An optimal endoreversible threeheatsource refrigerator

Z. Yan and J. Chen

Citation: Journal of Applied Physics 65, 1 (1989); doi: 10.1063/1.342570

View online: http://dx.doi.org/10.1063/1.342570
View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/65/1?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Optimized heat exchanger unit in a thermoacoustic refrigerator
AIP Conf. Proc. 1440, 1300 (2012); 10.1063/1.4704351

Equivalent combined systems of threeheatsource heat pumps

J. Chem. Phys. 90, 4951 (1989); 10.1063/1.456594

The effect of heattransfer law on performance of a twoheatsource endoreversible cycle

J. Chem. Phys. 90, 3740 (1989); 10.1063/1.455832

Optimal performance of an endoreversiblecombined refrigeration cycle

J. Appl. Phys. 63, 4795 (1988); 10.1063/1.340470

Optimal staging of endoreversible heat engines

J. Appl. Phys. 53, 1 (1982); 10.1063/1.331592

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
128.138.73.68 On: Sun, 21 Dec 2014 05:58:25
 An optimal endoreversible three-heat-source refrigerator
z. Van and J. Chen
Department of Physics, Xiamen University, Fujian, The People's Republic a/China
(Received 28 March 1988; accepted for publication 29 August 1988)
An endoreversible three-heat-source refrigerator only affected by thermal resistance, like a
reversible three-heat-source refrigerator, may be treated as a combined cycle of a two-heat-
source engine driving a two-heat-source refrigerator. The theory of finite time thermodynamics
in two-heat-source cycles is then used to analyze it and derive its basic optimum relation. Thus,
the fundamental effect of thermal resistance on the optimal performance of a three-heat-source
refrigerator is expounded. The conclusions obtained here are more realistic than those of
classical thermodynamics. They provide some new theoretical bases for further exploitation of
the three-heat-source refrigeration apparatus that applies to "cheap" heat sources, such as
solar energy, geothermal energy, waste heat, and so on.

dent of performance and the rate of refrigeration of an

I. INTRODUCTION
endoreversible three-heat-source refrigerator is derived. In
According to classical thennodynamics, the maximum Sec. IV, we discuss briefly the optimal performance of an
coefficient of performance of a three-heat-source refrigera~ endoreversible three-heat~source refrigerator, and obtain
tor is given by some important conclusions that have more realistic signifi-
€r = [( TIl - To) / T H ] [ TIl ( To - TJJ] , (1 ) cance than those of classical thermodynamics.
where T H , T L , and To are the temperature..'l of the high-
II. EQUIVALENT COMBINED SYSTEM OF AN
temperature reservoir, low~temperature reservoir, and envi-
ENDOREVERS!BLE THREE~HEAT~SOURCE
ronment, respectively. However, when the coefficient of per- REFRIGERATOR
formance of a three-heat-source refrigerator is demanded to
attain the bound €r of classical thermodynamics, the cyclic For a reversible three-heat~source refrigerator operat-
processes must be reversible. While the practical three~heat­ ing among reservoirs at temperatures T H' T L , and To, Eq.
source refrigerators are irreversible, none of their codfi- ( 1) shows that its coefficient of performance happens to be
cients of performance can attain €r' For example, in a solar the product of the efficiency of a reversible Carnot engine
energy refrigerator operating among reservoirs at tempera- operating between reservoirs at temperatures T H and To and
tures T H = 120 ·C, To = 40 °C, and TL = 15°C, Er comput- the coefficient of performance of a reversible Carnot refrig~
ed from Eq. (1) is equal to 2.35, while the practical coeffi- erator operating between reservoirs at temperatures To and
cients of performance are approximately 0.5-0.6. I It is thus T£. Hence, a reversible three-heat-source refrigerator may
dear that although the bound € r of classical thermodynam- be treated as a combined cycle of a reversible Carnot engine
ics is highly important in theory, it is usually too rough to driving a reversible Camot refrigerator, as shown in Fig. l.
predict the coefficient of performance of practical refrigera~ For an endoreversible three-heat-source refrigerator, al-
tors. Therefore, it is necessary to establish the bound of finite though it is different from a reversible three-heat~source re-
time thermodynamics in three-heat-source cycles. 2 •3 frigerator, its irreversibility only exists in the part of heat
In regard to finite time thermodynamics, many authors conduction between the working fluid and reservoirs, while
have carried out research in the last 10 years and made many the adiabatic processes are reversible. Consequently, its cy-
important conclusions. 4 - 13 Most of the conclusions concern cle may also be treated as a combined cycle that is similar to
the optimal theory of endoreversible two~heat~source cycles. the reversible cycle shown in Fig. 1, i.e., an engine driving a
They have important significance for research on the opti- refrigerator. But, owing to the irreversibility of heat conduc-
mal performance of endoreversible three-heat-source refri- tion existing in it, it must be treated as a combined cycle of an
gerators. endoreversible Carnot engine driving an endoreversible Car-
In this paper, the optimal theory of endoreversible two~ not refrigerator, as shown in Fig. 2. Where the temperatures
heat-source cycles is used to investigate the effect ofirrevers- of the working fluid in the Carnot engine at two isothermal
ibility of heat conduction on the performance of three-he at- processes are Tl and T z• the temperatures of the working
source refrigerators. Our plan is as follows. In Sec. II, it is
pointed out that an endoreversible three~heat~source refrig-
erator may be treated as a combined cycle of an endoreversi-
ble two-heat-source engine driving an endoreversible two-
heat-source refrigerator. The so-called endoreversible FIG. 1. Sketch of a reversible three-heat·
source refrigerator, treated as a reversible
three-he at-source refrigerator means that the only irreversi- w Camot engine driving a reversible Camot reo
bility in the refrigerator is the heat conduction between the frigerator.
working fluid and reservoirs, while a reversible three-heat-
source refrigeration cycle is still carried out inside the work-
ing fluid. 3 In Sec. III, the relation between the optimal coeffi~

J. Appl. Phys. 65 (1),1 January 1939 0021-8979/89/01 0001 ~04$02.40 © 1988 American Institute of Physics

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
128.138.73.68 On: Sun, 21 Dec 2014 05:58:25 ".:;<.•;..•..;•.•.•.....•.•...
"........ ·,············:·:·:·:;;:·:·:·:·:·:·;·x·:·:·:·:·;·:07·:O:O:·:.;.;.:0:.;.;.;.;.;.:.;.:;;;.:.;.;.;•.
0; •• ".;•••; •••• O;.............;~•••••••••••r.~•••••". • • • • • • • • • • • • • • • • • • • • • • • ••••• • .~ • • • • • • • • • • • ~_ ..
 where Kl = ayl(a l/2 + y1/2)2, while a and yare the heat
conductances between the working fluid and reservoirs at
temperatures TH and To, respectively.
Again, it is shown that under the circumstances of the
PIG. 2. Sketch of an cndoreversi-
given cycle time 72 and quantity of refrigeration QL' the
ble three-heat-source refrigerator, optimal coefficient of performance of an endoreversible Car-
treated as an endoreversible Car- not refrigerator operating between reservoirs at tempera-
not engine driving an endorcversi- tures To and TL is given byl2,l3
ble Carnal refrigerator.

¢; = T 1, - Ql.I(K27 2 ) ,
(5)
To - TL + QL I(Kz7 2 )
where K2 = py/(pllZ + y1/Z)2, while/3 is the heat conduc-
tance between the working fluid and reservoir at tempera-
ture Tr..
fluid in the Carnot refrigerator at two isothermal processes
Using Eqs. (3 )-( 5), and noting the relation Qfl = QL IE,
are T3 and T2 • That is, T I , T 2 , and T j represent the tempera-
tures of the working fluid in the three-heat-source refrigera-
we can find that under the circumstances of the given Qi. , 'I
and 72' the optimal coefficient of performance of an endore-
tor when the three isothermal processes are carried out, re- versible three-heat-source refrigerator is given by
spectively. They are respectively different from the
temperatures of the corresponding three reservoirs. More-
over, for convenience, we assume that the engine and refrig-
erator in the combined cycle operate alternately. Thus, the
(6)
combined cycle time 7 may be expressed as

(2) Now, let II and 72 vary, while QL and 7( = 71 + 7 2 )

remain unvaried. Under the above constraint conditions, we
where 71 and 72 are the cycle times of the engine and refrig- take E as objective function, and find its optimum. For this
erator in the combined cycle, respectively. reason, we introduce the Lagrangian
According to Fig. 2, we obtain the coefficient of perfor- L=E+A(I-7 1 - 7z )' (7)
mance of an endoreversible three-heat-source refrigerator
From the Euler-Lagrange equations
JL
(3) -=0 and (8)
a7 1

one obtains
where 1] and ¢; are the efficiency ofthe endoreversible Carnot
engine and the coefficient of performance of the endoreversi- (9)
bIe Carnot refrigerator in the combined cycle, respectively. a7, a'2
QH and QL are the heats absorbed from reservoirs at tem- Using Eq. (6), we can find
peratures T.H and TL by the working fluid per cycle, respec- K i/2 '2 TL - QLI(K212)
tively. Equation (3) shows that when both 1J and ¢; are opti- E = --- ------~--=-~ ( 10)
K i/27, Tlf - QLI(K1tl])
mum, E is also optimum under the corresponding conditions.
Thus, the optimal results of endoreversible two-heat-source from Eq. (9). Combining Eqs. (2) and (10), we find
engine and refrigerator may be used to establish the optimal
theory of endoreversible three-heat-source refrigerators.
The results below show that such an analysis is not only (11)
feasible but is also simpler.
and
1110 RELATION BETWEEN OPTIMAL COEFFICIENT OF K :12E7 TH - QL (iIK: 12 - llKi/2)
(12)
PERFORMANCE AND RATE OF REFRIGERATION K ;12ETJI + K ~I2TL
It is shown that under the circumstances of the given
cycle time 71 and quantity of supplying heat QH' the optimal Finally, substitution of Eqs. (1) and (12) into Eq. (6)
efficiency of an endoreversible Camot engine operating be- gives that under the given QL and 7 or the given rate of
tween reservoirs at temperatures TIl and To is given by4.8 refrigeration R ( = QL 17), the optimal coefficient of perfor-
mance E of an endoreversible three-heat-source refrigerator
(4) satisfies the equation

(13)

2 J. Appl. Phys., Vol. 65, No.1, 1 January 1989 z. Van and J. Chen 2

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
128.138.73.68 On: Sun, 21 Dec 2014 05:58:25
 We obtain from Eq. (13) R

K 1 T H (To - T L )(€, -E)

R= ,
(B + 1)2(1 + E)T H - B1TO + (I + €-')T[. FIG. 3. Sketch of the curve
(14)
of R varying with E for
where B=yIi2(a!/2_/3 1 / 2 }/LBl/2(a l12 +yI/2)], It is a =f3.
shown without difficulty that Eq, (14) determines the rela-
tion between the optimal coefficient of performance and the r E
given rate of refrigeration, as well as the relation between the
maximum rate of refrigeration and the given coefficient of
performance for endoreversib!e three-heat-source refrigera-
is larger than Em' Obviously, the larger Em is the only opti-
tors,14 Therefore, it is a fundamental optimum relation of
mal value of the coefficient of performance. It is thus obvious
endoreversible three-heat-source refrigerators. The various
optimum performance of endoreversible three-heat-source that although the coefficient of performance of an endore-
refrigerators may be derived from it. versible three-heat-source refrigerator is smaller than En the
refrigerator does not operate in the optimal working state if
IV. DISCUSSION it is smaller than Em' !n other words, the coefficient of per-
formance of an endoreversible three-heat-source refrigera-
(1) From Eg. (14), we can find that when the coeffi- tor should be situated between € m and E r • It is thus clear that
cient of performance is Em is an important performance parameter for the optimal
Ti? - T(~/2 design of three-heat-source refrigerators. The important sig-
em = nificance of €m lies in that it not only shows the optimal
T}?
coefficient of performance at the maximum rate of refrigera-
tion, but determines a lower limit to the value of the optimal
coefficient of performance as well for an endoreversible
(15) three-heat-source refrigerator. In addition, it is worthwhile
the rate of refrigeration R attains the maximum to note that Em is more approximate to realistic than <=7' For
example, when a = /3, Em of the solar energy refrigerator
Rmax =K1(T}f - Tf/2)2(TL + F)/(T H - TL + G), mentioned in the introduction is equal to 0.494 computed
(16)
from Eq. ( 15). It is considerably nearer to the realistic value.
where This shows dearly the -important significance of the bound-
F=BTL(Tll - To)/(Tll - TI.)' ary of finite time thermodynamics.
(3) When Tll --> 00, Eq. (14) is consistent with Eq. (5).
G = B ( T};2 - T A12) 2 [ B 2( TH - To) / (I'll - T1J
The physical meaning of this result is very clear, because the
12
+ Bn?(3TW + 2Tb - Tl/TW)/(Tl! - TL ) heat source at temperature T H ..." 00 corresponds to a work
source. Therefore, in such a case, endoreversible three-heat-
+ (3 T};2 + T b!2 )I ( T ~:2 - T i(2)] . source refrigerators become endoreversible Carnot refriger-
Rmax and € m are the two important performance param- ators. In a general way, work is "high-grade" energy, while
eters of a three-heat-source refrigerator. Like the maximum heat is much "cheaper" from the entropy-creation point of
power output Pmax and corresponding efficiency 11m (i.e., view. Therefore, in general, the optimal performance of an
the famous CA efficiency) first obtained by Curzon and endoreversible three-heat-source refrigerator is not as good
Ahlborn concerning a Carnot engine, IS they are conducive as those of an endoreversible Camot refrigerator. However,
to the further understanding of cyclic performance. How- this is because heat is much "cheaper" than work; therefore,
ever, it is worthwhile to notice that Eq. (15) shows that the idea of producing refrigeration by using heat is of large
when the heat conductances between the working fluid and significance.
three heat sources are different from each other, Em is not ( 4) It is seen from Eq. ( 1) iha t when T I . is very smaU, e r
only the function of reservoir temperatures, but is also de- or ¢c is quite small. In such a case, it is necessary to pay a
pendent on the heat conductances. It is different from the high price for lowering the temperature. This points up one
famous CA efficiency, which is only a function of reservoir essential serious difficulty in lowering the temperature to
temperatures. 4 - 9 •15 ultralow temperatures, It is also seen from Eqs. (5) and ( 14 )
(2) The curve of R(€) determined from Eq. (14) is that when thermal resistance exists, the rate ofrefrigeration
sketchily shown in Fig. 3. Figure 3 shows clearly that we can of Camot refrigerators or three-heat-source refrigerators
have € = €, if and only if R = O. That is to say that when may not attain K21~. Therefore, at ultralow temperatures,
thermal resistance exists, the coefficient of performance may the rate of refrigeration of endoreversible refrigerators is
not attain Er unless the rate of refrigeration is zero. This quite smalL This brings about another serious difficulty for
shows that the bound Er of classical thermodynamics does lowering the temperature to ultralow temperatures. That is
not have very large instructive significance for practical to say, the effect of thermal resistance brings about twofold
three-heat-source refrigerators. difficulties for lowering the temperature to uItralow tem-
It is also seen from Fig. 3 that when O<R < R max , one R peratures. This is an important theoretical problem in the
corresponds to two E; where one is smaller than em' the other development of cryogenic technology.

3 J. Appl. Phys., Vol. 65, No.1. i January 1SSg Z. Yan and J. Chen 3

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
128.138.73.68 On: Sun, 21 Dec 2014 05:58:25
 (5) When a = (:J = y, Eqs. (15) and ( 16) may be writ~ on the performance of three-heat-source refrigerators may
ten as be revealed from the above analysis, and conclusions more
realistic than those of classical thermodynamics are ob-
€m = (17) tained. This provides some new theoretical bases for the
further exploitation of the three-heat-source refrigeration
and apparatus that applies to "cheap" heat source, such as solar
energy, geothermal energy, waste heat, and so on.
Rmax = (a/4)(T}?- T6/2)2I~/(Tl! - Td, (18)
respectively. Equations (17) and (18) are the main results
obtained in Refs. 2 and 3. It is ofinterest to note that the form
of Eq. (17) is similar to that of Eq. (l), and € m is again a
function only of the reservoir temperatures. 'R. R. Washington, Chem. Eng. News 58,36 (1980).
2Z. Yan and S. Chen, Kexue Tongbao 31, 798 (1986) (Dull. Sci., in Chin-
It is seen clearly from the above results that it is a useful ese).
method to use a combined cycle to analyze the optimal per- 3Z. Yan and S. Chen, Cryo. No.4, 18 (1986) (in Chinese).
formance of endoreversible three-heat-source refrigerators. 4M. H. Rubin, Phys. Rev. A 19,1272 (1979); 19,1277 (1979).
It may simplify the problem in discussion, as well as reveal 'P. Salamon and A. Nitzan, J. Chem. Phys. 74, 3546 ( 1981 ).
OM. H. Rubin and B. Andresen, J. App!. Phys. 53,1 (1982).
the inherent relation between endoreversible three- and two- 7y. N. Orlov, Sov. Phys. Dok!. (USA) 30, 506 (1985).
heat~source refrigeration cycles. This is very beneficial to the "Z. Yan, J. Eng. Thermophys. 6, 1 (1985) (ill Chinese).
establishment of finite-time thermodynamics theory in 9A. D. Yos, Am. 1. Phys. 53, 570 (1985).
IIlC. H. Blanchard, J. App!. Phys. 51, 2471 (1980).
three~heat~source cycles. 16 It has been pointed out in Ref. 17
II Z. Yan, New Energy Sources 9, 42 (1987) (in Chinese).
that the absorption refrigerators can be discussed by the tri- I2Z. Yan, Wuli 13, 768 ( 1984) (ill Chinese).
cycle method, but no definite results have been derived. The "1. Chen and Z. Yan, J. App\. Phys. 63, 4795 (1988).
fundamental distinction between endoreversible and reversi- 14J. Chen and Z. Yan (unpublished).
"P. L. Curwn and B. Ahlborn, Am. J. Phys. 43,22 (1975).
ble three-heat~source refrigerators lies in the fact that the "'Z. Yan and J. Chen (unpublished).
irreversibility of heat conduction is considered in the former. '7So Andresen, P. Salamon, and R. S. Berry, 1. Chern. Phys. 66, 157!
Consequently, the fundamental effect of thermal resistance (1977).

4 J. Appl. Phys., Vol. 65, No.1, 1 January 1989 Z. Yan and J. Chen 4

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
128.138.73.68 On: Sun, 21 Dec 2014 05:58:25