(Reinhard Radermacher, Yunho Hwang) Vapor Compress
(Reinhard Radermacher, Yunho Hwang) Vapor Compress
(Reinhard Radermacher, Yunho Hwang) Vapor Compress
Compression
Heat Pumps
with Refrigerant
Mixtures
Reinhard Radermacher
Yunho Hwang
A CRC title, part of the Taylor & Francis imprint, a member of the
Taylor & Francis Group, the academic division of T&F Informa plc.
Published in 2005 by
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
Radermacher, Reinhard.
Vapor compression heat pumps with refrigerant mixes / Reinhard Radermacher, Yunho Hwang.
p. cm.
Includes bibliographical references and index.
ISBN 0-8493-3489-6 (alk. paper)
1. Heat pumps. 2. Refrigerants. I. Hwang, Yunho. II. Title.
TJ262.R33 2005
621.5′63--dc22 2005042099
∑Q + W = 0
i (1.1)
where Q stands for the amounts of heat exchanged at various temperature levels Ti
and W for the amount of work that is required in the process.
1
2 Vapor Compression Heat Pumps with Refrigerant Mixtures
The Second Law states that heat cannot flow from a lower to a higher temperature
without the expenditure of energy. One way of expressing this statement is shown
in the following equation.
i
(1.2)
Equation 1.1 is based on the assumption that the energy supplied to a system is
counted as positive.
Figure 1.1.1 illustrates the situation. The vertical axis is the temperature axis
with increasing temperature from the bottom to top, i.e., T1 < T2. A certain amount
of heat is removed by the device M at temperature T1 and, by using work that is
supplied to M, heat is rejected at T2. The amount of heat rejected is the sum of the
amount of work W required by M and the amount of heat removed at T1. This is a
consequence of the First Law,
Q1 + W = Q2 (1.3)
However, does this relationship of Figure 1.1.1 fulfill the Second Law, Equation 1.2?
Applying Equation 1.2 to the situation of Figure 1.1.1 yields
Q1 Q2
− =0 (1.4)
T1 T2
Q1 Q1 + W
− =0 (1.5)
T1 T2
T2
Q 2 = Q 1+ W
M W
Q1
T1
FIGURE 1.1.1 Heat Pumping — The process of lifting energy Q1 from a temperature level
T1 to T2, releasing Q2. The input of energy, in this case work W, is required.
Introduction 3
Q1 Q1 W
− − =0 (1.6)
T1 T2 T2
Since T2 > T1, the difference of the first two terms is still larger than zero. Equation
1.6 requires that the Second Law is fulfilled only if the work input is sufficiently
large. Thus, there is a minimum work requirement. This minimum amount of work
is required for the system to operate reversibly. The Second Law does not say
anything about an upper limit for the work, so the system may be very inefficient
and operate irreversibly if W becomes much larger than necessary. There is no upper
limit on the degree of irreversibility for the heat pumping operation.
The First and Second Laws state for lifting heat from one temperature level to
a higher one, the expenditure of energy is required and that the amount of energy
rejected at the higher temperature level is the sum of the heat removed from the
lower temperature level plus the energy added to accomplish the lift. The energy
required to drive the system can be supplied in the form of work, as assumed here,
or it can be supplied in the form of heat. Examples are absorption heat pumps,
desiccant systems and engine-driven heat pumps.
One of the most common implementations of a heat pump is shown in Figure
1.1.2. Liquid refrigerant evaporates in the evaporator at a given pressure level while
it absorbs heat. In other terms, this is the cooling effect. The compressor compresses
the vapor that exits from the evaporator and delivers the compressed, high-pressure
vapor to the condenser where the vapor is cooled and condensed. The liquefied vapor
leaves the condenser and enters the expansion valve. This valve reduces the pressure
level and meters the fluid so that as much refrigerant enters the evaporator as the
compressor removes.
The concept shown in Figure 1.1.2 represents a vapor compression heat pump.
It is the underlying concept of the type of heat pumps discussed in this text. All of these
systems have to fulfill the First and Second Law in the form of the Equations 1.1 and 1.2.
It should be noted that although both the First and Second Law must be observed
in heat pump design, the reader will notice that in subsequent chapters energy and
Condenser
Expansion
Va lve Compressor
Evaporator
FIGURE 1.1.2 Schematic of a vapor compression heat pump (or refrigeration system).
4 Vapor Compression Heat Pumps with Refrigerant Mixtures
mass balances are discussed but rarely entropy balances. This implies that the Second
Law could be violated in idealized situations. However, in all examples we are using
measured data of actual working fluids that do exist. These data inherently ensure
that the Second Law is not violated.
The following sections provide an overview of various heat pump applications.
Note that whenever energy is lifted from a lower temperature to a higher temperature,
we speak in general of a heat pumping process. However, for many applications, the
process is not as important as its result. In all refrigeration and air-conditioning
applications, energy is removed from a lower temperature level and rejected at a
higher temperature level. The main effect that interests the user is the refrigeration
or cooling effect, i.e., heat removal. On the other hand, in what is customarily called
heat pumping, the user is interested in the heat delivered at the high temperature level.
In both cases, the underlying thermodynamic process is the same; only the applica-
tion or the desired benefit are different. In this text, the term heat pumping is used as
the general description for the thermodynamic process independent of the application.
The terms refrigeration and air-conditioning refer specifically to the application.
the system is designed and sized for the air-conditioning function, the heat pumps
turn out to have insufficient heating capacity in most climates. Supplemental heat
in the form of electric or gas heat is usually required.
Another form of residential air-conditioning is the window air conditioner. It is
designed to cool just one room and the largest capacity usually does not exceed
7 kW (2 refrigeration tons). All the components are integrated within one chassis
that is usually mounted within the frame of a window.
in which food products such as fish, meats, dairy products and vegetables are cooled
or refrigerated from the moment of harvest or production through storage, transport,
wholesale and retail to the consumer, who uses the residential refrigerator as the
last storage facility in this chain. Accordingly, a wide range of refrigeration systems
have been developed that meet the various needs.
The domestic refrigerator is a unique device in that it has reached such a level
of maturity that some units last more than 20 years without ever failing or requiring
maintenance. In addition, the energy consumption has been reduced by more than
a factor of two over the past 30 years. The newest models of a typical refrigerator
for a U.S. household show a power consumption that is equivalent to that of a 50
W light bulb left on continuously.
Refrigeration systems in supermarkets usually have central compressor plants
with multiple compressors in parallel that serve the same set of evaporators. The
evaporators are located throughout the store in the display cabinets for frozen or
refrigerated foods. As a consequence, a single store can boast miles of refrigerant
piping. Due to the length of piping and the many inherent piping connections,
supermarket refrigeration systems tended to lose considerable amounts of refrigerant.
With the advent of environmental concerns and improved quality standards, these
leakages have been reduced considerably. New systems are under consideration in
which the refrigerant does not leave the machine room. Rather, a secondary refrigerant
is used to cool the display cabinets. This measure reduces refrigerant leakage and
increases system reliability, but may contribute to a reduction in energy efficiency.
Transport refrigeration systems, especially those that are trailer or container
mounted, must come with their own power supply, usually a small diesel or gas
engine. Systems must be especially robust to withstand the high vibration levels.
Also, these systems must be designed to cover a wide range of operating conditions.
The transported goods may require storage temperatures ranging from –40 to 20˚C
while the ambient temperature can span even a wider range of temperatures.
For food preservation, there are also a wide array of specialized required cooling
systems. Examples are milk coolers on farms, cooling systems for tank trucks and
refrigeration systems ranging from fish trawlers to wine coolers.
The air-conditioning of cars has become an expected feature in the United States
and many developed and developing countries. The cooling capacity varies over a
wide range because the compressor is directly coupled to the engine and the com-
pressor revolutions vary with that of the engine. Furthermore, due to vibration and
the need to use flexible hoses to connect the components that are mounted on the
car and those mounted on the engine block, these systems traditionally have a
relatively high rate of refrigerant leakage. With the advent of environmental con-
cerns, the leakage rates have been reduced dramatically.
working fluids, commonly termed refrigerants. They represent one of the key ingre-
dients of any vapor compression system and influence design, operation and cost
immensely. A short overview of the historical development of refrigerants follows
that provides insight into the development of the refrigeration systems. The following
is an excerpt of a paper presented by Calm and Didion in 1997.
Refrigeration goes back to ancient times, using stored ice and a number of
evaporative processes. Early in the 19th century, the use of a volatile fluid in a closed
cycle was proposed by Oliver Evans. He used evaporating ether, under a vacuum,
and then pumped the vapor to a water-cooled heat exchanger to be condensed for
reuse.
Other refrigerants were introduced in the 1830s with the invention of the vapor-
compression machine by Perkins. He designed a machine to use sulfuric (ethyl)
ether as the refrigerant (Perkins, 1834). The first machine actually used caoutchou-
cine, an industrial solvent that Perkins apparently utilized in his business as a printer.
It seems the first tradeoff in refrigerants — and one still driving selections — was
based on availability.
Table 1.1 summarizes the historic introduction of refrigerants. The years given
serve only as guidelines, since the period from the introduction of a new fluid to its
generally accepted use always spans a number of years. Table 1.1 summarizes early
refrigerants, namely those predating fluorinated chemicals and state-of-the-art fluoro-
chemicals that are under consideration today. Downing (1988), Nagengast (1989
and 1996) and Thévenot (1979) present further details.
The first century of refrigerant use was dominated by innovative efforts with
familiar fluids. The goals were to provide refrigeration and, later, to improve machine
durability. Use of blends was attempted where single-compound solutions could not
be found (Pictet, 1885). After World War I, attention turned to safety and performance
as well. Willis H. Carrier and R.W. Waterfill initiated one of the first documented
systematic searches (Carrier and Waterfill, 1924). They investigated a range of
candidate refrigerants for suitability for both positive displacement and centrifugal
compression machines. These analyses closely examined ammonia, ethyl ether,
carbon dioxide, carbon tetrachloride, sulfur dioxide and water. They finally selected
dielene (1,2-dichloroethene, R1130) for the first centrifugal machine, though an
international search was needed to find a source for it (Ingels, 1952).
Nearly all of the early refrigerants were flammable, toxic or both, and some
were also highly reactive. Accidents were common. For perspective, propane was
marketed as the odorless safety refrigerant (CLPC, 1922).
The discovery of fluorinated refrigerants began with Thomas Midgley, Jr., in
1928 (Midgley, 1937). Midgley and his associates, Albert L. Henne and Robert R.
McNary, scoured property tables to find chemicals with the desired boiling point.
They restricted the search to those known to be stable but not toxic or flammable.
They eventually used organic fluorides. While fluorine was known to be toxic by
itself, Midgley and his collaborators felt that compounds containing it could be
nontoxic.
Recognizing the deficiencies of the published literature, Midgley, Henne and
McNary turned to the periodic table of the elements. They quickly eliminated those
elements yielding insufficient volatility and then eliminated those resulting in unstable
8 Vapor Compression Heat Pumps with Refrigerant Mixtures
TABLE 1.1
Introduction of Refrigerants
Year Refrigerant Chemical Makeup, Formula
and toxic compounds as well as the inert gases, based on their low boiling points.
They were left with just eight elements: carbon, nitrogen, oxygen, sulfur, hydrogen,
fluorine, chlorine and bromine. These elements clustered at an intersecting row and
column of the periodic table of the elements, with fluorine at the intersection.
Midgley and his colleagues then made three interesting observations. First,
flammability decreases from left to right for the eight elements. Second, toxicity
generally decreases from the heavy elements at the bottom to the lighter elements
Introduction 9
at the top. And third, every known refrigerant at the time was made from combinations
of those elements. The historical fluids actually comprised only seven of the eight
Midgley elements, since there appears to be no record of prior use of refrigerants
containing fluorine.
The first publication on fluorochemical refrigerants shows how chlorination and
fluorination of hydrocarbons can be varied to provide desired boiling points (Midgley,
1930). This paper also shows how the composition of the molecule influences relative
flammability and toxicity. Commercial production of R12 began in 1931, followed
by R11 in 1932 (Downing, 1966 and 1984).
Other investigators have repeated Midgley’s search with newer methods and
modern databases and have come to similar findings. McLinden and Didion (1987)
documented an extensive screening of industrial chemicals. Of the chemicals meet-
ing their criteria, all but two — both highly reactive and toxic — contained the
Midgley elements.
Thus, refrigerants must be safe locally and globally. To be safe locally means
fluids should be nontoxic and nonflammable and pose no danger of explosion. They
must also not affect any construction materials. These criteria aim at safety during
10 Vapor Compression Heat Pumps with Refrigerant Mixtures
operation, manufacture, service, use and final disposal. To be safe globally means
refrigerants should not affect the environment in a significantly negative way.
In order to meet these requirements, an increasing effort is made to use refrig-
erant mixtures. They offer the potential to be the best compromise for fluids that
meet some of the thermodynamic requirements that are very important for the
performance of the equipment and hopefully all the other requirements that deter-
mine feasibility in many different aspects. Refrigerant mixtures are gaining impor-
tance as the acceptable working fluid of future generations of equipment. However,
the thermo-physical properties of mixtures are considerably more complex than
those of pure fluids. It is the focus of this text to comprehensively present the
characteristics of mixtures to the advanced student and practicing engineer.
TABLE 1.2
HCFCs Regulation Schedule
Jan. 1, 1996 2004 2010 2015 2020 2030
The base level is the sum of its 1989 ozone-depletion potential (ODP) weighted HCFCs
consumption plus 3.1% of its 1989 ODP weighted CFCs consumption.
REFERENCES
Albritton, D., 1997, “Ozone Depletion and Global Warming,” ASHRAE/NIST Refrigerants
Conference: Refrigerants for the 21st Century, ASHRAE. Atlanta, GA, pp. 1–5.
Allied Signal Chemicals, 1999, Genetron AZ-20 Product Brochure.
Angelino, G. and C. Invernizzi, 1988, “General method for the thermodynamic evaluation of
heat pump working fluids,” International Journal of Refrigeration, Vol. 11, No. 1,
pp. 16–25.
ANSI/ASHRAE Standard 15-1994: “Safety Code for Mechanical Refrigeration,” American
Society of Heating, Refrigerating, and Air-Conditioning Engineers. Atlanta, GA.
ANSI/ASHRAE Standard 34-1997: “Designation and Safety Classification of Refrigerants
(including addenda 34a through 34f),” American Society of Heating, Refrigerating,
and Air-Conditioning Engineers, Atlanta, GA.
ARI, 1992, “Centrifugal and Rotary Screw Water-Chilling Packages,” Standard 550-92, Air-
Conditioning and Refrigeration Institute, Arlington, VA.
ARI, 1996, “Inputs for AFEAS/DOE Phase 3 Study of Energy and Global Warming Impacts,”
Air-Conditioning and Refrigeration Institute, Arlington, VA.
ARI, 2000, “Statistical Release,” Air-Conditioning and Refrigeration Institute, Arlington, VA.
ASHRAE Handbook of Fundamentals, 1993, Chapter 16, American Society of Heating,
Refrigerating, and Air-Conditioning Engineers, pp. 2–6.
ASHRAE, 1993, “Addendum to Number Designation and Safety Classification of Refriger-
ants,” ANSI/ASHRAE Standard 34a-1993 (Addendum to ANSI/ASHRAE 34-1992),
American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Atlanta,
GA.
Atwood, T. and H.M. Hughes, 1990, “Refrigerants and energy efficiency,” Proceedings of the
1990 USNC/IIR-Purdue Refrigeration Conference and ASHRAE-Purdue CFC Con-
ference, West Lafayette, IN, pp. 80–89, July.
Bhatti, M.S., 1999, “A Historical Look at Chlorofluorocarbon Refrigerants,” ASHRAE Trans-
actions, Vol. 105, Pt. 1, pp. 1186–1208.
Bivens, D.B. and B.H. Minor, 1997, “Fluoroethers and Other Next Generation Fluids,”
ASHRAE/NIST Refrigerants Conference: Refrigerants for the 21st Century, The
National Institute of Standards and Technology, Gaithersburg, MD, pp. 119–131.
Bivens, D.B. and, B.H. Minor, 1998, “Fluoroethers and Other Next Generation Fluids,”
International Journal of Refrigeration, Vol. 21, No. 7, pp. 567–576.
Calm, J.M., 1993, “Comparative Global Warming Impacts of Electric Vapor-Compression
and Direct-Fired Absorption Equipment,” Report TR-103297, Electric Power
Research Institute (EPRI), Palo Alto, CA.
Introduction 13
Calm, J.M, 1994, “Refrigerant Safety,” ASHRAE Journal, Vol. 36, No. 7, pp. 17–26.
Calm, J.M., D.J. Wuebbles, and A. Jain, 1997, “Impacts on global ozone and climate from
2,2-dichloro-1,1,1-trifluoroethane (HCFC-123) emissions,” Journal of Climatic
Change, submitted, Car Lighting and Power Company (CLPC), 1922, Advertisement,
Ice and Refrigeration, p. 28.
Calm, J.M. and D.A. Didion, 1997, “Trade-Offs in Refrigerant Selections: Past, Present,
Future,” International Journal of Refrigeration, Vol. 21, No. 5, pp. 6–19.
Carrier, W.H. and R.W. Waterfill, 1924, “Comparison of Thermodynamic Characteristics of
Various Refrigerating Fluids,” Refrigerating Engineering.
Clayton, J.W., Jr., 1967, “Fluorocarbon Toxicity and Biological Action,” Fluorine Chemistry
Reviews, Vol. 1, pp. 197–252.
CLPC (Car Lighting and Power Company), 1922, Ice and Refrigeration, Advertisement, Nov.,
p. 28.
Corr, S., J.D. Morrison, F.T. Murphy and R.L. Powell, 1995, “Developing the hydrofluoro-
carbon range: Fluids for centrifugal compressors,” Proceedings of the 19th Interna-
tional Congress of Refrigeration, Paris, France: International Institute of Refrigeration
(IIR), IVa, pp. 31–138.
Dekleva, T.W., 1994, “Flammability Testing: Observations related to HFC systems,” Presen-
tation 2.5, ARI Flammability Workshop — Summary and Proceedings, Arlington,
VA: Air-Conditioning and Refrigeration Institute (ARI).
Didion, D.A., 1994, “The Impact of Ozone-Safe Refrigerants on Refrigeration Machinery
Performance and Operation,” Transactions of the Society of Naval Architects and
Engineers.
Didion, D.A., 1999, “The Application of HFCs as Refrigerants,” Proceedings of the Centenary
Conference of the Institute of Refrigeration, London, SW1H 9JJ.
Didion, D.A., 1999, “The Influence of the Thermophysical Properties of the New Ozone-Safe
Refrigerants on Performance,” International Journal of Applied Thermodynamics,
Vol. 2, No. 1, pp. 19–35.
Didion, D.A. and M.S. Kim, 1999, “NIST Leak and Recharge Simulation Program for
Refrigerant Mixtures,” NIST Standard Reference Database 73: REFLEAK, Version
2.0, National Institute of Standards and Technology, Gaithersburg, MD.
Dittus, F.W. and L.K.M. Boelter, 1930, University of Calif., Berkeley, Publ. Engr., Vol. 2,
p. 443.
Doerr, R.G., D. Lambert, R. Schafer and D. Steinke, 1993, “Stability studies of E-245
Fluoroether, CH3–CH2–O–CHF2,” ASHRAE Transactions, Vol. 99, Pt. 2, pp. 1137–1140.
Domanski, P.A., D.A. Didion and J.P. Doyle, 1994, “Evaluation of Suction Line — Liquid
Line Heat Exchange in the Refrigeration Cycle,” International Journal of Refriger-
ation, Vol. 17, No. 7, pp. 487–493.
Domanski, P.A., D.A. Didion and J.S.W. Chi, 1997, “CYCLE_D: NIST Vapor-Compression
Design Program,” Standard Reference Database 49, Gaithersburg, MD: National
Institute of Standards and Technology (NIST).
Domanski, P., 1997, “Theoretical Evaluation of the Vapor Compression Cycle With a Liquid-
Line/Suction-Line Heat Exchanger, Economizer, and Ejector,” NISTIR 5606, The
National Institute of Standards and Technology, Gaithersburg, MD.
Downing, R.C., 1966, “History of the organic fluorine industry,” Kirk-Othmer Encyclopedia
of Chemical Technology (2nd edition). New York, NY: John Wiley & Sons, Incorpo-
rated, Vol. 9, pp. 704–707.
Downing, R.C., 1984, “Development of chlorofluorocarbon refrigerants,” ASHRAE Transac-
tions. Atlanta, GA: American Society of Heating, Refrigerating, and Air-Conditioning
Engineers (ASHRAE), Vol. 90, Pt. 2B, pp. 481–491.
14 Vapor Compression Heat Pumps with Refrigerant Mixtures
Downing, R.C., 1988, Fluorocarbon Refrigerants Handbook, Englewood Cliffs, NJ, Prentice
Hall.
DuPont Fluorochemicals, 2000, “Technical Information for Suva Refrigerants,” http://www.
dupont.com/suva, Wilmington, DE.
Evans, O., 1805, The Abortion of a Young Steam Engineer’s Guide, Philadelphia, PA.
Fairchild, P. et al., 1991, “Total Equivalent Global Warming Impact: Combining Energy and
Fluorocarbon Emissions Effects,” Proceedings of the International CFC and Halon
Alternatives Conference, Baltimore, MD.
Fischer, S.K., P. Hughes and P. Fairchild, 1991, “Energy and Global Warming Impacts of
CFC Alternative Technologies,” Alternative Fluorocarbons Environmental Accept-
ability Study (AFEAS) and U.S. Department of Energy (DOE).
Fischer, S.K., J.J. Tomlinson and P.J. Hughes, 1994, “Energy and Global Warming Impacts
of Not-In-Kind and Next Generation CFC and HCFC Alternatives,” Alternative
Fluorocarbons Environmental Acceptability Study (AFEAS) and U.S. Department of
Energy (DOE).
Glamm, P.R., E.F. Keuper and F.B. Hamm, 1996, “Evaluation of HFC-245ca for Commercial
Use in Low Pressure Chillers,” Report DOE/CE/23810-67, Arlington, VA, Air-
Conditioning and Refrigeration Technology Institute (ARTI).
Godwin, D., 1993. “Results of System Drop-In Tests in ARI’s R-22 Alternative Refrigerants
Evaluation Program, Arlington, VA: ARI.
Gorenflo, D. and V. Bieling, 1986, “Heat transfer in pool boiling of mixtures with R-22 and
R-115,” XVII Int. Symposium on Heat and Mass Transfer in Cryogenics and Refrig-
eration, International Institute of Refrigeration, Paris.
Houghton, L.G. et al., 1995, “Climate Change 1994,” Radiative Forcing of Climate Change
and an Evaluation of the IPCC IS92 Emissions Scenarios, Cambridge University Press.
Huber, M.L., J.S. Gallagher, M.O. McLinden and G. Morrison, 1996, “NIST Thermodynamic
Properties of Refrigerants and Refrigerant Mixtures Database (REFPROP),” Standard
Reference Database 23, Version 5.0. Gaithersburg, MD, National Institute of Stan-
dards and Technology (NIST).
Ingels, M., 1952, Willis Haviland Carrier — Father of Air Conditioning. Syracuse, NY: Carrier
Corporation.
Intergovernmental Panel on Climate Change (IPCC), 1996, “Climate Change 1995 — Con-
tribution of Working Group I to the Second Assessment Report of the Intergovern-
mental Panel on Climate Change,” edited by J.T. Houghton, L.G. Meira Filho, B.A.
Callander, N. Harris, A. Kattenberg and K. Maskell. Cambridge, UK: Cambridge
University Press.
Johnson, J. and T. Watson, 1993, “R-134A Compatibility with Centrifugal Chillers,” Proceed-
ings of the International Seminar on New Technology of Alternative Refrigerants,
Keidanren Hall, Tokyo, Japan.
Kates, R.W. et al., National Academy Report, 1999, “Our Common Journey, a Transition
Toward Sustainability,” National Academy Press, Washington D.C., USA.
Kaul, M., M. Kedzierski and D.A. Didion, 1996, “Horizontal Flow Boiling of Alternative
Refrigerants within a Fluid Heated Micro-Fin Tube,” Process, Enhanced, and Multi-
phase Heat Transfer: A Festschrift for A.E. Bergles, Begell House, Inc., New York, NY.
Kedzierski, M.A., J.H. Kim and D.A. Didion, 1992, “Causes of the Apparent Heat Transfer
Degradation for Refrigerant Mixtures,” ASME HTD-197, Two Phase Flow and Heat
Transfer, American Society of Mechanical Engineers, New York, NY.
Kim, Man-Hoe, P.A. Domanski and D.A. Didion, 1997, “Performance of R-22 Alternative
Refrigerants in a System with Cross-Flow and Counter-Flow Heat Exchangers,”
NISTIR 5945, National Institute of Standards and Technology, Gaithersburg, MD.
Introduction 15
Kruse, H., 1993, “European Research and Development Concerning CFC and HCFC Substi-
tution,” Proceedings of ASHRAE/NIST Refrigerants Conference, Gaithersburg, MD,
pp. 41–57.
Lorenz, H., 1984, Die Ausnutzung der Brennstoffe in den Kühlmaschinen [Use of Flammable
Substances (commonly Fuels) in Refrigeration Equipment]. Germany: Zeitschrift für
die gesamte Kälte Industrie, Vol. 1, pp. 10–15.
Marques, M., S.A. Multibras and P.A. Domanski, 1998, “Potential Coefficient of Performance
Improvements due to Glide Matching with R407C,” Seventh International Refriger-
ation Conference at Purdue, West Lafayette, IN, pp. 101–108.
McLinden, M.O. and D.A. Didion, 1987, “Quest for alternatives,” ASHRAE Journal, Vol. 29,
No. 12, pp. 32–36, 38, 40 and 42.
McLinden, M.O., 1988, “Thermodynamic Evaluation of Refrigerants in the Vapor Compres-
sion Cycle Using Reduced Properties,” International Journal Refrigeration, Vol. 11,
No. 3, pp. 134–143.
McLinden, M.O., 1990, “Optimum refrigerants for non-ideal cycles — An analysis employing
corresponding states,” Proceedings of the 1990 USNC-IIR-Purdue Refrigeration Con-
ference and ASHRAE-Purdue CFC Conference, West Lafayette, IN, pp. 69–79.
McLinden, M., S.A. Klein, E.K. Lemmon and A.P. Perkin, 1998, “Thermodynamic and
Transport Properties of Refrigerants and Refrigerant Mixtures,” NIST Database 23:
NIST REFPROP, Version 6.0, National Institute of Standards and Technology, Gaith-
ersburg, MD.
Midgley, T., Jr. and A. L. Henne, 1930, “Organic fluorides as refrigerants,” Industrial and
Engineering Chemistry, Vol. 22, pp. 542–545.
Midgley, T., Jr., 1937, “From the periodic table to production,” Industrial and Engineering
Chemistry, Vol. 29, No. 2, pp. 239–244.
Molina, M. J. and F.S. Rowland, 1974, “Stratospheric sink for Chlorofluoromethanes: Chlorine
atom catalyzed destruction of ozone,” Nature, Vol. 249, pp. 810–812.
Morrison, G., 1974, “The shape of the temperature-entropy saturation boundary,” Interna-
tional Journal of Refrigeration, Vol. 17, No. 7, pp. 494–504.
Nagengast, B.A., 1989, “A history of refrigerants,” CFCs: Time of Transition, Atlanta, GA,
American Society of Heating, Refrigerating, and Air-Conditioning Engineers, pp. 3–15.
Nagengast, B.A., 1996, “History of sealed refrigeration systems,” ASHRAE Journal, Vol. 38,
No. 1, S37, S38, S42–S46 and S48.
NERC, 1995, Electric Supply and Demand Database (Version 2.0), Princeton, NJ: North
American Electric Reliability Council.
Perkins, J., 1834, Apparatus for Producing Ice and Cooling Fluids, Patent No. 6662, United
Kingdom.
Petterson, J., 1999, “Carbon Dioxide (CO2) as a Primary Refrigerant,” Proceedings of the
Centenary Conference of the Institute of Refrigeration, London, SW1H 9JJ.
Pictet, R., 1885, Aus Kohlensäure und schwefliger Säure bestehende Verflüchtigungsflüssigkeit
für Kältemaschinen [A Volatile Liquid Comprising Carbon Dioxide and Sulfur Diox-
ide for Refrigeration Machines], Patent 33733 (Class 17, Ice Production and Storage),
Germany: Kaiserliches Patentamp.
Pillis, J.W., 1993, “Expanding Ammonia Usage in Air Conditioning,” Proceedings of the
ASHRAE/NIST Refrigerants Conference: R-22/R-502 Alternatives, American Society
of Heating, Refrigerating, and Air Conditioning Engineers, Atlanta, GA, pp. 103–108.
Reed, J.W., 1993, “Environmental Overview: CFC and HCFC Regulatory Update,” Proceed-
ings of the 4th IEA Heat Pump Conference, Maastricht, the Netherlands, pp. 11–19.
Richard, R.G. and I.R. Shankland, 1992, “Flammability of alternative refrigerants,” ASHRAE
Journal, Vol. 34, No. 4, pp. 20 and 22–24, April.
16 Vapor Compression Heat Pumps with Refrigerant Mixtures
Ross, H., R. Radermacher, M. di Marzo and D. Didion, 1987, “Horizontal flow boiling of
pure and mixed refrigerants,” International Journal of Heat and Mass Transfer, Vol. 30,
pp. 972–992.
Sand, J.R. and S.K. Fischer, 1994, “Modeled Performance of Non-chlorinated Substitutes for
CFC-11 and CFC-12 in Centrifugal Chillers,” International Journal of Refrigeration,
Vol. 17, No. 1, pp. 40–48.
Sand, J.R., S.K. Fischer and V.D. Baxter, 1997, “Energy and Global Warming Impacts of
HFC Refrigerants and Emerging Technologies,” Alternative Fluorocarbons Environ-
mental Acceptability Study (AFEAS) and U.S. Department of Energy (DOE).
Smith, N.D., K. Ratanaphruks, M.W. Tufts and A.S. Ng, 1993, “R-245ca: A potential far-
term alternative for R-11,” ASHRAE Journal, Vol. 35, No. 2, pp. 19–23.
Smith, N.D. and M.W. Tufts, 1994, “Flammable properties of HFC refrigerants — Some
fundamental considerations,” Presentation 2.2, ARI Flammability Workshop — Sum-
mary and Proceedings, Arlington, VA: Air-Conditioning and Refrigeration Institute
(ARI).
Starner, K.E., 1993, “Heat Exchangers for Ammonia Water Chillers: Design Considerations
and Research Needs,” Proceedings of the ASHRAE/NIST Refrigerants Conference:
R-22/R-502 Alternatives, American Society of Heating, Refrigerating, and Air Con-
ditioning Engineers, Atlanta, GA, pp. 85–90.
Stera, A.C., 1993, “Developments in Transportation of Chilled Produce by Sea and Air,” IIR
Conference Proceedings: Cold Chain Refrigeration Equipment by Design, Interna-
tional Institute of Refrigeration, Paris.
Stoecker, W., 1989, “Expanded Applications for Ammonia — Coping with Releases to the
Atmosphere, CFCs: Today’s Options — Tomorrow’s Solutions,” Proceedings of
ASHRAE’s 1989 CFC Technology Conference, ASHRAE, Atlanta, GA.
Thévenot, R., 1979, A History of Refrigeration throughout the World, Translated from French
by J.C. Fidler. Paris, France: International Institute of Refrigeration.
Threlkeld, J.L., 1970, Thermal Environmental Engineering (2nd edition), Prentice Hall, Inc.
UNEP, 1987, “Montreal Protocol on Substances That Deplete the Ozone Layer,” United
Nations Environment Programme (UNEP).
UNEP, 1996, Handbook for the Montreal Protocol on Substances That Deplete the Ozone
Layer (4th edition), Nairobi, Kenya: United Nations Environment Programme
(UNEP) Ozone Secretariat.
WMO, 1994, “Scientific Assessment of Ozone Depletion: 1994,” Chaired by D.L. Albritton,
R.T. Watson and P.J. Aucamp, Global Ozone Research and Monitoring Project Report
37, Geneva, Switzerland, World Meteorological Organization (WMO).
Wood, B.D., 1982, Applications of Thermodynamics (2nd edition). Reading, MA, Addison-
Wesley Publishing Company, p. 40.
Wuebbles, D.J., 1981, “The Relative Efficiency of a Number of Halocarbons for Destroying
Stratospheric Ozone,” Report UCID-18924, Livermore, CA, Lawrence Livermore
National Laboratory (LLNL).
Wuebbles, D.J., 1995, “Weighing functions for ozone depletion and greenhouse gas effects
on climate,” Annual Review of Energy and Environment, Vol. 20, pp. 45–70.
Wuebbles, D.J. and J.M. Calm, 1997, “An environmental rationale for retention of endangered
chemical species,” Science.
2 Properties of
Working Fluids
The performance and efficiency of reversible cycles are independent of the properties
of any working fluids. However, the performance and efficiency of a real machine
are determined to a large degree by the properties of the working fluids. Furthermore,
both the first cost and the operating cost of a heat pump are strongly dependent on
the refrigerant properties.
17
18 Vapor Compression Heat Pumps with Refrigerant Mixtures
This variable includes values in the range 0.0 ≤ x ≤ 1.0, with 0.0 and 1.0 denoting
the respective pure components. The various definitions for variables describing
mixture compositions are discussed in Section 2.2.2.
50
30 Dew line
P = 1 MPa
2-p
has
e ra
ng e
20
10 T R32
Bubble line
Subcooled liquid
0
0.0 0.2 0.4 0.6 0.8 1.0
Pure R134a Mass Fraction of R32 [kg/kg] Pure R32
50
Superheated vapor
TR134a
40
Dew line
5"
Temperature [°C]
30
P = 1 MPa
T4 4' 4"
T3 3' 3 3"
20 T2 2"
2-ph
2' ase r
ange
TR32
10 1'
x2' x3' Bubble line
Subcooled liquid
x3" x3'
0 x3" x1,2 x3'
0.0 0.2 0.4 0.6 0.8 1.0
Pure R134a Mass Fraction of R32 [kg/kg] Pure R32
The vapor quality at state 3 is represented in Figure 2.1.2 by the ratio of the distances
(x3 – x3′)/(x3″ – x3′) and is expressed as a function of mass fractions only. This is a
unique feature of fluid mixtures.
It is noted that generally the temperature glide increases with increasing differ-
ence between the boiling points of the two pure components. Figure 2.1.3a shows
a temperature-mass fraction diagram for a mixture where the difference in boiling
points is relatively small, thus the temperature glide T is relatively small as well.
Figure 2.1.3b shows the temperature-mass fraction diagram of a mixture where the
difference in the boiling points is large. Accordingly, the temperature glide marked
as T is large as well. The size of the temperature glide is also a function of the mass
fraction. For small and large x, the glide is generally smaller than for intermediate
values of x. The mixtures that show the same characteristics as R32/R134a in Figures
2.1.1 through 2.1.3 are traditionally termed nonazeotropic mixtures or, in more recent
literature, zeotropic mixtures. The name implies that in phase equilibrium, the mass
fractions of the vapor and liquid phases are always different.
Some fluids form from azeotropic mixtures, such as a mixture of R12 and R152a,
or a mixture of water and ethanol. For an azeotropic mixture, the mass fractions of
the liquid and vapor phase are identical at a certain pressure and temperature as
shown in Figure 2.1.4. This state is called the azeotropic point and the mass fraction
of this point is called the azeotropic mass fraction. The temperature glide is zero at
this point. At all other mass fractions, the mixture exhibits zeotropic behavior. The
difference in mass fraction between the liquid and vapor phases changes its sign
when the overall mass fraction varies from a value less than the azeotropic mass
fraction to a value larger than the azeotropic mass fraction.
There are unique features of azeotropic mixtures that can make their use bene-
ficial. First, they provide another unique normal boiling point that might not be
covered by one of the pure compound refrigerants and therefore the possibility exists
of fitting some specific application more precisely. Second, they have all of the
benefits of pure compound refrigerants — constant boiling point and heat of vapor-
ization, without the fractionation. Third, they can make up for the physical or
chemical deficiency of one of the zeotropic components, like covering up the flam-
mability of R152a in R500 (Fairchild, 2000).
There are two types of azeotropes, distinguished from one another by the location
of the boiling points at the azeotropic mass fraction relative to the boiling points of
the pure fluids. The boiling point can be either higher than the boiling point of both
Properties of Working Fluids 21
Pressure = Constant
Dew line
T Two-phase
range
Boiling
line
x
x
(a)
Pressure = Constant
Dew line
T Two-phase
range
∆T
Boiling
line
∆x
x
(b)
of the two constituents of the mixture, as in Figure 2.1.4, or lower than the boiling
point of the two constituents, as illustrated in Figure 2.1.5.
Figure 2.1.6 shows the T-x diagram for the azeotrope R507A at a pressure of
0.1 MPa. R507A is a mixture of R125 and R143a with a mass fraction of 0.5 of
R125. Although Figure 2.1.6 shows that the mass fraction of the azeotropic point is
approximately 0.35 of R125 (the maximum in the saturation line), it is also noteworthy
that the temperature glide is very small throughout the entire mass fraction range.
22 Vapor Compression Heat Pumps with Refrigerant Mixtures
Pressure = const.
Temperature
Azeotropic point
A Mass Fraction B
FIGURE 2.1.4 T-x diagram for zeotropic mixture. (Tsat of the azeotropic point > Tsat of the
pure fluid.)
Pressure = const.
Temperature
Azeotropic point
A Mass Fraction B
FIGURE 2.1.5 T-x diagram for zeotropic mixture. (Tsat of the azeotropic point < Tsat of the
pure fluid.)
R125/R143a Mixture
−47.0
Superheated vapor
Sat. liquid
Sat. vapor
Dew line
P = 0.1 MPa
−47.5
Temperature [°C]
Subcooled liquid
−48.5
0.0 0.2 0.4 0.6 0.8 1.0
Pure R143a Mass Fraction of R125 [kg/kg] Pure R125
R125/R143a Mixture
17
Superheated vapor Sat. liquid
Dew line
Sat. vapor
16 P = 1 MPa
Temperature [°C]
15
Bubble line
14
Subcooled liquid
13
0.0 0.2 0.4 0.6 0.8 1.0
Pure R143a Mass Fraction of R125 [kg/kg] Pure R125
Figure 2.1.7 shows the T-x diagram for R507A at a pressure of 1 MPa. An
inspection and comparison of Figures 2.1.6 and 2.1.7 reveal that the azeotropic point
moves. In this case, the mass fraction of the azeotropic point decreases as the pressure
is increased from 0.1 MPa (Figure 2.1.6) to 1 MPa (Figure 2.1.7) and the azeotropic
point seems to disappear to the left. This is an interesting characteristic of azeotropes.
24 Vapor Compression Heat Pumps with Refrigerant Mixtures
The azeotropic point changes its mass fraction with the saturation pressure level and
can disappear altogether for certain conditions. A zeotropic mixture may have an
azeotropic point at other conditions and vice versa.
While an azeotropic point is more likely to exist when the boiling points of the
constituents are close together, it is certain to exist when the vapor pressure curves
Psat(T) of the two pure components cross. Such a point usually does not exist when
the boiling points are more than 30 K apart.
12
3
NH
10
CO2
8
Pressure [MPa]
R32
6
R1 opan22
34 e
Pr R
Bu ane
a
ne
t
bu
23
ta
25
R1
Iso
4
R1
ene
pyl
Pro
2
0
−50 0 50 100 150 200
Temperature [°C]
FIGURE 2.1.8 Vapor pressure curves for a number of pure fluids. The curves end at the
critical points of the respective fluids.
100
5 R22 NH 3
10 CO 2 R12
R32 ane
y le n e Prop4a buta
ne e
Prop R13 Iso Butan
R123
1
Pressure [MPa]
0.1
0.01
0.001
−50 0 50 100 150 200
Temperature [°C]
100
2
R2
3
10 NHpane
CO 2 2
R3 Pr o
25
R1
l ene
py
Pro
Pressure [MPa]
0.1
a
34
R1ane
t
bu ne
0.01 Iso uta
B
23
R1
0.001
−5.0 −4.5 −4.0 −3.5 −3.0 −2.5 −2.0
−1/Temperature × 1000 [1/K]
FIGURE 2.1.10 Vapor pressure curves on a ln(P)-(–1/T) diagram. (The lines are almost
straight.)
The advantage of this representation is the fact that the plots of bubble point
temperature versus saturation pressure are almost straight lines for most fluids and
fluid mixtures when the mass fraction is constant. These plots are also referred to
as vapor pressure curves. The following statements are valid for pure fluids only.
The area to the left of the vapor pressure curve of a given fluid represents the
subcooled liquid region. This is the region of higher pressures or lower temperatures
as compared to the saturation values. The area to the right represents superheated
vapor, the region of higher temperatures or lower pressures compared to the satu-
ration values. The two-phase range is represented by the vapor pressure line itself.
In some representations, the vapor pressure curves are drawn exactly as straight
lines. These diagrams were originally developed by Dühring and are now referred
to as Dühring plots.
For fluid mixtures, curves similar to vapor pressure curves can be obtained by
plotting the vapor pressure versus –1/T for the isosteres (lines of constant mass
fraction) of the saturated liquid phase (vapor quality is zero).
Figure 2.1.11 shows an example for R32/R134a. The space between the vapor
pressure curves of the pure constituents of the mixture is termed the solution field.
Pressure-temperature diagrams of mixtures are used extensively for the representa-
tion of cycle configurations in absorption technology.
Azeotropic mixtures yield vapor pressure lines similar to those for pure com-
ponents shown in Figure 2.1.11. However, while in zeotropic mixtures all vapor
pressure curves for mass fractions x with 0.0 < x < 1.0 are located between the lines
for the two pure components, the vapor pressure curve for the azeotropic mass
10
0.2 0. 6
1
n
Pressure [MPa]
ctio
fra
ass
m
34a
R1
0.0
0.1 0.4
0.8
1.0
0.01
−4.5 −4.0 −3.5 −3.0 −2.5
−1/Temperature × 1000 [1/K]
fraction of an azeotropic mixture is located outside the range limited by the vapor
pressure curves of the pure components.
00
0 m
00
500
0
80 70 kg/ 400
120
90
10
11
10 600
300
200
150
R-22
80
4 100
60
70
2 50
40
3
30 kg/m
Density =
1.0
20 20
Pressure (MPa)
15
T = 0°C
10
0.4
7.
−20
-K
kg
5.
0.2
kJ/
1.8
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
2.1
1.9
2.2
2.0
S=
3.
0.1 −40
2.
0.9
1.5
T = 160 °C
200
180
140
120
100
1.0
S = 0.8
80
0.04 −60
60
40
20
0.7
0
−20
−40
vapor
uid
0.5
0.02
d liq
x = 0.5
0.6
0.8
0.3
0.7
0.1
0.2
0.4
0.9
2.3
Saturated
2.4
2.2
rate
S=
Satu
0.01
100 150 200 250 300 350 400 450 500 550 600
Enthalpy (kJ/kg)
FIGURE 2.1.12 Pressure-enthalpy diagram of R22. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
4 60 60
50
40
40
2 30
Pressure (MPa)
24
20 20
16
1
12
T = 0°C
10
8
g-K
2
J/k
S = 1.5
1.20
0.96
1.04
1.12
1.60
1.28
1.36
1.44
1.76
1.68
0
8
2
6
6
0
4
4
6
2
4
8k
2.4
2.0
2.3
2.5
2.1
2.0
1.8
1.9
2.2
−20
2.7
2.6
0.4
2.4
5
S=
0.88
4
3.2
S = 0.8
200
180
0.2
T = 100°C
160
140
−40
120
2.4
vapor
80
2
60
quid
.5
40
X=0
20
0.1
0.7
0.2
0.6
0.8
0.9
Saturated
0.4
d li
0.3
1.6
0
0.72
−20
rate
−40
0.1 1.4
Satu
8
1.2
2.8
4
96
3.0
2.8
2.7
S=
2.
0.07
100 160 220 280 340 400 460 520 580 640 700 760
Enthalpy (kJ/kg)
FIGURE 2.1.13 Pressure-enthalpy diagram of R32. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
g/m 3
1600
00
1500
1550
0
0
00
50
900
50
R-123
00
800 600
0k
140
145
130
10
12
12
11
11
135
10
400
320
240
180
3
/m
4 = 160 kg
Density
160
120
140
80
2 60
120
40
100
1 32
Pressure (MPa)
24
80
T = 60°C
16
0.4 12
8
40
0.2 6
K
kg-
kJ/
0.85
0.90
0.94
0.98
1.02
1.06
1.10
1.14
4
1.18
1.34
1.26
1.22
1.30
1.38
1.62
1.42
1.74
1.82
1.46
1.50
1.78
1.86
1.58
1.66
1.70
8
1.54
1.94
4
1.9
2.0
2.0
2.1
2.1
0
3.2
1.9
0.1
20
S=
2.4
300
320
T = 180°C
280
260
240
uid
220
0
200
1.6
160
140
Liq
120
100
0.04 1.2
80
60
40
r
20
ted
−20
vapo
0.8
ura
Τ = 40°C
rated
Sat
0.02 0.6
.5
0.9
0.3
0.7
0.8
0.4
0.6
X=0
0.1
0.2
Satu
6
2.2
2.2
8
4
0.4
2.1
2.1
S=
0.01
160 200 240 280 320 360 400 440 480 520 560 600 640
Enthalpy (kJ/kg)
FIGURE 2.1.14 Pressure-enthalpy diagram of R123. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
/m 3
0 0 50
0 400
70 60
0
0
300
0 kg
0
0
80
100
90
110
1400
1300
1500
1600
240 R-125
120
180
4
65
150
55
120
45
96
35
2 72
25
60
3
48 kg/m
Density =
15
1 36
5
30
Pressure (MPa)
−5
24
−15
18
T = 25°C
15
0.4
12
3
9.6 kg/m
−35
S = 0.70 kj/kg-k
-K
7.2
0.2
J/kg
.30
0.80
0.75
0.85
0.90
1.00
0.95
6.0
5
1.40
1.20
1.45
1.50
1.55
1.25
5
1.05
1.10
.60
5
S=1
0
5
0k
0
0.8 1.3
.5 1.1
1.7
1.7
1.8
1.9
1.9
1.6
2.0
1
−45
1.8
4.8
apor
T = 115°C
175
165
or
S=
155
0.7
135
0.1
0.6
0.9
0.4
145
0.2
0.3
X=0
125
−55
vap
ated v
105
3.6
0.1
95
85
75
ted
65
T = 5°C
55
45
3.0
35
25
Satur
−65
ura
15
−5
−15
2.4
−25
Sat
−35
−45
−55
1.8
0
2.0
5
5
1.9
0
2.0
2.1
S=
0.04 1.2
0.03
120 160 200 240 280 320 360 400 440 480 520
Enthalpy (kJ/kg)
FIGURE 2.1.15 Pressure-enthalpy diagram of R125. (Copyright 2004, American Society of Heating, Refrigerating,
and Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
g/m 3
1500
1400
1300
0
1450
1350
300
0
0 0 400
00
50
600
00
500
125
0
70
0k
80
0
115
0
95
90
85
10
11
10
10
120
3
200 kg/m
Density =
160
120
100
4 90
80
70
R-134a 60
60
50
2
40
40
32
24
1
20
Pressure (MPa)
16
T = 0°C
12
0.4 8
6
−20
0.2 4
g-K
3.2
J/k
0.88
1.12
1.20
1.36
1.44
0.64
0.72
0.80
1.28
1.52
1.60
0.96
1.04
2.08
1.76
1.84
2.00
1.92
1.68
2.32
8
2.40
2.24
6k
2.4
2.4
2.1
0.1
S=
−40
1.6
1.2
320
300
por
280
260
T = 160°C
240
0.04
220
200
d va
0.8
vapor
180
−60
140
120
rate
100
0.6
80
d
60
Satu
Saturate
40
20
0
−20
0.02
−40
X = 0.5
0.4
0.4
0.9
0.8
0.3
0.6
0.2
0.1
0.7
0.3
6
2.5
8
2
4
2.4
2.7
2.6
S=
0.01
100 150 200 250 300 350 400 450 500 550 600 650 700 750
Enthalpy (kJ/kg)
FIGURE 2.1.16 Pressure-enthalpy diagram of R134a. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
240
8. 4 5 16 01 .00
01 01 0.00 0.0 0
230
0 0.003
0
0.0 0.0 0.004
0
3
105
011
220
012
0.001
01
0
6. 0.005 0
0.0
0.0
0.0
0.00
0.006
R-404A g
c.p. 80 m/k
4. O e = 0.00
Volum 0.010
60
60
50
0.015
40
50
2.
30
40 0.020
0.00095
0.00085
0.00090
0.00075
0.00080
30
20
0.030
20
10
1. 0.040
Temperature = 50 °C
10
10
40
30
20
0
70
60
0.8 0.050
Temperature = 0°C 0.060
0.6
Pressure (MPa)
−10 0.080
0.4 0.10
−20
0.15
uid
−30
liq
liquid
0.2 0.20
0.5
ted
Temperature = 140°C
200
0.6
0.7
0.9
210
0.1
0.2
0.4
0.8
190
0.3
180
170
160
150
ura
130
120
−40
rated
y=
110
100
Sat
0.30
80
90
70
60
alit
50
40
satu
30
0.1
Qu
20
10
0.40
−10
0
−50
−20
−30
−40
0.08
0.06 0.70
−60
0.04 1.0
-K
−70
kg
1.5
kJ/
.20
0.02 2.0
=2
1.90
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50
1.60
1.65
1.70
1.75
1.80
1.85
2.00
2.10
0
py
1.95
2.15
2.25
5
2.05
2.3
5
2.3
0
0
tro
2.4
2.5
2.4
En
0.01
100 150 200 250 300 350 400 450 500 550 600
Enthalpy (kJ/kg)
FIGURE 2.1.17 Pressure-enthalpy diagram of R404A. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
010
8.0
2
00 0. 02
0
80 0011
095
085
0 0
01
22
0. 5 0.0 .0030
090
0
0
0.0
0.0
21
2 0 40
0.00
1
0.00
0
0.
6.0 0.00
0.00
0
0
00 .
0. c.p. 0
20
0
0.005060
0
19
R-407C
70
O 0.0
3 /kg
4.0 080 m
e = 0.0
Volum
60
70 0.010
3.0
50
60
40
0.015
2.0 50
0.020
0.00070
0.00080
30
0.00075
40
20
30 0.030
1.0
10
20 0.040
0.8
Temperature−40 °C
0.050
−60
−50
−10
−30
−20
−70
10
0
0.060
0.6
Pressure (MPa)
Temperature = 0°C
0.080
0.4 −10
0.10
0.3 0.15
−20
uid
0.2 0.20
liq
−30
d
r
vapo
0.1
ate
0.2
0.3
0.30
0.4
tur
0.6
0.5
0.7
0.8
y=
0.9
rated
Sa
0.08
Qu
Satu
0.50
0.06 −50
0.70
Temperature = 120°C
180
170
160
150
140
130
.K 100110
90
0.04
80
70
1.0
60
50
−60
40
30
20
10
0.03
−10
−20
0
−30
−40
kg
−50
1.5
kj/
.35
0.02 −70 2.0
=2
py
0.80
tro
5
0.70
0.75
0.85
0.90
0
0.60
0
0.95
0.65
1.00
1.05
1.10
1.15
1.20
0
1.40
1.55
1.60
5
0
1.45
1.50
1.65
5
1.25
1.30
0
0
5
5
1.35
0
0
5
0
2.1
0
2.0
2.2
2.3
2.0
2.4
2.1
2.2
2.5
2.5
2.4
1.7
1.7
1.9
2.6
1.8
1.8
1.9
En
5
2.6
0.01
100 150 200 250 300 350 400 450 500 550 600
Enthalpy (kJ/kg)
FIGURE 2.1.18 Pressure-enthalpy diagram of R407C. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
10 0.040
1 1
0.050
0.8 Temperature = 0°C 0.8
re = 50°C
0.060
180
170
160
150
0.6 0.6
140
130
−10
120
110
100
90
80
70
0.10
Tempe tu60
40
−20
30
0.4 0.4
20
10
ra
−10
0
Pressure (MPa)
Pressure (MPa)
0.15
−30
0.20
0.2 0.2
−40
0.30
0.40
−50
id
g
0.1 0.50 m Y/k 0.1
liqu
0.5 Volume =
0.60
ted
0.08 0.08
ty =
ura
0.80
Quali
0.06 0.06
Sat
1.0
kg-K
2.0
0.02 −80 0.02
j/
3.0
1.8 k
2.2
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.9
2.0
2.1
2.4
2.5
20 2.3
2.6
4.0
py =
0.01 0.01
2.7
2.8
5.0
30
−90
40
Entro
50
6.0
60
0.008 70
80
0.008
0.006 8.0 0.006
100 150 200 250 300 350 400 450 500 550 600
Enthalpy (kJ/kg)
T= 0
T = 20°C
60
24
440
22
40
20
100
120
−20
180
−60
420
38
−40
36
80
14
160
60
0
34
32
0
28
40
0
0
26
30
40
0°C
0
0
0
0
0
0
10 30
Melting line
20
Pressure (MPa)
15
4.0 10
8
g -K
6
2.0 .0
kJ/k
7.0
S = 5.0
5.5
10.0
.5
6.0
6.5
7.5
8.0
8.5
9.5
4.5
11
9.0
.5
11 4
10
S=
1.0 12
.0 3
2
5
12. 1.5
0.4 1.0
0.8
0.6
0.2
uid
0.4 kg/m3
Vapor
Liq
0.1 0.3
420
T = 220°C
380
340
300
0.2
260
180
d
0.15
140
Saturated
rate
100
0.04
400
60
360
320
Satu
0.10
280
20
−20
240
T = 480°C
200
160
120
440
460
0.02
80
0.1
0.2
X = 0.5
0.3
.0
0.9
0.6
0.8
0.4
.5
0.7
40
13 14
0
.0
−40 13 S=
0.01
−1200 −1000 −800 −600 −400 −200 0 200 400 600 800 1000 1200 1400 1600 1800
Enthalpy (kJ/kg)
FIGURE 2.1.20 Pressure-enthalpy diagram of R717. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
50
00
50
500
00
90 0k
11
11
10
10
80
400
T=
420
400
380
360
80
200
280
100
340
160
−20
120
20
140
320
300
60
260
40
18
−40
220
240
0
40 300
°C
250
200
Melting line
K
g-
20 150
/k
S = 3.0
kJ
2.8
4.0
3.9
3.8
3.7
4.3
3.4
3.6
4.2
4.5
4.1
4.4
3.2
S=
4.6
100
4.7
10 Carbon dioxide 80
Pressure (MPa)
R-774 60
3
g/m
= 50 k
Density
40
4.0 30
25
20
420
380
2.0
340
300
15
260
T = 220 °C
400
180
360
320
140
280
Saturated vapor
10
240
100
200
d
rate
60
160
8.
1.0
120
20
Satu
5.2
−20
6. kg/m
80
5.1
X = 0.5
40
0.9
0.2
0.3
0.7
0.8
0.1
0.4
0.6
5.0
0
4.9
S=
−40
4.8
4.7
Triple point
0.4
300 400 500 600 700 800 900 1000 1100 1200 1300
Enthalpy (kJ/kg)
FIGURE 2.1.21 Pressure-enthalpy diagram of R774. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
0
70
0
kg/ 400 300
0
0
65
60
55
35
50
40 450
250
200
Melting line
20 150
70
T = 180 K
68
0
66
320
360
260
280
300
64
220
340
140
160
200
240
0
62
38
40
58
46
T=
42
56
60
44
54
48
100
52
0
0
0
0
0
0
0
0
0
0
0
50
0
10
0K
70
50
PROPANE 40
3
4. R-290 Density=30 kg
/m
20
2. 15
-K
Pressure (MPa)
kg
S= 4.0
kJ/
4.8
4.6
5.0
3.6
5.8
5.2
10
4.4
5.4
5.6
6.0
6.4
7.0
6.2
6.8
7.2
6.6
7.4
1. 7.
S=
7.6
5.
4.
0.4 3.
2.
0.2 1.5
1.0
.5
0.1
0.2
0.6
0.7
0.3
0.8
0.4
d vapor 0.9
X=0
uid
0.1 0.7
d liq
0.5
rate
0.4 kg/m3
Satu
Saturate
380
0.04 0.3
340
300
700
680
660
640
7.6 T=500 K
620
600
580
260
560
540
0.2
520
480
220
460
440
420
0.02
400
360
0
320
8.2
8.
7.8
280
=
240
S
0.01
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Enthalpy (kJ/kg)
FIGURE 2.1.22 Pressure-enthalpy diagram of R290. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
0
0
0
45 400 350 300
55
60
50
250
200
10 Normal butane 150
R-600
100
70
420
4.0 50
400
40
30
380
2.0
T = 340 K
360
20
15
Pressure (MPa)
1.0 3
Density = 10 kg/m
320
7.0
g-K
3.0
kJ/k
0.2
280
4.4
4.8
5.0
5.2
5.4
5.6
4.2
6.4
6.6
6.0
6.2
2.0
4.6
s = 4.0
5.8
S=
1.5
0.1
uid
r
Vapo
1.0
L iq
260 0.7
0.04
ated
d
0.5
rate
T = 440 K
Satur
420
0.4
Satu
560
580
380
600
540
520
340
500
480
460
300
240 0.3
400
260
360
X = 0.5
0.02 320
0.1
0.3
0.6
0.7
0.8
0.9
0.2
0.4
7.0
280
6.8
240
S=
0.01
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
Enthalpy (kJ/kg)
FIGURE 2.1.23 Pressure-enthalpy diagram of R600. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
0
250
0
350 300
0
Isobutane 45 400
60
55
50
R-600a 200
10 150
100
70
4.0
400
50
40
380
30
2.0
360
3
Density = 20 kg/m
340
15
Pressure (MPa)
1.0
320
T = 300 K
10
7.0
0.4 5.0
4.0
280
-K
J/kg
3.0
0.2
k
3.8
4.0
5.2
4.8
4.6
4.2
5.0
5.6
4.4
5.4
6.4
5.8
6.0
6.2
2.0
260
s = 3.6
S=
1.5
0.1 3
1.0 kg/m
uid
or 0.7
Liq
ted vap
600
580
560
380
540
520
T = 440 K
Satura
340
500
480
460
Satur
0.3
300
420
400
0.02
260
X = 0.5
360
320
6.6
0.1
0.3
0.6
0.7
0.8
0.9
0.2
0.4
280
S=
240
0.01
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
Enthalpy (kJ/kg)
FIGURE 2.1.24 Pressure-enthalpy diagram of R600a. (Copyright 2004, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
ln(P)
Isotherm
Critical point
e
op
ntr
I se
Isochore
Subcooled Superheated
liquid vapor region
region
Line of
constant
2-
2-phase range region
(vapor dome) vapor quality
ln(P)
hlatent h
ln(P)
Isotherm
Two-phase region
FIGURE 2.1.27 Various optional shapes for isotherms in the two-phase region of an ln(P)-h
diagram.
At first glance, the pressure-enthalpy diagrams for zeotropic mixtures have the
same general topology as those for pure fluids. However, there are two important
differences (Figure 2.1.18). R407C is chosen here as an example although it is a
ternary mixture. However, the characteristics of multicomponent mixtures and binary
mixtures in the ln(P)-h diagram are the same.
The first difference between the diagrams of pure fluids and mixtures is that the
zeotropic diagram is valid only for a given mass fraction. For a different mass
fraction, all the lines on this diagram shift. Thus, it is only possible to show one
diagram at a time with a full set of lines of constant properties. If one attempted to
superimpose a second diagram for a different mass fraction, the graphical represen-
tation would become incomprehensible. Only one mass fraction is typically circu-
lating in a vapor-compression cycle.
Second, due to the temperature glide in the two-phase region, the isotherms in
the two-phase region have a negative slope. Depending on the nature of the pure
fluids that constitute the mixture, the isotherm in the two-phase region may be
concave, convex, s-shaped or a straight line. The schematic in Figure 2.1.27 shows
examples of these isotherms. When one considers a constant pressure evaporation
process that commences with saturated liquid and ends with saturated vapor, it
becomes apparent that the beginning of the evaporation process starts at a given
isotherm and ends at an isotherm of higher temperature. Thus, the temperature glide
is correctly represented.
r
Isoba
Critical point
Subcooled
liquid
region Superheated
vapor region
2-
2-phase range region
(vapor dome)
Isenth
Line of a lp
constant
vapor quality
s
range (vapor dome) has a similar shape compared to the mixture diagram. Again, the
subcooled liquid range is located to the lower entropy (left) side of the vapor dome,
the superheated vapor region is to the right of the vapor dome and the lines constituting
the vapor dome are the saturated liquid and saturated vapor lines. They culminate in
the critical point on top of the vapor dome. The isobar is horizontal in the two-phase
region and ascends from the saturated vapor line with increasing slope into the super-
heated vapor region. From the saturated liquid line reaching into the subcooled liquid
region, the isobar drops very quickly and hugs the saturated liquid line, following it
much more closely than shown in Figure 2.1.28, especially when pressures are low.
In a T-s diagram for a zeotropic mixture, all lines are similar to the pure fluid
with the exception of the isobars. To correctly represent the temperature glide, the
isobars must show a positive slope in the two-phase region. Depending on the nature
of the pure fluids that constitute the mixture, the isotherm in the two-phase region
may be concave, convex, s-shaped or a straight line. The schematic in Figure 2.1.29
shows examples of these isotherms.
When considering a constant pressure evaporation process that commences with
saturated liquid and ends with saturated vapor, it becomes apparent that the beginning
of the evaporation process starts at a given temperature and ends at a higher tem-
perature on the same isobar.
r
Isoba
T
FIGURE 2.1.29 Various possible shapes of isobars in the two-phase range of a T-s diagram
for zeotropic mixtures.
means a finite specific heat in the two-phase range. Moreover, the slope varies with
respect to enthalpy; this means a variable specific heat resulting from a varying latent
heat of vaporization due to the changing liquid and vapor mass fractions of constit-
uents. The superheated region forms a narrow band as compared to other diagrams.
When the refrigeration cycle is superimposed together with temperature profiles
of secondary fluid heat source and sink on the T-h diagram, the benefit of zeotropic
mixtures is demonstrated graphically. Figure 2.1.31 shows the T-h diagram of a cycle
with heat exchange operating 400 kPa evaporation and 1600 kPa condensing pres-
sures for pure R22 and the zeotropic mixture R22/R114 (50/50 wt.%). In Figure
2.1.31, the cycle ABCD represents the refrigeration cycle (A–B: evaporation, B–C:
compression, C–D: condensation, D–A: isenthalpic expansion) and the lines Ae–Be
and Cc–Dc represent temperature profiles of heat source and heat sink. When the
temperature difference between the refrigerant and secondary fluid are compared
for two refrigerants, as shown in Figure 2.1.31, advantages of zeotropic mixtures
become distinct. For R22, the saturated vapor temperature during its condensing
process is closest to the secondary fluid temperature, showing a potential heat
exchange pinch-point. Moreover, R22 shows nonuniform approach temperatures
during its evaporation process, whereas R22/R114 shows almost uniform and smaller
approach temperatures than those of R22 during its evaporation and condensation
processes. Therefore, approach temperatures at the inlet and outlet of both R22 heat
exchangers are larger than those of R22/R114. Large approach temperatures mean
higher irreversible losses associated with heat exchange. Therefore zeotropic mix-
tures having temperature glides that are properly matched with the secondary fluids
have less irreversible losses than those of pure refrigerants.
100 100
a
saturation 2600 kP
Temperature (°C)
Temperature (°C)
2600 kPa a
incr 1600 kP
50 1600 kPa entr easing 50 1200 kPa
opy
saturation
1200 kPa 800 kPa
800 kPa
400 kPa
0 400 kPa 0
200 kPa
200 kPa
100 kPa
100 kPa
−50 −50
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Enthalpy (kJ/kg) Enthalpy (kJ/kg)
FIGURE 2.1.30 Temperature-enthalpy diagram (Granyrd et al., 1991). (a) R22, (b) R22/R114 (50/50 wt.%).
100 100
C
Temperature (°C)
Temperature (°C)
1600 kPa
Cc
50 1600 kPa 50 D
D Cc Bc
Dc
Dc Bc B
Ac 0 Ac 400 kPa
0 A
A 400 kPa B
−50 −50
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Enthalpy (kJ/kg) Enthalpy (kJ/kg)
(a) (b)
FIGURE 2.1.31 Temperature-enthalpy diagram of cycle with heat exchange (Granyrd et al., 1991). (a) R22, (b) R22/R114 (50/50 wt. %)
Temperature = Constant
Bubble line
Pressure
Subcooled liquid
Two-phase range
Superheated vapor
Dew line
Pure Pure
Fluid A Fluid B
Mass Fraction of B
beginning of this chapter. In principle, both diagrams are very similar except for the
important difference that the general outline is reversed.
Figure 2.1.32 shows a schematic of a P-x diagram. The subcooled liquid region
is now located above the two-phase region and the superheated vapor region is below.
Thus, all processes shown in the diagram are reversed compared to the T-x diagram.
Because this reversal can be quite confusing, it is recommended that readers pick
either the temperature-mass fraction diagram or the pressure-mass fraction diagram
as their favorite representation of mixture properties and then remain consistent. In
this text, the temperature-mass fraction diagram is used exclusively.
2.2.3 SOLUBILITY
Solubility is defined as the amount of a substance that dissolves in a specific amount
of the dissolving agent. Two other terms that can describe the solubility are miscible
and immiscible. Two gases form a homogeneous mixture when they are mixed.
Similarly, many liquids, when suitable, are mixed and form a homogeneous mixture;
these fluid pairs are miscible. Liquids that form an inhomogeneous mixture and do
not dissolve are termed immiscible. Substances with the same polarity tend to be
miscible and substances with opposite polarity tend to be immiscible. Good examples
of miscible and immiscible refrigerant mixtures and lubricant pairs are shown in
Table 2.3.
The miscibility depends on the substances that are mixed, the mass fraction and
temperature. Figure 2.2.1 shows the miscibility of refrigerant R134a in a specific
polyol-ester lubricant. The top of the curve point C indicates the minimum temper-
ature at which the refrigerant and lubricant are miscible in all mass fractions. In the
area to the left of the curve, the mixture is refrigerant rich and miscible, while in
the area right of the curve, the mixture is lubricant rich and miscible. In the area
below the curve the refrigerant and lubricant are present in two liquid phases. Each
phase is a homogeneous mixture of refrigerant and oil but of different mass fractions.
The area enclosed by the curve through A, B and C is termed the mixing gap. Their
TABLE 2.1
Types of Concentration for Binary Mixtures
Types of
Concentration Definition Physical Meaning
Note: M, m and V stand for the molecular mass, mass and volume, respectively.
mass fractions of two liquid phases are given for a certain temperature by the
horizontal limits of the curve for example points A and B at –60˚C. Thus, at –60˚C,
R134a and polyol-ester lubricant are miscible when the lubricant mass fraction is
either less than 4%, point A, or higher than 54%, point B. However, they are not
miscible when the lubricant mass fraction is between the mass fractions of points
A and B.
x
mB x M Ax ρx
x= x MB ρx
m A + mB (1 − x ) + x 1− x M B (1 − x ) MB
MA
Ψ ρΨ
mB M B M BΨ Ψ ρΨ
Ψ= MA Ψ MA
m A M A + mB M B (1 − Ψ ) + Ψ M A (1 − Ψ ) 1− Ψ (1 − Ψ ) + Ψ M A (1 − Ψ ) + M B Ψ
MB MB
β
mB β
β
M Aβ ρβ ρβ M B
β= MB
mA 1+ β +β MB 1+ β 1+ β
MA
χ ρχ ρχ M B
mB M B χ M Bχ
χ= MA χ MA MA
mA M A +χ 1+ χ MA +χ +χ
MB MB MB
mB M Ac c M Aυc c
c= υc c
V ρM B + ( M A − M B )c ρ−c M B (1 − υc) MB
mB M B M Acm M B cm M Acm
c = MBVCm MBcm cm
m V ρ + ( M A − M B ) cm ρ − M B cm ρ − M B cm
TABLE 2.3
Solubility of Refrigerant and Lubricant
Refrigerant Lubricant Solubility
80
60
40
Temperature [°C]
20
0
Homogeneous Mixture
−20 C
−40
A B
−60
Heterogeneous Mixture
−80
0 10 20 30 40 50 60 70 80 90 100
Lubricant Concentration [wt.%]
T1 T1 T2 = T1
mA kg m B kg mM kg
vA vB v 2 = ( v A+ v B)
(a) (b)
FIGURE 2.2.2 Liquid mixing phenomena. (a) Two liquids in one container, not mixed. (b)
Liquids are mixed. The total volume is reduced in this example.
Before mixing:
m1 = m A + m B = m M (2.5)
v1 = V1 m1 = (m AvA + mB vB ) mM (2.6)
h1 = (m AhA + m B hB ) m M (2.7)
150
50
Methyl cyclopentane
0
3-Methyl pentane
−100
2,3-Dimethyl butane
−150
0.0 0.2 0.4 0 .6 0.8 1 .0
Mass Fraction [kg/kg]
FIGURE 2.2.3 Volume changes at 25˚C for liquid mixtures (Cyclohexane with some other
C6 hydrocarbons). (Perry et al., 1997.)
s1 = (m As A + mB s B ) mM (2.8)
After mixing:
m2 = m1 = 1 (for the sake of simplicity, the total mass is set to 1 kg) (2.9)
v2 ≠ v1 (2.10)
h2 ≠ h1 (2.11)
s2 ≠ s1 (2.12)
The volume change in the mixing of liquids can be measured easily and is
usually small. Figure 2.2.3 shows the volume changes of liquid mixtures, cyclohex-
ane with some other C6 hydrocarbons, at 25˚C. As shown in Figure 2.2.3, the changes
of volume can be either positive or negative depending on the individual mixtures
that are sensitive to the effects of molecular size and shape and to the differences
in the nature and magnitude of intermolecular forces.
In general, the temperature of a mixture is different from the initial temperature
of its components when the mixing process is adiabatic. During the mixing process,
the mixture either rejects or absorbs energy called heat of mixing. When the heat of
mixing is positive, the temperature of the mixture decreases. On the other hand,
when the heat of mixing is negative, the temperature of the mixture increases. Figure
2.2.4 shows the heat of mixing for the H2O/NH3 mixture at 2 MPa and 40˚C. The
250
200
h1 h NH3
hH2O
150
Enthalpy [kJ/kg]
Heat of mixing
100
50
−50 h2
−100
0.0 0.2 0.4 0.6 0.8 1.0
Mass Fraction of NH3 [kg/kg]
FIGURE 2.2.4 Heat of mixing in liquid mixture (H2O/NH3 mixture at 2 MPa, 40˚C).
heat of mixing for the given mixture changes depending on the temperature and
pressure. The heat of mixing for the refrigerants that are the focus of this text is
quite small.
Mixture A Mixture B
mA kg/s mB kg/s
hB
TA, hA, vA, xA TB, hB, vB, xB mM
mA
Enthalpy
Mixing
mB hM
Mixture M
hA
mM kg/s
TM, hM, vM, xM
0 xA xM xB 1.0
Mass Fraction
(a) Mixing of Two Mixtures (b) Mixing Rule
of mixture B, are mixed adiabatically and form m M of the product mixture stream.
From the mass and energy balances, three conservation equations result:
m M = m A + m B (2.13)
Conservation of species:
m M x M = m A x A + m B x B = m M x A + m B ( x B − x A ) (2.14)
Conservation of energy:
From Equations 2.13, 2.14 and 2.15, the ratio m A /m B can be expressed in terms
of mass fractions or enthalpies as shown in Equation 2.16,
m A x B − x M hB − hM
= = (2.16)
m B x M − x A hM − hA
The mass fraction (xM) and enthalpy (hM) of the mixture after mixing can be re-
expressed as Equations 2.17 and 2.18 by using the ratio m B /m M ,
m B
xM = xA + (xB − x A ) (2.17)
m M
m B
hM = hA + ( hB − h A ) (2.18)
m M
As Equations 2.17 and 2.18 show, the mass fraction (xM) and enthalpy (hM) of the
product mixture after mixing are a linear function of the two mixtures A and B and
can be obtained by interpolation as shown in Figure 2.2.5(b). In an enthalpy-mass
fraction diagram, Equations 2.17 and 2.18 describe the so-called mixing rule.
Similarly, the temperature of the product stream can be found from the enthalpy
(h) and mass fraction (x) chart as shown in Figure 2.2.6. For example, two streams
of saturated liquid R32/R134a, stream state A at m = 2/3 kg/s, x = 0.11 kg/kg, T =
30˚C, and stream state B at m = 1/3 kg/s, x = 0.75 kg/kg, T = 10˚C, are adiabatically
mixed with each other. The state of mixture M can then be found from the mixing
rule. The mass fraction and enthalpy of M can be calculated from Equations 2.14
and 2.15.
600
P = Const. (1 MPa)
500
t line
Dewpoin
MV
Enthalpy [kJ/kg]
400
30°C 20°C 10°C
300
A Boiling point line
M
B
ML
200
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mass Fraction of R32 [kg/kg]
m B
xM = xA + ( x B − x A ) = 0.11 + 1/3(0.75 − 0.11) = 0.3223 [kg/kg]
m M
m B
hM = hA + (hB − hA ) = 218.0 + 1/3(244.1 − 218.0) = 226.7 [kJ/kg]
m M
The state of M can also be obtained from the h-x diagram by interpolation as shown
in Figure 2.2.5(b). In Figure 2.2.6, states A and B are connected by a straight line.
Then the state point M of the product stream is found at the location where the line
AB is divided in the ratio of AM:MB = 1:2. The temperature of the state point M
is found from the isotherm TM that passes through point M. As can be seen in Figure
2.2.6, the state point M is above the saturated liquid line. Therefore, the state M is
in the two-phase region.
Before mixing:
h1 = (m AhA + m B hB ) m M (2.19)
After mixing:
h2 = h1 + ∆h (2.20)
Heat of mixing:
m AhA + m B hB − q m (h − hA ) − q
hM = = hA + B B (2.23)
mM m M
When the initial temperatures of the two mixtures are the same before mixing and
maintained constant during the mixing process, the process is called isothermal. To
be an isothermal mixing process, the mixture needs to be heated or cooled to maintain
the initial temperature.
m AhA + m B hB
h = ∆h + = ∆h + (1 − x ) hA + xhB (2.24)
m M
The specific heat of the binary mixture can then be obtained by taking the derivative
of enthalpy with respect to the temperature,
∂h ∂∆h ∂h ∂h
cp = = + (1 − x ) A + x B (2.25)
∂T P ,x ∂T P ,x ∂T P ,x ∂T P ,x
From the definition of the specific heat of the individual components, Equation 2.25
becomes
∂∆h
cp = + (1 − x ) c p, A + xc p,B (2.26)
∂T P ,x
600
P = Const. (1 MPa)
500
Dewpoint line MV
Enthalpy [kJ/kg]
400
20.3°C 15°C
300
200
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mass Fraction of R32 [kg/kg]
from the intersection of the vertical mass fraction line and the saturated liquid and
vapor lines. The saturated liquid enthalpy is then 225.5 kJ/kg at 15˚C and the
saturated vapor enthalpy is 466.2 kJ/kg at 20.3˚C. Therefore, the latent heat of the
R32/R134a mixture at 1 MPa is 240.7 kJ/kg. In this way, the latent heat of the binary
mixture at different pressures and concentrations can be obtained from the h-x
diagram.
∂P
∂V =0 (2.27)
T =T
c
∂2 P
∂V 2 =0 (2.28)
T =T
c
However, the mixtures have different characteristics than pure substances. The
critical point of mixtures is not the inflection point. Critical point, critical temperature
and critical pressure of mixtures are all different. The critical point of mixtures is
defined as the point at which the dew line and bubble line meet. Figure 2.2.8 shows
the critical envelope of the binary mixture in the pressure-mass fraction diagram.
PCA and PCB represent the critical points of substances A and B, respectively. Critical
envelopes of maximum pressure and temperature are formed while connecting the
two critical points, PCA and PCB. When the pressure of a binary mixture is lower
than the critical pressure of either two substances, the dew and bubble lines at the
same temperature meet at the mass fractions of zero and one. This is the case for
temperatures T1 and T2 in Figure 2.2.8. When the pressure of a binary mixture is
between critical pressures of the two pure substances, the dew line and bubble lines
at the same temperature meet at the mass fraction of zero and less than one. For the
temperature T4, the dew and bubble lines meet at the mass fraction of 0 and x4
(Figure 2.2.8). The liquid and vapor phases at the same temperature and pressure
have the same mass fraction x4 at the critical point CP4. If the temperature is slightly
increased from point CP4, while the mass fraction and pressure are maintained such
that the temperature of the new critical point does not cross the critical envelope of
Properties of Working Fluids 59
Critic
PC,A al En
velop
Critica e
l Enve P4,c Supercritical
lope
region
CP4
T4 T4,c
T4 P3,c CP
p ,m
P4,A CP3 ax
T3 T3,cCPt,ma
x
T3 PC,B
Pressure
P2,B
P3,A st.
= con
for T
2
Line st.
Dew T 2 = con
e for
le Lin
Bubb P1,B
P2,A
onst.
T1 = c
ine for onst.
Dew
L 1=c
ine for T
le L
Bubb
P1,A
0 x4 1.0
A Mass Fraction B
maximum pressure, the mixture is still in the two-phase region and not in the
supercritical region. Therefore, the state of a mixture can be changed from the
subcritical region to the supercritical region only when the pressure and temperature
increase substantially such that the state point crosses the critical envelope of max-
imum pressure.
It should be noted that the condition where the second and third derivatives of
molar Gibbs free energy with respect to the mixture composition are both zero is
used in the REFPROP as the thermodynamic criteria for the critical loci of binary
mixtures. Pseudocritical parameters are then fitted to empirical functions from the
iteratively calculated critical loci (McLinden, 1998).
R32 R143a
66% flammable
60% 25°C 58% flammable
55%
100 60°C
R410A °C 51.5% R507A
R404A 10
0°C 36%
33% 33%
R407C
nonflammable
nonflammable
R125 R134a R125 R134a
(a) (b)
FIGURE 2.3.1 Flammability diagram for two ternary mixtures. (a) R407C. (b) R404A.
Properties of Working Fluids 61
TABLE 2.4
Thermophysical Properties of Pure Refrigerants
Property R32 R125 R134a R143a
Chemical formula CH2F2 CHF2CF3 CH2FCF3 CF3CH3
Critical temperature [˚C] 78.1 66.2 101.1 72.9
Latent heata [kJ/kg] 299.0 124.7 190.7 177.7
Specific heata [kJ/kg*K] 1.806/1.367 1.310/0.958 1.370/0.946 1.551/1.164
(liquid/vapor)
LFL to HFLb [%] 12.7 ~ 33.5 None None 7.1 ~ 16.1c
Molecular weight [kg/kmol] 52.0 120.0 102.0 84.0
Thermal conductivitya 0.147/0.0129 0.0663/0.0136 0.0876/0.0124 0.0765/0.0144
[W/m*K] (liquid/vapor)
Note: NIST REFPROP V6.0 (McLinden et al., 1998) unless otherwise noted.
a at 10˚C.
b Richard and Shankland (1992).
c Atofina Chemical (1999).
62 Vapor Compression Heat Pumps with Refrigerant Mixtures
TABLE 2.5
Environmental and Thermophysical Properties of Binary and Ternary Mixtures
Property R410A R507A R407C R404A
Components R32/125 R125/143a R32/125/134a R125/143a/134a
(50/50 wt.%) (50/50 wt.%) (23/25/52 wt.%) (44/52/4 wt.%)
ODP (R11 = 1.0) 0 0 0 0
GWP (CO2 = 1.0) 1890d HGWP 1.0d 1370b 3260c
Critical temperature [˚C] 70.2 70.8 86.1 72.1
Latent heata [kJ/kg] 208.4 152.2 199.2 156.0
Specific heata [kJ/kg*K] 1.565/1.207 1.427/1.057 1.449/1.009 1.434/1.057
(liquid/vapor)
Molecular weight [kg/kmol] 72.6 98.9 86.2 97.6
Thermal conductivitya 0.108/0.0129 0.0726/0.0141 0.0956/0.0121 0.0731/0.0140
[W/m*K] (liquid/vapor)
Note: NIST REFPROP V6.0 (McLinden et al., 1998) unless otherwise noted.
a At 10˚C.
b Calm (1995).
c Siva and Gustavo (1999).
d DuPont Fluorochemicals (2000).
well in heat pumps for all applications. Table 2.5 summarizes the environmental and
thermophysical properties of the selected binary and ternary mixtures as most prom-
ising alternatives of R22 and R502. Moreover, constituents of these mixtures are
two or three of R32, R125, R134a and R143a.
The properties of the new refrigerants should resemble as close as possible those
of the fluids being replaced. In particular, for R22 no such candidate has been found.
However, based on previous investigations that led to the introduction of R22, it was
known that R32 is the most desirable fluid. It has the highest efficiency of the pure
alternatives to R22 and, because of its high-pressure levels, also a very high volu-
metric capacity. Its only major shortcomings are its moderate flammability, the high
compressor discharge temperature (a consequence of R32’s low specific heat capac-
ity) and the high-pressure levels. The latter can be seen as an asset or challenge,
depending on the point of view.
The refrigerant mixtures R410A and R407C are compromises that aim at main-
taining the favorable characteristics of R32, especially its high efficiency, while
eliminating its shortcomings. R407C aims at achieving a fluid with properties resem-
bling R22 as closely as possible, the only major difference being that R407C has a
temperature glide. R410A, however, has almost no glide and it maintains the very
high-pressure levels of R32.
R410A consists of R32/R125, 50/50 wt.%. R125 has almost identical charac-
teristics to R32, especially the same pressure levels (Figure 2.1.8) but with a much
higher specific heat capacity and it is not flammable. As a pure fluid, R125 performs
quite poorly as refrigerant and the coefficient of performance (COP) is quite low.
Properties of Working Fluids 63
This is a consequence of the very low critical point of R125, while that of R32 is
rather high as compared to other refrigerants (Figure 2.1.8). The specific heat
capacity of R125 is rather high and thus it contributes to a reduced compressor
discharge gas temperature as compared to R32. Thus, R410A is nonflammable, has
an acceptable compressor discharge temperature, a high volumetric capacity, good
efficiency and a critical point that is lower than desirable. Under extreme air-
conditioning situations, the critical point can be reached or exceeded.
R407C consists of R32/R125/R134a, 23/25/52 wt.%. With the addition of R134a,
the vapor pressure is reduced and essentially matches that of R22. The mixture is not
flammable, the compressor discharge temperature is considerably lower than that of
R32 and the performance is acceptable. The critical point is higher than that of
R410A and it is not prone to be reached under extreme air-conditioning applications.
R507A consists of R125/R143a, 50/50 wt.%. This mixture is designed as a long-
term substitute for R502, a CFC-containing azeotrope. R404A consists of
R125/R143a/R134a, 44/52/4 wt.%. With the addition of R134a, the suction vapor
density matches that of R502. This mixture is designed to retrofit R502.
REFERENCES
Alefeld, G. and R. Radermacher, 1994, Heat Conversion Systems, CRC Press, Boca Raton, FL.
Allied Signal Chemicals, 1999, Genetron Products Brochure.
American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc., 1997,
ASHRAE Handbook — Fundamentals, pp. 19.8–19.60.
Atofina Chemical, 1999, Website: http://www.atofinachemicals.com/newelf/fluorochem/pdf/
6049.pdf
Bosnjakovic, F., 1965, Technical Thermodynamics, Holt, Rinehart & Winston, New York, NY.
Calm, J.M., 1995, Refrigerant Database, ARTI Report, DOE/CE23810-59C.
Downing, R.C., 1988, Fluorocarbon Refrigerants Handbook, Prentice Hall, Englewood Cliffs,
NJ.
DuPont Fluorochemicals, 2000, “Technical Information for Suva Refrigerants,” Website:
http://www.dupont.com/suva/na/usa/literature.
Fairchild, P.D., 2000, Personal communication.
Granyrd, E., J.C. Conklin and J.R. Sand, 1991, “Advantage of Enthalpy-Temperature Dia-
grams for Nonazeotropic Refrigerant Mixtures,” Proceedings of XVIIIth Congress
of Refrigeration, Montreal, Quebec, Canada.
Godwin, D.S., 1994, “Results of Soft-Optimized System Tests in ARI’s R-22 Alternative
Refrigerants Evaluation Program,” Proceedings of The 1994 International Refriger-
ation Conference, West Lafayette, IN, pp. 7–12.
Herold, K.E., R. Radermacher and S.A. Klein, 1996, Absorption Chillers and Heat Pumps,
CRC Press, Boca Raton, FL.
Hughes, M., 1997, “Contemporary Fluorocarbons, Refrigerant for the 21 Century,” Proceed-
ings of ASHRAE/NIST Refrigerants Conference, pp. 117–121.
Hwang, Y., J.F. Judge and R. Radermacher, 1995, “An Experimental Evaluation on Medium
and High Pressure HFC Replacements for R-22,” Proceedings of The 1995 Interna-
tional CFC and Halon Alternatives Conference, Washington D.C., pp. 41–48.
64 Vapor Compression Heat Pumps with Refrigerant Mixtures
McLinden, M.O., S.A. Klein, E.W. Lemmon and A.P. Peskin, 1998, “REFPROP, NIST
Thermodynamic Properties of Refrigerants and Refrigerant Mixtures Database,”
Version 6.0, Thermophysics Division of National Institutes of Standards and Tech-
nology, Gaithersburg, MD.
Perry, H.R., D.W. Green and J.O. Maloney, 1997, Perry’s Chemical Engineers’ Handbook
(7th edition), McGraw-Hill, New York, NY, pp. 4–12.
Radermacher, R. and D.S. Jung, 1992, “Theoretical Analysis of Replacement Refrigerants
for R-22 for Residential Uses,” U.S. Environmental Protection Agency Report, U.S.
EPA/400/1-91/041.
Richard, R. and L. Shankland, 1992, “Flammability of Alternative Refrigerants,” ASHRAE
Journal, Vol. 24, No. 4, pp. 20–24.
Sami, S.M., 1994, “A New Alternative HFC Blend for HCFC-22 and CFC-502,” Proceedings
of the 1994 International CFC and Halon Alternatives Conference, Washington D.C.,
pp. 56–64.
Siva, G. and D.R. Gustavo, 1999, “Total Equivalent Warming Impact of R22 Alternatives in
Air-Conditioning and Heat Pump Application,” ASHRAE Transactions, Vol. 105, Pt.
1, pp. 1228–1236.
Underwriters Laboratories Inc., 1994, UL Standard 2182, Refrigerants.
Vineyard, E.A., J.R. Sand and T.G. Statt, 1989, “Selection of Ozone-Safe, Nonazeotropic
Refrigerant Mixture for Capacity Modulation in Residential Heat Pumps,” ASHRAE
Transactions, Vol. 95, Pt. 1, pp. 34–46.
3 Vapor Compression
Cycle Fundamentals
The vapor compression cycle is the underlying thermodynamic process for most
refrigeration systems in stationary and many mobile applications. This cycle is also
referred to as the reverse Rankine cycle or the Perkins/Elmer cycle. For this text,
the term vapor compression cycle is used throughout. As a general introduction and
to establish the connection to the material taught in basic thermodynamics, the
Carnot cycle will be explained first and then the vapor compression cycle will be
discussed in full detail. This chapter also provides a description of the operation of
heat pumps within the larger context of energy conversion systems.
The signs of the quantities in Equation 3.1 are to be selected as indicated in Figure
3.1.1. All energies are counted as positive in the direction of the arrows. The
efficiency for power generation is defined as the amount of work produced divided
by the amount of heat supplied at the high temperature,
W
η= (3.2)
Q2
65
66 Vapor Compression Heat Pumps with Refrigerant Mixtures
Q2
T
A B
T2
Winput Woutput
T1
D C
Q
1
E F
0 s
FIGURE 3.1.1 The Carnot cycle for power generation on a temperature-entropy diagram.
For the Carnot cycle of Figure 3.1.1, the Second Law of Thermodynamics states
that for reversible operation the net entropy production is zero, so that
Q2 Q1
− =0 (3.3)
T 2 T1
Equation 3.2 can be modified to an expression that contains only temperatures using
Equations 3.1 and 3.3 by eliminating W,
T2 − T1
η= (3.4)
T2
The expression in Equation 3.4 is frequently termed the Carnot efficiency factor for
power generation.
Figure 3.1.2 shows a Carnot cycle that is operated as a heat pump cycle. The
direction of all processes is reversed compared to the power generation cycle. Here,
the temperatures are chosen such that heat Q0 is added to the working fluid at T0
along the process line GH, the fluid is compressed isentropically, HI, heat Q1 is
rejected at T1, IJ, and the fluid is expanded isentropically, JG. Again, the areas
represent energy transfers for reversible processes. The net amount of work input
required for this cycle is represented by the area GHIJ and the amount of heat
absorbed by the area GHKL. The sum of both areas (area IJLK) represents the
amount of heat rejected at T1.
The performance of a heat pump is described by the ratio of the benefit obtained —
the amount of heat available at the high temperature — divided by the expenditure,
the net work requirement. Since this value is always greater than 1.0, the term
coefficient of performance (COP) is customary,
Vapor Compression Cycle Fundamentals 67
Q1
J I
T1
Woutput Winput
T0
G H
Q0
L K
0 s
FIGURE 3.1.2 The Carnot cycle for heat pumping on a temperature-entropy diagram.
T1
COP = (3.5)
T1 − T0
This equation is derived by using the First and Second Laws for the heat pump
process of Figure 3.1.2 and by following the same procedure as for Equation 3.4.
The Carnot cycle of Figure 3.1.2 can also be used for cooling or refrigeration
applications. While the cycle itself does not change, the way we view the application
is different. For refrigeration, the heat removed at T0 is of interest and the COP is
defined as the ratio of the cooling capacity Q0 over the work input. Subscript R is
used to designate that the COP is defined for refrigeration,
T0
COPR = (3.6)
T1 − T0
Equation 3.6 is derived by applying the First and Second Laws as described above
for the power generation cycle. For the coefficients of performance of a heat pump
cycle and refrigeration cycle, the following relationship can be derived. This expres-
sion can be verified by substituting Equations 3.5 and 3.6 into Equation 3.7,
Q2
T
A B
T2
W
Q”1
J I
T1
D C
Q’1
T0
G H
F Q0
E L K
0 s
FIGURE 3.2.1 Carnot cycles for a combined power-generation/heat pumping facility such
as an engine driven or absorption heat pump.
cycle of Figures 3.1.1 and 3.1.2) are considered together. It is assumed that the
amount of work produced by the first cycle, Figure 3.1.1, is identical to the amount
of work required by the second cycle, Figure 3.1.2. The new device, Figure 3.2.1,
raises the temperature level of heat supplied at T0 to T1 by using the thermodynamic
availability of the high temperature energy supplied at T2. The waste heat of the
power generation portion of this combined cycle is also rejected at T1. Thus the total
amount of heat rejected at T1 has two contributions, Q′1 and Q″1 .
This combined cycle represents a heat pumping device that is driven by the input
of heat only. This model is a reflection of the fact that most of the power available
today is generated in a heat engine. Although this text focuses on the vapor com-
pression cycle that is driven by the input of power, in a larger context primary energy
made available mostly in the form of heat is the source of the power supplied to
heat pumps. Thus, Figure 3.2.1 is the ideal representation of several different heat
pumping concepts. Examples are an engine-driven vapor compression heat pump,
a combination of a steam or gas turbine power plant with an electrical vapor
compression heat pump or the representation of absorption or adsorption heat pump.
The coefficient of performance for the combined device is customarily defined as
Q0
COPAR = (3.8)
Q2
Q1′ + Q1′′
COPAHP = (3.9)
Q2
Vapor Compression Cycle Fundamentals 69
T2 − T0 T1
COPAHP = (3.10)
T2 T1 − T0
T2 − T1 T0
COPAR = (3.11)
T2 T1 − T0
Here, the distinction between a heat pump and a refrigeration system is only a
function of the application, not of the operating mode. Equation 3.7 is also valid for
any heat driven heat pump:
3
4 Condenser 2
Expansion
Compressor (a)
Va lve
Evaporator
5 1
ln (P)
4 3 2
(b)
5a 5 1
FIGURE 3.3.1 (a) Schematic of a vapor compression cycle. (b) Schematic of ln(P)-h diagram
with the state points of a vapor compression cycle superimposed.
plot the cycle on the ln(P)-h diagram, the saturation pressure or temperature for the
evaporator and the saturation temperature or pressure for the condenser.
To demonstrate this concept, the refrigerant ammonia is selected as an example.
The saturation temperature of the evaporator is assumed to be –20˚C and that of the
condenser 40˚C. The ln(P)-h diagram is shown in Figure 3.3.2. With the knowledge
of the evaporator saturation temperature, state point 1 is defined, located on the
saturated vapor line at the intersection with the isotherm of –20˚C. The saturation
pressure can be read from the pressure axis as about 0.19 MPa. Points 3 and 4 are
found from the knowledge of the condenser saturation temperature and the intersec-
tion with the saturated vapor and saturated liquid lines, respectively. The condenser
saturation pressure is read to be 1.7 MPa. Point 2 is found by following a line of
constant entropy from point 1 to the intersection with the isobar of the condenser
saturation pressure. Since the expansion process is isenthalpic, point 5 is found
vertically below 4, that is, the intersection of the line of constant enthalpy and the
evaporator isobar. With this, the ideal vapor compression cycle is completely defined
on the ln(P)-h diagram. In most cases, it is best to start plotting the ideal vapor
compression cycle on the ln(P)-h diagram by finding points 1 and 4; then all others
are fixed.
There seem to be two dominating reasons why the vapor compression cycle is
the cycle of choice for a wide range of refrigeration applications. One reason is the
fact that the latent heat during the liquid-vapor phase change process is used as the
mechanism to transfer heat to and from the working fluid. This greatly reduces the
mass flow rate in the cycle. But at least as important is the following consideration:
All other refrigeration cycles involve the exchange of work twice, once in a com-
pression process and once in an expansion process. The difference between these
two contributions of work is the net work input for a heat pump. When each of the
two processes that involve work has a realistic efficiency (for example, 0.8), the net
work is to be divided by an overall efficiency of 0.64. This efficiency of the cycle
4 3 2
5 1
FIGURE 3.3.2 ln(P)-h diagram of ammonia vapor compression cycle. (Copyright 2001, American Society of Heating, Refrigerating, and
Air-Conditioning Engineers, Inc. Reprinted by permission from ASHRAE 2001 Handbook — Fundamentals.)
m 1 = m 2 = m (3.13)
The energy balance states that all energy supplied to the compressor has to leave it
also under steady state condition without considering compressor losses. The mass
flow rate entering equals the mass flow rate leaving. In fact, since the ideal vapor
compression cycle is a closed system as can be seen from Figure 3.3.1(a), this
equation holds for all components of the cycle (W is replaced by Qi as appropriate),
Rearranging Equation 3.14 and using 3.13, the power input, W, can be calculated as
In a similar fashion, the amounts of heat exchanged in the evaporator, Qevap, and
condenser, Qcond, can be calculated as follows:
and
The energy balance must also be fulfilled for the overall cycle. It is left to the reader
to show that
Equation 3.18 serves as a good check to confirm that the above calculations are
internally consistent.
Figure 3.3.2 reveals additional useful information. The temperature of the vapor
leaving the compressor can be read from the chart, in this case about 141˚C. Further,
point 5 is located in the two-phase range. Using the lines of constant vapor quality,
x, the quality at the evaporator inlet, is determined to be 0.22. This indicates that
22% of the stream entering the evaporator is already evaporated. This evaporation
has to occur to ensure that the entire stream is cooled to the evaporator saturation
temperature.
T T
s s
(a) Carnot Cycle (b) Vapor Compression Cycle
FIGURE 3.4.1 Schematic of Carnot cycle (a) and vapor compression cycle (b) in a T-s
diagram.
74 Vapor Compression Heat Pumps with Refrigerant Mixtures
T
Isenthalpic
Expansion
Superheat
Horn
Isentropic
Expansion
FIGURE 3.4.2 Isenthalpic and isentropic expansion for a vapor compression cycle in a T-s
diagram.
Isenthalpic
ln(P) Expansion
Isentropic 4 3 2
Expansion
5a 5 1
FIGURE 3.4.3 Isenthalpic and isentropic expansion for a vapor compression cycle in an
ln(P)-h diagram.
Vapor Compression Cycle Fundamentals 75
in a ln(P)-h diagram. The enthalpy difference h5 – h5a represents the loss of work.
This energy, when not extracted from the fluid, results in an increased vapor quality.
Consequently, less liquid refrigerant is present to provide cooling.
It was shown in a previous chapter that the amount of work lost in the expansion
process is small and usually readily accepted. The superheat horn is not easily
avoided with the constraints imposed by conventional technology. Figure 3.4.4 puts
these contributions more clearly in perspective (Domanski, 1995). It shows the
relative penalties of the superheat horn and the expansion process. The term relative
penalty is defined as follows: The relative penalty for the loss of cooling capacity
is expressed as Qexp /QC, which stands for “capacity with isenthalpic expansion”
divided by the “capacity with isentropic expansion;” further, Wexp /WC expresses the
ratio of the respective work terms related to the expansion process and Wsup /WC the
additional work because of the superheat horn. In Figure 3.4.4, these losses are
graphed on the same bar chart as stacked bars for a large number of refrigerants.
The x-axis of Figure 3.4.4 shows the refrigerants sorted according to their vapor
specific heat capacity. For ammonia, the refrigerant with the lowest heat capacity,
the two work effects have more or less the same relative penalty. With increasing
heat capacity, the effect of the superheat horn vanishes while both of the remaining
penalties increase and the expansion process quickly becomes the most dominant
one. It should be noted that the relative penalties are not indicative of how much
each loss contributes to an overall degradation to COP. A more detailed analysis is
given in Domanski (1995). Figure 3.4.4 shows how the penalties vary with the
specific heat capacity of the refrigerant. This makes clear that the specific heat
capacity of a refrigerant has an important influence on system performance, resulting
in two effects. To better understand this, it is assumed for the following that a given
refrigerant has a high specific heat capacity. It is assumed also, without limiting the
general validity of the following consideration, that a positive displacement com-
pressor is used.
The temperature increase during the compression process is low. This results
from the fact that the compressor work is turned into the internal energy of the gas
within the compressor. There are several options to increase the internal energy of
a gas. Energy can be stored within molecules in the form of rotational or oscillatory
motion or in the form of translational energy, i.e., the speed at which the molecule
moves, its kinetic energy. Only the latter results in a temperature increase.
When the specific heat capacity of a gas is high, a considerable amount of the
increase in internal energy is stored within the molecule and only a small fraction
leads to an increase in kinetic energy, that is, its temperature. The temperature
increase requires increased compression work. For fluids of high specific heat capac-
ity, this effect is smaller than for fluids with low specific heat capacity. Thus, when
the compressor compresses a gas that stays relatively cool, the compressor work is
not as high as if the specific heat capacity were low and the temperature of the gas
increased more quickly.
On the other hand, during the expansion process of a fluid with a large specific
heat capacity, a relatively large amount of refrigerant has to evaporate to cool the
remaining liquid to the evaporator saturation temperature and less liquid is left to
provide cooling capacity.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
R-744
NH3
R-32
R-14
Qexp/QC
R-23
Wsup/WC
Wexp/WC
R-C270
R-22
R-13
R-13B1
R-114
R-236ea
R-227ea
I-C5
R-113
N-C5
R-218
E-245
R-C318
Vapor Compression Cycle Fundamentals 77
ln (P) ln (P)
h h
(a) (b)
FIGURE 3.4.5 Schematic of vapor compression cycle in a ln(P)-h diagram for a fluid of low
(a) and high (b) specific heat capacity.
This situation is better understood when the ln(P)-h diagram for a refrigerant of
high specific heat capacity is compared to one with low specific heat. Figure 3.4.5
shows the two diagrams side by side. Figure 3.4.5(a) shows the diagram of a low
heat capacity refrigerant; Figure 3.4.5(b), that of a high heat capacity fluid. Obvi-
ously, the vapor dome is tilted much more to higher enthalpies for the high capacity
fluid. As a consequence, the expansion process produces much more vapor and less
liquid and the evaporator capacity is reduced considerably. Furthermore, the com-
pression stays very close to the saturated vapor line and the compressed vapor is
almost not superheated at all. In fact, for some refrigerants, the state point of the
compressed vapor might be located in the two-phase range. In practice, avoiding a
two-phase stream leaving the compressor involves sufficiently superheating the
suction vapor. The penalty in terms of increased compressor work is small or even
desirable to avoid a two-phase condition at the compressor outlet.
4a 3
4 Condenser 2
(a)
Expansion
Compressor
Va lve
Evaporator
5 1a
1
ln (P)
2s 2
3
4a
4
(b)
5 1 1a
FIGURE 3.5.1 (a) Schematic of vapor compression cycle with state points to explain realistic
operating conditions. (b) Schematic of realistic vapor compression cycle on a ln(P)-h diagram.
Superheat (1–1a), subcooling (4a–4) and pressure drop (slant in line 2–4 and 5–1a) are
included.
contribute to the cooling capacity. Second, superheating usually occurs within the
evaporator for the purpose of ensuring that the compressor receives only vapor
without any liquid entrained that could damage the compressor.
Due to the pressure drop in the evaporator and the superheating at the evaporator
outlet, the compression process starts at a lower pressure and further into the
superheated vapor region than in Figure 3.3.1. This is a double penalty for the
compressor. A higher-pressure difference needs to be overcome by the compressor
and the work also increases due to the higher temperature of the suction vapor. An
inspection of the ln(P)-h diagrams, Figures 2.1.12 through 2.1.24, reveals that the
slope of the isentropes decreases as the enthalpy increases, i.e., as the state point
moves further away from the two-phase region. Thus, for isentropic compression,
the compressor work increases once again as the entering suctioned gas is super-
heated more.
Furthermore, the compressor must overcome the pressure drop of the refrigerant
in the discharge line and the condenser, yet another increase in compressor work
compared to the ideal cycle. The entire compression process is represented by the
process line 1a–2. Since the compression process is usually irreversible to some
degree due to internal friction of the mechanical parts, heat transfer and valve
pressure drop, point 2 lies to the right of the endpoint 2s that would be achieved
with isentropic compression. An additional penalty is increased volume, reducing
capacity for a given displacement.
The process line 2–3 represents the desuperheating process in the condenser and
line 3–4a the condensation process. Both exhibit the necessary pressure drop. Line
4a–4 represents the subcooled region, where the condensate is cooled to a lower
temperature.
Condensing
T 5 Refrigerant 6
8
7 Cooling Water
3
4 Chilled Water
1 2
Evaporating
Refrigerant
s
FIGURE 3.6.1 The evaporating temperature (1,2) and condensing temperature (5,6) are fixed
by the lowest chilled water (4) and highest cooling water (8) temperature, respectively. A
pinch point occurs at (1,4) and (5,8).
differences between both fluid streams become very small. As a consequence, the
heat transfer rate becomes small as well. A pinch point limits the heat transfer rate
due to very small approach temperatures. In order for a heat exchanger to still
facilitate heat exchange, the available heat transfer surface area must increase.
However, in this case the entropy production is also small due to the fact that heat
is transferred over a small temperature difference. In contrast, at points 2 and 3 and
also for 6 and 7, the temperature differences are large, the heat transfer rate is large
as well and so is the entropy production.
Note that even when the heat transfer area approaches infinity, the performance
improvement of the heat exchanger is limited (Rice, 1993a). While the temperature
difference at the pinch point approaches zero, the temperature differences at the
other end of the heat exchangers stay finite and the entropy production approaches
a finite value. In case of the evaporation and condensing processes in Figure 3.6.1,
the highest temperature of the evaporating refrigerant and the lowest temperature of
the condensing refrigerant are fixed by the respective water temperatures. In turn,
their pressures are fixed and thus the compressor power, so even with heat exchanger
areas approaching infinity, the entropy production in the heat exchangers cannot be
reduced to zero and the performance improvement of the entire heat pump is severely
limited.
The Lorenz cycle addresses this issue (Herold, 1989; Alefeld, 1987). When a
working fluid is available that changes its temperature during the course of the evapo-
ration or condensing process (i.e., it has a temperature glide as described in Chapter 2),
then the schematic of Figure 3.6.1 can be modified as shown in Figure 3.6.2.
In the ideal case that the change in temperature from the beginning to the end
of the evaporation process matches that of the chilled water, both process curves
become parallel and the pinch point is eliminated. Figure 3.6.2 shows the Lorenz
cycle. With an increase in heat transfer area, the temperature difference between the
two fluids continues to shrink everywhere in the exchanger and thus the overall
entropy production can theoretically approach zero. Assuming that the cooling and
chilled water temperatures are given, the suction and discharge pressures can
T Condensing Refrigerant
6
8
5
7 Cooling Water
4 Chilled Water 2
1 Evaporating Refrigerant
FIGURE 3.6.2 When the refrigerant saturation temperature has a glide, the pinch point can
be eliminated.
approach each other to the largest possible extent, certainly much further than in
the case shown in Figure 3.6.1. The overall efficiency of the cycle improves to the
largest possible extent, too. It should be noted that even for the same heat exchange
area, a Lorenz cycle that better matches the source and sink glides gives a higher
cycle efficiency because it operates at an improved mean temperature than the
corresponding pure refrigerant (Rice, 1993b).
The mixtures of Chapter 2 have the potential to approach the requirements of
the Lorenz cycle depending on the degree to which they match the application glide
in both the evaporator and the condenser. It is interesting to observe that for the
given chilled and cooling water temperatures of Figure 3.6.2, the Carnot cycle, also
symbolized by the process lines in Figure 3.6.1 (1–2–6–5), should have lower
efficiency than the Lorenz cycle of Figure 3.6.2 (1–2–6–5). After all, in a T-s diagram,
the area enclosed by the process represents the required work input as long as all
processes are reversible. This condition is fulfilled here. The reader might have the
expectation that no thermodynamic cycle can exceed the performance of the Carnot
cycle.
Both statements are correct. Upon closer examination, the Lorenz cycle does
not exceed the performance of Carnot cycles. This can be demonstrated with the
help of Figure 3.6.3. The Lorenz cycle as shown in Figure 3.6.3 can be approximated
by a series of Carnot cycles that are superimposed on the Lorenz cycle in a step-
like fashion. In the limiting sense that the Carnot cycles have an infinitesimal length
in the s-direction, they approach the Lorenz cycle.
Thus, the performance of the Lorenz cycle can be approximated through a series
of Carnot cycles and therefore the Second Law is not violated. In general, the
performance of many systems can be improved when a series of Carnot cycles (or
more realistic cycles) are used that more closely match the temperature profiles of
heat sinks and heat sources.
1
s
FIGURE 3.6.3 The Lorenz cycle (1–2–6–5) can be approached by many Carnot cycles
(small, vertical rectangles), when operated such that they approximate the Lorenz cycle.
3
4 Condenser 2
5 Evaporator 1
B
re
re
Vapor
Pu
Pu
3,4 4
Condenser 2
2 3
Evaporator
5,1 5 1
−1/T −1/T
(a) (b)
Low Side
Pressure Limit
FIGURE 3.7.2 When a vapor pressure curve of a given refrigerant is located within the
window, then it is suitable for the intended application. The window is specified through the
desired evaporator and condenser temperatures and the acceptable high and low pressure
limits.
the vapor pressure lines of the pure fluids A and B. This could imply that the outlet
condition is a two-phase fluid. It must be understood that each of the vapor pressure
lines is an isostere, a line of constant mass fraction. Each of the lines separates a
liquid and a vapor region for a mixture of that particular mass fraction. Thus, when
point 1 indicates saturated vapor leaving the evaporator, any point to the right of
that particular isostere represents superheated vapor with regard to that isostere.
In most pressure-temperature diagrams, there are no provisions to indicate the
dew points. The vapor curves shown indicate only bubble points. The user must find
the information about dew points 1 and 3 from another diagram or source. Thus,
the observer must understand the cycle and the diagram quite well to correctly
comprehend the meaning of point 2.
The pressure-temperature diagram is nevertheless a good tool to gain insight
about pressure and temperature relationships. For example, it allows a quick screen-
ing to determine which fluid is suitable for a given application. An application is
determined by the expected temperature levels of the evaporator and condenser and
by pressure limitations. The latter are dictated by two considerations: The first
limitation is the maximum pressure rating for commercially available components
and the second is given by what suction pressure is deemed acceptable for the
compressor. These values create a window as shown in Figure 3.7.2.
T3
T4 T4a
ln (P)
3 2
4 4a
T5a
5
1 1a X = Const
T1
T5
h
FIGURE 3.7.3 ln(P)-h diagram for a zeotropic mixture, with the state points of the vapor
compression cycle superimposed.
state points are the same as those for pure fluids. The difference lies in the underlying
fluid properties. Figure 3.7.3 shows, in addition to the cycle, the isotherms for the
state points in the two-phase region. In this way the temperature glide is represented
correctly. For example, in the evaporator the glide is T1–T5 and in the condenser
T3–T4a.
Upon closer inspection, Figure 3.7.3 reveals other important differences between
pure and mixed refrigerants in the vapor compression cycle. The temperature T5 is
higher than the liquid saturation temperature at the same pressure T5a. This indicates
that for mixtures, in contrast to pure fluids, the expansion process has two penalties.
There is less liquid to evaporate, which is the case for pure and mixed fluids, but
for mixtures, the temperature at the evaporator refrigerant inlet is not as low as it
could be without the flashing. This first loss, caused by the increased vapor quality,
affects the capacity and efficiency of the system and shows in the energy balance
for the evaporator. It is therefore termed a “First Law loss.” The second loss, the
temperature increase, does not show up in the energy balance but its effects are
clearly seen in a Second Law analysis. This situation is similar to the effects of the
glide. The fact that refrigerant mixtures have a glide is not immediately visible in
the equations of a First Law analysis using energy balances and is only apparent in
a Second Law analysis of the evaporator or condenser.
Nevertheless, when there is a Second Law benefit such as less entropy produc-
tion, it must show up in a better COP as well. With the glide, this benefit manifests
itself by an increase in suction pressure or a decrease in the discharge pressure as
discussed in Section 3.6; thus, the compressor work is reduced. The benefit that can
be attained when the amount of vapor produced during flashing is reduced is dis-
cussed in more detail later.
The reader should keep in mind that if the overall mass fraction changes for any
reason, this diagram is no longer valid and a new ln(P)-h diagram has to be generated
for the new mixture mass fraction.
T 2
P5
P3 3 P1
4
P4
5
1
X = Const
FIGURE 3.7.4 T-s diagram for vapor compression cycle. Pressure drop in the evaporator and
condenser is indicated through the slanted process lines 5–1 and 3–4.
2
T
3
P = Pcond
X3’ 4 4a
X4’
1
5 P = Pevap
5a
X1’
X5’
X5’’
0.0 X1,2,3,4,4a,5,5a 1.0
X
FIGURE 3.7.5 T-x diagram with the state points of the vapor compression cycle superimposed.
of vapor and liquid are different, depending on its location, a system leak can cause
a shift in the mass fraction of the remaining refrigerant mixture. The same holds for
periods when the system is turned off. A heat pump is filled to about 30% of its
total volume with liquid refrigerant. The remainder is refrigerant vapor with a mass
fraction different from that of the liquid phase. Again, a leak that allows only vapor
or only liquid to escape causes the mass fraction of the remaining refrigerant to
shift. Since the liquid phase holds the bulk of the refrigerant mass, which is close
to the charged mass fraction, vapor leaks are more effective in changing the mass
fraction. The influence of leaks is discussed in more detail in Section 9.9.
T
Hot Fluid 1
Hot Fluid 3
Hot Fluid 2
Cold Fluid
FIGURE 3.8.1 Matching of temperature glides. Hot fluid 3 has the best match with the cold
fluid.
Hot Fluid
Cold Fluid 2
Cold Fluid 1
FIGURE 3.8.2 Matching temperature glides. Cold fluids 1 and 2 leave nonlinear temperature
glides and pinch points can occur in the middle of the heat exchanger (Cold fluid 2) or in the
end (Cold fluid 1).
When the glides are parallel, i.e., they “match” as it is often termed, the entropy
production by the heat exchanger can be reduced to zero (in theory) by increasing
the heat exchanger area. When a pinch point exists, this is not possible any more.
When the heat exchanger area is increased, the approach temperature at the pinch
point becomes smaller but a large temperature difference continues to exist at the
other end of the heat exchanger. In the theoretical limit, for an infinitely large area,
the entropy production of this heat exchanger stays finite, even when the approach
temperature of the pinch point approaches zero. Thus, it is very important to strive
wherever possible for matching glides in heat exchanger design. With a nonlinear
glide, the pinch point can occur at any location within the heat exchanger. Figure
3.8.2 shows two examples with a concave and convex glide. For the concave case,
the pinch point occurs at the end of the heat exchanger; however, for the convex
case, the pinch point occurs somewhere in the middle. Its location can no longer be
controlled by selecting the thermal capacity alone. Also, the existence of this pinch
point cannot be observed when only inlet and outlet stream temperature are measured.
When refrigerant mixtures are used, a nonlinear glide can be thought of as caused
by two mixture constituents of significantly different latent heats. Figure 3.8.3
displays the latent heat of a two-component mixture over the mass fraction range.
The latent heat for component B is assumed to be significantly larger than that for
component A. Let us further assume that the fugacity of B is smaller than that of A
and that any effects of the heat of mixing are negligible. Thus, component A will
predominantly evaporate and the apparent latent heat for this portion of the evaporation
process will be dominated by the latent heat of A. This is indicated in Figure 3.8.3
with the latent heat of the mixture staying at levels similar to that of A. In this region,
a small amount of heat evaporates a relatively large amount of refrigerant. With the
accompanying change in composition, this accounts for a significant temperature
glide as shown in Figure 3.8.4, while the remaining liquid is evaporated with a
significantly higher latent heat.
Latent
Heat (q)
0 1.0
X
A B
FIGURE 3.8.3 Component A has a low latent heat, component B a high one. When the mass
fraction of A is high in the vapor phase, its latent heat dominated that of the mixture for a
wide range of x.
Hot Fluid
Mixture of A & B
FIGURE 3.8.4 A mixture with the constituents A, B of Figure 3.8.3 shows a nonlinear glide
with a pinch point in the middle of the heat exchanger.
During this first portion of the evaporating process, the remaining liquid phase
will be enriched in B only when the evaporation process approaches its end and the
remaining portion of B must evaporate as well with the latent heat of the mixture
approaching that of B. The curve of the latent heat in Figure 3.8.3 rises to the level
of the latent heat of B. In this portion of the evaporation process, the amount of heat
absorbed for a given mass fraction change is larger. The resulting temperature interval
is larger than before and the glide decreases in slope as shown in Figure 3.8.4. The
mixture AB has characteristics that produce a convex temperature glide with a pinch
point in the middle of the heat exchanger.
When the latent heat of A is much higher than that of B, the effects are reversed
and curves with the opposite behavior were generated. The result is a mixture with
a concave glide. Nonlinear glides have certain challenges and opportunities in heat
pump applications. When water or dry air is the heat transfer medium for an
evaporating or condensing mixture, the mixture should have a linear glide of equal
size for best glide matching when the refrigerant side pressure drop is accounted
for. When the heat transfer medium has a nonlinear glide on its own, as happens in
the case of dehumidification of air, the temperature glide of the refrigerant mixture
should match that of the air stream. Another example where a nonlinear glide is
welcome is for systems with two evaporators or two condensers that operate on
different temperature levels. A more detailed description of these is given in Chapter 5.
It should be noted that for best heat pump performance, the temperature glides
must match in both heat exchangers, the evaporator and condenser. The situation is
complicated by the fact that most systems require in the evaporator a certain degree
of superheat and in the condenser a certain degree of subcooling. These character-
istics add their own glide to that of the mixture two-phase range. Finally, a significant
superheat is also present in the condenser. This challenge is never met by any heat
transfer fluid and the designer is forced to find a suitable compromise.
Case A Case B
Tcond
Mixture Pure Refrigerant
Refrigerant Temperature
Tevap
Case C Case D
Refrigerant Flow
Tcond
Tevap
Case A
6
5
Heating COP Case D
4 Case C
3 Case B
2
0 0.2 0.4 0.6 0.8 1.0
Composition (mole fraction of R22)
1. When the temperature glide of the refrigerant inside the heat exchanger
matches that of the external fluid stream, the potential COP improvement
is largest.
2. If the glide of the refrigerant is much smaller or much larger than that of
the external fluid, the COP is penalized.
3. The larger the glide (when refrigerant glide and external glide are
matched) and the smaller the lift, the higher the improvement in COP.
4. The theoretical limit for the improvement in COP is the doubling of the
value for a pure fluid (Case A in Figure 3.9.1). In this case, the temperature
of the refrigerant leaving the evaporator is the same as that of the refrig-
erant leaving the condenser. Any further increase in the evaporator outlet
temperature, or decrease in the condenser outlet temperature, would allow
for direct heat exchange between the respective external fluid streams and
eliminate the need for a heat pump in this range of the temperature glide.
5. Since the glide of a mixture varies with mass fraction, only certain mass
fractions lead to the best match and therefore performance improvement.
When a mixture has a large temperature glide and the application only a
narrow one, there may be two mixture mass fractions that match the glide:
a low mass fraction and a high mass fraction. At other mass fractions, the
performance could be reduced.
6. The capacity changes primarily with the vapor pressure of the refrigerant
at evaporator conditions. When mixtures have a large glide, the change in
capacity with mass fraction can be considerable as discussed in Chapter 7.
7. The highest performance improvement is gained when the glide in the
condenser and the evaporator are matched simultaneously.
impacts not only the heat transferred, but also the amount of refrigerant in the system.
In general there are two types of heat exchanger models: One is a relatively simple
UA model and the other is a detailed model based on distributed parameter heat
exchanger simulation that solves the continuity, species, momentum and energy
equations simultaneously for air (or water) to refrigerant heat exchangers while using
local heat transfer and pressure drop correlations.
Prominent among the previous work on this subject is the tube-by-tube analysis
developed by Domanski (1989 and 1995). In this analysis, a wet (moisture from air
conditioning on the outside surface) or dry evaporator was modeled by assuming
the air passes directly from one tube to another. The heat and mass transfer of each
tube was evaluated based on its own air and refrigerant inlet states; this allowed for
investigation of air maldistribution. The simulation, restricted to pure refrigerants,
was experimentally verified for wet coils and found to predict the overall capacity
within –8.8 to +23.3% (Chawalowski et al., 1989).
Using the work by Domanski, Conde and Sutter (1992) developed a similar
simulation with several modifications. One of the improvements included the ability
to determine the refrigerant inventory within the evaporator. This code was part of
an overall steady state cycle simulation and was experimentally verified with R22.
The verification showed that the simulation frequently overpredicted the capacity
by more than 10%.
Oskarsson et al. developed three steady state models for dry, wet and frosted
evaporators. All of the models assumed that the entire heat exchanger could be
simulated with one equivalent tube in crossflow. Furthermore, these models were
developed for pure refrigerants. One of the models divided the evaporator into three
different regions: the two-phase region, the transitional region and the superheated
region. The most complex model divided the heat exchanger into finite elements,
while the simplest model was parametric and required one of the other models to
generate constants for it. The three-region model was experimentally verified and
found to predict the capacity of a dry coil between –17.7 and +11.3%. Furthermore,
the finite element model agreed with the three regions model within 2%.
Nyers has developed a transient evaporator model (Nyers, 1994). This model
was developed for air to pure refrigerant applications. This model was fully distrib-
uted and simultaneously solved the continuity, momentum and energy equations but
was not compared to any transient or steady state experimental data.
Judge and Radermacher (1997) developed a finite element transient and steady
state model that was verified experimentally within 8%. The authors used this model
to quantify the effects of heat exchange geometry on the performance of R407C.
For the conventional shell-and-tube heat exchangers, Hwang and Radermacher
(1998) developed a water-to-refrigerant evaporator and condenser models that
divided the heat exchanger into finite elements. These models are verified with
experimental data for R22 and CO2: a 5% deviation for the evaporator model and
a 7% deviation for the condenser model were reported.
UA Model
When analyzing the heat exchanger, either log-mean-temperature-difference
(LMTD) or the effectiveness-NTU (number of transfer units) method can be used.
In this section, only the LMTD model is explained. If the reader is interested in the
effectiveness-NTU method, the detailed explanation and examples can be found
from Incropera and DeWitt (1990).
In the UA model, the heat exchanger performance is defined as Equation 3.19,
Q = UA × LMTD (3.19)
where
U = overall heat transfer coefficient
A = total heat transfer area
LMTD = log-mean-temperature-difference
In this model, three parameters in Equation 3.19 are given and the other variable
is calculated by Equation 3.19. If the heat exchanger geometry (A) and the opera-
tional conditions (U and ∆T) are given, the heat exchanger performance (Q) is
calculated. On the other hand, if the target performance (Q) is given, only one
parameter among the heat exchanger geometry (A) and the operational conditions
(U and ∆T) is calculated while the other two parameters are given.
Applying the UA model for each flow regime can further advance the UA model.
An evaporator should be divided into two-phase and superheated regions. A con-
denser should be divided into superheated, two-phase and subcooled regions. The
total heat exchanged is then the sum of the heat exchanged at each region, as shown
in Equation 3.20,
Q= ∑ Q = ∑U A × LMTD
i i i i (3.20)
where
Ui = overall heat transfer coefficient for each segment
Ai = total heat transfer area for each segment
LMTDi = log-mean-temperature-difference for each segment
When applying the LMTD, some considerations must be accounted. First, the LMTD
is designed for counterflow heat exchange configuration. However, it can be used
for many different HX (heat exchanger) geometries by using the correction factor
(CF) as shown in Equation 3.21,
∆Tin − ∆Tout
LMTD = CF ×
( )
(3.21)
ln ∆Tin / ∆Tout
where
CF = correction factor for heat exchanger geometry (examples are given in
Figure 3.10.1)
∆Tin = temperature difference between hot fluid inlet and cold fluid outlet
(Thf,in – Tcf,out)
∆Tout = temperature difference between hot fluid outlet and cold fluid inlet
(Thf,out – Tcf,in)
1.1
p/q=0.2
1.0
0.4
FIGURE 3.10.1 LMTD correction factor for single pass crossflow heat exchanger (one fluid
mixed, other unmixed).
When one fluid undergoes phase change or two fluids have a very similar heat
capacity, then CF = 1. A correction factor for single pass crossflow HX, one fluid
mixed, other unmixed, as is the case for most of the air-to-refrigerant heat exchan-
gers, is available from Bowman (1940) as shown in Equation 3.22,
p− q
CF = r /r0 = r
( )
1− q
(3.22)
ln 1− p
LMTDhx = ∑ f LMTD
i
i i (3.23)
where
LMTDhx = LMTD for overall heat exchanger
fi = heat transfer fraction for individual phase zone (Qi /Qhx)
LMTDi = LMTD for individual phase zone
By assuming the constant heat transfer coefficient (U) for all phases, Rice and
Sand (1990) and Domanski and McLinden (1990) introduced the harmonic weighted
LMTD as defined in Equation 3.24,
1
LMTDhx
= ∑ LMTD
i
fi
i
(3.24)
Nu a = 0.5 Gz 0.86 (Pt /D c )0.11(s/D c )−0.09 (Sd /Pl )0.12 (2Sp /Pl )−0.34 for Gz ≤ 25
(3.26)
Nua = 0.83 Gz 0.76 (Pt /Dc )0.13 (s/Dc )−0.16 (Sd /Pl )0.25 (2Sp /Pl )−0.43 for Gz > 25
where
Gz = Graetz number (RePrDH/L)
s = space between fins (fin pitch – fin thickness)
Dc = fin collar outside diameter
Sd = peak-to-vallet wave depth
Pl = longitudinal tube pitch
Sp = one-half wave length
NR = number of tube rows
Nu =
(
( f /8) Re − 1000 Pr )
( Pr )
(3.27)
1 + 12.7( f /8) 0.5 2/ 3
−1
where
f = (0.79 ln(Re) – 1.64)–2
As for the two-phase heat transfer coefficients, many correlations are available.
In this section, one preferred boiling and one condensation correlation are intro-
duced, which provided the smallest error with experimental data. Details of boiling
heat transfer correlations are described in Section 8.3; details of condensation heat
transfer correlations are described in Section 8.6.
Jung et al. (1989b) developed the following flow boiling correlation for horizontal
tubes for the binary mixtures R22/R114 and R12/R152a as shown in Equation 3.28,
where
hl = heat transfer coefficient for the liquid phase flowing alone
F = forced convection enhancement parameter
where
F = 2.37 (0.29 + 1/Xtt)0.85
S = forced convection suppression parameter
where
S = 4048 Xtt1.22 BO1.13 for Xtt < 1
S = 2 – 0.1 Xtt–0.28 BO–0.33 for 1 < Xtt < 5
Cme = 1 – 0.35 |Xv – Xl|1.56
where
Xl = liquid phase mass fraction
Xv = vapor phase mass fraction
Cun = |1 + (b2 + b3)(1 + b4)| × (1 + b5)
hpool = pool boiling heat transfer coefficient
where
hpool = hi /Cun
hi = ha hb /(hb Xl + ha Xv)
ha hb = pool boiling heat transfer coefficient of pure refrigerants a and b given
by Stephan and Abdelsalam’s (1980) correlation
b2 = (1 – Xl) ln |(1.01 – Xl)/(1.01 – Xv)| + Xl ln (Xl /Xv) + |Xl – Xv|1.5
b3 = 0 for Xl µ 0.01
b3 = (Xl /Xv)0.1 – 1 for Xl < 0.01
b4 = 152 (P/Pca)3.9
b5 = 0.92 |Xl – Xv|0.001 + (P/Pca)0.66
Xl /Xv = 1 for Xl = Xv = 0
where
Pca = critical pressure of more volatile component
Sweeney (1996) and Sweeney and Chato (1996) correlated Kenney et al.’s (1994)
data for R407C in a smooth tube using Dobson’s correlations,
0.3
G
h = 0.7
300
kl
D
0.8 0.4
( )
−0.89
0.023 Rel Prl 1 + 2.22 Xtt for annual flow regime (3.29)
where
Nu f = 0.0195 Rel0.8 Prl0.4 φl ( X tt )
where
C1
φl ( X tt ) = 1.376 +
X tt c2
where
c1 = 4.12 + 0.48 Frl – 1.564 Frl2, c2 = 4.12 + 0.48 Frl – 1.564 Frl2, for 0 < Frl ≤ 0.7
c1 = 7.242, c2 = 1.655 for Frl > 0.7
where
Frl = liquid Froude number
θl = angle subtended from the top of the tube to the liquid level
θl arccos(2α − 1)
1− ≅ by Jaster and Kosky (1976)
π π
tanh(mL )
ηf = (3.31)
mL
where
m = (hP/kAc)0.5
h = air-side heat transfer coefficient
P = fin perimeter
k = fin thermal conductivity
Ac = fin cross-sectional area
L = fin height
Af
ηo = 1 − (1 − η f ) (3.32)
At
where
Af = fin surface area
At = total surface area (fin surface area + fin base area)
1
U= (3.33)
Ao 1 Ao ln(ro /ri ) 1
+ +
Ai hr 2πkL ηo ha
where
hr = refrigerant-side heat transfer coefficient
ha = air-side heat transfer coefficient
ro = tube outside radius
ri = tube inside radius
Ao = tube outside surface area
Ai = tube inside surface area
k = thermal conductivity of tube
L = tube length
0.0791
f= for 2.1 × 10 3 < Re < 10 5 (3.34)
Re 0.25
As for the two-phase pressure drop, many correlations are available. In this
section, the most preferred boiling pressure drop and condensation pressure drop
correlations are introduced, having the smallest error with experimental data. Details
of boiling pressure drop correlations are described in Section 8.4; details of conden-
sation pressure drop correlations are described in Section 8.7.
Jung and Radermacher (1989) studied the boiling pressure drop of R12, R22,
R114, R152a and R12/R152a. The authors suggested a modified version of Marti-
nelli-Nelson correlation as shown in Equation 3.35,
2 f f 0G 2 L 1
∫
x2
∆Ptp = φtp2 dx (3.35)
Dρl ∆x x1
where
ffo = 0.046 Re–0.2
G = mass flux
D = tube diameter
ρl = liquid density
where
Pr = reduced pressure (P/Pc)
where
∆P = total two-phase pressure drop
∆Pf = frictional pressure drop
∆Pm = pressure drop due to the momentum change
G2 x 2 ρ (1 − x )2 x 2 ρ (1 − x )2
∆Pm = + − +
l l
(3.37)
ρl α ρ (1 − α ) α ρ (1 − α ) 1
v 2 v
where
G = mass flux
ρl = liquid density
ρv = vapor density
x = vapor quality
α = void friction
−1
x ρl /ρv + 0.4(1 − x )
( )
0.5
ρv 1 − x
α = 1 + 0.4 + 0.6 (3.38)
ρl x
x + 0 .4 (1 − x )
(
∆Pf /∆z = ∆P /∆z v Φ 2v ) (3.39)
where
Φv = Lockhart-Martinelli parameter
∆P 2 fvG 2 x 2
∆z = (3.40)
v
Dρv
where
∆Pv = vapor flow pressure drop
D = tube diameter
fv = frictional coefficient considering two-phase flow as vapor flow
TABLE 3.1
Effect of Heat Exchanger Geometry in Cooling Mode
TABLE 3.2
Effect of Heat Exchanger Geometry in Heating Mode
for the counter flow as compared to the crossflow configuration. It should be noted
that the penalty of parallel flow is a little bit more severe than the benefit of counter
flow since the parallel flow encountered a pinch point.
The compressor is the heart of the vapor compression cycle and there are many levels
of complexity among the compressor models in the literature. In general, compressor
models are classified into four versions: efficiency model, map-based model, lumped
parameter model and distributed parameter model. There are six representative com-
pressor efficiencies: overall isentropic efficiency (ηoverall,isen), volumetric efficiency
(ηvol), motor efficiency (ηmot), mechanical efficiency (ηmec), isentropic efficiency (ηisen)
and suction gas superheating efficiency (ηsuperheat) (Shaffer and Lee, 1976). The def-
inition of these six efficiencies is as follows:
shaft power
ηmoot =
actual motor power
(3.42)
indicated power
ηmec =
shaft power
where
indicated power = work delivered to the gas
An isentropic analysis model is the simplest model and is independent of the physical
characteristics of compressors and various types of refrigerants. Typically, two
efficiencies, isentropic and volumetric efficiency, are given as inputs in this model.
These two efficiencies can be directly obtained from experimental data or varied as
a design parameter.
The second model is a polytropic analysis model. Some investigators prefer to
use the polytropic efficiency instead of the isentropic efficiency. The polytropic
efficiency is defined by Equation 3.43,
n −1
n Pdis n 1
Wp = PsucVsuc − (3.44)
n -1 P
suc
n − 1 ( γ − 1) /η pol
= (3.45)
n γ
(Murphy and Goldschmidt, 1985). Although the map-based model is very accurate
for a given fluid at steady state, it is not generally applicable to refrigerants and
compressors other than the ones upon which it was based. Compressor manufacturers
provide coefficients based on ARI Standards for the positive displacement compres-
sor, ARI Standards 520 (1997) and 540 (1999). Revised ARI Standard 540 (1999)
used suction- and discharge-dew point temperatures as its test conditions while
AREP (1993) and Rice (1993) used mean temperature of bubble and dew temperature
for the evaporating and condensing temperatures. Pure refrigerants, azeotropic mix-
tures and near-azeotropic mixtures are not affected by dew point- and mean tem-
perature-approaches. However, zeotropic mixtures, especially having a large tem-
perature glide, can show differences when these two approaches are compared. Since
mean evaporating and condensing temperatures of zeotropic mixtures are higher and
lower than dew temperatures at the suction and discharge, respectively, the dew point
approach will result in higher COP due to the reduced pressure ratio than the mean
temperature approach.
The most complex compressor model is a distributed parameter compressor
model. In this model the complete set of multidimensional energy, momentum and
continuity equations is somewhat simplified and solved. MacArthur developed the
distributed parameter model that deals with seven different state points (MacArthur,
1984). The compression process is assumed to be polytropic. Heat transfer between
different compressor components is accounted for by utilizing constant heat transfer
coefficients, as is the effect of thermal storage. Judge (1996) developed a distributed
parameter compressor model that accounts for valve dynamics and heat exchange
between various components of the reciprocating compressor. Other studies using
this type of model include: Ng et al. (1976), Recktenwald et al. (1986), Perez-Segarra
et al. (1994) and Lio et al. (1994).
From this literature review, it is clear that there is a wide range of methods
available for modeling compressors. The simple method, the efficiency model, is
most convenient for design point analysis but is of less accuracy over a range of
operating conditions. However, this method is a good approach for parametric
studies. The distributed parameter model provides better accuracy but needs larger
computing time.
The purpose of the expansion device is to reduce the pressure of the refrigerant. The
temperature of the refrigerant during expansion is also reduced so that the refrigerant
can absorb heat in the evaporator. The simplest expansion device model is a ther-
modynamic expansion model that assumes an isenthalpic expansion with negligible
heat loss. The expansion device inlet enthalpy then equals the outlet enthalpy and
this model can avoid any hardware dependence.
The four common expansion devices are capillary tubes, short tube orifices
(STs), thermostatic expansion valves (TEVs) and metering valves. In actual systems,
the capacity and operating mode determine the preferred expansion device. Capillary
tubes are used for small capacity cooling applications; STs are used for medium
capacity cooling applications; TEVs are used for larger capacity cooling and heating
applications. Among the expansion devices, the ST garners attention recently because
of its simplicity, low cost and serviceability.
Short Tube Models
Though the geometry of the STs is very simple, the expansion process within the
STs is so complex that it is difficult to accurately correlate experimental data from
various refrigerants and boundary conditions. Three modeling efforts are introduced
as follows:
Mei (1988) developed a ST model based on his experimental data with R22. He
concluded that the degree of upstream subcooling and the pressure difference were
the major parameters of STs. He then suggested using a general orifice type model
but the orifice constant should be a function of those two parameters. Mei reported
that his correlation matched experimental data 3 and 4% in average for STs opening
and length, respectively.
Aaron and Domanski (1990) developed a ST model based on their experimental
data with R22. In their research, they found that the mass flow rate varied only 3 to
8% depending on downstream pressure in “choked” flow condition. The second
finding was that the mass flow rate depended upon the degree of subcooling: the
mass flow rate increased approximately 23% when the subcooling increased by
8.3˚C. They also reported that doubling the length of the ST reduced the mass flow
rate by 5% and inlet chamfering increased the mass flow rate by 5 to 25%, while
the exit chamfering had no impact on the flow rate. Based on the orifice equation
(Equation 3.46), they modeled the ST:
m = Cc Ac 2ρ( P1 − P2 ) (3.46)
where
Cc = ST chamfer coefficient
Ac = ST cross-sectional flow area
ρ = density of upstream fluid entering ST
P1 = upstream pressure
P2 = downstream pressure
Kuehl and Goldschmidt (1992) and Kim and O’Neal (1993 and 1994a) modified
the P2 of the form of the Aaron’s equation to correlate their data. Kornhauser (1993)
developed a more physically based model with Aaron’s experimental data by con-
sidering the metastable conditions occurring within the ST. However, this model is
applicable only to “choked” flow regions.
Since the correlations developed by Aaron, Kuehl and Kim were based on R22,
their correlations must be compared to experimental data with mixtures. Judge
(1996) compared his experimental data for R22 and R407C with Aaron and Kim’s
correlations. He reported that Aaron’s model predicted the low side pressure of R22
and R407C within 6 and 11% errors, respectively. He also reported that Kim’s model
had errors about twice as large as those of Aaron’s when it was compared with R22
experimental data.
Most recently, Payne and O’Neal (1998 and 1999) investigated mass flow char-
acteristics of R407C and R410A with and without POE oil through ST orifices 12.7,
19.1 and 25.4 mm in length with diameters of 1.09, 1.34, 1.71 and 1.94 mm. Payne
and O’Neal then fitted their data to a previously developed semi-empirical model
by Kim and O’Neal (1994a and 1994b), essentially the same form as Equation 3.46.
Payne and O’Neal’s correlation for R407C and R410A can handle two-phase inlet
conditions as shown in Equation 3.47. The effects of oil on the refrigerant mass flow
rate was reflected in the two-phase quality correction factor (Ctp) and adjusted down
stream pressure (P2a),
where
m = mass flow rate [kg/s]
Ctp = two-phase quality correction factor
(Ctp = 1 for single phase; use Equation 3.48 for two-phase)
γc = constant (Table 3.3)
ρ = density of upstream fluid entering ST [kg/m3]
P1 = upstream pressure [kPa]
P2a = adjusted downstream pressure [kPa] (Equation 3.49)
{( )( )}
−1
C tp = 1 + a1x1 1 + a2 LD a3Y a4 Ln ( LD) (3.48)
where
LD = L/D
L = length of ST [mm]
D = ST diameter [mm]
a1, a2, a3, a4 = constant (Table 3.4)
0.5
x ρ
Y= 1 l
1 − x1 ρv
where
x1 = upstream vapor quality
ρl = saturated liquid density at upstream pressure
ρv = saturated vapor density at upstream pressure
{
P2a = Psat b1 + b2 PRAb3 LD b4 SUBC b 5 + b6 PRAb7 +
(3.49)
b8 exp(b9 DR × LDb10 ) + b11EVAP }
TABLE 3.3
Critical Parameters for Equation (3.47)
Critical Parameter R407C R410A
TABLE 3.4
Coefficients of Correction Factors for Equations (3.48) and (3.49)
R407C R410A
Coefficients Pure 1.1% POE Oil Pure 2.2% POE Oil
where
Psat = saturated liquid pressure corresponding to upstream temperature [kPa]
PRA = P1/Pc
DR = D/1.524
SUBC = (Tsat – T1)/Tc
EVAP = (Pc – P2)/Pc
b1 through b10 = constants (Table 3.4)
TABLE 3.5
Equations of State Available in REFPROP
Equation of State Applicable Fluids Source
Carling-Starling-DeSantis (CSD) All of the pure and mixture fluids DeSantis (1976)
except ammonia
Modified Bennedict-Webb-Rubin 13 fluids (R32, R123, R124, R125, Bennedict, Webb and
(MBWR) R134a, R143A, R152A, R290, Rubin, (1940)
R600, R600A, N-C5, I-C5, R236fa)
Extended corresponding states Thermodynamic properties of pure Huber and Ely (1992)
(ECS) and mixture fluids without an Huber et al. (1992)
MBWR EOS (Equation of State)
and transport properties of all fluids
except R134a
Haar-Gallagher (HG) Ammonia Haar and Gallagher (1978)
As a serious effort to evaluate the performance of possible alternatives for the existing
refrigerants, researchers at NIST developed several cycle simulation models. Ini-
tially, McLinden (1987) developed a cycle model Cycle 7 which consists of seven
cycle state points as shown in Figure 3.10.2(a).
Cycle 7 then evolved by adding a suction-line heat exchanger and by improving
the compressor model, resulting in Cycle 11 which consists of eleven state points
as shown in Figure 3.10.2(b) (Domanski and McLinden, 1992).
3
4
Temperature
7
6
Entropy
(a)
5
6
Temperature
7
8
2
1
11
10
9
Entropy
(b)
FIGURE 3.10.2 State points of two cycle models. (a) State points of Cycle 7. (b) State points
of Cycle 11.
Pannock and Didion (1991) used Cycle 11 to find the most promising binary zeotropic
mixtures for residential heat pump applications with the following hydrofluorocarbons
(HFCs): R23, R32, R125, R134a, R143a and R152a. In their work, the following
assumptions were used:
• Isenthalpic expansion.
• Suction-line heat exchanger (SLHX) was not implemented.
• Fixed cycle parameters such as the compressor speed, refrigerant pressure
drops in the heat exchangers, polytropic compressor efficiency, degree of
subcooling at the condenser outlet, degree of superheating at the com-
pressor outlet and compressor cylinder inlet and outlet pressure drop and
heat transfer coefficients.
• Heat transfer fluid inlet and outlet temperatures are based on four cooling
test conditions, A, B, C, and D, as specified in ASHRAE Standard 116
(1995).
• Heat transfer fluid is water/glycol mixture. Temperature change of the
heat transfer fluid is set to 12.2 and 8.2°C for the evaporator and condenser,
respectively, to simulate evaporator- and condenser-air flow rate 400 CFM
and 800 CFM, respectively.
After a series of simulation runs for binary mixtures of the above pure refrigerants,
the two most promising mixtures, R32/R152a and R134a/R152a, were selected
among 15 binary mixtures (R23/R32, R23/R125, R23/R143a, R23/R134a,
R23/R152a, R32/R125, R32/R143a, R32/R134a, R32/R152a, R125/R143a,
R125/R134a, R125/R152a, R143a/R134a, R143a/R152a and R134a/R152a). The
R32/R152a mixture showed the best performance. The second and third best per-
forming mixtures were R32/R134a and R125/R134a, respectively.
Figures 3.10.3 and 3.10.4 show the relative volumetric capacity and COP versus
the R32 mass fraction to those of R22. Due to the large number of graphs generated
by this study, the authors refrained from appending all diagrams and restricted
themselves to the best performing mixtures: R32/R152a and R32/R134a. The volu-
metric capacity of R32/R152a reached that of R22 at 37 and 44% of R32 mass
fraction for cooling and heating, respectively. The maximum COP of R32/R152a for
cooling and heating was reached at about 40 to 50% of R32 mass fraction. The R32
60
R32/R134a
(a)
20
15
COP Improvement over R22 [%]
R32/R152a
10
R32/R134a
5
−5
0 0.2 0.4 0.6 0.8 1.0
R32 Mass Fraction
(b)
mass fraction should be about 50% for R32/R152a to have better cooling and heating
performance than R22. One interesting result of R32/R152a is that the heating COP
has two peak COPs as the R32 mass fraction increases. Pannock and Didion (1991)
explained this phenomenon by the temperature glide of the mixture: the first maxi-
mum is reached at the point of best glide matching and the COP decreases by the
further increase in R32 mass fraction due to the excessive temperature glide of the
mixture. However, the temperature glide of the mixture starts to decrease by further
increase of the R32 mass fractions, which results in the second peak in the COP by
80
40
20
R32/R152a
0
−20
−40
−60
0 0.2 0.4 0.6 0.8 1.0
R32 Mass Fraction
(b)
10
R32/R152a
COP Improvement over R22 [%]
R32/R134a
5
−5
0 0.2 0.4 0.6 0.8 1.0
R32 Mass Fraction
(b)
FIGURE 3.10.4 Performance of R32/R152a and R32/R134a at ASHRAE Heating 47S con-
ditions (Pannock and Didion, 1991). (a) Relative volumetric capacity. (b) Relative COP.
R32/R152a for the same R32 mass fraction. R134a is not flammable, so R32/R134a
is definitely less flammable than R32/R152a. Experimental results of these two
mixtures are referred to the same study explained in Section 5.1.2.
This investigation of binary zeotropic mixtures showed their potential as alter-
native refrigerants for R22. It was also found that matching the temperature glide
of the refrigerant and air are important to show zeotropic mixtures’ potential. In
actual residential air conditioners, the indoor heat exchanger can approach cross-
counter flow but the outdoor heat exchanger cannot since the number of rows is
typically one or two. The temperature glide of mixtures then is not utilized in the
outdoor heat exchanger. Another situation to be noted is the refrigerant flow reversal
case by changing operating mode from one to another. The temperature glide of
mixtures is then not utilized in the cross-counter flow type indoor coil during the
heating mode, as will be discussed in Section 5.2.
Domanski and Didion (1993) also reviewed the performance of nine R22 alternatives,
as listed in Table 3.6. The number in the first column of this table designates each
refrigerant in Figures 3.10.5 through 3.10.7. Using these nine refrigerants, three cycle
options were investigated: drop-in, system modification, and system with SLHX.
In this model analysis, the following assumptions were made:
• Overall compressor polytropic efficiency: 0.7
• Pure crossflow heat exchangers for both evaporator and condenser
• Zero degree of superheating at the evaporator outlet
• Zero degree of subcooling at the condenser outlet
• Fixed pressure drop across heat exchangers (both evaporator and con-
denser): 34.5 kPa
• Fixed heat transfer fluid inlet and outlet temperatures (Table 3.7)
TABLE 3.6
Refrigerants Investigated
Molecular Mass Tdew [˚C]
No. Refrigerant Mass Fraction [%] [kg/kmol] at 1 atm Tglide [˚C]
(COP–COPR22)/COPR22
0.4
(Q–QR22)/QR22
0.2
0
0
−0.2
−0.1
−0.4
−0.6 −0.2
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Refrigerant/Mixture Refrigerant/Mixture
(a) (b)
FIGURE 3.10.5 Drop-in performance (Domanski and Didion, 1993). (a) Relative volumetric capacity. (b) Relative COP.
(COP–COPR22)/COPR22
(Qvol–Qvol,R22)/Qvol,R22
0.4
0.2
0
0
−0.2
−0.1
−0.4
−0.6 −0.2
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Refrigerant/Mixture Refrigerant/Mixture
(a) (b)
FIGURE 3.10.6 Performance of modified system (Domanski and Didion, 1993). (a) Relative volumetric capacity. (b) Relative COP.
0.1
(COP–COPR22)/COPR22
(Qvol–Qvol,R22)/Qvol,R22
0.4
0.2
0
0
−0.2
−0.1
−0.4
−0.6 −0.2
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Refrigerant/Mixture Refrigerant/Mixture
(a) (b)
FIGURE 3.10.7 Performance of system with SLHX (Domanski and Didion, 1993). (a) Relative capacity. (b) Relative COP.
TABLE 3.7
Heat Transfer Fluid (Air) Temperatures
Mode Cooling A Cooling B Heating 47S Heating 17L
In the drop-in performance analysis, it was assumed that there was no system
modification except the expansion device. The compressor RPM, displacement vol-
ume and heat exchanger areas were fixed as those of the R22 system. Figure 3.10.5
shows drop-in performance results. The refrigerant that has the higher saturation
pressure has the higher volumetric capacity. R32/R125 has the highest volumetric
capacity and R134a has the lowest volumetric capacity. However, the COP results
were reversed: the refrigerant that has higher saturation pressure has the lower COP.
The second analysis was the performance comparison under equal heat
exchanger loading as suggested by McLinden and Radermacher (1987). In this
analysis, the overall heat capacities of evaporator and condenser (UAe and UAc) and
average evaporation temperature (Te,avg) were selected such that the R22 saturation
temperatures at the evaporator and condenser outlet were 7.8 and 46.1˚C, respec-
tively, at cooling A test condition. The system was then modified for each refrigerant
such that the two constraints of Equations 3.50 and 3.51 were satisfied. Equation
3.50 requires the same heat exchanger area loading and Equation 3.51 requires a
specific size of the evaporator.
Qe
= const as R22 value (3.50)
UAe + UAc
Figure 3.10.6 shows the performance of the refrigerants in the modified system.
This figure shows that the modified system has a larger change in volumetric capacity
than the drop-in case since a larger heat exchanger area is required as the volumetric
capacity increases (Equation 3.50). This effect on the volumetric capacity minimized
the penalty in COP, so the modified system has less change in COP than the drop-
in case. One important result is that the COPs of all nine refrigerants are lower than
that of R22.
The third analysis was conducted for the modified system with a SLHX. This
analysis was done on cooling for ASHRAE A and the 17L heating conditions. It was
further assumed that the SLHX effectiveness was 45%. Figure 3.10.7 shows results
of an analysis for the system with SLHX. The COP of the modified system with the
SLHX improved over that without the SLHX. Most refrigerants except R32/R227ea
showed a comparable COP with R22 within 5%.
(a)
ORNL-DWG 83-1887A
10
8 Same
6 Design Capacity
% Capacity Difference Relative to R-22
(Same Nominal Design Capacity)
4
2
0 R-22
−2
−4
R-32/R-125 [60/40]
−6
−8 R-32/R-134a [30/70]
−10
R-32/R-125/R-134a [30/10/60]
−12
−14 R-32/R-227ea [35/65]
−16
−18 Heating Cooling
−20
Low-Temp High-Temp Low-Temp High-Temp
Heat Pump Ambient Conditions
(b)
ORNL-DWG 83-1886A
10
8 R-32/R-125 [60/40] R-32/R-125/R-134a [30/10/60]
6 R-32/R-134a [30/70] R-32/R-227ea [35/65]
% COP Difference Relative to R-22
4
2
0 R-22
−2
−4
−6
−8
−10
−12
−14
−16 Domanski and
−18 Heating Didion (1993) Cooling
−20
Low-Temp High-Temp Low-Temp High-Temp
Heat Pump Ambient Conditions
FIGURE 3.10.8 Performance of four refrigerant mixtures (Rice et al., 1993). (a) Relative
capacity and COP. (b) Relative heating seasonal performance factor (HSPF).
lower heating capacity but 4 to 5% higher COP for all tests conditions. R32/R227ea
is the only one that showed lower heating capacity and COPs for more test conditions
than those of R22. This degradation is because of the excessive refrigerant temper-
ature glide resulted in a heat exchanger temperature pinch that forced the mean
refrigerant temperature down in the evaporator and up in the condenser.
As shown in Figure 3.10.8 the low temperature heating capacity of the low
temperature glide mixtures is 4 to 6% lower than that of R32/R125. To investigate
the potential for improvement of the low heating performance of the small temper-
ature glide mixtures, passive and active composition shifting was examined for these
mixtures. The utilization of the suction line accumulator enables passive control
over the composition shift by varying the amount of the liquid refrigerant depending
on the ambient temperature. For the passive control, it was assumed that the mass
fraction of R32/R134a was 36/64 wt.%, based on Shiflett’s (1993) results, and the
mass fraction of R32/R125/R134a was 36/12/52 wt.%. A cycle with rectifier offers
active composition shift control. For the active control, it was assumed that the mass
fraction of R32/R134a was 48/52 wt.% and the mass fraction of R32/R125/R134a
was 48/16/36 wt.%. Details of capacity control by the composition shift are discussed
in Chapter 7. The results of passive and active composition shift are shown in Figure
3.10.9. The results showed that R32/R134a gained 5 and 16% heating capacity over
its original composition for the passive and active composition shift, respectively.
For the R32/R125/R134a, the capacity gains were 7 and 19%, respectively. However,
the COP of R32/R134a was reduced by 2 and 5% for the passive and active
composition shift, respectively. For the R32/R125/R134a, the COP loss was 7 and
19%, respectively. Figure 3.10.9(b) shows the heating seasonal performance factor
(HSPF) of three mixtures with the composition shift options for the climate region
IV and V. R32/R125 showed about 2% higher HSPF than that of R22. R32/R134a
and R32/R125/R134a also showed 2% better HSPF for the passive composition shift
and 5 to 8% better HSPF for the active composition shift than that of R22. The
predicted cooling seasonal energy efficiency ratio (SEER) of RR32/R125 was about
4% less than that of R22 while those of R32/R134a and R32/R125/R134a was 5 to
6% higher than R22.
Design Cycle Model: Mark V
Oak Ridge National Laboratory (ORNL) has developed and distributed a series of
versions of DOE/ORNL Heat Pump Design Model (HPDM) that is a hardware-based,
steady-state performance simulation model since 1978 as summarized by Rice (1997).
Ellison and Creswick (1978) developed the first ORNL version of a cycle model
based on an MIT model developed by Hiller and Glicksman (1976). Fisher and Rice
(1983) then released the first major revision called Mark I. Fisher et al. (1988) released
Mark III as a PC DOS version. In 1991, Spatz reported using a modified version of
the Mark III model to first identify the system potential of R410A, considering
transport and thermodynamic property effects on heat exchanger performance. Rice
(1987 and 1988) added charge inventory capability and Rice (1991) completed a
variable-speed model of the Mark III by adding electronically commutated motors
and design parametric capability, called Mark IV. Recently, ORNL released the Mark
V version (1994). The latest version of this model includes the variable-speed
(a)
ORNL-DWG 83 -1864
20
COP Low-Temp Heating
15 Capacity
10
5
relative to R-22
% Difference
0 R-22
−5
O O P A O P A
−10
O = Original Composition
−15 P = Passive Composition Shift (6% Wt. Shift R-32)
A = Active Composition Shift (18% Wt. Shift R-32)
−20
R-32/R-125 R-32/R-134a R-32/125/134a
[60/40] [30/70] [30/10/60]
R-22 Alternatives
(b)
ORNL-DWG 83 -1863
10
DOE Region IV
8
DOE Region V
6 (Maximum DHR)
4
relative to R-22
2
% Difference
0 R-22
−2 O O P A O P A
−4
−6 O = Original Composition
P = Passive Composition Shift (6% Wt. Shift R-32)
−8 A = Active Composition Shift (18% Wt. Shift R-32)
O
−10
R-32/R-125 R-32/R-134a R-32/125/134a R-32/R-227ea
[60/40] [30/70] [30/10/60] [35/85]
R-22 Alternatives
FIGURE 3.10.9 Performance of four refrigerant mixtures with composition shift (Rice et al.,
1993a).
Table 3.8 shows the list of refrigerants investigated as R22 alternatives. It should be
noted that the constituents of the binary mixtures were selected such that one of low
boiling point refrigerants (R143a, R32 and R125) and one of high boiling point
refrigerants (R152a, R134, R134a and R124) were mixed to minimize the operating
pressure deviation from that of R22. The refrigerant components of the ternary
mixtures were selected such that R32 and R152a were mixed with nonflammable
refrigerants (R134, R134a and R124) to reduce the overall flammability. The mass
fraction of each component was determined to maximize the cooling COP at 35˚C
outdoor conditions. The maximum COP of R143a/R134a and R125/R134a were
obtained when the mass fraction of R134a was zero. These two binary mixtures
were eliminated from the further analysis.
Table 3.9 shows the performance comparison of pure refrigerants tested at 35˚C
outdoor conditions. Results indicated that R32 would help to increase the volumetric
capacity whereas R124, R134 and R134a would decrease the volumetric capacity.
R152a and R134 would help to increase COP, whereas the others would diminish
the COP.
TABLE 3.8
List of Refrigerants Investigated
Pure Refrigerant Binary Refrigerant Ternary Refrigerant
TABLE 3.9
Performance Comparison of Pure Refrigerants Investigated
Pure Refrigerant COP COP Changes [%] Volumetric Capacity [kJ/m3]
COP
4 (2878) (3377) (3439) (3809) (3512) (3865)
(2511) (3820) (2170) (2413) (2308) (VCR kJ/m3)∗
(3735) (1940)
(2342)
(6227)
(1382)
(3441)
3
(3305)
0
R-22 R-32/ R-32/ R-32/ R-32/ R-143a/ R-134 R-32 R-143a
R-124 R-152a R-152a/ R-134a R-152
R-134a
R-32/ R-32/ R-32/ R-125/ R-143a/ R-152a/ R-134a R-124 R-152
R-152a/ R-152a/ R-134 R-152a R-124
R-124 R-134
∗Numbers in parentheses represent the volumetric capacity (VCR kJ/m3) for each refrigerant
ASHRAE Standard 116 (1995). The HSPF was evaluated for six climate regions.
Results of SPFs are shown in Figure 3.10.11. As shown, R32/R152a exhibited the
best SPFs, followed by R32/R134 and R32/R134a.
In order to examine the potential capacity control capability through the con-
centration shift, it was assumed that the active concentration control could achieve
a 20% shift. Then the SPF was evaluated for R32/R152a and R32/R124. As shown
in Figure 3.10.12, CSPF was not affected by the concentration shift but the HSPF
was increased by 10% for R32/R124 and 6% for R32/R152a.
Detailed Transient Cycle Model: HPCYCLE
Judge (1996) developed a fully implicit, distributed parameter simulation, HPCYCLE,
capable of modeling the transient and steady state aspects of an air conditioner and
heat pump. For the steady state, the simulation solved the complete continuity,
species, energy and momentum equations while transiently only the momentum
equation was omitted. The cycle simulation consisted of cycle component simula-
tions. The simulation was applied for both R22 and R32/R134a (30/70 wt.%) to
study several different system configurations transiently and at steady state.
From the experimental results, it was found that there was essentially no perfor-
mance difference between R22 and R32/R134a. In addition, most of the R32/R134a
quantities except the refrigerant temperature and the refrigerant concentration were
almost identical to R22. Therefore, these two aspects will be explained further.
Vapor Compression Cycle Fundamentals 127
SPF
6
0
CSPF HSPF I HSPF II HSPF III HSPF IV HSPF V HSPF VI
FIGURE 3.10.11 Seasonal performance factor (SPF) comparison (Radermacher and Jung,
1993).
SPF
6
0
CSPF HSPF I HSPF II HSPF III HSPF IV HSPF V HSPF VI
Figure 3.10.13 shows the refrigerant temperature of R22 and R32/R134a at the
inlet of the four-cycle components over time. Two different aspects of the refrigerant
temperature of R32/R134a as compared to R22 are as follows: The temperature
entering the evaporator closely reproduces the trends of the low side pressure since
128 Vapor Compression Heat Pumps with Refrigerant Mixtures
380
360 Condenser
Temperature (K)
340
320 STR
300
Compressor
280
260 Evaporator
240
0 60 120 180
Time (seconds)
(a) R22
380
360
Condenser
340
Temperature (K)
320
STR
300
Compressor
280
260 Evaporator
240
0 60 120 180
Time (seconds)
(b) R32/R134a
it is in the two-phase region. The temperature leaving the evaporator follows the
inlet temperature of the evaporator until the exiting refrigerant becomes superheated
vapor. The temperature of the refrigerant leaving the compressor abruptly increases
after it has leveled off. This occurs because in the beginning there is liquid refrigerant
in the compressor shell that supplies the compressor with saturated vapor. The abrupt
increase occurs when there is no longer any liquid in the compressor shell and, as
a result, the compressor receives superheated vapor. The temperature of the refrig-
erant leaving the condenser and entering the expansion device changes little when
compared to the other temperatures. This is primarily due to the large heat transfer
area of the condenser that allows the outlet refrigerant temperature to closely
approach the outdoor temperature that remains constant. Although very similar to
the plot for R22 there are a few points that make it unique. The most notable
difference is at the compressor inlet. The temperature at the compressor inlet does
not follow the temperature of the evaporator inlet, as did R22 in the early transients.
This is a result of the temperature glide that occurs during the phase change of a
zeotropic mixture. Another difference is the considerably lower condenser inlet
temperature. This is a consequence of the higher specific heat of R32/R134a. Also
noteworthy is the slightly lower evaporator inlet temperature of R32/134a as com-
pared to R22.
The refrigerant concentration is the other quantity that makes R32/R134a unique
from R22. Figure 3.10.14 shows the R32 concentration of R32/R134a versus time
Vapor Compression Cycle Fundamentals 129
0.5
0.4
R-32 Concentration (kg/kg)
Circulated Concentration
0.3
Charged Concentration
0.2
0.1
Compressor Condenser STR Evaporator
0
0 60 120 180
Time (seconds)
at six different locations. Some important aspects of the R32 concentration changes
of R32/R134a are listed below.
Initially, the refrigerant leaving the evaporator and entering the compressor is
depleted of the more volatile component, R32. This is because at start-up, the
refrigerant in the evaporator is in the two-phase region and contains less R32 than
the charged concentration. On the other hand, the vapor, initially in equilibrium with
the refrigerant in the evaporator, is rich in R32 and occupies the condenser. Hence,
the condenser acts like a reservoir of R32 and the evaporator acts like a reservoir
of R134a. As time proceeds, the two-phase, R134a-rich, refrigerant enters the com-
pressor. There the compressor shell separates the liquid from the vapor so that the
compressor compresses only R32-rich vapor. As a consequence the refrigerant leav-
ing the compressor is still relatively rich in R32. However, as the liquid leaves the
evaporator and the amount of liquid in the compressor shell decreases, the R32
concentration leaving the compressor starts to decline. This subsequently causes the
decline in the R32 concentration leaving the condenser. These processes continue
until the circulated concentration is reached at the inlet to each component.
REFERENCES
Aaron, D.A. and P.A. Domanski, 1990, “Experimentation, analysis, and correlation of refrig-
erant 22 flow through short-tube restrictors,” ASHRAE Transactions, Vol. 96, Pt. 1,
pp. 729–742.
Air-Conditioning and Refrigeration Institute, 1997, “Positive Displacement Refrigerant Com-
pressors,” Compressor Units and Condensing Units, ARI Standard 520.
Air-Conditioning and Refrigeration Institute, 1999, “Method for Presentation of Compressor
Performance Data,” ARI Standard 540.
130 Vapor Compression Heat Pumps with Refrigerant Mixtures
Air-Conditioning and Refrigeration Institute, 1993, “Results from ARI’s R22 Alternative
Refrigerants Evaluation Program (AREP),” Compressor Calorimeter, System Drop-
in, Soft Optimized System Test Reports, Vol. 4, Report #157.
Alefeld, G., 1987, “Efficiency of Compressor Heat Pumps and Refrigerators derived from the
Second Law of Thermodynamics,” International Journal of Refrigeration, Vol. 10,
No. 6, pp. 331–341.
Alefeld, G. and R. Radermacher, 1994, Heat Conversion Systems, CRC Press, Boca Raton, FL.
American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc., 1995,
“Methods of Testing for Seasonal Efficiency of Unitary Air-conditioners and Heat
Pumps,” ASHRAE ANSI/ASHRAE 116-1995.
Bennedict, M., G.B. Webb, and L.C. Rubin, 1940, “An Empirical Equation for the Thermo-
dynamic Properties of Light Hydrocarbons and Their Mixtures,” J. Chem. Phys.,
Vol. 8, pp., 334–348.
Blasius, H., 1913, “Das Ahnlichkeitsgesetz bei Reibungsvorgangen in Flussikeiten,” Forsch.
Arb. Ing. Wes., 34, Berlin, Germany.
Bowman, R., A. Mueller, and W. Nagle, 1940, “Mean Temperature Difference in Design,”
ASME Transactions, Vol. 62, No. 283.
Chawalowski, M., D.A. Didion, and P.A. Domanski, 1989, “Verification of evaporator models
and analysis of performance of an evaporator coil,” ASHRAE Transactions, Vol. 95,
Pt. 1, pp. 1229–1236.
Chen, J.C., 1966, “A Correlation for Boiling Heat Transfer to Saturated Fluids in Vertical
Flow,” Ind. Eng. Chem. Proc. Design Dev., Vol. 5, No. 3, pp. 322–339.
Chen, S.L., F.M. Gerner, and C.L. Tein, 1987, “General Thin Film Condensation Correlations,”
Journal of Experimental Heat Transfer, Vol. 1, pp. 93–107.
Conde, M.R. and P. Sutter, 1992, “The Simulation of Direct Expansion Evaporator Coils for Air-
Source Heat Pumps,” International Congress of Refrigeration, Montreal, pp. 1459–1463.
David, S.G., 1994, “Results of Soft Optimized System Tests in ARI’s R22 Alternative Refrig-
erants Evaluation Program,” Proceedings of the 1994 International Refrigeration
Conference at Purdue, West Lafayette, pp. 7–12.
David, S.G. and M.S. Menzer, 1993, “Results of Compressor Calorimeter Tests in ARI’s R22
Alternative Refrigerants Evaluation Program,” Proceedings of the 1993 International
CFC and Halon Alternatives Conference.
DeSantis, R., F. Gironi, and L. Marrelli, 1976, Ind. Eng. Chem. Fund., Vol. 15: 183.
Dobson, M.K., J.C. Chato, D.K. Hinde, and S.P. Wang, 1994, “Experimental Evaluation of
Internal Condensation of Refrigerants R12 and R134a,” ASHRAE Transactions, Vol.
100, Pt. 1, pp. 744–754.
Dobson, M.K., J.C. Chato, J.C. Wattelet, J.A. Gaibel, M. Ponchner, P.J. Kenney, R.L. Shimon,
T.C. Villaneuva, N.L. Rhines, K.A. Sweeney, D.G. Allen, and T.T. Hershberger, 1994,
“Heat Transfer and Flow Regimes during Condensation in Horizontal Tubes,” ACRC
Report 57.
Dobson, M.K. and J.C. Chato, 1998, “Condensation in Smooth Horizontal Tubes,” Journal
of Heat Transfer, Vol. 120, pp. 193–213.
Domanski, P.A., 1989, “HVSIM — An evaporator simulation model accounting for refrigerant
and one dimensional air distribution,” NISTIR 89-4133.
Domanski, P.A., 1995, “Theoretical Evaluation of the Vapor Compression Cycle With a
Liquid-Line/Suction-Line Heat Exchanger, Economizer, and Ejector,” NIST Report,
NISTIR 5606.
Domanski, P.A. and M.O. McLinden, 1992, “A Simplified Cycle Simulation Model for the
Performance Rating of Refrigerants and Refrigerant Mixtures,” International Journal
of Refrigeration, Vol. 15, No. 2, pp. 81–88.
Vapor Compression Cycle Fundamentals 131
Domanski, P.A. and D.A. Didion, 1993, “Thermodynamic Evaluation of R-22 Alternative Refrig-
erants and Refrigerant Mixtures,” ASHRAE Transactions, Vol. 99, Pt. 2, pp. 636–648.
Ellison, R.D. and F.A. Creswick, 1978, “A Computer Simulation of Steady State Performance
of Air-to-Air Heat Pumps,” ORNL/Con-16.
Fischer, S.K. and C.K. Rice, 1983, “The Oak Ridge Heat Pump Models: I. A Steady-State
Computer Design Model for Air-To-Air Heat Pumps,” ORNL/CON-80/Rl.
Fisher, S.K. et al., 1988, “The Oak Ridge Heat Pump Model: Mark III Version Program
Documentation,” ORNL/TM-10192.
Gnielinski, V., 1976, “New Equation for Heat and Mass Transfer in Turbulent Pipe and
Channel Flow,” Int. Chemical Engineer, Vol. 16, pp. 359–368.
Gray, D.L. and R.L. Webb, 1986, “Heat Transfer and Friction Correlations for Plate Finned-
Tube Heat Exchangers having Plain Fins,” Proc. of Eighth Int. Heat Transfer Con-
ference, San Francisco, pp. 123–131.
Gungor, K.E. and R.H.S. Winterton, 1986, International Journal of Heat and Mass Transfer,
Vol. 29, No. 3, pp. 351–358.
Haar, L. and J. Gallagher, 1978, “Thermodynamic Properties of Ammonia,” J. Physical
Chemical Ref. Data, Vol. 7, pp. 635–792.
Haar, L., J.S. Gallagher, and G.S. Kell, 1984, NBS/NRC Steam Tables, Hemisphere Publishing
Co., New York, pp. 301–316 and 635–792.
Herold, K.E., 1989, “Performance Limits for Thermodynamic Cycles,” Proceedings of ASME
Winter Annual Meeting, AES-Vol. 8, pp. 41–45.
Herold, K.E., R. Radermacher, and S.A. Klein, 1996, Absorption Chillers and Heat Pumps,
CRC Press, Boca Raton, FL.
Hiller, C.C. and L.R. Glicksman, 1976, “Improving Heat pump Performance via Compressor
Capacity Control-Analysis and Test,” Vol. 1 & 2, MIT Laboratory Report No. MIT-EL
76-001.
Huber, M.L., D.G. Friend, and J.F. Ely, 1992, “Prediction of the Thermal Conductivity of
Refrigerants and Refrigerant Mixtures,” Fluid Phase Equilibria, Vol. 80, pp. 249–261.
Huber, M.L. and J.F. Ely, 1992, “Prediction of the Viscosity of Refrigerants and Refrigerant
Mixtures,” Fluid Phase Equilibria, Vol. 80, pp. 239–248.
Hwang, Y. and R. Radermacher, 1998, “Emerging Refrigerants,” Proceedings of 1998 IIR
Conference, pp. 1.1–1.14.
Incropera, F.P. and D.P. Dewitt, 1990, Fundamentals of Heat and Mass Transfer, John Wiley
& Sons, New York, NY, pp. 645–678.
Jaster, H. and P.G. Kosky, 1976, “Condensation in a Mixed Flow Regime,” International
Journal of Heat and Mass Transfer, Vol. 19, pp. 95–99.
Judge, J., 1996, “A Transient and Steady State Study of Pure and Mixed Refrigerants in a
Residential Heat Pump,” Ph.D. Dissertation, University of Maryland, College Park, MD.
Judge, J. and R. Radermacher, 1997, “A Heat Exchanger Model for Mixtures and Pure
Refrigerant Cycle Simulations,” International Journal of Refrigeration, Vol. 20, No. 4,
pp. 244–255.
Jung, D.S., M. McLinden, R. Radermacher, and D.A. Didion, 1989, “A Study of Flow Boiling
Heat Transfer with Refrigerant Mixtures,” International Journal of Heat Transfer,
Vol. 32, No. 9, pp. 1751–1764.
Jung, D.S. and R. Radermacher, 1989, “Prediction of Pressure Drop during Horizontal Annular
Flow Boiling of Pure and Mixed Refrigerant,” International Journal of Heat Transfer,
Vol. 32, No. 12, pp. 2435–2446.
Jung, D.S., R. Radermacher, and D.A. Didion, 1989, “Horizontal Flow Boiling Heat Transfer
Experiments with a Mixture of R22/R114,” International Journal of Heat Transfer,
Vol. 32, No. 1, pp. 131–145.
132 Vapor Compression Heat Pumps with Refrigerant Mixtures
Kandlikar, S.G., 1991, “Correlating Flow Boiling Heat Transfer Data in Binary Systems,”
Proceedings of ASME National Heat Transfer Conference.
Kenney, P. et al., 1994, “Condensation of a Zeotrophic Refrigerant R-32/R-125/R-134a
(23%/25%/52%) in a Horizontal Tube.” ACRC Report. TR-062.
Kim, Y. and D.L. O’Neal, 1993, “An experimental study of two-phase flow of HFC-134a
through short tube orifices,” ASHRAE Transactions, Vol. 98, Pt. 1, pp. 122–124.
Kim, Y. and D.L. O’Neal, 1994a, “Two-phase flow of Refrigerant-22 through short tube
orifices,” ASHRAE Transactions, Vol. 100, Pt. 1, pp. 323–334.
Kim, Y. and D.L. O’Neal, 1994b, “A semi-empirical model of two-phase flow of HFC-134a
through short tube orifices,” AES Vol. 29, ASME Winter Annual Meeting.
Kornhauser, A.A., 1993, “Physically Realistic Closed Form Modes for Flashing Flow in Short
Tubes,” Proceedings of the 28th Inter-Society Energy Conversion Engineering Con-
ference.
Kuehl, S. and V.W. Goldschmidt, 1992, “Flow of R-22 through short tube restrictors,” ASHRAE
Transactions, Vol. 78, pp. 59–64.
Lio, M., A. Doria, and M. Bucciarelli, 1994, “Numerical Analysis of the Dynamics of Reed
Valves Taking into Account the Acoustic Coupling with the Fluid,” Purdue Compres-
sor Conference, Vol. 2, pp. 229–234.
MacArthur, J.W., 1984, “Theoretical analysis of the dynamic interactions of vapor compres-
sion heat pumps,” Energy Conversion Management, Vol. 24, pp. 49–66.
McLinden, M.O., 1987, “Theoretical Vapor Compression Cycle Model, Cycle 7,” National
Bureau of Standards, Gaithersburg, MD.
McLinden, M.O. and R. Radermacher, 1987, “Methods for Comparing the Performance of
Pure and Mixed Refrigerants in the Vapor Compression Cycle,” International Journal
of Refrigeration, Vol. 10, No. 6, pp. 318–325.
McLinden, M.O., J.G. Gallagher, H.L. Huber, and G. Horrison, 1993, Refprop Refrigerant
Properties Database: Capabilities, Limitations, and Future Directions, 1982, Proceed-
ings of ASHRAE/NIST, Refrigerant Conference, Gaithersburg, MD.
Mei, V.C., 1988, “Short-Tube Refrigerant Flow Restrictors,” ASHRAE Transactions, Vol. 88,
Pt. 2, pp. 157–169.
Murphy, W.E. and Goldschmidt, V.W., 1985, “Cyclic Characteristics of a typical Residential
Air Conditioner-Modeling of Start-Up Transients,” ASHRAE Transactions, Vol. 92,
Pt. 1A, pp. 186–202.
Ng, E., A. Tramschek, and J. Maclaren, 1976, “Computer Simulation of Reciprocating Com-
pressor Using a Real Gas Equation of State,” International Journal of Refrigeration,
Vol. 15, pp. 33–42.
Nyers, J. and G. Stoyan, 1994, “A Dynamical Model Adequate for Controlling the Evaporator
of a Heat Pump,” Intl. J. of Refrigeration, Vol. 17, p. 2.
ORNL Website, 1998, DOE/ORNL Heat Pump Design Model.
Oskarsson, S.P., K.I. Krakow, and S. Lin, 1990, “Evaporator models for operation with dry,
wet, and frosted finned surfaces Part II: Evaporator Models and Verification,”
ASHRAE Transactions, Vol. 96, Pt. 1, pp. 373–380.
Pannock, J. and D. Didion, 1991, “The Performance of Chlorine-Free Binary Zeotropic
Refrigerant Mixtures in a Heat Pump,” NISTIR 4748, National Institute of Standards
and Technology, Gaithersburg, MD.
Payne, W.V. and D.L. O’Neal, 1998, “Mass Flow Characteristics of R407C Through Short
Tube Orifices,” ASHRAE Transactions, Vol. 104, Part 1A, pp. 197–209.
Payne, W.V. and D.L. O’Neal, 1999, “Multiphase Flow of Refrigerant R410A Through Short
Tube Orifices,” ASHRAE Transactions, Vol. 105, Part 2.
Vapor Compression Cycle Fundamentals 133
Perez-Segarra, C.D., F. Escanes, and A. Oliva, 1994, “Numerical Study of the Thermal and
Fluid-Dynamic Behavior of Reciprocating Compressors,” Purdue Compressor Con-
ference Proceedings, pp. 145–150.
Rabas, T.J. and P.W. Eckels, 1985, “Heat Transfer and Pressure Drop Performance of Segmented
Extended Surface Tube Bundles,” ASME Journal of Heat Transfer, Vol. 75, pp. 1–8.
Radermacher, R. and D.S. Jung, 1993, “Theoretical Analysis of Replacement Refrigerants
for R-22 for Residential Uses,” ASHRAE Transactions, Vol. 99, Pt. 1, pp. 333–343.
Recktenwald, G.W., J.W. Ramsey, and S.V. Patankar, 1986, “Predictions of Heat Transfer in
Compressor Cylinders,” Purdue Compressor Conference Proceedings, West Lafay-
ette, pp. 159–174.
Rice, C.K., 1987, “The Effect of Void Fraction Correlation and Heat Flux Assumption on
Refrigerant Charge Inventory Predictions,” ASHRAE Transactions, Vol. 93, Pt. 1, pp.
341–367.
Rice, C.K., 1988, “Capacity Modulation Component Characterization and Design Tool Devel-
opment,” Proceedings of the 2nd DOE/ORNL Heat Pump Conference: Research and
Development on Heat Pumps for Space Conditioning Applications, CONF-8804100,
Washington D.C., pp.23–33.
Rice, C.K., 1991, The ORNL Modulating Heat Pump Design Tool — User’s Guide,
ORNL/CON-343.
Rice, C.K., 1993a, “Influence of HX Size and Augmentation on Performance Potential of Mix-
tures in Air-To-Air Heat Pumps,” ASHRAE Transactions, Vol. 99, Pt. 2, pp. 665–679.
Rice, C.K., 1993b, “Compressor Calorimeter Performance of Refrigerant Blends-Comparative
Methods and Results for a Refrigerator/Freezer Application,” ASHRAE Transactions,
Vol. 99, Pt. 1, pp. 1447–1466.
Rice, C.K., L.S. Wright, and P.K. Bansal, 1993, “Thermodynamic Evaluation Model for R22
Alternatives in Heat Pumps-Initial Results and Comparisons,” 2nd International Con-
ference on Heat Pumps in Cold Climates, New Brunswick, Canada, Canada Research,
pp. 81–96.
Rice, C.K. and W.L. Jackson, 1994, “PUREZ — The Mark V ORNL Heat Pump Design
Model for Chlorine-Free, Pure and Near-Azeotropic Refrigerant Alternatives,”
Preliminary Documentation Package, Oak Ridge National Laboratory.
Rice, C.K., 1997, “DOE/ORNL Heat Pump Design Model, Overview and Application to R22
Alternatives,” 3rd International Conference on Heat Pumps in Cold Climates, Nova
Scotia, Canada.
Sami, S.M., J. Schnotale, and J.G. Smale, 1992, “Prediction of the Heat Transfer Character-
istics of R22/R152a/R114 and R22/R152a/R124,” ASHRAE Transactions, Vol. 98,
Pt. 2, pp. 51–58.
Sand, J.R., E.A. Vineyard, and R.J. Nowark, 1990, “Experimental Performance of Ozone Safe
Alternative Refrigerants,” ASHRAE Transactions, Vol. 96, Pt. 2, pp. 173–182.
Shaffer, R.W. and W.D. Lee, 1976, “Energy Consumption in Hermetic Refrigerator Compres-
sors,” Proceedings of the 1976 Purdue Compressor Technology Conference at Purdue,
West Lafayette.
Shah, M.M., 1979, “A General Correlation for Heat Transfer During Film Condensation Inside
Pipes,” Journal of Heat and Mass Transfer, Vol. 22, pp. 547–556.
Shah, M.M., 1982, “Chart Correlation for Saturated Boiling Heat Transfer: Equations and
Further Studies,” ASHRAE Transactions, Vol. 88, Pt. 1, pp. 185–196.
Sheffield, J.W., R.A. Wood, and H.J. Sauer, 1989, “Experimental Investigation of Thermal
Conductance of Finned Tube Contacts,” Experimental Thermal and Fluid Science,
Elsevier Science Publishing Co. Inc., New York, NY, pp. 107–121.
134 Vapor Compression Heat Pumps with Refrigerant Mixtures
Shiflett, M.B., 1993, “HCFC-22 Alternatives for Air Conditioning and Heat Pumps,” ASHRAE
Seminar Presentation, DuPont Fluorochemicals, Wilmington, DE.
Spatz, M.W., 1991, “Performance of Alternative Refrigerants From a System Perspective,”
Proceedings of International CFC and Halon Alternative Conference, Baltimore, MD,
pp. 352–363.
Stephan, K. and M. Abdelsalam, 1980, “Heat Transfer Correlations for Natural Convection
Boiling,” International Journal of Heat Transfer, Vol. 23.
Sweeny, K.A., 1996, “The Heat Transfer and Pressure Drop Behavior of a Zeotropic Refrig-
erant Mixture in a Microfin Tube,” M.S. Thesis, University of Illinois at Urbana-
Champaign.
Sweeny, K.A. and J.C. Chato, 1996, “The Heat Transfer and Pressure Drop Behavior of a
Zeotropic Refrigerant Mixture in a Microfin Tube,” ACRC Technical Report 95,
University of Illinois at Urbana-Champaign.
Tandon, T.N., H.K. Varma, and C.P. Gupta, 1986, “Generalized Correlation for Condensation
of Binary Mixtures Inside a Horizontal Tube,” International Journal of Refrigeration,
Vol. 9, pp. 134–136.
Traviss, D.P., W.M. Rohsenow, and A.B. Baron, 1973, “Forced-Convection Condensation
Inside Tubes: A Heat Transfer Equation for Condenser Design,” ASHRAE Transac-
tions, Vol. 79, Pt. 1, pp. 157–165.
Vineyard, E.A., J.R. Sand, and R.J. Nowark, 1989, “Experimental Performance of Ozone-
Safe, Alternative Refrigerants,” ASHRAE Technical Data Bulletin, Vol. 6, No. 1, pp.
66–75.
Vineyard, E.A., J.R. Sand, and T.G. Statt, 1989, “Selection of Ozone-Safe, Nonazeotropic
Refrigerant Mixture for Capacity Modulation in Residential Heat Pumps,” ASHRAE
Transactions, Vol. 95, Pt. 1, pp. 34–46.
Webb, R.L., 1990, “Airside Heat Transfer Correlations for Flat and Wavy Plate Fin-and-Tube
Geometries,” ASHRAE Transactions, Vol. 96, Pt. 2, pp. 445–449.
Welsby, P., S. Devotta, and P.J. Diggory, 1988, “Steady- and Dynamic-State Simulations of
Heat Pumps, Part II: Modeling of a Motor Driven Water-to-Water Heat Pump,”
Applied Energy, Vol. 31, pp. 239–262.
4 Methods for Improving
the Cycle Efficiency
Several options for the incremental efficiency improvement of the vapor compression
cycle are available. In addition, based especially on the understanding of the refrig-
erant properties and how they affect cycle performance, very significant additional
improvements are available when the cycle is modified and new components are
added. Three measures are discussed in this chapter, the suction line to liquid line
heat exchanger, the economizer and the isentropic expansion device. While these
devices also benefit cycles with pure fluids, they have unique effects for mixtures
when considered in conjunction with the temperature glide. For mixtures alone, one
additional option exists — the three-path evaporator.
135
136 Vapor Compression Heat Pumps with Refrigerant Mixtures
d Water
Chille
t
hea
Evaporation without
per
Superheat
Su
Evaporation with
Superheat
FIGURE 4.1.1 The need for superheat can lower the evaporation temperature.
ln (P)
FIGURE 4.1.2 The need for superheat can require a lower suction pressure and thus lower
compressor capacity.
ln (P)
4a 4
T5
5a
T5a
ln (P)
4a 4
TB 2 TA
5a
TB
FIGURE 4.1.4 Effects of subcooling for two refrigerant mixtures having different isotherms.
When zeotropic mixtures are used, the subcooling has two effects on the evap-
orator inlet conditions. As with pure fluids, subcooling reduces the enthalpy and
increases the capacity of the evaporating fluid. At the same time, as a result of the
glide, the inlet temperature to the evaporator is reduced, as shown in Figure 4.1.3.
With increased subcooling from 4 to 4a, the evaporator inlet state point changes
from 5 to 5a. As can be seen in Figure 4.1.3, the temperature T5a is lower than T5.
However, depending on the slope of the two-phase isotherm in this portion of the
two-phase region, the decreased inlet temperature effect can be more or less pro-
nounced. Figure 4.1.4 shows examples of isotherms for two different refrigerant
mixtures, A and B. The isotherm TA of refrigerant mixture A has an almost horizontal
slope in the low quality region close to the saturated liquid line. TB and TB2 of mixture
B have a relatively steep slope. Thus, the effect of subcooling on reducing the
evaporator inlet temperature as shown in Figure 4.1.3 is much less pronounced for
a mixture that exhibits an isotherm of type TA. There is essentially no difference in
temperature between points 5a and 5. For an isotherm of type TB, however, point 5a
is located on the lower temperature isotherm TB while point 5 lies on TB2, TB < TB2.
This reduction of the evaporator inlet temperature has interesting consequences.
The suction pressure can be increased slightly in order to maintain a constant
evaporator capacity as long as the glide is still well matched. The compressor power
138 Vapor Compression Heat Pumps with Refrigerant Mixtures
4 Condenser 2
3 Compressor
1a (a)
4a Suction Line to
Expansion Liquid Line Heat
Valve Exchanger
Evaporator
5, 5a 1
ln (P)
T4a T4
4a 4 3 2
(b)
5a 5 1 1a
h 5a h1 h
FIGURE 4.2.1 (a) Schematic of vapor compression cycle with suction line to liquid line heat
exchanger. (b) The state points according to (a) on an ln(P)-h diagram. The arrow indicates
the direction of heat transfer.
Methods for Improving the Cycle Efficiency 139
difference h4 to h4a from the subcooled liquid to the suction gas (h1a – h1). Thus, the
suction gas is superheated and its enthalpy increased by
while the inlet enthalpy to the evaporator is lowered by the same qslhx. This transfer
of heat enables one to transfer cooling capacity de facto from the evaporator satu-
ration temperature level via the suction heat exchange to the temperature level of
the (warmer) suction gas as it is being superheated.
This benefit comes with a challenge. The suction gas is superheated and the
state point for the superheated gas moves into a region of the ln(P)-h diagram where
the isentropes become flatter and the compressor work input increases. Thus, the
benefit of the suction to liquid line heat exchanger depends strongly on whether or
not the benefit outweighs the penalty of increased compressor work. This strongly
depends on the working fluid used.
As a rule, the following consideration applies: The higher the specific heat
capacity of the refrigerant molecules, the higher the benefit of the suction to liquid
line heat exchange. A high specific heat capacity of the refrigerant has two effects:
The temperature increase during the compression process is low, the compressor
compresses a gas that heats up less and the compressor work does not increase as
much because of the lack of superheat. However, during the expansion process a
relatively large amount of refrigerant must evaporate to cool the remaining liquid
to the evaporator saturation temperature.
This situation is better understood when the ln(P)-h diagram for a refrigerant of
high specific heat capacity is compared to a refrigerant of low specific heat. Figure
4.2.2 shows two diagrams side by side; (a) shows the diagram of a low capacity
refrigerant, (b) that of a high heat capacity fluid. Obviously, the vapor dome is much
tilted more to higher enthalpies for the high heat capacity fluid as shown in Figure
4.2.2(b). As a consequence, the expansion process produces much more vapor and
ln (P) ln (P)
h h
(a) (b)
FIGURE 4.2.2 (a) ln(P)-h diagram for a refrigerant of low specific heat capacity. (b) ln(P)-
h diagram for a refrigerant of large specific heat capacity.
140 Vapor Compression Heat Pumps with Refrigerant Mixtures
less liquid. The evaporator capacity is reduced considerably. Furthermore, the com-
pression process stays very close to the saturated vapor line and the compressed
refrigerant vapor is almost not superheated at all. In fact, for some refrigerants, the
state point of the compressed vapor can be located in the two-phase range. In this
case, the penalty in terms of increased compressor work is small or even desirable
to avoid a two-phase condition at the compressor outlet. The benefit of improved
evaporator capacity is very important because otherwise the refrigerant would not
be very efficient.
For refrigerants of low specific heat capacity, the suction to liquid line heat
exchanger can have the opposite effect. Examples for fluids of low specific heat
capacity are ammonia, R22, R32 and other fluids with a small number of atoms in
the molecule. Refrigerants of high specific heat capacity are R123, R134a and other
such molecules with a large number of atoms within the molecule. When the number
of atoms is small, any addition of energy can only be transferred into translational
movement, increasing the temperature. When the number of atoms in a molecule is
large, additional energy is first transferred into internal degrees of freedom, such as
oscillations and rotations of the atoms within the molecule, and very little energy
into the translational movement. The resulting increase in temperature is small.
Figure 4.2.3 shows a plot of COP improvement when a suction to liquid line
heat exchanger is used with three levels of heat exchange efficiency, ηhx = 0.0
(baseline), ηhx = 0.5 and ηhx = 1.0. All COP values refer to the Carnot COP and all
values are evaluated at the same reduced temperature. The refrigerants on the x-axis
are sorted according to their vapor specific heat capacity beginning with the lowest
specific heat capacity. For low heat capacity fluids, the suction to liquid line heat
exchanger has an obvious detrimental effect. This is followed by a range where there
is no particular benefit and the higher specific heat capacity of the refrigerant
becomes more pronounced in the improvement resulting from the heat exchanger.
It should be noted that the fluids with a high specific heat capacity reach higher
fractions of the Carnot COP. From this point of view, refrigerants with high molecular
weights should be preferred. The trade-off is that the vapor pressure tends to be
very low for these fluids and system size increases considerably. Thus, the low
pressure refrigerants are best suited for large capacity machines.
For zeotropic mixtures, the suction to liquid line heat exchanger has an additional
effect as shown in the ln(P)-h diagram of Figure 4.2.4. By reducing the enthalpy at
the evaporator inlet, the temperature of the entering mixture is reduced because of
the influence of the glide. T5a is lower than T5, assuming that the suction pressure
level did not change because of the heat exchanger. This decrease in temperature at
the evaporator inlet increases the evaporator capacity because the temperature dif-
ference that drives the heat transfer is increased.
Alternatively, the suction pressure can be increased slightly and the penalty of
increased compressor work somewhat reduced. Depending on the size of the glide
and its linearity, this effect can be significant (steep glide close to the liquid line)
or nonexistent (flat glide close to the liquid line). See Figures 4.1.3 and 4.1.4 with
examples in Chapter 5 where experimental results are discussed.
COPhx/COPC
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
R-744
NH3
R-32
R-14
R-23
ηhor = 0.0
R-C270
R-22
R-13B1
R-152a
R-21
ηhx = 1.0
R-290
R-143a
R-12
R-143
R-134a
R-142b
R-125
R-11
R-134
R-600a
R-124
R-141b
E-134
R-115
R-600
R-123
FIGURE 4.2.3 Plot of COP improvement using suction to liquid line heat exchanger (Domanski, 1995).
R-123a
R-245cb
R-114
R-236ea
R-227ea
I-C5
R-113
N-C5
R-218
E-245
R-C31B
142 Vapor Compression Heat Pumps with Refrigerant Mixtures
ln (P)
4a 4 3
2
Transfer
of Heat
5a 5 1 1a
T5a T5 T1
FIGURE 4.2.4 For zeotropic mixtures, the suction to liquid line heat exchanger lowers the
evaporator inlet temperature from T5 to T5a.
3
Condenser
Expansion 2
Valve 1 4 7 Compressor
(a)
Expansion 5
Valve 2 1
Evaporator
6
ln (P)
3 2
5 4 78 9 (b)
6 6a 1
FIGURE 4.3.1 (a) State points superimposed on a ln(P)-h diagram. (b) Schematic of a vapor
compression cycle with economizer.
Q evap,Eco = m 3 (1 − x 4 )( h1 − h5 ) (4.3)
where x4 is the fraction of the total mass flow rate that is extracted on the intermediate
pressure level.
The difference in evaporator capacity can be calculated as
where
x4 is the vapor quality at point 4
x4 can be expressed in terms of enthalpy as Equation 4.6
After using Equation 4.6 for x4, the following final expression Equation 4.7 is
obtained:
h − h1
∆Q = m 3 (h3 − h5 ) 7 (4.7)
h7 − h5
144 Vapor Compression Heat Pumps with Refrigerant Mixtures
Mass Fraction
T at Condenser
Exit
54 7
6
1
0 1.0
X
3
4 Condenser 2
Power Compressor
Power Input
5a Evaporator
1
ope
ln (P)
Isentr
3 2
4
5a
5 1
∆W = ∆Qevap h
FIGURE 4.4.1 Vapor compression cycle with expander to recover expansion work.
∆W = m ( h5 − h5 a ) (4.8)
This same amount of energy, since it is extracted from the refrigerant stream, also
represents an increase of evaporator cooling capacity as shown in Figure 4.4.1. The
new evaporator capacity is now calculated as
146 Vapor Compression Heat Pumps with Refrigerant Mixtures
This benefit is available even when the work produced by the expander is not utilized.
However, to maximize the benefit of an expander, both the increase in evaporator
capacity as well as the reduction in compressor work should be utilized. Since the
efficiency of the heat pump is defined as evaporator capacity over net compressor
power input, this parameter is improved by two contributions.
The benefit of using the expander depends strongly on the overall efficiency of
the device. Current designs reach an overall efficiency gain of the heat pump on the
order of 5%. When refrigerant mixtures are used, the expander causes the evaporator
inlet temperature of the refrigerant to be reduced, an additional benefit.
Condenser
Compressor
Water
T5a < T5
ln (P)
T5a T5
4a 4
FIGURE 4.5.2 The three-path evaporator lowers the evaporator inlet temperature. Alterna-
tively, the suction pressure can be raised.
capacity is increased and the inlet temperature lowered, also beneficial for pure
fluids. In the case of the three-path heat exchanger, no benefit is derived for pure
fluids because there is no net gain in evaporator capacity while there is a reduction
in the evaporator inlet temperature of the two-phase refrigerant stream for mixtures.
As discussed above, this benefit again can be translated into a somewhat higher
compressor suction pressure when the conditions in the evaporator allow. This
condition is fulfilled when the evaporator outlet temperature after the increase in
suction pressure is still lower than the incoming water temperature.
The main difference between improved subcooling at the condenser outlet and
the three-path evaporator is the following: When the liquid refrigerant is subcooled
in the condenser, the heat of subcooling is rejected to the exterior where it leaves
the cycle. With the three-path evaporator, the heat of subcooling remains within the
cycle. Therefore, no capacity benefit is achieved but a secondary benefit (reduced
evaporator inlet temperature) is acquired in the case of zeotropic mixtures. The
benefit of the three-path evaporator is available in addition to that of the suction to
the liquid line exchanger and there is no operational penalty associated with it.
Ideally, and especially for mixtures with high specific heat capacities, a designer
should consider employing all three measures to the maximum beneficial extent: sub-
cooling, suction line to liquid line exchange and the three-path evaporator configuration.
REFERENCES
Domanski, P.A., 1995, “Theoretical Evaluation of the Vapor Compression Cycle With a
Liquid-Line/Suction-Line Heat Exchanger, Economizer, and Ejector,” NIST Report,
NISTIR 5606.
Radermacher, R., 1988, “Vapor Compression Heat Pump with Non-azeotropic Refrigerant
Mixture and Absorber Desorber Heat Exchange,” U.S. Patent No. 4,724,679.
Radermacher, R. and D.S. Jung, 1993, U.S. Patent No. 5,207,077.
Vakil, H.B., “Means and Methods for the Recovery of Expansion Work in Vapor Compression
Cycle; Subcooling in Tube Parallel to Evaporator,” U.S. Patent No. 4,304,099.
148 Vapor Compression Heat Pumps with Refrigerant Mixtures
Vakil, H.B., “Vapor Compression Cycle Device with Multicomponent Working Fluid Mixture;
Receiver Behind Condenser and in Between the Two Evaporators,” U.S. Patent No.
4,283,919.
Vakil, H.B., “Vapor Compression Cycle Device with Multicomponent Working Fluid Mixture
and Method; Receivers Behind Condenser and Evaporators,” U.S. Patent No.
4,217,760.
Vakil, H.B. and J.W. Flock, “Vapor Compression Cycle Device with Multicomponent Working
Fluid Mixture and Method; Pseudo Counter Flow Heat Exchanger,” U.S. Patent No.
4,281,890.
Vakil, H.B., “Vapor Compression Cycle Device with Multicomponent Work Fluid Mixture;
Receiver and Separate Separation Device Behind Cond. Accumulator,” U.S. Patent
No. 4,179,898.
149
Vineyard and Conklin, 1991; Conklin and Vineyard, 1991; Vineyard, Conklin, and
Brown, 1993).
The breadboard heat pump I is shown schematically in Figure 5.1.1. Water was
selected as a heat source- and sink-fluid because capacity measurements are simple
and because counterflow and intracycle heat exchange can be achieved. A constant
evaporator heat flux was desired to compare the different refrigerants (McLinden
and Radermacher, 1987). Because the refrigerants to be examined had differing
properties, achieving this goal required use of a variable speed compressor. An open
two piston reciprocating compressor was selected together with a variable speed
motor.
Figure 5.1.2 shows a side view of the setup. The condenser heat exchangers are
coiled aluminum extrusions, as shown in Figure 5.1.3. The evaporator heat exchang-
ers (Figure 5.1.4) were chosen to allow for temperature measurements all along the
flow channel. Two 6 m long extruded copper tubes were selected for the evaporator
heat exchanger. The heat exchanger tubes consisted of only two passages, a center
path with “microfins” on the walls and a set of “rectangular” channels around this
inner tube. The fins and channels are spiraled, changing top to bottom every 150
mm. A third passage was added on the outside of the evaporator to obtain a heat
exchanger for a third fluid, thus allowing for heat exchange between the evaporator
refrigerant and the condensed liquid refrigerant throughout the length of the evap-
orator. The two heat exchanger tubes are connected in series, resulting in a total
evaporator heat exchanger length of 12 m.
Thermocouples were soldered to the outside of the evaporator heat exchangers
every 0.6 m, thereby allowing measurement of the temperature profile in the outer-
most channel. For the rectangular channels, measurement was limited to the crossover
between 3 m sections of the 12 m heat exchanger. This resulted in five thermocouples
for these passages. For the center passage, thermocouples are located every 0.6 m
in a sealed 3.2 mm thin wall, stainless steel tube. These thermowells are mounted
in the center of the innermost channel to measure the in-stream temperature of the
flowing fluid. Additional measurement devices were several other thermocouples,
thermopiles and pressure transducers located at critical points in the cycle.
Another important feature of the breadboard heat pump is a liquid/vapor-sepa-
rating accumulator in front of the compressor serving as a refrigerant storage vessel
as well (Figure 5.1.5).
The capacity of both heat exchangers was measured using redundant methods for
verification and the compressor power was measured using a torque meter and
.
m1
T4
Condenser Pump Refrigerant
T3
T6, P9
Flow Meter Water/Glycol
T5, P5
.
Oil Separator Electric Heater
m2 T7, P7
T11 T12 . .
Qcond = Q1 (T4−T3)/(T1−T2)
Qcond = m2 (h5−h6)
.
m 3
. . = Q. + Q. + Q.
Evaporator Capacity:
Qpump . = m. C (T − T )
Qevap 2 pump calor
. = m. (hp −h )
Q2
Qevap 3 12 11
T13 . = Q. (T − T )/(T
Qevap 2 10 9
Qevap 2 12 11 13 − T11)
Enclosure . = m. (h − h )
Qcomp
. . = Q. −Q.
Q
Q
comp 2 7 8
FIGURE 5.1.2 As-built drawing of breadboard heat pump I (Kauffeld et al., 1990).
FIGURE 5.1.4 Schematic of evaporator extruded heat exchanger (Kauffeld et al., 1990).
Wet Dry
Vapor Vapor
A A
Section A-A
Liquid
Oil Return
tachometer. All the different methods employed throughout the test program agreed
closely (within 6%), confirming the validity of the test results.
Two zeotropic binary refrigerant mixtures R22/R114 and R13/R12 were tested in
the breadboard heat pump. Tests were conducted at the same capacity for each
mixture to achieve equal average heat flux per unit area of the heat exchangers for
all tests as one of the major criteria for comparability of refrigerants. One exception
was pure R114, which would have required an excessively high compressor speed
to reach the desired capacity. The capacity was chosen to be 4.1 kW, obtained with
pure R22 at 500 RPM compressor speed. Heat sink and heat source entry and exit
water temperatures were kept constant for all tests as the second major criterion for
comparability (McLinden and Radermacher, 1987). The temperatures were 28 and
47˚C entering and leaving the condenser, and 27 and 13˚C entering and leaving the
evaporator. The expansion device was set simultaneously to give marginal subcooling
at the condenser outlet and to provide evaporation throughout the entire evaporator.
The subcooling was determined by the disappearance of vapor bubbles in a sight
glass in the liquid line upstream of the expansion valve. The condition of flooded
evaporator was determined by the presence of liquid droplets leaving the evaporator
in a sight glass in this pipe and by the presence of liquid in the accumulator.
The mixture mass fractions examined cover the whole range from 0 to 100%
for the R22/R114 test series, whereas for the second refrigerant mixture (R13/R12)
mass fractions were limited to less than 45% of R13 by weight because of excessive
pressures in the condenser that exist with more R13.
TABLE 5.1
Evaporator Configurations
Configuration No. 1 2 3 4
Evaporating Refrigerant C B A A
Liquid Refrigerant B C B C
Water A A C B
A major task of this project was to analyze the influence of different heat exchangers
on the performance of a heat pump employing zeotropic refrigerant mixtures. Four
different configurations of the three passages in the evaporator were tested. Detailed
instrumentation and observation are limited to the evaporator because this heat
exchanger has a greater effect on system performance than the condenser. According
to the nomenclature of Figure 5.1.4, the four evaporator configurations are shown
in Table 5.1.
Configurations 2 and 4 were expected to perform better than 1 and 3 since the
evaporating refrigerant and water (which constituted the majority of the transferred
heat) were in the channels that had enhanced surfaces. Furthermore, these channels
were adjacent, limiting the losses due to heat passing through the intermediate
passage.
The performance of configurations 1 and 3 is strongly influenced by the surface
characteristic of the outermost channel. The smooth copper surface provides a poor
heat exchange area, resulting in large temperature differences between the water and
the refrigerant for these two configurations. The relative heat exchanger performance
is shown in the temperature plots for the four configurations for pure R22 in Figure
5.1.6. It is observed that R22 uses the full length of the heat exchanger for config-
urations 1 and 3 but only about one half to three quarters of the heat exchanger for
configurations 2 and 4. Figure 5.1.7 shows the evaporator temperature profiles of
R22/R114. As shown, the heat exchangers for configurations 2 and 4 have smaller
approach temperatures than that of R22 due to the temperature glide of R22/R114.
Figures 5.1.8 and 5.1.9 show the performance of R22/R114.
For the mixture of R22/R114, a maximum increase in coefficient of performance
(COP) of 32% was found at a mass fraction of 65% of R22. At lower R22 mass
fraction, the capacity dropped off rapidly and the COP of R114 was determined to
be about 40% lower than that of R22. However, this large decline is a reflection of
the fact that the compressor is not designed for such a low-pressure refrigerant and
valve pressure losses become dominant. For R22/R114 there is no significant dif-
ference in performance whether or not the suction-line heat exchanger is used (Figure
5.1.8). The best performance was achieved with evaporator configuration 4. Config-
uration 3 performed worst but still showed an increase in performance of 5 to 10%.
Temperature, °C
Temperature, °C
Temperature, °F
Temperature, °F
20 20
60 60
10 10
45 45
C.O.P = 4.81 C.O.P = 5.15
30 0 0
30
Distance Through Evaporator Distance Through Evaporator
(a) Configuration 1. (b) Configuration 2.
90 90
30 30
Water Water
Refrigerant Refrigerant
75 75
Temperature, °C
Temperature, °C
Temperature, °F
Temperature, °F
20 20
60 60
10 10
45 45
C.O.P = 4.85 C.O.P = 5.18
30 0 0
30
Distance Through Evaporator Distance Through Evaporator
(c) Configuration 3. (d) Configuration 4.
FIGURE 5.1.6 Evaporator temperature profiles for R22 (Kauffeld et al., 1990).
90 90
30 30
Water R22/R114 (73/27) Water R22/R114 (57/43)
Temperature, °F
Temperature, °C
Temperature, °F
Temperature, °C
75 Refrigerant 75 Refrigerant
20 20
60 60
10 10
45 Refrigerant glide = 8.4°F 45 Refrigerant glide = 16.3°F
C.O.P = 5.78 C.O.P = 5.02
0 0
30 30
Distance Through Evaporator Distance Through Evaporator
(a) Configuration 2, best COP. (b) Configuration 2, best glide.
90
30 90 R22/R114 (57/43) 30
Water R22/R114 (75/25) Water
Temperature, °F
Temperature, °C
Temperature, °F
Temperature, °C
75 Refrigerant 75 Refrigerant
20 20
60 60
10 10
45 45
Refrigerant glide = 14.5°F Refrigerant glide = 25.1°F
C.O.P = 5.42 0 C.O.P = 4.91 0
30 30
Distance Through Evaporator Distance Through Evaporator
(c) Configuration 3, best COP. (d) Configuration 3, best glide.
90 90
30 30
Water R22/R114 (66/34) Water R22/R114 (60/40)
Temperature, °F
Temperature, °C
Temperature, °F
Temperature, °C
75 Refrigerant 75 Refrigerant
20 20
60 60
10 10
45 Refrigerant glide = 18.9°F 45 Refrigerant glide = 23.3°F
C.O.P = 5.80 C.O.P = 5.80
0 0
30 30
Distance Through Evaporator Distance Through Evaporator
(e) Configuration 4, best COP. (f ) Configuration 4, best glide.
FIGURE 5.1.7 Evaporator temperature profiles for R22/R114 (Kauffeld et al., 1990).
Figures 5.1.10 and 5.1.11 show the performance of R13/R12. The improvement
in cycle efficiency for the mixture R13/R12 compared to the efficiency obtained
with pure R22 and pure R12 is shown in Figure 5.1.11 to be 16 and 40%, respectively,
with internal heat exchange and 8 and 27%, respectively, without internal heat
exchange. For the test set performed with R13/R12, it was not possible to perform
tests with high concentrations of R13 since the pressure would have been far beyond
the highest acceptable pressures.
Another observation is that at any temperature glide less than the maximum
possible, there will be two mass fractions that will give the required glide. The
choice of which of these two mass fractions to use would be based on other
refrigerant properties. For instance, with the R22/R114 mixture, the composition
high in R22 would be chosen because of the better compressor efficiency and
capacity with higher R22 mass fractions. With the R13/R12 mixture, the mass
fraction low in R13 would be chosen to avoid the high discharge pressures with
high mass fractions of R13.
R22/R114
6
C.O.P.
5
W/O liquid line HX
Configuration 2
W/O liquid line HX
4 Configuration 3
W/O liquid line HX Configuration 4
With liquid line HX Configuration 2
With liquid line HX Configuration 3
With liquid line HX Configuration 4
3
100 80 60 40 20 0
Weight % R22
7
R22/R114
6 32%
C.O.P.
4
With liquid line heat exchanger
Without liquid line heat exchanger
3
100 80 60 40 20 0
Weight % R22
The capacity per compressor revolution for both mixtures in evaporator config-
uration 4 is shown in Figure 5.1.12. Compared to pure R22, the R22/R114 mixture
(65/35 wt.%) that showed 32% efficiency improvement (Figure 5.1.10) resulted in
a capacity loss of 8%. The R13/R12 mixture (34/66 wt.%) that showed a 16%
efficiency improvement over pure R22 (Figure 5.1.11) resulted in a capacity increase
of 21%.
An interesting feature of temperature plots is that the location of the pinch points
varies for different refrigerant mixtures. The pinch point, the closest temperature
6 R13/R12 4
Without liquid line HX
Configuration 2
Without liquid line HX
4
C.O.P.
5 Configuration 4
With liquid line HX
3
Configuration 2
With liquid line HX
Configuration 3 2
4 With liquid line HX 2
Configuration 4
100 80 60 40 20 0
Weight % R13
7
With liquid line heat exchanger
Without liquid line heat exchanger
6
C.O.P.
16%
C.O.P. for R22 8%
5
40% 27%
R13/R12
4
100 80 60 40 20 0
Weight % R13
0.70
Volumetric Capacity, Btu/Revolution
0.60 R1
3/
R1
0.50 2
0.40 R22
/ R1
14
0.30
0.20
0.10
0
100 90 80 70 60 50 40 30 20 10 0
Weight % Low Boiler (R22 or R13)
FIGURE 5.1.12 Volumetric capacity of R22/R114 and R13/R12 for evaporator configuration
4 (Kauffeld et al., 1990).
90
R22/R114 (66/34)
Water
75 Refrigerant
Temperature, °F
60
45
30
90
R13/R12 (32/68)
Water
75 Refrigerant
Temperature, °F
60
45
30
Distance Through Evaporator
FIGURE 5.1.13 Comparison of evaporator temperature profiles for R22/R114 and R13/R12
(Kauffeld et al., 1990).
approach between the refrigerant and the heat exchange fluid, is at the two ends of
the heat exchanger for the R22/R114 mixture, whereas for the R13/R12 mixture the
pinch point is in the center of the heat exchanger. This is shown in Figure 5.1.13,
where a line connects the points of entry and exit to emphasize this nonlinearity.
The nonlinear shape of the temperature profile during evaporation results from
the differing properties of the pure refrigerants in combination with mixture effects.
The dominant effect in determining the shape of the profiles appears to be the relative
amounts of the more and less volatile components. With the R22/R114 mixture of
60% R22, the bulk of heat necessary to vaporize the mixture is required at temper-
atures close to the bubble point temperature (i.e., at temperatures closer to the boiling
point of the more volatile R22). As a result, the temperature versus enthalpy profile
is concave. With the R13/R12 mixture, the mass fraction of the more volatile R13
is only 20% and thus the bulk of the heat of vaporization is concentrated at the
higher temperatures, resulting in a convex profile.
A secondary effect is the different heats of vaporization for the two components.
This is most pronounced with R22/R114, where the heat of vaporization of R22 is
substantially higher than that of R114. Because the first bubbles of vapor generated
are enriched in R22, the higher heat of vaporization for this refrigerant further
contributes to the concave shape. This behavior is important for the prediction of
system performance by the use of simple computer programs which assume a linear
temperature glide throughout the heat exchanger, since the actual glide may deviate
quite far from linearity. These errors are greatest with efficient heat exchangers where
the temperature differences are small, the application seen as most favorable to
mixtures.
The convex shape of the temperature profile of R13/R12 tends to be amplified
by a pressure drop as present in the evaporator, whereas the concave shape of
R22/R114 tends to be leveled. Another aspect of this nonlinearity is discussed in
the next section.
The final part of this investigation was concerned with the validity of the proposed
improvement obtained by using an intermediate heat exchange between the evapo-
rating refrigerant and the condensed liquid (Vakil, 1981). Liquid precooling in
evaporator (LPE) improves efficiency with mixtures by shifting a portion of the
evaporator capacity to a lower vapor quality and thus to a lower temperature. Tests
have been conducted with both refrigerant mixtures R22/R114 and R13/R12 to
evaluate this theoretical proposal. As one can see from Figures 5.1.10 and 5.1.11,
noticeable improvement in efficiency is only obtained for the mixture of R13/R12.
The reasons for this different behavior are seen in the pressure-enthalpy diagrams
for the two refrigerant mixtures.
First, the two-phase isotherms are nearly flat at low quality for R22/R114 (Figure
5.1.14) while those for R13/R12 (Figure 5.1.15) show a substantial slope throughout
the two-phase region. Because of this flatness, the heat exchanger has little value
Enthalpy (Btu/lbm)
25 50 75 100 125 150
10,000
5000 750
100°
80°
Pressure (kPa)
40°
1 4,4'
1000 1' 150
20°
2' 3,3'
2 75
500
s = 0.2 kJ/kg°C
0°
.2
0.4
=0
0.4
0.6
0.6
ty
0.8
ali
0.8
Qu
−20° 30
200
50 100 150 200 250 300
Enthalpy (kJ/kg)
FIGURE 5.1.14 Effect of suction line heat exchanger for R22/R114 (Kauffeld et al., 1990).
Enthalpy (Btu/lbm)
25 50 75 100 125 150
10,000
5000 750
60°
1' 1
4, 4'
Pressure (psia)
Pressure (kPa)
40°
2000 300
2' 20°
3'
2 3
s = 0.4 kJ/kg°C
10000 150
0°
0.6
0.8
75
0.2
500 −20°
y=
0.4
alit
0.6
Qu
0.10
0.8
200 30
50 100 150 200 250 300
Enthalpy (kJ/kg)
FIGURE 5.1.15 Effect of suction line heat exchanger for R13/R12 (Kauffeld et al., 1990).
for R22/R114 for the same reasons that cause it to have little value for pure refrig-
erants with their truly horizontal isotherms. Referring to Figures 5.1.14 and 5.1.15,
the shift of the evaporator entering condition from point 2 to 2 can be seen to result
in a higher evaporator pressure (which improves efficiency) at the same average
evaporator temperature for R13/R12, while the shift from point 2 to 2 has no effect
on evaporator pressure for R22/R114. Second, the shape of the two-phase dome
results in flashing more of the R13/R12 mixture between the approximate operating
points (30˚C condensing, 10˚C evaporating) at which the tests were run, indicating
more potential for reducing entering vapor quality.
5.1.1.7 Discussion
5.1.1.8 Conclusion
• For all heat exchanger configurations for which a full range of mass
fractions was tested, the optimum mass fraction of each mixture performed
better than pure R22.
• The best efficiency measured with the mixture R22/R114 (at a mass
fraction of 65% R22) was approximately 32% higher than the best effi-
ciency measured with R22.
• The best efficiency measured with the mixture R13/R12 (at a mass fraction
of 35% R13) was approximately 16% better than the best efficiency
measured with R22.
• The better the heat exchanger effectiveness, the greater the improvement
the mixtures showed over pure R22.
• In the evaporator configurations having poorer heat exchanger effective-
ness, the gliding temperature effect of mixtures was less important to
cycle efficiency than other refrigerant properties.
• Pressure drop in the evaporator reduced the temperature glide and hence
reduced the cycle efficiency improvement by mixtures.
• Nonlinearity of temperature with enthalpy was observed for the two mix-
tures tested.
• Intracycle heat exchange between the condensed liquid and evaporating
refrigerant streams benefited the R13/R12 mixture but not the R22/R114
mixture.
• Intracycle heat exchange between the condensed liquid and evaporating
refrigerant streams is beneficial for those mixtures showing a substantial
loss of potential temperature glide as a result of flashing through the
expansion device.
• Efficiency improvement by using mixtures does not imply loss of capacity
as a trade-off.
Figure 5.1.16 shows a schematic of the laboratory breadboard heat pump II. The
refrigerant side consists of the compressor, condenser, SLHX, expansion device and
evaporator. To minimize the charge, the system components were limited to the basic
components without any large volume auxiliary equipment such as an oil separator
or an accumulator. The compressor was two cylinder reciprocating compressor. All
heat exchangers used in this breadboard heat pump II (evaporator, condenser and
SLHX) were in a counterflow heat exchange configuration: the refrigerant flows in
the inner tube and a water-ethylene glycol mixture flows in the annulus. The inner
tube and annulus side were equipped with turbulators and spine fins in order to
increase the heat transfer coefficient. This heat transfer enhancement resulted in
large flow pressure drops but had other benefits such as the compactness of the
system design and reduced refrigerant charge. The SLHX was installed between the
liquid line after the condenser outlet and compressor suction line with a bypass line.
The expansion valve was a metering valve controlled manually.
For all tests, two cycle parameters were controlled such that the degree of subcooling
at the condenser outlet and degree of superheating at the evaporator outlet were
positive in order to get comparable test results. The degree of subcooling was
especially limited as much as possible (not greater than 2 K) and the superheating
region was limited within 5% of the evaporator length. These conditions were
satisfied by adjusting the refrigerant charge and expansion valve opening. The
capacity was maintained for all tests by adjusting the compressor speed from
1500 RPM to 1620 RPM. A refrigerant sample was taken from the discharge line
during the steady state phase of the test and then evaluated with the gas chromato-
graph to measure the mass fraction of the mixtures.
Using this test apparatus II, R32/R134a and R32/R152a mixtures at various mass
fractions of R32 were conducted at various test conditions defined in Table 5.2.
T
W/G
FM MM
C T P
M S Co
Cooling
Brine
I X
Charging
System
IX
Three-way W/G
valve
Thermostatic P
T T
valve, dioxide T
T Thermostatic T E
valve FM MM
Hand valve
HC HV
Filter, dryer
(interground) C - Condenser HC - Heater, constant P - Pump
CO - Compressor HV - Heater, water S - Longer & speed motor
Sight glass E - Evaporater I - Inverter W/G - water/glycol tank
FM - Volume flow meter IX - Inverter heat exchanger X - External heat exchanger
Filter (water) MM - Mass flow meter M - Mirror
FIGURE 5.1.16 Schematic of breadboard heat pump II (Pannock and Didion, 1991).
TABLE 5.2
Test Conditions for Heat Transfer Fluid
Evaporator Evaporator Condenser Condenser
Test Inlet (˚C) Outlet (˚C) Inlet (˚C) Outlet (˚C)
25
20
COP improvement over R22 (%)
15
10
R32/R152a R32/R134a
−5
−10
0.15 0.2 0.25 0.3 0.35 0.4 0.45
Mass fraction R32
Test condition 1A
FIGURE 5.1.17 Cooling COP for Test A (Pannock and Didion, 1991).
The R32 mass fraction of the mixture affects the suction and discharge pressure.
Lowering R32 mass fraction can reduce the operating pressures but it also reduces
volumetric capacities. R32/R152a mixtures showed lower pressures than R32/R134a
mixtures at the same R32 mass fraction. At a mass fraction of 28 and 36%, the
suction pressures of R32/R134a and R32/R152a are equal to that of R22. A similar
trend to the suction pressures was observed for the discharge pressures. The discharge
pressures of R32/R134a and R32/R152a vary from –12 to +8% and from –24 to
–6% with respect to R22 in the measured mass fraction range, respectively.
For both mixtures, the suction temperature was constant (within ±1 K) for all
tests but the discharge temperature was different. For both mixtures, the discharge
temperature was lower than that of the R22. The R32/R152a mixtures showed a
decrease of 3 K to 5 K with increasing R32 mass fraction and the R32/R134a mixture
showed a decrease of 5 K to almost 10 K as compared to R22.
The operating pressures and temperatures of both mixtures are within acceptable
ranges as expected from the test condition.
Low Temperature Cooling Test (Test B)
The cooling capacity was kept constant at 3.7 kW for these tests. Figure 5.1.18
shows the COP improvement over R22 for Test B. Figure 5.1.18 shows that the COP
improvement increases steadily when the R32 mass fraction increases for both
mixtures as similar to Test A case.
For R32 mass fraction below 17% in the R32/R152a mixture, the COP is lower
than that of R22. The COP improvement was over 20% when the R32 mass fraction
30
20
COP improvement over R22 (%)
R32/R152a R32/R134a
10
−10
−20
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Mass fraction R32
Test condition 1B
FIGURE 5.1.18 Cooling COP for Test B (Pannock and Didion, 1991).
was higher than 35%. For R32/R134a, the COP is lower than that of R32/R152a for
R32 mass fractions lower than about 30%. The maximum COP increase was 22%
over that of R22 at the same capacity at 38% R32 mass fraction.
At a mass fraction of 28 and 36%, the suction pressures of R32/R134a and
R32/R152a are equal to that of R22. A similar trend to the suction pressures was
observed for the discharge pressures. The discharge pressures of R32/R134a and
R32/R152a vary from –12 to +8% and from –24 to –6% with respect to R22 in the
measured mass fraction range, respectively.
Similar to Test A, the suction and discharge pressures were again within accept-
able limits with respect to R22. The suction pressures of the R32/R134a mixture
were 10 to 15% higher than those of the R32/R152a mixture, meaning about 100 kPa
higher suction pressure. The discharge side pressure difference was about doubled
to 200 kPa. The discharge pressures of R32/R134a were lower than those of R22
for R32 mass fraction of less than 30%. For R32/R152a, all measured discharge
pressures were below that of R22.
For both mixtures, the suction temperature was constant (within ±1 K) for all
tests except one data point but the discharge temperature was different. For both
mixtures, the discharge temperature was lower than that of the R22. The R32/R152a
mixtures showed a decrease of 3 to 5 K for 20 to 30% of R32 mass and the
R32/R134a mixture showed a decrease of 10 K with more than 30% R32 mass
fraction.
High Temperature Heating Test (Test 47S)
The heating capacity was kept constant at 3.1 kW for these tests. Figure 5.1.19
shows the COP improvement over R22 for Test 47S. Again, the COP improvement
increases steadily when the R32 mass fraction increases for both mixtures as similar
to the cooling mode. Different from the cooling mode, the heating COPs of the
R32/R134a mixtures were higher than those of the R32/R152a mixtures (Figure
5.1.20). The COPs of both mixtures were higher than that of R22 when the R32
mass fraction was more than 20%. The R32/R152a mixture also shows a COP
improvement against R22 when the R32 mass fraction was more than 20%. At the
mixture compositions of same speed and same capacity, R32/R134a and R32/R152a
showed 4 and 8% improvement, respectively.
Low Temperature Heating Test (Test 17L)
The heating capacity was kept constant at 2.0 kW for these tests. Figure 5.1.20
shows the COP improvement over R22 for Test 17L similar to the Test 47S case.
The COP improvement increases steadily when the R32 mass fraction increases for
both mixtures, as with the other cases. Similar to Test 47S, the heating COPs of the
R32/R134a mixtures were 5 to 10% higher than those of the R32/R152a mixtures
at the same R32 mass fraction (Figure 5.1.20). The COPs of both mixtures were
lower than that of R22 for most of the R32 mass fractions investigated. Only the
R32/R134a mixture showed a 5% COP improvement against R22 at 40% of R32
mass fraction. The same COP as R22 was reached at about 30% of R32 mass
fractions for the R32/R134a mixtures and at above 40% of R32 mass fractions for
the R32/R152a mixtures.
15
10
COP improvement over R22 (%)
5 R32/R152a
R32/R134a
0
−5
−10
−15
−20
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Mass fraction R32
Test condition 1B
FIGURE 5.1.19 Heating COP for Test 47S (Pannock and Didion, 1991).
10
5
R32/R134a
COP improvement over R22 (%)
−5
−10
R32/R152a
−15
−20
−25
−30
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Mass fraction R32
Test condition 10
FIGURE 5.1.20 Heating COP for Test 17L (Pannock and Didion, 1991).
25
20
COP improvement over R22 (%)
15
R32/R152a
10
5
R327/R134a
−5
−10
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Mass fraction R32
Test condition 1A-SLHX
FIGURE 5.1.21 Cooling COP for Test SLHX (Pannock and Didion, 1991).
5.1.2.3 Conclusion
• For R32/R152a mixtures, the mass fraction of R32 should be higher than
40% to ensure a heating capacity comparable to R22 capacity for all
operating conditions.
• R32/R152a mixtures showed a significant increase in COP (14 to 20%)
for cooling conditions with a counterflow heat exchange configuration.
• R32/R152a mixtures have the lowest GWP possible of the tested mixtures,
about one fourth of the R22 value, and are flammable in the whole mass
fraction range.
• R32/R134a mixtures showed better or equivalent performance than
R32/R152a when the mass fraction of R32 was higher than 35%. Specif-
ically, the heating performance of R32/R134a mixtures is better than that
of R32/R152a due to its larger volumetric capacity at the same R32 mass
fraction.
• R32/R134a mixtures are “controllable-flammable” mixtures since they are
the mixture of flammable R32 and nonflammable R134a.
• The potential benefits of SLHX are a higher performance for the given
R32 mass fraction. Therefore, the R32 mass fraction can be lowered until
the same volumetric capacity is obtained. This reduces the flammability
of R32 rich mixtures. In fact, the COP increases for both mixtures over
R22 up to 20% when the mass fraction of R32 was 40% and counterflow
heat exchanger is used.
• R32/R134a mixtures with SLHX showed a higher COP increase over the
case without SLHX than did R32/R152a mixtures.
• R32/R134a and R32/R152a are potential alternatives for R22 since both
mixtures showed a comparable performance if the appropriate mass frac-
tion mixture is chosen and counterflow heat exchange with the heat transfer
fluid is possible.
TABLE 5.3
Test Conditions
Variable Cooling Mode Heating Mode
16 Resistance 4.7
Cooling Capacity (KBtu/h)
4 1.2
0 0
0 0
(a)
1.0
4.0
R152a
R152a
Measured COP
Measured COP
2.5 25% R1381
R22 3.6
50% R1381 50% R1381
75% R1381
2.0 3.0
R1381 C
1.5 2.5
75 85 95 105 0 20 40 60
Outdoor temperature (°F) (b) Outdoor temperature (°F)
FIGURE 5.1.22 Cooling performance of R13B1/R152a (Vineyard, 1988). (a) Cooling. (b) Heating.
TABLE 5.4
Test Results
R13B1/R152a R13B1/R152a R13B1/R152a
Refrigerant R22 R152a (25/75 wt.%) (50/50 wt.%) (75/25 wt.%)
P
e
do di
47° do di
f
(b) Finned Surface
23°
do di
f
(c) Microfin Surface
FIGURE 5.1.23 Three enhanced surface tubes (Vineyard and Conklin, 1991).
TABLE 5.5
Geometry of Enhanced Tubes
Smooth
Tube Type Tube Fluted Tube Finned Tube Microfin Tube
7
27.8°C DOE Rating
Condition
6 R22-Dotted Lines
Coefficient of performance (COP)
1
5 6 7 8 9 10 11 12 13 14
Capacity (kW/h)
(45˚C saturation temperature) for the condenser. Evaporator inlet quality was approx-
imately 0.25 and the degree of the superheating at the evaporator outlet was approx-
imately 3 to 5˚C.
Test results of R22 and R143a/R124 with the smooth tube and three enhanced
tubes are shown in Figure 5.1.24 and summarized as follows:
• Enhanced surface tubes significantly improved the COPs of both R22 and
R143a/R124. COP improvement of R143a/R124 was larger than that of R22.
• When the capacity exceeded approximately 7 kW, the COP of R143a/R124
dropped rapidly due to the greater pressure drop and refrigerant side heat
transfer degradation associated with zeotropic mixtures.
• Different enhanced tubes worked better for the evaporator and condenser.
Using a fluted tube for the evaporator and a finned tube for the condenser
resulted in the best performance improvement (11 to 23%) over R22.
As a continuing effort from Vineyard and Conklin (1991) to improve system COP,
Vineyard et al. (1993) experimentally investigated three methods of improving heat
transfer enhancements, inserts, finned tube and a segmented evaporator.
Inserts
Perforated foil inserts were used inside of the inner tube. The perforated brass foil
(w = 0.23 mm, t = 0.05 mm, 50 holes/cm2) was used for the second through fourth
evaporator passes to enhance the nucleate boiling for the low quality region. The
aluminum bent tabs, shown in Figure 5.1.25, were inserted in the second through
seventh condenser passes.
30°
t
FIGURE 5.1.25 Geometry of bent-tab insert (w = 13.33 mm, t = 0.8 mm, = 2.75 m long)
(Vineyard et al., 1993).
Annulus Wall
Water Flow
47° do di
Refrigerant
Flow f
Water Flow
FIGURE 5.1.26 Geometry of finned tube (do = 15.8 mm, di = 13.7 mm, f = 0.38 mm)
(Vineyard et al., 1993).
Finned Tube
A commercially available finned tube was used for the refrigerant side tube, as
shown in Figure 5.1.26.
Segmented Evaporator
The segmented evaporator was designed to enhance the nucleate boiling heat transfer
for the low vapor quality region and the convective boiling heat transfer for the high
vapor quality region. Evaporator passes 1 through 4 used a rolled-fin surface as
shown in Figure 5.1.27 and evaporator passes 5 through 8 used a finned tube surface
as shown in Figure 5.1.26. Two sections then formed the segmented evaporator, as
shown in Figure 5.1.28.
Test results of four tube combinations for the evaporator and condenser are
shown in Figure 5.1.29. As shown in (a), the bent-tab inserts and the finned tube
cases were compared to the smooth tube case for R22 and R143a/R125. The smooth-
tube performance of R143a/R125 was 13 to 27% lower than that of R22. The first
Annulus Wall
Refrigerant Flow
Reentrant
Cavity
Surface
Tube
Wall
Water Flow
LC
set of tube enhancements for R143a/R125, the perforated foil evaporator and bent-tab
condenser, showed no COP improvement due to the high pressure drop. The second
set of tube enhancements for the R143a/R125, the perforated foil evaporator and
finned tube condenser, showed a 15% COP improvement over the R22 at the lowest
capacity and 6% lower COP than that of R22 at the highest capacity. This COP
decrease for the higher capacity was due to the increased pressure drop at the higher
refrigerant mass flow rate. The third set of the combination was the segmented
evaporator and finned condenser. For the third set, R143a/R124 showed a 5 to 9%
COP improvement over R22 and R32/R124 showed a 9 to 17% COP improvement
over R22. When the third set was compared to the smooth tube case, R22 showed
a 22 to 27% improvement and R143a/R124 showed a 62 to 80% improvement.
In summary, the following conclusions were obtained.
Refrigerant in
T/C
Water Out Pressure
RTD Transducer H20 T/C
REF T/C
REF T/C
Water
REF T/C
REF T/C
H20 T/C
Refrigerant Out
T/C
Pressure RTD
Transducer Water in
A test facility called a psychrometric calorimeter was set up to measure the steady
state performance of an air-to-air heat pump. The test facility consisted of a closed
indoor loop, an environmental chamber and a data acquisition system. The closed indoor
loop was equipped with devices that measured the dry bulb and dew point temper-
atures of the air and a nozzle that measured the air flow rate downstream of the test
unit, as shown in Figure 5.2.1. The test chamber was equipped with an air handling
unit, as shown in Figure 5.2.2, so it could simulate the summer and winter climate
conditions defined by ASHRAE Standard 116, Table 5.6 (ASHRAE, 1995).
Segmented, Finned
6
6
Foil, Finned 5
5
4 4
1 1
12 16 20 24 28 32 36 40 44 48 12 16 20 24 28 32 36 40 44 48
Capacity (1000 Btu/h) Capacity (1000 Btu/h)
(a) (b)
INDOOR
Condensed
Water out TEST Wet- and dry-bulb
temperature
Diffusion Receiving
UNIT Manometer Baffle Chamber
Manometer
AIR HANDLER
OUTDOOR TEST UNIT
T with 5
thermocouples Wet-and dry-bulb
Extended temperature
Duct
Ref. out
P T
Coil
P T
FLOW
Code for outdoor unit:
T T Temperature grid unit is shown during cooling mode
Compressor T : Temperature measurement
with 9 thermocouples P : Pressure measurement
Ref. in
TABLE 5.6
ASHRAE Test Conditions
Indoor Indoor Outdoor Outdoor
Dry Bulb Wet Bulb Dry Bulb Wet Bulb
Test (˚C) (˚C) (˚C) (˚C)
The test unit was a 7.0 kW capacity split heat pump system using a reciprocating
compressor and two expansion devices. One expansion device was a thermostatic
expansion valve (TEV) for the heating mode and the other was a short tube restrictor
(ST) for the cooling mode. Figure 5.2.3 shows the heat exchange configuration for
the indoor coil. The indoor unit was installed at an inclination of 20˚ from the vertical
with downward airflow, as shown in Figure 5.2.3. The test unit was designed for
Refrigerant Flow
Cooling
Heating
Cooling
R22; therefore, the original components were used for the evaluation of the refrig-
erants having a similar vapor pressure with R22. Except for the ST, the R22 tests
were carried out without making any modifications to the system while the mixture
tests were carried out after changing the lubricant for the compressor from mineral
oil to ester oil and changing the filter drier. When testing higher vapor pressure
refrigerants than R22, the compressor and TEV designed for higher operating pres-
sures were used. As compared to the original compressor, calorimeter tests performed
by the compressor manufacturer showed that the high pressure compressor had 4%
lower efficiency at 54.4˚C condensing temperature and 7% higher efficiency at
37.8˚C condensing temperature.
TABLE 5.7
Environmental and Safety Properties of Refrigerants Investigated
Refrigerant R22 R407C R410A R410B
TABLE 5.8
Thermodynamic Properties of Refrigerants Investigated
(Hwang et al., 1995)
Refrigerant R22 R407C R410A R410B
Source: Gallagher et al., 1993, unless otherwise noted. Hwang et al., 1995.
10
3
Vapor Pressure [MPa]
0.5
0.3 R-22
R407C
0.2 R410A
R410B
0.1
−20 −10 0 10 20 30 40 50 60 70
Temperature [°C]
of these refrigerants is approximately 44% higher than that of R22, which also
translates into higher capacities for the same size compressor. To test the fluids with
the original heat exchangers, the displacement of the compressor had to be reduced
to a similar capacity. These binary mixtures have a higher vapor pressure by approx-
imately 60% compared to R22 but the slope of the vapor pressure curve is quite
similar to that of R22 in the temperature range of interest (Figure 5.2.4). Therefore,
the pressure ratio of these refrigerants is expected to be similar to that of R22.
Capacity Measurement
The experiments to measure the capacity and COP were performed based on
ASHRAE Standard 116 (ASHRAE, 1995) and ARI Standard 210/240 (ARI, 1994).
In this study, the air side capacity and the refrigerant side capacity were measured.
The loop air enthalpy method was used to measure the air side capacity: the air flow
rate and air enthalpy difference between inlet and outlet of the indoor coil were
measured. A nozzle apparatus measured the air flow rate. In measuring the refrigerant
side capacity, the refrigerant mass flow rate and the refrigerant enthalpies at the inlet
and outlet of the heat exchanger were used. The mass flow rate was measured by a
Coriolis-type mass flowmeter and the refrigerant enthalpies were calculated by
REFPROP Version 4.0 (Gallagher et al., 1993) from the temperature and pressure
measurements.
ASHRAE Standard 116 (1995) requires that the capacities determined using
these two methods agree within 6% of each other. The two methods agreed within
3% for all tests conducted in this study. The reported capacity and COP values were
based on air side values. The refrigerant side values were used only to check the
total energy balance and to prove the validity of the test.
Test Procedures
The experimental tests were carried out with refrigerants R22 and its alternatives,
R407C, R410A and R410B. First, the R22 “baseline” test and the “retrofit” test for
R407C were performed. The baseline R22 tests were conducted without making any
modifications to the system. In the retrofit tests, only the refrigerant charge was
optimized to find the optimum charge that gave the maximum COP at ASHRAE
cooling test A condition. After the optimum charge was found, ASHRAE tests were
carried out to compare the performance of each refrigerant and mixture.
Second, the “soft optimization” was carried out for all refrigerants tested to
maximize both the cooling and heating COP. The combination of the expansion
devices and the refrigerant charge was chosen to maximize both the cooling COP
for ASHRAE cooling test B and the heating COP for ASHRAE heating test 47S.
At first, the optimum charge for heating was obtained by adjusting the refrigerant
charge at the 47S heating test conditions. The optimum charge was chosen such that
the maximum COP could be obtained. To find the ST for the cooling mode that
corresponds with the optimum charge in the heating mode, several charge optimi-
zation tests were run with different size STs at the cooling test conditions.
7.6 3.15
3.10
7.4
Cooling Capacity [kW] 3.05
3.00
COP
7.2
2.95
R-22 COP
2.90
7.0
R-407C Capacity
2.85
R-22 Capacity R-407C COP
6.8 2.80
3.6 3.9 4.1 4.3 4.5
Charge [kg]
FIGURE 5.2.5 Charge optimization of R22 and R407C (Hwang et al., 1995).
TABLE 5.9
Comparison of ASHRAE Cooling A and Heating 47S Test Results
Test Type R22 Baseline R407C Retrofit
Test Mode Cooling A Heating 47S Cooling A Heating 47S
TABLE 5.10
Comparison of ASHRAE Cooling B and Heating 17L Test Results
Test Type R22 Baseline R407C Retrofit
Test Mode Cooling B Heating 17L Cooling B Heating 17L
Also, the cycle shifts to higher evaporating and condensing pressures. The higher
pressure ratio lowers the compressor efficiency. Although R407C has a slightly
higher capacity; it has a lower COP.
R407C retrofit shows more degradation for heating than for cooling. This can
be explained by the heat exchange configuration of the indoor coil as shown in
Figure 5.2.3. The heat exchanger configuration contributes to the capacity degrada-
tion in the heating mode, especially for mixtures that have a temperature glide.
Soft Optimization Test Results
To improve the steady-state performance of the mixture, the soft optimization was
carried out for R22, R407C, R410A and R410B that could reconcile the imbalance
between the cooling and heating optima. As already mentioned, the optimum charge
was found for the heating test 47S, then the ST was adjusted so that the optimum-
cooling COP occurs at the same charge. When the optimum charge was the same
for different sizes of STs as for R410B, the subcooling and superheating were
evaluated to choose the optimum ST size. For R410B, a 1.55 mm ST has the wider
range of refrigerant charge with acceptable subcooling and superheating than those
of 1.60 mm ST. Therefore, the 1.55 mm ST was chosen for R410B and the result
TABLE 5.11
Optimum Charge for Each Operating Mode
Refrigerant R22 R407C R410A R410B
of this procedure is shown in Table 5.11. Except for R410B, it was possible to adjust
the charge and ST so the optimum charges for heating and cooling test conditions
were the same. For R410B, the average between the optimum heating and cooling
charges was used. The larger ST, 1.65 mm, was also tested but the reasonable values
of subcooling and superheating were not obtained. This mismatch could have been
eliminated or minimized if the more finite increases for the size of STs and refrigerant
charges were tested.
Steady State Performance Results
After the soft optimization tests, ASHRAE tests were carried out with the optimum
refrigerant charge and ST. Test results for capacity and COP are compared with
other data in Tables 5.12, 5.13 and 5.14. R407C had approximately the same cooling
and heating capacity (within 3% of R22) but also a slight degradation in cooling
and heating COP (4 to 8%). The existing heat exchanger was not suited for utilizing
the temperature glide of this mixture and contributed to the slight degradation in
capacity. As expected, this refrigerant had a larger pressure ratio and this also might
contribute to the degradation in COP.
The two binary mixtures showed higher cooling capacity by up to 7% and COP
up to 4% compared to those of R22 despite the smaller compressor. These refrig-
erants also had better heating capacity by 3 to 6% at 47S and up to 20% in the 17L
TABLE 5.12
ASHRAE Cooling A Test Results
Refrigerant R22 R407C R410A R410B
TABLE 5.13
ASHRAE Heating 47S Test Results
Refrigerant R22 R407C R410A R410B
TABLE 5.14
ASHRAE Cooling B and Heating 17L Test Results
Refrigerant R22 R407C R410A R410B
test. R410A and R410B showed similar cooling and heating performances when
compared with each other.
As expected, these refrigerants have higher evaporating and condensing pres-
sures by approximately 60% compared with those of R22 and the pressure ratio was
nearly the same as that for R22.
The results of this study for R407C and R410B are compared with the results
presented by AREP (Alternative Refrigerants Evaluation Program) (Godwin, 1993)
in Table 5.15, while the results of R410B are compared with the results by Kol-
liopoulos (1994). Overall, the data of this study agree well with the AREP data. For
R410B, the data show good agreement with those of Kolliopoulos except for the
low temperature heating capacity, which is higher than that of Kolliopoulos. This
significant gain in capacity arises because at higher ambient temperatures and there-
fore higher capacities, the heat exchangers are slightly undersized for the high
pressure fluids. This occurs because it was not possible to exactly match the high
pressure compressor’s capacity to that of the existing system. Support for this
conclusion comes from the minimal superheating.
For the data presented here, the TEV maintained the superheating as low as 1
to 2˚C while in the Kolliopoulos data, the superheating was maintained at 11.1˚C.
TABLE 5.15
Comparison of Results with AREP Data
Cooling Cooling Heating Heating
Refrigerant Source Capacity COP Capacity COP
Note: The above data are the ratios of each refrigerant performance to R22 performance.
Another difference is that a reciprocating compressor was used in this study and a
scroll compressor was used in the Kolliopoulos study.
Seasonal Performance Results
When evaluating the overall performance of heat pumps, including the steady state
and cyclic performances, the seasonal performance factors seasonal energy efficiency
ratio (SEER) and heating seasonal performance factor (HSPF) are used. These factors
are important because they approximate operation over an entire season. The sea-
sonal performance was calculated based on ASHRAE Standard 116 (ASHRAE,
1995). Eight different ASHRAE test results, including the steady state and cyclic
performances, were used in this calculation.
The seasonal performance factors for each refrigerant are calculated and com-
pared in Table 5.16. Relative to R22, R407C has a 6.8% lower SEER and a 4.7%
lower HSPF, while R410A has a 2.6% higher SEER and nearly the same HSPF and
R410B has a 1.7% higher SEER and a 1.8% higher HSPF. Two binary mixtures
have lower COP for 47S test conditions but their better cyclic performance contrib-
utes to having a slightly better HSPF.
R407C, having similar operating pressures as R22, has the advantage of being
a retrofit, which means minimal changes to the system for a quick switch to alter-
natives. Only the expansion devices and the refrigerant charge must be chosen
carefully to have performance within 7.8% of R22. R407C has a 6.8% lower cooling
seasonal performance and a 4.7% lower heating seasonal performance as compared
to R22. The reduction in performance as compared to R22 is relatively small and
can be overcome with systems specifically designed for R407C.
The two binary refrigerants require careful design considerations, especially heat
exchangers, pipes and compressors, due to the much higher operating pressures.
Although these binary refrigerants need significant system modification, they have
the benefit of higher steady state and seasonal performances. These benefits could
be even greater if the system is designed specifically for these fluids. Overall, the
binary fluids perform better than R22. R410A has a cooling seasonal performance
TABLE 5.16
Seasonal Performance Test Results
Refrigerant R22 R407C R410A R410B
Notes:
1. Conditions in SEER calculation.
(a) Oversize factor: 10%.
(b) Design temperature of 35˚C (needs eight temperature
bins ranging from 18 to 41˚C).
(c) The bin hours are based on the U.S. national average
climate.
(d) The building load (BL) was selected from the R22
baseline case.
2. Conditions in HSPF calculation.
(a) The regional outdoor temperature: –22˚C.
(b) Location requires 18 temperature bins ranging from –32
to 18˚C.
(c) The bin hours were based on Region V in ARI Standard
210/240 (1994).
that is 2.6% better than R22 and a heating seasonal performance that is essentially
the same as R22. R410B has cooling and heating seasonal performances that are
1.7 and 1.8% better than R22, respectively.
shown in Figure 5.2.3. The heat exchange pattern between air and refrigerant is near
counter flow for the cooling mode but it is near parallel flow for heating. Installing
four check valves, as shown in Figure 5.2.6, changed the indoor coil path to supply
the refrigerant to the indoor coil in the same direction regardless of the action of
the four-way valve. With this path modification, the refrigerant could be supplied
into the indoor coil from bottom to top for both cooling and heating modes to
maintain a near counter flow heat exchange between air and refrigerant at all times.
This option is important when using zeotropic mixtures that have a temperature
glide because the refrigerant path is reversed depending on the operating mode in
a heat pump.
Path modification results show a 3.7% lower capacity and a 5.5% lower COP
for cooling test A as compared to that of R22, as shown in Table 5.17. The four
additional check valves cause an additional pressure drop, resulting in a lower mass
flow rate and a lower evaporating pressure. But it shows a significant improvement
in the heating COP (10.3%) for heating test 47S, as compared to the retrofit case
shown in Table 5.18. The greatest improvement for path modification can be
observed in the low temperature heating test 17L: 15.6% for capacity and 17.6%
for COP, as shown in Table 5.19.
The refrigerant temperatures along the cycle for retrofit and path modification
are compared in Figures 5.2.7 and 5.2.8 for cooling test A and heating test 47S. In
these figures, the x-axis is not to scale; it indicates just the sequence of each
temperature probe as the refrigerant passes on its way through the cycle. The first
point is the compressor discharge and the last point is the compressor inlet. The
profile for the cooling of R407C is similar to that of R22. The heating case shows
a higher compressor discharge temperature and a lower evaporating temperature than
that of R22, resulting in a high pressure ratio for R407C. The temperature profiles
in the evaporator and condenser show large differences. These temperature profiles
explain the performance degradation of the retrofit case, which had the large tem-
perature difference because of the near parallel flow configuration. They also explain
the performance improvement due to path modification, which resulted in a small
temperature difference because of the near counter flow configuration. The path
modification can improve the performance by reducing the thermodynamic irrevers-
ibility. Although path modification improved the performance of the R407C over
the retrofit, it still experiences degradation as compared to R22, especially in the
heating mode: a 4.2% lower capacity and a 7.9% lower COP at heating test 47S.
The reason for this degradation is the retrofit was performed after the charge was
optimized at cooling test A conditions without optimization at the heating test
conditions, so degradation of the heating performance is natural. These results
indicate that more extensive optimization is necessary, such as the soft optimization.
Switch
Over Valve
Indoor Check Valve
Coil
Comp Outdoor
Coil
Short Tube
Thermostatic
FLOW DIRECTION Expansion
Cooling Valve
Heating
Path Modification Heating
Check Valve
FIGURE 5.2.6 Schematic cycle diagram of path modification (Hwang et al., 1995).
TABLE 5.17
Comparison of ASHRAE Cooling A Test Results
Case R22 Base R407C Retrofit R407C Path Modification
TABLE 5.18
Comparison of ASHRAE Heating 47S Test Results
Case R22 Base R407C Retrofit R407C Path Modification
TABLE 5.19
Comparison of ASHRAE Cooling B and Heating 17L Test Results
Case R22 Base R407C Retrofit R407C Path Modification
100
R-22 Baseline
R-407C Retrofit
R-407C Path Modification
80
60
Temperature [°C]
40
20
0
Compressor Condenser Condenser Condenser Condenser Condenser TEV Evaporator Evaporator Compressor
Discharge -Inlet -1/4 -Middle -3/4 -Outlet -Inlet -Inlet -Outlet -Suction
FIGURE 5.2.7 Refrigerant temperature along the cycle in the cooling mode (Hwang et al.,
1995).
100
R-22 Baseline
R-407C Retrofit
R-407C Path Modification
80
Temperature [°C]
60
40
20
0
Compressor Condenser Condenser Condenser Condenser Condenser TEV Evaporator Evaporator Compressor
Discharge -Inlet -1/4 -Middle -3/4 -Outlet -Inlet -Inlet -Outlet -Suction
FIGURE 5.2.8 Refrigerant temperature along the cycle in the heating mode (Hwang et al.,
1995).
In this section, two sets of drop-in tests conducted by NIST are introduced. The first
set of tests was conducted with a zeotropic refrigerant, a R13B1/R152a (65/35 wt.%)
mixture in a relatively unmodified residential heat pump designed for R22 (Mulroy
and Didion, 1986). The second set of tests was conducted with two R32 mixtures,
The purposes of the first set of tests were to investigate the potential benefits of
zeotropic mixtures in terms of thermodynamic efficiency by utilizing gliding tem-
peratures during phase change and capacity modulation by using mass fraction shift.
To fulfill these purposes, the following four tests were conducted with a nominal 3
refrigeration ton, split system, air-to-air unitary heat pump:
1. Pure retrofit tests without any modification of the test heat pump unit.
2. Simple cycle optimization tests in terms of a refrigerant charge and orifice
tube choice.
3. Effect of the zeotropic mixture mass fraction shift on the performance.
4. System with a receiver to evaluate its effect on the zeotropic mixture mass
fraction shift.
32000
22000
Capacity (Btu/hr)
34000
Capacity (Btu/hr)
Capacity (Btu/hr)
30000
32000 21000
28000
30000 20000
8.0 1.9
2.7
COP
COP
FIGURE 5.2.9 Charge and expansion device optimization test results (Mulroy and Didion, 1986). (a) R22. (b) R32/R134a (34/66 wt.%).
(a) Test A and Test B. (b) Test 47S. (c) Test 17L.
TABLE 5.20
Comparison of Operating Pressures at Test A
Refrigerant R22 R13B1/R152a
Expansion device Capillary tube Orifice 2.0 mm Capillary tube Orifice 2.1 mm
Discharge pressure [kPa] 1917 1924 2816 2048
Discharge temperature [˚C] 111 117 105 98
Suction pressure [kPa] 572 628 571 627
Pressure ratio 3.35 3.06 3.83 3.27
and 6% greater than at Test B for R22 and R13B1/R152a, respectively. However,
those conditions providing the maximum COP at Test B determine the optimum
orifice size and charge since Test B has more bin hours. Another purpose of the
optimization of an expansion device is to obtain a minimal high side pressure and
subcooling. Table 5.20 compares the operating pressures of two refrigerants. As can
be seen from Table 5.20, the cycle with an optimized orifice has a pressure ratio
reduced by 9 and 15% as compared to the drop-in tests with a capillary tube for
R22 and R13B1/R152a, respectively. The suction pressure increases for the orifice
cases due to the reduced degree of superheating at the evaporator outlet as an outcome
of an orifice size optimization. The discharge temperature of R13B1/R152a is about
10 K lower than that of R22, as shown in Table 5.20.
As can be seen from Figure 5.2.9, the performance of the R13B1/R152a mixture
was lower than that of R22 in general: a 3% loss in COP for Test 47S, a 12% loss
in capacity and 11% loss in COP for the cooling test, except a 6% improvement in
capacity and no change in COP for Test 17L (–8.3˚C). One of the reasons for this
performance degradation is an inability to utilize a temperature glide of the zeotropic
mixtures. There are two barriers to utilize the temperature glide of zeotropic mixtures
in an air-to-air heat pump or air conditioner: the crossflow heat exchange configu-
ration for the most widely accepted heat exchangers and the refrigerant flow reversal
when switching the operating mode from cooling to heating for heat pumps. The
second reason is that these tests were conducted on a drop-in basis without any
major system modification.
To increase the heating capacity at low ambient temperatures, it was necessary
to sacrifice the subcooling at the condenser outlet. To investigate the performance
potential of the R13B1/R152a mixture with subcooling and a large mass fraction
shift, a high mass fraction test was conducted by charging a premixed mixture at
86% of R13B1. This investigation of the mass fraction shift on the capacity modu-
lation showed that the heating capacity at –8.3˚C increased 18% above that with
R22 when the R13B1 mass fraction was 86% (Table 5.21). This mass fraction shift
strategy would contribute to increased seasonal performance in cold climates by
reducing the need for electric heat.
To realize this benefit, a receiver was added in the liquid line to create a mass
fraction shift at low air temperatures. The volume of the receiver was 2.3 l, which
TABLE 5.21
Performance of High R13B1 Mass Fraction
Refrigerant R22 R13B1/R152a (86/14 wt.%)
Moreover, the performance of this binary mixture has shown an equivalent perfor-
mance to that of R22 with a counter flow heat exchange configuration (Rothfleisch
and Didion, 1993).
Retrofit Test of R32/R134a
Similar to the tests for R13B1/R152a, a retrofit test was conducted with an optimi-
zation of the expansion device opening and refrigerant charge at ASHRAE Test A
condition. The optimum combination of expansion device opening and refrigerant
charge were determined when that combination yields the maximum COP. The
system was then tested at ASHRAE Test B condition with the optimum combination
determined from Test A.
In general, the residential heat pumps have two expansion devices, one for
cooling and the other for heating. The second optimization for the heating mode
was conducted at the ASHRAE Test 47S condition similar to the cooling mode
optimization.
The results of the cooling mode optimization are shown for R22 and R32/R134a
in Figure 5.2.10. In these tests, the degree of subcooling was adjusted by changing
the refrigerant charge and the degree of superheating at the evaporator outlet was
adjusted by the expansion valve opening between 1.7 and 2.8˚C. The highest capacity
was obtained at a minimum of superheating. Therefore, for each value of subcooling,
the expansion valve opening was optimized. The optimum values of subcooling were
approximately 5.6 and 2.8˚C for R22 and R32/R134a, respectively, as shown in
Figure 5.2.10.
The same optimization procedures applied for the heating mode. The results of
the heating mode optimization for R22 showed that there was no definite maximum
at various values of the subcooling, as shown in Figure 5.2.11. As can be seen from
this result, the maximum COP is achievable when the subcooling becomes zero.
However, the degree of subcooling decreases as the ambient temperature drops.
Therefore, if the subcooling were set to zero at Test 47S conditions, the condenser
outlet would be two-phase at Test 17L conditions. When the ambient temperature
drops, the heating load increases but the heating capacity decreases. If the condenser
outlet became two-phase, the heating capacity would further decrease. Therefore,
the heating optimization should be conducted while balancing the heating COP at the
high ambient temperature and the heating capacity at the low ambient temperature.
Figures 5.2.12 through 5.2.14 show the effect of subcooling on the capacity,
COP and condensing pressure. As the subcooling increases, the capacity increases
since the latent heat of evaporation increases due to the lower evaporator inlet
enthalpy. As the subcooling increases, the power increases since the condensing
pressure increases as shown in Figure 5.2.14. The reason the condensing pressure
increases is that as the subcooling increases, a smaller portion of the condenser is
used for the two-phase and superheated gas heat transfer.
The only way to maintain the same capacity or even increase capacity is to
increase the temperature difference if the condenser UA is constant. In other words,
the condensing temperature and condensing pressure increase and result in an
increased pressure ratio. When the compression pressure ratio increases, the com-
pressor delivers less mass flow rate and the power per unit refrigerant mass increases.
200 Vapor Compression Heat Pumps with Refrigerant Mixtures
3.46
3.44
3.42
COP
3.4
3.38
3.36
3.34
1.7 2.8 3.9 5 6.1 7.2 8.3 9.4
(3) (5) (7) (9) (11) (13) (15) (17)
Subcooling °C (°F)
(a)
3.1
3.08
3.06
3.04
3.02
COP
2.98
2.96
2.94
2.92
1.1 2.2 3.3 4.4 5.8 6.7 7.8 8.9 10
(2) (4) (6) (8) (10) (12) (14) (16) (18)
Subcooling °C (°F)
(b)
3.92
3.9
3.88
3.86
3.84
COP
3.82
3.8
3.78
3.76
3.74
1.7 2.6 3.9 5 6.1 7.2
(3) (5) (7) (9) (11) (13)
Subcooling °C (°F)
FIGURE 5.2.11 Heating mode optimization for R22; x-flow: crossflow, x-para: cross-parallel
flow (Rothfleisch and Didion, 1993).
In summary, there are two competing effects to be balanced: a positive effect on the
COP by a larger latent heat and a negative effect by an increased condensing pressure.
One important characteristic of the R32/R134a mixture obtained from these
figures is that the capacity of R32/R134a increases slower than that of R22 while
the power increases faster than that of R22 when the subcooling increases. This fact
means the optimum COP of R32/R134a occurs at a lower subcooling than that of
R22. This feature can be understood from the vapor pressure curve of these two
refrigerants. R32/R134a has a greater slope than that of R22; therefore, the greater
slope of R32/R134a causes a faster condensing pressure increase than R22 at the
same temperature increase.
Since the retrofit approach was used in this study, the heat exchanger configu-
ration was the same for the two refrigerants but it changed from crossflow (referred
to as x-flow) to cross-parallel flow (referred to as x-para) when the operating mode
switched from cooling to heating. There was a definite optimum point for R22
cooling optimization results but no optimum point for the R22 heating optimization
results, as shown in Figures 5.2.10 and 5.2.11.
Since these tests were done for the same R22, the different results are believed
due to the difference in the condenser configuration. In fact, with help from Figures
5.2.12 through 5.2.14, one can explain this phenomenon. The outlet refrigerant of
the cross-parallel flow heat exchanger will approach the exiting air temperature while
that of the crossflow heat exchanger will approach the entering air temperature.
202 Vapor Compression Heat Pumps with Refrigerant Mixtures
(9) 2.64
Capacity kW (kBtu/h)
(8.8) 2.58
(8.6) 2.52
(8.4) 2.46
(8.2) 2.40
(8) 2.34
1.7 2.8 3.9 5 6.1 7.2 8.3 9.4 10.6
(3) (5) (7) (9) (11) (13) (15) (17) (19)
Subcooling °C (°F)
R22 Cool (x-flow) R22 Heat (x-para) mtx Cool(x-flow)
FIGURE 5.2.12 Capacity vs. subcooling; x-flow: crossflow, x-para: cross-parallel flow (Roth-
fleisch and Didion, 1993).
840
820
800
780 A
Power (watts)
760
B
740
720
700
680 C
660
1.7 2.8 3.9 5 6.1 7.2 8.3 9.4 10.6
(3) (5) (7) (9) (11) (13) (15) (17) (19)
Subcooling °C (°F)
R22 Cool (x-flow) R22 Heat (x-para) Mix Cool (x-flow)
FIGURE 5.2.13 Power vs. subcooling; x-flow: crossflow, x-para: cross-parallel flow (Roth-
fleisch and Didion, 1993).
Therefore, the condenser outlet refrigerant temperature in the heating mode should
be higher than that in the cooling mode since the approach temperature cannot be
negative. The condensing temperature in the heating mode should then be higher
than that in the cooling mode when the same degree of subcooling is maintained.
This reason is why the power and condensing pressure in the heating mode increases
faster than in the cooling mode.
Experimental Performance Measurements 203
(310) 2137
(270) 1862
B
(250) 1724
(230) 1586
1.7 2.8 3.9 5 6.1 7.2 8.3 9.4 10.6
(3) (5) (7) (9) (11) (13) (15) (17) (19)
Subcooling °C (°F)
R22 cool (x-flow) R22 heat (x-para) Mix cool (x-flow)
FIGURE 5.2.14 Condensing pressure vs. subcooling (Rothfleisch and Didion, 1993).
TABLE 5.22
Retrofit Performance of R32/R134a as compared to R22
Test Condition [˚C] Normalized Capacity Normalized COP Normalized Power
The retrofit performance data of R32/R134a normalized by R22 results for the
same test are summarized in Table 5.22. This table shows the lower capacity and
COP by about 6 to 10% but the heating capacity increases 4% at the –8.3˚C heating
test due to the passive mass fraction shift.
One of the reasons for the lower performance of R32/R134a is due to the lower
volumetric capacity of R32/R134a. The latent heat of R32/R134a was 17% higher
than that of R22 but the suction density was 21% lower than that of R22 for the
cooling Test A condition, which resulted in an 8% lower volumetric capacity. Figure
5.2.15 compares two refrigerant cycles in the ln(P)-h diagram. From Figure 5.2.15,
a higher condensing pressure and larger latent heat of evaporation of R32/R134a
than those of R22 can be easily recognized.
SLHX Test of R32/R134a
To investigate the cycle options for the R32/R134a mixture performance improve-
ment, further tests with SLHX were conducted in the cooling mode. The low-pressure
204 Vapor Compression Heat Pumps with Refrigerant Mixtures
(300) 2068
(220) 1517
(180) 1241
(140) 965
(100) 689
(60) 414
(20) 138
0 47 93 140 186 233 279 326 372
(0) (20) (40) (60) (80) (100) (120) (140) (160)
Enthalpy kJ/kg (Btu/lb)
34% R32/66% R134a R22
FIGURE 5.2.15 P-h diagram of R22 and R32/R134a (34/66 wt.%) (Rothfleisch and Didion,
1993).
Bottom Circuit EVAP Temp Profile Top Circuit EVAP Temp Profile
35°C (95°F) Cooling 35°C (95°F) Cooling
(60) 15.6 (74) 23.3
(58) 14.4
(70) 21.1
Temperature °C (°F)
(56) 13.3
Temperature °C (°F)
FIGURE 5.2.16 Evaporator temperature profile of R22 and R32/R134a (34/66 wt.%). (a)
Bottom circuit. (b) top circuit. (Rothfleisch and Didion, 1993).
inlet side of the SLHX can be either single-phase or two-phase by adjusting the
expansion valve. Both test results at Test A and B conditions are summarized in
Table 5.23. From the test results, the SLHX cycle with optimized expansion valve
opening at Test A condition showed a performance degradation at Test B condition,
even worse than without SLHX.
For the cycle without SLHX, the refrigerant migrates from the evaporator to the
condenser when the ambient temperature drops since the compressor delivers more
Experimental Performance Measurements 205
TABLE 5.23
SLHX Performance of R32/R134a as compared to R22
Refrigerant Phase
Entering Low-pressure Normalized Normalized Normalized
Test Condition [˚C] Side of SLHX Capacity COP Power
Note: The expansion valve was adjusted at the 35˚C test condition.
a The expansion valve was adjusted at the 27.8˚C test condition.
refrigerants to the condenser, while the expansion device further limits the refrigerant
flow due to the reduced pressure ratio. For the cycle with SLHX, the refrigerant
flow through the expansion valve is limited due to the increased subcooling by the
SLHX, which contributes to the lower performance. When the expansion valves
were adjusted at Test B (27.8˚C) condition, the performance was improved for Test B
while Test A had an identical performance.
Conclusions
• The heating optimization should be conducted while balancing the heating
COP at the high ambient temperature and the heating capacity at the low
ambient temperature.
• The compressor power increases as the degree of subcooling increases.
The capacity of R32/R134a increases slower than that of R22, while the
power increases faster than that of R22 when the subcooling increases.
This fact means the optimum COP of R32/R134a occurs at lower sub-
cooling than that of R22.
• The retrofit performance of R32/R134a showed the lower capacity and
COP than those of R22 by about 6 to 10%. The heating capacity increased
by 4% at the –8.3˚C heating test due to the passive mass fraction shift.
• The SLHX cycle with optimized expansion valve opening at Test A
condition showed a performance degradation at Test B condition, even
worse than without SLHX. When the expansion valves were adjusted at
Test B (27.8˚C) condition, the performance was improved for Test B while
Test A had an identical performance.
206 Vapor Compression Heat Pumps with Refrigerant Mixtures
REFERENCES
Air-Conditioning and Refrigeration Institute, 1994, “Unitary Air-Conditioning and Air Source
Heat Pump Equipment,” ARI Standard 210/240.
Allied Signal Chemicals, 1995, Genetron AZ-20 Product Brochure.
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., 1995,
“Methods of Testing for Seasonal Efficiency of Unitary Air-conditioners and Heat
Pumps,” ASHRAE STANDARD ANSI/ASHRAE 116.
ASHRAE Handbook of Fundamentals, 1985, Chapter 16, “Refrigerants,” American Society
of Heating, Refrigeration and Air Conditioning Engineers, Atlanta, GA.
Calm J.M., 1995, “Refrigerants and Lubricants Data for Screening and Application,”
Proceedings of the 1995 International CFC and Halon Alternatives Conference, Wash-
ington D.C., pp. 169–178.
Conklin, J.C. and E.A. Vineyard, 1991, “Cycle Performance Comparison between a Non-
azeotropic Refrigerant Mixture and R22,” Proceedings of XVIIIth International
Congress of Refrigeration, Vol. III, pp. 1270–1274.
Didion, D.A. and D. Bivens, 1989, “The Role of Refrigerant Mixtures as Alternatives,” CFC
Technical Conference ’89 at the National Institute of Standards and Technology,
American Society of Heating, Refrigeration and Air Conditioning Engineers, Atlanta,
GA.
Domanski, P.A. and M.O. McLinden, 1992, “A Simplified Cycle Simulation Model for the
Performance Rating of Refrigerants and Refrigerant Mixtures,” International Journal
of Refrigeration, Vol. 15, No. 2.
DuPont, 1994, Suva Refrigerants Technical Information.
DuPont, 1995, Suva Alternative Refrigerants Technical Information.
Gallagher, J., M. McLinden, and M. Huber, 1993, REFPROP NIST Thermodynamic Proper-
ties of Refrigerants and Refrigerant Mixtures Database, Version 4.01, Thermophysics
Division of National Institute of Standards and Technology, Gaithersburg, MD.
Godwin, D., 1993, “Results of System Drop-In Tests in ARI’s R-22 Alternative Refrigerants
Evaluation Program,” ARI, Arlington, VA.
Hwang, Y., J. Judge, and R. Radermacher, 1995, “An Experimental Evaluation of Medium
and High Pressure HFC Replacements for R22,” Proceedings of the 1995 International
CFC and Halon Alternatives Conference, Washington D.C., pp. 41–48.
Kauffeld, M., W. Mulroy, M. McLinden, and D. Didion, 1990, “An Experimental Evaluation
of Two Nonazeotropic Refrigerant Mixtures in a Water-To-Water, Breadboard Heat
Pump,” NISTIR 90-4290, National Institute of Standards and Technology, Gaithers-
burg, MD.
Kolliopoulos, K.G., 1994, “Update on Developmental HFC Replacements for HCFC-22,”
Presented at the 1994 International CFC and Halon Alternatives Conference, Wash-
ington, D.C.
Kruse, H., 1980, “The Advantages of Non-Azeotorpic Refrigerant Mixtures for Heat Pump
Application,” Mons (Belgium) Meeting of Commissions D1, D2, E1 and E2 of the
IIR.
McLinden, M. and R. Radermacher, 1987, “Methods for Comparing the Performance of Pure
and Mixed Refrigerants in the Vapor Compression Cycle,” International Journal of
Refrigeration, Vol. 10, No. 6.
Morrison, G. and M. McLinden, 1986, “Application of a Hard Sphere Equation of State to
Refrigerants and Refrigerant Mixtures,” NBS Technical Note 1226, National Institute
of Standards and Technology, Gaithersburg, MD.
Experimental Performance Measurements 207
Mulroy, W. and D.A. Didion, 1986, “The Performance of a Conventional Residential Sized
Heat Pump Operating with a Nonazeotropic Binary Refrigerant Mixture,” NBSIR
86-3422, National Institute of Standards and Technology, Gaithersburg, MD.
Mulroy, W., M. Kauffeld, M. McLinden, and D.A. Didion, 1988, “An Evaluation of Two
Refrigerant Mixtures in a Breadboard Air Conditioner,” Proceedings of the IIR Con-
ference at Purdue, West Lafayette, IN.
Mulroy, W., M. Kauffeld, M. McLinden and D.A. Didion, 1988, “Experimental Evaluation
of Two Refrigerant Mixtures in a Breadboard Air Conditioner,” Proceedings of the
2nd DOE/ORNL Heat Pump Conference, CONF-8804100.
Pannock, J. and D.A. Didion, 1991, “The Performance of Chlorine-Free Binary Zeotropic
Refrigerant Mixtures in a Heat Pump,” NISTIR 4748, National Institute of Standards
and Technology, Gaithersburg, MD.
Quast, U. and H. Kruse, 1986, “Experimental Performance Analysis of a Reciprocating
Compressor Working with Non-Azeotropic Refrigerant Mixtures,” Proceedings of
the International Compressor Engineering Conference at Purdue, West Lafayette, IN.
Radermacher, R., 1986, “Advanced Versions of Heat Pump with Zeotropic Refrigerant Mix-
tures,” ASHRAE Transactions, Vol. 92, Pt. 2, pp. 52–59.
Rothfleisch, P.I. and D.A. Didion, 1991, “A Performance Evaluation of a Variable Speed,
Mixed Refrigerant Heat Pump,” NISTIR 93-5321, National Institute of Standards and
Technology, Gaithersburg, MD.
Rothfleisch, P.I. and D.A. Didion, 1993, “A Study of Heat Pump Performance Using Mixtures
of R32/R134a and R32/R125/R134a as Drop-In Working Fluids for R22 with and
without a Liquid-Suction Heat Exchanger, NISTIR 5321, National Institute of Stan-
dards and Technology, Gaithersburg, MD.
Vakil, H., 1981, “Means and Method for the Recovery of Expansion Work in a Vapor
Compression Cycle Device,” United States Patent No. 4034099.
Vineyard, E.A, 1988, “Laboratory Testing of a Heat Pump System Using an R13B1/R152a
Refrigerant Mixture,” ASHRAE Transactions, Vol. 94, Pt. 1, pp. 292–303.
Vineyard, E.A. and J.C. Conklin, 1991, “Cycle Performance Comparison between a Non-
azeotropic Refrigerant Mixture and R22,” Proceedings of XVIIIth International Con-
gress of Refrigeration, Vol. II, pp. 543–547.
Vineyard, E.A. and J.C. Conklin, 1993, “Cycle Performance Testing of Nonazeotropic
Mixtures of HFC32/HFC124 with Enhanced Surface Heat Exchangers,” ASHRAE
Transactions, Vol. 99, Pt. 2, pp. 97–103.
Vineyard, E.A., J.C. Conklin, and A.J. Brown, 1993, “Cycle Performance Testing of Non-
azeotropic Mixtures of HFC-143a/HCFC-124 and HFC32/HCFC124 with Enhanced
Surface Heat Exchangers, ASHRAE Transactions, Vol. 99, Pt. 2, pp. 97–103.
6 Refrigerant Mixtures
in Refrigeration
Applications
As the global economy and technology have improved dramatically since World
War II, the demand for human comfort has resulted in an increased demand for
household refrigerators and freezers. Energy usage has increased also by the
increased use of energy-consuming products. This change of lifestyle generates the
demand to provide additional refrigeration capacity with increased energy efficiency.
The introduction of refrigerant mixtures was considered mainly due to the increased
demand of the larger refrigeration capacity, the better control of two evaporator
temperature levels and the potential of improved energy efficiency. Refrigerators
with two temperature levels were considered good applications for refrigerant mix-
tures because of one main inherent thermodynamic irreversibility: in conventional
refrigerators and freezers, the higher temperature food compartment is cooled by an
evaporator operating on the freezer temperature level.
209
210 Vapor Compression Heat Pumps with Refrigerant Mixtures
Condenser
3 2'
3A 2'A
Expansion Compressor
Device 4'A 5A
Airflow paths
Damper 4'A 6A
4'A 1A
4' 1
Evaporator
FIGURE 6.1.1 Single evaporator refrigeration cycle with two temperature level compartments.
Refrigerating capacity
Lost expansion work
2'
Work due to superheating
P1
Temperature (T)
3 2
3A 2'A
1A
4'A
1
4 4'
P2 Q
Entropy (s)
FIGURE 6.1.2 T-s diagram of single evaporator refrigeration cycle with two temperature
level compartments.
To maintain the temperature in the freezer compartment at the desired level, the
low evaporation temperature must also be used to cool the food compartment. Thus,
the temperature lift, the temperature difference between the condenser and the
evaporator, is necessarily greater for the SETC than for the appliances with SEOC.
The inherent problem with SETC is that it does not take advantage of the smaller
temperature lift available for the food compartment, resulting in a lower overall
COP. The main reasons for employing the SETC concept are space, cost savings
and excellent reliability. Since the Montreal Protocol declared a ban on CFCs
(chlorofluorocarbons) used as refrigerants and foaming agents of refrigeration sys-
tems, there has been an effort to find CFC substitutes for the single evaporator
refrigeration cycle.
Refrigerant Mixtures in Refrigeration Applications 211
TABLE 6.1
Simulation Results of Refrigerants Examined
Volumetric Mixed Volumetric
Pure Refrigerant COP Capacity [kg/m3] Refrigerant COP Capacity [kg/m3]
Stocker and Walukas simulated the performance of the SETC with R12/R114
mixtures (1981). In their model, the condenser outlet and the evaporator outlet were
assumed to be a saturated liquid and vapor, respectively. In their study, they showed
when the mass fraction of R114 is 0.5, the power consumption is 11% less than
when the mass fraction of R114 is 0.01.
As a preliminary study to screen potential alternative refrigerants, Jung and
Radermacher (1991) presented their simulation results for 15 pure and 21 mixed
refrigerants. A list of refrigerants examined with results are shown in Table 6.1.
Based on these results, they concluded that there is no drop-in pure refrigerant
because R134, R134a and R152a have a 7 to 30% lower capacity than R12. More-
over, R22, R32, R152a and R142b have comparable COPs to R12 but have far
different capacities than R12. As drop-in refrigerants, they recommended R22/R142b,
R22/R152a and R32/R142b.
In an experimental effort, Vineyard et al. (1989) investigated the energy-saving
potential of five refrigerants: R12, R500, R12/dimethylether (DME), R22/R142b
and R134a. From their drop-in tests for these five refrigerants in a 0.5 m3 top-mount
refrigerator-freezer, they found that azeotropic mixtures, R500 and R12/DME saved
212 Vapor Compression Heat Pumps with Refrigerant Mixtures
Condenser
3 2'
3A 2'A
Expansion Compressor
Device 4'A 5A 6A 1A
4' 1
5
1st Evaporator 2nd Evaporator
FIGURE 6.2.1 Dual evaporator refrigeration cycle with two temperature level compartments.
Refrigerant Mixtures in Refrigeration Applications 213
Refrigerating capacity
Lost expansion work 2'
Work due to superheating
Additional work
P1
Temperature (T)
3 2
3A 2'A
6 6A 7A
4'A 5A 7
P3
4 4' 5 1
P2 Q
Entropy (s)
FIGURE 6.2.2 T-s diagram of dual evaporator refrigeration cycle with two temperature level
compartments.
the first evaporator and the food compartment with the second evaporator. A T-s
diagram of SETC is shown in Figure 6.2.2. Dotted lines (5A–4′A and 7A–6A) are
air temperatures of the freezer compartment and the food compartment, respectively,
that vary along the length of the first and second evaporators. The dotted line
(3A–2′A) is the air temperature along the condenser. The requirement of having two
separate compartment temperatures results in an additional loss for dual-evaporator
refrigeration systems as clearly represented by the area P3. As for the SETC, the
lower evaporation temperature must be used for the food compartment also in order
to maintain the lower temperature in the freezer compartment. The inherent problem
with DETC is the same as that of SETC, resulting in a low overall COP.
Refrigerating capacity
Lost expansion work 2'
Work due to superheating
P1
Temperature (T)
2
3
2'A
P3
3A 6A
5A 1A
4'A 1
4 4' 5
P3 Q
Entropy (s)
FIGURE 6.3.1 T-s diagram of dual evaporator Lorenz cycle with two temperature level
compartments.
Condenser
3 2'
3A 2'A
3I
3II Compressor
Internal Suction-Line
Heat Exchanger Heat Exchanger
Expansion 1
Device
4'A 5A 6A 7A
4' 7
56
1st Evaporator 2nd Evaporator
The expanded two-phase refrigerant from the expansion device enters the evap-
orator and then passes through the internal heat exchanger, the food temperature
evaporator and the suction heat exchanger, sequentially. Therefore, the internal heat
exchanger and suction line heat exchanger offer more subcooling at the condenser
outlet and at the same time take advantage of and provide an appropriate temperature
glide for both evaporators. Figure 6.3.3 shows the T-s diagram of the Lorenz-
Meutzner cycle. Here the additional work requirement, area P3 of Figure 6.3.1, is
not present because the refrigerant entering the refrigerator evaporator is heated by
Refrigerant Mixtures in Refrigeration Applications 215
Refrigerating capacity
Lost expansion work 2'
Work due to superheating
P1
Te mperature (T)
2
3
3I 2'A
3II 3A
7A
4'A 5A 6A
1
4' 5 6 7
4
P2 Q1 Q2
Entropy (s)
the second heat exchanger from T5 to T6. Therefore, an approach temperature between
the air and refrigerant of the second evaporator is reduced. With the addition of a
food compartment evaporator and the use of a well-chosen zeotropic refrigerant
mixture, evaporation at two temperature levels in a single refrigeration cycle can be
achieved with a higher COP than the conventional vapor compression refrigeration
cycle. This is the concept upon which the Lorenz-Meutzner cycle is based.
Lorenz and Meutzner reported a 20% energy savings for a Lorenz-Meutzner
cycle with a mixture of R22/R11 (50/50 wt.%) compared to a conventional refrig-
erator with R12. Stoecker and Walukas (1981) showed no energy savings for a
mixture of R12/R114 (50/50 wt.%) but they reported that the simulated power of
the Lorenz-Meutzner cycle with the R12/R114 mixture (50/50 wt.%) was 12% less
than that of R12. In their study, they showed that when the mass fraction of R114
is 0.5, the power consumption is 11% less than when the mass fraction of R114 is
0.01. Tiedemann et al. (1991) tested various mixtures but did not find any savings.
The experimental results of Rose et al. (1992) showed a 9% reduction in energy
consumption with a zeotropic mixture of R22/R123. Sand et al. (1993) found that
the Lorenz-Meutzner cycle with R32/R124 mixtures (15/85 wt.%) saved 6 to 7% energy
use over R12. They also emphasized the importance of counterflow, refrigerant-to-
air heat exchangers.
Condenser
3 2'
3A 2'A Compressor
Suction-Line
1
Heat Exchanger
4'A 5A 6A 7A
5 6
4' 7
Expansion 8
Device 10 9
11 2nd Internal
1st Evaporator 2nd Evaporator
Heat Exchanger
Refrigerating capacity
Lost expansion work 2'
Work due to superheating
P1
Te mperature (T)
2
3
2'A
98
10 3A
11 7A
4'A 5A 6A
1
4' 5 6 7
4
P2 Q1 Q2
Entropy (s)
shown in Figure 6.4.1. This is an example for implementing the three paths evapo-
rator design that is discussed in Chapter 4. The difference between the Lorenz-
Meutzner cycle and the modified Lorenz-Meutzner cycle is shown in the T-s diagram
of the modified Lorenz-Meutzner cycle refrigerator, Figure 6.4.2. The modified Lorenz-
Meutzner cycle refrigerator has a four step subcooling process, while the
Lorenz-Meutzner cycle refrigerator has only two steps of subcooling. In the modified
Lorenz-Meutzner cycle, the first and second evaporators serve as three-way heat
exchangers made by a smaller tube (liquid line) inserted into a larger tube (suction
line and evaporator). The compartment air surrounding the bigger tube and subcooled
liquid in the smaller tube rejects heat to the two-phase refrigerant running counterflow
Refrigerant Mixtures in Refrigeration Applications 217
through the annulus of the two tubes. Zhou et al. (1994) reported 16.6, 14.6 and
16.7% energy savings from their experiments with a modified-Lorenz-Meutzner
cycle refrigerator with zeotropic mixtures R22/R123, R290/n-pentane and R290/R600,
respectively. The conventional SETC uses a damper to control the food compartment
temperature and a thermostat to control the freezer compartment temperature. The
main purpose of the dual evaporators is to provide better temperature control.
For the Lorenz-Meutzner cycle, however, independent temperature control of
the two compartments is not as simple since there is no damper. The temperature
conditions of both compartments compared to their temperature settings determine
the operation of the compressor. Therefore, one compartment temperature can be
dropped further than the set temperature when the other compartment temperature is
higher than the temperature set point. In this case, failure of implementing indepen-
dent temperature control results in an energy loss. Simmons et al. (1995) suggested
an independent temperature control of compartments in a modified Lorenz-Meutzner
cycle by using a bistable solenoid valve as shown in Figure 6.4.3. The suggested
control of a modified Lorenz-Meutzner cycle directs refrigerant through alternate
evaporator paths with a bistable solenoid valve that regulates the entire flow of
refrigerant to only the freezer evaporator or to both evaporators.
When the bistable solenoid valve is at position A (Mode A), the cycle becomes
a modified Lorenz-Meutzner cycle as shown in Figure 6.4.3. When the bistable
solenoid valve is at position B (Mode B), the cycle becomes SEOC with two serial
suction line heat exchangers. When both compartments need cooling, the bistable
valve switches to Mode A. When only the freezer compartment needs cooling, the
bistable valve switches to Mode B. Therefore, the suggested control method for a
modified Lorenz-Meutzner cycle provides a similar independent temperature control
to the SETC with the damper.
Condenser
3 2'
3A 2'A
Bistable Solenoid Valve B A
3I Compressor
3II
5I
1st Internal Suction-Line
1
Expansion Heat Exchanger Heat Exchanger
5"
Device B
4'A 5A
6A 7A
4" 5 6
4' 7
Expansion 8
Device A 10 9
11 2nd Internal
1st Evaporator 2nd Evaporator
Heat Exchanger
REFERENCES
Ashley, C., 1945, U.S. Patent No. 2492725.
Dmitriev, W. et al., 1974, “Comparative Tests of Hermetic Compressor FR-O.15 on R12 and
Non-azeotropic Mixture of R12/R143,” Cholodilnaja Technika, Vol. 51, No. 6, pp.
25–27.
Jung, D.S. and R. Radermacher, 1991, “Simulation of a Single-Evaporator Refrigerator,”
International Journal of Refrigeration, Vol. 14, pp. 223–232.
Kuijpers, L.J.M., J.A. de Wit, A.A.J. Benschop and M.J.P. Janssen, 1990, “Experimental
Investigations into the Ternary Blend HCFC-22/124/152A as a Substitute in Domestic
Refrigeration,” IIR Purdue Refrigeration Conference.
Liu, Z., I. Haider, and R. Radermacher, 1995, “Simulation and Test Results of Hydrocarbon
Mixtures in a Modified-Lorenz-Meutzner Cycle Domestic Refrigerator,” HVAC&R
Research, Vol. 1, No. 2, pp. 127–142.
Lorenz, A. and K. Meutzner, 1975, “On application of non-azeotropic two component refrig-
erants in domestic refrigerators and home freezers,” XIV International Congress of
Refrigeration, Moscow.
Phillipp, L., 1948, U.S. Patent No. 2483842.
Radermacher, R. and D.S. Jung, 1992, “Refrigeration System,” U.S. Patent No. 5,092,138.
Radermacher, R. and D.S. Jung, 1993a, “Refrigerator System for Two-Compartment Cooling,”
U.S. Patent No. 5,235,820.
Radermacher, R. and D.S. Jung, 1993b, “Subcooling System for Refrigeration Cycle,” U.S.
Patent No. 5,243,837.
Radermacher, R., H. Ross and D.A. Didion, 1983, “Experimental Determination of Forced
Convection Evaporative Heat Transfer Coefficients for Non-Azeotropic Refrigerant
Mixtures,” ASME Winter Annual Meeting, Boston, MA, Paper No 83-WA/HT-54,
Nov., pp. 13–18.
Radermacher, R., K. Kim, W. Kopko and J. Pannock, 1995, U.S. Patent No. 5,406,805.
Rose, B., D.S. Jung and R. Radermacher, 1992, “Testing of Domestic Two-evaporator Refrig-
erators with Zeotropic Refrigerant Mixtures,” ASHRAE Transactions, Vol. 98, Pt. 2,
pp. 216–226.
Sand, J.R., E.A. Vineyard and V.D. Baxter, 1993, “Laboratory Evaluation of An Ozone-Safe
Nonazeotropic Refrigerant Mixture in a Lorenz-Meutzner Refrigerator Freezer
Design,” ASHRAE Transactions, Vol. 93, Pt. 1, pp. 1467–1481.
Schubert, A. and Herrick, C., 1964, U.S. Patent No. 3,019,614.
Stoecker, W.F. and D.J. Walukas, 1981, “Conserving Energy in Domestic Refrigerators
through the Use of Refrigerant Mixtures,” International Journal of Refrigeration,
Vol. 4, pp. 201–208.
Tiedemann, T., M. Kauffeld, K. Beermann and H. Kruse, 1991, “Evaluation of ozone-safe,
low greenhouse warming potential zeotropic refrigerant mixtures in household refrig-
erators/freezers,” XVIIIth International Congress of Refrigeration, Montreal, Canada.
Refrigerant Mixtures in Refrigeration Applications 219
221
222 Vapor Compression Heat Pumps with Refrigerant Mixtures
storage of one component by exploiting the mass fraction difference in the two-
phase state provides the opportunity to control the mixture mass fraction. Changes
of mixture mass fraction affect the density and latent heat of the circulating mixture,
resulting in capacity changes. Some options of capacity control by using refrigerant
mixtures are introduced here.
Using the method of modifying mixture mass fraction to adjust capacity is most
likely only suitable in a limited range around the ratio 1:2. Since the compressor
speed remains unchanged, the amount of work needed just to operate the compressor
(friction, oil pump and other sources) is essentially constant. Thus, during low
capacity operation, the fraction of lost work increases relative to compression work
done on the refrigerant and the overall system efficiency decreases.
3A 2'A
3
2'
Condenser
3I Receiver
Va lve C
Compressor
Heater
3II
Expansion
Device Evaporator Va lve D
4' 1
4'A 1A
Assume that a binary zeotropic mixture of R12 and R22 is charged in the cycle.
Normal boiling points of R12 and R22 are –29.8 and –40.8˚C, respectively, and R22
has a higher saturation pressure and larger latent heat than those of R12. For example,
the saturation pressures of R12 and R22 are 423 kPa and 681 kPa, respectively, at
10˚C saturation temperature. The latent heats of R12 and R22 are 148 kJ/kg-K and
196 kJ/kg-K, respectively, at 10˚C saturation temperature. Therefore, an R22-rich
mixture works at higher operating pressures and provides more capacity. When the
system capacity is larger than the load, there is a need for reducing capacity, so
valve C is opened and valve D is closed. Thus, R22 is selectively absorbed into the
molecular sieve. In this case, the R22 mass fraction of the circulating mixture
becomes lower than the originally charged mass fraction and the system capacity is
reduced due to the increase in R12 content. When the system capacity is smaller
than the load, there is a need for restoring system capacity, so valve C is closed and
valve D is opened. An electric heater helps to release the R22 that is absorbed in
the molecular sieve. In this case, the R22 mass fraction of circulating mixture
becomes the same as the originally charged mass fraction and the system capacity
is restored.
A similar effect can be achieved with a single accumulator located at the evap-
orator outlet only. Whenever a two-phase mixture leaves the evaporator, the liquid
phase is stored in the accumulator while the vapor continues to circulate through
the system. The stored liquid phase is rich in the high boiler (A) and the mass fraction
of the circulating mixture is shifted toward a higher content of the low boiler (B)
and the system capacity is increased.
Many air conditioners that also operate as heat pumps require a suction accu-
mulator to store excess charge during the heat pumping mode. Thus, this accumulator
causes a shift in mass fraction, increasing capacity during heating only. Using only
one accumulator causes a shift in mass fraction that is smaller than that possible
with the two accumulator configuration.
3A 2'A
Receiver
3
2'
Condenser
3I
Compressor
Expansion
Device
Evaporator Accumulator
4' 1
5
6
4'A 5A
is filled with a liquid mixture and the low pressure accumulator stores a mixture
that is predominantly rich in component B because most of component A was already
released in the evaporator. Thus, the low pressure accumulator provides a mixture
rich in component A to the compressor and reduces the degree of superheating.
For example, Figures 7.1.3 and 7.1.4 show the saturated vapor density and latent
heat of condensation for the binary mixtures of R22/R114. Boiling points of R22
and R114 are –40.8 and 3.8˚C, respectively. Therefore, R22 works as component A
and R114 works as component B. These figures show the higher mass fraction of
30
80/20 wt.%
50/50 wt.%
25
20/80 wt.%
Density [kg/m3]
20
15
10
−10 −8 −6 −4 −2 0 2 4 6 8 10
Evaporating Temperature [°C]
170
80/20 wt.%
20/80 wt.%
150
140
130
120
40 42 44 46 48 50 52 54 56 58 60
Condensing Temperature [°C]
R22 provides a higher saturated vapor density and latent heat of condensation at the
same temperature. Therefore, the first case of filled high pressure accumulator
provides smaller capacity due to the lower R22 mass fraction in the circulating
mixture. The second case of a filled low pressure accumulator provides a larger
capacity due to the higher R22 mass fraction in the circulating mixture. In this way,
the system capacity can be adjusted by the control of mixture mass fraction. Figure
7.1.5 shows the results of capacity modulation relative to the load variation.
nt
ref led
era
int f il
rig
po tor
ng la
igh e a
n h sur
t
fri d
ran
re ille
h o res
ge
int r f
ric w p
po ato
Lo
ng ul
ili um
bo acc
ow er
n l ssu
h o re
ric gh p
Hi
3A 2'A
Receiver
7
3
2'
Condenser
Separator
8
3I
Compressor
Expansion
Device
Evaporator
Accumulator
4' 1
5
6
4'A 5A
TABLE 7.1
Comparison of Capacity Control (R22/R114 with 50/50 mole%)
Minimum Minimum
Capacity Maximum for TAC for MTAC
8A
Sub-Condenser
8 7 3A 2'A
Parallel
Receiver 2'
9 Condenser
3
Receiver
Compressor
3I
Expansion
Device Evaporator Accumulator
4' 1
5
6
4'A 5A
from 38 to 100% using an R22/R114 binary mixture (50/50 mol.%) and from 28 to
100% using an R23/R22/R114 ternary mixture (10/60/30 mol.%).
6A 5A 3A 2'A
Sub-Condenser
5 3
6 2'
7 Condenser
Rectifier
Overflow
Tank
3II 3I Compressor
3III
3-way
Valve
Heater Evaporator
1
3IV 4'
Expansion
Device
4'A 1A
In normal operation, the three-way valve connects the rectifier heat exchanger
outlet (3III) and expansion valve inlet (3IV) as indicated by the dotted arrows. In
this case, the overflow tank delivers liquid refrigerant from the condenser into the
rectifier. The liquid refrigerant in the rectifier is fed into the expansion valve. Through
this process, the mass fraction of the mixture in the rectifier and that circulating in
the cycle becomes the same. If a large system capacity is required, the direction of
the three-way valve is reversed and the electric heater is turned on. The refrigerant
from the overflow tank (3II) is now connected to the expansion valve inlet (3IV).
The low boiling point refrigerant is then vaporized and fed into the subcondenser
where it is condensed again. This condensed low boiling point refrigerant enters the
overflow tank and is mixed with the circulating refrigerant. Through this process, the
mass fraction of low boiling point refrigerant in the circulating mixture increases and
the system capacity increases. Jakobs and Kruse (1978) examined this cycle option
with a binary mixture, R12/R114 (60/40 wt.%). They confirmed the circulating mass
fraction of R12 changed from 50 to 90% and achieved 30% more heating capacity.
8
9 Vapor Condenser
3 2'
Rectifier
3A 2'A
Compressor
10
Reservoir Capillary Expansion
Tube Device
Liquid Evaporator 1
Accumulator
4' 5 7
Capillary
Tube 6
4'A 5A
exchange heat and mass. The high boiling point refrigerant in the rising vapor (R22)
is then condensed and leaves the rectifier with the liquid stream of the rectifier’s
lower end. The low boiling point refrigerant in the falling liquid (R13B1) is vaporized
and rises. In this way, the low boiling point refrigerant is stored in the reservoir and
the high boiling point refrigerant is circulating in the cycle. Therefore, the system
capacity is reduced. When a large capacity is required, the expansion valve opens
and the stored low boiling point refrigerant is released from the reservoir and
circulates in the cycle, resulting in a large capacity.
3A 2'A
3
2'
Condenser
Expansion Compressor
Device
1
Evaporator
4'
5
4'A 5A Rectifier
Accumulator
Heater
the rectifier that was not released in the evaporator drops through the rectifier to the
accumulator. These two streams encounter each other in the rectifier and exchange
heat and mass. The high boiling point refrigerant in the rising vapor (R134a) is then
condensed and leaves with the liquid. The low boiling point refrigerant in the falling
liquid (R32) is vaporized and rises with the vapor. In this way, the high boiling point
refrigerant (R134a) is stored in the accumulator and the low boiling point refrigerant
(R32) is circulating in the cycle. The advantage of the CRA compared to the cycle
with an accumulator only is a reduction in the size of the accumulator. Moreover,
CRA is able to extend the composition change by eliminating the thermal equilibrium
constraint.
Rothfleisch (1995) investigated the performance of the CRA with a binary
mixture, R32/R134a (30/70 wt.%), and a ternary mixture, R32/R125/R134a (23/25/52
wt.%). He confirmed that the maximum mass fraction shift of the circulating mixture
amounts to 54/46 wt.% for R32/R134a and 36/36/28 wt.% for R32/R125/R134a.
He also found that the heating seasonal performance factor was improved by about
3% with the CRA in use.
TABLE 7.2
Performance Comparison of Pure Refrigerants Investigated by ORNL
Cooling “A” Cooling “B” Heating “47S” Heating “17L”
Capacity Capacity Capacity Capacity
Refrigerant [kW] COP [kW] COP [kW] COP [kW] COP
TABLE 7.3
Temperature Glides of Binary Mixtures for 50/50 wt.%
R32 3 3 4 4 8 8 23 23 23 28 44
R125 3 3 4 6 6 19 19 19 23 40
R143a 3 4 5 5 17 17 17 23 38
R22 3 4 4 11 11 11 15 29
R218 3 3 7 7 7 10 23
R134a 3 3 3 3 5 11
R152a 3 3 3 4 8
R124 3 3 3 4
R124a 3 3 4
R142b 3 4
C318 3
R143
flammable flammable
H
H
R50 R170
R160 R161
R40 R41
toxic toxic R150 R151 R152a
R30 R31 R32
R140a R141b R142b R143a
TABLE 7.4
Characteristics of Potential Pure Component
Refrigerant Characteristics Normal Boiling Point [˚C]
TABLE 7.5
R22 Alternative Refrigerants Investigated by AREP
Type Refrigerant
TABLE 7.6
R502 Alternative Refrigerants Investigated by AREP
Type Refrigerant
TABLE 7.7
R22 Alternative Refrigerants Investigated by AREP
Cooling Cooling Heating Heating
Refrigerant Capacity Efficiency Capacity Efficiency
14
R22
13
R410A
12
Capacity [kW]
11
10
8
28 35 46 52 57
Ambient Temperaure [°C]
FIGURE 7.2.2 Ambient temperature vs. performance changes (Chin and Spatz, 1999).
4.5
R22 (Chin)
4.0 R410A (Chin)
R22 (Fluor)
3.5 R410A (Flour)
3.0
COP
2.5
2.0
1.5
1.0
25 30 35 40 45 50 55 60
Ambient Temperaure [°C]
FIGURE 7.2.3 Ambient temperature vs. performance changes (Chin and Spatz, 1999).
REFERENCES
Chen, J. and H. Kruse, 1997, “Mass Fraction Shift Simulation for the Mixed Refrigerants
R404A, R32/R134A, and R407C in an Air-conditioning System,” International Jour-
nal of HVAC&R Research, Vol. 3, No. 2, pp. 149–157.
Chin, L. and M. W. Spatz, 1999, “Issues Relating to the Adoption of R410A in Air Condi-
tioning Systems,” Proceedings of 20th International Congress of Refrigeration,
IIR/IIF, No. 179.
David, G. and M. Menzer, 1993, “Results of Compressor Calorimeter Tests in ARI’s R22
Alternative Refrigerants Evaluation Program,” Proceedings of ASHRAE/NIST Refrig-
erants Conference for R22/R502 Alternatives, pp. 1–18.
Fisher, S.K. and J.R. Sand, 1993, “Screening Analysis for Chlorine-Free Alternative Refrig-
erants to Replace R22 in Air-Conditioning Applications,” ASHRAE Transactions,
Vol. 99, Pt. 2, pp. 627–634.
236 Vapor Compression Heat Pumps with Refrigerant Mixtures
237
238 Vapor Compression Heat Pumps with Refrigerant Mixtures
Pure Liquid
Binary Mixture
Film
Boiling
Critical
Heat Flux
Minimum
Surface Heat Flux Heat Flux
Onset of
Nucleate
Boiling
Wall Superheat
FIGURE 8.1.1 Schematic diagram of pool boiling curve for binary mixture.
Heat Transfer Coefficient [kW/m2-K]
30
Plain tube
25 Turbo-B
High flux
Ideal HTC
20
15
10
0
0 0.25 0.5 0.75 1.0
R11 Mass Fraction
P = 1 bar q = 30 kW/m2
FIGURE 8.1.2 Pool boiling heat transfer coefficients of R11/R113 mixtures (Trewin et al.,
1994).
the bubble is no more than the temperature difference between the wall temperature
and the bulk liquid saturation temperature for the given overall mass fraction. Since
the liquid layer surrounding the bubble is enriched in the low vapor pressure com-
ponent, its saturation temperature is higher than that of the bulk liquid. This reduction
of the effective superheating is the main and most complex contribution to the heat
transfer coefficient degradation.
The reasons for the heat transfer coefficient degradation of zeotropic mixtures
is summarized as follows:
Heat Transfer of Refrigerant Mixtures 239
Bubble Liquid
Surrounding
Bubble
R
Diffusion
Area
Ttp Tsl
Radius
TABLE 8.1
Nucleate Pool Boiling Research
Authors Refrigerant Psat or Tsat Test Surface Oil [%]
Webb and Pais R123, R134a, R11, R12, 27˚C PT, FT, TB, SE None
(1991) R22, R123
Gorenflo and Sokol R134a 1.4–36.5 bar PT, FT, TX None
(1991) R227 1.0–26.4 bar
Shi and Hahne R134a 2.0–25.0 bar 0.1 mm wire None
(1991) R152a 2.2–23.0 bar
Webb and Pais R123, R134a, R11, R12, 4.4 and 26.7˚C PT, FT, TB, TX, None
(1992) R22 SE
Gupte and Webb R123, R11 4.4 and 26.7˚C SE None
(1992)
Memory et al. R114 2.2˚C PT, FT, TB, HF 0–10
(1993)
Webb and McQuade R123, R11 4.4˚C PT, TB, SE 0–5
(1993)
Muller et al. (1993) R134a 4.4 and 32˚C 0.1 mm wire 0–10
2.0 mm strip
Gupte and Webb R123 4.4 and 26.7˚C TB, SE None
(1994) R134a SE
R11 FT, TB, SE
Schomann et al. Propane 0.7–17.1 bar PT None
(1994) R114 2.3–29.3 bar
Buschmeier at al. Propane/n-butane 3.8–21.1 bar PT None
(1994)
Ohta and Fujita R11/R113 1.0–15.0 bar Flat plate None
(1994)
Trewin at al. (1994) R11/R113 1.03 bar PT, TB, HF None
Schmidt (1995) R22/R142b 0.05–0.20a Flat disk None
Propane/n-butane 0.03–0.90a
Ethane/n-butane 0.10–0.60a
Gorenflo et al. Propane/i-butane/n-butane 1.6–12.6 bar PT None
(1995)
Palm (1995) R134a 0 and 20˚C PT, FT, TB, TX 0–3
R22 0 and 25˚C and Sevac
Note: PT: Plain tube, FT: Low-finned tube, TB: Turbo B tube, TX: Gewa TX 19 tube, SE: Gewa SE
tube, HF: High flux,
a Reduced pressure.
Source: Thome, 1996.
quality increases and the liquid film thickness is diminished, evaporation of molecules
from the liquid surface becomes a more important contribution. When the liquid
film becomes thinner than the diameter of a bubble before it detaches, the evaporation
process dominates (suppression of nucleate boiling). Most flow boiling correlations
try to tie these two contributions to the heat transfer process by having a boiling
term and an evaporation term. The fact that the mass fraction of liquid and vapor
phases are quite different for zeotropic mixtures affects these heat transfer processes
quite significantly and must be accounted for in the correlations as well. Table 8.2
summarizes research on flow boiling heat transfer coefficients of mixtures based on
Wang and Chato’s (1995) summary, Thome’s (1996) summary and an additional
survey of the latest literature.
Ross et al. (1987) investigated convective boiling heat transfer coefficient of
R152a/R13B1. Their results can be summarized as follows:
Jung et al. (1989a,b) investigated the convective boiling heat transfer coefficient
of R22/R114 and R12/R152a. Their results are summarized here:
TABLE 8.2
In-tube Flow Boiling Research
Mass Flux
Authors Refrigerant [kg/m2-s] Tube/ID [mm]
2
0.0 0.2 0.4 0.6 0.8 1.0
R22 Mass Fraction
m=23 g/sq = 17 kW/m2
FIGURE 8.2.1 In-tube boiling heat transfer coefficients of R22/R114 mixtures (Jung et al.,
1989a,b).
Liquid layer
Heat Transfer Coefficient, HTCtp
q5
Va por core
Increasing heat flux
q4 Bubble
hnbc
q3
hcec
q2
hcec
q1
Quality
• However, convective boiling is dominant when the vapor quality is high. For
the convective boiling regime, the heat transfer coefficient is not affected
by the heat flux since the nucleate boiling is suppressed (Figure 8.2.2).
• The mixture shows a different circumferential variation in wall tempera-
ture and heat transfer coefficient from that of pure fluids. Due to gravity,
the liquid film thickness at the top is thinner than that at the bottom
(assuming a constant heat flux). Therefore, for pure fluids, the top part
has better heat transfer than the bottom part due to reduced heat transfer
resistance. This heat transfer variation results in a lower wall temperature
at the top than at the bottom. However, this film thickness variation causes
a mass fraction variation for zeotropic mixtures. Due to the thinner film
thickness at the top, the mass fraction of the more volatile component in
the top film portion is reduced and causes a heat transfer coefficient
reduction due to an increased mass transfer resistance. Therefore, the
bottom portion of the liquid film provides better heat transfer than the top
portion for zeotropic mixtures. This heat transfer variation results in a
lower wall temperature at the bottom than at the top.
• Similar to Jung et al.’s (1989) results, the heat transfer coefficient increases
as the heat flux increases in this partial boiling regime and the heat transfer
coefficient is not affected by the heat flux for the convective boiling regime.
• The heat transfer coefficient increases when the quality is higher than
20% as the refrigerant mass flux increases due to the increased mean
velocity of the refrigerant.
• Similar to Jung et al.’s (1989) results, the local heat transfer coefficient of
the zeotropic mixture at the top is lower than that at the bottom. However,
the local heat transfer coefficient of the azeotropic mixture at the top is
higher than that at the bottom as is the case for pure fluids.
Many other investigators confirmed the boiling heat transfer coefficient degra-
dation as compared to the ideal heat transfer coefficient as follows:
• Torikoshi and Ebisu (1993) studied the boiling heat transfer coefficient
of R32/R134a (30/70 wt.%) and reported the degradation of the boiling
heat transfer coefficient as compared to the ideal boiling heat transfer
coefficient was 30 and 20%, respectively, for the mass flux of 91 kg/m2-s
and 182 kg/m2-s when the saturation temperature was 0ºC.
• Rohlin (1994) reported the degradation of the boiling heat transfer coef-
ficient of R22/R142b (60/40 wt.%) as 40% when the saturation temper-
ature was 0˚C.
Some researchers investigated the boiling heat transfer coefficient of binary and
ternary mixtures that are considered as replacements of existing pure refrigerants,
R12 and R22:
Heat Transfer of Refrigerant Mixtures 245
summary of the most referenced correlations for flow boiling. Webb and Gupte
classified heat transfer models into three categories: superposition, enhancement and
asymptotic models. In this section, twelve nondimensional general correlations are
examined.
Chen, Bennett-Chen, Gungor-Winterton and Jung’s correlations utilize a super-
position model. In this model, the total heat transfer coefficient is the sum of nucleate
boiling and bulk convection contributions. Shah, Schrock-Grossman, Mishra, Sami
and Shin’s correlations utilize an enhancement model. In this model, the two-phase
heat transfer coefficient enhances the single-phase heat transfer coefficient of a
flowing liquid by a two-phase enhancement factor. Liu-Winterton, Wattelet and
Bivens-Yokozeki’s correlations utilize an asymptotic model. In this model, the total
heat transfer coefficient is of the form of Kutateladze’s (1961) power-type addition.
Chen (1966) proposed a general correlation for saturated boiling in a vertical tube.
His correlation is based on the superposition model, which divides the heat transfer
into two parts: a nucleate pool boiling contribution (hnb) based on Foster and Zuber’s
pool boiling equation (1955) and a bulk convective contribution (hbc) based on Dittus
and Boelter’s (1930) equation, as shown in Equations 8.1 through 8.3,
kl 0.8 0.4
hbc = 0.023 Retp Prl (8.2)
D
where
Retp = two-phase Reynolds number
G (1 − x ) D
Retp = F ( X tt )1.25
µl
where
F(Xtt) = two-phase correction factor
where
S = two-phase correction factor
( ) ( )
−1
S Re tp = 1 + 2.56 × 10 −6 Re1tp.17
Bennett and Chen (1980) modified the Chen correlation to account for the analysis
of nucleate boiling near the wall and for the effects of a liquid Prandtl number higher
than 1 on the bulk convection. Moreover, this correlation is based on experiments
with an aqueous mixture of ethylene glycol and water in a vertical channel. The
bulk convection term (hbc) was modified, as shown in Equation 8.4. The two-phase
correction factor was also modified, as shown in Equation 8.5,
where
hl = liquid phase heat transfer coefficient
kl 0.8 0.4
hl = 0.023 Rel Prl
D
− F ( X tt )hl X o
1− exp
kl
S= (8.5)
F ( X tt )hl X o
kl
0.5
σ
where X o = 0.04
g(ρl − ρv )
248 Vapor Compression Heat Pumps with Refrigerant Mixtures
As shown in Equation 8.5, this correlation includes the surface tension and
density difference between liquid and vapor phases in the nucleate boiling suppres-
sion factor. The heat transfer enhancement factor (F) includes thermal characteristics
of the fluid.
Ross et al. (1987) suggested the use of Bennett and Chen correlation with the
two-phase correction factor S set to zero. They proposed this idea based on their
experiment on the convective boiling heat transfer coefficient of R152a/R13B1. Since
the nucleate boiling is suppressed, the evaporation is the main heat transfer mech-
anism for the binary mixtures.
Gungor and Winterton (1986) developed the following correlation for flow boiling
in horizontal and vertical tubes. They used the basic form of the Chen correlation
in superposition form, as shown in Equation (8.6):
where
hl = heat transfer coefficient for the liquid phase flowing alone
hpool = pool boiling heat transfer coefficient
E= forced convection enhancement parameter
S= forced convection suppression parameter
hpool = 55P γ 0.12 (− log10 P γ )−0.55 M −0.5q 0.67
E = 1 + 24000 Bo1.16 + 1.37 X tt−0.86
S = (1 + 1.15 × 10 −6 E 2 Re1l .17 )−1
Jung et al. (1989a) developed the following flow boiling correlation for horizontal
tubes for the binary mixtures R22/R114 and R12/R152a. They used the basic form
of the Chen correlation in superposition form similar to Gungor and Winterton’s as
shown in Equation 8.7,
where
Cme = 1 – 0.35|Xv – Xl|1.56
F = forced convection enhancement parameter
Heat Transfer of Refrigerant Mixtures 249
F = 2.37(0.29 + 1/Xtt)0.85
hl = heat transfer coefficient for the liquid phase flowing alone (Equation 8.5)
S = forced convection suppression parameter
S = 4048Xtt 1.22BO1.13 for Xtt < 1
S = 2 – 0.1Xtt –0.28BO–0.33 for 1 < Xtt < 5
Cun = |1 + (b2 + b3)(1 + b4)|(1 + b5)
hpool = pool boiling heat transfer coefficient
hpool = hi /Cun
where
hi = ha hb /(hb Xl + ha Xv)
where
Pca = critical pressure of more volatile component
Jung et al. reported that this correlation matches their experimental data for
R22/R114 (23/77, 48/52, 77/23 wt.%) over 1700 data points and R12/R152a (21/79,
88/12 wt.%) within 9.6% deviation.
For saturated flow boiling in vertical and horizontal tubes, Shah (1976) proposed
the enhancement model, which enhances the single-phase heat transfer coefficient
of flowing liquid by a two-phase enhancement factor as shown in Equation 8.8. He
classified the flow-boiling regime into three regions by utilizing three dimensionless
parameters, shown in Equation 8.8. They are the pure nucleate boiling regime, the
bubble suppression regime and the convective boiling regime:
(
ψ s = h /hl = f Co, Bo, Frle ) (8.8)
250 Vapor Compression Heat Pumps with Refrigerant Mixtures
where
Co = [(1 – x)/x]0.8 [ρv /ρl]0.5
Bo = (q11)/(Ghlv)
Frle = (G2)/(ρl2gD)
Wattelet et al. (1994a) examined Shah’s correlation with their experimental data
for R22/R124/R152a and reported that Shah’s correlation overestimates the exper-
imental data by more than 50% in the wavy-stratified flow regime, but matches well
with the experimental data in the annual flow regime.
(
h /hl = C1 Bo + C2 X tt−0.66 ) (8.10)
where
Bo = boiling number
Cl = 7.39 × 10–3
C2 = 1.5 × 10–4
Mishra et al. (1981) developed the following correlation based on their data fit to
R12/R22. This correlation is an enhancement model like Shah’s and very similar to
Schrock and Grossman’s, as shown in Equation 8.11,
Heat Transfer of Refrigerant Mixtures 251
where
Cl = 5.64 for 23–27 wt.% of R12 and 21.75 for 41–48 wt.% of R12
m = 0.23 for 23–27 wt.% of R12 and 0.29 for 41–48 wt.% of R12
n = 0.15 for 23–27 wt.% of R12 and 0.23 for 41–48 wt.% of R12
Sami et al. (1992) developed their correlation Equation 8.12 for mixtures in an
enhanced tube — that is, an enhancement model in a form of the Dittus and Boelter
equation. However, Sami’s correlation uses a thermal conductivity term instead of
the Prandtl number.
where
kl = thermal conductivity of liquid
A = 1 – C(∑|XL – XV|)0.82
C = 0.79 for mixtures with HFCs
C = 0.68 for mixtures with HCFCs
Shin et al. (1996) developed the following correlation Equation 8.1 based on a fit to
R32/R134a and R290/R600a data of three compositions (75/25, 50/50, 25/75 wt.%).
This correlation is an enhancement model like Shah’s and used a correction factor
to account for the mass transfer effect due to the composition difference between
liquid and vapor phases, considered a driving force for mass transfer at the interface:
h /hl = (1 − CF ) F (8.13)
where
F = Chen’s two-phase correction factor
CF = A|Y – X|n
where
A = curved fitted coefficients (0.569 for R32/R134a; 0.533 for R290/R600a)
n = curved fitted coefficients (0.860 for R32/R134a; 0.828 for R290/R600a)
252 Vapor Compression Heat Pumps with Refrigerant Mixtures
Liu and Winterton (1991) proposed the following correlation for flow boiling inside
tube and annuli. They utilized an asymptotic model based on Kutateladze’s (1961)
power-type addition, as shown in Equation 8.14,
where
E = [1 + xPrl (ρl /ρv – 1)]0.35
S = (1 + 0.055E0.1 Rel0.16)–1
Wattelet et al. (1994b) proposed an asymptotic model based on their data for R12
and R134a. The Wattelet correlation accounted for the decrease in heat transfer due
to the wavy-stratified flow pattern in the low mass fluxes by adding a Froude
dependence to the convective term, as shown in Equation 8.15,
1/ 2.5
h = hnb 2.5 + hbc 2.5 (8.15)
where
where
M = molecular weight
q = heat flux [W/m2]
Pr = reduced pressure
hbc = Fhl R
Heat Transfer of Refrigerant Mixtures 253
where
F = 1 + 1.925Xtt–0.83
hl = 0.23Rex0.8 Pr0.4 λl /D
R = 1.32Frl0.2 if Frl < 0.25
R = 1 if Frl ≥ 0.25
where
Frl = (G/ρl)2/(9.80665D)
Bivens and Yokozeki (1994) suggested a modified Jung correlation and Wattelet
correlation for R32/R125 and R32/R125/R134a to account for the mass transfer
resistance, as shown in Equation 8.16,
{
h = htp / 1+ htpTint /q } (8.16)
where
q = heat flux [W/m2]
Tint = temperature at the liquid-vapor interface
q
Tint = 0.175(Td − Tb ) 1 − exp − −4
1.3 × 10 ρl h fg
where
Td = dew point temperature [˚C]
Tb = bubble point temperature [˚C]
ρl = liquid density [kg/m3]
hfg = heat of vaporization [J/kg]
htp = heat transfer coefficient
1/ 2.5
htp = hnb 2.5 + hbc 2.5
where
hnb = 55 M −0.5q 0.67 Pr0.12 [− log10 Pr ]−0.55
hbc = Fhl R
where
F = (0.29 + 1/Xtt)0.85
R = 2.838Frl0.2 if Frl ≤ 0.25
R = 2.15 if Frl > 0.25
254 Vapor Compression Heat Pumps with Refrigerant Mixtures
2 f f 0G 2 L 1 G2 x 2 ρl (1 − x )2
∫
x
∆PMN = φ2f 0 dx + + − 1 (8.17)
Dρl x 0 ρl α ρv (1 − α )
Heat Transfer of Refrigerant Mixtures 255
14 +5% 14 +5%
12 -5% 12 -5%
Predicted Capacity [kW]
8 8
6 6
4 4
2 JUNG 2 GUNGOR
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Measured Capacity [kW] Measured Capacity [kW]
14 +5%
12
−5 %
Predicted Capacity [kW]
10
4
SHAH
2
0
0 2 4 6 8 10 12 14
Measured Capacity [kW]
14 +5% +5%
14
12 −5 % 12 −5 %
Predicted Capacity [kW]
10 10
8 8
6 6
4 4
CHEN KANDLIKAR
2 2
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Measured Capacity [kW] Measured Capacity [kW]
FIGURE 8.3.1 Flow boiling heat transfer correlation results (Judge, 1996).
where
∆PMN = Martinelli and Nelson total two-phase pressure drop
ff0 = frictional coefficient considering two-phase flow as liquid flow
G = mass flux
x = vapor quality
10
Heat Transfer Coefficient [kW/m-K] Jung Chen
9
Gungor Kandlikar
8 Shah
7
6
5
4
3
2
1
0
0 0.2 0.4 0.6 0.8 1
Relative Heat Exchanger Length
FIGURE 8.3.2 Evaporation heat transfer coefficient vs. length (R22) (Judge, 1996).
10
Heat Transfer Coefficient [kW/m-K]
Jung Chen
9
Gungor Kandlikar
8 Shah
7
6
5
4
3
2
1
0 0.2 0.6 1
0 0.4 0.8
Relative Heat Exchanger Length
FIGURE 8.3.3 Evaporation heat transfer coefficient vs. length (R407C) (Judge, 1996).
L = length
D = tube diameter
ρl = liquid density
ρv = vapor density
φ2f 0 = two-phase frictional multiplier
α = void fraction
The Blasius friction factor relation for the turbulent flow in a smooth pipe as
shown in Equation 8.18 can be used to obtain the frictional coefficient (ff 0):
The frictional pressure drop when the total flow is liquid is given in Equation 8.19,
2 f f 0G 2 L 2
∆Pf ,MN = φf0 (8.19)
Dρl
To obtain the total pressure drop during the evaporation, the Martinelli and
Nelson correlation requires the evaluation of the local two-phase frictional multiplier
(φ2f0) and void fraction (α). Wallis (1969) gave the following relations (Equations
8.20 and 8.21) for these two parameters:
2
1
where φl = 1 + 0.5
x
Ross et al. (1987) compared their experimental data for R13B1/R152a (18/82,
38/62 and 80/20 wt.%) with Chisholm’s (1967) correlation, a modified version of
the Martinelli and Nelson method. They reported that the Chisholm correlation
matches with their data within a 30% variation and there is no need to modify the
Chisholm correlation.
Singal et al. reported that these correlations match their data within a 30%
deviation. However, these correlations cannot be used for R13 mass fractions larger
than 25%.
( x − x )D G2 L
∆Ptp = favg + 2 1 (8.24)
xavg L Dρavg
where
favg = 0.43 Kf 0.3 Re–0.28
Kf = Bo Pierre number
Kf = ∆x hfg /L
ρavg = ρv ρl [xavg ρl + (1 – xav) ρv]
xavg = (xi + xo)/2
2 f f 0G 2 L 1
∫
x2
∆Ptp = φtp2 dx (8.25)
Dρl ∆x x1
where
f fo = 0.079 Re −0.25
2 f f 0G 2 L 1
∫
x2
∆Ptp = φtp2 dx (8.26)
Dρl ∆x x1
where
f fo = 0.079 Re −0.25
where
0.5 0.125
ρ µv
Γ= l µ
ρv l
5
G=91 [kg/m2 -s]
G=182 [kg/m2 -s]
Pressure Drop [kPa/m] 4 R22, G=91
R22, G=182
0
0.0 0.2 0.4 0.6 0.8 1.0
R-32 Composition
Ts = 5°C
FIGURE 8.4.1 Pressure drop of R32/R134a during evaporation (Torikoshi and Ebisu, 1993).
TABLE 8.3
In-tube Condensation Research
Mass Flux
Authors Refrigerant [kg/m2-s] Tube/ID [mm]
mixtures based on Wang and Chato’s (1995) summary and an additional survey of
the latest literature. Wang and Chato pointed out that the main focus of the recent
research on condensation heat transfer with mixtures is on:
During the mixture condensation process, the higher boiling component transfers
into condensate at the vapor-liquid interface more so than the lower boiling com-
ponent because in the condensate, the mass fraction of the high boiling point
component is higher than in the vapor phase. More of the lower boiling point com-
ponent diffuses into the vapor mixture away from the interface. This mass fraction
gradient causes the deterioration of condensation heat transfer at the interface.
Therefore, the greater volatility difference between the mixture components results
in lowering the condensation heat transfer. This resistance to condensation heat
transfer is similar to that of evaporation heat transfer. The resistance is seen as being
caused by the originating phase.
Stoecker and Kornota (1985) investigated the condensation heat transfer coeffi-
cients of R12/R114. They divided the condenser into two parts, a high temperature
condenser and a low temperature condenser. Their results can be summarized as
follows:
• The local condensing heat transfer coefficients of the mixtures are smaller
than those of pure refrigerants.
• The condensing heat transfer coefficients of a zeotropic mixture are
enhanced in a microfin tube by about 30% as compared to a plain tube.
This enhancement is only half as much compared to that of the pure
refrigerant enhancement.
• The maximum degradation of the condensing heat transfer coefficient of
R22/R114 is about 20% at 34 to 43 wt.% of R22 mass fraction.
• The circumferential wall temperature distribution is somewhat different
from that of the pure refrigerants. For R22, the circumferential wall
temperature variation is small at the entrance but gradually increases in
the flow direction, indicating the flow pattern changes from the annular
flow to semiannular flow. The circumferential wall temperature of R22
decreases from the right, top, bottom and left since the thin liquid film at
the top moves to the right and the thick liquid film at the bottom moves
to the left due to the internal groove. For R22/R114 mixtures, the circum-
ferential wall temperature is uniform, which might be due to the opposite
changes of liquid-vapor interfacial temperature distribution.
Conklin and Vineyard (1991) studied the condensing heat transfer coefficient of
R143a/R124 (30/70, 51/49, 73/27 and 77/23 wt.%) in a finned tube and fluted tube.
They reported that the condensing heat transfer coefficients of theses binary mixtures
are higher in a finned tube than in a fluted tube. The condensing heat transfer
coefficients of R143a/R124 (30/70 wt.%) are about 15% higher than that of R22.
The pressure drop in the finned tube is only half of that in the fluted tube.
Sami et al. (1992) studied the condensing heat transfer coefficient of
R22/R152a/R114 (53/19/28 wt.%) and R22/R152a/R124 (50/20/30 wt.%) in a fluted
tube and reported the condensing heat transfer coefficient of these two ternary
mixtures is lower than R22 and higher than R12. No significant difference was found
between the condensing heat transfer coefficient of R22/R152a/R114 and
R22/R152a/R124. By using the same facility, Sami and Schnotale (1993) studied
the condensing heat transfer coefficient of R22/R114 and R22/R152a in a fluted
tube. As the R114 mass fraction increases, the condensing heat transfer coefficient
of R22/R114 increases almost linearly. However, the condensing heat transfer coef-
ficient of R22/R152a decreases as the R152a mass fraction increases. They also
found that the heat transfer coefficients of R22/R114 mixtures are lower than those
of R22 but the heat transfer coefficients of R22/R152a mixtures are higher than
those of R22.
Torikoshi and Ebisu (1993) studied the condensing heat transfer coefficient of
R32/R134a (30/70 wt.%). They reported the degradation of the condensing heat
transfer coefficient compared to the ideal condensing heat transfer coefficient as 36
and 29% for a mass flux of 91 kg/m2-s and 182 kg/m2-s, respectively, when the
saturation temperature was 50˚C.
TABLE 8.4
Ratio of In-tube Condensing Heat Transfer Coefficient to R22
R32/R125 R32/R134a R32/R134a
Condensing (60/40 (10/90 (25/75 R32/R125/R134a R125/R143a/R134a
Temperature wt.%) wt.%) wt.%) (30/10/60 wt.%) (44/53/4 wt.%)
Wijaya and Spatz (1995) reported that the condensing heat transfer coefficient
of R410A is slightly higher (about 3 to 6%) than that of R22 at low refrigerant
quality since the liquid thermal conductivity of R410A is as much as 8% lower than
that of R22 and its viscosity is 30 to 36% lower than that of R22.
Wang et al. (1996) compared the condensing heat transfer coefficient of R22
and R407C. They reported that the heat transfer coefficient of R407C is only 30 to
50% that of R22. The heat transfer coefficient of R407C is not very dependent upon
the changes of condensing pressure though the heat transfer coefficient of R22
increases as the condensing pressure increases.
Eckels and Tesene (1999) compared the condensing heat transfer coefficients of
R22, R134a, R410A and R407C in a smooth tube and microfin tubes. Their results
are:
• The condensing heat transfer coefficients for the microfin tube are signif-
icantly higher than those for the smooth tube for all refrigerants tested.
• R134a has the highest heat transfer coefficients in both types of tubes
while R407C has the lowest HTCs in both types of tubes.
• The heat transfer coefficient enhancement of the microfin tube over the
smooth tube varies from 2.2 to 2.5 at the lowest mass flux and from 1.2
to 1.6 at the highest mass flux. However, only R407C has a lower enhance-
ment at the lowest mass flux. This enhancement can be explained by the
turbulence enhancement by the internal fins that delay the flow pattern
transition from the annular flow to wavy flow.
kl 0.8 1/ 3
h = 0.0265 Reeq Prl ,Reeq > 5 × 10 4
D
(8.27)
k
h = 5.03 l Re eq
0.8
Prl1/ 3 ,Re eq < 5 × 10 4
D
where
Reeq = Reynolds number at equivalent mass flux given by Equation 8.28
0.5
µ ρ
Reeq = Rels + Re vs v l (8.28)
µ l ρv
where
Rels = GD (1 – x)/µl
Revs = GDx/µv
kl Re 0.9 Prl
h= F ( X tt ) l (8.29)
D F2
where
F2 = 5 Prl + 5 ln 1 + Prl (0.09636 Rel 0.585 − 1) for 50< Re l < 1125 (8.31)
F2 = 5 Prl + 5 ln(1 + 5 Prl ) + 2.5 ln(0.00313 Rel0.812 ) for Rel > 1125
where
hl = single-phase liquid heat transfer coefficient by Dittus and Boelter equation
h=
kl
D
( )
0.023 Rel0.8 Prl0.4 1 + 2.22 X tt−0.89 for annual-flow regime (8.33)
where
Nu f = 0.0195 Rel0.8 Prl0.4 φl ( X tt )
where
C1
φl ( X tt ) = 1.376 +
X tt c2
where
c1 = 4.12 + 0.48Frl – 1.564Frl2
c2 = 4.12 + 0.48Frl – 1.564Frl2 for 0 < Frl ≤ 0.7
c1 = 7.242
c2 = 1.655 for Frl > 0.7
where
Frl = liquid Froude number
θl = angle subtended from the top of tube to the liquid level
θl arccos(2α − 1)
1− ≅ by Jaster and Kosky (1976)
π π
kl −0.365
h = 2.82 Ph Re v 0.146 Prl1/ 3 (8.35)
D
where
h fg
Ph =
(
cPl Tsat − Tw )
8.6.6 KOYAMA CORRELATION
Koyama et al. (1990) proposed the following correlation (Equation 8.36) based on
their experimental data with R22/R114 mixtures in a horizontal microfin tube. Their
correlation is a modified version of Fujii and Nagata’s (1973).
0.4 0.8
kl D Re Pr
h = 0.53 Ph −0.6 l l (1 − 0.73X + 0.37 X 2 + 0.36 X 3 ) (8.36)
D L R
0.4 0.6
k D −0.6 Rel Prl
h = 6.4 l Ph R (8.37)
D L
where
0.5
ρµ
R= l l
ρvµ v
Later, Sami et al. (1995) reported that this correlation with a coefficient 0.212
instead of 6.4 matches with their experimental data for R23/R22/R134a (5/90/5 wt.%),
R23/R22/R152a (5/90/5 wt.%), R23/R125/R152a (8/62/30 wt.%) and R23/R22/R134a
(5/65/30 wt.%) within a 20% deviation.
0.3
G
h = 0.7 h Eq. (8.36) for annual-flo
ow regime (8.38)
300
0.3
G
h= h Eq. (8.37) for wavy-flow reg
gime (8.39)
300
+10% +10%
14 14
+5% +5%
−5% −5 %
12 12
−10% −10%
Predicted Capacity [kW]
8 8
6 6
4 4
DOBSON TRAV ISS
2 2
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Measured Capacity [kW] Measured Capacity [kW]
14 +10%
+5%
−5%
12
Predicted Capacity [kW]
−10%
10
2 SHAH
0 0 2 4 6 8 10 12 14
Measured Capacity [kW]
14 +10% 14 +10%
+5% +5%
12 −5% −5%
12
Predicted Capacity [kW]
−10% −10%
Predicted Capacity [kW]
10 10
8 8
6 6
4 4
2 CHEN TANDON
2
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Measured Capacity [kW] Measured Capacity [kW]
Moreover, the accelerational pressure drop is generally much smaller than the fric-
tional pressure drop; therefore, most investigators only considered the frictional
pressure drop. Two major parameters of the frictional pressure drop are the liquid
viscosity and mass flux.
6
Heat Transfer Coefficient [kW/m-K] Tandon
5 Traviss
Shah
4 Chen
Dobson
3
0
0 0.2 0.4 0.6 0.8 1
Relative Heat Exchanger Length
FIGURE 8.6.2 Condensation heat transfer coefficient vs. length (R22) (Judge, 1996).
8
Heat Transfer Coefficient [kW/m-K]
Tandon
7
Traviss
6 Shah
Chen
5
Dobson
4
3
2
1
0
0 0.2 0.4 0.6 0.8 1
Relative Heat Exchanger Length
FIGURE 8.6.3 Condensation heat transfer coefficient vs. length (R407C) (Judge, 1996).
Wavy
Gx (kg/m s)
2
Condensation
10 (Indoor Hx)
Bubbly
Slug
λ
Condensation
(Outdoor Hx)
Stratified
1.0
R-22 Plug
0.1
1.0 10 100 103 104
G(1−x)R
FIGURE 8.6.4 Baker’s flow map: Conditions for evaporation and condensation processes
are superimposed (Judge, 1996).
where
∆P = total two-phase pressure drop
∆Pf = frictional pressure drop
∆Pm = pressure drop due to the momentum change
G 2 x 2 ρl (1 − x )2 x 2 ρl (1 − x )2
∆Pm = + − + (8.41)
ρl α ρv (1 − α) α ρv (1 − α)
2 1
where
G = mass flux
ρl = liquid density
ρv = vapor density
x = vapor quality
α = void friction
−1
x ρl /ρv + 0.4(1 − x )
( )
0.5
ρv 1 − x
α = 1 + 0.4 + 0.6 (8.42)
ρl x
x + 0 .4 (1 − x )
( )
∆Pf /∆z = ∆P /∆z v Φ 2v (8.43)
272 Vapor Compression Heat Pumps with Refrigerant Mixtures
where
Φv = Lockhart and Martinelli parameter
∆P 2 fvG 2 x 2
∆z = Dρ (8.44)
v v
where
∆Pv = vapor flow pressure drop
fv = frictional coefficient considering two-phase flow as vapor flow
D = tube diameter
The Colburn friction factor relation for the turbulent flow in a smooth pipe (as
shown in Equation 8.45) can be used to obtain the frictional coefficient (fv),
The Sami correlation matches their experimental data within 30%. Later, Sami
et al. (1995) examined their earlier correlation against their data for other mixtures
of R23/R22/R152a (5/90/5 and 5/65/30 wt.%), R23/R22/R290 (5/90/5 wt.%),
R23/R22/R134a (5/65/30 wt.%) and R23/R125/R152a (8/62/30 wt.%) and reported
that the mean deviation is 15% for most of the data.
REFERENCES
Akers, W.W., H.A. Deans and O.K. Crosser, 1959, “Condensing Heat Transfer within Hori-
zontal Tubes,” Chemical Engineering Progress Symposium series, Vol. 55, No. 29,
pp. 171–176.
Baker, O., 1954, “Design of Pipe Lines for Simultaneous Flow of Oil and Gas,” Journal of
Oil and Gas, July, No. 26.
Bennett, D.L. and J.C. Chen, 1980, “Forced Convective Boiling in Vertical Tubes for Saturated
Pure Components and Binary Mixtures,” Journal of AIChE, Vol. 26, pp. 454–461.
Bivens, D.B. and A. Yokozeki, 1994, Proceedings of 1994 International Refrigeration Con-
ference at Purdue, West Lafayette, IN.
Bivens, D.B. and A. Yokozeki, 1994, A Heat Transfer of Zeotropic Refrigerant Mixtures Heat
Pumps for Energy Efficiency and Environmental Progress, Elsevier, Amsterdam.
Buschmeier, M., P. Sokol, A.D. Pinto, and D. Gorenflo, 1994, “Pool Boiling Heat Transfer
of Propane/n-Butane Mixtures at a Single Tube with Superimposed Convective Flow
of Bubbles or Liquid,” Proceedings of 10th International Heat Transfer Conerence,
Vol. 5, pp. 69–74.
Chen, J.C., 1966, “A Correlation for Boiling Heat Transfer to Saturated Fluids in Vertical
Flow,” Ind. Eng. Chem. Proc. Design Dev., Vol. 5, No. 3, pp. 322–339.
Chen, S.L., F.M. Gerner, and C.L. Tein, 1987, “General Film Condensation Correlations,”
Experimental Heat Transfer, Vol. 1, pp. 93–107.
Chisholm, D., 1967, “A Theoretical Basis for the Lockhart-Martinelli Correlation for Two-
Phase Flow,” International Journal of Heat and Mass Transfer, Vol. 10, pp.
1767–1778.
Chisholm, D., 1973, “Pressure Gradients Due to Friction During the Flow of Evaporating
Two-phase Mixtures in Smooth Tubes and Channels,” International Journal of Heat
and Mass Transfer, Vol. 16, pp. 347–358.
Collier, J.G. and J.R. Thome, 1994, Convective Boiling and Condensation, Clarendon Press
Oxford, England, pp. 535–584.
Conklin, J.C. and E.A. Vineyard, 1990, “Tube-side Evaporation of Non-Azeotropic Refrig-
erant Mixtures from Two Enhanced Surfaces,” ASME Winter Meeting.
Conklin, J.C. and E.A. Vineyard, 1991, “Tube-side Condensation of Non-Azeotropic Refrig-
erant Mixtures for Two Enhanced Surfaces,” ASME/JSME Thermal Engineering
Proceedings, Vol. 3, pp. 251–256.
Cooper, M.G., 1984, “Saturation Nucleate Pool Boiling, A Simple Correlation,” 1st U.K.
National Conference on Heat Transfer, Vol. 2, pp. 785–793.
Darabi, J., M. Salehi, M.H. Saeedi and M.M. Ohadi, 1995, “Review of Available Correlations
for Prediction of Flow Boiling Heat Transfer in Smooth and Augmented Tubes,”
ASHRAE Transactions, Vol. 101, Pt. 1, pp. 965–975.
Dittus, F.W. and L.M.K. Boelter, 1930, University of California Publications on Engineering,
Vol. 2, p. 443.
274 Vapor Compression Heat Pumps with Refrigerant Mixtures
Dobson, M.K., J.C. Chato, D.K. Hinde and S.P. Wang, 1994, “Experimental Evaluation of
Internal Condensation of Refrigerants R12 and R134A,” ASHRAE Transactions, Vol.
100, Pt. 1, pp. 744–754.
Dobson, M.K. and J.C. Chato, 1998, “Condensation in Smooth Horizontal Tubes,” Journal
of Heat Transfer, Vol. 120, pp. 193–213.
Doerr, T. and M.B. Pate, 1993, “In-Tube Evaporation and Condensation Alternatives for R22
and R502,” ASHRAE/NIST Refrigerants Conference, R22/R502 Alternative Session
II, Gaithersburg, MD.
Doerr, T.M., S.J. Eckels and M.B. Pate, 1994, “In-Tube Condensation Heat Transfer of
Refrigerants Mixtures,” ASHRAE Transactions, Vol. 100, Pt. 2, pp. 547–557.
Eckels, S.J. and M.B. Pate, 1991, “In-tube Evaporation and Condensation of a Ternary
Refrigerant Blend and HFC-134a,” ERI Project 3315, Iowa State University.
Eckels, S.J. and B.A. Tesene, 1999, “A Comparison of R22, R134A, R410A, and R407C
Condensation Performance in Smooth and Enhanced Tubes: Part I, Heat Transfer,”
ASHRAE Transactions, Vol. 105, Pt. 2.
Foster, H.K. and N. Zuber, 1955, Journal of AIChE, Vol. 1, pp. 531–535.
Fujii, T., T. Nagata, 1973, “Condensation Vapor in a Horizontal Tube,” Report of Research
Institute of Industrial Science, Kyushu University, Vol. 52, pp. 35–50.
Gorenflo, D., P. Kaupman, R. Koster, and M. Buschmeier, 1995, “Pool Boiling Heat Transfer
of Propane/i-Butane/n-Butane Mixtures,” Proceedings of the 19th International Con-
gress of Refrigeration, IVa, pp. 238–245.
Goto, M., N. Inoue, and K. Koyama, 1995, “Evaporation Heat Transfer of HCFC22 and Its
Alternative Refrigerants inside an Internally Grooved Horizontal Rube,” Proceedings
195th International Congress of Refrigeration, IVa, pp. 246–253.
Gungor, K.E. and R.H.S. Winterton, 1986, International Journal of Heat and Mass Transfer,
Vol. 29, No. 3, pp. 351–358.
Hihara, E., K. Tanida, and T. Saito, “1989, Forced Convective Boiling Experiments of Binary
Mixtures,” JSME International Journal, Series II, Vol. 32, No. 1, pp. 98–106.
Jaster, H. and P.G. Kosky, 1976, “Condensation in a Mixed Flow Regime,” International
Journal of Heat and Mass Transfer, Vol. 19, pp. 95–99.
Judge, J., 1996, “A Transient and Steady State Study of Pure and Mixture Refrigerants in a
Residential Heat Pump,” Ph.D. Dissertation, University of Maryland, College Park, MD.
Judge, J. and R. Radermacher, 1997, “A Heat Exchanger Model for Mixtures and Pure
Refrigerant Cycle Simulation,” International Journal of Refrigeration, Vol. 20, No.
4, pp. 244–255.
Jung, D.S. and R. Radermacher, 1989, Prediction of Pressure Drop during Horizontal Annular
Flow Boiling of Pure and Mixed Refrigerants, International Journal Heat Mass
Transfer, Vol. 32, pp. 2435–2446.
Jung, D.S., M. McLinden, R. Radermacher and D.A. Didion, 1989a, “A Study of Flow Boiling
Heat Transfer with Refrigerant Mixtures,” International Journal of Heat Transfer,
Vol. 32, No. 9, pp. 1751–1764.
Jung, D.S., M. McLinden, R. Radermacher and D.A. Didion, 1989b, “Horizontal Flow Boiling
Heat Transfer Experiments with a Mixture of R22/R114,” International Journal of
Heat Transfer, Vol. 32, No. 1, pp. 131–145.
Kattan, N., J.R. Thome, and D. Favrat, 1994, “In Tube Flow Boiling of R-502 and Two Near-
Azeotropic Alternatives,” Proceedings 10th International Heat Transfer Conference,
Brighton, Vol. 7, pp. 455–460.
Kenney, P.J., R.L. Shinon, T.C. Villaneuva, N.L. Rhines, K.A. Sweeney, D.G. Allen, and T.T.
Hershberger, 1994, “Heat Transfer Flow Regimes of Refrigerants in a Horizontal
Tube Evaporator, ACRC Technical Report 55, Univ. of Illinois.
Heat Transfer of Refrigerant Mixtures 275
Koyama, S., A. Miyara, H. Takamatsu and T. Fujii, 1990, “Condensation Heat Transfer of
Binary Refrigerant Mixtures of R22 and R114 Inside a Horizontal Tube with Internal
Spiral Grooves,” International Journal of Refrigeration, Vol. 13, pp. 256–263.
Kutateladze, S.S., 1961, “Boiling Heat Transfer,” International Journal of Heat and Mass
Transfer, Vol. 4, pp. 31–45.
Liu, Z. and R.H.S. Winterton, 1991, “A General Correlation for Saturated and Subcooled
Flow Boiling in Tubes and Annuli Based on a Nucleate Pool Boiling,” International
Journal of Heat and Mass Transfer, Vol. 34, pp. 2759–2765.
Martinelli, R.C. and Nelson, D.B., 1948, “Prediction of Pressure Drop During Forced Circu-
lation Boiling of Water,” Transactions of ASME, Vol. 70, pp. 695–702.
Melin, P. and L. Vamling, 1995, “Flow Boiling Heat Transfer and Pressure Drop for Zeotropic
Mixtures in a Horizontal Tube, Two-Phase Flow Modeling and Experimentation,”
Vol. 1, pp. 609–616.
Mulroy, W., M. Kaulfeld, M. McLinden, and D. Didion, “An Evaluation of Two Refrigerant
Mixtures in a Breadboard Air Conditioner Status of CRCs — Refrigerant Systems
and Refrigerant Properties,” IIR, Paris, pp. 27–34.
Murata, K. and K. Hashizume, 1990, “An Investigation on Forced Convection Boiling of
Nonazeotropic Refrigerant Mixtures, Heat Transfer-Japanese Research, Vol. 19, No.
2, pp. 95–109.
Murata, K. and K. Hashizume, 1993, “Forced Convective Goiling of Nonazeotropic Refrig-
erant Mixtures Inside Tubes,” Journal Heat Transfer, Vol. 115, pp. 680–689.
Niederkruger, M., D. Steiner, and E.U. Sehliinder, 1992, “Horizontal Flow Boiling Experi-
ments of Saturated Pure Components and Mixtures of R846-R12 at High Pressures,”
International Journal Refrigeration, Vol. 15, No. 1, pp. 48–58.
Ohata, H. and Y. Fujta, 1994, “Nucleate Pool Boiling of Binary Mixtures,” Proceedings of
the 10th International Heat Transfer Conference, Vol. 5, pp. 129–134.
Pierre, B., 1956, “Coefficient of Heat Transfer for Boiling Freon-12,” Horizontal Tubes,
Heating and Air Treatment Engineer, Vol. 19, pp. 302–310.
Radermacher, R., H. Ross, and D. Didion, 1983, “Experimental Determination of Forced
Convective Evaporative Heat Transfer Coefficients for Non-azeotropic Refrigerant
Mixtures,” ASME Paper, 83-WA/HT54.
Ro, S.T., H.S. Jun, Y.S. Chang and J.Y. Shin, 1994, “Condensation Heat Transfer of a Heat
Pump System using the Refrigerant Mixtures R32/R134a,” ASHRAE Transactions,
Vol. 100, Pt. 2, pp. 715–734.
Rohlin, P., 1994, “Heat Transfer Coefficients in Horizontal Flow Boiling of Some Pure
Refrigerants and Their Zeotropic Mixtures: Experimental and Theoretical Results,”
Proceedings of IIR Conference, CFCs, The Day After, pp. 583–590.
Ross, H., R. Radermacher, M. di Marzo, and D. Didion, 1987, “Horizontal Flow Boiling of
Pure and Mixed Refrigerants,” Int. J. Heat Mass Transfer, Vol. 30, pp. 979–992.
Sami, S.M. and T.N. Duong, 1991, “Experimental Study of the Heat Transfer Characteristics
of Refrigerant Mixture R-22/R-114 in the Annulus of Enhanced Surface Tubing,”
International Communication Heat Mass Transfer, Vol. 18, pp. 547–558.
Sami, S.M. and T.N. Duong, 1991, “Flow Boiling Characteristics of Refrigerant Mixture
R-22/R-114 in the Annulus of Enhanced Surface Tubing Heal Recovery Systems &
CHP,” Vol. 12, No. 1, pp. 39–48.
Sami, S.M. and J. Scbnotale, 1992, “Comparative Study of Two Phase Flow Boiling of
Refrigerant Mixtures and Pure Refrigerants Inside Enhanced Surface Tubing,” Inter-
national Communication Heat Mass Transfer, Vol. 19, pp. 137–148.
276 Vapor Compression Heat Pumps with Refrigerant Mixtures
Sami, S.M., J. Schnotale and J.G. Smale, 1992, “Prediction of the Heat Transfer Character-
istics of R22/R152a/R114 and R22/R152a/R124,” ASHRAE Transactions, Vol. 98,
Pt. 2, pp. 51–58.
Sami, S.M. and J. Schnotale, 1993, “Prediction of Forced Convective Condensation of Non-
Azeotropic Refrigerant Mixtures Inside Enhanced Surface Tubing,” Applied Scientific
Research, Vol. 50, pp. 149–168.
Sami, S.M., P.J. Tulej and B. Song, 1995, “Study of Heat and Mass Transfer Characteristics
of Ternary Nonazeotropic Refrigerant Mixtures Inside Air/Refrigerant-Enhanced Sur-
face Tubing,” ASHRAE Transactions, Vol. 101, Pt. 1, CH-95-23-4.
Schmidt, J.G. 1995, Blasenbildung beim siedon von stoffgemischen, Ph.D. Thesis, University
of Karlsruhe.
Schrock, V.E. and L.M. Grossman, 1959, “Forced Convective Boiling Studies,” University of
California, Institute of Engineering Research, Report No. 73308-UCX-2182, Berke-
ley, California.
Shah, M.M., 1976, “A New Correlation for Heat Transfer During Boiling Flow through Pipes,”
ASHRAE Transactions, Vol. 82, No. 2, pp. 185–196.
Shah, M.M., 1979, “A General Correlation for Heat Transfer During Film Condensation Inside
Pipes,” Journal of Heat and Mass Transfer, Vol. 22, pp. 547–556.
Shah, M.M., 1982, “Chart Correlation for Saturated Boiling Heat Transfer: Equations and
Further Studies,” ASHRAE Transactions, Vol. 88, Pt. 1, pp. 185–196.
Shin, J.Y., M.S. Kim and S.T. Ro, 1996, “Correlation of Evaporative Heat Transfer Coefficients
for Refrigerant Mixtures,” Proceedings of the 1996 International Refrigeration Con-
ference at Purdue, West Lafayette, IN, pp. 151–156.
Shin, J.Y., M.S. Kim and S.T. Ro, 1997, International Journal of Refrigeration, Vol. 20, No. 4,
pp. 267–275.
Singal, L.C., C.P. Sharma and H.K. Varma, 1983a, “Pressure Drop During Forced Convection
Boiling of Binary Refrigerant Mixtures,” International Journal of Multiphase Flow,
Vol. 9, No. 3, pp. 309–323.
Singal, L.C., C.P. Sharma, H.K. Varma, 1983b, “Experimental Heat ransfer Coefficient for
Binary Refrigerant Mixtures of R13 and R12,” ASHRAE Transactions, pp. 165–188.
Soliman, H.M., J.R. Schuster, and P.J. Berenson, 1968, “A General Heat Transfer Correlation
for Annular Flow Condensation,” ASME Journal Heat Transfer, Vol. 90, pp. 267–276.
Souza, A.L. and M.M. Pimenta, 1995, “Prediction of Pressure Drop During Horizontal Two-
phase Flow of Pure and Mixed Refrigerants,” ASME Conference on Cavitation and
Multiphase Flow, FED-Vol. 210, pp. 161–171.
Stephan, K. and P. Preusser, 1979, “Heat Transfer and Critical Heat Flux in Pool Boiling of
Binary and Ternary Mixtures,” German Chemical Engineering, Vol. 2, pp. 161–169.
Stephen, K. and M. Abdelsalam, 1981, “Heat Transfer Correlations for Natural Convection
Boiling,” International Journal Heat Mass Transfer, Vol. 23, pp. 73–87.
Sternling, C.V. and L.J. Tichacek, 1961, “Heat Transfer Coefficients for Boiling Mixtures,”
Chemical Engineering Science, Vol. 16, pp. 297–337.
Stoecker, W.F. and E. Kornota, 1985, “Condensing Coefficients When Using Refrigerant
Mixtures,” ASHRAE Transactions, Vol. 91, Pt. 2B, pp. 1351–1367.
Sundaresan, S.G., M.B. Pate, T.M. Doerr and D.T. Ray, 1996, Proceedings of 1996 Interna-
tional Refrigeration Conference at Purdue, West Lafayette, IN, pp. 187–192.
Sweeny, K.A., 1996, “The Heat Transfer and Pressure Drop Behavior of a Zeotropic Refrig-
erant Mixture in a Microfin Tube,” M.S. Thesis, University of Illinois at Urbana-
Champaign.
Heat Transfer of Refrigerant Mixtures 277
Sweeny, K.A. and J.C. Chato, 1996, “The Heat Transfer and Pressure Drop Behavior of a
Zeotropic Refrigerant Mixture in a Microfin Tube,” ACRC Technical Report 95,
University of Illinois at Urbana-Champaign.
Tandon, T.N., H.K. Varma and C.P. Gupta, 1986, “Generalized Correlation for Condensation
of Binary Mixtures Inside a Horizontal Tube,” International Journal of Refrigeration,
Vol. 9, pp. 134–136.
Thome, J.R., 1982, “Latent and Sensible Heat Transfer Rates in the Boiling of Binary
Mixture,” Journal of Heat Transfer, Vol. 104, pp. 474–478.
Thome, J.R., 1983, “Prediction of Binary Mixture Boiling Heat Transfer Coefficients Using
Only Equilibrium Data,” International Journal of Heat and Mass Transfer, Vol. 26,
No. 7, pp. 965–974.
Thome, J.R., 1996, “Boiling of New Refrigerants: A State-of-Art Review,” International
Journal of Refrigeration, Vol. 19, No. 7, pp. 435–457.
Torikoshi, K and T. Ebisu, 1993, “Heat Transfer and Pressure Drop Characteristics of R-134a,
R-32, and a Mixture of R-32/R134a Inside a Horizontal Tube,” ASHRAE Transactions,
Vol. 99, Pt. 2, pp. 90–96.
Torikoshi, K. and T. Ebisu, 1993b, “In-Tube Heat Transfer and Characteristics of Azeotropic
Mixtures of HFC-32 and HFC-125,” ASHRAE/NIST Refrigerants Conference,
R22/R502 Alternatives Session II, Gaithersburg, MD.
Traviss, D.P., W.M. Rohsenow and A.B. Baron, 1973, “Forced-Convection Condensation
Inside Tubes: A Heat Transfer Equation for Condenser Design,” ASHRAE Transac-
tions, Vol. 79, Pt. 1, pp. 157–165.
Trewin, R.r., M.K. Jensen, and A.E. Bergles, 1994, “Pool Boiling from Enhanced Surfaces
in Pure and Binary Mixtures of R113 and R11,” Proceedings of 10th International
Heat Transfer Conference, Vol. 5, pp. 165–170.
Van Wijk, W.R., A.S. Vos and S.J.D. Van Stralen, 1956, “Heat Transfer to Boiling Binary
Liquid Mixtures,” Chemical Engineering Science, Vol. 5, pp. 68–80.
Wallis, G.B., 1969, One-Dimensional Two-Phase Flow, McGraw-Hill, New York, NY, pp.
29–58.
Wang, C.C., C.S. Kuo, Y.J. Change and D.C. Lu, 1996, “Two-Phase Flow Heat Transfer and
Friction Characteristrics of R22 and R407C,” ASHRAE Transactions, Vol. 102, Pt. 1,
pp. 830–838.
Wang, S.P. and J.C. Chato, 1995a, “Review of Recent Research on Heat Transfer with Mixtures-
Part 1: Condensation,” ASHRAE Transactions, Vol. 101, Pt. 1, pp. 1376–1386.
Wang, S.P. and J.C. Chato, 1995b, “Review of Recent Research on Heat Transfer with
Mixtures-Part 2: Boiling and Evaporation,” ASHRAE Transactions, Vol. 101, Pt. 1,
pp. 1387–1401.
Wattelet, J.P., J.C. Chato, A.L. Souza, and B.R. Christofferson, 1993, Initial Evaporative
Comparison of R-22 with Alternative Refrigerant R134a and R-32/R-125, Air Con-
ditioning and Refrigeration Center Report TR-39, University of Illinois.
Uchida, M., M. Itoh, N. Shikazono and M. Kudoh, 1996, “Experimental Study on the Heat
Transfer Performance of a Zeotropic Refrigerant Mixture in Horizontal Tubes,”
Proceedings of the 1996 International Refrigeration Conference at Purdue, West
Lafayette, IN, pp. 133–138.
Wattelet, J.P., J.C. Chato, B.R. Christoffersen, J.A. Gaibel, M. Ponchner, P.J. Kenny, R.L.
Shimon, T.C. Villaneuva, N.L. Rhines, K.A. Sweeny, D.G. Allen and T.T. Hershberger,
1994a, “Heat Transfer Flow Regimes of Refrigerants in a Horizontal-Tube Evapora-
tor,” ACRC Report TR-55.
278 Vapor Compression Heat Pumps with Refrigerant Mixtures
Wattelet, J.P., J.C. Chato, A.L. Souza and B.R. Christoffersen, 1994b, “Evaporative Charac-
teristics of R12, R134a, and a Mixture at Low Mass Fluxes,” ASHRAE Transactions,
Vol. 100, Pt. 1, pp. 603–615.
Webb, R.L. and N.S. Gupte, 1992, “A Critical Review of Correlations for Convective Vapor-
ization in Tubes and Tube Banks,” Heat Transfer Engineering, Vol. 13, No. 3, pp.
58–81.
Whalley, P.B., 1994, Boiling, Condensation, and Gas-liquid Flow, Clarendon Press Oxford,
England, pp. 167–173.
Wijaya, H. and M.W. Spatz, 1995, “Two-Phase Flow Heat Transfer and Pressure Drop
Characteristics,” ASHRAE Transactions, Vol. 101, Pt. 1, pp. 1020–1030.
9 Operational Issues
This chapter discusses operational and handling issues that are unique to refrigerant
mixtures. It is based on laboratory experience with refrigerant mixtures and guide-
lines by the U.S. Environmental Protection Agency (EPA) and the Air-Conditioning
and Refrigeration Institute (ARI) for the handling of refrigerants.
9.1 REGULATION
9.1.1 CLEAN AIR ACT AMENDMENTS OF 1990
In 1990, the Clean Air Act was amended to control the production and use of
refrigerants in the United States. The goal of these amendments was to reduce the
total use and emissions of chlorofluorocarbons (CFCs) and hydrochlorofluorocar-
bons (HCFCs) and to maximize recycling. Two pertinent sections to the Clean Air
Act, Sections 608 and 609, restrict the emissions of class I (CFCs) and II (HCFCs)
substances from service, repair and disposal operations on stationary equipment and
mobile applications, respectively. It is unlawful in the United States for any person
to knowingly vent, release or dispose of any class I and II substances used as a
refrigerant in appliances and industrial process refrigeration while maintaining,
servicing, repairing or disposing, effective July 1, 1992. The EPA has the authority
to write necessary regulations to implement the Clean Air Act.
Violations of the Clean Air Act are punishable by fines up to $25,000 per
violation per day and the possible loss of technician certification. However, the EPA
allows some minimal emissions in the course of handling of refrigerants and oper-
ation of equipment in the following cases:
279
280 Vapor Compression Heat Pumps with Refrigerant Mixtures
• De minimis release if they occur, though all requirements for the proper
handling of refrigerants of allowed and approved recovery and recycling
machines are used.
• Refrigerant release from the equipment while in the normal operation,
such as purging and leaks.
• Nitrogen mixed with a small mass fraction of R22 for leak testing.
• Refrigerant release from hose purging or connections while charging or
servicing.
• Check refrigerant and oil compatibility: R22 and mineral oils are com-
patible, as are R407C and POE oils. However, R407C is not compatible
with mineral oils. Therefore, a flushing of the system is needed.
• Old refrigerant removal: Recover old refrigerant charged in the system
using a recovery machine.
• Initial oil drain from the compressor: Drain all old oil from the compressor
through the drain port of the compressor, if any is available. If no oil drain
port is available, separate the compressor from the system and drain the
oil from the compressor through the ports.
• Initial oil drain from the heat exchangers: Drain as much of the oil from
the cycle by blowing off with compressed nitrogen gas through the service
ports.
• Reassemble the compressor if it was disassembled.
• Evacuate the system using a vacuum pump.
• Fill the system with new lubricant (POE) to the minimum quantity
required and charge the system with R407C (use recovered R407C if
available).
• Run the system for approximately 10 minutes to help any residual old
refrigerant gas and oil to mix with new refrigerant and oil.
• Repeat above procedures except initial oil removal from the heat exchang-
ers until the refractive index difference between the oil sample and new
oil is within 1%.
• Prepare two containers charged with pure components, R32 and R125,
and one empty container for the mixture. A mixture container must have
two valves, liquid and gas.
• Evacuate the empty container using a vacuum pump to remove any
remaining gas and moisture in the container.
Operational Issues 283
• Calculate the total mass of the mixture to blend and its component mass.
For example, use 5 kg of R32 and 5 kg of R125 to make 10 kg of R32/R125
mixture with 50/50 wt.% mass fraction.
• Charge lower vapor pressure refrigerants first. If the higher vapor pressure
refrigerant is charged first, it may not be easy to charge the lower vapor
pressure fluid to the mixture container.
• To enhance the refrigerant transfer, place the container for the mixture in
a cold place (such as in a bucket filled with dry ice or ice water).
• Place a R125 container on the weighing scale and connect the R125
container and the mixture container with a refrigerant hose.
• While weighing the R125 container, transfer R125 to the mixture con-
tainer by adjusting container valves until the required amount of R125 is
transferred.
• Repeat the same procedure to transfer R32 from the R32 container to the
mixture container.
• Mix well by shaking.
• Take a liquid sample of the refrigerant mixture from the mixture container
and check the mass fraction of the mixture by using a gas chromatograph.
• Adjust the mass fraction. If the desired mass fraction is not achieved, then
calculate the amount to add from the measured mass fraction, add the
insufficient component and measure the mass fraction again. Repeat this
procedure until the desired mass fraction is achieved within 1%.
• Prepare refrigerant sample bottle. The volume of the sample bottle is sized
such that after the refrigerant sample expands into the bottle, only vapor
exists (approximately 1 l bottle is used as an example).
• Clean the sample bottles using acetone or another solvent to remove any
residual refrigerant and oil that might accumulate in the bottles: Evacuate
the bottles and pour approximately 100 ml of acetone in the bottle at
atmospheric pressure. Violently agitate the sample bottle for approxi-
mately 30 seconds. Charge nitrogen into the bottle to about 350 kPa
(nitrogen is used to rapidly push the acetone–oil solution out of the sample
bottle). Turn the bottle upside down and open the valve to allow the
acetone–oil solution to flow out.
• Evacuate the (preferably heated) sample bottle for several hours to remove
any remaining acetone.
• Attach the sample bottle to the sample tube already installed to a system
tube. The sample tube must be inserted into the middle of the refrigerant
stream in the liquid and vapor lines to ensure that bulk samples are
284 Vapor Compression Heat Pumps with Refrigerant Mixtures
Sample Bottle
Upper Valve
Needle Valve
Vacuum Pump
Sample Tube
Lower Valve
System Pipe
Refrigerant Flow
• Choose the size of vacuum pump that is appropriate for the system size.
Make sure the system pressure is not higher than 1 atm and that all
refrigerant that may have been in the system is reclaimed. Figure 9.6.1
shows a typical 1/2 hp deep vacuum pump.
• Make sure the space around the pump exhaust is well ventilated since the
vacuum pump will emit residual gases from the system and sometimes
discharge oil mist.
• Connect the middle port of the manifold gauge set to the vacuum pump.
• Open both system’s service ports (liquid and gas line ports).
• Open both low and high pressure side valves of the manifold gauge set.
• Turn on the vacuum pump. Keep evacuating until the vacuum pressure
reaches 500 µm of Hg, then run the vacuum pump at least 30 minutes
more to remove moisture left in the system or dissolved in the oils. The
longer the time for evacuation and the higher the temperature of all internal
surfaces of the system, the more complete is the removal of moisture.
• Shut down procedure for vacuum: Close valve to the system to be evac-
uated, bring the inlet port pump to atmospheric pressure (open valve or
disconnect hose), then pump can be turned off.
• Vacuum pump oil change: After every 20 to 30 operation hours, the oil
of the vacuum pump must be changed as it picks up moisture and con-
tamination. Keep a service sheet for every vacuum pump where operating
hours and oil changes are recorded.
not make any difference. However, containers for mixtures such as R407C and
R410A have two discharge ports, a liquid port and a gas port. When charging
refrigerant mixtures to the system, the liquid port should be used to maintain the
mass fraction. Other than this precaution, the same refrigerant charging procedure
can be used for both pure refrigerants and mixtures, explained as follows:
Pressure: Constant
Dewpoint Line
Superheated
Ta Vapor
Saturated
Tsv Vapor
Two Phase
Temperature
Range
Ttp
(Tsv-Tsl):
Temperature
Glide
Tsl
Saturated
Liquid Tb
Boiling Point
Line
Subcooled
Liquid
when the temperature is between its saturated liquid and vapor temperature (Tsl and
Tsv). This mass fraction difference brings new design issues:
3
2'
Condenser
Evaporator
Refrigerant and
4' 1 Oil Mixture
and bulk-vapor, the average liquid mass fraction and the average vapor mass fraction
are different from the circulating mass fraction due to the vapor-liquid volume
fractionation. Moreover, the circulating mass fraction shifts from the charged mass
fraction because the more volatile component accumulates predominantly in the
condenser vapor phase.
Corr and Murphy (1994) assumed that the subcooled liquid mass fraction is
similar to the average saturated liquid mass fraction of the condenser and the
superheated suction vapor mass fraction is similar to the average saturated vapor
mass fraction of the evaporator. With this simplified assumption, Corr and Murphy
showed that –1 to 3% of R32 mass fraction is shifted in the circulating refrigerant
when R32/R125 (60/40 wt.%) is charged.
To illustrate the effect of mass fraction shift on the cycle performance, the
changes of volumetric capacity and coefficient of performance (COP) are calculated
when R32 mass fractions of two binary mixtures of R32/R134a (30/70 wt.%) and
R32/R125 (50/50 wt.%) are changed with 10% variation from the charged concen-
tration, as shown in Figures 9.9.3 and 9.9.4. R32/R134a (30/70 wt.%) exhibits
approximately twice larger changes in volumetric capacity and COP due to the mass
fraction shift than that of R32/R125 (50/50 wt.%). If the circulating mass fraction
15
Volumetric Capacity Changes [%]
−5
−10
−15
−0.1−0.08 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 0.08 0.1
R32 Mass Fraction Changes
Te = 5°C, Tc = 45°C
FIGURE 9.9.3 Performance changes due to mass fraction shift (Corr and Murphy, 1994).
6
R32/R134a (30/70 wt.%)
4
R32/R125 (50/50 wt.%)
COP Changes [%]
−2
−4
−6
−0.1 −0.08 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 0.08 0.1
R32 Mass Fraction Changes
Te = 5 °C,Tc = 45°C
FIGURE 9.9.4 Performance changes due to mass fraction shift (Corr and Murphy, 1994).
Operational Issues 291
30
R32
25 R125
R134a
Solubility [wt. %]
20
15
10
0
−20 10 0 10 20 30 40 50
Temperature [°C]
FIGURE 9.9.5 Solubility of HFC refrigerants in ISO 32 polyol-ester (Corr and Murphy, 1994).
shift ranges 10%, R32/R134a (30/70 wt.%) undergoes volumetric capacity and COP
variation within 10% and 2%, respectively. This result is due to the larger temperature
glide of R32/R134a (30/70 wt.%), a zeotropic mixture, than that of R32/R125 (50/50
wt.%), almost an azeotropic mixture. Therefore, the mass fraction shift in the cycle
should be accounted for in the cycle design, especially for the zeotropic mixtures
with large temperature glides.
Figure 9.9.5 shows an example for the different HFC refrigerant solubility in
the ISO 32 polyolester oil (Corr and Murphy, 1994). As shown in Figure 9.9.5, the
solubility of R32 is much lower than that of R125 and R134a due to its low molecular
weight. Therefore, R32 resides in the heat exchangers or circulates through the cycle,
more so than the other components do if binary mixtures, R32/R125 or R32/R134a,
are charged. As the compressor sump temperature decreases, the difference between
the refrigerant solubility increases. Even more R32 then migrates to the heat
exchangers or circulates in the cycle. However, the impact of the different refrigerant
solubility in the oil is not so significant because for large systems, the oil charge is
usually smaller than that of the refrigerant charge and the compressor sump tem-
perature is usually high enough to maintain a lower refrigerant solubility. In small
systems such as domestic refrigerators, the oil charge can equal or exceed the
refrigerant charge and these effects are more important.
In 1995, Sumida et al. also investigated the circulating mass fraction shift of a
zeotropic mixture in the cycle. Sumida et al. distinguished between “circulating mass
fraction” on a mass flux basis from “existing mass fraction” on a mass basis, defined
as: The circulating mass fraction (α) is the ratio of the mass flux of the more volatile
component to the total mass flux of the refrigerant mixture in the cycle, as shown
in Equation 9.1. The existing mass fraction (β) is the ratio of the mass of the more
volatile component to the total mass of the refrigerant mixture in the cycle, as shown
in Equation 9.2.
wl (1 − fv )ρl ul + wv fvρvuv
α= (9.1)
(1 − fv )ρl ul + fvρvuv
292 Vapor Compression Heat Pumps with Refrigerant Mixtures
wl (1 − fv )ρl + wv fvρv
β= (9.2)
(1 − fv )ρl + fvρv
where
w = local mass fraction of the more volatile component
ρ = density
u = mean velocity
fg = void fraction
F = fv /(1 − fv ) (9.3)
D = ρv /ρl (9.4)
S = uv /ul (9.5)
wl + wv FDS
α= (9.6)
1 + FDS
wl + wv FD
β= (9.7)
1 + FD
FD(S − 1)( wg − wl )
δ = α −β = (9.8)
(1 + FDS )(1 + FD)
Therefore, the circulating mass fraction is different from the existing mass
fraction except in the case of either a zero slip between gas and liquid or the same
liquid and vapor mass fraction. Since two-phase flow of any fluid has a positive slip
ratio and since zeotropic mixtures have a higher gas mass fraction than liquid mass
fraction, the mass fraction difference δ always becomes positive for the two-phase
flow of a zeotropic mixture.
Consider the mass fraction difference between the charged mass fraction (γ) and
the circulating mass fraction (α). From the mass conservation of the more volatile
component in the total cycle, γ can be expressed in terms of β as shown in Equation 9.9,
Operational Issues 293
γWtotal = ∑ (β W )i i (9.9)
φ=α−γ = ∑ δ WW
i
i
i
total
(9.10)
To determine the mass fraction difference between the circulating mass fraction
and charged mass fraction φ, the mass fraction difference between the circulating
mass fraction and existing mass fraction δ must be determined for each component.
However, δ for single phase becomes zero because of zero slip; therefore, only the
contribution of the two-phase components remains. From the void fraction correla-
tion for fg, parameters F and S are determined from the Equations 9.3 and 9.11,
1 − fv
S= (9.11)
fv (1/x − 1)(ρv /ρl )
Sumida et al. used Smith’s correlation for void fraction and showed that the
circulating mass fraction of R32/134a (30/70 wt.%) shifts from 0.32 to 0.42, depend-
ing on the amount of refrigerant in the accumulator. They also showed that the
capacity is affected up to 10% because of this mass fraction shift; however, the COP
is maintained.
In another study of this issue, Chen and Kruse (1995) also investigated the
circulation mass fraction shift. Chen and Kruse recommended Highmark’s void
fraction model after comparing seven different void fraction models with their
experiment. Chen and Kruse examined the mass fraction shift of R23/R152a (20/80
and 25/75 wt.%) and found the mass fraction shifts of 0.06 and 0.05 for each mass
fraction. Chen and Kruse also showed the mass fraction shifts and their performance
effects for three additional refrigerant mixtures, as shown in Table 9.1.
TABLE 9.1
Mass Fraction Shift and Capacity Changes of Three Refrigerant Mixtures
Refrigerant Mixture Mass Fraction Shift [wt.%] Capacity Changes
Liq
eak
uid
rL
po
Le
Va
ak
xi,l (1-xq)+xi,v (xq-Rleak ) x i,l (1-xq){1-Rleak /(1-xq)} + xi,v xq
1 - R leak 1 - R leak
xi = xl ,i (1 − xq ) + xv,i xq (9.12)
xl ,i (1 − xq ) + xv ,i x q (1 − Rleak /x q )
xi ,new = (9.13)
1 − Rleak
Therefore, the change of the overall mass fraction of the i-th component in the tank
or system can be expressed as Equation 9.14,
Rleak
∆xi = xi ,new − xi = (1 − x q )( xl ,i − x v ,i ) (9.14)
1 − Rleak
Similarly, if there is a liquid leak, the change of the overall mass fraction of the i-th
component in the tank or system can be expressed as Equation 9.15,
Rleak
∆xi = xi ,new − xi = x q ( x v ,i − xl ,i ) (9.15)
1 − Rleak
The isothermal vapor leak is of more concern than the liquid leak because it
occurs for a longer period of time and sometimes is harder to detect; therefore, a
partial recharging may be considered after the vapor leak. If there is a liquid leak,
it occurs over a short period of time and the loss of total refrigerant is significant;
therefore, a complete recharging may be considered after a liquid leak. However,
the isothermal vapor leak also needs to be treated carefully when the components
of the mixture have a large vapor pressure difference because the mass fraction shift
becomes the issue in this case.
Kim and Didion (1995a) developed a leakage simulation model, REFLEAK,
using the approach described above. Figure 9.9.7 shows the simulation results of an
isothermal vapor leak cases for the selected refrigerants R410A, R32/R125 (30/70
wt.%) and R407C by using REFLEAK. The boiling points of R32, R125, and R134a
are –51.7, –48.1, and –26.1˚C, respectively. As shown in Figure 9.9.7, the mass
fraction of the more volatile component decreases and the mass fraction of the less
volatile component increases as a result of a vapor leak. If there is 50% vapor lost,
R410A loses 2% of R32 liquid and 1% of R32 vapor. For the same case, R32/R125
(30/70 wt.%) loses 2% of R32 liquid and vapor and R407C loses 7% of R32 liquid
and vapor. Therefore, the impact of the vapor leak will be more severe for R407C
and the next one is R32/R125 (30/70 wt.%). R410A has the least impact.
Bivens et al. (1996) also developed a similar model by integrating the differential
equations for mass fraction shifts and obtained a similar result as Kim and Didion.
296 Vapor Compression Heat Pumps with Refrigerant Mixtures
0.60
Tamb = 30°C, xq =0.5
R32 Liquid R32, Vapor
0.50
0.45
0.40
0 20 40 60 80 100
Leakage Rate [%]
(a) R410A
1.0
Tamb = 30°C,xq =0.5
R32 Liquid R32, Vapor
0.8
R134a Liquid R134a Vapor
Mass fraction
0.6
0.4
0.2
0.0
0 20 40 60 80 100
Leakage Rate [%]
(b) R32/125 (30/70 wt.%)
1.0
0.9 R32 Liquid R32, Vapor Tamb = 30°C, xq =0.5
0.8 R125 Liquid R125 Vapor
0.7
R134a Liquid R134a Vapor
Mass fraction
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0 20 40 60 80 100
Leakage Rate [%]
(c) R407C
FIGURE 9.9.7 Isothermal vapor leak process (Kim and Didion, 1995b).
Operational Issues 297
TABLE 9.2
Mass Fraction Shift during Liquid Charging from
Container (25˚C)
R410A R407C
Liquid Level [%] R32 R125 R32 R125 R134a
REFERENCES
Air-Conditioning and Refrigeration Institute, 1994, Applicant Study Guide for Refrigerant
Containment, Section 608, Clean Air Act, 2nd Edition.
Bivens, D.B., M.B. Shiflett, C.C. Allgood and A. Yokozeki, 1996, “Zeotropic Mixture Sepa-
rations Analyses,” Proceedings of the 1996 International Refrigeration Conference at
Purdue, West Lafayette, IN, pp. 113–118.
Bivens, D.B. and A. Yokozeki, 1998, “Mass Fraction Changes during Container Transfers for
Multi-component Refrigerants,” Proceedings of the 1998 International Refrigeration
Conference at Purdue, West Lafayette, IN, pp. 81–86.
Chen, J. and H. Kruse, 1995, “Calculating Circulation Concentration of Zeotropic Refrigerant
Mixtures,” International Journal of HVAC&R Research, Vol. 1, No. 3, pp. 219–231.
298 Vapor Compression Heat Pumps with Refrigerant Mixtures
Chen, J. and H. Kruse, 1997, “Concentration Shift Simulation for the Mixed Refrigerants
R404A, R32/134a, and R407C in an Air-Conditioning System,” International Journal
of HVAC&R Research, Vol. 3, No. 2, pp. 149–157.
Corr, S. and F.T. Murphy, 1994, “Mass Fraction Shifts of Zeotropic HFC Refrigerants in
Service,” Proceedings of IIR Conference for CFCs, The Day After, pp. 29–40.
Kim, M.S. and D.A. Didion, 1995a, “Simulation of Isothermal and Adiabatic Leak Processes
of Zeotropic Refrigerant Mixtures,” International Journal of HVAC&R Research, Vol.
1, No. 1, pp. 3–20.
Kim, M.S. and D.A. Didion, 1995b, “Simulation of a Leak/Recharge Process of Refrigerant
Mixtures,” International Journal of HVAC&R Research, Vol. 1, No. 3, pp. 242–254.
Sumida, Y., N. Tanaka and T. Okazaki, 1995, “Prediction of the Circulating Mass fraction of
a Zeotropic Mixture in a Refrigerant Cycle,” Proceedings of 19th International Con-
gress of Refrigeration, Vol. IVb, pp. 1013–1020.