You are on page 1of 46

Technical Paper #6

Method for Determining


Best Economic Pipe Size for
Ammonia Refrigeration Piping

Robert A. Sterling, P.E.


Ammonia Refrigeration Foundation

Abstract

Economic pipe size is an important factor in project design and analysis for any piping system. In the
industrial ammonia refrigeration industry, where energy usage by, and maintenance of, the mechanical
refrigeration system is often one of the largest operating costs, a flexible method for analysis to
minimize the total life-cycle cost of a piping system is desirable.

A method for determining a minimum header life-cycle cost has been developed to meet this need.
Through broad-based piping labor and materials data aggregated by RSMeans (a company that
aggregates construction cost data), data on energy usage by equipment in ammonia refrigeration
systems, and well-established calculations for pressure drop in piping, the method produces a series
of results that allow comparison of life-cycle costs for several pipe sizes in the same service. The
method is flexible enough to apply over a large range of conditions and pipe sizes and is suitable for
incorporating into an automatic software calculation package.

© IIAR 2017 1
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Introduction

The ammonia refrigeration industry currently uses various methods to determine


appropriate pipe sizes for a given service. These range from rules of thumb to table
sizing to, in some cases, sophisticated proprietary calculation software that provides
much information regarding a selected pipe size and its performance in the specified
service.

What does not currently exist is a detailed method, based on verifiable and easily
updatable data, to look at and compare the life-cycle cost of several pipe sizes side
by side, with sufficient flexibility to fit a wide variety of circumstances. Some data do
exist, in many forms, but few are specifically applicable to ammonia refrigeration in
general, and those that are specific are based on assumptions that may or may not be
true.

In light of these circumstances, an updated method has been proposed for


determining a lowest-life-cycle-cost pipe size for ammonia refrigerant piping.
The scope of this method is applicable within a bounded set of rules, which will
be described later in this paper. The method is intended as a tool for use by the
industrial refrigeration industry where such a tool is desired and is thus intended
to be relatively simple for use by a wide audience. The approach is stepwise, and
although it involves significant data lookup and calculation, it is well suited for
performing with a computer.

History

The history of economic pipe size calculation spans the history of industrial piping
itself. A landmark work on the subject was written by Nolte and published in 1978.
This work, of course, builds on the work of others, but is essentially a compendium
of the knowledge current at the time of writing. Nolte’s approach was based on the

Technical Paper #6 © IIAR 2017 3


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

premise that, for most metallic piping materials, costs could be scaled by assuming
them proportional to diameter to the 1.5 power, and the work cites studies from the
era that show this to be true. In the absence of the ubiquitous personal computer,
such an approach was convenient, because it allowed fairly rapid calculation without
the need for extensive knowledge of the project or system.

Life-cycle costs in Nolte’s work were based on factors that could be used to adjust
the results from the base formulas. This allowed flexibility to adjust to changing
conditions. The underlying philosophy of the method also recognized that when an
optimum lay between two pipe sizes, moving upward to the next pipe size is not
always the best choice from an economic standpoint.

Specific to the industrial ammonia refrigeration industry, Richards (1984) began


research on the topic of economic pipe size in the early 1980s. ASHRAE published
other research regarding refrigerant pipe sizing in general (not specific to ammonia;
Richards 1984) much earlier than this, but focused on pressure drop above other
factors.

Much of Richards’ efforts focused on the assertion that the installed cost of piping
is a constant and is proportional to the diameter and length of a particular section.
Thus, Richards’ method applied simple scaling, based on pipe diameter, to find
and compare installed cost of a section (1984). This was in direct contrast to
Nolte’s earlier published work. The reasons for this are not now known, but such
assumptions may prove unnecessary in the presence of the data available today.
The method considered here will avoid making any assumptions regarding scaling
through the use of published data for many pipe sizes.

4 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Need for an Updated Approach

Given the wide availability of real cost data for piping and lingering questions on
fundamental concepts in previous IIAR efforts to determine economic pipe size in
industrial ammonia refrigeration piping, an opportunity exists to investigate a data-
based approach that removes some of the assumptions from calculating an economic
ammonia refrigeration pipe size.

It was therefore determined that research be undertaken to provide such an


approach. The advantages of doing so include
• Removing some simplifying assumptions from the process of determining
economic pipe size and replacing them with data;
• Allowing a comparison to be made between the proposed data-based method and
existing results, at least for ammonia refrigeration piping specifically; and
• Allowing the ready application of such a method to the use of computers to
calculate specific results based on various inputs; this is not an extreme departure
from the previously available methods, but is certainly a necessary outcome of the
vast increase in the use of cost data over the previous methods.

With these advantages in mind the work was undertaken.

Development of the Method

The revised approach, which builds on the previous work (particularly that done by
Richards) while avoiding some of its pitfalls, uses as its basis a set of assumptions,
much like the previous work, but that rely more heavily on data to compare available
choices for pipe diameter.

Technical Paper #6 © IIAR 2017 5


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

The assumptions necessary to perform the analysis are laid out as follows:

1. First cost for any pipe size and schedule may or may not be linearly scalable
with pipe diameter. Therefore, data on the components of first cost are necessary
to guess the aggregated first cost of the representative pipe section accurately.
These components include the following items:
a. Metallic pipe cost, based on diameter and wall thickness;
b. Labor cost for welded piping;
c. Material and labor costs of insulation; and
d. Material and labor costs of paint and coatings (where applicable).

2. Energy consumption through pressure loss in a section of piping can be


calculated based on the inlet conditions and assumed mass flow rate through
the section. This may be similar to the approach in Richards (1984), but the
details of Richards’ analysis are not available as background to understand the
foundation of those results. The background on how the actual values for this
effort are calculated is presented in detail within this paper.

3. The present value sum of items 1 and 2 that results in the lowest present value is
the best economic pipe size.

No method can be a true and complete model for the real cost of ownership for
any particular section of pipe, but these steps represent an approximation based on
the maximum expected flow within that pipe section. Not only are conditions in
any particular section unlikely to be exactly those of the engineering analysis, but
providing more than a guess of what energy costs will be at any point in the future
is not possible. This is an inherent and unavoidable limitation of an economic pipe
size analysis. Therefore, making reasonable assumptions about ownership cost is
considered sufficient, so long as they are uniformly applied in the analysis.

With regard to first cost, other project costs might certainly skew the analysis. Items
such as varying welding techniques (Gas Metal Arc Welding vs. Gas Tungsten Arc
Welding for root pass), shop fabrication versus field fabrication, and nondestructive

6 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

examination could all affect the outcome. However, such factors differ from those
included in the analysis in that, at least at present in the refrigeration industry, they
are not considered standard for field-erected systems. In addition, their costs can vary
widely on a case-by-case basis (i.e., based on the skill of the GTAW welder, number
of joints observed using Non-Destructive Examination, [NDE], etc.). It also bears
mentioning that separate data for these particular items, based on many steel pipe
diameters and wall thicknesses, are not currently available.

Other items that are not considered in the analysis, with explanations of the decision
to exclude them, are as follow:
• Pipe supports (including wind and seismic bracing)
• The range in support techniques and costs is too vast to encompass within the
scope of the analysis. The adequacy of any support arrangement is affected by
a huge number of factors that can’t be envisioned at the level of granularity
involved in sizing a pipe. Industrial refrigeration headers are rarely supported
in a single-pipe-single-support arrangement.
• Varying pipe size by one standard size up or down is unlikely to appreciably
affect budgetary costs for piping supports and would likely not show up as a
measurable impact on cost.
• Valves
• Header piping in typical industrial ammonia plants does not contain a
significant number of valves, if any, other than those in isolating accumulator
and recirculator vessels, which are outside the scope of the analysis.
• Energy users other than pumps and compressors
• Costs for running air units and condensers, while pertinent in an overall
system cost analysis, will have little or no bearing on an individual economic
pipe size compared with that of the compressor and pump energy.
• Identification for piping
• The number and type of labels used in industrial refrigeration piping varies
widely based on location (roof vs. indoors) and arrangement and is assumed
not to be a controlling factor for a general pipe sizing analysis of adjacent-size
piping comparisons.

Technical Paper #6 © IIAR 2017 7


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

The listed items are not expected to skew the analysis in any appreciable way based
on a single stepwise change in pipe size.

With these framework assumptions in place, and assuming that sufficient data can be
obtained for those items included for consideration, then a straightforward analysis
becomes possible. This analysis consists of determining the listed material and
labor costs and summing these with cost of energy, as a net present value, absorbed
throughout the lifetime of the system by a section of piping (through pressure
loss). This sum can be compared with that calculated for adjacent pipe sizes so
that the user can see the net present cost of ownership of the section of piping is in
comparison with other sizes and choose the best one.

Of course, economic pipe size based on the listed factors cannot replace specific
engineering analysis for the project. The method demonstrated here is only intended
to help in determining the proper sizing of horizontal header piping. Branch piping,
vertical risers, and the like are subject to other constraints with regard to pipe size.
The following considerations should be observed when using the analysis:

1. The inputs provided by a user to obtain a result based on economic pipe size
cannot guarantee a working system. Where considerations other than simple
operating cost based on maximum load weigh more heavily on the design,
engineers must attend to these considerations. Operation of the process,
available pressure drop, future flexibility of the plant, and a host of other factors
can affect the final choice of pipe size for any given header.

2. A means to determine whether or not to decrease pipe size based on a smaller


design load downstream of a branch takeoff in the header is not available with
the level of detail proposed here. This is a complex notion depending on how
the reduction is to be constructed and the cost of associated labor and materials.
The method cannot provide guidance to determine how to proceed in such cases.
Information is available in Nolte (1978), but may be antiquated.

8 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

3. The proposed method does not determine whether the section being analyzed
is appropriate for the design pressure, design temperature, and arrangement to
ensure that allowable stresses in accordance with the applicable piping code
are not exceeded. Engineers should be cautious in analyzing only sections
appropriate for the intended service.

4. While pipe sizes selected through use of this method are typically within normal
acceptable limits for flow, pipe sizes may be economically attractive while being
subject to hammer, erosion, choked flow, or cavitation. The user must evaluate
such conditions separately.

Prior to demonstrating the analysis method, some expanded commentary on the


guiding principles of the analysis is appropriate. The following details are pertinent to
explaining the exact approach undertaken here:
• Material cost and labor hours for welded piping:
• The RS Means Facilities Construction Cost Data Book contains data for
welded steel water distribution piping, in ASTM A53/ASTM A106 and 304
SS materials. While the way that individual refrigeration contractors manage
quality and inspection is unknowable, the case is made here that water
distribution piping, which would presumably fall under the purview of ASME
B31.9-2008, is quite similar to ammonia piping with respect to how the latter
is fabricated using ASME B31.5-2013 rules. In fact, many contractors that weld
ammonia refrigerant piping and hot and cold water utility water piping use
the same procedures and assembly techniques for both. In addition, chilled
water piping for refrigeration falls under the requirements of ASME B31.5, and
such piping is likely aggregated into the RS Means numbers. Requirements
for welded piping in ASME B31.9 are well aligned with those in ASME B31.5,
with the only major difference being acceptance of defects where a small
amount of lack of fusion is acceptable under ASME B31.9, but no lack of
fusion is acceptable in ASME B31.5. Both standards handle weld preparation,
tacking, etc., in the same way, and both require visual examination only.
These data are therefore considered appropriate for industrial ammonia piping

Technical Paper #6 © IIAR 2017 9


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

in accordance with ASME B31.5. In addition to this reasoning, the notes in RS


Means specifically state that any mechanical discipline (including heating and
air conditioning) can use the piping data.
• Carbon steel and 304 stainless steel pipes are the only materials considered.
• Material pricing and labor hours in RS Means represent a national average,
which is in the interest of the analysis (the intent is that the analysis be
general, industry-wide, and for the United States).
• RS Means numbers on labor may be somewhat conservative when considering
long, branchless header runs, as welds (or threaded couplings) are assumed
to be on 10 ft centers instead of the 20 ft centers common in industrial
refrigeration systems. However, when considering typical runs of piping
containing branches, the extra welds can account for those real-world piping
arrangements and so are deemed appropriate here.
• Labor hour values have embedded in them the handling, setting, preparation,
and fitting costs, which can vary based on pipe schedule. The available
data are broken down by pipe schedule, so these numbers can be accurately
referenced and little or no extrapolation is necessary. Numbers used in the
analysis will reflect the absence of pipe supports (raw numbers include clevis
supports, but factors are provided for removing these component costs).
• Insulation costs:
• The RS Means data book also provides piping insulation data for mechanical
piping. While the insulation types common in industrial refrigeration
(polystyrene and polyisocyanurate) are not specifically listed, data are
available for calcium silicate, which, for the purposes of this paper, has
been chosen as a substitute in the absence of other data. These data, along
with additional cost and labor data for aluminum jackets (vinyl will not
be considered because most header piping is located outdoors and uses
aluminum jacket), will be used to provide cost comparisons for insulation.
The information is presented in the RS Means data book with limited size
ranges for insulation thickness. Therefore, while this will neither provide a
solution sensitive to temperature and thickness suitable for every user, nor

10 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

allow individual users to tailor their insulation costs to high degree, it avoids
complexity and error in attempting to parse out the insulation costs to a fine
level of detail, for which most users will not be willing to sacrifice ease of use.
• Ammonia properties will be obtained from REFPROP version 9.12 (Lemmon, et al,
2013).
• Energy rates are used as an input to the calculation only as a simple, average $/
kWh rate. Considerations such as demand charges are assumed to be built into
this average rate.
• The charge per kWh is also not adjusted year over year for inflation in this
analysis. This may or may not reflect future energy costs, which are unknowable.
• Concerning pressure drop, installation labor, pipe material costs, and insulation
costs, the analysis considers a 100-ft section of pipe. This fixes the number of
welds and the section length and allows an equal comparison of two or more
sizes.
• Net present value is compounded annually.
• Compressor energy has been calculated using the following assumptions:
• Oil-flooded screw compressor,
• Externally oil cooled, and
• No suction or discharge superheat or pressure drop.
These assumptions were used with two manufacturers’ software programs
to generate energy curves based on the brake-horsepower per ton (BHP/TR)
values for various suction temperatures. The results have been averaged, and a
mathematical equation generated to allow BHP/TR calculations to be performed
within the analysis and used to generate energy penalties based on varying
pressure drops. A saturated condensing temperature of +85°F (for ammonia,
not considering other refrigerants) is used as an average value. In a range of
+75°F to +95°F saturated condensing temperature, BHP/TR results for a
range of suction temperatures, when compared with an assumed condensing
temperature of +85°F, have a maximum of +20% error at 95°F and -18% error
at 75°F for a single-stage (no economizer) compression process and a maximum
of +10% at +95°F and -8.4% at +75°F for two-stage intercooled arrangement

Technical Paper #6 © IIAR 2017 11


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

at a fixed intermediate temperature of +20°F. These results suggest that the use
of an average condensing temperature is appropriate within the accuracy of the
calculations. See Table 11 in the Appendix for compressor horsepower values
that may be used with suction temperatures for this analysis and for curve fits
suitable for use in a software package. Note that curve fits are based on the ratio
of suction to discharge pressure, not temperature.
• Compressor motor efficiency has been assumed at 95%.
• Power consumption for liquid lines is based on calculated hydraulic horsepower
at the given operating temperature divided by an assumed pump efficiency (60%)
and motor efficiency (95%).
• Given that safety factors would be uniformly applied as a multiplier for a total
unit cost of ownership for each pipe size, which would only serve to exaggerate
calculated differences, safety factors are not considered.

Analysis

With this explanation of background materials complete, the analysis can be


demonstrated. As mentioned previously, a stepwise approach is used. It is described
as follows.

Step 1 – Define analysis parameters

The following parameters are required for input to the analysis:

1. Mass flow rate: This can be either directly input or determined from mass
flow required for a particular heat absorption and service (evaporator suction,
compressor suction, liquid feed, compressor discharge, etc.);

2. Flow type:

a. Liquid,

12 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

b. Vapor, or

c. Two-phase (with overfeed rate specified by user);

3. Pipe material;

4. Energy cost (per kWh);

5. Hours of operation per year;

6. Labor cost (per hour); and

7. Assumed rate of inflation (on a yearly basis).

These inputs form the basis of any calculation for optimum pipe size.

Step 2 – Size the baseline case (pressure drop and power consumption)

Determine an appropriate baseline size based on a guess for allowable pressure


drop in a horizontal header. For this analysis, this baseline is determined using the
following limits:

1. For liquid piping, less than or equal to 2.00 psi drop through a 100-ft section of
piping;

2. For vapor and two-phase piping at temperatures above -20°F, less than or equal
to 0.50 psi pressure drop through a 100-ft section of piping; and

3. For vapor and two-phase piping at temperatures at or below -20°F, less than or
equal to 0.25 psi pressure drop through a 100-ft section of piping.

Note that these baseline limits do not always provide for a base case that is within
one size of the most economical pipe size for the service, but this is heavily
dependent on usage, energy rates, and system life, the combination of which is
difficult to correct for on the front end of the analysis.

Technical Paper #6 © IIAR 2017 13


header.  For  6.  
this  Labor   cost  
analysis,   (per  
this   hour);  ias  nd  
baseline   determined  using  the  following  limits:  
7.   Assumed  rate  of  inflation  (on  a  yearly  basis).  
1.   For  liquid  piping,  less  than  or  equal  to  2.00  psi  drop  through  a  100-­‐ft  section  of  piping;  
These  
2.   For   inputs  
vapor   and  form   the  basis  
two-­‐phase   of  any  
piping   at  ctemperatures  
alculation  for  aobove  
ptimum   pipe  
-­‐20°F,   size.  
less   than  or  equal  to  0.50  ps
pressure  drop  through  a  100-­‐ft  section  of  piping;  and  
Step  
2017 IIAR 2  –  Size  
Natural the  baseline  
Refrigeration Conferencecase  &(pressure  
Heavy Equipmentdrop  Expo,
and  pSan ower  Antonio, consumption)  
TX
3.   For  vapor  and  two-­‐phase  piping  at  temperatures  at  or  below  -­‐20°F,  less  than  or  equal  to  0.
Determine  an  appropriate  baseline  size  based  on  a  guess  for  allowable  pressure  drop  in  a  h
pressure  drop  through  a  100-­‐ft  section  of  piping.  
header.  For  this  analysis,  this  baseline  is  determined  using  the  following  limits:  
Note  that  these  baseline  limits  do  not  always  provide  for  a  base  case  that  is  within  one  size  of  the  m
Liquid 1.   For  liquid  piping,  less  than  or  equal  to  2.00  psi  drop  through  a  100-­‐ft  section  of  pip
economical  pipe  size  for  the  service,  but  this  is  heavily  dependent  on  usage,  energy  rates,  and  syste
2.   For  vapor  and  two-­‐phase  piping  at  temperatures  above  -­‐20°F,  less  than  or  equal  to
life,  the  combination  of  which  is  difficult  to  correct  for  on  the  front  end  of  the  analysis.  
pressure  drop  through  a  100-­‐ft  section  of  piping;  and  
Liquid pressure Liquid  
drop is based on vpure
3.   For   apor  aliquid flow, without
nd  two-­‐phase   piping  at  flashing.
temperatures   Liquid at  or  flow is-­‐20°F,  less  than  or  eq
below  
determined by first using
Liquid   the dColebrook
pressure   pressure   equation
drop  
rop  is  based   otn  
hrough   ato find
 100-­‐ft  
pure  liquid   flow,  sthe friction
ection   factor,
of  pflashing.  
without   iping.   ƒ. The
Liquid   flow  is  determined  by  fir
Colebrook equation
using  has
the  Cthe form:equation  to  find  the  friction  factor,  𝑓𝑓.  The  Colebrook  equation  has  the  form:  
olebrook  
Note  that  these  baseline  limits  do  not  always  provide  for  a  base  case  that  is  within  one  size
economical  
" pipe  size  
+ for  the  service,  but  this  is  heavily  dependent  on  usage,  energy  rates,  a
1.2"
=
life,   t −2
he   c log ,
ombination  
"* +f  which    is   (1)
o (1)  
difficult  to  correct  for  on  the  front  end  of  the  analysis.  
# -./ 34 #

Where Where   Liquid  
Liquid  pressure  drop  is  based  on  pure  liquid  flow,  without  flashing.  Liquid  flow  is  determine
𝑓𝑓   using  tis  
he  the   friction  feactor  
Colebrook   quation  to  find  the  friction  factor,  𝑓𝑓 .  The  Colebrook  equation  has  the
ƒ is the friction 𝜖𝜖   factoris  the  pipe  roughness  (in  units  of  length)  
+
ε is the pipe roughness
D   (inpipe  
is  t"he   units of length)
diameter   (in  same  
1.2"units  as  roughness)  
= −2 log"* , +  
D is the pipe Re  
diameteris  t(in he  
# R eynolds  
same units number  
-./ roughness)
as 34 #

Re Note  that  tnumber


is the Reynolds he  Colebrook  equation  is  implicit  for  the  friction  factor,  f.  Despite  this,  it  is  the  most  wid
Where  
used  equation  for  finding  the  friction  factor  and  is  suitable  for  all  Reynolds  numbers  (Lindeburg  200

Note that the Colebrook 𝑓𝑓  


equation
The  Colebrook   isis  implicit
requires  
the  friction  
forfactor  
the  calculation   the of  tfriction factor,
he  Reynolds   ƒ. Despite
number   for  the  fthis,
low  in  itquestion.  The  Reyno
𝜖𝜖   is  the  pipe  roughness  (in  units  of  length)  
number  
is the most widely usedis  equation
the  ratio  ofor
f  inertial  
finding forces  
theto  friction
viscous  ffactor
orces  and  
and is  gisiven   by  
suitable for all
D   is  the  pipe  diameter  (in  same  units  as  roughness)  
Reynolds numbers (Lindeburg 89:2001).
𝑅𝑅𝑅𝑅 =Re     is  the  Reynolds  number   (2)  
;
Note  that   the  Colebrook  equation  is  implicit  for  the  friction  factor,  f.  Despite  this,  it  is  the  m
The Colebrook requires the ecalculation
Where   used   of thethe  
quation  for  finding   Reynolds number
friction  factor   and  ifor the flow
s  suitable   for  ain
ll  Reynolds  numbers  (Linde
question. The Reynolds number is the ratio of inertial forces to viscous forces and is
The  Colebrook  requires  the  calculation  of  the  Reynolds  number  for  the  flow  in  question.  Th
given by number  is  the  ratio  of  inertial  forces  to  viscous  forces  and  is  given  by  
8  
  89:
𝑅𝑅𝑅𝑅 = ;
  (2)

Where  
Where

P is the density 8  
v is the velocity   in the pipe
D is the “hydraulic diameter,” in this case the pipe diameter
M is the dynamic viscosity

14 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Ρ  
is  the  density  
𝑣𝑣   is  the  velocity  in  the  pipe  
D   is  the  “hydraulic  diameter,”  in  this  case  the  pipe  diameter  
Both the Reynolds number Μ
and  
the friction factor are dimensionless. The user must
is  the  dynamic  viscosity  
be careful to employ units that are consistent and that cancel to ensure the correct
results. Ρ   is  the  dBoth   the  Reynolds  number  and  the  friction  factor  are  dimensionless.  The  user  must  be  careful  to
ensity  
𝑣𝑣   is  the  vunits   that  
elocity   in  athe  
re  cponsistent  
ipe   and  that  cancel  to  ensure  the  correct  results.    

Flow in a D  
pipe is is  
a tspecial
he  “Flow  
hydraulic  
in  a  pdipe  
case iameter,”  
of the in  this  
is  a  sReynolds
pecial   case  
case   of  tthe  
number, he  Rpeynolds  
ipe  
and diameter  
number,  
where theand  
mass where  
flow the  m
isass  flow  is  known,  it  ca
Μ   is  the  drewritten  
ynamic  viscosity  
as  follows:  
known, it can be rewritten as follows:
Both  the  Reynolds  number  and  the  friction  
ABCCDEFG factor  are  dimensionless.  The  user  must  be  careful  to  employ  
:∙ ∙8
,I ∙J
units  that  are  consistent  and  that  cancel  
H∙ to  ensure  L∙MNOO#PQR
the  correct  results.  
(3)  
𝑅𝑅𝑅𝑅 = K
=   (3
; ;∙8∙S

Flow  in  a  pipe  is  a  special  case  of  the  Reynolds  number,  and  where  the  mass  flow  is  known,  it  can  be  
As isrewritten  
apparentas  from follows:   As  is  apparent  
Equation (3), the from   Equation  
pipe diameter(3),  the  
must pipe  be
diameter  
estimated must  bto
e  ecalculate
stimated  to  a calculate  a  Reynolds  n
Reynolds number. ThisThis   ABCCDEFG
complicates  the
complicates the  method
method  bby
y  forcing  
forcing an  aan
dditional  
additional calculation,   but  is  necessary  to  begin  the  a
calculation,
:∙ ∙8
,I ∙J
but is necessary to begin the analysis. H∙ As  an  example,  
L∙MNOO#PQRassume  a  header  pipe  is  required  to  carry  400  tons  of  refrigeration  (TR)  of  +0°F  
𝑅𝑅𝑅𝑅 = = K
  (3)  
; ;∙8∙S recirculated  ammonia  liquid  at  a  3:1  overfeed  rate.  The  properties  of  ammonia  in  Table  1  apply.
As anAs  example, assume
is  apparent   a header
quation   (3),  tpipe is required
he  pipe   diameter  mtoust  
carry 400 tons
be  estimated   to  of
from  ETable  1.  Properties  of  saturated  liquid  ammonia  at  +0°F   refrigeration
calculate   a  Reynolds  number.  
(TR) This  
of +0°F recirculated
complicates   ammonia
the  method   liquid
by  forcing   at a 3:1calculation,  
an  additional   overfeed brate. ut  is  nThe properties
ecessary   to  begin  of
the  analysis.  
ammonia in Table 1 apply. Temperature  (°F)   0.0  
As  an  example,  assume  a  header  pipe  is  required  to  carry  400  tons  of  refrigeration  (TR)  of  +0°F  
Saturated  pressure  (psi)   30.40  
recirculated  ammonia  liquid  at  a  3:1  overfeed   2 rate.  The  properties  of  ammonia  in  Table  1  apply.  
Liquid  density  (lb/ft )   41.34  
Temperature (°F) 0.0
Liq.  dynamic  viscosity  (lbm/ft-­‐s)  
Table  1.  Properties  of  saturated  liquid  ammonia  at  +0°F   1.402e-­‐04  
Saturated pressure (psi) 30.40
Liq.  enthalpy  (BTU/lb-­‐R)   112.61  
Liquid density (lb/ft
Temperature   (°F)   ) Vapor  enthalpy  (BTU/lb-­‐R)  
2
41.34 0.0  
681.39  
Liq. dynamic
Saturated  viscosity (lbm/ft-s)
pressure    (psi)   1.402e-04 30.40  
Liq. enthalpy (BTU/lb-R)
Liquid  density   (lb/ft2)   112.6141.34  
VaporLiq.  
enthalpy
dynamic  (BTU/lb-R) As   a   f irst  
viscosity  (lbm/ft-­‐s)   g uess   f or   p ipe  1.402e-­‐04  
681.39 diameter,  assume  that  a  2  in.  schedule  40  pipe  is  chosen.  The  internal  di
of  this  pipe  is  2.067  in.  With   this  information  and  that  in  Table  1,  calculating  the  Reynolds  numb
Table 1.Liq.   enthalpy  
Properties of(BTU/lb-­‐R)  
saturated liquid ammonia at +0°F 112.61  
possible.  The  mass  flow  can  
Vapor  enthalpy  (BTU/lb-­‐R)   be  calculated  as  follows:  
681.39  
  1**^_` Pc Pc
As a first guess for pipe diameter, assume
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 = that a 2 in.
400𝑇𝑇𝑇𝑇× schedule
× 40 pipe is chosen.
×3 = 421.96  
Mab∙_3 de".-fg""1.d" ^_` Mab
The internal
As  a  first  gdiameter of dthis
uess  for  pipe   pipe ais
iameter,   2.067
ssume   in.a  With
that   this information
2  in.  schedule   and that
40  pipe  is  chosen.   The  iin Tablediameter  
nternal  
of  this  pipe  the
1, calculating is  2.067  
Reynolds The  
in.   WRith  
eynolds  
number number   can  
this  information   then  
and  
is possible. be  
that  
The in  cmass
alculated  
Table  1flow per   Equation  
,  calculating  
can bethe  (R3).  
eynolds  nas
calculated umber  is  
possible.  The  mass  flow  can  be  calculated  as  follows:  
follows: 𝑙𝑙𝑙𝑙
4×421.96
1**^_`𝑅𝑅𝑅𝑅×= Pc 𝑚𝑚𝑚𝑚𝑚𝑚 Pc = 370,786  
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 = 400𝑇𝑇𝑇𝑇× 𝑙𝑙𝑙𝑙 ×3 = 421.96 1𝑓𝑓𝑓𝑓   60𝑠𝑠  
Mab∙_3 de".-fg""1.d" ^_` Mab
0.0001402 ×2.067𝑖𝑖𝑖𝑖× ×𝜋𝜋×
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 12𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚
The  Reynolds  number  can  then  be  calculated  per  Equation  (3).  
With  the  Reynolds  number  determined,  the  pipe  roughness  provides  the  final  piece  of  informati
needed  to  determine   the  friction  𝑙𝑙𝑙𝑙
4×421.96 factor.  For  steel  pipe,  the  average  roughness  is  0.0002  ft  (0.002
𝑅𝑅𝑅𝑅 = 𝑚𝑚𝑚𝑚𝑚𝑚 = 370,786  
𝑙𝑙𝑙𝑙 1𝑓𝑓𝑓𝑓 60𝑠𝑠
Technical Paper #6 9   0.0001402 𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠© ×2.067𝑖𝑖𝑖𝑖×
IIAR 2017
12𝑖𝑖𝑖𝑖
×𝜋𝜋×
𝑚𝑚𝑚𝑚𝑚𝑚
15
 
With  the  Reynolds  number  determined,  the  pipe  roughness  provides  the  final  piece  of  information  
Vapor  enthalpy  (BTU/lb-­‐R)   681.39  
 

As  a  first  guess  for  pipe  diameter,  assume  that  a  2  in.  schedule  40  pipe  is  chosen.  The  internal  diameter  
of  this  pipe  is  2.067  in.  With  this  information  and  that  in  Table  1,  calculating  the  Reynolds  number  is  
2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX
possible.  The  mass  flow  can  be  calculated  as  follows:  

1**^_` Pc Pc
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 = 400𝑇𝑇𝑇𝑇× × ×3 = 421.96    
Mab∙_3 de".-fg""1.d" ^_` Mab

The
The   Reynolds
Reynolds   number
number   canbe  then
can  then   be calculated
calculated   per(3).  
per  Equation   Equation (3).

𝑙𝑙𝑙𝑙
4×421.96
𝑅𝑅𝑅𝑅 = 𝑚𝑚𝑚𝑚𝑚𝑚 = 370,786  
𝑙𝑙𝑙𝑙 1𝑓𝑓𝑓𝑓 60𝑠𝑠
0.0001402 ×2.067𝑖𝑖𝑖𝑖× ×𝜋𝜋×
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 12𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚

With  the  Reynolds  number  determined,  the  pipe  roughness  provides  the  final  piece  of  information  
needed  
With tthe o  determine  
Reynolds the  
numberfriction  factor.  
determined,For  steel  the
pipe,  
pipe the  aroughness
verage  roughness  
provides is  0.0002   ft  (0.0024  
the final piece in.).    
of information needed to determine the friction factor. For steel pipe, the average
9  
roughness is 0.0002 ft (0.0024 in.).
 

0.0024𝑖𝑖𝑖𝑖
1 2.51
= −2 log"* 2.067𝑖𝑖𝑖𝑖 +  
𝑓𝑓 3.7 371,315 𝑓𝑓
0.0024𝑖𝑖𝑖𝑖
1
With  f  implicit  in  the  equation,  calculating   2.067𝑖𝑖𝑖𝑖 2.51
= −2 logt"* he  friction   factor  
+via  the  Colebrook     equation  requires  an  
iterative   p rocess.   T his   i s   d one   𝑓𝑓
b y   f irst   s lightly   r 3.7 the  equation  
ewriting   371,315 i n   t 𝑓𝑓
he   f orm:  
With ƒ implicit in the equation, calculating the friction factor via the Colebrook
equation requires an iterative process. This is done by first slightly
g1 rewriting the
With  f  implicit  in  the  equation,  calculating  the  0.0024𝑖𝑖𝑖𝑖
friction  factor  via  the  Colebrook  equation  requires  an  
equation in the form:𝑓𝑓 = −2 log 2.067𝑖𝑖𝑖𝑖 + 2.51
iterative  process.  This  is  done  by  first  slightly  
"* rewriting  the  equation  in  the  form:    
3.7 371,315 𝑓𝑓
g1
0.0024𝑖𝑖𝑖𝑖
A  friction  factor  can  now  b𝑓𝑓e  g=uessed   2.067𝑖𝑖𝑖𝑖 2.51
−2 (log start  with  
"*
0.02)  and  
+plugged  into  the  right     side  of  the  equation.  
3.7
The  next  step  is  to  solving  for  f  on  the  left  side,  then  use  the  371,315 𝑓𝑓 back  into  the  right  side  and  
result  to  plug  
continue  this  process  until  the  error  is  within  acceptable  limits  (four  significant  figures  is  assumed  
acceptable  
A   in  this  
friction  factor   case).  
can   now  be  guessed  (start  with  0.02)  and  plugged  into  the  right  side  of  the  equation.  
The   next  step  factor
A friction is  to  solving  
can now for  f  obe
n  the  
guessedleft  side,  
(startthen  uwith
se  the  
0.02) result  and
to  plug  
pluggedback  into  
into the  the
right  
right side  side
and  
Performing  this  iteration,  the  liquid  friction  factor  is  revealed  to  be  0.02109.  With  this  information,  
continue  
of the equation.this  process  The until  next
the  error  
stepis  iswithin   acceptable  
to solving for lƒimits  
on (the four  left
significant  
side, then figures   is  athe
use ssumed  
result
calculating  the  liquid  pressure  drop  in  a  100-­‐ft  section  of  piping  is  now  possible.  
acceptable  in  this  case).  
to plug back into the right side and continue this process until the error is within
The   Darcy  equation  
acceptable limits is  (four
commonly   used  to  cfigures
alculate  isfriction   loss  for   liquid  laminar  athis
nd  turbulent  flow  
Performing   this   iteration,   the  significant
liquid  friction   assumed
factor  is  revealed   to  bacceptable
e  0.02109.  Win
ith  this  icase).
nformation,  
(Lindeburg  2001).  It  has  the  form  of  
calculating  the  liquid  pressure  drop  in  a  100-­‐ft  section  of  piping  is  now  possible.  
Performing #w9 I
this   iiteration, the liquid friction factor isfor  
revealed to be a0.02109. With this
The   ℎ#e=
Darcy   quation   s  commonly   used   to  calculate   friction   loss   liquid  laminar   (4)  
nd  turbulent   flow  
18x
information,
(Lindeburg   calculating
2001).   theof  liquid pressure drop in a 100-ft section of piping is now
It  has  the  form  
Where  
possible.
#w9 I
ℎ# =   (4)  
hf   18x
is   the  head  loss  (in  length  dimensions)  
𝑓𝑓   is  the  friction  factor  
Where  
L   is  the  equivalent  length  of  the  pipe  section    
hf    
𝑣𝑣 is  the  hvelocity  
ead  loss  (in  length  dimensions)  
16 D  
𝑓𝑓   is  the  fpriction  
ipe  diameter  
factor   © IIAR 2017 Technical Paper #6
g  
L   is  the  eacceleration   due  to  
quivalent  length   of  gtravity  
he  pipe  section    
  𝑣𝑣   is  the  velocity  
D   is  the  pipe  diameter  
371,315 𝑓𝑓
g1
0.0024𝑖𝑖𝑖𝑖
g1
0.0024𝑖𝑖𝑖𝑖 2.51
A  friction  factor  
2.067𝑖𝑖𝑖𝑖 𝑓𝑓 =be  2.51
can  now   g−2
uessed  
log"* (start  2.067𝑖𝑖𝑖𝑖
with  0.02)  +and  plugged  into  the    right  side  of  th
𝑓𝑓 = −2 log"* +   3.7 371,315
The  next  step  3.7
is  to  solving  
371,315for  f  on  t𝑓𝑓
he  left  side,  then  use   the  result  𝑓𝑓
to  plug  back  into  the  r
continue  this  process  until  the  error  is  within  acceptable  limits  (four  significant  figures  is  a
A  friction  
Method for Determining Best Economic Pipe factor   cfor
Sizein  
acceptable   an   now  
Ammonia
this   be  gRefrigeration
case).   uessed  (start  Piping
with  0.02)  and  plugged  into  the  right  side  of  the  eq
A  friction  factor  can  now  be  guessed  (start  with  0.02)  and  plugged  into  the  right  side  of  the  equation.  
The  next  step  is  to  solving  for  f  on  the  left  side,  then  use  the  result  to  plug  back  into  the  right  
The  next  step  is  to  solving  for  f  on  the  left  side,  then  use  the  result  to  plug  back  into  the  right  side  and  
Performing  
continue   this  iteration,  
this  process   until  the  tehe   liquid  
rror   is  wfithin  
riction   factor  is  rlevealed  
acceptable   to  b
imits  (four   e  0.02109.  
significant   With  tihis  
figures   inf
s  assu
continue  this  process  until  the  error  is  within  acceptable  limits  (four  significant  figures  is  assumed  
calculating  
acceptable   the  
in  this   liquid  pressure  drop  in  a  100-­‐ft  section  of  piping  is  now  possible.  
case).  
acceptable  in  this  case).  
The Darcy equation is commonly used to calculate friction loss for liquid laminar and
The  Darcy  
Performing   equation  
this   iteration,   is  tche  
ommonly   used  to  
liquid  friction   calculate  
factor   friction  to  
is  revealed   loss  
be  f0or   liquid  W
.02109.   laminar   and  
ith  this   turb
informa
Performing  
turbulentthis  
flowiteration,  
(Lindeburg the  liquid  
2001). friction  
It fhas
actor  
theis  revealed  
form of to  be  0.02109.  With  this  information,  
(Lindeburg  
calculating   the  l2iquid  
001).  pIressure  
t  has  the   form  
drop   in  oaf    100-­‐ft  section  of  piping  is  now  possible.  
calculating  the  liquid  pressure  drop  in  a  100-­‐ft  section  of  piping  is  now  possible.  
I
The  Darcy  equation  
#w9 is  c  ommonly  used  to  calculate  friction  loss  for  liquid  laminar  and  turbulen
ℎ# = friction  

The   Darcy  equation  is  commonly  used  to  calculate  
18x l (4)
oss  for  liquid  laminar  and  turbulent  flow  
(Lindeburg  2001).  It  has   the  form  of  
(Lindeburg  2001).  It  has  the  form  of  
Where   #w9 I
Where #w9 I ℎ# =  
ℎ# =   hf  
18x (4)  
18x is  the  head  loss  (in  length  dimensions)  
hf is the head loss (inWhere   length 𝑓𝑓dimensions)
  is  the  friction  factor  
Where  
ƒ is the friction factor hf   L   is  the  equivalent  length  of  the  pipe  section    
is   t he  
is  hthe  
ead   loss  (in  length  dimensions)  
velocity  
ofd𝑣𝑣imensions)  
hf   is  the   head  loss  (length
in  length    
L is the equivalent 𝑓𝑓 the pipe fsection
  D   is  the   riction   factor  
𝑓𝑓   is  the  friction  factor   is  the  pipe   diameter  
v is the velocity L   g   is  the   e quivalent   length  doue  
f  tthe  
L   is  the  equivalent  length  of  the  pipe  sis   the  acceleration  
ection     o  gpravity  
ipe  section    
D is the pipe diameter   𝑣𝑣   is  the  velocity  
𝑣𝑣   is  the  velocity  
is  the  pipe  diameter   D  
g is the acceleration due to gravity is  the  pipe  diameter  
D  
The   v elocity  
is  tm ust   be  calculated   to   dgetermine  
ravity   friction  loss.  Using  the  mass  flow,  pipe  diam
is  the  acceleration  due  g  
he   acceleration   due   to  
g   to  gravity  
  density,  velocity  can  be  calculated  as  
  The velocity must be calculated to determine friction loss. Using the mass flow, pipe
L∙yNOO#PQR "
diameter, and density, velocity can be
The  velocity  
𝑣𝑣m= calculated
ust   ∙ as   to  determine  friction  loss.  Using  the  mass  flow,  pipe  diamete
be  cIalculated  
The  velocity  must  be  calculated  to  determine  friction  
8 lS oss.  Using  
: the  mass  flow,  pipe  diameter,  and  
density,  velocity  can  be  calculated  as  
density,  velocity  can  be  calculated  as  
𝑙𝑙𝑙𝑙
L∙yNOO#PQR " 4×421.96 𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓
L∙yNOO#PQR " 𝑣𝑣 = ∙   𝑚𝑚𝑚𝑚𝑚𝑚 = 7.291  
𝑣𝑣 = ∙   8I S : 𝑣𝑣 = × = 438.0
(5)  
I 8 S : (5) 1
0.02967𝑓𝑓𝑓𝑓 ×𝜋𝜋 41.34𝑙𝑙𝑙𝑙 𝑚𝑚𝑚𝑚𝑚𝑚 𝑠𝑠
𝑙𝑙𝑙𝑙
𝑙𝑙𝑙𝑙 above  flow  of  400  
For  the   TR  at  3:1  𝑚𝑚𝑚𝑚𝑚𝑚
4×421.96 recirculation  𝑓𝑓𝑓𝑓 rate  in  a  100-­‐ft  
𝑓𝑓𝑓𝑓 section  of  𝑓𝑓𝑓𝑓
pipe,  the  frictio
4×421.96 𝑓𝑓𝑓𝑓 𝑣𝑣 = 𝑓𝑓𝑓𝑓 × 𝑓𝑓𝑓𝑓 = 438.0 = 7.291  
𝑣𝑣 = be   𝑚𝑚𝑚𝑚𝑚𝑚 × = 438.0 = 7.291
1  
0.02967𝑓𝑓𝑓𝑓 ×𝜋𝜋 41.34𝑙𝑙𝑙𝑙 𝑚𝑚𝑚𝑚𝑚𝑚 𝑠𝑠
0.02967𝑓𝑓𝑓𝑓 1 ×𝜋𝜋 41.34𝑙𝑙𝑙𝑙 𝑚𝑚𝑚𝑚𝑚𝑚 𝑠𝑠
For  t10   he  above  flow  of  400  TR  at  3:1  recirculation  rate  in  a  100-­‐ft  section  of  pipe,  the  friction  lo
For  the  above  flow  of  400  TR  at  3:1  r   ecirculation  rate  in  a  100-­‐ft  section  of  pipe,  the  friction  loss  would  
be  
be  
For the above flow of 400 TR at 3:1 recirculation rate in a 100-ft section of pipe, the
friction loss would be 10  
10  
 
 
𝑓𝑓𝑓𝑓 1
0.02109×100𝑓𝑓𝑓𝑓×(7.29 )
ℎ# = 𝑠𝑠 = 10.132𝑓𝑓𝑓𝑓  
𝑓𝑓𝑓𝑓
2×0.17225𝑓𝑓𝑓𝑓×32.2 1
𝑠𝑠

The  Darcy  equation  produces  a  result  with  dimensions  of  length,  which  must  be  converted  to  pressure.  
Head  
Theloss   can  bequation
Darcy e  converted   to  pressure  
produces loss  via  with
a result Equation   (6),  
dimensions of length, which must be
converted
∆𝑝𝑝 = to pressure. Head loss can be converted to pressure loss via Equation
ℎ ×𝜌𝜌×𝑔𝑔   (6)  (6),
#

𝑙𝑙𝑙𝑙 𝑓𝑓𝑓𝑓 𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1 1𝑓𝑓𝑓𝑓 1


∆𝑝𝑝 = 10.132𝑓𝑓𝑓𝑓×41.34 ×32.2 × × = 2.91𝑝𝑝𝑝𝑝𝑝𝑝  
𝑓𝑓𝑓𝑓 - 𝑠𝑠 1 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 144𝑖𝑖𝑖𝑖1
Technical Paper #6 © IIAR 2017 17
Notice  that  in  U.S.  customary  units,  an  additional  factor  to  correct  for  lbf  vs.  lb  is  required.  Clearly,  as  a  
baseline  case  of  2  psi/100  ft,  a  2  in.  pipe  is  undersized.  The  process  must  be  repeated  with  2  1/2  in.  
diameter  pipe,  the  results  of  which  show  a  calculated  pressure  drop  of  1.161  psi.  This  pressure  fits  the  
𝑓𝑓𝑓𝑓 1
0.02109×100𝑓𝑓𝑓𝑓×(7.29)
ℎ# = 𝑠𝑠 = 10.132𝑓𝑓𝑓𝑓  
𝑓𝑓𝑓𝑓
2017 IIAR2×0.17225𝑓𝑓𝑓𝑓×32.2
Natural Refrigeration𝑠𝑠Conference
1 & Heavy Equipment Expo, San Antonio, TX

The  Darcy  equation  produces  a  result  with  dimensions  of  length,  which  must  be  converted  to  pressure.  
Head  loss  can  be  converted  to  pressure  loss  via  Equation  (6),   𝑓𝑓𝑓𝑓
0.02109×100𝑓𝑓𝑓𝑓×(7.29 )1
ℎ# = 𝑓𝑓𝑓𝑓 1 𝑠𝑠 = 10.132𝑓𝑓𝑓𝑓  
0.02109×100𝑓𝑓𝑓𝑓×(7.29 ) 𝑓𝑓𝑓𝑓
∆𝑝𝑝 = ℎ# ×𝜌𝜌×𝑔𝑔   ℎ# = 𝑠𝑠 = 2×0.17225𝑓𝑓𝑓𝑓×32.2
10.132𝑓𝑓𝑓𝑓   (6)   (6)
𝑓𝑓𝑓𝑓 𝑠𝑠 1
2×0.17225𝑓𝑓𝑓𝑓×32.2 11
𝑙𝑙𝑙𝑙 equation   𝑓𝑓𝑓𝑓 roduces   𝑙𝑙𝑙𝑙𝑙𝑙 ∙ a𝑠𝑠𝑠𝑠  result  w 1𝑓𝑓𝑓𝑓 1dimensions  of  length,  which  
∆𝑝𝑝 = 10.132𝑓𝑓𝑓𝑓×41.34 - ×32.2 1p× The   D arcy   × ith  0.02109×100𝑓𝑓𝑓𝑓×(7.29
1
= 2.91𝑝𝑝𝑝𝑝𝑝𝑝   𝑓𝑓𝑓𝑓 1 must  be  converted  to  
)
The  Darcy  equation  produces   Head   𝑓𝑓𝑓𝑓
loss  cw
a  result   an   be  
ith   𝑠𝑠
dimensions   32.2𝑙𝑙𝑙𝑙
converted  oto   ∙ 𝑓𝑓𝑓𝑓
f  lpength,  
ressure   144𝑖𝑖𝑖𝑖
ℎ#which  =loss  vmia  ust   Equation  
be  converted   (6),   to  𝑓𝑓𝑓𝑓 p𝑠𝑠ressure.   = 10.132𝑓𝑓𝑓𝑓  
Head   l oss   c an   b e   c onverted   t o   p ressure   l oss   v ia   E quation   ( 6),   2×0.17225𝑓𝑓𝑓𝑓×32.2 1
Notice  that  in  U.S.  customary  units,  an  additional   ∆𝑝𝑝 = ℎ# ×𝜌𝜌×𝑔𝑔   factor  to  correct  for  lbf  vs.  lb  is  required.  Clearly,  a𝑠𝑠s  a  
baseline   case  o
Notice f  that
2=  psi/100  
ℎin U.S. ft,  acustomary
 2  in.  pipe  is  uunits, ndersized.  additional The  process  factor must  be  torepeated   wfor ith  2lbf  1/2   in.  lb
∆𝑝𝑝 # ×𝜌𝜌×𝑔𝑔   The  Darcy  an equation  produces   a  result  correct with  dimensions   vs. (6)  is which  must  be  conve
o1f  length,  
diameter  pipe,  the  results  of  which  show  a  calculated  pressure  drop  of  1.161   𝑙𝑙𝑙𝑙 psi.  This   𝑓𝑓𝑓𝑓pressure   𝑠𝑠 the   1𝑓𝑓𝑓𝑓 1
𝑙𝑙𝑙𝑙𝑙𝑙 ∙fits  
required. Clearly, as a baseline Head   case loss   of
∆𝑝𝑝 can  
= 2 b10.132𝑓𝑓𝑓𝑓×41.34
e  converted  
psi/100 ft, ato   2pressure  
in.- pipe ×32.2 loss  
is 1vundersized.
ia  
× Equation  (6),   The
× = 2.91𝑝𝑝𝑝𝑝𝑝𝑝  
criterion  of  being  the  smallest  pipe  size  with  a  p𝑙𝑙𝑙𝑙 ressure  drop   𝑓𝑓𝑓𝑓 less   𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1or  e𝑓𝑓𝑓𝑓
than   qual   1𝑓𝑓𝑓𝑓to  1 2  p 𝑠𝑠si  per   100   ∙ 𝑓𝑓𝑓𝑓 144𝑖𝑖𝑖𝑖1
32.2𝑙𝑙𝑙𝑙
process must∆𝑝𝑝 be=repeated 10.132𝑓𝑓𝑓𝑓×41.34
equivalent  ft,  and  thus  2  1/2  in.  pipe  is  chosen  𝑓𝑓𝑓𝑓
with 2 1/2 in.base  
×32.2
as  -∆𝑝𝑝
the  =
diameter ×
1case.   pipe, × the results of which show
= 2.91𝑝𝑝𝑝𝑝𝑝𝑝  
ℎ#𝑠𝑠×𝜌𝜌×𝑔𝑔   32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 144𝑖𝑖𝑖𝑖1
a calculated pressure drop Notice  of that  1.161 in  U.S.  psi. customary  
This pressure units,  an  fits additional  the criterion factor  to  of correct  
being for  the
lbf  vs.  lb  is  required.  Cle
This  pNotice  
ressure   drop   in  cU
an   bsizee   used   baseline  
to   caalculate   caase   of  2  psi/100  
n  haydraulic   ft,  tao  
 2c  iorrect  
n.   pipe  for  pis  ower  
ulbf  ndersized.   Talculated  
he  p𝑓𝑓𝑓𝑓 rocess   must   be  
a  repeated  
1𝑓𝑓𝑓𝑓 1 with  2  1
smallest that   pipe .S.   customary  
with units,  
pressure drophless
dditional   orsepower.  
factor  
∆𝑝𝑝 than
= or Pump  
10.132𝑓𝑓𝑓𝑓×41.34 equal to can  
v2s.  psi b  bis  e  
l𝑙𝑙𝑙𝑙 rcequired.  
per
×32.2 100 equivalent
×
p𝑙𝑙𝑙𝑙𝑙𝑙
er  
Clearly,   ∙ 𝑠𝑠a1s  
× = 2.91𝑝𝑝
Equation   (7).   case  of  2  psi/100  fdiameter  
baseline   t,  a  2  in.  ppipe   ipe,  
is  tuhe   results  of  The  
ndersized.   which   show  maust  
process    calculated  
be  𝑓𝑓𝑓𝑓 - pressure  
repeated   1 d232.2𝑙𝑙𝑙𝑙
w𝑠𝑠ith   rop  
 1/2  oif  n.  
1∙.161  
𝑓𝑓𝑓𝑓 p144𝑖𝑖𝑖𝑖
si.  This  
1 pressure
ft, and thus 2 1/2 in. pipe is chosen as the base case.
diameter  pipe,  the  results  ocriterion   f  which  sohow   f  being   the  smallest  
a  calculated   pipe  sdize  
pressure   rop  woith   f  1a.161  
 pressure  
psi.  This   drop   less  than  
pressure   otr  he  
fits   equal  to  2  psi  per  10
∆Ä×yNOO#PQR
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃o=
criterion   f  being  t:he  smallest       equivalent   Notice  
pipe  size   t hat  
ft,  waith   i n  
nd  at  hus   U .S.   c ustomary  
2  1/2  din.  
pressure   rop   pipe   u nits,  
is  tchan  
less   hosen   a n   a dditional  
or  eaqual  
s  the  to  base   f actor  
p(7)  
case.  
2  psi   t o  
er  100   c orrect   for  lbf  vs.  lb  is  requ
This pressure
equivalent   ft,  and  drop thus  2can  1/2  ibe n.  pused baseline  
ipe   is  cto case  
calculate
hosen   as  othe  
f  2  bphydraulic
si/100  
ase   ft,  a  horsepower.
case.   2  in.  pipe  is  undersized.   Pump power The  process   can must  be  repeated
Using  the   This   p ressure  
diameter   d rop   c an  
,  power  pconsumption   b e   u sed  
ipe,  the  results   t o   c alculate  
in  tohe  
f  w1hich   h
00-­‐ft  show   ydraulic   h orsepower.  
a  calculated  
section  of  p2ressure   P ump  
 1/2  in.  drop   power  
of  1c.161  
an  bpe  si.  
calculaThis
be pcalculated
reviously  calculate   per Equation information   (7). sample  
piping  This  
is   pressure  drop  can  be  used  to  ccriterion   Equation   ( 7).  
alculate  hoydraulic   f  being  thhe   orsepower.  
smallest  pPipe   ump   power  
size   with  ac  an   be  calculated  
pressure   drop  less   per  than  or  equal  to  2  ps
Equation  (7).   equivalent  ∆Ä×yNOO#PQR ft,  and  thus  2  1/2  in.  pipe  is  chosen  as  the  base  case.  
𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃1 =     (7)
1.161 1 ×421.96
𝑖𝑖𝑖𝑖 ∆Ä×yNOO#PQR 𝑚𝑚𝑚𝑚𝑚𝑚 144𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚 : ℎ𝑝𝑝
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 =     × This  1pressure   × d× rop  can  be  used   =to   calculate  h=
0.05173ℎ𝑝𝑝 ydraulic  
0.03858𝑘𝑘𝑘𝑘   horsepower.   (7)   Pump  power  can  b
𝑙𝑙𝑙𝑙: 𝑓𝑓𝑓𝑓 60𝑠𝑠 𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙
41.34 - Using  the   p reviously   c alculate   i nformation  
𝑠𝑠 consumption in the 100-ft sample sample  section  o
,   p ower   c onsumption   i n   t he   1 00-­‐ft  
Equation  (7).   550  
Using the previously 𝑓𝑓𝑓𝑓 calculate information , power
Using  the  previously  calculate   piping   is  
information   ,  power  consumption  in  the  100-­‐ft  sample  section  of  2  1/2  in.  
section
This  result   of 2 1/2 in. piping
is  its  he  ideal  hydraulic  horsepower  consumed   is 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃by  
∆Ä×yNOO#PQR
=this  section  o  f     piping.  It  must  be  divided  by  
piping   :
𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙
the  assumed  efficiencies  of  the  pump,  motor,  a1.161 nd  power   f actor  
×421.96 t o   give  the   true  e1stimated  
144𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚 consumption   ℎ𝑝𝑝
𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑖𝑖𝑖𝑖 1 𝑚𝑚𝑚𝑚𝑚𝑚
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 Using   = t he   p reviously   c alculate   ×
i nformation   × ,   p ×
ower   c onsumption   = 0.05173ℎ𝑝𝑝 = 0.038
of  power  at  the  meter. 1.161     1 ×421.96 𝑚𝑚𝑚𝑚𝑚𝑚 144𝑖𝑖𝑖𝑖
1
𝑚𝑚𝑚𝑚𝑚𝑚 𝑙𝑙𝑙𝑙 ℎ𝑝𝑝 𝑓𝑓𝑓𝑓 1 60𝑠𝑠 𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙 in  the  100-­‐ft  sample  
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 𝑖𝑖𝑖𝑖 piping   × is   1 × 41.34 × - = 0.05173ℎ𝑝𝑝 = 0.03858𝑘𝑘𝑘𝑘   550  
𝑠𝑠
𝑙𝑙𝑙𝑙
41.34 is  t-aken  to  be  6𝑓𝑓𝑓𝑓 60𝑠𝑠𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙
For  this  analysis,  pump  efficiency   0%.  Motor  efficiency   550   is  taken  to  be  95%.  The  actual  
𝑓𝑓𝑓𝑓 𝑠𝑠
power  consumption  then  becomes   This  result  is  the  ideal  h1.161 ydraulic   𝑙𝑙𝑙𝑙𝑙𝑙horsepower   𝑙𝑙𝑙𝑙consumed  1by  this  section  of  piping.  It  must  be  div
×421.96
𝑖𝑖𝑖𝑖 1 𝑚𝑚𝑚𝑚𝑚𝑚 144𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚 ℎ𝑝𝑝
This  result  is  the  ideal  hydraulic   the  ahssumed  orsepower   𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃
efficiencies   = of  tb
consumed   he  
y  tphis  ump,  s𝑙𝑙𝑙𝑙 motor,  
ection   of  ap× nd   power  
iping.   I1t   m ×fust  
actor  
b t×
e   d o  ivided  
give  𝑓𝑓𝑓𝑓the   rue  =
by  ∙ t𝑙𝑙𝑙𝑙𝑙𝑙 0.05173ℎ𝑝𝑝
estimated   con
ÑQR4Ö
41.34 𝑓𝑓𝑓𝑓 60𝑠𝑠
𝑃𝑃
the   =
Éassumed    
efficiencies   of  of   power  
thydraulic
he   pump,   at  m the  
otor,  meter.and    p   ower   factor  𝑓𝑓𝑓𝑓 to  - give   (8)   550  
This result
η η Ñà á A is the ideal horsepower consumed by tthis he  true  section estimated   consumption  
of piping. It 𝑠𝑠
of  
must power   beat  divided
the  meter. by     the assumed
For   this  aThis  
nalysis,   efficiencies
result   pump  
is  the  
of theis  pump,
efficiency  
ideal  hydraulic   taken   to  motor,
be  60%.  and
horsepower   cM
power
otor  
onsumed  
factor
efficiency  
by  this  is  
to too  f  bpe  
taken  
section   95%.  It  Tm
iping.   heu
Where  
give
For   this  the true estimated
analysis,   consumption
power  
pump  efficiency   ics  onsumption  
taken  
the   to  bof
assumed   epower
e  6then   at ethe
bMecomes  
0%.   otor  
fficiencies   meter.
officiency  
f  the   is  taken  
pump,   to  bae  
motor,   nd  95%.   The  
power   actual  
factor   to  give  the  true  estim
ηpower  
Ä   c Pump   e
onsumption   fficiency  
t hen   b ecomes   of   p ower   a t   t he   m eter.    
ÑQR4Ö
ηMFor   Motor   e fficiency   𝑃𝑃É =  
this analysis, ÑQR4Ö pump efficiency áisAtaken to be 60%. Motor efficiency is taken to be
η η Ñà
PF   𝑃𝑃Power  
É = η Fηactor     (typically  taken   For  ats  his  
1)  analysis,  pump  efficiency  is  taken  to  be  60%.  Motor   efficiency  is  taken  to  be  9
(8)  
95%. The actual á A Ñà power consumption then becomes
  Where   power   c onsumption   t hen   b ecomes  
Where   ÑQR4Ö
11   ηÄ   Pump  
𝑃𝑃É =efficiency  
 
ηá ηA Ñà (8)
  ηÄ   Pump  efficiency   ηM   Motor  efficiency  
ηM   Motor  efficiency   PF  Where  
Power  Factor  (typically  taken  as  1)  
PF   Power  Factor     (typically  taken  as  1)  
18   ηÄ©   IIAR Pump  
2017 efficiency   Technical Paper #6
11   ηM   Motor  efficiency  
11    
PF   Power  Factor  (typically  taken  as  1)  
   
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Where

ηp Pump efficiency
ηm Motor efficiency
PF Power Factor (typically taken as 1)

0.05173ℎ𝑝𝑝
𝑃𝑃É = = 0.091ℎ𝑝𝑝 = 0.06768𝑘𝑘𝑘𝑘  
0.6×0.95
 
This  is  the  power  consumed  in  this  section  of  piping  at  full  load.  It  can  be  used,  along  with  the  system’s  
estimated  
This is the usage  
power and  lifetime,  
consumed to  calculate  
in thisasection
 net  present   value  of  at
of piping the  
full energy  
load. consumed  
It can be in  tused,
he  section  
over   t he   l ife   o f   t he   s ystem   i n   S tep   3 .  
along with the system’s estimated usage and lifetime, to calculate a net present value
of the energy consumed in the section over the life of the system in Step 3.
Vapor  
Pressure  drop  in  vapor  lines  incurs  an  energy  cost  at  the  compressor,  much  like  liquid  line  pressure  drop  
incurs  a  pump  energy  penalty.  This  requires  a  pressure  drop  calculation  for  vapor  lines,  very  similar  to  
that  of  liquid.  
Vapor
Vapor  lines  differ  from  liquid  in  that  the  fluid  is  compressible.  The  Darcy  equation  is  considered  
adequate  
Pressurein  drop cases  in
where  
vapor the  lines
pressure  
incurs drop  an is  <energy
10%  of  the  
cost absolute  
at theinlet   pressure  and  
compressor, much the  Mlike ach  number  
is  
liquid<0.3  (Lindeburg  
line pressure 2001).  drop
However,  
incursconsidering  
a pumpthe   flow  limits  
energy penalty.established   previously  
This requires afor   vapor  lines  
pressure drop
(limits  on  pressure  drop  to  0.5  psi/100  ft  in  vapor  lines  above  -­‐20°F  and  0.25  psi/100  ft  in  vapor  lines  -­‐
calculation for vapor lines, very similar to that of liquid.
20°F  and  below),  the  Darcy  equation  will  provide  satisfactory  results  down  to  a  temperature  of  
approximately  -­‐80°F,  arguably  well  below  industrial  refrigeration  applications  for  ammonia.  
Vapor lines differ from liquid in that the fluid is compressible. The Darcy equation
Note  that  pure  vapor  flow  will  be  based  on  saturated  (suction)  and  superheated  (discharge)  conditions,  
is considered adequate in cases where the pressure drop is <10% of the absolute
again  using  the  Colebrook  equation  to  find  the  friction  coefficient,  f.  The  process  of  determining  
inlet pressure and the Mach number is <0.3 (Lindeburg 2001). However, considering
pressure  drops  in  vapor  lines  then  becomes  quite  similar  to  that  in  liquid  lines.  
the flow limits established previously for vapor lines (limits on pressure drop to 0.5
Table  2.  Saturated  ammonia  vapor  properties  at  -­‐30°F  
psi/100 ft in vapor lines above -20°F and 0.25 psi/100 ft in vapor lines -20°F and
below), the Darcy
Temperature   (°F)   equation will provide-­‐30  
satisfactory results down to a temperature
Saturated  
of pressure  (-80°F,
approximately psia)   arguably well below
13.89   industrial refrigeration applications for
3
Density  (lb/ft )  
ammonia. 0.05268  
Enthalpy  (btu/lb)   670.10  
Dynamic  viscosity  (lbm/ft-­‐s)   5.390E-­‐6  
Note that pure vapor flow will be based on saturated (suction) and superheated
Entropy  (btu/lb-­‐R)   1.581  
(discharge)ated (for suction)
Const.  pressure  specific  heat  
and conditions,
0.51768  
again using the Colebrook equation to
(BTU/lb-­‐R)  (60°F,  14.7  psia)  
Const.  volume  specific  heat   0.39217  
(BTU/lb-­‐R)  (60°F,  14.7  psia)    
 
Technical Paper #6 © IIAR 2017 19
Using  the  same  process  discussed  under  the  liquid  piping  calculation,  the  pressure  drop  can  be  
determined  for  a  vapor  flow.  In  this  case,  assume  a  mass  flow  rate  of  350.0  lb/min  of  saturated  vapor  
2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

find the friction coefficient, ƒ. The process of determining pressure drops in vapor
lines then becomes quite similar to that in liquid lines.

Temperature (°F) -30


Saturated pressure (psia) 13.89
Density (lb/ft3) 0.05268
Enthalpy (btu/lb) 670.10
Dynamic viscosity (lbm/ft-s) 5.390E-6
Entropy (btu/lb-R) 1.581
Const. pressure specific heat 0.51768
(BTU/lb-R) (60°F, 14.7 psia)
Const. volume specific heat 0.39217
(BTU/lb-R) (60°F, 14.7 psia)
Table 2. Saturated ammonia vapor properties at -30°F

Using the same process discussed under the liquid piping calculation, the pressure
drop can be determined for a vapor flow. In this case, assume a mass flow rate
of 350.0 lb/min of saturated vapor generated in a bank of flooded blast cell coils
being fed with +0°F pumped liquid (corresponding to a heat absorption rate in the
evaporator of 977.2 TR). To initiate the calculation, Reynolds number, friction factor,
and velocity are determined using the information listed in Table 2 for saturated
ammonia vapor at -30°F. The initial guess of pipe size is given as 12 in. standard
weight, which has an internal diameter of 12.0 in. (1.0 ft). The Reynolds number is
calculated from Equation 3.

𝑙𝑙𝑙𝑙
4×350
𝑅𝑅𝑅𝑅 = 𝑚𝑚𝑚𝑚𝑚𝑚 =  1,377,965  
𝑙𝑙𝑙𝑙 1𝑓𝑓𝑓𝑓 60𝑠𝑠
0.000005390 ×12.0𝑖𝑖𝑖𝑖× ×𝜋𝜋×
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 12𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚

Following  Reynolds  number,  the  friction  factor  f  is  determined  using  Equation  1  (starting  with  an  
estimate  
Following of  0.02,  
Reynolds three  iterations  
number,are   sufficient  
the friction to  factor
find  the  f ais
nswer   in  this  case).  
determined using Equation 1
(starting with an estimate of 0.02, three iterations are sufficient to find the answer in
0.0024𝑖𝑖𝑖𝑖
1 2.51
this case). = −2 log"* 12.00𝑖𝑖𝑖𝑖 + , 𝑓𝑓 = 0.01445  
𝑓𝑓 3.7 1,377,965 𝑓𝑓

 
20 The  velocity  is  determined  next  using  Equation  
© IIAR 5:  2017 Technical Paper #6

𝑙𝑙𝑙𝑙
4×350 𝑓𝑓𝑓𝑓 - 𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓
𝑣𝑣 = 𝑚𝑚𝑚𝑚𝑚𝑚 × = 8,459 = 140.99  
1 1
𝑙𝑙𝑙𝑙
4×350 𝑙𝑙𝑙𝑙
𝑅𝑅𝑅𝑅 = 4×350 𝑚𝑚𝑚𝑚𝑚𝑚 =  1,377,965  
𝑅𝑅𝑅𝑅 = 0.000005390 𝑙𝑙𝑙𝑙 ×12.0𝑖𝑖𝑖𝑖× 𝑚𝑚𝑚𝑚𝑚𝑚 1𝑓𝑓𝑓𝑓 60𝑠𝑠 =  1,377,965  
Method for Determining Best Economic 𝑓𝑓𝑓𝑓 𝑙𝑙𝑙𝑙 Size for Ammonia
Pipe ×𝜋𝜋×
1𝑓𝑓𝑓𝑓Refrigeration 60𝑠𝑠 Piping
0.000005390 ∙ 𝑠𝑠 ×12.0𝑖𝑖𝑖𝑖× 12𝑖𝑖𝑖𝑖 ×𝜋𝜋× 𝑚𝑚𝑚𝑚𝑚𝑚
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 12𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚
𝑙𝑙𝑙𝑙
Following  Reynolds  number,  the  friction  factor  f  is  determined  using  Equation   4×350 1  (starting   with  an  
Following  Reynolds  number,  the  friction  factor  f  is  d𝑅𝑅𝑅𝑅 etermined  
= 𝑚𝑚𝑚𝑚𝑚𝑚
using  Equation  1  (starting  with  an   =  1,377,965  
estimate  of  0.02,  three  iterations  are  sufficient  to  find  the  answer  in  this  case).   𝑙𝑙𝑙𝑙 1𝑓𝑓𝑓𝑓 60𝑠𝑠
estimate  of  0.02,  three  iterations  are  sufficient  to  find  the  0.000005390 answer  in  this  case).   ×12.0𝑖𝑖𝑖𝑖× ×𝜋𝜋×
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 12𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚
0.0024𝑖𝑖𝑖𝑖
1 0.0024𝑖𝑖𝑖𝑖 𝑙𝑙𝑙𝑙 2.51
1 = −2 log"*Following   12.00𝑖𝑖𝑖𝑖
4×350
Reynolds   + number,  
2.51 the  friction  
, 𝑓𝑓 = f0.01445
actor  f  is  d   etermined  using  Equation  1  (starting  w
𝑅𝑅𝑅𝑅 =𝑓𝑓 = −2 log"* 12.00𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚 =0.01445
 1,377,965  
3.7 + 1,377,965
𝑙𝑙𝑙𝑙 of  0.02,  three  1𝑓𝑓𝑓𝑓
estimate   𝑓𝑓 , 𝑓𝑓 =
re  sufficient  to  f  ind  the  answer  in  this  case).  
iterations  a60𝑠𝑠
𝑓𝑓 0.000005390 3.7×12.0𝑖𝑖𝑖𝑖× 1,377,965 ×𝜋𝜋× 𝑓𝑓
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 12𝑖𝑖𝑖𝑖 𝑚𝑚𝑚𝑚𝑚𝑚
  0.0024𝑖𝑖𝑖𝑖
1 2.51
 Following   R eynolds   n umber,   t he   f riction   f actor  
The  velocity  is  determined  next  using  Equation  5:   f   i s   d = −2 log"* 12.00𝑖𝑖𝑖𝑖
etermined   u sing   E quation   1  (starting  
+ with  an   , 𝑓𝑓 = 0.01445  
The  The velocity
velocity  
estimate   of  0is   isthree  
determined
determined  
.02,   next
next  using  
iterations   using Equation
are  sEufficient  
quation   5to   𝑓𝑓 5:
:   find  the   3.7
answer  in  this  case).   1,377,965 𝑓𝑓
𝑙𝑙𝑙𝑙
4×350 𝑙𝑙𝑙𝑙 𝑓𝑓𝑓𝑓-- 𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓
𝑣𝑣 = 4×350  
0.0024𝑖𝑖𝑖𝑖
𝑚𝑚𝑚𝑚𝑚𝑚 × 𝑓𝑓𝑓𝑓 = 8,459 𝑓𝑓𝑓𝑓 = 140.99 𝑓𝑓𝑓𝑓  
1 𝑣𝑣 = 11 𝑓𝑓𝑓𝑓 1𝑚𝑚𝑚𝑚𝑚𝑚×𝜋𝜋 12.00𝑖𝑖𝑖𝑖
× 0.05268𝑙𝑙𝑙𝑙 2.51
= 8,459 𝑚𝑚𝑚𝑚𝑚𝑚 =, 140.99 𝑠𝑠  
= −21log 1The  
1 𝑓𝑓𝑓𝑓"* ×𝜋𝜋 velocity   is  d+etermined  n𝑚𝑚𝑚𝑚𝑚𝑚
0.05268𝑙𝑙𝑙𝑙 𝑓𝑓 E=
ext  using   0.01445
𝑠𝑠 5:    
quation  
𝑓𝑓 3.7 1,377,965 𝑓𝑓
As  a  final  check  of  the  adequacy  of  the  Darcy  equation,  the  Mach  number,  the  ratio  of  the  actual  velocity  
As  a  final  check  of  the  adequacy  of  the  Darcy  equation,  the  Mach   number,   𝑙𝑙𝑙𝑙 the  ratio  o-f  the  actual  velocity  
to  the  speed  of  sound  in  the  fluid,  a,  is  checked  to  ensure  that  it   4×350
is  less  t𝑚𝑚𝑚𝑚𝑚𝑚
han  or  equal   𝑓𝑓𝑓𝑓 to  0.3.  To  simplify   𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓
  As
to   the  sfinal
peed  ocheck
f  sound   in  the
the  adequacy
fluid,  a,  is  checked   tDarcy
o  ensure   𝑣𝑣 =it  is  1less  
that   than   ×
or  equal   =the
to  0.3.   T8,459 implify   = 140.99  
o  sratio
the  caalculation,   the  aof ssumption   of  ideal  gof as  the
behavior   is  uequation,
sed,  allowing   the
1 𝑓𝑓𝑓𝑓 ×𝜋𝜋 1t Mach
he   M ach   number,
n
0.05268𝑙𝑙𝑙𝑙 umber   t o   b e   a djusted  
𝑚𝑚𝑚𝑚𝑚𝑚 𝑠𝑠
The  cvalculation,  
the   elocity  is  dtetermined  
he   assumption   next  ouf  sing  
ispeed
deal   Equation  
gas  behavior   5:   is  used,  allowing  the  Mach  number  to  be  adjusted  
of theonly  
based   actual velocity
on  absolute   to the
temperature   from  of the  sound
reference.   in the The  fluid,
speed  o a,f  sisound  
checked in  an  ideal   to ensure
gas  is   that
based  only  on  absolute  temperature   As  f𝑙𝑙𝑙𝑙
rom   the  
a  final   reference.  
check   of  the  Tahe   speed  oof  f  tshe  
dequacy   ound   in  aen  quation,  
Darcy   ideal  gas   is   Mach  number,  the  ratio  of  the
the  
it is less than or equal to 4×3500.3. To simplify- the calculation, the assumption of ideal gas
â3 ∗ _ _å to   the  speed  
𝑚𝑚𝑚𝑚𝑚𝑚 𝑓𝑓𝑓𝑓of  sound  in  the  f𝑓𝑓𝑓𝑓 luid,  a,  is  checked   𝑓𝑓𝑓𝑓 to  ensure  that  it  is  less  than  or  equal  to  0.
behavior 𝑎𝑎 = is â3used, ∗_ _  𝑣𝑣 = 1 the
allowing 1 Mach × number=to8,459 be adjusted = 140.99 based  
only on absolute (9)  
𝑎𝑎 = yã _   å 1 𝑓𝑓𝑓𝑓 the   ×𝜋𝜋calculation,  
0.05268𝑙𝑙𝑙𝑙 the  assumption   𝑚𝑚𝑚𝑚𝑚𝑚of  ideal  gas  b𝑠𝑠ehavior   is  used,   (9)  allowing  the  Mach  number  t
temperatureyã from_the reference. based   The only   speed
on  absolute  of sound in an ideal
temperature   from  gas the  ris eference.  The  speed  of  sound  in  an  ideal  g
As   a  final  check  of  the  adequacy  of  the  Darcy  equation,  the  Mach  number,  the  ratio  of  the  actual  velocity  
Where  
Where  
to  the  speed  of  sound  in  the  fluid,  a,  is  checked  to  â3 ensure  
∗_ _åthat  it  is  less  than  or  equal  to  0.3.  To  simplify  
K   is   t he   r atio   o f   s pecific   h eats   a𝑎𝑎t  t=he   r eference     t (9)
emperature,   5M 20  ach  
R   number  to  be  adjusted  
the  calculation,  
K   is  tthe  
he  aratio  
ssumption   o f   s of  ideal  
pecific   h gas  
eats   a btehavior  
t   he   r yã is  u_sed,  
eference   t allowing  the  
emperature,   5 20   R  
R*   is   the  universal   gas  constant,   1.986   BTU/lbmol-­‐R  
based   only  on  is  
R*   absolute  
t he   u temperature  
niversal   g as   c from  the  
onstant,   1 reference.  
.986   B The  speed  of  sound  in  an  ideal  gas  is  
TU/lbmol-­‐R  
MW   is  the  molecular  weight   Where   of  the  gas  
Where MW   is  t∗he  molecular  weight  of  the  gas  
T   is  
â3the  _ r_ eference   temperature,  520  R  
å
T   𝑎𝑎 = is  the  reference     temperature,   K   520  Ris     the  ratio  of  specific  heats  at  the  reference  t(9)   emperature,  520  R  
TA   Is  
yã the  absolute  
_ temperature  of  the  flowing  vapor  
T   isIs  the
the  aratio
bsolute   oftspecific
emperature   R*   of  the   flowing   vapor   gas  
In  Kthis  case,  the   heats as  atis  
the the  reference
universal   constant,  1.986  
temperature, 520 BTU/lbmol-­‐R  
R
A
speed  of  sound   is  calculated  
Where  
In   tR*
his  case,  the   s peed   o f   s ound  
is the universal gas constant, 1.986 i s   c alculated   MW   a s   is  the  mBTU/lbmol-R
olecular  weight  of  the  gas  
T   is   t he   r eference   temperature,  520  R  
MWK   isis  the
the   rmolecular
atio  of  specific  
0.51768 weight 𝐵𝐵𝐵𝐵𝐵𝐵of
heats   at  tthe gas
he  reference   temperature,  
𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙 520  R  
0.51768 ×1.986 𝐵𝐵𝐵𝐵𝐵𝐵 T   ×520𝑅𝑅×778.17
Is  the  absolute   𝑓𝑓𝑓𝑓t𝐵𝐵𝐵𝐵𝐵𝐵
emperature  
∙ 𝑙𝑙𝑙𝑙𝑙𝑙 430𝑅𝑅 of  the  flowing   𝑓𝑓𝑓𝑓vapor  
T R*   𝑎𝑎isis  
=the 0.39217
the   ureference
niversal   ×1.986 temperature,
gas   𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙A ∙1𝑅𝑅.986  
constant,   520 R
BTU/lbmol-­‐R  
×520𝑅𝑅×778.17 × = 1,288  
0.39217 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
In   this   ∙ 𝑅𝑅 the   𝐵𝐵𝐵𝐵𝐵𝐵 430𝑅𝑅 𝑓𝑓𝑓𝑓
𝑙𝑙𝑙𝑙case,   𝑙𝑙𝑙𝑙𝑙𝑙speed  
∙ 𝑠𝑠 of  sound  
1 is  calculated  
× 520𝑅𝑅as   = 1,288 𝑠𝑠  
TA MW   𝑎𝑎Is= is  the
the  m absolute
olecular  wtemperature eight  𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
17.03 of  𝑙𝑙𝑙𝑙
the  g×as  of the
𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠flowing
1 vapor 520𝑅𝑅 𝑠𝑠
17.03 × 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓
T   is  the  reference  temperature,   𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 520   32.2𝑙𝑙𝑙𝑙R   ∙ 𝑓𝑓𝑓𝑓
TA   0.51768 𝐵𝐵𝐵𝐵𝐵𝐵 𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙
Is  the  absolute  temperature  of  the  flowing  vapor  ×1.986 ×520𝑅𝑅×778.17
In this
The   Mach  case,number   theof  speed
the  vapor   of sound
flow  is  tis hen  calculated
determined   as as  
0.39217 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑅𝑅 𝐵𝐵𝐵𝐵𝐵𝐵 × 430𝑅𝑅 = 1,288
The   Mach  
In  this   number  
case,   of  the  
the  speed   of  vsound  
apor  filow  
s  calculated   s  𝑎𝑎 =
is  then  daetermined   as   𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1 520𝑅𝑅
140.99 17.03 ×
𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓
𝑀𝑀 = 140.99 = 0.109  
0.51768 𝐵𝐵𝐵𝐵𝐵𝐵 𝑀𝑀 =  1,288 = 0.109  𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙
×1.986  1,288
×520𝑅𝑅×778.17 430𝑅𝑅 𝑓𝑓𝑓𝑓
𝑎𝑎 = 0.39217 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
The   Mach   ∙ 𝑅𝑅number  of  the  vapor   𝐵𝐵𝐵𝐵𝐵𝐵 flow  × is  then   determined  
= 1,288 as    
This  result  is  well  below  the  0.3  maximum   𝑙𝑙𝑙𝑙 Mach  𝑙𝑙𝑙𝑙𝑙𝑙 number  ∙ 𝑠𝑠 1 for  applying  these  520𝑅𝑅 equations.   𝑠𝑠
This  result  is  well  below  the  0.3  17.03 maximum  Mach   × number  for  applying  these  equations.  
𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 140.99
13   𝑀𝑀 = = 0.109  
13    1,288
 
 The  Mach  number  of  the  vapor  flow  is  then  determined  as  
This  result  is  well  below  the  0.3  maximum  Mach  number  for  applying  these  equations.  
140.99
𝑀𝑀 = = 0.109  
13    1,288
Technical Paper #6 © IIAR 2017 21
 
This  result  is  well  below  the  0.3  maximum  Mach  number  for  applying  these  equations.  
is  the  molecular  weight  of  the  gas  
T   is  the  reference  temperature,  520  R  
TA   Is  the  absolute  temperature  of  the  flowing  vapor  
In  this  case,  the  speed  of  sound  is  calculated  as  

2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX
0.51768 𝐵𝐵𝐵𝐵𝐵𝐵 𝑓𝑓𝑓𝑓 ∙ 𝑙𝑙𝑙𝑙𝑙𝑙
×1.986 ×520𝑅𝑅×778.17
𝑎𝑎 = 0.39217 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑅𝑅 𝐵𝐵𝐵𝐵𝐵𝐵 × 430𝑅𝑅 = 1,288 𝑓𝑓𝑓𝑓  
𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1 520𝑅𝑅 𝑠𝑠
17.03 ×
𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓
The Mach number of the vapor flow is then determined as
The  Mach  number  of  the  vapor  flow  is  then  determined  as  

140.99
𝑀𝑀 = = 0.109  
 1,288

This  result  is  well  below  the  0.3  maximum  Mach  number  for  applying  these  equations.  
This result is well below the 0.3 maximum Mach number for applying these
13  
 
equations.

The pressure drop can then be determined in the section using Equation 4 multiplied
The  pressure  drop  can  then  be  determined  in  the  section  using  Equation  4  multiplied  by  the  density  and  
by the density and gravitational acceleration.
gravitational  acceleration.  

𝑓𝑓𝑓𝑓 1 𝑓𝑓𝑓𝑓
0.01445×100𝑓𝑓𝑓𝑓×(140.99 ) 32.2 1
∆𝑝𝑝 = ℎ# 𝜌𝜌𝜌𝜌 = 𝑠𝑠 ×0.05268 𝑙𝑙𝑙𝑙 × 𝑠𝑠 1 × 𝑓𝑓𝑓𝑓 = 0.1633𝑝𝑝𝑝𝑝𝑝𝑝  
𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓 - 𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 144𝑖𝑖𝑖𝑖1
2×1𝑓𝑓𝑓𝑓×32.2 32.2
𝑠𝑠 1 𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1

The  pressure  drop  ics  an   acceptable  


then  be  dtetermined  
o  the  baseline   criterion  
in  the   section  fu or   pressure  
sing   Equation  drop   (although  
4  m ultiplied  tbhe  
y  tche  
alculation  
density  ais  nd  
not   shown  
gravitational  
The pressure here,   1drop
0  in.  pisipe  
acceleration.   does  not  allow  
acceptable to the for  abaseline
 0.25  psi  ocriterion
r  less  pressure  
for pressuredrop  in  the  drop
100-­‐ft   sample  
(although
section),  and  therefore  12  in.  pipe  is  the  baseline  size  for  the  mass  flow  and  temperature  in  question.  
the calculation is not shown here, 10𝑓𝑓𝑓𝑓in. pipe does not allow for 𝑓𝑓𝑓𝑓 a 0.25 psi or less
0.01445×100𝑓𝑓𝑓𝑓×(140.99 )1 𝑙𝑙𝑙𝑙 32.2 𝑓𝑓𝑓𝑓 1
pressure
Note  
∆𝑝𝑝 t=hat   ℎ#t𝜌𝜌𝜌𝜌drop
he   in the
p=revious   100-ft sample
calculation   𝑠𝑠 ×0.05268
is  for  a  ssection),
aturated   and therefore
suction   line.  ×The  m12 𝑠𝑠 1 fpipe
in.
ethod   or  
× is the=baseline
calculating   pressure  
0.1633𝑝𝑝𝑝𝑝𝑝𝑝  
𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓 - 𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 144𝑖𝑖𝑖𝑖1
drop  
sizein   for a  sthe
uperheated  
mass flow discharge  
and ltemperature
2×1𝑓𝑓𝑓𝑓×32.2 ine  would   be  exactly   the  same,  using  
in question. 32.2superheated   properties  for  
𝑠𝑠 1 𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1
ammonia.  
The  pressure  drop  is  acceptable  to  the  baseline  criterion  for  pressure  drop  (although  the  calculation  is  
Note
The   thatconsumption  
power   the previous calculation
difference   for  this   ispressure  
for a saturated suction
drop  can  then   line. The
be  calculated   method
using   for
the  pressure  
not  shown  here,  10  in.  pipe  does  not  allow  for  a  0.25  psi  or  less  pressure  drop  in  the  100-­‐ft  sample  
ratios  
calculating of  the  nominal  
pressure and  drop
reduced   in sauction  
superheatedpressure,  discharge using  the  curve  
linefits  
would shown  be in  texactly
he  notes  the for  Tsame,
able  
section),  and  therefore  12  in.  pipe  is  the  baseline  size  for  the  mass  flow  and  temperature  in  question.  
11.   A ssuming   a n   o
using superheated properties for ammonia.verall   8 5°F   c ondensing   t emperature   ( 166.51   p sia)   t he   d ifference   c an   b e   c alculated   as   Deleted:
Note  that  the  previous  calculation  is  for  a  saturated  suction  line.  The  method  for  calculating  pressure  
g*.2*"f g*.2*"f
drop  i𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃
n  a  superheated   13.89line  
discharge   − 0.1633 would  be  exactly   he  same,  u13.89
t0.53567 sing  superheated  𝐵𝐵𝐵𝐵𝐵𝐵 properties   for  
=
The power consumption difference 0.53567 −
for this pressure drop can then be𝑇𝑇𝑇𝑇 ×977.2𝑇𝑇𝑇𝑇
calculated
ammonia.   166.51 166.51
using the pressure=ratios 10.84ℎ𝑝𝑝 of =the8.084𝑘𝑘𝑘𝑘  
nominal and reduced suction pressure, using the
The  
curve fits shown in the notes forthis  
p ower   c onsumption   d ifference   f or   Table pressure  
11. Assuming drop  can  then   anboverall
e  calculated   85°F using   the  pressure  
condensing
As   w ith   t he   p ump   p ower   f or   r ecirculated   l iquid,   t his   m ust   b e   d ivided  
ratios  of  the  nominal  and  reduced  suction  pressure,  using  the  curve  fits  shown  in  the  notes  for   b y   t he   a ssumed   s ystem   efficiency  
T able  
temperature
of   t he   d rive   m (166.51
otor,   h ere   apsia)
ssumed   the
a t   difference
0 .95.   can be calculated as
11.  Assuming  an  overall  85°F  condensing  temperature  (166.51  psia)  the  difference  can  be  calculated  as   Deleted:

8.084𝑘𝑘𝑘𝑘
13.89 − 0.1633 g*.2*"f= 8.509𝑘𝑘𝑘𝑘   13.89 g*.2*"f 𝐵𝐵𝐵𝐵𝐵𝐵
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.53567 0.95 − 0.53567 ×977.2𝑇𝑇𝑇𝑇
166.51 166.51 𝑇𝑇𝑇𝑇
This  section  of  piping  c=auses   an  increase  
10.84ℎ𝑝𝑝 in  power  consumption  in  the  system,  as  shown  above.  This  can  
= 8.084𝑘𝑘𝑘𝑘  
be  used  in  later  steps  to  provide  a  net  present  value  of  the  increased  power  consumption  over  the  life  of  
As  with  the  pump  power  for  recirculated  liquid,  this  must  be  divided  by  the  assumed  system  efficiency  
the  system  that  can  be  compared  with  that  of  other  sizes.  
of  the  drive  motor,  here  assumed  at  0.95.  
22 Note  that  although  the  given  information  8.084𝑘𝑘𝑘𝑘 f©
or  IIAR
the  f2017
low  lists  +0°F  as  the  intermediate  temperature,  
Technical Paper the  #6
power  consumption  listed  assumes  +20°F.  This  difference   is  considered  acceptable  and  should  not  be  
= 8.509𝑘𝑘𝑘𝑘  
0.95
interpreted  as  skewing  the  analysis,  both  because  the  error  is  not  particularly  large  (Jekel  and  Reindl  
2008)   and  because  
This  section   all  ccauses  
of  piping   ompared   pipe  sizes  
an  increase   in  w ill  be  u
power   sed  with  the  isn  ame  
consumption   the  scystem,  
orrelations.  
as  shown  above.  This  can  
ammonia.  

The  power  consumption  difference  for  this  pressure  drop  can  then  be  calculated  using  the  pressure  
ratios  of  the  nominal  and  reduced  suction  pressure,  using  the  curve  fits  shown  in  the  notes  for  T able  
11.  Assuming  an  overall  85°F  condensing  temperature  (166.51  psia)  the  difference  can  be  calculated  as   Deleted:  Table
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping
13.89 − 0.1633 g*.2*"f 13.89 g*.2*"f
𝐵𝐵𝐵𝐵𝐵𝐵
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.53567 − 0.53567 ×977.2𝑇𝑇𝑇𝑇
166.51 166.51 𝑇𝑇𝑇𝑇
= 10.84ℎ𝑝𝑝 = 8.084𝑘𝑘𝑘𝑘  
As with the pump power for recirculated liquid, this must be divided by the assumed
As  with  the  pump  power  for  recirculated  liquid,  this  must  be  divided  by  the  assumed  system  efficiency  
system efficiency of the drive motor, here assumed at 0.95.
of  the  drive  motor,  here  assumed  at  0.95.  

8.084𝑘𝑘𝑘𝑘
= 8.509𝑘𝑘𝑘𝑘  
0.95

This  section  of  piping  causes  an  increase  in  power  consumption  in  the  system,  as  shown  above.  This  can  
be  
This used   in  later  of
section steps  
piping to  provide  
causes a  net  
anpresent   value  
increase inopower
f  the  increased  
consumptionpower  consumption  
in the system,over  tas
he  life  of  
the  system  that  can  be  compared  with  that  of  other  sizes.  
shown above. This can be used in later steps to provide a net present value of the
Note   that  although  
increased powerthe   given  information  
consumption overfor   the  
the flow  
life oflists  
the+0°F   as  the  that
system intermediate   temperature,  
can be compared with the  
power  
that of consumption  
other sizes. listed  assumes  +20°F.  This  difference  is  considered  acceptable  and  should  not  be  
interpreted  as  skewing  the  analysis,  both  because  the  error  is  not  particularly  large  (Jekel  and  Reindl  
2008)  and  because  all  compared  pipe  sizes  will  be  used  with  the  same  correlations.  
Note that although the given information for the flow lists +0°F as the intermediate
The   aforementioned  
temperature, the cpower
orrelation   for  energy  usage  
consumption listed in  tassumes
he  section  +20°F.
is  only  suitable  
This fdifference
or  suction  lines.  
is
Discharge  
considered acceptable and should not be interpreted as skewing the analysis,oboth
l ine   e nergy   a lso   r equires   t he   u se   o f   s implifying   a ssumptions.   G iven   t he   c omplexity   f  varying  
suction  temperatures  and  discharge  temperatures,  an  approximation  is  used  here  such  that,  for  
because the error is not particularly large (Jekel and Reindl 2008) and because all
pressure  ratios  of  suction  to  discharge  less  than  0.2,  assume  that  power  increases  by  0.02  hp/psi.  For  
compared
pressure   pipe
ratios   sizes will
of  suction   be used
to  discharge   of  with
0.2  or  the same
greater,   correlations.
assume   that  power  increases  by  0.01  hp/psi.  

Two-­‐Phase  Flow  
The aforementioned correlation for energy usage in the section is only suitable for
suction lines. Discharge line energy also requires the use of simplifying assumptions.
14  
  Given the complexity of varying suction temperatures and discharge temperatures, an
approximation is used here such that, for pressure ratios of suction to discharge less
than 0.2, assume that power increases by 0.02 hp/psi. For pressure ratios of suction
to discharge of 0.2 or greater, assume that power increases by 0.01 hp/psi.

Two-Phase Flow

The pressure drop for two-phase flow cannot be treated the same as a single-phase
flow. The accuracy of one method versus another is outside the scope of this analysis,
and rather than attempt to prove a case for which one should be used, a method has
simply been chosen and will be uniformly applied to all two-phase flows.

Technical Paper #6 © IIAR 2017 23


2017 IIAR Natural
The   Refrigeration
pressure   drop  for  tConference
wo-­‐phase  f&low  
Heavy Equipment
cannot   Expo,the  
be  treated   Sansame  
Antonio,
as  a  TX
single-­‐phase  flow.  T
of  one  method  versus  another  is  outside  the  scope  of  this  analysis,  and  rather  than  attemp
case  for  which  one  should  be  used,  a  method  has  simply  been  chosen  and  will  be  uniforml
two-­‐phase  flows.    
The  pressure  drop  for  two-­‐phase  flow  cannot  be  treated  the  same  as  a  single-­‐phase  flow.  The  accur
The method presented here is that of Friedel as presented in Thome (1984). The
of  one  method   The  mvethod   ersus  paresented  nother  is  o hutside  
ere  is  tthat   he  socope   of  this  
f  Friedel   as  paresented  
nalysis,  and   in  Trhome  
ather  t(han   attempt  to  prov
21984).  The  pressure  
pressure drop is case   assumed for  wassumed   to be
hich  one  tso   the liquid
hould   be  liquid   pressure
used,  paressure   drop
 method   multiplied
has  m simply   been   by a factor, Φfr .
be  the   drop   ultiplied   by  cahosen  
 factor,   and  
Φfrw 2 ill  be  uniformly  applied  
.  The  pressure  drop  due  
The  pressure  
The drop  for  tdrop
pressure wo-­‐phase   due flow  
to cannot  bis
friction e  tcalculated
reated  the  same   using as  aEquation  single-­‐phase   10. flow.  The  accuracy  
The  pressure  drop  for  two-­‐phase   two-­‐phase   flow   f
c lows.  
annot  
calculated     b e   t reated  
using   t he   s ame   a s   a   s ingle-­‐phase   f low.   T he   a ccuracy  
of  one  method  versus  another  is  outside   the  scope   oEf  quation  
this  analysis,   10.   and  rather  than  attempt  to  prove  a  
of  one  method  versus  another  is  outside  the  scope  of  this  analysis,  and  rather  than  attempt  to  prove  a  
case  for  which  one  should  The   be  um ethod  
sed,   a  mpethod  resented   has  shimply  
ere  is  bthat   2 ocf  
een   Friedel  
hosen   and   as  wpresented   in  Thome  
ill  be  uniformly   (1984).  
applied   to  aTll  he  pressure  drop  is  
case  for   which  one  should  be  used,  a  method  ∆𝑝𝑝 has  =simply  
∆𝑝𝑝w Φb𝑓𝑓𝑓𝑓een     (10)
chosen  and  will  be  uniformly  applied  to  all  
two-­‐phase  flows.    
two-­‐phase  flows.     assumed  to  be  the  liquid  pressure  drop  multiplied  by  a  factor,  Φfr .  The  pressure  drop  due  to  friction 2

calculated  Where   using  Equation  10.  


The  method  presented  here  is  that  of  Friedel  as  presented  in  Thome  (1984).  The  pressure  drop  is  
The  method  Where presented  here  is  that  of  Friedel  as  presented  in  Thome  (1984).  The  pressure  drop  is  
assumed  to  be  the  liquid  pressure  drop   mΔp ultiplied   2 by  a  factor,  Φfr2.  The  pressure  drop  due  to  friction  is  
assumed  to  be  the  liquid  pressure  d∆𝑝𝑝 rop  =m∆𝑝𝑝 wL  Φ𝑓𝑓𝑓𝑓 b
ultiplied    is  
y  tahe   pressure  
 factor,   Φfr2.  dThe   rop  pfor   the  total  
ressure   mdass  
drop   ue  ftlow   flowing  
o  friction   is  as  liquid  only   (10)  
calculated  using  Equation  10.  
0.   pressure Φ fr     is   t he   t wo-­‐phase   m ultiplier  
2
calculated   ΔpuLsing  Equation   is 1the drop for the total mass flow flowing as liquid only
Where   The   l iquid-­‐phase   pressure  drop  is  calculated  as  follows:  
2
Φ ∆𝑝𝑝fr2= ∆𝑝𝑝w Φ𝑓𝑓𝑓𝑓2 is   the two-phase multiplier (10)  
∆𝑝𝑝 = ∆𝑝𝑝w Φ𝑓𝑓𝑓𝑓   ΔpL   is  the  pressure  dwrop  for  the  total  1m"ass  flow  flowing  (10)  
as  liquid   only  
∆𝑝𝑝 w = 4𝑓𝑓 w 8 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 îQî 1:  
Where  
Where   The liquid-phase pressure Φfr    
2
is  the  two-­‐phase  ëmultiplier   ï
drop is calculated as follows:
The  liquid-­‐phase  pressure  drop  is  calculated  as  follows:  
ΔpL   is  the  pressure  drop  for   Where  the  total  mass  flow  flowing  as  liquid  only  
ΔpL   is  the  pressure  drop  for  the  total  mass  flow  flowing  as  liquid  only  
Φfr2     is  is  tthe  
2
he  two-­‐phase  multiplier   ∆𝑝𝑝w = 4𝑓𝑓
w
fL   w 8ë 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
1 "
  (11) (11)  
Φfr     two-­‐phase  multiplier   is  the  liquid   îQîfriction  1:ï factor  (as  calculated  in  Equation  12)  
The  liquid-­‐phase  pressure  drop  is  calculated  as  follows:  
The  liquid-­‐phase  pressure  drop  is  calculated   L  as  follows:   is  the  length  of  the  section  
Where   " Di   is  the  internal  diameter  of  the  pipe  
Where w
∆𝑝𝑝w = 4𝑓𝑓w w 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀îQî11 "   (11)  
∆𝑝𝑝w = 4𝑓𝑓w 8 ë 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀îQî 1:   ρL  
ï is  the  liquid  density   (11)  
8ë fL   1: ï is  the  liquid  friction  factor  (as  calculated  in  Equation  12)  
Massfluxtot   is  the  total  mass  flows  divided  by  flow  area  
Where   ƒL is the liquid friction L   is  the  length   factorof  (as the  calculated
section   in Equation 12)
Where    
is the length D   is   t
ofl(iquid  
the he   i nternal   d iameter   o f   t he   p ipe  
L fsection
i
fL   is  the  liquid  friction   f The  
actor   as   c riction  factor  
alculated   i n   E is  given  
quation   1 by  
2)  
fL   is  the  liquid  friction   ρLf  actor  (as   is  tche   liquid  diensity  
alculated  
L  Di is  the  length   is the f  tinternal
oMassflux
he  section   diameter of n   theEquation  12)  
pipe
L   is  the  length  of  the  section   tot   is  the   𝑓𝑓 total  mass  
=
*.*/f
  f flows  
or   R e  >  2d,000  
ivided  by  flow  area  
Di  ρL is thediameter  
is  the  internal   liquid density of  the  pipe  34 ñ.Ió
w
Di    
is  the  internal  diameter  of  the  pipe  
ρ  Massflux is  the   is the
liquid   density   total
liquid  mass friction  flows divided by
8ë flow area
ρL  L is  the  
totliquid   dThe   ensity   factor  yNOO#Pòô
𝑅𝑅𝑅𝑅 =
is  given  öFö by  
 
Massflux   is  the  total  mass  flows  divided  by  flow  area   ;
Massfluxtottot   is  the  total  mass  flows  divided   *.*/f b y   f low   a rea  
  ñ.Ió  for  
  The liquid friction factor 𝑓𝑓w is = given byRe  >  2,000   (12)  
The  liquid  friction  factor  is  given  by   Where  
34
The  liquid  friction  factor  is  given  by   yNOO#PòôöFö 8ë
*.*/f 𝑅𝑅𝑅𝑅 = fL  
; is  t   he  liquid  friction  factor  (as  calculated  in  Equation  12)   (13)  
𝑓𝑓w = *.*/f  for  Re  >  2,000   for Re > 2,000 (12) (12)  
𝑓𝑓w = 34ñ.Ió  for  Re  >  2,000  
ñ.Ió
𝜇𝜇   is  the  liquid  dynamic  viscosity   (12)  
34
yNOO#PòôöFö 8ë
Where   Di   Is  the  internal  diameter  of  the  pipe  
𝑅𝑅𝑅𝑅 =yNOO#PòôöFö 8ë   (13)  
𝑅𝑅𝑅𝑅 = ;   (13) Massfluxtot   Is  the  sum  of  the  liquid  and  vapor  mass  flows   (13)   divided  by  flow  area  
; fL   is  the  liquid  friction  factor  (as  calculated  in  Equation  12)  
 
Where   𝜇𝜇   is  the  liquid  dynamic  viscosity  
Where   For  an  example  -­‐20°F  wet  suction  flow  of  400  TR,  at  a  recirculation  rate  of  3:1  (returning  2
Di   Is  the  internal  diameter  of  the  pipe  
fL   is  the  liquid  friction   f part  v(apor)  
actor   as  calculated  in  a  100-­‐ft   in  Esquation  
ection  of   10  in.  schedule  40  pipe  (D  =10.02  in.  or  0.835  ft  with  an  in
12)  
fL   is  the  liquid  friction   Massflux factor   tot   (as  
Is  tche   sum  of  the  liquid  and  vapor  mass  flows  divided  iby  flow  area  
2alculated  in  Equation  12)  
𝜇𝜇   is  the  liquid  dynamic  viscosity   0.5476   f t ),   t he   f ollowing   c alculation   determines  the  Friedel  liquid-­‐phase  pressure  drop.  M
𝜇𝜇   is  the  liquid  d   ynamic  viscosity  
D   Is  the  internal  diameter   been   of  cthe  
alculated  
pipe   based  on  saturated  pumped  liquid  from  a  -­‐20°F  recirculator.  
Di  i Is  the  internal   For   an  example  
diameter   of  the  -­‐20°F  
pipe   wet  suction  flow  of  400  TR,  at  a  recirculation  rate  of  3:1  (returning  2  parts  liqu
Massfluxtot   Is  the  sum  of  the  liquid  and  vapor  mass  flows  divided  by  flow  area  
Massfluxtot   Is  the  sum  of   part  
the  vliquid  
apor)  ain   a  v1apor  
nd   00-­‐ft  m section  
ass  flows   of  1d0   in.  schedule  
ivided   by  flow  4a0  rea   pipe  (Di  =10.02  in.  or  0.835  ft  with  an  internal  ar
  215  
  0.5476   f t ),   t he   f ollowing   c alculation   d etermines   t he   Friedel  liquid-­‐phase  pressure  drop.  Mass  flow  h
For  an  example  -­‐20°F  wet  suction  flow     of  400  TR,  at  a  recirculation  rate  of  3:1  (returning  2  parts  liquid,  1  
For  an  example  -­‐20°F  wet  suction  
24 been  calculated  
flow  of  400   based  
TR,  © aot  IIAR
n  
a  sraturated   2017 pumped  
ecirculation   rate  olf  iquid  3:1  (freturning  
rom  a  -­‐20°F  
2  prarts  
ecirculator.  
Technical liquid,   1   #6
Paper
part  vapor)  in  a  100-­‐ft  section  of  10  in.  schedule  40  pipe  (Di  =10.02  in.  or  0.835  ft  with  an  internal  area  of  
part  vapor)   in  a  100-­‐ft  section  of  10  in.  schedule  40  pipe  (Di  =10.02  in.  or  0.835  ft  with  an  internal  area  of  
0.5476  ft22),  the  following  calculation   15   determines  the  Friedel  liquid-­‐phase  pressure  drop.  Mass  flow  has  
0.5476  ft ),  the  following  calculation   determines  the  Friedel  liquid-­‐phase  pressure  drop.  Mass  flow  has  
been  calculated  based  on  s  aturated  pumped  liquid  from  a  -­‐20°F  recirculator.  
been  calculated  based  on  saturated   pumped  liquid  from  a  -­‐20°F  recirculator.  
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Where

ƒL is the liquid friction factor (as calculated in Equation 12)


μ is the liquid dynamic viscosity
Di Is the internal diameter of the pipe
Massfluxtot Is the sum of the liquid and vapor mass flows divided by flow area

For an example -20°F wet suction flow of 400 TR, at a recirculation rate of 3:1
(returning 2 parts liquid, 1 part vapor) in a 100-ft section of 10 in. schedule 40
pipe (Di =10.02 in. or 0.835 ft with an internal area of 0.5476 ft2), the following
calculation determines the Friedel liquid-phase pressure drop. Mass flow has been
calculated based on saturated pumped liquid from a -20°F recirculator.

0.079
𝑓𝑓w = *.12 = 0.004954  
𝑙𝑙𝑙𝑙
411.286
𝑚𝑚𝑚𝑚𝑚𝑚 ×0.835𝑓𝑓𝑓𝑓
0.079
𝑓𝑓w = 0.5476𝑓𝑓𝑓𝑓 1 1𝑚𝑚𝑚𝑚𝑚𝑚
× *.12 = 0.004954  
𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 60𝑠𝑠
411.286
0.000162𝑚𝑚𝑚𝑚𝑚𝑚 ×0.835𝑓𝑓𝑓𝑓
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠
0.5476𝑓𝑓𝑓𝑓 1 1𝑚𝑚𝑚𝑚𝑚𝑚
×
𝑙𝑙𝑙𝑙 60𝑠𝑠
0.000162
𝑓𝑓𝑓𝑓𝑙𝑙𝑙𝑙∙ 𝑠𝑠 1
100𝑓𝑓𝑓𝑓 411.286 1 𝑙𝑙𝑙𝑙𝑓𝑓 ∙ 𝑠𝑠 1 𝑚𝑚𝑚𝑚𝑚𝑚1 𝑓𝑓𝑓𝑓 1
∆𝑝𝑝w = 4×0.00496 𝑚𝑚𝑚𝑚𝑚𝑚 × × ×
0.835𝑓𝑓𝑓𝑓 0.5476𝑓𝑓𝑓𝑓 1 1 2×42.23 𝑙𝑙𝑙𝑙 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 3600𝑠𝑠 1 144𝑖𝑖𝑖𝑖1
𝑙𝑙𝑙𝑙 𝑓𝑓𝑓𝑓 -
100𝑓𝑓𝑓𝑓 411.286 1 𝑙𝑙𝑙𝑙𝑓𝑓 ∙ 𝑠𝑠 1 𝑚𝑚𝑚𝑚𝑚𝑚1 𝑓𝑓𝑓𝑓 1
∆𝑝𝑝w = 4×0.00496 = 0.0009503  𝑝𝑝𝑝𝑝𝑝𝑝   𝑚𝑚𝑚𝑚𝑚𝑚 × × ×
0.835𝑓𝑓𝑓𝑓 0.5476𝑓𝑓𝑓𝑓 1 𝑙𝑙𝑙𝑙 1 144𝑖𝑖𝑖𝑖1
2×42.23 - 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 3600𝑠𝑠
𝑓𝑓𝑓𝑓
The  calculated  liquid  pressure  drop  is  used  with  the  Friedel  multiplier.  This  multiplier  is  calculated  by  use  
= 0.0009503  𝑝𝑝𝑝𝑝𝑝𝑝  
of  the  following  family  of  equations.  
The  calculated  liquid  pressure   drop  is  used  with  the  Friedel  multiplier.  This  multiplier  is  calculated  by  use  
3.24𝐹𝐹𝐹𝐹
1
Φ = 𝐸𝐸 +
of  the  following  family     (14)  
#Ö 𝐹𝐹𝐹𝐹o𝐻𝐻f  ∙ equations.  
0.045
𝑊𝑊𝑊𝑊𝐿𝐿 0.035

1 yNOO#PòôöFö 3.24𝐹𝐹𝐹𝐹
I
Φ
𝐹𝐹𝐹𝐹#Ö
ü =
= 𝐸𝐸 + I 0.045
    (14)  
(15)  
x8ë :𝐹𝐹𝐹𝐹
†𝐻𝐻 ∙ 𝑊𝑊𝑊𝑊𝐿𝐿 0.035

I : #
yNOO#Pòô
𝐸𝐸 ü= =1 − 𝑥𝑥 1 +IöFö𝑥𝑥 1   ï °  
𝐹𝐹𝐹𝐹 (16)  
(15)  
x8ë :† : # ° ï

= 𝑥𝑥1*./e
𝐹𝐹 =
𝐸𝐸
*.11L
− 𝑥𝑥11 −+𝑥𝑥𝑥𝑥 1 ï  °  
: # (17)  
(16)  
:° #ï
: *.f" ; *."f ; *./
𝐻𝐻 =
𝐹𝐹
ï
= 𝑥𝑥 *./e ° *.11L °
  1 − ;ï   (18)  
(17)  
:° 1 − 𝑥𝑥

Technical Paper *.f" ; *."f


#6
:ïyNOO#Pòô I ; *./ © IIAR 2017 25
öFö 8ë
𝐻𝐻
𝑊𝑊𝑊𝑊=w =:
°
  1 − ;
°
  (18)  
(19)  
° ¢:;†ï ï

g" I
yNOO#Pòô
ô "gô öFö 8ë
𝑊𝑊𝑊𝑊
𝜌𝜌ü w== + ¢:     (19)  
(20)  
𝑙𝑙𝑙𝑙 60𝑠𝑠
0.000162
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠

𝑙𝑙𝑙𝑙 1
100𝑓𝑓𝑓𝑓 411.286 1 𝑙𝑙𝑙𝑙𝑓𝑓 ∙ 𝑠𝑠 1 𝑚𝑚𝑚𝑚𝑚𝑚1
∆𝑝𝑝wIIAR
2017 = 4×0.00496 𝑚𝑚𝑚𝑚𝑚𝑚
Natural Refrigeration Conference & Heavy1Equipment Expo, San ×Antonio, TX × ×
0.835𝑓𝑓𝑓𝑓 0.5476𝑓𝑓𝑓𝑓 𝑙𝑙𝑙𝑙 1
2×42.23 - 32.2𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 3600𝑠𝑠
𝑓𝑓𝑓𝑓
= 0.0009503  𝑝𝑝𝑝𝑝𝑝𝑝  

The  calculated  
The calculated liquid pressureliquid  
droppis ressure  
used dwith
rop  is  
the used  
Friedelwith  tmultiplier.
he  Friedel  mThis
ultiplier.  
multiplierThis  multiplier  is  calcu
is calculated byof  
use the  
offollowing   family  ofamily
the following f  equations.  
of equations.
3.24𝐹𝐹𝐹𝐹
Φ#Ö 1 = 𝐸𝐸 +
(14)  
𝐹𝐹𝐹𝐹𝐻𝐻 ∙0.045 𝑊𝑊𝑊𝑊𝐿𝐿 0.035

I
yNOO#Pòô
𝐹𝐹𝐹𝐹ü =
(15)
öFö
 
x8ë :† I

: #
𝐸𝐸 = 1 − 𝑥𝑥 1 + 𝑥𝑥 1 ï °  
(16) :° #ï

𝐹𝐹 = 𝑥𝑥 *./e 1 − 𝑥𝑥 *.11L  
(17)
: *.f" ; *."f ; *./
𝐻𝐻 = ï °
1− °  
: ;
(18)
° ï ;ï

I
yNOO#Pòô 8
𝑊𝑊𝑊𝑊w = öFö ë
 
(19)
¢: †

ô "gô g"
𝜌𝜌ü = +   (20)
:° :ï

Where  
Where
G   is  the  acceleration  due  to  gravity  
G ρL  
is the acceleration is  the  
due liquid   density  
to gravity
ρL ρG  
is the liquid is  the  vapor  density  
density
ρG μL  
is the vapor is  the  liquid  dynamic  viscosity  
density
μL μG  
is the liquid is  the  vapor  
dynamic viscositydynamic  viscosity  
μG Di  
is the vapor is  the  internal  
dynamic viscosity diameter  of  the  pipe  
Di is Massflux tot   diameter
the internal is  the  sum  of
of  the
the  pipe
liquid  and  vapor  mass  flows  divided  by  flow  area  
Massfluxtot is the sum x   of the is  the  
liquid liquid/vapor  
and vapor quality  
mass(between   0  and  1by
flows divided )   flow area
x σ  
is the liquid/vapor is  the  quality
surface  (between
tension     0 and 1)
σ is   the surface tension

Table 3 shows the properties of ammonia for this flow.


16  
 

26 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping
Table  3  shows  the  properties  of  ammonia  for  this  flow.  

Table  3.  Properties  for  two-­‐phase  ammonia  flow  at  400TR,  3:1  recirculation  rate  

Temperature  
Temperature (°F)  
(°F) -­‐20.0   -20.0
Saturated  
Saturated pressure  
pressure (psia)  
(psia) 18.28   18.28
3
Liquid  density  (lb/ft )   3 42.23  
Liquid density (lb/ft ) 42.23
Liq.  dynamic  viscosity  (lbm/ft-­‐s)   0.0001616  
Liq. dynamic viscosity 3 (lbm/ft-s) 0.0001616
Vapor  density  (lb/ft )   3 0.06810  
Vapor density (lb/ft ) 0.06810
Vap.  dynamic  viscosity  (lbm/ft-­‐s)   5.500E-­‐6  
Vap. dynamic viscosity
Vapor  quality  (ratio  of  vapor   (lbm/ft-s) 0.3333  
5.500E-6
Vapor quality (ratio of vapor
mass  to  total  mass  in  the  control   mass to total mass in the control volume) 0.3333
Surface tension (lbf/ft)
volume)   0.002267
Surface  
Table 3. tProperties
ension  (lbf/ft)  
for two-phase ammonia 0.002267  
flow at 400TR, 3:1 recirculation rate
 

Thevarious  
The   various factors
factors   in Equations
in  Equations   14 through
14  through   20  may  be  20 may bewcalculated
calculated   within  
ith  the  properties   the properties
Table   3.  
Performing  
in Table 3. these   calculations,  
Performing these calculations,
g"

0.3333 1 − 0.3333 𝑙𝑙𝑙𝑙


𝜌𝜌ü = + = 0.2037  
𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑓𝑓𝑓𝑓 -
0.06810 - 42.23 -
𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓

𝑙𝑙𝑙𝑙
411.286 𝑙𝑙𝑙𝑙
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀îQî = 𝑚𝑚𝑚𝑚𝑚𝑚 = 750.7  
0.5479𝑓𝑓𝑓𝑓 1 𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝑓𝑓𝑓𝑓 1

𝑙𝑙𝑙𝑙
(750.7 )1 ×0.8350𝑓𝑓𝑓𝑓 𝑚𝑚𝑚𝑚𝑚𝑚1
𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝑓𝑓𝑓𝑓 1
𝑊𝑊𝑊𝑊w = × = 8,809  
𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 ∙ 𝑓𝑓𝑓𝑓 3600𝑠𝑠 1
0.002267 ×0.2037 - ×32.20
𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓 𝑙𝑙𝑙𝑙𝑙𝑙 ∙ 𝑠𝑠 1
*.f" *."f *./
𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙
42.23 0.000005500 0.000005500
𝑓𝑓𝑓𝑓 - 𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠
𝐻𝐻 = 1− = 178.52  
𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙
0.06810 - 0.0001616 0.0001616
𝑓𝑓𝑓𝑓 𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠

𝐹𝐹 = 0.3333*./e 1 − 0.3333 *.11L


= 0.3876  

0.079
𝑓𝑓x = *.12 = 0.002128  
𝑙𝑙𝑙𝑙
751.07 ×0.8350𝑓𝑓𝑓𝑓
𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝑓𝑓𝑓𝑓 1
𝑙𝑙𝑙𝑙 60𝑠𝑠
0.000005500 ×
𝑓𝑓𝑓𝑓 ∙ 𝑠𝑠 𝑚𝑚𝑚𝑚𝑚𝑚

17  
  Technical Paper #6 © IIAR 2017 27
2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

𝑙𝑙𝑙𝑙
42.23
- ×0.002128
1 1 𝑓𝑓𝑡𝑡
𝐸𝐸 = 1 − 0.3333 + 0.3333 = 30.483  
0.06810 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 ×0.00496
42.23 𝑓𝑓𝑓𝑓 ×0.002128
-
𝑓𝑓𝑡𝑡 -
𝐸𝐸 = 1 − 0.33331 + 0.33331 = 30.483  
𝑙𝑙𝑙𝑙
𝑙𝑙𝑙𝑙 0.06810
1 ×0.00496
751.07 𝑓𝑓𝑓𝑓 -
𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝑓𝑓𝑓𝑓 1 𝑚𝑚𝑚𝑚𝑚𝑚1
𝐹𝐹𝐹𝐹ü = × = 140.53  
𝑓𝑓𝑓𝑓 𝑙𝑙𝑙𝑙 1
𝑙𝑙𝑙𝑙 1 3,600𝑠𝑠 1
32.2 1 ×0.835𝑓𝑓𝑓𝑓×
751.07 0.2037
𝑠𝑠 𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝑓𝑓𝑓𝑓 1 𝑓𝑓𝑓𝑓 - 𝑚𝑚𝑚𝑚𝑚𝑚1
𝐹𝐹𝐹𝐹ü = 𝑙𝑙𝑙𝑙 1 × 3,600𝑠𝑠 1 = 140.53  
𝑓𝑓𝑓𝑓 42.23𝑙𝑙𝑙𝑙 - ×0.002128
32.21 1 ×0.835𝑓𝑓𝑓𝑓×
1 0.2037 𝑓𝑓𝑡𝑡
3.24×0.3876×178.5
𝐸𝐸 = 1Φ −#Ö0.3333 𝑠𝑠= 30.44 ++ 0.33331 𝑓𝑓𝑓𝑓 - = 30.483  
0.045 𝑙𝑙𝑙𝑙0.035 = 161.07  
 140.53   8
0.06810 - ×0.00496 809
3.24×0.3876×178.5 𝑓𝑓𝑓𝑓
1
Φ
With  the  calculated  Friedel  factor,  
#Ö = 30.44 +
multiplying  the  liquid  
0.045 friction  pressure  = 161.07  
drop  to  attain  the  overall  
 140.53 1  88090.035
pressure  drop  of  the  flow  is  now  possible.   𝑙𝑙𝑙𝑙
With the calculated Friedel factor, 751.07multiplying the liquid friction
𝑚𝑚𝑚𝑚𝑚𝑚 1 pressure drop to
𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝑓𝑓𝑓𝑓 1
With  the  calculated  F𝐹𝐹𝐹𝐹
riedel  
ü = f actor,   m ultiplying   t he   l iquid   f riction   1 ×p ressure  
1
d=
rop   to  attain  the  overall  
140.53  
attain the overall pressure drop
𝑓𝑓𝑓𝑓 of the flow is now
𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 possible. 3,600𝑠𝑠
pressure  drop  of  the  flow  
∆𝑝𝑝 i= s  32.2
n∆𝑝𝑝
ow  Φ
possible.  
2
×0.835𝑓𝑓𝑓𝑓× 0.2037 ×161.07
w𝑠𝑠 1𝑓𝑓𝑓𝑓 = 0.0009503 𝑓𝑓𝑓𝑓
2
- = 0.153𝑝𝑝𝑝𝑝𝑝𝑝  
𝑖𝑖𝑖𝑖
2 𝑙𝑙𝑙𝑙𝑙𝑙
3.24×0.3876×178.5
The  resulting  pressure  ∆𝑝𝑝 ∆𝑝𝑝 Φ𝑓𝑓𝑓𝑓criterion  
=fits  
drop   = 0.0009503 ×161.07 = 0.153𝑝𝑝𝑝𝑝𝑝𝑝  
Φ#Ö 1wthe  
= 30.44 + for  piping   at  2this  
𝑖𝑖𝑖𝑖
0.045
temperature  
0.035
for  pressure  
= 161.07   drop  per  100  ft  
 140.53  8809
of  piping  not  to  exceed  0.25psi  as  a  baseline  case.  Pressure  drop  for  this  flow  through  an  8  in.  pipe  
The   resulting  
exceeds   ressure  ft.   drop   fits  the  criterion   for  bpaseline  
iping  at  ize  
this  for  
temperature   for  pressure     dthe  
rop  per  100  ft  
With   the  0c.25   psi/100  
alculated   Friedel   Therefore,  
factor,  m1ultiplying   0  in.  is  the  
the  liquid  fsriction   this  
p two-­‐phase  
ressure   drop  tfo  low.  
attain   overall  
of  
The piping   not  to  exceed  
resulting pressure 0.25psi   drop as  afits
 baseline  
the criterion case.  Pressure   for piping drop  for  atthis   flow  
this through  an  8  for
temperature in.  pipe  
pressure  
Again,  as  0w d rop   o f  
ith  ptsi/100  t he  
he  pressure  f low   i s   n ow   p ossible.  
drops  calculated   for  bdaseline  
ry  vapor   flow,  
exceeds   .25   ft.  Therefore,   10  in.  is  the   size   for  tthis  
his  ptwo-­‐phase  
ressure  drop  
flow.  can  
  be  used  with  the  
pressure
numbers   or  drop
equations   per i100 ft of piping
n  T able  11   not toincreased  
to  determine   exceed 0.25psi p ower   c as a baseline
onsumption   a t   t case.
he   Pressurefor  
compressors   Deleted:  
2 𝑙𝑙𝑙𝑙𝑙𝑙
Again,  
drop
the   flow   afor
s  listed.  
with  
this the  
Aflow ssuming   ∆𝑝𝑝
pressure  
through =
85°F   ∆𝑝𝑝
drops   Φ
w an c𝑓𝑓𝑓𝑓
saturated   =
alculated   0.0009503
8 in.condensing  
pipe for  dexceeds
ry  vcapor   ×161.07
flow,  
0.25this  
ondition,   psi/100 = 0.153𝑝𝑝𝑝𝑝𝑝𝑝
pressure  
ft. dTherefore,  
rop  can  be  u10 sed  in.
with  
isthe  
𝑖𝑖𝑖𝑖2
numbers  
the baseline or  equations  
size for in  Tthis
able  11   two-phaseto  determine   flow.increased  power  consumption   at  the  compressors  for   Deleted:  
g*.2*"f g*.2*"f
The  
the  frlow  
esulting  
listed.   pressure  
Assuming   18.8
drop  
85°F   −
fits   t0.153
he  criterion  
saturated   for  piping  
condensing   at  this  18.8
condition,   temperature  for   𝐵𝐵𝐵𝐵𝐵𝐵
pressure  drop  per  100  ft  
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.53567 − 0.53567 ×400𝑇𝑇𝑇𝑇 = 2.50ℎ𝑝𝑝
of  piping  not  to  exceed  0.25psi   166.51 as  a  baseline  case.  Pressure  d166.51 rop  for  this  flow  t𝑇𝑇𝑇𝑇hrough  an  8  in.  pipe  
Again,0as
exceeds   .25  
with the
= pressure
ft.  T18.8
1.86𝑘𝑘𝑘𝑘   − 0.153 drops g*.2*"fcalculated for dry
size  f18.8
vapor flow,𝐵𝐵𝐵𝐵𝐵𝐵
g*.2*"f this pressure drop
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 =p0.53567
si/100   herefore,   10  in.  is  the  − baseline  
0.53567 or  this  two-­‐phase   flow.    
×400𝑇𝑇𝑇𝑇 = 2.50ℎ𝑝𝑝
can be used with the numbers 166.51 or equations in Table 166.51 11 to determine 𝑇𝑇𝑇𝑇 increased
These  laiquid,  
Again,   s   w dtry  
ith   he   vapor,  
p ressure   a
= 1.86𝑘𝑘𝑘𝑘   nd  d t wo-­‐phase  
rops   c c
alculated  alculations  
f or   d m
ry   v ust  
apor   tfhen   b
low,   te   r epeated  
his   p f or  
ressure   d the  csan  
rop   izes   immediately  
be   used   with  the  
power consumption at the compressors for the flow listed. Assuming 85°F saturated
adjacent   t o   t he   b aseline   c ase.  
numbers  or  equations  in  T able  11  to  determine  increased  power  consumption  at  the  compressors  for   Deleted:  
condensing
These   liquid,   condition,
dry   vapor,  and   two-­‐phase   calculations  
the  flow   listed.   Assuming   85°F   saturated   condensing  mcust   then  be  repeated  for  the  sizes  immediately  
ondition,  
Table  4   s hows   t he  
adjacent  to  the  baseline  case.   r esults   o f   t his   a nalysis   f or   a ll   t hree   f low   types.  The  power  consumption  will  be  used   Deleted:  
in  Step  3  to  calculate  a  present-­‐value   18.8 − 0.153 l ifecycle  
g*.2*"fcost  for  the  piping  sections  
18.8 g*.2*"f analyzed  
𝐵𝐵𝐵𝐵𝐵𝐵 (vapor  piping  is  used  
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃
Table  4  
as  the   shows  
example   = 0.53567
tfhe  
low  results  
for  those   of  this   analysis  for  all  
calculations).   −three  
0.53567 flow  types.  The  power  consumption   ×400𝑇𝑇𝑇𝑇w= ill  2.50ℎ𝑝𝑝
be  used   Deleted:  
166.51 166.51 𝑇𝑇𝑇𝑇
in  Step  3  to  calculate  a  present-­‐value  lifecycle  cost  for  the  piping  sections  analyzed  (vapor  piping  is  used  
= 1.86𝑘𝑘𝑘𝑘  
Table  4.  Comparison  of  pressure  drop  and  energy  consumption  for  liquid,  vapor,  and  two-­‐
as   the  example  flow  for  those  calculations).  
phase  flows  
These   liquid,  dry  vapor,  and  two-­‐phase  calculations  must  then  be  repeated  for  the  sizes  immediately  
Table  4.  Comparison  of  pressure  drop  and  energy  consumption  for  liquid,  vapor,  and  two-­‐
adjacent  to  the  baseline  case.  
phase  flows   Line  Size   Pressure  Drop  (psi/100   Power  Consumption  
ft)   (kW)  
Table  4  shows  the  results  of  this  analysis  for  all  three  flow  types.  The  power  consumption  will  be  used   Deleted:  
Liquid  
in  Step  3  Line   Size   a  present-­‐value  
to  calculate   Pressure  Dlifecycle   rop  (psi/100  
cost  for  the  Power   Consumption  
piping   sections  analyzed  (vapor  piping  is  used  
2  in.  schedule  40   2.911  
ft)   0.1696  
(kW)  
as  t2  he   example  
1/2   in.  schedule   flow  f4or   0  those  calculations).   1.161  © IIAR 2017 0.06767  
28 Liquid   Technical Paper #6
3  in.  schedule  40  
2   0.380  
2.911   0.02216  
0.1696  
Table  4.  Comparison  of  pressure  drop  and  energy  consumption  for  liquid,  vapor,  and  two-­‐
2  1/2  in.  schedule  40   1.161   0.06767  
phase  flows  
18   3  in.  schedule  40   0.380   0.02216  
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

These liquid, dry vapor, and two-phase calculations must then be repeated for the
sizes immediately adjacent to the baseline case.

Table 4 shows the results of this analysis for all three flow types. The power
consumption will be used in Step 3 to calculate a present-value lifecycle cost for
the piping sections analyzed (vapor piping is used as the example flow for those
calculations).

Line Size Pressure Drop (psi/100 ft) Power Consumption (kW)


Liquid
2 in. schedule 40 2.911 0.1696
2 1/2 in. schedule 40 1.161 0.06767
3 in. schedule 40 0.380 0.02216
Vapor
10 in. Std (schedule 40) 0.4118 21.75
12 in. Std (schedule 40) 0.1633 8.509
14 in. Std (schedule 40) 0.1085 5.634
Two-Phase
8 in. schedule 40 0.404 8.727
10 in. Std (schedule 40) 0.153 3.263
12 in. Std (schedule 40) 0.0682 1.448
Table 4. Comparison of pressure drop and energy consumption for liquid, vapor, and two-phase flows

Step 3 – Determine first costs

The analysis now only requires that first costs be determined, along with a present
value of the power consumption of each piping section, to determine a best economic
size for each case. These costs are determined in accordance with the rules stated
earlier.

Technical Paper #6 © IIAR 2017 29


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Tables 12 through 15 provide information for determining first costs associated with
the piping chosen in Step 2. These costs are essentially summed from the tables and
used to provide the first cost for each size in each service.

As an example, assume that all three pipe sizes for the vapor piping shown in
Table 4 are of carbon steel, are schedule 40, and are insulated with 3-in. insulation
with aluminum jacket. From Table 12, Table 14, and Table 15, the following can be
developed to sum these costs.

Table 5 assumes an all-inclusive labor rate (labor, burden, etc.) of $50/hour for
piping assembly and $40/hour for insulation labor. As mentioned previously, material
costs are taken directly from the data source, RS Means, and are assumed to be U.S.
national averages. Note that all costs are assumed to be for a 100-ft length, so all
costs from respective tables are multiplied by 100.

Note also that the insulation jacket is based on jacket inside diameters of 21 in., 23
in., and 24 in. for the 10 in., 12 in., and 14 in. pipes respectively (pipe OD plus 2X
insulation thickness of 3 in.). In the interest of a concise table, labor and materials
for painting of the piping are included in the piping numbers.

Insul. Total
Line Pipe Insul. Insul. Total
Pipe Pipe Insul. Jacket Labor Total First
Size Labor Labor Jacket Mat’l
Mat’l Cost ($) Cost Cost Cost Cost ($)
(in.) Hours Hours Labor Cost ($)
($) ($)
10 A53B $6,080 88 $2,400 21.3 $425 17.2 $8,905 $5,940 $14,845
12 A53B $8,312 111.14 $2,650 22.9 $474 17.8 $11,436 $7,185 $18,621
14 A53B $10,282 144 $3,000 24.6 $499 18 $13,781 $8,904 $22,685
Table 5. First costs for vapor piping comparison

This comparison is straightforward and needs now only the energy costs from the
pressure losses calculated in Step 2 to give a picture of the cost of ownership for each
segment of piping. Note that, for this limited data set, first costs do not appear to

30 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

track linearly with pipe size. Using the 10 in. pipe as a baseline, while size increases
20% and 40% respectively to 12 in. and 14 in., the first costs increase 24% and 52%.
However, this is not enough information to determine whether the behavior is or is
not roughly linear.

Step 4 – Determine present value of energy costs

For this portion of the analysis, four pieces of information are required to calculate
the present value of the energy costs used by the system:

1. Yearly hours at the assumed condition,

2. Life of the system (years),

3. Energy cost per kWh, and

4. Inflation rate.

The choice of a value for item 1 is difficult at times, given that the design of a
system pipe size is always for peak loading, but in many situations, peak load
only occurs for a few days per year, such as in distribution centers where load is
heavily dependent on ambient conditions. In process applications, peak load in a
particular header may occur each day for one or two full shifts. Particular care must
be exercised in choosing the yearly run time as this value has the largest chance of
skewing the analysis.

As an example, the pure vapor flows given previously (in Step 2) are analyzed here
with the assumption that these loads are from processing equipment running for 16
hours per day, 6 days per week. This provides a value for yearly run time of 4,992
hours and can be considered an extreme loading case.

Technical Paper #6 © IIAR 2017 31


4.   Inflation  rate.  

The  choice  of  a  value  for  item  1  is  difficult  at  times,  given  that  the  design  of  a  system  pipe  siz
for  peak  loading,  but  in  many  situations,  peak  load  only  occurs  for  a  few  days  per  year,  such
distribution  centers  where  load  is  heavily  dependent  on  ambient  conditions.  In  process  appl
2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX
peak  load  in  a  particular  header  may  occur  each  day  for  one  or  two  full  shifts.  Particular  care
exercised  in  choosing  the  yearly  run  time  as  this  value  has  the  largest  chance  of  skewing  the

As  an  example,  the  pure  vapor  flows  given  previously  (in  Step  2)  are  analyzed  here  with  the  
that  these  
The following will be given loads  ato
as inputs re  fthe
rom  lifecycle
processing  
costequipment  
analysis:running  for  16  hours  per  day,  6  days  per  we
provides  a  value  for  yearly  run  time  of  4,992  hours  and  can  be  considered  an  extreme  loadin
• A system life of 25 years,
• An annual inflation The  
ratefollowing  
of 4%, wandill  be  given  as  inputs  to  the  lifecycle  cost  analysis:  
• A cost of energy of $0.11/kW.
•   A  system  life  of  25  years,  
•   An  annual  inflation  rate  of  4%,  and  
The net present value of the system can then be determined as follows.
•   A  cost  of  energy  of  $0.11/kW.    

The  net  of


The present value of a series present  
monetary value  ovalues,
f  the  system   can  then  be  yearly
compounded determined  
is givenas  follows.  
in
Equation 21. The  present  value  of  a  series  of  monetary  values,  compounded  yearly  is  given  in  Equation  21

"§a • g"
𝑃𝑃𝑃𝑃 = 𝐴𝐴î   (21)
a∙ "§a •

Where  
Where
 PW   is  the  present  worth  of  an  alternative  
I   is  the  annual  inflation  rate  
PW is the present worth of an alternative
N   is  the  system  life  in  years  
I is the annual inflation rate
T   is  the  year  of  interest  in  the  sum  
N is the system lifeAin years
is  the  yearly  cost  of  energy  (see  Development  of  the  Method)  
t  
T is the year of
  interest in the sum
At is the yearly cost of energy (see Development of the Method)
20  
 
The multiplier on the right side of the equation is easily calculated, although tables of
The  are
such values multiplier   on  the  right  in
also compiled side  
variousof  the  eresources
quation  is  e(e.g.,
asily  calculated,  
White etalthough   tables  
al. 1998). Forof  asuch  
4%values  are  
inflation also  
rate,compiled  
the PWin  factor
various  resources  (e.g.,  White  et  al.  1998).  For  a  4%  inflation  rate,  the  PW  factor  for  a  
for a 25-year series of annual cash flows is 15.6221. The
25-­‐year  series  of  annual  cash  flows  is  15.6221.  The  present  value  of  the  energy  cost  (At)  for  the  three  
present value of the energy cost (At) for the three alternative vapor pipes becomes
alternative  vapor  pipes  becomes  

$*.""
(21.75𝑘𝑘𝑘𝑘×4,992ℎ𝑟𝑟× )×15.6221 = $186,580.36  (10  in.  schedule  40  pipe)  
âã∙ßÖ

$*.""
(8. 509𝑘𝑘𝑘𝑘×4,992ℎ𝑟𝑟× )×15.6221 = $72,993.67   (12  in.  schedule  40  pipe)  
âã∙ßÖ

$*.""
(5.634𝑘𝑘𝑘𝑘×4,992ℎ𝑟𝑟× )×15.6221 = $48,330.75   (14  in.  schedule  40  pipe)  
âã∙ßÖ

Clearly,  from  these  results,  where  energy  cost  and  system  life  are  relatively  high,  the  cost  of  ownership  
favors  a  larger  pipe.  T able  6  compares  total  cost  of  ownership  for  the  vapor  pipe  alternatives  in  
question.  

Table  6.  Summary  of  vapor  piping  lifecycle  cost,  25-­‐year  system  life,  4,992  hr/yr,  
$0.11$/kWh    
32 © IIAR 2017 Technical Paper #6
Size   First  Cost   Present  Value  of   Total  Cost  of  the  
Energy  Cost   Alternative  
10  in.   $14,845   $186,580   $201,425  
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Clearly, from these results, where energy cost and system life are relatively high, the
cost of ownership favors a larger pipe. Table 6 compares total cost of ownership for
the vapor pipe alternatives in question.

Size First Cost Present Value of Energy Cost Total Cost of the Alternative
10 in. $14,845 $186,580 $201,425
12 in. $18,621 $72,994 $91,615
14 in. $22,685 $48,331 $71,016
Table 6. Summary of vapor piping lifecycle cost, 25-year system life, 4,992 hr/yr, $0.11$/kWh

The cost of ownership falls rapidly with increased pipe size in this particular case.
However, note that a lower peak-load duty, lower lifecycle, or lower energy costs
can skew the analysis. With the same piping and load, but for a 15-year system life
and loads near peak for only 1/8 of the year’s operating hours (1,095) (Table 7), the
results still favor the 14 in. pipe section (total the cost is essentially identical to that
of a 12” pipe section). Note that the present value factor for a 15-year life at 4% is
11.1184.

Size First Cost Present Value of Energy Cost Total Cost of the Alternative
10 in. $14,845 $29,128 $43,973
12 in. $18,621 $11,395 $30,016
14 in. $22,685 $7,545 $29,525
Table 7. Summary of vapor piping lifecycle cost, 15-year system life, 1,095 hr/yr, $0.11/kWh

If the same analysis is performed with the cost of energy lowered to $0.08/kWh, the
favorable pipe diameter changes to 12 in., as shown in Table 8.

Technical Paper #6 © IIAR 2017 33


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Size First Cost Present Value of Energy Cost Total Cost of the Alternative
10 in. $14,845 $21,184 $36,029
12 in. $18,621 $8,288 $26,909
14 in. $22,685 $5,487 $28,172
Table 8. Summary of vapor piping lifecycle cost, 15-year system life, 1,095 hr/yr, $0.08/kWh

Clearly, in this particular scenario, the breakeven point between 12 in. schedule 40
and 14 in. schedule 40 piping is around $0.11/kWh. In reality, peak load operation
for even 1/8 of the hours in a year is likely to be a high estimate for many facilities,
but quite low for continuous process operations (like spiral freezers, plate freezers,
and cooling tunnels), and careful consideration of the process is necessary to ensure
the yearly operating hours are realistic.

To broaden the exercise a bit further, Table 5 can be expanded by adding the first cost
for 16 in. pipe, again with 3 in. insulation, with the 4,992 hr/yr, 25-year life inputs
applied (see Table 9).

Insul.
Line Pipe Insul. Insul. Total Total Total
Pipe Pipe Insul. Jacket
Size Labor Labor Jacket Mat’l Labor First
Mat’l Cost ($) Cost Cost
(in.) Hours Hours Labor Cost ($) Cost ($) Cost ($)
($)
10 A53B $6,080 88 $2,400 21.3 $425 17.2 $8,905 $5,940 $14,845
12 A53B $8,312 111.14 $2,650 22.9 $474 17.8 $11,436 $7,185 $18,621
14 A53B $10,282 144 $3,000 24.6 $499 18 $13,781 $8,904 $22,685
16 A53B $11,931 166.14 $3,300 26.7 $550 18.6 $15,781 $10,119 $25,900
Table 9. First costs for vapor piping comparison, 25-year system life, 4,992 hr/yr, $0.11/kWh,
including 16 in. diameter

The pressure drop and energy costs for a 16 in. pipe with 977 TR of -30°F dry vapor
are 0.0604 psi and 3.641 kW (using the methods described in Step 2). Expanding
Table 6 to include 16 in. pipe, the lifecycle cost comparison is shown in Table 10.

34 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Size First Cost Present Value of Energy Cost Total Cost of the Alternative
10 in. $14,845 $186,580 $201,425
12 in. $18,621 $72,994 $91,615
14 in. $22,685 $48,331 $71,016
16 in. $25,900 $23,265 $49,165
Table 10. Summary of vapor piping lifecycle cost, 25-year system life, 4,992 hr/yr, $0.11$/kWh

Clearly, on a heavily used process pipe, investment in a large header provides ample
payback and should be considered where conditions are favorable, such as in the
aforementioned case. In plants with diverse loads, installation of large pipes may
not be the correct approach for every header in the plant, but can certainly be cost-
effective for certain targeted areas of heavy use.

The aforementioned methodology provides a clear, stepwise means of comparing the


economic impact of increasing or decreasing pipe size. While the particular items
shown may or may not reflect actual economic conditions for any particular project,
its importance lies in the ability of a user, armed with the proper information, to use
engineering judgement to ascertain what sizes may provide the highest economic
benefit to the end user. Clearly, a “cheap” project has the potential to cost a great
deal of money over the life of the system.

While the “best economic” pipe, chosen in accordance with the criteria presented
cannot always be used in situations where operational considerations require other
pipe sizes, the method is acceptable under various conditions to aid the designer in
providing the best lifecycle cost plant to the owner/end user.

Technical Paper #6 © IIAR 2017 35


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Conclusion

A method of analysis for economic pipe sizing in ammonia headers was developed,
and examples were provided illustrating its use. It was shown that the economic pipe
size depends heavily on the cost of energy and the anticipated system usage and life.

While the exact examples provided may not be applicable to any particular
installation, they are indicative of what results the analysis can have based on
varying certain parameters.

Work still remains in determining, for real systems, what anticipated life and
operating hours make sense in a general analysis, but such is outside the scope of
this analysis. Guidelines for these parameters may be discussed and agreed upon by
those using the method.

The analysis is limited by the available data, which are in the Appendix. Additional
options for the analysis become available with additional data, but the methodology
of the analysis remains as presented.

References

American Society of Mechanical Engineers, ASME B31.9, Building Services Piping,


2008 Edition, Chapter VI.

American Society of Mechanical Engineers, ASME B31.5, Refrigerant Piping and Heat
Transfer Components, 2013 edition, Chapter VI.

Jekel, T.B., and Reindl, D.T. (2008). “Two-stage compression.” ASHRAE Journal,
August, pp. 46–51.

36 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Lindeburg, M.R. (2001). Mechanical engineering reference manual for the PE exam,
11th Edition, Professional Publications, Inc., Belmont, CA, pp. 17-1–17-40, 18-6–18-7.

Nolte, C.B. (1978). “Least annual cost” In Optimum Pipe Size Selection, 1st Edition,
Trans Tech Publications, Clausthal, Germany,pp. 11–31.

Richards, W.V. (1984). “Practical pipe sizing for refrigerant vapor lines.” IIAR Annual
Conference, San Francisco, CA, pp. 39C–65C.

RSMeans Facilities Construction Cost Data Book (2012), 27th Annual Edition.
Norwell, MA, RSMeans.

Thome, J.R. (2004). Wolverine engineering data book III, Ch. 13. ORGANIZATION,
LOCATION, pp. 13-5–13-6. Wolverine Division of UOP.

White, J.A., et al. (1998). Principles of engineering economic analysis, 4th Edition.
New York, John Wiley & Sons, Inc., pp. 110, 418.

Lemmon, E.W., Huber, M.L., and McLinden, M.O. (2013). NIST standard reference
database 23: Reference fluid thermodynamic and transport properties-REFPROP,
Version 9.11. National Institute of Standards and Technology, Standard Reference
Data Program, Gaithersburg, MD.

Technical Paper #6 © IIAR 2017 37


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Appendix

Compressor Horse Power Values (BHP/TR)


Temperature Single Stage Two-Stage
-60°F - 3.078797
-55°F - 2.835703
-50°F - 2.617779
-45°F - 2.421913
-40°F - 2.245435
-35°F - 2.086041
-30°F - 1.941745
-25°F 2.314706 1.810824
-20°F 2.072987 -
-15°F 1.861748 -
-10°F 1.676574 -
-5°F 1.513767 -
0°F 1.370212 -
+5°F 1.24328 -
+10°F 1.130746 -
+15°F 1.030716 -
+20°F 0.941579 -
+25°F 0.861953 -
+30°F 0.790659 -
Table 11. Compressor power at various saturated suction temperatures

38 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Notes:

1. Common saturated condensing temperature is +85°F.

2. 1.  Two-stage values are calculated at a +20°F intermediate temperature.


Common  saturated  condensing  temperature  is  +85°F.  
1.  
2.   Common   saturated  
Two-­‐stage   condensing  
values  are   calculated  taemperature   is  +85°F.   temperature.  
t  a  +20°F  intermediate  
3. 2.  The curve fit used toalculated  
generate the single-stage power values within the
3.   Two-­‐stage  
The  curve  fvit  alues  
used  atre  
o  gcenerate   at  saingle-­‐stage  
the    +20°F  intermediate   temperature.  
power  values   within  the  temperature  ranges  
3.  temperature
The   curve  
listed   fit  uranges glisted
sed  tao  re  
is  (pressures   enerate   isthe  
(pressures
single-­‐stage  
in  absolute):  
are in absolute):
power   values  within  the  temperature  ranges  
listed  is  (pressures  are  in  absolute):  
𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃. g*.e"L
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.34321 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃. g*.e"L  
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.34321 𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃  
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃
4.   The  curve  fit  used  to  generate  the  two-­‐stage  power  values  within  the  temperature  ranges  listed  
4. 4.  The
The   curve
curve  ffit used
it  aure  
is  (pressures   sed   to
in  to  
generate
generate  
absolute):   the  tthe two-stage
wo-­‐stage   power  powervalues  wvalues
ithin  the   within the temperature
temperature   ranges  listed  
is  (pressures  
ranges listedais re  (pressures
in  absolute):  are in absolute):
𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃. g*.2*"f
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.53567 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃. g*.2*"f  
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 0.53567 𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃  
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎  𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃
5.   These  curve  fits  provide  for  an  R2  value  above  0.996.  
2
5.   These   curve  fits   provide  ufsed  or  ain  
s  R1an  value   asia,  
bove   0.996.  
5. 6.  These
The  discharge  
curve fits pressure  
provide for 66.51  
R2 pvalue which  
above represents  
0.996.a  saturated  condensing  
6.   The   d ischarge   p
temperature  of  +85°F  ressure   u sed   i s   1 66.51   p sia,   w hich   r epresents   a  saturated  condensing  
temperature  
6. 7.  The
The  discharge of  pressure
table  values   +a85°F  
nd  curve  used fits  are  isfor  
166.51 psia,
the  raw   which
shaft   power  represents a saturated
at  the  compressor   shaft  and  do  not  
7.   The   table  
include   m values  
otor   e and  curve  
fficiency   o fpits  
r   are  ffactor.  
ower   or  the  raw  shaft  power  at  the  compressor  shaft  and  do  not  
condensing temperature of +85°F
include  motor  efficiency  or  power  factor.  
7. The table values and curve fits are for the raw shaft power at the compressor
shaft and do not include motor efficiency or power factor.

Technical Paper #6 © IIAR 2017 39


40
Raw Raw Separate
Material Labor Final
Assumed Material Labor Welding Final Labor
Size Schedule Subtraction Subtraction Material
Material Cost, (hours/ Labor (hours/ft)
Factor Factor Cost ($/ft)
($/ft) ft) (hours/ft)
0.5 80 A106 Gr B 9.95 0.131 0.31 0.54 0.0593 6.8655 0.11956
0.75 80 A106 Gr B 11.5 0.14 0.28 0.49 0.0696 8.28 0.141
1 80 A106 Gr B 13.65 0.157 0.28 0.49 0.08 9.828 0.16007
1.25 80 A106 Gr B 18.05 0.188 0.28 0.49 0.0842 12.996 0.18008
1.5 80 A106 Gr B 28 0.208 0.28 0.49 0.0889 20.16 0.19498
2 80 A106 Gr B 33.5 0.262 0.21 0.4 0.1067 26.465 0.2639
2 40 A53 Gr B 12.35 0.262 0.09 0.21 0 11.2385 0.20698
2.5 40 A53 Gr B 21 0.34 0.09 0.21 0 19.11 0.2686
3 40 A53 Gr B 25.5 0.372 0.09 0.21 0 23.205 0.29388
4 40 A53 Gr B 35 0.432 0.05 0.12 0 33.25 0.38016
5 40 A53 Gr B 38 0.5 0.05 0.12 0 36.1 0.44

© IIAR 2017
6 40 A53 Gr B 50.5 0.667 0.05 0.12 0 47.975 0.58696
8 40 A53 Gr B 55 0.828 0.05 0.12 0 52.25 0.72864
10 40 A53 Gr B 64 1 0.05 0.12 0 60.8 0.88
12 40 A53 Gr B 87.5 1.263 0.05 0.12 0 83.125 1.11144
14 40 A53 Gr B 106 1.6 0.03 0.1 0 102.82 1.44
16 40 A53 Gr B 123 1.846 0.03 0.1 0 119.31 1.6614
18 40 A53 Gr B 144 2.182 0.03 0.1 0 139.68 1.9638
20 40 A53 Gr B 169 2.667 0.03 0.1 0 163.93 2.4003
22 40 A53 Gr B 200.8016 2.8335 0.03 0.1 0 194.777552 2.55015
24 40 A53 Gr B 235 3 0.03 0.1 0 227.95 2.7
Table 12. Carbon steel pipe material and labor (RSMeans)

Technical Paper #6
2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Notes:

1. For labor and materials listed for sizes ½ in. through 2 in., schedule 80, the raw
labor and materials represent those of threaded A106 Gr B piping.

2. Material and labor subtraction factors represent

a. For Schedule 80 piping, deletion of the coupling and hanger.

b. For Schedule 40 piping, deletion of the hanger.

This is necessary due to data availability, in that RS Means does not present
data specifically for piping that is assembled with socket-welded couplings as is
common in the refrigeration industry. Note that these factors are a percentage of
the raw numbers presented.

3. Separate welding labor factors are intended to add back in labor for welding
schedule 80 piping together, which is assumed to be butt welding labor. Note
that RS Means presents this as hours/joint, where in Table 12, it is presented as
hours per ft with an assumed joint spacing of 10 ft.

4. Final material and labor costs, as listed in the last two columns of Table 12, are
those used in the analysis.

Technical Paper #6 © IIAR 2017 41


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Raw Raw Final Final


Material Labor
Assumed Material Labor Material Labor
Size Schedule Subtraction Subtraction
Material Cost, (hours/ Cost (hours/
Factor Factor
($/ft) ft) ($/ft) ft)
0.5 10 A304 10 0.128 0.09 0.22 9.1 0.09984
0.75 10 A304 10.2 0.142 0.09 0.22 9.282 0.11076
1 10 A304 10.25 0.16 0.04 0.15 9.84 0.136
1.25 10 A304 11.75 0.176 0.04 0.15 11.28 0.1496
1.5 10 A304 13.15 0.193 0.04 0.15 12.624 0.16405
2 10 A304 16.1 0.239 0.04 0.15 15.456 0.20315
2.5 10 A304 21 0.314 0.03 0.08 20.37 0.28888
3 10 A304 25.5 0.348 0.03 0.08 24.735 0.32016
4 10 A304 31 0.381 0.03 0.08 30.07 0.35052
5 10 A304 40.5 0.457 0.03 0.08 39.285 0.42044
6 10 A304 49.5 0.6 0.03 0.04 48.015 0.576
8 10 A304 86.5 0.727 0.03 0.04 83.905 0.69792
10 10 A304 121 0.96 0.03 0.04 117.37 0.9216
12 10 A304 151 1.143 0.03 0.04 146.47 1.09728
Table 13. Stainless steel pipe material and labor (RSMeans)

Notes:

1. Material and labor subtraction factors represent deletion of the hanger in the
raw data.

42 © IIAR 2017 Technical Paper #6


Thickness of Insulation 1-1/2" 2" 3"
Pipe Size Pipe OD Materials Labor Overall Materials Labor Overall Materials Labor Overall
(rounded to ($/ft) (Hrs/ft) Size ($/ft) (Hrs/ft) Size ($/ft) (Hrs/ft) Size
1/2")
0.5 1 2.95 0.1 4 4.67 0.11 5 8.45 0.139 11

Technical Paper #6
0.75 1 3 0.1 4 4.92 0.11 5 8.5 0.139 11
1 1.5 3.28 0.1 4.5 5.2 0.11 5.5 8.6 0.139 11.5
1.25 1.5 3.53 0.103 4.5 5.55 0.114 5.5 8.7 0.145 11.5
1.5 2 3.79 0.103 5 5.8 0.114 6 8.75 0.145 12
2 2.5 4.18 0.107 5.5 6.1 0.119 6.5 9.05 0.152 12.5
2.5 3 4.56 0.107 6 7.3 0.119 7 10.65 0.152 13
3 3.5 4.76 0.11 6.5 7.35 0.123 7.5 10.75 0.16 13.5
4 4.5 5.5 0.119 7.5 8.5 0.128 8.5 13.7 0.168 14.5
5 5.5 6.2 0.123 8.5 0.7 0.133 9.5 14.85 0.178 15.5
6 6.5 6.35 0.133 9.5 10.6 0.139 10.5 16.7 0.178 16.5
8 8.5 8.4 0.152 11.5 12.6 0.152 12.5 19.9 0.188 18.5
10 11 11.15 0.168 14 15.85 0.168 15 24 0.213 21

© IIAR 2017
12 13 13.2 0.178 16 17.6 0.178 17 26.5 0.229 23
14 14 14.95 0.188 17 19.35 0.188 18 30 0.246 24
16 16 16.7 0.2 19 21.5 0.2 20 33 0.267 26
18 18 18.5 0.213 21 23.5 0.213 22 36 0.291 28
20 20 23 28.5 0.246 24 43 0.32 30
Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

22 22 25 32 0.267 26 46.5 0.356 32


24 24 27 32.5 0.291 28 50 0.4 34
Table 14. Insulation material and labor (RSMeans)

Notes:

1. The labor and material numbers listed in Table 14 are for calcium silicate with no fabric mesh.

43
2. Jacket materials and labor are not included. See Table 15.
2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Jacket ID Materials ($/ Labor Jacket ID Materials ($/ Labor


(in.) ft) (hr/ft) (in.) ft) (hr/ft)
3 0.82 0.089 20 4.99 0.18
3.5 0.94 0.093 20.5 5.12 0.182
4 1.06 0.097 21 5.25 0.184
4.5 1.19 0.102 21.5 5.375 0.185
5 1.31 0.107 22 5.5 0.186
5.5 1.43 0.112 22.5 5.6 0.188
6 1.55 0.119 23 5.7 0.19
6.5 1.67 0.122 23.5 5.825 0.1915
7 1.79 0.125 24 5.95 0.193
7.5 1.92 0.13 24.5 6.08335 0.2038
8 2.04 0.132 25 6.2058 0.2061
8.5 2.17 0.134 25.5 6.32825 0.2084
10 2.54 0.143 26 6.4507 0.2107
10.5 2.66 0.145 26.5 6.57315 0.213
11 2.78 0.15 27 6.6956 0.2153
11.5 2.9 0.152 27.5 6.81805 0.2176
12 3.02 0.155 28 6.9405 0.2199
12.5 3.14 0.16 28.5 7.06295 0.2222
13 3.27 0.162 29 7.1854 0.2245
13.5 3.395 0.1625 29.5 7.30785 0.2268
14 3.52 0.163 30 7.4303 0.2291
14.5 3.64 0.165 30.5 7.55275 0.2314
15 3.76 0.167 31 7.6752 0.2337
15.5 3.885 0.1675 31.5 7.79765 0.236
16 4.01 0.168 32 7.9201 0.2383
16.5 4.13 0.17 32.5 8.04255 0.2406
17 4.25 0.172 33 8.165 0.2429
17.5 4.335 0.173 33.5 8.28745 0.2452
18 4.42 0.174 34 8.4099 0.2475
18.5 4.58 0.176 34.5 8.53235 0.2498
19 4.74 0.178 35 8.6548 0.2521
19.5 4.865 0.179
Table 15. Insulation jacket labor and materials (RSMeans)

44 © IIAR 2017 Technical Paper #6


Method for Determining Best Economic Pipe Size for Ammonia Refrigeration Piping

Notes:

1. For noninteger ID sizes above 12.5 in., data are interpolated from whole sizes.

2. Data are only available for aluminum jacket up to 24 in. jacket ID. Larger IDs are
generated using the following curve fits:

a. Material cost = 0.2449 x ID + 0.0833, ID in in.

b. Labor hours = 0.0046 x ID + 0.0911, ID in in.

Size Range Material Cost ($/ft) Labor Hours


One Coat (Insulated Piping)
1–4 in. 0.12 0.013
5–8 in. 0.31 0.026
10–12 in. 0.31 0.039
14–16 in. 0.42 0.052
Two Coats (Uninsulated Piping)
1–4 in. 0.2 0.021
5–8 in. 0.41 0.042
10–12 in. 0.61 0.062
14–16 in. 0.82 0.082
Table 16. Material and Labor for Painting of Piping (RSMeans)

Notes:

1. Values are for brush-on painting of piping.

2. The assumption of one coat being for insulated piping denotes a prime coat
without a top coat where insulation is used. A second coat of paint is assumed
where piping remains uninsulated.

Technical Paper #6 © IIAR 2017 45


2017 IIAR Natural Refrigeration Conference & Heavy Equipment Expo, San Antonio, TX

Notes:

46 © IIAR 2017 Technical Paper #6

You might also like