United States
Environmental Protection
Agency
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-97/058
July 1997
SEPA Project Summary
Comparison of CFC-114 and
HFC-236ea Performance in
Shipboard Vapor Compression
Systems
D. T. Ray, M. B. Pate, and H. N. Shapiro
In compliance with the Montreal Pro-
tocol and Department of Defense di-
rectives, alternatives to the refrigerant
1,2-dichloro-tetrafluoroethane (CFC-114)
are being investigated by the U.S. Navy
and the U.S. EPA for use in shipboard
chillers. The refrigerant 1,1,1,2,3,3-
hexafluoropropane (HFC-236ea) has
emerged as a candidate for drop-in re-
placement.
A computer model was developed for
comparing these two refrigerants in a
simulated 440-kW centrifugal chiller
system. Equations for modeling each
system component were developed and
solved using the Newton-Raphson
method for multiple equations and un-
knowns. Correlations were developed
for CFC-114 and HFC-236ea boiling and
condensing coefficients taken at the
Iowa State Heat Transfer Test Facility.
The model was tested for a range of
inlet condenser water temperatures and
evaporator loads. The results are pre-
sented and compared with data pro-
vided by the Naval Surface Warfare
Center (NSWC) in Annapolis, MD.
The experimental data provided by
the NSWC sufficiently validate the
model, and the simulation model pre-
dicts that HFC-236ea would perform fa-
vorably as a drop-in substitute for CFC-
114.
Several recommendations are dis-
cussed which may further improve the
performance of HFC-236ea in Navy chill-
ers. Recommendations include adjust-
ing the load of the evaporator to achieve
positive gage pressure, use of a purge
device, use of a variable speed com-
pressor, further testing with azeotropic
mixtures, and use of high performance
tubes in the heat exchangers.
The work represented by the report
was funded by the Department of
Defense's Strategic Environmental Re-
search and Development Program
(SERDP).
This Project Summary was developed
by EPA's National Risk Management
Research Laboratory's Air Pollution
Prevention and Control Division, Re-
search Triangle Park, NC, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
Background
Fully halogenated chlorofluorocarbons
(CFCs) are manufactured chemicals with
properties that make them useful for such
applications as aerosol propellents, foam
blowing agents, solvents, and refrigerants
for automotive, residential, and commer-
cial applications. CFCs became popular
in part because they were chemically
stable, nonflammable, and nontoxic.
Ironically, the chemical stability of CFCs
is the cause for their present perceived
threat to the environment. Scientific evi-
dence suggests that the harmful alterations
of the Earth' s atmosphere occurring from
the use of CFCs are of regional and glo-
bal proportions. As early as 1974, con-
cerns about the potential harmful environ-
mental effects associated with the use of
CFCs were raised when it was suggested
that the chlorine from these compounds
could efficiently destroy stratospheric
ozone. Additionally, there is a growing con-
sensus among scientists that CFCs may
contribute to global warming.
-------
CFC-114 has been in use on Navy ships
since 1969 and has demonstrated excel-
lent reliability. However, design improve-
ments have often lagged behind commer-
cial advancements in compressor tech-
nology, advanced heat transfer surface
technology, and intelligent control system
technology. Trichlorofluoromethane (CFC-
11), used extensively on commercial ships,
was found to be unsatisfactory to the Navy
because of problems unique to surface
craft and submarine applications. For ex-
ample, CFC-11 decomposes at high tem-
peratures causing toxicity problems on
submarines as the air is recycled in high-
temperature air purification equipment. In
contrast, CFC-114 remains stable at high
temperatures.
Other requirements unique to the Navy
include the need for small inventory and
small components due to space con-
straints. Energy efficiency has been a low
priority in the past, but with shrinking de-
fense budgets, it has become more im-
portant. Additionally, surface craft and sub-
marines need to operate silently in tactical
situations, and recycle air in living spaces.
Cooling systems must be able to operate
at as low as 10% of maximum capacity
during normal peacetime operations yet
handle dramatic increases in load when
firing weapons in combat or training situa-
tions. Fully halogenated refrigerants, such
as CFC-114, generally exhibit the best
compatibility and impose the least restric-
tion in choice of materials; a suitable re-
placement must display similar material
compatibility. Other requirements for a suit-
able replacement include meeting safety
and environmental standards for toxicity,
flammability, ozone depletion potential, and
global warming potential.
A need exists for a suitable near-term
replacement for existing equipment using
CFC-114 . Because industry attention has
been focused on CFC-11 and dichlorodif-
luoromethane (CFC-12) replacements, the
Navy must devote substantial resources
to address the CFC-114 problem. Current
potential alternatives for CFC-114 are not
well developed, and significant modifica-
tions to system equipment will likely be
necessary in order to accommodate them.
At one time, 1,1,1-trifluoro-2,2-
dichloroethane (HCFC-123) and 1,1,1,2-
tetrafluoro-2-chloroethane (HCFC-124)
were leading alternatives for CFC-11 and
CFC-114, respectively. When it became
apparent that these HCFCs would also be
phased out as environmentally unsuitable,
the EPA began investigating "backup" al-
ternatives. As a result, a series of propanes
have emerged as candidate replacements
for CFC-114.
HFC-236ea is a promising candidate
for replacing CFC-114 for several reasons.
First, a commercial production route is
available for large quantities through the
use of hexafluoropropylene. Second, ini-
tial modeling conditions appear favorable
as a drop-in substitute, with modeled per-
formance being within 1% of CFC-114
and operating capacities, pressures, and
temperatures matching closely. Flamma-
bility tests, materials compatibility tests,
and oil miscibility tests appear favorable.
Acute inhalation test results indicate lower
acute toxicity than CFC-114. In addition,
estimates predict that HFC-236ea has a
short atmospheric lifetime.
Objective
The objective of this work is to evaluate
HFC-236ea as a potential near-term re-
placement for CFC-114 in shipboard chill-
ers using a computer-simulated system
with data from the Iowa State University
Heat Transfer Test Facility and also using
data provided by the NSWC. With the
information gathered, appropriate design
recommendations to accommodate the use
of HFC-236ea in shipboard chillers are
offered.
Scope
A computer program has been devel-
oped that simulates the performance of a
440-kW capacity, single-stage, centrifugal,
chilled-water air-conditioning plant. The
design conditions shown in Table 1 are
based on the design of a typical air-condi-
tioning plant in use on Navy surface craft
and submarines.
Given the entering and leaving tem-
peratures of the chilled water, the enter-
ing temperature of the condenser water,
and the flowrates of the chilled water and
condenser water, the model predicts the
required compressor power and the re-
frigerant saturation temperatures in the
heat exchangers. With knowledge of fluid
properties and tube geometries, the per-
formance of the system with different re-
frigerants and enhanced surface tubes can
be compared under similar operating con-
ditions. For this study, the model is used
to compare refrigerants CFC-114 and
HFC-236ea using 10.23 fins per centime-
ter (fpc) tubes in the condenser and evapo-
rator. The results are presented with de-
sign recommendations.
Model and Experimental
Results Comparison
The NSWC, in cooperation with the U.S.
EPA, has tested CFC-114 and HFC-236ea
in a 440-kW laboratory centrifugal chiller
representative of those used in the U.S.
Navy's fleet of surface craft and subma-
rines. The laboratory chiller is fully instru-
mented.
Modeled and measured performance
values were compared for both refriger-
ants. The measured compressor power
provided by the NSWC was calculated
from measurements of torque and angu-
lar velocity of the compressor shaft. The
modeled value of compressor power is
the rate of work performed directly on the
fluid and does not include the mechanical
or heat losses as the power is transferred
from the compressor shaft to the impeller
and ultimately to the working fluid. A lin-
ear relationship was found to exist be-
tween the shaft power and the power trans-
ferred to the working fluid for the data
provided by the NSWC. The correlation
was applied to the results of the model as
an assumed mechanical efficiency.
Even with an efficiency factor applied,
the model underpredicts the amount of
compressor power required to meet the
specified load by 6 to 26%. This could be
related to the use of inlet guide vanes to
the compressor which are not modeled in
this study.
Table 1. Simulated Design Conditions
Component Design Condition
Value
Evaporator
Evaporator
Evaporator
Condenser
Condenser
Chilled Water Flowrate
Entering Chilled Water Temperature
Leaving Chilled Water Temperature
Water Flowrate
Entering Water Temperature
28.4 L/s
10.7°C
7.0°C
31 .5 L/s
31 .4°C
-------
A comparison of the system coefficient
of performance calculated using NSWC's
measurements and that predicted using
the model developed in this study shows
that the model overpredicts the coefficient
of performance for HFC-236ea by as much
as 38%. The result is consistent with what
one would expect when comparing mod-
eled with measured results. Since models
often make use of simplifying assump-
tions, the results tend to be idealized.
One would expect to see the actual re-
sults to be less favorable than modeled
results.
A comparison of modeled and mea-
sured refrigerant temperatures in the con-
denser shows that the model predictions
compare well with CFC-114 data; how-
ever, it underpredicts the condenser tem-
perature for most HFC-236ea data by as
much as 9°C. This difference could be
caused, in part, by poor heat transfer in
the condenser with HFC-236ea. If this were
the case, it would take a higher condenser
temperature to overcome whatever ther-
mal resistance is present. More compres-
sor power would be required to provide
this additional temperature lift, resulting in
lower system performance. Because the
refrigerant temperature in the condenser
is closely tied to the condensing coeffi-
cient by use of the log mean temperature
difference equation for heat transfer in the
condenser, one would also expect to see
a difference in a comparison of the mea-
sured and modeled condensing coeffi-
cients. As the condensing temperature in-
creases while entering and leaving, cool-
ing water temperatures remain constant
and the log mean temperature difference
increases. This would result in a modeled
decrease in the condensing coefficient.
An example of when conditions may
exist in the condenser that hamper heat
transfer is when noncondensable gases,
left unpurged, accumulate in the upper
vapor space of the condenser. This is a
plausible explanation for the differences
in condenser saturation temperatures ob-
served. For HFC-236ea, both the mea-
sured and modeled evaporator tempera-
tures are near 2°C. The corresponding
saturation pressures for these saturation
temperatures are less than atmospheric
pressure. This could cause nonconden-
sable gases to leak into the evaporator
due to negative gage pressure. These
gases would migrate and collect in the
condenser and could significantly degrade
the performance of the condenser and the
entire system. If air, in fact, was present
in the condenser, it would drive the out-
side heat transfer coefficient down, result-
ing in a high condenser saturation tem-
perature.
A comparison of modeled and mea-
sured cooling water temperatures leaving
the condenser shows that modeled val-
ues are within 0.2°C of measured values.
This is consistent with trends observed for
condenser capacity. This is expected,
since the rate of heat transfer and the
temperature of the water leaving the con-
denser are the two variables in the water-
side heat transfer equation for the con-
denser.
A comparison of modeled and mea-
sured evaporator temperature shows that
modeled and measured boiling coefficients
compare well despite some variance. One
would therefore expect to see a variance
in a comparison of boiling coefficients since
these variables must balance in the log
mean temperature difference equation for
the heat transfer in the evaporator.
A comparison of modeled and mea-
sured evaporator capacity shows that the
values match because the capacity is cal-
culated based on three common inputs:
entering and leaving evaporator water tem-
peratures, and evaporator water flowrate.
Thus, the evaporator capacity remains
fixed as long as these three input values
are held constant.
A comparison of modeled and mea-
sured refrigerant flowrate shows that the
model consistently predicts the flowrate
for both refrigerants within +5%. This sug-
gests that the enthalpy differences also
compare favorably since the rate of heat
transfer in the evaporator is constant and
is equal to the refrigerant mass flowrate
times the enthalpy difference across the
evaporator.
Comparison of CFC-114 and
HFC-236ea Performance
The model is used to predict the perfor-
mance of both refrigerants, CFC-114 and
HFC-236ea, through a range of operating
conditions. This is done by using the fleet
design point as the default and varying
one parameter at a time over an appropri-
ate range to see the effects on the sys-
tem. The results yield additional insight as
to the possible suitability of HFC-236ea
as a drop-in substitute for CFC-114.
Entering Condenser Water
Temperature
As the Navy operates its fleet around
the world, ships encounter a wide range
of condenser water temperatures because
sea water is used directly in the heat
exchanger to remove heat from the work-
ing fluid. Chillers for Navy ships are de-
signed for a condenser water temperature
of 31.4°C; however, temperatures encoun-
tered may range from -1.3 to 35.3°C de-
pending on where the ship is operating.
Since heat transfer in the condenser is
driven by the temperature difference be-
tween heat transfer fluids, a condenser
water temperature that is too high could
hamper the performance of the condenser
and subsequently the entire refrigeration
cycle. Thus, the entering condenser water
temperature is significant to the perfor-
mance of the overall system.
The predicted power required to drive
the compressor more than doubles for
both refrigerants as the water tempera-
ture entering the condenser increases from
16 to 38°C. The trend is expected since
better heat transfer occurs as the tem-
perature of the cooling water entering the
condenser decreases. The efficiency of
the refrigeration cycle should thereby im-
prove, resulting in less power input re-
quired to the compressor. Additionally, the
model predicts that HFC-236ea used as a
drop-in substitute for CFC-114 may result
in energy savings. At the design point of
operation, the predicted power required to
drive the compressor using HFC-236ea is
91.4% of the power required using CFC-
114. The model predicts that, for any cool-
ing water temperature, the power required
for a refrigeration cycle using HFC-236ea
as a drop-in will be significantly less than
the same cycle using CFC-114 as the
working fluid. The predicted savings in
power consumption by using HFC-236ea
at the design point of operation is 550
kW.
The NSWC data for CFC-114 show
nearly constant compressor power over
the range of entering condenser water
temperatures. The data for HFC-236ea
show significant scatter. Both the CFC-
114 and HFC-236ea measured values of
required compressor power are above the
predicted values for the range of entering
condenser water temperatures.
The coefficient of performance is pre-
dicted to decrease as the inlet condenser
water temperature increases. Additionally,
at the design point of 31.4°C, the pre-
dicted coefficient of performance for HFC-
236ea is 4.25 compared to 3.91 for CFC-
114. The model predicts better perfor-
mance using HFC-236ea over the range
of condenser water temperatures simu-
lated.
The measured values of coefficient of
performance for both CFC-114 and HFC-
236ea are less than predicted values. As
the entering condenser water temperature
increases, the measured and predicted
values of the coefficient of performance
move toward better agreement.
-------
Modeled trends predict that both perfor-
mance indicators—compressor power re-
quirement and coefficient of performance—
improve as the inlet condenser water tem-
perature decreases from 38 to 16°C since
the required power consumption decreases
and the coefficient of performance in-
creases. These are expected trends since
lower condenser water temperatures pro-
vide a higher temperature difference be-
tween the heat transfer fluids, resulting in
increased cooling potential in the con-
denser. HFC-236ea is also predicted to
outperform CFC-114 over a range of inlet
condenser water temperatures, partly be-
cause measured heat transfer coefficients
for HFC-236ea were found to be greater
than those of CFC-114.
The predicted values for CFC-114 and
HFC-236ea saturation temperature in the
condenser for a given temperature of the
cooling water entering the condenser are
nearly identical. The measured values of
condenser saturation temperature for CFC-
114 agree with the predicted values, while
the HFC-236ea data show the same trend
but are generally higher than predicted.
The predicted saturation temperature for
HFC-236ea in the evaporator as a func-
tion of the entering condenser water tem-
perature is higher than predicted for CFC-
114 over the range of entering condenser
water temperatures. The HFC-236ea data
compare well with predicted values, while
there appears to be less agreement be-
tween measured and modeled values for
CFC-114.
The evaporator capacity is compared
with the temperature of the water entering
the condenser. Since the evaporator ca-
pacity is fixed by holding the water-side
conditions constant, the predicted capac-
ity values for HFC-236ea and CFC-114
are identical. Scatter is shown in the mea-
sured values of HFC-236ea, while the
measured values of CFC-114 agree well
with predicted values.
The condenser capacity is also com-
pared with the entering condenser water
temperature. Predicted values for CFC-
114 and HFC-236ea are nearly equal. For
both CFC-114 and HFC-236ea, measured
capacity values are higher than predicted,
with significant scatter observed in the
HFC-236ea data.
The refrigerant mass flowrate is com-
pared with the temperature of the water
entering the condenser. The measured
and predicted values for CFC-114 are in
close agreement, while there is significant
scatter in the data for the flowrate of HFC-
236ea. As condensing water temperature
increases, an increasing trend is predicted
in refrigerant mass flowrate.
Entering Evaporator Water
Temperature
In this situation, the evaporator load is
defined by a constant chilled water flowrate
of 28.4 L/s, a chilled water inlet tempera-
ture of 7°C, and a chilled water exit tem-
perature ranging from 9.2 to 12.6°C. Addi-
tionally, the temperature of the water en-
tering the condenser is held constant at
31.4°C, and the flowrate of the condenser
water is held constant at 31.5 L/s. As the
cooling load is systematically varied, the
effect on various performance indicators
may be observed.
As the temperature of the chilled water
returning from the load and entering the
evaporator increases while other design
operating conditions remain constant, in-
creasing power is required to drive the
compressor. This is an expected trend
because, as the water temperature enter-
ing the evaporator increases, the load is
increased in the evaporator. In order to
accommodate this increased load, either
the refrigerant mass flowrate must increase
or the enthalpy difference across the
evaporator must increase in order to pro-
vide enough heat transfer to maintain a
constant chilled water exit temperature.
The result is the need for more power to
drive the compressor. A comparison of
HFC-236ea and CFC-114 shows that, for
the compared range of chilled water tem-
peratures entering the evaporator, HFC-
236ea always required less compressor
power than CFC-114 when modeled as
the working fluid.
A comparison of the coefficient of per-
formance as a function of chilled water
temperature entering the evaporator shows
that as the temperature increases the co-
efficient of performance decreases. This
means that as temperature increases, the
rate of increase of power required by the
compressor is greater than the increase
in cooling capacity. Additionally, the coef-
ficient of performance for HFC-236ea is
higher than that for CFC-114 for the range
of temperatures modeled.
Design Recommendations
One way to improve the performance of
the fleet's 440-kW chiller and allow the
use of HFC-236ea as an alternative work-
ing fluid is to reduce the load on the
evaporator by increasing the temperatures
of the chilled water entering and leaving
the evaporator by a few degrees. This
would allow the refrigerant temperature in
the evaporator to rise slightly, which, in
turn, would result in an evaporator pres-
sure that is above atmospheric pressure.
With a positive gage pressure in the evapo-
rator, there is less possibility of
noncondensable gases and contaminants
leaking into the system where they can
accumulate in the condenser and reduce
performance. The low evaporator tempera-
tures and high condenser temperatures
reported by the NSWC suggest the possi-
bility of this occurrence. This solution
avoids the cost of redesigning system com-
ponents.
A purge device at the high point of the
condenser would allow purging of
noncondensable gases that might accu-
mulate there. If noncondensables are a
persistent problem, the purge unit may be
malfunctioning or the system may have
an air leak larger than the purge unit can
handle.
A variable-speed compressor would
eliminate the need for hot gas by-pass or
the extensive use of inlet guide vanes in
the compressor to control the refrigerant
flow. A variable-speed chiller would allow
maximum system performance to be real-
ized over a broad range of operating con-
ditions, resulting in maximum energy sav-
ings.
Another possible improvement might be
to mix HFC-236ea with other non-CFC
refrigerants to form an azeotropic mixture
with properties that allow the saturation
point in the evaporator to stay above at-
mospheric pressure. The mixture could
be chosen so as to maintain the desired
properties of HFC-236ea.
Additionally, better performance in the
Navy's fleet air-conditioning units could
be realized by investing in commercially
available high performance heat exchanger
tubes. While not reported in this study,
Turbo B tubes were simulated with CFC-
114 and HFC-236ea under fleet design
conditions and were predicted to perform
significantly better than 10.23 fpc tubes in
both the evaporator and the condenser.
Finally, the model predicts that HFC-
236ea used as a drop-in substitute for
CFC-114 without any design modifications
may result in energy savings. The model
predicts that, for any set of conditions, the
power required for a refrigeration cycle
using HFC-236ea as a drop-in will be sig-
nificantly less than the same cycle using
CFC-114 as the working fluid. The pre-
dicted savings in power consumption by
using HFC-236ea at the design point of
operation is 8.6%. If HFC-236ea is to be
used only as a near-term replacement, it
may be appropriate to use it without mak-
ing any significant design changes to the
system.
-------
Conclusion
The Montreal Protocol began a world-
wide drive to eliminate the production of
CFCs and other chemicals which are
thought to be harmful to the environment.
As a result of the restrictive legislation
that followed, there is an immediate need
to replace CFC-114 which is used exten-
sively in U.S. Navy surface craft and sub-
marines. Preliminary research conducted
by the EPA suggested that HFC-236ea
might perform suitably as a near-term drop-
in replacement for CFC-114. However, at
the time of this study, heat transfer data
for HFC-236ea were not available.
For this reason, a computer model was
developed for comparing CFC-114 and
HFC-236ea in a simulated 440-kW labo-
ratory centrifugal chiller system represen-
tative of those found in the U.S. fleet. The
model is semi-empirical, combining ther-
modynamic and heat transfer theory, as
well as boiling and condensing heat trans-
fer coefficient data measured at the Iowa
State University Heat Transfer Test Facil-
ity.
The NSWC in Annapolis, MD, also pro-
vided data for this study. A 440-kW labo-
ratory centrifugal air-conditioning plant and
HFC-236ea were used for the data collec-
tion. The experimental data provided by
the NSWC were compared with the mod-
eled results.
The model was tested for a range of
inlet condenser water temperatures, en-
tering and leaving chilled water tempera-
tures, and evaporator and condenser wa-
ter flowrates. The simulation model pre-
dicts that HFC-236ea would perform fa-
vorably as a drop-in substitute for CFC-
114 .
Additionally, several recommendations
were provided for improved performance
using HFC-236ea in centrifugal chiller sys-
tems. Design recommendations discussed
in this study include manipulating the
evaporator load to achieve positive gage
pressure in the evaporator, ensuring the
absence of noncondensable gases in the
system, using a variable-speed compres-
sor with a fixed inlet guide vane angle to
the impeller, conducting further research
using azeotropic mixtures with HFC-236ea
as the major component, and installing
high-performance enhanced surface tubes
in both the evaporator and the condenser.
In summary, the simulation developed
in this study provides results that are con-
sistent with the expected behavior of a
440-kW refrigeration system. The results
provided by the NSWC sufficiently vali-
date the model. Finally, the results sug-
gest that HFC-236ea would perform well
in existing CFC-114 centrifugal chillers,
although design modifications should be
considered for optimal performance.
-------
D. T. Ray, M. B. Pate, and H. N. Shapiro are with Iowa State University, Ames, IA
50011.
Theodore G. Brna is the EPA Project Officer (see below).
The complete report, entitled "Comparison of CFC-114 and HFC-236ea Perfor-
mance in Shipboard Vapor Compression Systems," (Order No. PB97-178735;
Cost: $25.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use $300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
EPA/600/SR-97/058
------- |