Purdue University

Purdue e-Pubs
International Refrigeration and Air Conditioning
School of Mechanical Engineering
Conference

1996

A Performance Comparison of Secondary

Refrigerants
S. W. Inlow
Purdue University

E. A. Groll
Purdue University

Follow this and additional works at: http://docs.lib.purdue.edu/iracc

Inlow, S. W. and Groll, E. A., "A Performance Comparison of Secondary Refrigerants" (1996). International Refrigeration and Air
Conditioning Conference. Paper 349.
http://docs.lib.purdue.edu/iracc/349

This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact epubs@purdue.edu for
additional information.
Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/
Herrick/Events/orderlit.html
 A PERF ORM ANCE COM PARI SON OF SECO NDAR Y REFR IGER
ANTS

Scott V/. Inlow and Eckhard A. Groll

Ray W. Herrick Laboratories, Purdue Univers ity
West Lafayett e, IN 47907-1077, USA.

ABSTR ACT
One potentia l alternat ive refrigera tion system for superma rket applicat ions
is the seconda ry-loop
system. This system uses a conventional direct-expansion refrigeration
system to cool a thermof luid
that is pumped through out the superma rket to provide the necessary cooling.
In order for this system to
be a ,,.iable alternat ive. the loss in perform ance resultin g from the added level
of heat exchang e and the
addition of a pump must be minimized through proper secondary refrigerant
selection and system design.
Two gener2.lized compute r models of a conventional refrigeration system
and a secondary-loop system
have been developed to aid in this design process. These models were used
to investigate the perform ance
of secondary-loop and conventional refrigeration systems. Possible seconda
ry refrigerants are identified.
A detailed parame tric study of the refrigeration coil and intercon necting
pipe diamete rs was conduct ed
to achieYe thebes<: system perform ance for each fluid and for HCFC-22 in
the conventional system. The
compari son of these secondary fluids relative to HCFC-22 in a conYentional
system is presente d. The
results indicate that through proper secondary refrigerant selection and system
design, the secondary-
loop refrigeration system could provide equivalent perform ance to existing
systems.

1 INTRO DUCT ION

Increasing concern in recent years over the environ mental effects of chlorofl
uorocarbon ( CFC) and hydroch lo-
rofiuoro carbon (HCFC) refrigerants has led to internat ional agreements concern
ing their phase out. This phase out
calls for a stop to product ion of all CFC's by the year 1996, and an end
to product ion of all HCFC's by the year
2030. The phase out of these refrigerants is of particul ar concern to superma
rkets. Superm arkets are estimat ed
to account for 4% of the national electric power and 30 to 50% of that
power is used for refriger ation purpose s
[McDowell, Klein & Mitchell 1995]. Current superma rket refrigeration
systems are direct expansi on systems in
which each display case contains its own evapora ting unit. The direct expansio
n refrigerant is circulat ed through -
out the superma rket to each of the display cases providin g the necessary
cooling. A typical system will use on the
order of 900 to 1130 kg (2000 to 2500 lb;.,) of refrigera nt charge. It is
estimat ed that up to 30% of this charge
can be lost each year [Harrison, Keeney & Nelson 1995]. This can represen
t a significant cost to the superma rket.
This expense could increase due to the added cost of the hydrofiuorocarbo
n (HFC) mixture s that are expecte d to
replace current refrigerants in most applicat ions. In addition to the cost concerns
, there is the added environ mental
concerns related to the impact of HFC's on global warming .
One potentia l alternat ive to using HFC mixture s, is the secondary-loop
refrigeration system. This system
uses a direct-e xpansio n refrigeration system (primary -loop) to cool a thermof
luid. The thermof luid is then pumped
through out the superma rket to provide the necessary cooling. Leakage rates
of the primary refriger ant are reduced
by containi ng the direct-expansion system in a compac t sealed unit, thereby
elimina ting the long lengths of inter-
connecting piping through which the primary refrigera nt must flow. It is
anticipa ted that such a system may have
a lower overall coefficient of perform ance (COP) due to the added level of
heat exchange and the addition al power
costs associat ed with the pump. In order for secondary-loop systems to
provide a viable alternat ive, the loss in
perform ance must be minimiz ed through proper seconda ry refrigerant selection
and system design.
Two generalized compute r models have been developed to simulat e a convent
ional system and a seconda ry-
loop system using either volatile or single phase seconda ry .refrigerants.
These models were used to compar e the
perform ance of heat transfer fluids in the secondary-loop system to a convent
ional system using HCFC-2 2 at various
capacities. In order to obtain a fair compari son, the size of the refrigera
tion coil was kept constan t. A detailed
parame tric study of the refrigeration coil geometry, evapora tion tempera ture,
and the intercon necting pipe diamete r
was conduct ed to determi ne the best system perform ance for each fluid
at a given capacity. The results of this
study are presented here.

357
 2 REFRIG ERANT SELECT ION

perfor-
Proper selection of the primary and secondary refrigerants is important in order to achieve the best
be contained in a compact sealed unit, more options are available when
mance_ Because the primary cycle can
includes, in addition to the HFC's, the use of ammonia or hvdrocarbo ns_
choosing the primary refrigerant. This
investigati on as the choice for the primary refrigerant . Stoecker [1989J cites several
Ammonia was considered in this
re ranges and
advant.ages to using ammonia over HCFC-22, including, a higher cycle efficiency over most temperatu
amic and
higher heat transfer coefficients_ These advantage s can be attributed to ammonia"s superior thermodyn
direct expansion
transport properties . Comparin g ammonia and HCFC-22 at a given capacity in a conventional
Derformance is
system, with the same geometry and operating under identical conditions. a 7~12% increase in
-
achieved with ammonia depending upon the capacity chosen.
and non-
Uniike the primary refrigerant. it is necessary for the secondary heat transfer fluid to be non-toxic
to minimize pumping costs. the fluid should also
flammable because it is circulated throughou t the store. In order
Because of these restrictions , the available options are
have excellent thermodyn amic and transport. properties.
single-pha se heat transfer fluids which could be used in
limited. There are a variety of commercially developed,
secondary-loop systems.
temperatu re
• Propylene Glycol.- Propylene glycol/wat er solutions are a common hea~: transfer fluid for medium
at the expense of
application s, T > -20°C( -4° F)- The freezing point of this solution can be lowered.
volume solution can
performance, by increasing the percentage of propylene glycol in the solution. A 50% by
propylene glycol as
provide freeze protection to about -2-3°C. The Food and Drug Administr ation recognizes
a safe food additive. It is non-flamm able and non-corrosive to mosr materials.
glycol
• Ethylene Glycol: Ethylene glycol/wa ter solutions provide superior transport properties to propylene
As a result, special precaution s
solutions and provide a slightl:y·lower freezing point. However, it is orally toxic.
can be obtained to -50°C (-58°F).
would be required for use in a supermark et systems. Freeze protection
transfer
However, like propylene glycol. lowering the freezing point comes at the expense of decreased heat
performan ce and increased pumping power.
. The
• Hydrofi1wroether: A ne>v fluid, referred to as a hydrofluoroether (HFE), has recently been introduced
making it suited for both low and medium temperatu re
freezing point of this fluid is listed to be below -100°C,
the fluid to be orally non-toxic, non-flamm able, and it has been
applications_ Preliminar y tests have shown
found to be compatibl e 'vith most common materials.
these fluids
• Synthetic Organic Fluzd: Several synthetic organic heat transfer fluids are also available. One of
to -73 .3°C (-1 00°F), thereby providing another option for low temperatu re
was found to offer freeze protection
superior low temperatu re viscosity when compared to ethylene glycol.
systems. This synthetic fluid provides
Like ethylene glycol it is orally toxic and would therefore require special precautions.
Figure 1.
The viscosity and thermal conductivi ty of each of these fluids versus the temperatu re is shown in
For a given pipe diameter and
A low viscosity fluid reduces the pumping requireme nts in the secondary-loop.
Reynolds number leading to higher heat transfer
fluid velocity, a fluid with a. lower viscosity will also have a higher
seen, the HFE
coefficients. Higher Yalues ofthe thermal conductivi ty also lead to improved heat transfer. As can be
while the glycols are seen to have higher thermal conductivi ties.
and synthetic fluid have superior viscosity values,
factor in the performan ce of the glycols, whereas, heat transfer
This suggests pumping power may be the critical
the fluids versus
may be the critical factor for the HFE and synthetic fluid. Figure 2(a) shows the Prandtl number of
Prandtl numbers can be expected to have higher heat transfer coefficient s. The
temperatu re. Fluids with higher
The transition
Nusselt number is shown versus the Reynolds number in Figure 2(b) for a constant temperatu re.
flow is seen to occur at a Reynolds number of 2300. The heat transfer coefficient
between laminar and turbulent
with a correlation by Jakob [ASHRAE 1993]. The turbulent heat transfer
in the laminar regime was calculated
Dittus-Bo elter correlation [lncropera & DeWitt 1990]. Because the Prandtl number
coefficient was calculated by the
considerab ly higher, it is possible to have laminar flow and still achieve sufficient heat transfer in
of the glycols is
the refrigeration coil.
secondary
In addition to the above mentioned single phase heat transfer fluids, carbon dioxide as a volatile
s. Carbon dioxide has recently received
refrigerant was investigate d at low and medium temperatu re application
1994, Kauffeld 1995, Pearson & Fellow
increased attention as a secondary fluid [Hesse 1995, Enkemann & Arnemann

358
 [()" ft---,.........-----,--------,,----;::::::::==:
:::==~l 0.4,,---.--- ---.--- ---,--- --,---. .-,---- ---;,
l - 50°/" E'thylene Gly~::ol

- -- ~~~ Pmpylene Glyool

--i
t Ij (} 35 rl--~--------:::-=-:-::-::-:-=--:-:~::-=-:-=-=--=--=-=-=--=-=-==-=-=---::-=-:_-~-,
10·1 ~ - ... .....,...., _ ~ _..... Synthetrc.OrganJc ~
' --- 'l ~ 0.3
J :i'
~0.:!5~
,
1
- f - 50% Ethylene Glycol
- - SO"!.. Propyl eM Glycol
~ 11.2 ~- HF=E
Synti'!O!IG Organ1e

= 11.15

0./l
J O - ' ' - - - - - - - - - J -_ _ __...._ _ _ _ _ _ _ __,
____ ~
1--
I
-:on -:s -:w -!5 -1o 11.115 ----.... ,.-,,--- ------'- -----'-- ---'---- _j
-5 -30 -25 -~() -15
Tcmpt!rat~Jrc IC I -[() -5
Tc:mpcrJrure ICi

(a) (b)

Figure 1: Transpon. propert y comparison of single phase heat transfer fluids.

(a) Yiscosity (b) thermal conduct ivity

NIX/,
3/IOr;::::=:::::!:===::c::=:::::::::=:::::=;----,----.---,---r-i
-5!l'-•Ethyl.:u cClycol ~ 50o/, Ethylene Giy~ol, Pr .. 115.4
71Xl ' - - Tcmp=-IOC(!4 F')
~[)';; Pmpyh!!Jc Gly1.ol
'' ~ - HF'E 250
- - 50~ Propylene Glycol. Pr;;;; 348.6
'' - • HFE. Pr .. 15.1
6(XJ Synthetic Orgimn:;, Pr= 18.1

51Xl 200

""
~ISO
31Xl z
/00

21Xli-----

·....:.·-
IIXJ ~-------------J
() (,.__::._.__-_._- _c-c___-_·-___.-_·-----'-'------ -...
----·_-_·-_,_·_-~--·-.._·_- _.-_
.._-_,__--_·----·.._-__:.____J
-20 -IN -11\ -14 -12 -10 -N -6 -4 -2 o'-S<.._X),--/--:0'-{)()-1_,5<-XJ-2:c!X~I-0-2-5.._0_0----:3-'oot-l-35-'I-J0--41-"){)(,-l_4_5:'-0-0-5:ctX"'X-J_5__5.,(lO:---'
Temper.•nure (C)
Reynolds Number

(a) (b)

Figure 2: (a) Prandtl number compari son (b) Nusselt number versus Reynold
s number at constan t tempera ture

359
 because of its
1992/93, Inlow & Groll1996 ]. It was a popular refrigerant in the early part of this century, primarily
& Arnemann
safety. Hs use diminished however with the advent of the CFC refrigerants in the 1930's. Enkemann
heat capacity,
[1994] site several advantage s to using a volatile secondary refrigerant , including, higher volumetric
as compared to using HFCs
constant temperatu re evaporatio n, and lmver viscosity. The advantage s to using C0 2
enthalpy of evaporatio n, it is
in a conventional system include an extremely low global warming potentia.!, a high
the use of C0 is its relatively low
non-flamm able and non-toxic, and its low cost and availability. One drawback to 2
psi). The high saturation pressure of
critical temperatu re (31.1°C, 88°F) and high critical pressure (73.7 bar, 108:3
r approach room temperatu re.
C02 presents a problem during cycle defrost when the temperatu res in the evaporato
Solutions to this problem are given by Enkemann & Arnemann [1994] and Inlow & Groll [1996].

3 SYSTEM MODELS

of the
The secondary-loop and conventional system models predict the capacity and outlet air conditions
only) power based on a user defined
refrigerati on coil, a.s well as, the compressor, fan, and pump (sec_ondary-loop
are, the refrig-
system aud operating conditions. There are three main componen ts to each of the models. These
tional system, and the analysis of the
eration coil analysis. a thermodyn amic analysis of the primary-lo op/conven
the pump in the secondary -loop system
pressure losses and heat transfer in the interconnecting piping (including
here. This
model). A multistage compression cycle with flash tank is simulated by t.he computer models described
temperatu res which are required in supermar-
cycle is necessary in order to allow the systems w produce the low
the outlet refrigerant temperatu res from
ket display cases. The economize r cycle vYas chosen in order to reduce
analysis is based on the procedure outlined in
the second compressor as much as possible. The refrigeration coil
A det.ailed description of the models can be found in
Chapter 6 of ASHRAE Equipmen t volume [ASHRAE 1988].
models are: a steady-sta te system, outlet states of the
Inlow & Groll [1996j. The major assumptions used in the
the intermedia te
condenser and flash tank are saturated. adiabatic expansion devices, negligible pressure drops in
using an isentropic efficiency, condensat ion occurs
heat e..xchanger and condenser. compressor power is calculated
over the entire refrigerati on coil, and the effects of frost buildup are neglected.

4 PERFOR MANCE COMPA RISON

loop
The computer models described were used to conduct an investigati on of the performance of the secondary-
re application s, a medium temperatu re with an air inlet temperatu re
and conventional systems at two temperatu
temperatu re with an air inlet temperatu re to the coil of -20.0°C (-4.0°F).
to the coil of -5.-56°C (42°F), and a low
At the low
A 50% propylene glycol solution, the HFE, and C0 2 were investigate d at the medium temperatu re.
was also
temperatu re, the HFE, synthetic fluid, and C02 were used. A conventional system using HCFC-22
synthetic fluid
modeled at both temperatu res. Ethylene glycol was not considered because of its toxicity. The
low temperatu re
>vas considered, despite also being toxic, in order to provide a third secondary refrigerant at the
application . The following general system parameter s were kept constant in the analysis:
• Coil air face velocity was 1.5 m/s (300 ft/min) with an inlet relative humidity of 50%
• The refrigerati on coil used copper tubes in a staggered layout with aluminum fins

• Fin spacing was 78 fins/meter (2 fins/inch)

exchanger
• 3°C (5.4°F) temperatu re difference between the outlet refrigerant streams at the intermedia te heat
(secondary-loop simulation only)

• Condensing temperatu re of 40.56°C (105°F)

• Compressor and pump efficiencies were 70%, fan efficiency of 40%

• 60 meters (196 ft) of interconnecting piping with 2.54 em (1.0 in) of insulation
geometry was
The fluids were compared at the same capacity and using the same size refrigerati on coil. The coil
fixed to be 26.4 em (10.4 in) long, 15.2 em (6.0 in) high, and 86 em (33.9 in) wide.
the refrig-
In order to obtain the best performance with each fluid a detailed parametric study was done on
eration coil and interconne cting piping to determine the optimum value of the coil tube diameter, Do; the number

360
 Table 1: Secondary-loop geometri c paramete rs
Fluid D0 [em] Nr NL D,z [em] Dzz [em]
50% Propylene Glycol 1.50 (5/8 in) 5 10 1.90-5 (3/4 in) 1.905
Synthetic Fluid 1.59 5 8 1.00-5 1.905
HFE 1.59 .5 8 1905 1.905
co2 1.27 (1/2 in) 6 8 1.27 0 ..535 (1/4 in)
HCFC-22 1.59 5 8 1.59 0.535

l.X,----------,---------.---------,

l.6f-------

,.~; 1

0. -----------------------
- i
0v ,_
·- ----- J
~--_,

I
![-(:m
I
-Cfl~

I Ht'JC-'2:2
HCn.. -:::

- • Hll;
'"
- ·I-III'

1.5 nr,l;==
1.5
====: ------- --::'.,- -----_j ~.5
C:lfliLl:llylkWJ

(a) (b)

Figure :3: COP versus refrigeration coii capacity for (a) medium temperat ure system,
(b) low temperat ure system

of tubes high, Nt, and deep, _T\[L; inlet refrigeration temperat ure; suction line (intercon
necting piping leading from
the coil to the intermed iate heat exchanger) diameter , D,z; and the liquid line (intercon
necting piping leading from
the intermed iate heat exchanger to the coil) diameter , Dzz. With the coil volume fixed
the only way to increase or
decrease !Vt and/or NL is to increase or decrease the transvers e and/or longitudi nal tube
spacing. The value chosen
for each of these paramete rs represents a trade-off between increased or decreased heat
transfer and increased or
decreased pressure drop. Table 1 shows the values of these paramete rs which were
found to provide the highest
performa nce. The optimum inlet refrigerant temperat ure depends upon the coil capacity.
The values of these
paramete rs remain relatively constant for each fluid with the exception of C0 . Because
3 the vapor density of C0 2
is relatively high, the pressure drop is smaller. Therefore smaller tube diameter s can
be used.
Figure 3 shows the COP ofthe secondary-loop and conventional systems versus capacity_
The COP is defined
as,
COP= Qair
WP+Wc +WJ (1)
where Qair is the refrigeration coil capacity, and Wp, We, and W1 represent the pump
(secondary-loop model only),
compressor, and fan power respectively. At the medium temperat ure, C02 and HCFC-22
are within 4% of each
other over the entire range. The propylene glycol and HFE provide a similar COP.
These two single-phase fluids
are within approxim ately 20% of the HCFC-22 and C02- The exceptional performa
nce of the secondary-loop
system using C0 2 can be attribute d to three factors. One, its high heat transfer coefficien
ts in the refrigera tion
coil provide a greater capacity at a higher evaporat ion temperat ure. The second
and third factors are its low
viscosity and relatively high vapor density. These two factors reduce pressure losses
througho ut the system. The
lower performance of the single phase fluids can be contribu ted to the lower heat transfer
coefficients and increased

361
pumping power.
At the low temperature , C02 provides a definite advantage over the other fluids. Again, the single-phase fluids
are similar in performance and are within approximate ly 20% of the C0 2 system. The performance of HCFC-22
is seen to drop off sharply. This due to the extremely low temperature s which are required to achieve the higher
capacities. An evaporation temperature of almost -40°C (-40°F) is requited to achieve a capacity of 2.0 k\V. The
low vapor density of HCFC~22 also leads to high pressure drops in the refrigeration coil, further reducing the COP.
The nearly identical performance of the HFE and the synthetic fluid could be expected due to the similarity in the
transport properties previously shown. However, unlike the synthetic fluid. preliminary tests have shown the HFE
to be non-toxic. This makes the HFE a more favorable heat transfer fluid.

5 CONCLUS IONS

The results presented here shmv that a secondary-loop system. using C02 as a volatile secondary refrigerant
and ammonia as the primary refrigerant. provides a COP equivalent to existing systems using HCFC-22. If this
system can be designed to overcome the high vapor pressures produced using C02, then it could provide a long
term. environment al and economic solution for supermarket refrigeration. Secondary-loop systems using single-
phase refrigerants do not perform as well due to higher pressure losses and lower heat transfer performance. The
COP of these systems could possibly be improved by increasing the refrigeration coil size. This would require
cabinet redesign and would reduce product. display areas. Despite these shortcomings, secondary-loop systems
using single-phase secondary refrigerants ,,·ould be simpler to design and may be more easily implemented. Further
improvements in both secondary-loop systems may still be possible through improvements in the secondary-loop
and the primary-loop. Additional secondary refrigerants should also be identified and investigated.
References
ASHRAE [1988]. 1988 ASHRAE Handbook - Equipment, Atlanta: American Society of Heating, Refrigerating,
and Air-Conditioning Engineers, Inc.
ASHRAE [1993]. 1993 ASHRA.E Handbook- Fundamental s, Atlanta: American Society of Heating, Refrigerating,
and Air-Conditioning Engineers, Inc. 0

Enkemann, T. & Arnemann, M. [1994]. Investigation of C02 as a Secondary Refrigerant, IIR Conference, Hanover,
Germany.
Godwin, D. S. [1994]. Results of Soft-Optimized System Tests in ARI's R-22 Alternative Refrigerants Evaluation
Program, Internationa l Refrigeratio n Conference at Purdue University, Purdue University, West Lafayette, IN
47907, pp. 7-12.
Harrison, M. R., Keeney, R. C. & Nelson, T. P. [1995]. Pilot Survey of Refrigerant Use and Emissions From Retail
Food Stores, ASHRAE Transactions 101(1): 25-33.
Hesse, U. [1995]. Secondary Refrigerant System Options For Supermarke t Refrigeration, Internationa l CFC and
H alan Alternatives Conference.
Incropera, F. P. & DeWitt, D.P. [1990]. Fundamenta ls of Heat and Mass Transfer, 3rd ed., John Wiley and Sons,
Inc.
Inlow, S. W. & Groll, E. A. [1996]. Analysis of Secondary-Loop Refrigeration Systems Using Carbon Dioxide as a
Volatile Secondary Refrigerant, Interna·tiona l Journal of Heating, Ventilating, Air-Conditio ning and Refrigera-
tion Research 2(2): 107-121.
Kauffeld, M. [1995]. - Neue N H3-Technol ogie- N H3 mit C02 als Kaltetrager, DIE KALTE und Klimatechni k
pp. 931-932.
McDowell, T. P., Klein, S. A. & Mitchell, J. W. [1995]. Investigation of Ammonia Equipment Configurations for
Supermarke t Refrigeration Applications, Epa report, University of Wisconsin-Madison.
Pearson, S. & Fellow, B. [1992/93]. Development of Improved Secondary Refrigerants, Technical report, Institute
of Refrigeration, Glasgow.
Stoecker, W. F. [1989]. Growi!lg Opportuniti es for Ammonia Refrigeration, Technical report, Internationa l Institute
of Ammonia Refrigeration.

362