimension).
For the gear lubrication of the condition (high tooth
load Fig. 9b slow speed operation) is not easy for
safety operation, so that some attentions written below
(1)-(3) are useful:
(1) The teeth are treated with the hardening by
nitride process, then polished by lapping finish to
make 0.5 μm flatness on the contact teeth surfaces.
The error of the tooth profile and the pitch are not
important in case of slow rotation (Under 100 rpm).
(2) For the teeth strength, especially to the contact
pressure, we should pay attention carefully (oil film
thickness at the contact point etc.).
(3) For the gear design, the contact points motion
by a pure rolling each other teeth as perfect as possible.
This kind of care makes steady lubrication by EHL
(Elastic & Hydro-Dynamic Lubrication). This Hybrid
H.S.T. idea can be used also for Wind Turbine. The
reason is, the case of wind turbine drives high speed
generator with very slow speed turbine [11].
8. Piston Pump
Fig. 10 is the pump designed for the Hybrid H.S.T.
for study purpose of a large Pendulor. It belongs to
non-rotating axial piston pump group of five pistons
in a pump considering the abnormal slow speed
operation. The pump can be used as a variable type or
fixed one when the suction check valves are kept in
open or in close. In case of the large Pendulor (and Fig.
9 is the Hybrid pump for it [2]), the displacement Dp
required for the Hybrid piston pump is shown by Eq.
(6).
9. Study on the Piston and the Shoe
The large Pendulor with the hybrid H.S.T. faces a
situation that tells us the Wave Plant would be instable
more if the piston assembly is exposed under an
abnormal slow speed and a high load operation with
the conventional type slipper. Therefore, we decided
to make the Hybrid H.S.T. be stable better with the
new idea to improve the static bearing for the axial
piston slipper (Patent Pending).
When a large displacement (DP = 100 lit/rev) pump
is running with speed: np = 100 rpm, it delivers oil,
Qth = Dp × np = 10,000 l/m (7)
where, Qth is the theoretical oil delivery. Qth is
proportional to speed: np, and the actual flow: Qp is
shown by Eq. (8),
Qp = Qth Δq (8)
where, Δq: leak flow.
Leak flow Δq is proportional to the delivery
pressure: p and time: t of the pump operation. Since
the clearance located between pressure difference in
the pump where being a path of the leakage, the
passing area would be proportional to the pump
dimension.
The author experienced an observation of the swash
plate type piston machine which happened metal contact
whenever starting. The friction by the metal contact
accompanied with a fair amount of oil leak which
made the condition worse, too. Fig. 11 is a piston and
a shoe of the axial piston pump/motor invented to
overcome the energy loss by the metal contact at slow
motion. The principle is applied both in the piston and
on the shoe as shown in Fig. 11. On the surface of the
static bearing (piston and shoe), there are some thrust
control pools of which pressure is controlled by the
outer load, irrespective of running speed. The oil
pressure at a place is the function of flow-in and
flow-out oil volume, the external load controls the
pressure of the control pools, and the pool pressure
adjusts the lateral clearance of the bearing pads
non-uniformly. So, it produces the bearing action
change, no metal contact (Patent Pending). The thrust
pools locate on the piston surface, too. The pool
pressure is controlled when the load balance approves
the lateral load and the reaction relationship caused
by the oil pressure. Fig. 11 shows that the piston
 Wave Power Converter Pendulor with Hybrid H.S.T. 9

Fig. 10 Piston pump for hybrid H.S.T.

Fig. 11 Piston and shoe with self-balancer (Pat. Pending).

Table 1 The wind-waves grown by several wind speeds.

Case No. terms 1 2 3 4
Wave height, H1/3 (m) 3.0 4.5 5.4 6.6
Wave period, T1/3 (s) 7.0 9.0 9.5 10.5
Wind speed (m/s) 15.5 20 22.5 25.0
Duration (H) 10 10 10 10
Minimum fetch length (km) 40 52 58 65
10 Wave Power Converter Pendulor with Hybrid H.S.T.
clearance is controlled to move the piston center.
Therefore, when the pressure is bigger, the lateral
force becomes the bigger. The principle:
The piston configuration is located right to left, ball
formed joint, clown formed seal, pools located on the
taper formed guide. The piston can rotate in the lateral
gap by the tapered guide which affects pressure
change of the pools.
When a normal vector pushes the shoe at the sliding
center shown in Fig. 11, one force vector component
of up direction, drives the piston by anti-clock wise
moment. Therefore, on the taper part of the piston,
upper side gap becomes larger and lower side one
becomes minimum.
Then, this change makes the pressure of upper side
pool the higher and lower side pool the smaller. The
pressure difference between up and down side pushes
the piston down direction which acts force to the
piston with the opposite direction of the load. If the oil
pressure cancels the piston load, the minimum friction
loss would be possible applied with this technology
(Patent Pending).
10. Conclusions
The idea of Pendulor with the hybrid H.S.T. can be
concluded as shown bellow.
(1) The new idea was reported here that the
Pendulor can be survival against the storm with the
hybrid H.S.T. assembled into the Pendulor system.
(2) This idea contributes to the practical wave
energy conversion not only being survival in the storm
but also opening a new possibility to develop MW
class large Pendulor by EHL (Elasto-Hydro Dynamic
Lubrication) technology for the high torque and slow
speed operation.
(3) The axial piston pump with a small moment of
inertia, can improve the Pendulor efficiency at high
torque slow speed operation. This idea on the hybrid
H.S.T. is suitable for the large wind turbine
application, too.
(4) The converter on the moving body type, pelamis
and oyster, the cause of failure was studied to
overcome the problem in the future.
Acknowledgment
Ms. Chiyoko NAGAISHI Japan donated 1.5 million
yen for the research and Saga University Japan
supported the study.
References
[1] Watabe, T. 2008. Utilization of the Ocean Wave Energy.
Fuji Print Press.
[2] Gunawardane, S. D. G. S. P. 2016. Study on the
Pendulor Wave Energy Converter in Array, Configuration,
Energy.
[3] Watabe, T. 2015. “Challenge to Low Cost Wave Energy
Converter with Hybrid H.S.T.” OTEC 20.
[4] Mitsuyasu, H. 1994. Physics of Ocean Waves. IWATANI
Book Co.
[5] Watabe, T. 2010. Floating wave energy converter.
Japanese Patent, No. 4449872.
[6] Watabe, T. 2001. Wave energy converter. Japanese Patent,
No. 3218462.
[7] Watabe, T., Shin, H. S., and Kim, B. K. 2015. “Large Oil
Pump for Wave Energy Conversion.” Fluid Power (in
Japanese) 54 (2): 3-8.
[8] US Army. 1973. Shore Protection Manual, US
Government Printing Office.
[9] Barth, S., and Eecen, P. J. 2006. “Description of the
Relation of Wind, Wave.” OWEZ.
[10] Watabe, T., Kondo, H., and Yano, K. 1986. “Pendulor
Type Wave Energy Converter.” BHRA.
[11] Watabe, T. 2014. Hybrid trans. mitt. Jap. Patent, App No.
2014-173194.
[12] Watabe, T. 2017. Piston and shoe of excellent low friction
under a high torque/slow speed operation, assembled in a
piston pump. Jap. Patent, App No. 2017-108656.
Journal of Energy and Power Engineering 12 (2018) 11-15
doi: 10.17265/1934-8975/2018.01.002
D DAVID PUBLISHING

An Investigation to Assess Storage Stability of Pomelo

Seed Oil Biodiesel

Madhurjya Saikia, Dilip K. Bora and Kalyan Kalita

Mechanical Engineering Department, Assam Engineering College, Guwahati 781001, India

Abstract: Biodiesel is a biomass-based renewable and clean fuel. It can be used directly in existing diesel engines without any
modification. In spite of having so many advantages, it has an issue regarding long-term storage stability. Biodiesel oxidizes when it
comes in proximity of light, temperature, humidity, metals etc. The oxidation causes precipitation and sedimentation in the fuel. Such
biodiesel is unfit to use in engines. Hence, this study aims to assess the storage stability of pomelo seed oil (Citrus maxima) biodiesel
over a period of five months upon storage both in open and closed condition. Storage stability of biodiesel is determined in terms of acid
value, peroxide value and viscosity at regular interval of time. It is observed that there is a strong relationship between oxidation and
outside factors such as light, temperature fluctuations and humidity level. Upon exposure to these factors, oxidation in biodiesel
accelerated greatly.

Key words: Biodiesel, storage stability, acid value, peroxide value, viscosity.

1. Introduction 20].
Hence, quality should be maintained all the time to
Bio-diesel refers to a biomass based fuel having
ensure longevity of engines. Some of the tests based on
short chain alkyl esters, made by transesterification of
which storage stability can be ascertained are peroxide
vegetable oil or animal fat which can be used in
value, acid value and viscosity of the fuel. The present
existing diesel-engines [1-3]. It is considered to be a
study tries to assess storage stability of pomelo seed oil
naturally oxygenated fuel with inherent oxygen in itself
biodiesel (Citrus maxima) over a period of five months
about 9-10% [3, 4]. Recent researches suggest that
both in open and closed condition.
biodiesel is a clean fuel with less contribution to the net
increase of carbon monoxide, carbon dioxide, other 2. Materials and Methods
hydrocarbons and sulphur dioxide [5-7]. Moreover,
The experimental study is carried out in Mechanical
biodiesel from various non-edible oil resources has
Engineering Department Engine Laboratory,
been used successfully without requiring any engine
Dibrugarh University, India.
modification [8-10]. As a fuel, biodiesel is engine
friendly but its storage stability is a concern. When 2.1 Biodiesel Production
biodiesel is stored for a long time, it degrades rapidly in
The vegetable oil from pomelo seeds had been
contact with sunlight, moisture, and humidity. This is a
extracted by using a mechanical expeller. Pomelo seed
common problem faced by all biomass-based liquid
oil is converted to biodiesel via transesterification
fuels [11-19]. Ultimately, oxidized biodiesel becomes
process. Fig. 1 depicts the steps of laboratory scale
unfit to use in modern engines and causes trouble by
biodiesel production steps. Two hundred mL of methyl
clogging fuel injection and multiple another failures
alcohol was mixed with 14.5 gm of KOH (catalyst) to
due to precipitation and sedimentation problem [12, 13,
form a potassium methoxide solution. Then, this
solution of sodium methoxide was mixed with pomelo
Corresponding author: Madhurjya Saikia, research scholar,
ME Deptt., research fields: solid & liquid fuels and biodiesel. oil (1,000 mL). The mixture was held at a temperature
12 An Investigation to Assess Storage Stability of Pomelo Seed Oil Biodiesel

Fig. 1 Biodiesel production process.

Table 1 Physical and chemical properties of pomelo biodiesel samples. The biodiesel samples were stored
biodiesel (Citrus maxima). in sealed glass vessels. One of these glass vessels was
Pomelo placed normally in a room. The other one was placed in
Sl No. Properties Diesel
biodiesel
1 3
Density (kg/m ) 822 834
a wooden cabinet to resemble closed condition. Storage
Kinematic viscosity stability of the samples is assessed in terms of acid
2 1.92 2.6
(cSt)
value, peroxide value and viscosity at an interval of one
3 Pour point (°C) 15 16
4 Cloud point (°C) 4 4 month for five months.
5
Calorific value
27.1 53.54
2.2.1 Acid Value
(MJ/kg)
It is the mass of potassium hydroxide (KOH) in
6 Flash point (°C) 151 66
7 Fire point (°C) 200 76 milligrams that is required to neutralize one gram of
8 FFA% 0.959 - chemical substance. Precisely, five gram oil was taken
9 Acid number 0.479 - in a conical flask and isopropyl alcohol was added till
of 60 °C for 90 minutes with a constant agitation speed the oil dissolves completely. This solution was titrated
of 300 rpm. Then, the mixture was allowed to cool and against 0.1 N KOH solution after adding 1 or 2 drops of
settle in a separating flask for 12 hours. Two phenolphthalein indicator. The appearance of pink
layers were formed in the separating flask; bottom colour indicates the endpoint. Then, the acid value was
layer of glycerol and top layer of biodiesel. The top calculated using the formula given below.
layer of biodiesel has been separated by using a Acid value = (A B) × N × 56.1/W
separating flask. The properties of biodiesel are listed where, A = amount of 0.1 N KOH required for titration
in Table 1. with oil;
B = amount of 0.1 N KOH required for titration
2.2 Determination of Storage Stability of Biodiesel
without oil;
Storage stability studies were conducted on two W = weight of oil, N = normality of KOH.
 An Investigation to Assess Storage Stability of Pomelo Seed Oil Biodiesel 13

2.2.2 Peroxide Value (2) A = 0.59, B = 0.00247 for t ≥ 100.

Peroxide value is a measure of the concentration of
peroxides and hydroperoxides formed in the initial
stages of lipid oxidation. Milliequivalents of peroxide
per kg of fat are measured by titration with iodide ion.
To measure peroxide value, 5 gm of oil was taken in a
flask. A mixture of 20 mL of acetic acid and 10 mL of
chloroform was added to oil sample. The mixture was
swirled at low speed. After proper mixing, 0.5 mL of
freshly prepared potassium iodide solution was added
to the solution and stirred at a low speed. Then, the
solution was treated with 30 mL of distilled. This
solution was titrated against 0.1 N sodium thiosulphate,
using 0.5 mL starch solution as indicator. The amount
of sodium thiosulphate solution required for titration
was recorded. Peroxide value was determined by the
following method:
S = Titration of sample;
B = Titration of blank;
N = Normality of sodium thiosulphate solution.
Peroxide value:
2.2.3 Determination of Viscosity
The viscosity was measured in redwood viscometer.
Precisely, 50 mL of biodiesel was heated at 80 ºC in the
viscometer. By opening the aperture at the bottom of
the viscometer, the oil was allowed to flow and the time
was recorded. The kinematic viscosity was determined
by using the formula,
Kinematic viscosity = At B/t.
The constants:
(1) A = 1.79, B = 0.00260 for 34 ≤ t ≤ 100;
3. Results and Discussions
The acid values, peroxide values and viscosities had
been determined both conditions at an interval of one
month for five month duration. During the
experimentation period, the average ambient
temperature and humidity had been recorded and given
in Table 2.
The variations of acid values, peroxide values and
viscosities with storage time have been shown in Figs.
2-4 for both open and closed conditions. The results
were distinctly divided in two phases; first from
September to November 2016 and second from
December 2016 onwards till the end of the experiment.
During the first phase, the acid values, peroxide
values and viscosities for both conditions tended to
decrease as the ambient temperature, humidity and
sunlight tended to fall progressively on the advent of
winter. This indicated the slowing of oxidations
reactions and improvement in stability. The lowest acid
values, peroxide values and viscosities reached by both
the open and close samples had been 1.22 & 0.561, 20
& 2 mEq/kg and 1.26 cSt & 0.85 cSt respectively for
the month of November 2016. In the second phase, the
ambient temperature and humidity levels rose. This
expedited oxidation reactions in biodiesel samples. As
result, acid values, peroxide values and viscosities had
increased with storage time for both the samples in
open and closed conditions. Another important
observation during this phase was that the biodiesel in
close condition degraded at a higher rate than that of

Table 2 Stability status.

Peroxide value
Acid Value Viscosity (cSt)
Days (mEq/kg) Temp (°C) Humidity (%)
Open Close Open Close Open Close
30/09/2016 2.95 2.95 40 40 1.27 1.27 32.2 57
30/10/2016 2.24 2.44 30 30 1.50 1.50 28.2 48
30/11/2016 1.22 0.561 20 2 1.32 0.85 23.5 35
30/12/2016 1.875 2.453 27 27 1.47 1.05 24 64
30/01/2017 2.287 3.927 22 22 1.46 1.32 26 47
14 An Investigation to Assess Storage Stability of Pomelo Seed Oil Biodiesel

Acid value open Acid value close

4.5
4.0
3.5
3.0
Acid value

2.5
2.0
1.5
1.0
0.5
0.0
30/9/2016 30/10/2016 30/11/2016 30/12/2016 30/1/2017
Days

Fig. 2 Acid value vs. Days.

45
Peroxide value open Peroxide value close
40
35
Peroxide value (mEq/kg)

30
25
20
15
10
5
0
30/09/2016 30/10/2016 30/11/2016 30/12/2016 30/01/2017
Days
Fig. 3 Peroxide value vs. Days.

Viscosity open Viscosity close

1.6
1.4
Viscosity(cSt)

1.2
1.0
0.8
0.6
0.4
0.2
0.0
30/09/2016 30/10/2016 30/11/2016 30/12/2016 30/01/2017
Days
Fig. 4 Viscosity vs. Days.
 An Investigation to Assess Storage Stability of Pomelo Seed Oil Biodiesel
open one. Unlike the open condition, the temperature
and humidity level were intact in the closed condition
and were slightly higher than that of open one. This
small temperature gradient resulted higher oxidation
that sample.
4. Conclusions
In this study, storage stability of pomelo seed oil
biodiesel is evaluated on terms of acid value, peroxide
value and viscosity. The behaviour of parameters can
be divided into two phases. During the first phase
(September to November 2016), the parameters
showed a decreasing trend indicating lower oxidation
due to low ambient temperature and humidity level. As
the surrounding temperature and humidity increased
slightly in the second phase, there was rapid oxidation
in biodiesel and acid values, peroxides values and
viscosities rose with storage time. The oxidation of
biodiesel can be reduced by applying antioxidants such
as TBHQ, PrG, BHA and BHT.
References
[1] Peterson, C. L., Feldman, M., Korus, R., and Auld, D. L.
1991. “Batch Type Transesterification Process for Winter
Rape Oil.” Applied Engineering in Agriculture 7 (6):
711-6.
[2] Ma, F., Clements, L. D., and Hanna, M. A. 1998. “The
Effects of Catalyst, Free Fatty Acids, and Water on
Transesterification of Beef Tallow.” Trans. Am. Soc. Agric.
Eng. 41 (5): 1261-4.
[3] Balaji, G., and Cheralathan, M., 2013. “Potential of
Various Sources for Biodiesel Production.” Energy
Sources, Part A 35: 831-9.
[4] Ilkilic, C., and Behçet, R. 2010. “The Reduction of
Exhaust Emissions from a Diesel Engine by Using
Biodiesel Blend.” Energy Sources, Part A 32 (9): 839-50.
[5] Canakci, M., and Van Gerpen, J. H. 2003. “Comparison of
Engine Performance and Emissions for Petroleum Diesel
Fuel, Yellow Grease Biodiesel, and Soybean Oil
Biodiesel.” ASAE 46 (4): 937-44.
[6] Karaosmanoglu, F. 1999. “Vegetable Oil Fuels: A
Review.” Energy Sources 21 (3): 221-31.
[7] Canakci, M., and Hooz, M., 2006. “Energy and Exergy
Analyses of a Diesel Engine Fuelled with Various
Biodiesels.” Energy Sources, Part B 1 (4): 379-94.
15
[8] Rajan, K., and Kumar, K. R. S. 2009. “Improvement of
Performance and Emission Characteristics of a DI Diesel
Engine with Turbulence Induced Piston (Internal Jet
Piston) Using Biodiesel Blends.” SAE International 28
(December): 35-40.
[9] Bora, D. 2009. “Performance of Single Cylinder Diesel
Engine with Karabi Seed Biodiesel.” Journal of Scientific
& Industrial Research 68 (11): 960-3.
[10] Bora, D. 2014. “Biofuel Production from Mesua Ferrea L
Seed Oil.” International Journal of Engineering and
Technical Research (IJETR) 2 (9): 45-50.
[11] Du Plessis, L. M., De Villiers, J. B. M., and Van der Walt,
W. H. 1985. “Stability Studies on Methyl and Ethyl Fatty
Acid Esters of Sunflower Seed Oil.” Journal of the
American Oil Chemists Society 62 (4): 748-52.
[12] Bondioli, P., Gasparoli, A., Lanzani, A., Fedeli, E.,
Veronese, S., and Sala, M. 1995. “Storage Stability of
Biodiesel.” Journal of the American Oil Chemists Society
72 (6): 699-702.
[13] Bondioli, P., Gasparoli, A., Della Bella, L., and Tagliabue,
S. 2002. “Evaluation of Biodiesel Storage Stability Using
Reference Methods.” European Journal of Lipid Science
and Technology 104 (12): 777-84.
[14] Bondioli, P., Gasparoli, A., Della Bella, L., Tagliabue, S.,
and Toso, G. 2003. “Biodiesel Stability under Commercial
Storage Conditions over One Year.” European Journal of
Lipid Science and Technology 105 (12): 735-41.
[15] Middlemarch, M., and Schober, S. 2003. “The Influence
of Antioxidants on the Oxidation Stability of Biodiesel.”
Journal of the American Oil Chemists Society 80 (8):
817-23.
[16] Serrano, M., Bouaid, A., Martinez, M., and Aracil, J. 2013.
“Oxidation Stability of Biodiesel from Different
Feedstocks: Influence of Commercial Additives and
Purification Step.” Fuel 113 (November): 50-8.
[17] Serrano, M., Martinez, M., and Aracil, J. 2013.
“Long-Term Storage Stability of Biodiesel: Influence of
Feedstock, Commercial Additives and Purification Step.”
Fuel Processing Technology 116 (December): 135-41.
[18] Knothe, G., and Dunn, R. O. 2003. “Dependence of Oil
Stability Index of Fatty Compounds on Their Structure
and Concentration and Presence of Metals.” Journal of the
American Oil Chemists Society 80 (10): 1021-6.
[19] Dilip, B. 2009. “Storage Stability of Mahua Oil Methyl
Ester.” Journal of Scientific & Industrial Research 68 (2):
149-52.
[20] Madhurjya, S. 2017. “Effect of Antioxidant on Storage
Stability of Citrus Maxima Biodiesel.” International
Journal of Mechanical and Production Engineering 5 (2):
49-51.
Journal of Energy and Power Engineering 12 (2018) 16-25
doi: 10.17265/1934-8975/2018.01.003
D DAVID PUBLISHING

Behavior of Hydrogen and Nitrogen in Tungsten, as

Divertor Wall of a Fusion Reactor

Sergio Beneitez1, Juana L. Gervasoni1, 2 and Juan Carlos Furnari3

1. Bariloche Atomic Center, CNEA (National Atomic Energy Commission), Bariloche 8400, Argentina
2. CONICET (National Council of Scientific and Technical Research), Argentina
3. Ezeiza Atomic Center, CNEA (National Atomic Energy Commission), Ezeiza 1802, Argentina

Abstract: The analysis of the interaction of hydrogen, nitrogen (and their isotopes) with tungsten is important, since this material is a
strong candidate to form the first wall of fusion reactors for both magnetic and inertial confinement, and these atoms have a very
sensitive (desired and unwanted) interaction with it. For this purpose, we study the effects and electronic state densities of atomic
hydrogen and nitrogen in pure tungsten, in order to analyze some important properties such as the density of states of the system.
Focusing on this application, this work is a preliminary study of the behavior of atoms of hydrogen and nitrogen, on a surface of
tungsten on the three sites of the cell: top, hollow and bridge. We use a program simulation based on the DFT (density functional
theory) implemented in the Open-Source Code Quantum Espresso, in order to obtain the adsorption energy and the density of states
of the systems.

Key words: Tungsten, electronic properties, fusion.

1. Introduction capacity at high temperature (11.3 kW/m at 1,000 °C),

is characterized by a high thermal conductivity and a
The new reactor concepts, including fusion, are
low coefficient of thermal expansion, so that it
characterized by greater efficiency, better utilization of
presents a good resistance to thermal shock. Tungsten
nuclear fuel and waste minimization. To achieve high
has a high threshold for physical spray energy, so it is
performance parameters, ongoing research and testing
a low erosion material when irradiated with different
of new materials is necessary [1].
types of particles; has a low solubility of H2, low
The recent conceptual designs of a fusion reactor,
retention of tritium and its activation is low [1]. The
both confinements as magnetic inertial [2] consider
conditions in which the fusion is conceived prioritize
tungsten (W) as one of the functional materials [3-5],
the study of the damage on tungsten caused by
candidate to be part of several engineering units, in
hydrogen and its isotopes under extreme conditions
particular the first wall or first containment barrier, the
(pressure, temperature). However, relevant
divertor [5, 6], subjected to extreme conditions,
experimental results were obtained for mechanisms of
temperatures and irradiation, even higher than those in
formation of molecules of NH3 on tungsten surface [7],
a fission reactor in terms of fluxes, spectrum hardness
due to the interaction of it with H, N, H2, N2 particles
and number of source terms (neutrons, gamma, H, H2,
4 of the plasma at low temperatures [8]. Taken this into
He, 3He). Of all metals and alloys, tungsten has the
account, we will analyze the mechanisms of
highest melting and boiling point (PF-3,422 °C,
adsorption of H, H2, N, N2 as precursors of the
PE-5,930 °C), exhibits high hardness and good heat
formation of NH3 in tungsten. The term adsorption is
introduced in Ref. [9] to connote the condensation of
Corresponding author: Juana L. Gervasoni, Ph.D., research gases on surfaces, produced by the forces of
fields: metallic and nanostructured, of applied research
management. interaction between the solid and the molecules or
 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor
atoms of the gas. Basically these forces are of two
kinds: physical, with predominant interactions of Van
der Waals type, and chemical interactions that have
taken into account the chemical bonds of the system.
Thus, adsorbate is defined as the unit (atom or
molecule) adsorbed by the metal surface.
2. Materials and Methods
In this article we employ the DFT (functional
density theory) [10-14] implemented in the
Open-Source Code Quantum Espresso [15]. In order
to calculate the adsorption energy of the system, we
solve the Kohn-Sham equations in the frame of the
GGA (generalized gradient approximation), using the
PBE (Perdew-Burke-Ernzerhof) functional [13] with
ultrasoft Vanderbilt pseudopotential. These equations
were solved in a self-consistent way, for a periodic
system of symmetry of translation in 3D for the case
of the calculation of the bulk. A slab was defined for
calculations with 2D symmetry with the extended
third dimension (in the z direction), in order to
simulate a surface by increasing the distance between
cells.
The Kohn-Sham orbitals are expanded in plane
waves and the cut-off value for the energy is 650 eV.
The geometric optimization is performed using the
BFGS (Broyden-Fletcher-Goldfarb-Shanno) algorithm.
The adsorption energy is defined as [9]:
Fig. 1 Scheme of the considered system W-atom: (1) top interaction site, (2) bridge interaction site, (3) hollow interaction
sites.
17
E E ⁄ E E (1)
where Emolecule/surface is the energy of the
surface-molecule (or atom) system, Esurface is the
energy of the surface of the adsorbate and Emolecule is
the energy of the isolated molecule (or atom). According
to this definition, positive adsorption energy corresponds
to an attractive interaction between the adsorbate and
the surface. The dimensions of the super cell were
defined by parameterizing the increase of the
corresponding dimension considering the variation of
the adsorption energy Eads, taking as a criterion a
variation of less than 1% of it. The cutoff and the
mesh of the first Brillouin zone were determined in
the same way. We define ecutwfc and ecutrho as the
energy cutoff (Ry) and the kinetic energy cutoff for
charge density and potential. We selected for this
system: ecutwfc = 30, ecutrho = 10 * ecutwfc. The
size of the k-mesh is 10 × 10 × 1. After the relaxation
of the parameters to determine Eads, and the
subsequent selfconsistent calculation, the calculations
of the DOS (density of states) were implemented.
3. Results
Considering Eads as the parameter to be stabilized,
we take a slab of 10 atomic layers in the direction
(001) with a lattice of 74 atoms. The same size is
taken in the perpendicular directions (010), and (100),
as shown in Fig. 1. In each of the three crystallographic
18 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor
Table 1 Energy of adsorption for the tree tipes if
interaction sites, for H and N, measured in eV.
Atom Top Hollow Bridge
H 3.630487 3.645677 4.729204
N 8.919015 11.51710 10.69447
Table 2 Distances W-atom in A.
Atom Top Hollow Bridge
d (W-H) [A] 1.773792 2.205204 1.918391
d (W-N) [A] 1.72925 1.849850 1.894001
directions of tungsten, we study different sites and
angular dependence: Top: the atom interacting directly
with an atom of the surface (Fig. 1(1)). Hollow: the
atom interacts with 4 atoms of the surface (Fig. 1(2)).
Bridge: The atom interacting with two atoms on the
surface (Fig. 1(3)). The atoms were fixed at a distance
between d atom-W = 2 Å, with respect to the surface.
The positions were allowed to fully relax. The
network parameter for the optimized system, a =
3.162 Å, diverging from the reference value a = 3.164
Å obtained from Ref. [16], in less than 0.1%, was
used in the calculations.
Table 1 shows the adsorption energy of the
hydrogen and nitrogen atoms, in both it was observed
that the energy of adsorption does not vary
appreciably according to the site of interaction. Table
2 lists the distances resulting from the relaxation
calculation, between the adsorbated atoms and metal
surface atoms (datom-W).
3.1 Density of States
3.1.1 Hydrogen on W
In this sub section we show the DOS of H on W, for
the three sites: hollow, bridge and top.
Let us remember that the DOS of a system
describes the number of states per interval of energy
(at each energy level) available to be occupied. PDOS
(projected density of states) is related with DOS and is
the number of available states for electrons to occupy
projected on atomic orbitals. It means that it is
possible to see which shell (s, p, d, f) and orbital for
specified atom has the biggest share in DOS.
Note that for the following graphics, EFermi for the
pure W is shown in green, and for the WH system is
shown in red.
In Fig. 2a we show DOS on W, and W-H system for
bridge site: it can be noticed that near the Fermi level,
the W-H system continues to be conductive although
to a lower extent with respect to the W (pure) system.
For the projected state density (Fig. 2b) it can be noted
that near the Fermi level the 3d orbital is the one that
contributes the most to the transport properties or to
the conductive character: hydrogen does not
contribute.
For H on hollow site: we see in Fig. 3a that for this
interaction site it can be noticed that near the Fermi
level the density of the state does not change, so the
system’s characteristics do not change. In Fig. 3a, we
see that, as for the bridge interaction site, the projected
state density near the Fermi level, the 3d orbital gives
the principal contribution to the transport properties or
to the conductive character. In this case, hydrogen
does not change the behavior of DOS.
Finally, in Fig. 4a we show the properties for top
site: just as for the bridge and hollow sites, for this
interaction site it may be noted that near the Fermi
level, the density of the state does not change, so the
system keeps its driver’s property unalterable or does
not change.
On the other hand, from Fig. 4b it is observed that
near the Fermi level for this site, it is the orbital s of
the hydrogen atom that governs or contributes slightly
to the conductive character.
3.1.2 Nitrogen on W
In this subsection we show the DOS of N on W, for
the three sites: hollow, bridge and top, following the
scheme of the precedent subsection.
In Fig. 5a we show DOS on W, and W-N system for
bridge site: it may be noted that the Fermi level of the
W-N system shifts slightly for higher frequencies (or
energies) and decreases near the EFermi.
For the PDOS, from Fig. 5b it can be noted that the
orbital 2p of the tungsten competes with the nitrogen
 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor 19

(a)

(b)
Fig. 2 (a) DOS of H-W on bridge site; (b) PDOS (projected DOS) of the same system.
20 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor

(a)

(b)
Fig. 3 (a) DOS of H-W on hollow site; (b) PDOS of the same system.
 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor 21

(a)

(b)
Fig. 4 (a) DOS of H-W on top site; (b) PDOS of the same system.
22 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor

(a)

(b)
Fig. 5 (a) DOS of N- bridge site; (b) PDOS of the same system.
 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor 23

(a)

(b)
Fig. 6 (a) DOS of N-W on hollow site; (b) PDOS of the same system.
24 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor

(a)

(b)
Fig. 7 (a) DOS of N-W on top site; (b) PDOS of the same system.

orbital 3d for defining the conductive character of the
system.
From Fig. 6a, we see that the energy EFermi moves
even more than for the interaction of the bridge site.
From Fig. 6a the contribution is observed very similar
near the level of the 2p orbital of nitrogen and 3d of
the tungsten.
From Fig. 7a a lower variation of the DOS is
observed near the Fermi level that is almost equal. For
the projected state density (PDOS, Fig. 7a), it can be
noted that the 2p orbital of the nitrogen competes with
the 3d orbital of the tungsten to define the conductive
character, but it is indisputable that the one who
contributes or governs the conductive character to a
 Behavior of Hydrogen and Nitrogen in Tungsten, as Divertor Wall of a Fusion Reactor
greater extent is the orbital 2d of nitrogen.
4. Discussion and Conclusions
In this paper, we studied the density of states of
atomic H and N in W, pointing to analyze the behavior
of this system in the extreme condition of it in a fusion
reactor. We used the Quantum Expresso method,
applying the ultrasoft Vanderbilt pseudopotential to
calculate adsorption energies of H, as precursors of
the NH3 molecule on the tungsten surface in the three
sites: hollow, bridge and top. In order to achieve
adsorption energies with a higher accuracy, one needs
to go beyond the Quantum Expresso approximation
with the consideration of more general kernel within
the DFT formulation. But the goal of this work was to
determine the process of adsorption in order to explain
the existence of NH3 in the divertor. This development
began with the simulation of ion implantation
experiments, validated against existing laboratory
experimental results [7]. The findings and challenges
of developing a 3-D predictive capability for ammonia
transport in a Tungsten divertor wall are under way.
Acknowledgments
The present work was partially subsidized by the
National University of Cuyo (Project UNC 06/C367
and UNC 06-C345) and CNEA’s Controlled Fusion
Program.
References
[1] Zinkle, S. J. 2005. “Advanced Materials for Fusion
Technology.” Fusion Engineering and Design 74 (1-4):
31-40.
[2] Ehrlich, K. 1999. “The Development of Structural
Materials for Fusion Reactors.” Transactions of the Royal
Society of London A 357 (1752): 595-617.
[3] Tanaka, S., Matera, R., Kalinin, G., Barabash, V., and
Mohri, K. 1999. “ITER Materials R & D Data Bank.”
Journal of Nuclear Materials 271-272 (May): 478-85.
[4] Kalinin, G., Barabash, V., Cardella, A., Dietz, J., Ioki, K.,
Matera, R., Santoro, R. T., and Tivey, R. 2000.
25
“Assessment and Selection of Materials for ITER
In-Vessel Components.” Journal of Nuclear Materials
283-287 (1): 10-9.
[5] Bloom, E. E., Busby, J. T., Duty, C. E., Maziasz, P. J.,
McGreevy, T. E., Nelson, B. E., Pint, B. A., Tortorelli, P.
F., and Zinkle, S. J. 2007. “Critical Questions in Materials
Science and Engineering for Successful Development of
Fusion Power.” Journal of Nuclear Materials 367-370
(A): 1-10.
[6] https://www.iter.org/.
[7] Drenik, A., Vesel, A., and Mozetic, M. 2006.
“Measurements of Probability for Heterogeneous
Recombination of Hydrogen Atoms on Surfaces of
Fusion Relevant Materials.” 33rd EPS Conference on
Plasma Phys. Rome, 19-23 June 2006 ECA Vol.30I,
P-5.170.
[8] Decreton, M., Shikama, T., and Hodgson, E. R. 2004.
“Performance of Functional Materials and Components in
a Fusion Reactor.” Journal of Nuclear Materials 329-333
(4): 125-32.
[9] Kaiser, H. 1881. “Condensation of Gases on Free
Surfaces.” Wield. Ann. 14: 451.
[10] Kohn, W., and Sham, L. J. 1965. “Self-Consistent
Equations Including Exchange and Correlation Effects.”
Physical Review 140: A1133-8.
[11] Hohenberg, P., and Kohn, W. 1964. “Inhomogeneous
Electron Gas.” Physical Review B 136 (3): 864-71.
[12] Stott, M. J., and Zaremba, E. 1980. “Linear-Response
Theory within the Density-Functional Formalism:
Application to Atomic Polarizabilities.” Physical Review
A 21 (1): 12-23.
[13] Perdew, J. P., Burke, K., and Ernzerhof, M. 1996.
“Generalized Gradient Approximation Made Simple.”
Physical Review Letters 77 (18): 3865-8.
[14] Gervasoni, J. L., and Furnari, J. C. 2014. “Density of
Electron States and Volume of Solution of Hydrogen in
SiC Alloys.” Journal of Energy and Power Engineering 8
(2): 390-3.
[15] http://www.quantum-espresso.org/.
[16] Shikama, T., Tsuchiya, B., and Hodgson, E. R. 2007.
“Effects of Neutron Irradiation on the Properties of
Functional Materials for Fusion Applications: Role of
Hydrogen in Radiation Effects on Oxide Ceramics.”
Journal of Nuclear Materials 367-370: 995-1002.
[17] Van Helden, J. H., Wagemans, W., Yagci, G., Zijlmans, R.
A. B., and Schram, D. C. 2007. “Detailed Study of the
Plasma-Activated Catalytic Generation of Ammonia in
N2-H2 Plasmas” Journal of Applied Physics 101 (4):
043305.
Journal of Energy and Power Engineering 12 (2018) 26-34
doi: 10.17265/1934-8975/2018.01.004
D DAVID PUBLISHING

A Study on Control of Interconnection Inverter with FRT

and Islanding Detection Functions

Shoji Kawasaki and Masahiro Ise

Department of Electronics and Bioinformatics, School of Science and Technology, Meiji University, Kanagawa 214-8571, Japan

Abstract: In this study, the authors aim to develop the interconnection inverter of PV (photovoltaic generation) system with FRT (fault
ride thorough) function and islanding detection function, and analyze the interaction between the both functions during the momentary
voltage drop by using an analytical model of distribution system interconnected plural PV systems. Moreover, the authors propose a
cooperated control method of the inverters with the islanding detection function and FRT function, and carry out a numerical
calculation in order to verify the validity of the proposed method.

Key words: Interconnection inverter, islanding, islanding detection function, momentary voltage drop, FRT function.

1. Introduction inverter with an FRT (fault ride thorough) function to

continue outputting during the momentary voltage
In recent years, for a global warming prevention and
drop.
breakaway from fossil energy resources dependence,
In this study, the authors propose a cooperated
the introduction of the renewable energy progresses
control method of the inverter with FRT function and
worldwide. Particularly, the introduction of the PV
the frequency change rate detection method which is
(photovoltaic generation) has been spreading in Japan
one of the islanding detection functions. The frequency
[1]. The PV system is composed of the interconnection
change rate detection method is used in the previous
inverter and solar panel. However, there is a possibility
study [3]. The detection system sends an off-signal to
that the problem of frequency or voltage occurs to an
the inverter when frequency change rate exceeds the
electric power system by widespread of PVs in the
threshold in succession to avoid unnecessary stop. In
future [2]. When an accident occurs in the electric
this study, in addition, the authors verify the usefulness
power system and the distributed generation is
of the proposed control method by the numerical
paralleled off from electric power system, it becomes
simulation using a distribution system model
islanding. Since there are the problems on preservation
interconnected plural PV systems.
of the electric power system when it is in a condition of
islanding, it is necessary to prevent islanding. 2. Islanding Detection Function
Therefore, the inverters are provided with an islanding
2.1 Frequency Change Rate Detection Method
detection function. On the other hand, if PV systems
are paralleled off by this function when the momentary This section describes the frequency change rate
voltage drop occurs by an accident such as thunderbolt, detection method which is one of the passive methods
demand-supply balance of the electric power system of islanding detection [4]. Fig. 1 shows the control
collapses and it might become the massive blackout. block of frequency change rate detection method. The
Therefore, recent years, it has been developed the method senses a sudden change of frequency of grid
voltage by accidents. When the frequency of voltage
Corresponding author: Shoji Kawasaki, associate professor,
Eng. Dr., research fields: advancement of electric power system, changes suddenly and the frequency change rate ∆
improvement of electric power quality. exceeds the threshold , the PV system is paralleled
 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions
off. The standard of the threshold is ±0.1% ~ ±0.3%.
2.2 Frequency Feedback and Step-Injection Method
This section describes the frequency feedback and
step-injection method which is one of the active
methods of islanding detection [4, 5]. Fig. 2 shows the
control block of frequency feedback and step-injection
method. This method has two functions, the frequency
feedback part and step-injection part. And the reactive
current is injected by these two parts when the
frequency changes suddenly.
Fig. 3 shows the characteristics of frequency
deviation—reactive power of frequency feedback part
[6]. In the frequency feedback part, when the frequency
changes suddenly, the reactive power is injected
rapidly in the direction that promotes a change of
frequency. As shown in Fig. 3, the amount of injected
reactive power is decided depending on the amount of
change of frequency.
On the other hand, when the change of frequency is
small (within ±0.01 Hz) and the harmonic voltage and
the fundamental voltage is changed, the reactive power
in the direction in which the frequency decreases is
injected by the step-injection part as shown in Figs. 4
and 5.
Fig. 1 Control block of frequency change rate detection.
Fig. 2 Control block of frequency feedback and step-injection.
27
In this method, since the reactive power is injected in
the direction that promotes a change of frequency,
there is not mutual interference when plural PV
systems are interconnected. And since the reactive
power is injected depend on change of frequency; PV
systems do not give the influence to the electric power
system.
3. FRT Function
Recent years, it has been developed the inverter
with FRT function that continues outputting during the
momentary voltage drop. This chapter describes the
FRT requirement and proposed control method of FRT
function.
3.1 FRT Requirement
Fig. 6 shows the FRT requirement of the voltage
reduction tolerated dose [7]. If the residual voltage
is over 20% and the duration of voltage drop is
within 1.0 sec, the inverter is required to continue
outputting.
Fig. 7 shows the FRT requirement of the frequency
variation tolerated dose for step variation. Even if the
frequency increases stepwise, the inverter is required to
continue outputting when the rise is less than 50.8 Hz.
28 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions

Fig. 3 Characteristics of frequency deviation—reactive power of frequency feedback part.

Fig. 4 Step-injection in case of harmonic voltage variation. Fig. 7 Frequency variation tolerated dose (for step
variation).

Fig. 5 Step-injection in case of fundamental voltage Fig. 8 Frequency variation tolerated dose (for ramp
variation. variation).

Fig. 8 shows the FRT requirement of the frequency

variation tolerated dose for ramp variation. Even if the
frequency increases, the inverter is required to continue
outputting when the frequency change rate is less than
±2 Hz/s.

3.2 FRT Control Logic

When it becomes islanding, the frequency increases

Fig. 6 Voltage reduction tolerated dose. as shown in Fig. 9 by the reactance of load and by the
 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions 29

block of islanding detection and FRT function. The

authors propose the FRT function that avoids
unnecessary stop during the momentary voltage drop.
First, the larger threshold is set for the frequency
change rate (±0.6% (±0.3 Hz)). And send an off-signal
to the inverter when the frequency change rate ∆
exceeds the threshold consecutively from the
standard value of frequency (50 Hz or 60 Hz) and
the number of frequency change continuation cycle N

Fig. 9 Image of frequency fluctuation during islanding.

frequency feedback and step-injection method. On the

other hand, when the momentary voltage drop occurs,
the frequency fluctuates for a brief time as shown in
Fig. 10. If the inverter is carried with the frequency
change rate detection method, PV system is paralleled
off when the frequency increases like Fig. 9 or when
the frequency fluctuates like Fig. 10, even though PV
system must ride through.
Fig. 11 shows the control block of proposed FRT Fig. 10 Image of frequency fluctuation during momentary
function and Fig. 12 shows the cooperated control voltage drop.

Fig. 11 Control block of proposed FRT function.

Fig. 12 Cooperated control block of islanding detection and proposed FRT function.
30 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions
exceeds the threshold . Hence, in the inverter which
is carried by the proposed cooperated control method,
PV system is paralleled off when the frequency keeps
increasing or decreasing. And when the frequency
exceeds the setting value of over frequency relay
or the settingvalue of under frequency relay
, PV system is paralleled off.
4. Numerical Calculation Example
4.1 Analytical Model of Distribution System and
Simulation Condition
The authors carry out the numerical calculations in
order to verify the validity of the proposed method. Fig.
13 shows the analytical model of distribution system.
The PV system and load are interconnected in the
low-voltage distribution system. The momentary
voltage drop will occur due to the short circuit accident.
The islanding will occur due to separating from
high-voltage system. The load capacity of consumer is
1 kW. The output of PV is 1 kW. The impedance of
distribution lines is shown in Table 1 [8]. Fig. 14 shows
the analytical model of distribution system in which
plural PV systems are interconnected. The condition of
load capacity, output of PV, and impedance of
distribution lines are the same as the condition of Fig.
13.
Table 1 Impedance of distribution lines.
Unit
High voltage wire Ohm/km
Low voltage wire Ohm/km
Fig. 13 Analytical model of distribution system.
4.2 Simulation Results
Firstly, the authors carried out the numerical
calculations in case of PV system without FRT
function and in case of PV system with the proposed
FRT function, using the analytical model of Fig. 13.
Figs. 15-17 show the simulation results in case of
PV system without FRT function during the
momentary voltage drop. Figs. 18-20 show the
simulation results in case of PV system with the
proposed FRT function during the momentary voltage
drop. As shown in Figs. 15-17, in case of PV system
without FRT function, PV system was paralleled off
during the momentary voltage drop by the frequency
change rate detection method and the output of PV
became zero. On the other hand, in case of PV system
with the proposed FRT function, PV system rode
through (see Fig. 19).
Figs. 21-23 show the simulation results in case of
PV system with the proposed FRT function during the
islanding. In case of PV system with the proposed
FRT function, since the frequency exceeded the
setting value of over frequency relay during the
islanding, PV system was paralleled off and the output
of PV became zero (see Figs. 21-23).
Secondly, the authors carried out the numerical
calculation in case of interconnection of the plural PV
Resistance Reactance
0.0616 0.1504
0.2638 0.2438
 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions 31

Fig. 14 Analytical model of distribution system interconnected plural PV systems.

Fig. 15 Voltage variation during momentary voltage drop Fig. 17 Frequency variation during momentary voltage
without FRT function. drop without FRT function.

Fig. 16 Current variation during momentary voltage drop Fig. 18 Voltage variation during momentary voltage drop
without FRT function. with proposed FRT function.
32 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions
Fig. 19 Current variation during momentary voltage drop
with proposed FRT function.
Fig. 20 Frequency variation during momentary voltage
drop with proposed FRT function.
Fig. 21 Voltage variation during islanding with proposed
FRT function.
Fig. 22 Current variation during islanding with proposed
FRT function.
Fig. 23 Frequency variation during islanding with
proposed FRT function.
systems using the analytical model of Fig. 14. All three
PV systems were implemented the proposed FRT
function and the momentary voltage drop occurred.
Figs. 24-26 show the simulation results. All PV
systems rode through and prevented malfunction that
occuring due to the frequency change rate detection
method during the momentary voltage drop.
In addition, numerical calculation was carried out in
case that only PV system 3 in Fig. 14 was
implemented the proposed FRT function and the
momentary voltage drop occurred. Figs. 27-29 show
the simulation results. As shown in Fig. 28, PV
system 1 and PV system 2 malfunctioned and
paralleled off during the momentary voltage drop. PV
system 3 rode through and prevented malfunction
occuring due to the frequency change rate detection
method. Since PV system 1 and PV system 2
paralleled off during the momentary voltage drop, the
frequency of PV system 3 changed more than the case
of all three PV systems with the proposed FRT
function.
Fig. 24 Voltage variation in case of interconnection of
plural PV systems.
 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions
Fig. 25 Current variation in case of interconnection of
plural PV systems.
Fig. 26 Frequency variation in case of interconnection of
plural PV systems.
Fig. 27 Voltage variation in case of only PV system 3 is
implemented with proposed FRT function.
Fig. 28 Current variation in case of only PV system 3 is
implemented with proposed FRT function.
33
Fig. 29 Frequency variation in case of only PV system 3 is
implemented with proposed FRT function.
5. Conclusions
In this study, the authors proposed the cooperated
control method of the interconnection inverter of PV
system with FRT function and islanding detection
function that avoids unnecessary stop. In these
functions, PV system is paralleled off when the
frequency change rate exceeds the threshold
consecutively from the standard value of frequency (50
Hz or 60 Hz). PV systems with these proposed
functions can ride through and prevent malfunction
that is occurred by the frequency change rate detection
method during the momentary voltage drop. The
authors showed the usefulness of the proposed control
method by the numerical calculations.
In the future, the authors will investigate the
interaction between the islanding detection function
and FRT function during the momentary voltage drop
on various conditions. And the authors will implement
the proposed method to the actual inverter and will
verify the usefulness of the proposed method using
experimental equipment.
References
[1] Kurokawa, K. 2008. “The State-of-the Art of Solar
Photovoltaic and Its Future Direction.” IEEJ Trans. PE
128 (7): 904-7.
[2] Ishikawa, T. 2006. “Outline of Grid-Interconnection
Technologies of Distributed Generation Systems.” IEEJ
Trans. PE 126 (10): 964-8.
[3] Kobayashi, H. 2009. “Development of Islanding
Detection Measure for PV Power Generation to Achieve
Stable Operation in Grid Frequency Variation.” CRIEPI
Research Reports. No. R08013.
34 A Study on Control of Interconnection Inverter with FRT and Islanding Detection Functions

[4] Kai, T. 2015. “Thorough Explanation Illustration Power
System Interconnection.” Ohmsha, Ltd.
[5] Endo, H., Takuma, T., Mizukawa, Y., Hayashi, K., and
Shinoda, Y. 2016. “Development of Photovoltaic
Generation System Line Back α IV.” GS Yuasa Technical
Report 13 (1).
https://www.gs-yuasa.com/jp/technic/vol13/pdf/13_01_0
19.pdf.
[6] Tsutsumi, K., Kimura, R., Sano, M., Taoka, H., Takano,
H., and Sakata, T. 2016. “A Study on Influence of
Islanding Detection Time Caused by Response Speed of
PV Reactive Power.” Annual Meeting of IEEJ 6: 201.
[7] Interconnection of Power System Expert Committee.
2016. “Grid-Interconnection Code JEAC9701-2016.” The
Japan Electric Association.
[8] Kondoh, J. 2010. “Evaluation on Reduction of Output
Suppression Loss by Cooperative Control of Voltage
Profile in a Distribution System with a Large Amount of
Photovoltaic Power Generation.” IEEJ Trans. PE 130 (11):
981-8.
Journal of Energy and Power Engineering 12 (2018) 35-43
doi: 10.17265/1934-8975/2018.01.005
D DAVID PUBLISHING

Evaluation of Height of Tsunami Induced by Submarine

Landslide

Yoshikane Murakami1, Takemi Shikata2 and Koji Tonomo2

1. Centers for Civil Engineering and Architectures, Kansai Electric Power Co., Inc., Fukushima Ward, Osaka, Osaka 553-0003,
Japan
2. Port and Coastal Engineering Group, Newjec Inc., Kita Ward, Osaka, Osaka 531-0074, Japan

Abstract: Under the new regulatory requirements for nuclear power plants in Japan, which were enacted in response to the
nuclear accident associated with the Great East Japan Earthquake Tsunami that occurred on 11 March 2011, it is a requirement to
establish a site-specific “standard tsunami” based on numerical analysis considering non-seismic factors in addition to general
seismic faults. It is necessary to establish a consistent evaluation scheme for estimation of tsunami height induced by submarine
landslide, since a standard framework for evaluation has not yet been established even though several models for calculation have
been proposed and applied in practice. In this study, we estimated the scale of submarine landslide from a literature survey and
showed examples of tsunami height evaluation using multiple schemes. As a result of evaluation of tsunami height using three
schemes, the Watts model, the KLS model, and the modified-KLS model, the result obtained by the KLS model was comparatively
large for every case.

Key words: Submarine landslide, landslide-induced tsunami, tsunami height evaluation, new regulatory requirements for nuclear
power plants in Japan, standard tsunami.

1. Introduction the review of conformity to the new regulatory

requirements for nuclear power plants, it is a
In cases of tsunami induced by offshore submarine
requirement to establish a site-specific “standard
landslide, it is reported that although frequency of
tsunami” based on numerical analysis, allowing for
occurrence is exceptionally low compared with the
uncertainties such as source location, scale and
more common seismic tsunami induced by marine
occurrence timing etc., considering general seismic
fault activity, these may nonetheless on rare occasions
faults, non-seismic factors and combinations thereof
inflict extensive damage, such as in the 1771
selected as the source of the tsunami.
Meiwa-Yaeyama earthquake tsunami, et cetera. In
On the other hand, it is recognized as an important
addition, since traces of numerous offshore submarine
issue to establish a consistent evaluation scheme for
landslides have been discovered in the sea around
submarine landslide, since a standard framework for
Japan, the possibility of future occurrences cannot be
evaluation has not yet been established although
excluded. Therefore, it is expedient to consider
several models for calculation have been proposed and
appropriate offshore submarine landslides as a
applied in practice. In this study, we propose one
necessary factor in evaluating tsunami impacts on
consistent method of tsunami height evaluation
coastal structures, based on topographical
applicable to offshore submarine landslide, based on
characteristics of the evaluated site. For example, in
the review on conformity to the new regulatory
requirements relating to the Takahama nuclear power
Corresponding author: Yoshikane Murakami, master of
plant (Takahama NPP) owned by Kansai Electric
engineering, manager, research fields: tsunami engineering,
dam engineering. Power Co., Inc.
36 Evaluation of Height of Tsunami Induced by Submarine Landslide
2. Procedure for Evaluating Submarine
Landslide Tsunami
In the evaluation of tsunami induced by the
submarine landslides in the waters surrounding
Wakasa Bay in the Japan-Sea coast, we carried out the
following procedure based on available scientific
evidence. Tsunami height estimation was carried out
in consideration of the uncertainty of the location and
scale of the submarine landslide, which may have the
potential for a tsunami to affect the location site of
power plant.
(1) Literature survey
 Confirm the distribution of submarine landslide
based on the results of a review of literature, such as
the seabed geological map.
(2) Specify the features of landslide topography
 Interpret possible landslide terrain features from
the submarine topography prepared by commercial
digital data such as M7000 series etc.
 Interpret the marine acoustic exploration record
and classify the strata based on the classification of
Ref. [1]; ascertain presence of submarine landslide.
(3) Scale estimation of submarine landslide
topography
 Divide the location area of the extracted
submarine landslide topography with consideration of
its position and orientation, and approximately
evaluated by the estimated vertical cross-sectional
area of the collapsed part of the landslide.
 To calculate the detailed topographic change
due to the submarine landslide, interpret the collapsed
part and sedimentary part of the sea bottom
topography judged to be the largest submarine
landslide in each divided area, based on the reanalysis
of the acoustic exploration record in the relevant sea
area.
(4) Tsunami height estimation
 Estimate the initial sea level distribution of the
tsunami using multiple procedures.
 Calculate tsunami propagation with nonlinear
long wave theory.
3. Setting of Submarine Landslide
Properties to Be Evaluated
3.1 Detection and Extraction of the Submarine
Landslide Traces
According to Ref. [2], a group of submarine
landslide traces is shown near the Oki Trough of the
Japan Sea. Also, several possible slump scarps are
shown extensively around the Oki Trough in the
marine geological map provided by National Institute
of Advanced Industrial Science and Technology (e.g.
Ref. [3]). According to the sedimentological map (e.g.
Ref. [4]), the facies suspected of submarine landslides
are shown around Oki Trough in the same way. On the
other hand, slump scarps and facies suspected of
submarine landslides are not shown in other areas.
Based on the results of these literature surveys, we
ascertained the presence of submarine landslide
topography by reanalysing results of high resolution
marine acoustic exploration records for all sea areas
around the Oki Trough where there might be a
submarine landslide topography.
As a result, a total of 38 submarine landslide traces
were extracted on the continental slope near the
southeastern and southwestern parts of the Oki Trough,
about 500-1,000 m in depth.
3.2 Selection of Submarine Landslide Topographies to
Be Evaluated
The 38 submarine landslide traces extracted as a
study target were broadly classified into three areas
(areas A to C) depending on their locations and
orientations, as shown in Fig. 1.
Here, with regard to the submarine landslide ES-G3
of area A, located in the north of Gentatsu-se (a part of
the Echizen Bank Chain), Ref. [5] mentioned that
“This submarine landslide trace (Es-G3) was formed
by two slope failures. The collapse time is thought to
be at the time of declining sea level in the last ice age,
and the trigger for the collapse may have been an
increase in the inclination angle due to developing
anticline structure, and an increase in the supply of
 Evaluation of Height of Tsunami Induced by Submarine Landslide 37

Based on the above, as shown in Fig. 2, Es-G3 in

area A, Es-K5 in area B and Es-T2 in area C were
selected as the largest submarine landslide topography
of the respective areas to be evaluated.

3.3 Estimation of Topographic Change Caused by

Submarine Landslide

The sea bottom topography change due to landslide

is estimated by the following procedure (Fig. 3) in
accordance with Ref. [5] with high resolution marine
acoustic exploration records, since it needs to be
Fig. 1 Location map of 38 extracted submarine landslide
estimated accurately as it is considered to directly
traces.
affect tsunami height.
sediment to the slope area.” In other words, it is (1) Reanalyze the high resolution marine acoustic
impossible to accurately estimate a future submarine exploration records and interpret the width, length,
landslide, since the geological environment around the altitude etc. of the collapsed part and sedimentary part
Oki Trough is totally different from that in the past.
Therefore, we decided to evaluate, for each area, the
tsunami induced by the largest one from the extracted
traces of earlier submarine landslides. Using this
approach, we expected that we could obtain a
comprehensive estimate, on the safe side, for tsunami
height evaluation, even allowing for the
uncertainties—the possibility that a future submarine
landslide will occur on a different scale and in a
different place.
Since vertical water level fluctuations have a high Fig. 2 Selected largest submarine landslide topography of
impact on tsunami height estimation, we considered respective areas.
the largest submarine landslide to be the submarine
landslide topography with the largest vertical cross
section of the collapsed part, estimated from the high
resolution marine acoustic exploration records for the
respective areas. Regarding the validity of the method
of scale evaluation based on the cross-sectional area
applied in this study, it is confirmed that the same
landslide topography is selected as the largest scale,
even in the case where the estimated volume of the
collapsed part is calculated by multiplying the
projected area of the slump scarp by the maximum
thickness, interpreted from the high resolution marine Fig. 3 Typical example of interpretation of the sea bottom
acoustic exploration records. topography change due to submarine landslide.
38 Evaluation of Height of Tsunami Induced by Submarine Landslide

from the features of the facies of multiple sub-bottom

profiles passing through the submarine landslide
topography.
(2) To restore the terrain contour before the collapse,
extend the outline of the scarp edge and connect it
smoothly, cross-check between each of the survey
lines, and confirm three-dimensional consistency of
the restored terrain contour.
(3) Considering the balance between the volume of
collapsed part and sedimentary part, estimate a
Fig. 4 Estimated distribution of the sea bottom change due
distribution of the volume of topographic change in
to submarine landslide in area B.
the sea bottom.
Applying the above procedure on the submarine 4.2 KLS Model
landslide of Es-G3 in area A interpreted by Ref. [5],
Based on the calculation model proposed by Ref.
produced substantial consistency with Yamamoto’s
[6] as shown in Fig. 5, the topographic change before
results in the restored topography before the collapse,
and after the landslide is considered as propagating at
the distribution of the sea bottom topography change
sliding velocity “U” from the single rupture starting
and the collapse volume, and confirmed the
point (peak of the collapsed part), and the topography
reasonability of the procedure. The above procedure
change of each calculation point is considered as
was then also applied for areas B and C to estimate
continuing for duration time “Tz”. Next, the tsunami
the change in the sea bottom topography. Fig. 4 shows
propagation simulation is carried out with the
the estimated distribution of the sea bottom change
condition that this topographical change amount
due to the submarine landslide in area B, as one
(change amount per unit calculation time) is
instance.
considered as directly reflecting in the sea bottom
4. Tsunami Height Estimation Induced by topography and sea level.
Submarine Landslide Currently, there is no established knowledge to set
the required parameters “U” and “Tz” in general.
4.1 Approach for Tsunami Height Estimation
Therefore, in accordance with our previous study [9],
For evaluation of the tsunami height by the we applied the maximum landslide velocity obtained
submarine landslide, tsunami propagation simulation by the estimation formula, shown as Umax in Eq. (1),
was carried out with the condition that the initial water proposed by Ref. [8] as sliding velocity “U” in this
level distribution was calculated based on the original study. As for the duration time “Tz”, we applied
sea bottom topography and the collapsed topography
after the landslide.
Calculation of the initial water level distribution
was carried out by more than one forecasting approach
in consideration of uncertainty. The KLS (kinematic
landslide) model proposed by Ref. [6], and Watts
model proposed by Refs. [7] and [8] were applied to
set the initial condition for the tsunami propagation
simulation. Fig. 5 Conceptual overview of the KLS model proposed.
 Evaluation of Height of Tsunami Induced by Submarine Landslide 39

the calculated value obtained from Eqs. (2) and (3)
based on the landslide velocity Umax so that the
duration time would be the fastest in the range in
which the topographic change is considered
reasonable. Also, the completion time of the landslide,
“Tend”, can be calculated as shown in Eq. (4).
b sin   (  1)  tan  
U  gd 1   (1)
d 2Cd  tan  
TZ  DZ U Z
DZ  T cos , 　U Z  U sin 
Tend  TZ  (h  d ) U Z
where, b: length of sliding mass, d: initial depth of
sliding mass, T: thickness of sliding mass, θ: slope
angle, Ψ: friction angle, γ: specific gravity of sliding
mass, Cd: drag coefficient, DZ: maximum collapse
depth, UZ: sliding velocity (vertical component), h:
water depth.
4.3 Watts Model
Whereas the above-mentioned KLS model
calculates the time series change of the sea bottom
topography and sea surface fluctuation simultaneously
for every time unit, the Watts model is a relatively
simple scheme, in that it calculates tsunami
(2) propagation by using as the initial condition the
(3) initial spatial distribution of water level due to the
landslide.
(4)
To adopt the Watts model, it is necessary to
properly define the initial water level distribution. In
this study, we considered applying the proposed
formula of Refs. [8] and [7] based on Ref. [9]. The
proposed equations are shown in Eqs. (5) and (6).

0,3 D  y  y0     
  x  x  x0   
2 2
  x  x0  
  x, y    sech 2     exp       '
exp 
    (5)
min  w  0     0  0 
  
  
  

(= λ0/2), which are required for the proposed equation

 w 
 0,3 D   0, 2 D   (6) are obtained from the result of interpretation of
 w  0 
collapsed topography, or calculated from the
where, η0,3D: maximum water level decrease prediction equation of the tsunami amplitude as shown
(three-dimensional), w: width of SMF (submarine below.
mass failure), ηmin: minimum of the function on the R   Cm
t0  (7)
right-hand side of (5) excluding the amplitude, κ; κ’: g  1
shape parameters (κ = 3). In addition, the parameters
w, η0, D, λ0: characteristic tsunami wave length and Δx 0  t0 gd (8)

 0.131  T  b sin  
1.25 0.63
b
0,2 D  S0     1.47  0.35(  1) (  1)
0.39
     (9)
 sin   b  d  R

where, b: length of SMF, d: initial depth of SMF, T:
thickness of SMF, θ: slope angle, γ: specific gravity of
SMF, Xg: coordinates of the position where the initial
water depth is d, Cm: additional mass coefficient (= 1),
S: moving distance, S0: characteristic distance (= S/2),
R: curvature radius (= b2/8T), ΔΦ: rotation angle (=
2S0/R), t0: characteristic time, η0,2D: maximum water
level decrease amount at X = Xg. In addition, the
symbols of the parameters applied for the Watts model
used here are as shown in Fig. 6.
4.4 Calculation Conditions and Detailed Parameters
Tsunami propagation simulation was carried out on
the above introduced calculation schemes, applying
40 Evaluation of Height of Tsunami Induced by Submarine Landslide

Table 2 Calculation conditions applied to the KLS model

in area B.
Parameters Settings
Grid spacing 450 m
Vertical rupture propagation velocity 1.0 m/s
Collapse duration time 2 min

Fig. 6 Initial water level distribution of submarine

landslide in area B.

Table 1 Parameters applied to the Watts model in area B.

Submarine
Parameter Remarks
landslide area B
γ (-) 1.4 According to literature
Length of sliding mass
b (m) 9,600 Fig. 8 Time series variation of topography change (above)
(map-reading)
Depth at the peak of collapsed and tsunami propagation (below) calculated by KLS model
T (m) 130
part (map-reading) in area B.
Width of collapsed part
w (m) 6,200
(map-reading)
Initial depth of sliding mass
d (m) 700
(map-reading)
Inclination of sliding mass
θ (deg) 1.7
(map-reading)
g (m/s2) 9.8
Cm 1
Distance between peak of
S 15,200 collapsed part and peak of
sedimentary part (map-reading)
S0 7,600 = S/2
Cn 0.086 = S0/(Rcosθ)
R (m) 88,615 = b2/8T
a0 (m2/s) 0.140 = S0/t02
t0 (sec) 233 From Eq. (7)
λ0 (m) 19,292 From Eq. (8) Fig. 9 Distribution of topography change amount input to
ΔΦ (rad) 0.172 2S0/R the KLS model in area B.
umax (m/s) 32.63 = S0/t0
the parameters of each of the submarine landslides in
ΔX (m) 9,646 = λ0/2
κ’ 0.625 areas A to C established in Chapter 3. The parameters
η0,2D (m) 9.79 From Eq. (9) and initial water level distribution of area B in the
η0,3D (m) 2.38 From Eq. (6) Watts model are shown in Table 1 and Fig. 7. Table 2
shows the calculation conditions applied to the KLS
model in area B, and Fig. 8 shows the distribution of
topography change amount input to the KLS model.
Fig. 9 shows the time series variation of topography
change, and tsunami propagation from immediately
after the topography change induced by the sliding
until about 16 minutes after the topography change is
Fig. 7 Symbol of the parameters applied for the Watts
model. completed.
 Evaluation of Height of Tsunami Induced by Submarine Landslide
According to this, it can be found that until about 4
minutes after the beginning of the landslide, only the
collapse advances, without any sedimentation, and
correspondingly the water level is unilaterally
depressed. The water level then begins to rise, after 8
minutes, synchronously with the beginning of
sedimentation.
For each calculation scheme, the general method of
non-linear long wave theory (shallow water theory)
computed by differentiating the staggered Leap-frog
scheme was applied to calculate tsunami propagation.
The minimum space grid size: 3.125 m, time step
for calculation: 0.05 s, computation time: 3 h,
Manning’s roughness coefficient: 0.03, tide level for
tsunami height evaluation: T.P. + 0.49 m (considering
mean HWL for upside calculation) and T.P.-0.01 m
(considering mean LWL for downside calculation) are
applied as calculation conditions respectively.
5. Calculation Result and Consideration
The results of the tsunami height evaluation on the
front of the intake of the Takahama NPP are
summarized in Table 3. This shows that the evaluation
result by the KLS model has a high impact on the
evaluation point in both cases.
Here, regarding the validity of the tsunami height
evaluation, we conducted a reproduction calculation
of the submarine landslide model experiment
implemented by Ref. [10] using the KLS model and
the Watts model [9]. According to this, the KLS model
has an excessive result of about 1.5 to 3 times the
tsunami height compared with the experimental result
in all verification cases. On the other hand, the
reproducibility of the Watts model was good, especially
in the case where the slope angle was relatively small
(6° or less). In this study, based on the slope angle of
area B being as little as 1.7°, it is considered that the
Watts model is more appropriate than the KLS model
from the viewpoint of tsunami height evaluation.
The reason for the calculation result using the KLS
model being excessive compared with the
41
experimental result, is presumed to be that the
sediment volume between the collapsed part and
sedimentary part is imbalanced in the landslide
duration time. In the case where the KLS model was
applied to the submarine landslide, collapse proceeds
until the front edge of the sliding mass reaches the
sedimentary part, since a single rupture starting point
is set in the collapsed part. This is due to the fact that
the KLS model is fundamentally a computational
model devised for tsunami height estimation induced
by onshore landslide (collapse of volcanic edifice).
In order to solve this matter, we proposed the
modified-KLS model in Ref. [9]. This is a model
designed such that collapse and sedimentation proceed
simultaneously, to keep the balance of the
sedimentation volume between the collapsed part and
the sedimentary part in the landslide duration, by
setting the rupture starting point at two locations both
in the collapse part and in the sedimentary part, where
the landslide velocity applied to the modified-KLS
model was set to 1/2 of the KLS model from the
consideration that the landslide completion time
should be the same for both models. As a result, the
modified-KLS model showed that the experimental
results can be reproduced well from the viewpoint of
tsunami height evaluation.
Based on the above findings, the calculation results
obtained by applying the modified-KLS model to the
submarine landslide in Area B are summarized in
Table 4. According to this result, tsunami height
calculated by the modified-KLS model has the
equivalent result as the Watts model at both the
highest and the lowest water level, and only the
result calculated by the KLS model obviously became
large.
It is not easy to reasonably set the landslide velocity
and the rupture duration time in practice, since there
are no detailed observation records of the submarine
landslide, as pointed out by Ref. [9]. On the other
hand, applying the sliding velocity and the rupture
duration time estimated by the method of Ref. [8], it
42 Evaluation of Height of Tsunami Induced by Submarine Landslide

Table 3 Results of the tsunami height evaluation by the (3) As a result of applying the modified-KLS model
modified-KLS model.
to the study case where the maximum tsunami height
Tsunami height
Evaluation was obtained, the result obtained was almost
Tsunami source Highest Lowest
model
(T.P. m) (T.P. m) equivalent to the result from the Watts model, and
Submarine Modified
Area B 1.9 -1.0 only the KLS model was obviously larger.
landslide KLS
From the above, the KLS model, applying the
Table 4 Results of the tsunami height evaluation. maximum sliding velocity calculated by the Watts
Tsunami height model as a parameter, was considered to provide a
Evaluation
Tsunami source Highest Lowest
model larger result than the other two models or
(T.P. m) (T.P. m)
Watts 0.6 -0.1 experimental results. Therefore, it is considered that
Area A
KLS 1.7 -0.9 evaluation results on the safe side can be obtained by
Submarine Watts 1.9 -1.0
landslide
Area B adopting the KLS model for evaluation of tsunami
KLS 3.4 -3.0
Watts 1.6 -1.0 height induced by submarine landslide.
Area C
KLS 2.8 -1.7
References
was suggested that from the view point of tsunami [1] Ikehara, K., Katayama H., and Sato, M. 1990.
height evaluation, approximately reasonable results Sedimentological Map Offshore of Tottori, Marine
are obtained by the Watts model and the Geological Map, No. 36, Geological Survey of Japan.
[2] Tokuyama, E., Honza, E., Kimura, M., Kuramoto, S.,
modified-KLS model. Also, sufficiently safe results
Ashi, J., Okamura, Y., Arato, H., Itoh, Y., Soh, W.,
were considered to be obtained by the KLS model. It Hino, R., Nohara, T., Abe, H., Sakai, S., and
can be said that the evaluation scheme for submarine Mukaiyama, K. 2001. “Tectonic Development in the
landslide-induced tsunami height proposed in this Regions around Japan since Latest Miocene.” Journal of
Japanese Society for Marine Survey and Technology 13
study is practically applicable.
(1): 27-53.
6. Conclusions [3] Yamamoto, H., Joshima, M., and Kishimoto, K. 1989.
Geological Map Offshore of Tottori, Marine Geological
In this study, we estimated the scale of submarine Map, No. 35, Geological Survey of Japan.
landslide from a literature survey and showed [4] Katayama, H., Sato, M., and Ikehara, K. 1993.
Sedimentological Map Offshore of Kyō-ga-Misaki,
examples of tsunami height evaluation using multiple
Geological Map, No. 38, Geological Survey of Japan.
schemes. The main conclusions are as follows. [5] Yamamoto, H. 1991. “A Submarine Sediment Slide on
(1) A procedure was proposed to set submarine the Continental Slope off Fukui Prefecture, Southern
landslide properties, taking into account uncertainties Japan Sea.” Bulletin of the Geological Survey of Japan 42
(5): 221-32.
of location and scale. For the evaluation target of
[6] Satake, K., and Kato, Y. 2002. “The 1741 Japan Sea
tsunami induced by submarine landslide in the waters Tsunami Caused by Debris Avalanche in the
around Wakasa Bay, the largest landslide topography Toshima-Oshima Volcano.” Kaiyo Monthly Special (28):
was selected in each of three areas. 150-60.
[7] Grilli, S. T., and Watts, P. 2005. “Tsunami Generation by
(2) Two schemes were applied to calculate the
Submarine Mass Failure. I: Modeling, Experimental
initial tsunami distribution, and the setting procedure Validation, and Sensitivity Analysis.” Journal of
for the necessary parameter was proposed for each Waterway, Port, Coastal, and Ocean Engineering 131 (6):
scheme. As a result of the tsunami height evaluation 283-97.
[8] Watts, P., Grilli, S. T., Tappin, D. R., and Fryer, G. J. 2005.
by the two schemes of the Watts model and the KLS
“Tsunami Generation by Submarine Mass Failure. II:
model, the result obtained by the KLS model was Predictive Equations and Case Studies.” Journal of
comparatively large for every case. Waterway, Port, Coastal, and Ocean Engineering 131 (6):
 Evaluation of Height of Tsunami Induced by Submarine Landslide 43

298-310.
[9] Tonomo, K., Shikata, T., and Murakami, Y. 2015.
“Accuracy Study of Numerical Simulation of Tsunami
Applied to the Submarine Landslide Model.” Journal of
JSCE, Ser. B3 (Civil Engineering in the Ocean) 71 (2):
I_557-I_62.
[10] Hashimoto, T., and Dan, K. 2008. “Experimental Study
on Submarine Landslide Tsunami in Various Landslide
Mass Shapes.” Abstract of JSCE Annual Meeting 63 (2):
395-6 (in Japanese).
Journal of Energy and Power Engineering 12 (2018) 44-56
doi: 10.17265/1934-8975/2018.01.006
D DAVID PUBLISHING

Characterization of Photovoltaic Modules under Arid

Environments

Abubaker Younis1, 2, Esam Elsarrag1, Yousef Alhorr1 and Mahmoud Onsa2

1. Gulf Organization for Research and Development, Qatar Science and Technology Park, Tech 1 Level 2, 203, Qatar
2. Department of Mechanical Engineering, Faculty of Engineering, University of Khartoum, Khartoum, Sudan

Abstract: Performance characteristics data of solar photovoltaic (PV) cell/module are conventionally obtained under standard testing
conditions. In the present work, the performance of PV modules under extreme temperatures and insolations experienced in State of
Qatar was utilized to aid presenting a simplified characterization approach for the special case of arid environmental conditions. The
chosen model was the well-known single diode model with both series and parallel resistors for greater accuracy. The modeling
technique was validated by comparing the numerically calculated electrical characteristics with experimentally obtained data using two
approaches: a single indoor fixed monocrystalline PV module inside a solar simulation chamber which physically simulated different
weather scenarios by changing irradiation intensities and temperature, and a set of outdoor fixed polycrystalline PV modules. The result
of the indoor experiment was presented in form of performance curves, and the outdoor experiment results in a monthly accumulated
power production chart format. Both illustration types showed acceptable tolerance.

Key words: Photovoltaics, electrical characterization, mathematical modeling.

1. Introduction Elias et al. [7] developed a model for an ideal solar

cell of a PV module using MATLAB software. The
A solar photovoltaic (PV) cell/module has been
simulation was carried out to evaluate the influence of
always defined as the semiconductor device that
the variation of solar cell temperature, solar irradiation,
converts sunlight into electricity, and the modeling and
diode ideality factor and energy gap on I-V and P-V
simulation of its electrical behavior is common practice,
characteristics of the PV cell. Petkov et al. [8] used
and has been widely reported. In reported literature,
classical and modified single-diode models to model
various approaches have been employed to express the
the electrical behavior of PV cells while using Mathcad
electrical behavior of solar cells/modules.
software. A similar approach was also used by Anku et
Rauschenbach [1], Townsend, [2], Eckstein [3], and
al. [9] who introduced an optimization model to ensure
Schroder [4] reviewed several models and their utility
efficient use of PV modules by building a single diode
for system design purposes. Single diode model for PV
model using blocks from the MATLAB/Simulink
cell electrical behavior modeling has been studied
library. Patel and Sharma [10] and Fares et al. [11] also
intensively among authors [5], work was focused on
depended on MATLAB/SIMULINK for modeling,
single-diode model and I-V and P-V characteristics
simulation and implementation of the solar
were presented in function of series resistance, parallel
photovoltaic cell. The characteristics were obtained for
resistance, temperature and the irradiation. Bikaneria et
both the single cell module and the whole PV module.
al. [6] presented one diode PV cell model, the theory,
The effect of the surrounding environmental
the construction, and working of PV cells.
conditions was also tackled numerously in the
literature with Muralidharan [12], in particular, Ref.
Corresponding author: Esam Elsarrag, Ph.D., professor,
[12] analyzed the effect of varying physical and
research fields: building and energy, renewable energy, built
environment. Email: elsarrag@hotmail.com. environmental factors on the I-V characteristics of a
 Characterization of Photovoltaic Modules under Arid Environments 45

PV cell. Ghosh et al. [13] defined a circuit-based

simulation model for a PV module to obtain the
electrical behavior of a PV module with respect to
changes in environmental parameters of temperature
and irradiation.
In recent years, substantial progress has been made
Fig. 1 The equivalent circuit of single diode model for a
in the development of mathematical models which solar PV cell.
sufficiently describe the electrical system of PV
Then the photocurrent equation representing any
modules. However, for the accurate prediction of the
operational conditions is [15]:
electrical behavior of PV modules, a comprehensive
and precise model is of interest thus far. Particularly , ∆ (2)
one which could be used in a variety of climates. This
where is reference value of solar irradiation
article presents a model and a solution algorithm,
equals to 1,000 W/m2, , is the photocurrent at
which can be used to characterize the performance of a
reference conditions (Standard Testing Conditions)
PV cell/module under extreme environmental
and it could be approximated to the reference value of
conditions of irradiation and temperature. The
the short circuit current , when PV cell is short
described approach was continually validated against
circuited according to Bellia et al. [15], then:
experimentally obtained data to further improve the
model and its accuracy. , , (3)

2. Modeling and Solution Algorithm Temperature difference ∆ , , where

the reference temperature , 298 .
The equivalent electrical circuit of a PV cell in the
The well-known Diode reverse saturation current
single diode model is a single diode circuit shown in
equation is according to Chan et al. [14]:
Fig. 1. The circuit consists of photocurrent source,
diode, shunt or parallel resistor and a series resistor in (4)
the load branch.
where is called the diode diffusion factor and is
Then the current and voltage (I-V) characteristics of
the material band gap energy (1.11 eV for Si at 300 K
the single diode model-derived using Kirchhoff law are
[16]).
given by Chan et al. [14]:
When dividing the reverse saturation current Eq. (4)
by itself at STC conditions, it becomes:
(1)
1
,
where light or photocurrent , diode reverse ,
(5)
saturation current , and shunt or parallel resistance ,
are unknown parameters. The respected equation
Ahmad et al. [17] have introduced approximations
for each of the three parameters is a function of cell
to , :
temperature, and absorbed solar radiation.
Light current or photocurrent equation is a ,
(6)
, ,
function of solar irradiation , cell temperature
and short circuit current temperature coefficient . Then merging Eq. (6) into Eq. (5) will result in:
46 Characterization of Photovoltaic Modules under Arid Environments

, where is Boltzmann’s constant 1.381×10-23 J/K and

,
, is the electronic charge 1.602 10 .
(7)
is the cell temperature. Ahmad et al. [17] have
,
suggested that the chosen value of when
is the ideality factor and its value differs based on substituted should automatically make the computed
photovoltaic cell technology. The chosen values for maximum power equals the experimental maximum
purpose of this study are 1.2 and 1.3 for monocrystalline power at its reference conditions or standard test
and polycrystalline respectively by Tsai et al. [18]. conditions, then the next equation is:
is the number of cells in series, is called the , , ,
thermal voltage and is given by the equation according , ,
1
to: (9)
, ,
(8)

And hence will be:

, ,
,
, (10)
, , ,
, , ,
,

Electrical and thermal characteristics values are has been considered by the equations:
taken from manufacturer data sheet provided for
, 1 , (11)
RNG-50D module (50W Monocrystalline Solar Panel)
demonstrated in the following tables.
, 1 , (12)
Effect of temperature on the open circuit voltage,
short circuit current, and the optimum operating power

, , 1 , (13)

Then updating Eqs. (7) and (10) by Eqs. (11)-(13):

, (14)
,

, ,

, , , (15)
,

Preliminary analysis to voltage readings obtained approximated by natural logarithm function of time
from the experimental testing showed that voltage :
variation profile could be approximated to the closest (16)
known nonlinear function profile which is the natural For numerical solution simplification purposes, and
logarithmic function. due to its small magnitude compared with the parallel
Then the voltage generating function could be resistance , series resistance is neglected.
 Characterization of Photovoltaic Modules under Arid Environments 47

Table 1 Renogy’s RNG-50D electrical, thermal and Table 2 Solarturk energi STR60 electrical and thermal
mechanical characteristics. characteristics.
Electrical Data Electrical Data
Maximum power at STC 50 W Maximum power at STC 250 W
Optimum operating voltage (Vmp) 18.5 V Optimum operating voltage (Vmp) 29.98 V
Optimum operating current (Imp) 2.7 A Optimum operating current (Imp) 8.34 A
Open circuit voltage (Voc) 22.7 V Open circuit voltage (Voc) 37.41 V
Short circuit current (Isc) 2.84 A Short circuit current (Isc) 8.79 A
Module efficiency 14.67% Module efficiency 15.3%
Thermal characteristics Thermal characteristics
Operating module temperature 40 to +90 Operating module temperature 40 to +120
Nominal operating cell temperature (NOCT) 47 2 Nominal operating cell temperature
47 2
Temperature coefficient of Pmax ( ) 0.23 %⁄°C (NOCT)
Temperature coefficient of Pmax ( ) 0.43 %⁄
Temperature coefficient of Voc ( ) 0.33 %⁄
Temperature coefficient of Voc ( 0.32 %⁄
Temperature coefficient of Isc ( ) 0.05 %⁄ )
Temperature coefficient of Isc ( ) 0.05 %⁄
Number of cells 36 (4 × 9)
Number of cells 60 (6 × 10)
Solar cell type Monocrystalline
Solar cell type Polycrystalline
Löper et al. [19] have suggested adding a constant characteristics converging to the experimental ones.
offset to the open circuit voltage as fit parameter The constant offset is open circuit voltage at STC
so to obtain from the numerical solution electrical conditions , , then Eq. (11) will be:

* *
, 1 , 11

Updating Eqs. (14) and (15) by Eq. 11 , the new statement of the diode reverse saturation current is:
*
*
, 14
,

And for the resistance in parallel:

, ,

, (15) *
, ,
,

Numerical solution for the implicit nonlinear I-V
characteristics Eq. (1) is of interest, so to compute the
relevant unknown parameters and then to evaluate
current and voltage characteristics.
The numerical solver used here is a combination of
bisection, secant and inverse quadratic interpolation
methods, coded in MATLAB software by the
command FZERO Mathworks (2016).
The equations’ solving pattern or solution algorithm
used to numerically compute I-V characteristics is then
translated to MATLAB command lines which displays
results in form of figures and graphs. Fig. 2
demonstrates the solution algorithm flow chart.
3. Test Setup and Procedure
3.1 Indoor Setup
A solar/weather simulation chamber, which is a
controlled environment is used for characterization of
the PV module and validation of the numerical solution
results.
The testing was conducted in an Atlas SEC 1100
Solar Simulator Chamber located in GORD’s research
facility (TechnoHub) at Qatar. The SEC 1100 has a
chamber size of 1,100 liters with an effective radiation
48 Characterization of Photovoltaic Modules under Arid Environments

Fig. 2 Flow chart demonstrates the solution sequence of the single diode model to obtain electrical characteristics.

area of 5,600 cm2 and acts as a perfect test chamber to
characterize small to medium sized PV panels. The
integrated Metal Halide Global (MHG Model: 4,000 W
Luminary) conforms with IEC 60904 Class A
guidelines and provides quality solar simulation with a
close spectral match to natural sunlight combined with
high irradiation efficiency and spatial uniformity in the
test area. The MHG lamp provides adjustable output up
to a maximum of 1,200 W/m2 with a spectral range of
280-3,000 nm.
The parameters that can be controlled in the chamber
include air (ambient) temperature, irradiation, relative
humidity via a touch screen control panel located on
the front of the chamber. The device also has logging
capabilities for theses parameters along with a 6
channel DAQ for temperature sensors, irradiation
measurement and Black Standard Temperature control
in real time. The output is saved in csv format as well as
displayed on screen graphically. Multi-stage safety
features ensure that the chamber operates in the preset
limits and ensure prolonged testing durations.
The response of the PV at two different irradiation
levels (700 W/m2 & 1,200 W/m2) of the sun was tested.
In each set of irradiation values, the temperature of the
chamber was varied to achieve different operating
temperatures for the PV module, thus dividing the test
to multiple tests with different combinations of
Irradiation and PV Temperatures. Each test consisted
of measuring the open circuit voltage Voc, short circuit
current Isc and output parameters (voltage & current) of
the PV at varying load resistances ranging from 1 Ohm
to 100 Ohm.
The load resistance was simulated using 50 W power
resistors of values 1 Ohm, 4 Ohm and 8 Ohm with
 Characterization of Photovoltaic Modules under Arid Environments
multiple resistors being connected together in series
combinations via a custom-made micro-controller
based relay network to achieve different progressive
resistive values from 1 to 100 Ohm in steps of 1 Ohm.
The output of the PV was measured for consistency by
keeping each value of resistance constant for 5 minutes
under the same conditions. The value was limited at
100 Ohm due to the reason that the PV performance
degraded exponentially after the 10 Ohm mark and
adding more resistance was not relevant to the test
which was trying to validate a model which predicts a
practical scenario, where the maximum possible
resistance values would not go beyond a dozen or so
ohms.
The temperature of the PV was monitored at 6
different points of the PV backside and the average
temperature was considered as the PV temperature.
The measurement was done by Class B PT 100
temperature sensors connected to the monitoring
system of the Atlas SEC 1100 solar simulator chamber
logger which gave out real time outputs of the
parameters being monitored except the power output of
the PVs. Fig. 3 shows the monocrystalline PV inside
the simulator and the PT100 sensors attached at its
backside. The power output of the PV was measured
using a separate micro-controller based system linked
to the resistance simulator and used an
ACS712ELECTR-05B-T based current sensor and
voltage divider coupled with 10 bit ADC of the
microboard, which provided resolutions of 0.009 A for
the current and 0.005 V for the voltage. The data were
logged in a csv file format for analysis and the power
curves obtained at different test parameters were
validated against the simulated curves from the model.
The humidity was set constant at 30%.
The irradiation at the plane of the PV panel was
measured using an ISO 9060 first class compliant Kipp
& Zonen CMP 6 pyranometer (in the SEC) which can
measure up to 2,000 W/m2 in a spectral range of 285 to
2,800 nanometers with a sensitivity of 12.29×10-6
µV/W·m-2 and a 180° field of view. Fig. 4 summarizes
49
the schematic diagram of the full setup.
3.2 Outdoor Setup
A set of four 250 W Polycrystalline PV modules was
mounted outdoor in the GORD TechnoHub research
facility. Fig. 5 photographically demonstrates the setup.
The monitoring and logging of the electrical
characteristics were done by designing and building a
custom-made microcontroller based Smart Data
Logging system which measured each parameter with
specific sensors. The logging was done in real time
high frequency samples and saved to a memory card in
CSV format for easy analysis. The current was
measured using ACS712 based current sensors, which
can measure DC current in both directions with a range
of 20 Amps, and an accuracy of 1.5%. The recording
and monitoring was done automatically by the logger
and the data were saved in a memory card with all
information per day saved in one file with the date and
timestamp. The recording was done from sunrise to
sunset using the RTC information and was turned on
during the day and off during the night. The weather
data for the year of 2017 were collected using weather
station fixed in the same research facility.
4. Results and Discussions
The following set of figures present many selected
current vs. voltage (I-V) and power vs. voltage (P-V)
curves, which compare experimental characteristics
belonging to indoor test with the simulated
characteristics. The values of temperatures and
irradiation intensities are determined randomly
according to suitability and availabilities at the testing
facility. Graphs are generated by MATLAB software.
The process of validating the new model results
against the experimentally obtained results is visually
demonstrated in Figs. 6 and 7 which present data in I-V
curve format. A decent identicality between the
validation and simulation curves is observed, also the
difference in magnitude of short circuit current
between the two cases is barely distinguishable.
50 Characterization of Photovoltaic Modules under Arid Environments

(a) (b) (c)

Fig. 3 Photographic views of RNG-50 PV module placed inside the solar simulator and the PT100 sensors are attached to its
backside.

Fig. 4 Schematic diagram for the PV module testing setup at GORD’s Technohub.
 Characterization of Photovoltaic Modules under Arid Environments
Fig. 5 Photographic view of outdoor setup for testing
polycrystalline PV module at GORD’s Technohub.
Complying with the I-V figures trend, Figs. 8 and 9
show almost identical simulation and validation curves
Fig. 6 Current vs. voltage at 95 °C and 1,100 W/m2.
Fig. 7 Current vs. voltage at 66 °C and 900 W/m2.
51
for the pre- and post- maxima region, while a
non-drastic displacement between the two curves
occurred at the maxima region. It is also clearly
illustrated that the precision of the results is within an
acceptable range, although further considerations
would be of interest to maximize the model accuracy,
so when including other environmental parameters
there would be no effect on the total tolerance
percentage.
Typical P-V behavior of the PV cell/module was
maintained as illustrated in Figs. 10-12, which show
performance curves generated from the new model
following the same trend of the conventional theoretical
curves with wider range of temperatures and fixed
52 Characterization of Photovoltaic Modules under Arid Environments

Fig. 8 Power vs. voltage at 95 °C and 1,100 W/m2.

Fig. 9 Power vs. voltage at 66 °C and 900 W/m2.

Fig. 10 Power vs. voltage at different temperatures and 250 W/m2.

 Characterization of Photovoltaic Modules under Arid Environments 53

Fig. 11 Power vs. voltage at different temperatures and 500 W/m2.

Fig. 12 Power vs. voltage at different temperatures and 1,000 W/m2.

Fig. 13 Current vs. voltage at different temperatures and 500 W/m2.

54 Characterization of Photovoltaic Modules under Arid Environments

Fig. 14 Current vs. voltage at different temperatures and 250 W/m2.

Fig. 15 Current vs. voltage at different temperatures and 1,000 W/m2.

160
140
Total kWh per month

120
100
80
60
40
20
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Measured Simulation

Fig. 16 Total power produced by 4 polycrystalline PV modules in GORD’s Technohub for 2017.
 Characterization of Photovoltaic Modules under Arid Environments
1,800
1,600
1,400
Power (kWh per year)
1,200
1,000
800
600
400
200
0
Measured
Fig. 17 Annual measured power production compared with the simulated power for the polycrystalline PV set in the year
2017.
irradiance. The effect of temperature demonstrated in
the previous figures complies with the governing
mathematical equations which explicitly include
temperature terms that by its turn, reduce
proportionally the open circuit voltage and the
maximum output power with the temperature increase.
As the mentioned typicality of P-V curves also
implies on the computationally generated I-V
performances charts in Figs. 13-15, it is distinctly
presented that the temperature has almost no effect on
the short circuit current magnitude and that perfectly
matches the literature.
The real-life applicability of the visualized
convergence in the corresponding curves presented
earlier was tested by using the outdoor setup data
collected for the year of 2017. Fig. 16 shows bar chart
of the monthly accumulated power produced by the 4
PV panels drawn against the simulated power
production.
As shown in Fig. 17 the difference between annual
simulated and measured power performance rests
within acceptable range of exactly 8%.
5. Conclusions
The I-V and P-V characteristics resulted from the
numerical solution of the electrical model presented in
this work were validated against experimentally
55
Simulated
obtained data from both monocrystalline and
polycrystalline PV modules placed indoor and outdoor
respectively at multiple combinations of the input
parameters the model could accept, namely, the
irradiance and temperature. The presented graphs have
shown sufficient convergence and very acceptable for
non-delicate applications of the PV panels,
annual-wise the prediction capability of the modeling
method was responsive as it clearly takes in
consideration the heat and light intensity effects. Hence
the model and the solution algorithm could be
described as reliable ones.
References
[1] Rauschenbach, S. 1980. Solar Cell Array Design
Handbook—The Principle Technology of Photovoltaic
Energy Conversion.
[2] Townsend, T. U. 1989. “A Method for Estimating the
Long-Term Performance of Direct-Coupled Photovoltaic
Systems.” MS Thesis.
doi:10.1016/0038-092X(84)90010-0.
[3] Eckstein, J. 1990. “Detailed Modelling of Photovoltaic
System Components.” MS Thesis.
[4] Schroder, D. K. 2006. Material and Device
Semiconductor Material and Device. 3rd ed. Physics
Today. doi:10.1063/1.2810086.
[5] Azzouzi, M., Popescu, D., and Bouchahdane, M. 2016.
“Modeling of Electrical Characteristics of Photovoltaic
Cell Considering Single-Diode Model.” Journal of Clean
Energy Technologies 4 (6): 414-20.
56 Characterization of Photovoltaic Modules under Arid Environments
doi:10.18178/JOCET.2016.4.6.323.
[6] Bikaneria, J., Prakash Joshi, S., and Joshi, A. 2013.
“Modeling and Simulation of PV Cell Using One-Diode
Model.” International Journal of Scientific and Research
Publications 3 (10): 2250-3153.
[7] Elias, B. H., AlSadoon, S. H. M., and Abdulgafar, S. A.
2014. “Modeling and Simulation of Photovoltaic Module
Considering an Ideal Solar Cell.” International Journal of
Advanced Research in Physical Science 1 (August): 9-18.
[8] Petkov, M., Markova, D., and Platikanov, S. 2011.
Modelling of Electrical Characteristics of 2.
doi:10.5767/anurs.cmat.110202.en.171P.
[9] Anku, N. E. L., Kankam, A., Takyi, A., and Amponsah, R.
2015. “A Model for Photovoltaic Module Optimisation.”
Journal of Mechanical Engineering and Automation 5 (2):
72-9. doi:10.5923/j.jmea.20150502.02.
[10] Patel, J., and Sharma, G. 2013. “Modeling and Simulation
of Solar Photovoltaic Module Using Matlab/Simulink.”
International Journal of Research in Engineering
Technology 2 (3): 225-8.
[11] Fares, M. A., Atik, L., Bachir, G., and Aillerie, M. 2017.
“Photovoltaic Panels Characterization and Experimental
Testing.” Energy Procedia 119 (July): 945-52.
doi:10.1016/j.egypro.2017.07.127.
[12] Muralidharan, R. 2014. “Mathematical Modeling,
Simulation and Validation of Photovoltaic Cells.”
International Journal of Research in Engineering and
Technology 3 (10): 170-4.
[13] Ghosh, S., Shawon, M., Rahman, A., and Abdullah, R.
2013. “Modeling of PV Array and Analysis of Different
Parameters.” International Journal of Advancements in
Research & Technology 2 (5): 358-63.
[14] Chan, D. S. H., Phillips, J. R., and Phang, J. C. H. 1986.
“A Comparative Study of Extraction Methods for Solar
Cell Model Parameters.” Solid-State Electronics 29 (3):
329-37. doi:10.1016/0038-1101(86)90212-1.
[15] Bellia, H., Youcef, R., and Fatima, M. 2014. “A Detailed
Modeling of Photovoltaic Module Using MATLAB.”
NRIAG Journal of Astronomy and Geophysics 3 (1): 53-61.
doi:10.1016/j.nrjag.2014.04.001.
[16] Kittel, C. 2005. Introduction to Solid State Physics, 8th ed.
New York: John Wiley.
[17] Ahmad, T., Sobhan, S., and Nayan, F. 2016. “Comparative
Analysis between Single Diode and Double Diode Model
of PV Cell: Concentrate Different Parameters Effect on Its
Efficiency.” Journal of Power and Energy Engineering 4
(3): 31-46. doi:10.4236/jpee.2016.43004.
[18] Tsai, H., Tu, C., and Su, Y. 2008. “Development of
Generalized Photovoltaic Model Using
MATLAB/SIMULINK.” Proceedings of the World
Congress on Engineering and Computer Science 2008
(WCECS 2008), October 22-24, 2008, San Francisco,
USA.
[19] Löper, P., Pysch, D., Richter, A., Hermle, M., Janz, S.,
Zacharias, M., and Glunz, S. W. 2012. “Analysis of the
Temperature Dependence of the Open-Circuit Voltage.”
Energy Procedia 27: 135-42.
doi:10.1016/j.egypro.2012.07.041.