Analysis of a hybrid compression—absorption cycle using lithium bromide and water as the working fluid

Analysis of a hybrid compression—absorption cycle using lithium bromide and water as the working fluid

Analysis of a hybrid compression-absorption cycle using lithium bromide and water as the working fluid Keith E. Herold, Lawrence A. Howe* and Reinhard...

991KB Sizes 0 Downloads 60 Views

Analysis of a hybrid compression-absorption cycle using lithium bromide and water as the working fluid Keith E. Herold, Lawrence A. Howe* and Reinhard Radermacher Department of Mechanical Engineering, University of Maryland, College Park, M D 20742, USA *Energy Concepts Co., 627 Ridgely Ave., Annapolis, M D 21401, USA R e c e i v e d 5 J a n u a r y 1990," revised 3 O c t o b e r 1990

Hybrid refrigeration cycles which combine a mechanical compressor and an absorption cycle in such a way that they share a single evaporator were analysed. The motivation for the investigation of hybrid cycles was the need to more efficiently utilize the output of an internal combustion engine. The hybrid cycles make efficient use of both the work and the heat output of an engine. Performance calculations are reported for a promising cycle which utilizes LiB~H20 as the working fluid. For this working fluid, the refrigerant is water. Owing to the potential sensitivity of the absorption cycle components to oil contamination, the cycle was analysed assuming an oil-free steam compressor (screw design). Although oil-free steam compressors are available, they are used only sparingly in the industry. The capital cost for such a compressor is very high and the isentropic efficiency of the available units is low. This combination of high cost and low performance results in poor economics for the hybrid cycle based on the available technology. However, the cycle has significant potential from a thermodynamic viewpoint and it provides an incentive for compressor manufacturers to refine the oil-free steam compression technology.

(Keywords: hybrid compression-absorptioncycles; steam compressors;water-lithium bromide workingfluid)

Analyse d'un cycle hybride fi compression-absorption au bromure de lithium et eau comme fluide actif On a analysk des cycles .[rigorifiques hybrides combinant un cycle h compresseur mdcanique et un cycle [1 absorption de telleJbgon qu 'ils se partagent un mOme kvaporateur. Cette recherche sur les cycles hybrides a ~t[" motivde par le besoin d'utiliser de jbqon plus efficace un moteur (i combustion interne. Les cycles hybrides utilisent de Jafon eJficace le travail et la production de chaleur d'un moteur. On rapporte les calculs de la perJ%rmance d'un cycle promeneur au bromure de lithium-eau comme fluide act(f, I'eau ktant le JHgorigg'ne. Compte tenu de la sensibilitd ~ventuelle des composants du cycle h absorption h la contamination par l 'huile. on a supposd, pour l'analyse, que le compresseur de vapeur est sans huile (d vis). Bien que les compresseurs de vapeur sans huile soient disponibles, ils ne sont utilisks que rarement dans l'industrie. L'investissement pour un tel compresseur est trOs Olevk et le rendement isentropique des machines disponibles est faible. A cause de cette combinaison co~t klevk//bible efficacit~, la rentabilitk d'un cycle hybride, Jbnd~ sur une telle technologic, est .[bible. Cependant, le cycle off?e des perspectives int~ressantes d'un point de rue thermodynamique, et son utilisation devrait inciter les fabricants de compresseurs h approJondir la technologie de la compression de vapeur sans huile.

(Mots cl6s: cycles hybrides compression-absorption; compresseur de vapeur; fluide actif bromure de litbium-eau)

The motivation for this investigation of hybrid compression-absorption cycles was the need to utilize both the heat and work available simultaneously from one source. An example of such a source is an internal combustion engine, which produces shaft work and which typically rejects energy to the environment as heat. If such an engine is used to drive the compressor in a vapour compression cycle, the system efficiency is relatively low because the potential of the heat rejected from the engine is not realized. Improved system efficiencies are possible by utilizing the rejected heat to drive a refrigeration cycle that complements the output of the vapour compression cycle. An absorption cycle, driven by heat, is a natural choice for this complementary cycle. One option is to drive a conventional vapour compression chiller with the shaft of the engine and drive a conventional absorption chiller 0140 7007/91/0502644)9 :.~') 1991 Butterworth Heinemann Ltd and llR

264

Int. d. Refrig. 1991 Vo114 September

with the heat from the engine cooling jacket and the exhaust gas. In this option, the two cycles are completely separate. An alternative option, which forms the subject of this paper, is to integrate the two cycles so that they circulate a common refrigerant and utilize a single evaporator. Such a hybrid cycle has several advantages over the separate cycle option. As the hybrid system shares heat exchangers between the two sides of the cycle, it utilizes fewer heat exchangers, which has potential economic benefits compared to the separate cycle option. The choice of working fluids for hybrid cycles is constrained by the same factors that constrain the choice for the individual cycles. Absorption chiller technology has developed around two primary absorption pairs: lithium bromide---water and water-ammonia. Although much research has gone into identifying alternatives to these two pairs, no other pair has gained widespread use.

Hybrid compression-absorption cycle: K. E. Herold et al. C3

Water-ammonia hybrid cycles have been discussed previouslyL This paper deals with lithium bromide water hybrid cycles. For these cycles, the refrigerant is water. Water possesses several very desirable properties as a refrigerant, including high temperature stability and a very high latent heat of vaporization. However, it also imposes certain limitations on the cycle application as a result of a relatively high freezing temperature, and tends to cause corrosion in the presence of oxygen, which can lead to maintenance problems.

23 .~--W 22

Related work

The idea of using an engine to drive a refrigeration cycle is well established, as shown, for example, by the eight papers which were presented at an A S H R A E symposium on that subject held in 1987. Two of those papers are of interest here as they deal with an overview of the current technology 2.3. A US company that currently offers engine-driven heat pump packages has carried out research on a compression-absorption cycle where the two cycles are separate 4. The concept of a hybrid compression-absorption cycle where the cycles share a single refrigerant has been considered by a number of previous workers. A variety of possible configurations for the cycle have been explored by Alefeld and co-workers 5,6. A related cycle, which is a vapour compression cycle with a solution loop in place of the condenser, expansion valve and evaporator, has been studied by Ahlby v, Radermacher e t al. 8 and Stokar 9. Another variation by Sunye e t a l ? ° incorporates a compressor with an absorption heat transformer. Description of the cycle

In this study, numerous cycle variations were considered. The cycle which was found to have the best combination of features consists of a double-effect LiBr-H20 absorption cycle combined with a vapour compression cycle in such a way that the two cycles share a common evaporator. A schematic diagram of this hybrid cycle is shown in F i g u r e 1. The cycle diagrams presented in this paper are superimposed on pressure-temperature axes so that the relative position of the heat exchangers indicates the internal operating conditions. The absorption cycle consists of eight heat exchangers. Component E is the evaporator and component A is the absorber. Components D1 and D2 are low and high temperature desorbers, respectively. Components CI and C2 are condensers. The cycle also utilizes two solution heat exchangers (recuperators) labelled R 1 and R2. The vapour compression cycle introduces the compressor and the condenser C3. The cycle is fired by heat at two temperature levels (Tin and T i n ) , in desorbers D1 and D2, and powered simultaneously by a work input to the compressor. The compressor suction is taken from condenser C2, The vapour flow consists of the vapour from desorber D2 plus a portion of the vapour which flashes in the expansion valve between condensers C3 and C2. The remainder of the vapour is condensed in C2. The vapour from D2 is desuperheated in condenser C2 before it flows to the compressor. The high side pressure, point 23, is chosen such that the heat of the condensing refrigerant, released in C3, is supplied to the high temperature desorber, D2,

'I

i

i

I

TO

Tt

- I ,' T " ' ~ TDI

TD 2

Figure 1 Hybridvapour compression-absorptionchiller. Design consists of LiBr-H20 double-effectabsorption chiller with steam compressor. Compressorsuction is taken from the condenserand condensate injection to the compressor is provided. (D1,D2) desorbers; (A) absorber; (CI,C2,C3) condensers; (E) Evaporator; (RI,R2) recuperative heat exchangers;(O) pump: ( X ) expansion valve; (.v~.)heat; (-~) work; (@) compressor Figure 1 Groupe reJ?oidisseur hybride ~ compression de vapeur et absorption . ll s'agit d'un groupe ~ absorption bromure de lithium-eau double effet, avec un compresseur de vapeur. L'aspiration au compresseur se Jait h partir du condenseur et l'injeetion du condensat au compresseur est assurbe. (D1, D2) dbsorbeurs," (A) absorbeur; (C1,C2,C3) condenseurs; ( E) kvaporateur; ( R1, R2) ~changes de chaleur de r~cupkration; ( 0 ) pompe; ( ~ ) d&endeur," (~-)chaleur; (--~) travail," (@) compresseur

by internal heat exchange. The exhaust heat of the engine is supplied to the high temperature desorber as well. The engine cooling water fires the low temperature desorber, D1, together with heat provided by condenser C2. This arrangement provides a good temperature match between the heat sources and heat requirements, which leads to efficient energy conversion. In the hybrid cycle, both the vapour compression side and the absorption side share a common evaporator. The utilization of the heat which would normally be rejected adds to the coefficient of performance and capacity of the engine-driven chiller. The portion of the evaporator load attributable to the vapour compression side of the hybrid can be determined by comparing the compressor mass flow-rate (at point 22) with the evaporator mass flowrate (point 1). For typical operating conditions this value is approximately 80%. Thus the hybrid cycle is dominated by the vapour compression side. This result is consistent with the relatively low availability associated with the waste heat rejected from the engine. Referring to F i g u r e 1, components C2 and D I can be seen to operate at approximately the same temperature (TD0. As the desorber, D1, requires heat and the condenser, C2, needs to reject heat, it is possible to achieve both with internal heat exchange. This internal heat exchange is a standard feature of a double-effect absorption cycle. Heat from the engine is supplied to each of the Rev. Int. Froid 1991 Vol 14 Septembre

265

Hybrid compression-absorption cycle: K. E. Herold et al. desorbers, Dl and D2, to power the cycle. Heat is rejected by the cycle from the condenser, C1, and the absorber, A. In a commercial installation, this heat rejection would typically be achieved with a cooling tower. The cooling effect, which is the 'product' of the cycle, is represented by the heat transfer into the evaporator, E. The cooling effect is provided at the lowest temperature in the cycle (To). In addition to the energy transfers mentioned above, a small electrical energy input, of the order of 0. 1% of the compressor work input, is needed to drive the two solution pumps. The line on Figure 1 which includes point 25 returns a portion of the condensate to the compressor. The condensate is injected into the compression chamber primarily to limit the temperature rise of the vapour being compressed. A portion of the injected liquid evaporates and cools the two-phase mixture leading to lower compressor temperatures. High temperatures in the compressor must be avoided due to close tolerances between the moving parts. An added benefit of condensate injection is improved sealing of the compression chamber that results when a liquid is present. Oil-flooded screw compressors achieve this liquid seal with oil. A drawback of condensate injection is that a portion of the working fluid is recirculated and not sent to the evaporator. However, the performance advantages far outweigh this drawback.

Computer model The performance of the hybrid refrigeration cycles was calculated using a classical cycle modelling approach. Similar approaches have been used previously ~ ~3 for modelling LiBr H20 absorption systems. The vapour compression model is based on an isentropic efficiency for the compressor. The thermodynamic properties of the hybrid cycle working fluid were obtained from two sources. The properties of solutions of LiBr H20 were obtained from the correlation published by McNeely 14. The properties of pure water, in both liquid and vapour states, were obtained from the values published by Keenan et al. ~. Both sources utilize identical reference states for enthalpy and therefore no reference state corrections are needed. Three major cycle variations were considered initially, including the preferred cycle shown in Figure 1. The two others are shown in Figures 2 and 3. The cycle in Figure 2 is very similar to that in Figure 1, except that the absorption cycle is a single-stage configuration. As a result, the pressure of the steam is lower at the compressor suction and the specific volume is larger. The cycle in Figure 3 is based on a double-effect absorption cycle but the compressor suction is taken from the intermediate pressure level. Once again, the specific volume of the steam is very large. Both of these cycles are impractical as a result of the physical size of the compressor needed to accommodate the large specific volumes. By taking the compressor suction from the high pressure level as in the preferred cycle in Figure 1, the compressor size is smaller by a factor of approximately eight times compared to the cycles in Figures 2 and 3. The engine was modelled by curve-fitting the manufacturer's performance data for a representative engine. The engine characteristics used for this purpose are given in Figure 4. The resulting routines calculate efficiency and

266

Int. J. Refrig. 1991 Vo114 September

C2

25 22

C1

[ D

Figure 2 Hybrid vapour compression absorption chiller based on a single-effect LiB~H20 absorption cycle. Compressor suction is taken from the condenser and condensate injection to the compressor is provided. This cycle operates at relatively low suction side pressure and consequently requires a larger compressor than the cycle of Figure 1. Key as in Figure 1 Figure 2 Groupe re/?oidisseur hybride fi compression de vapeur-absorption, fondg" sur un cycle ~t absorption bromure de fithium-eau • simple effet. L'a.spiration au compresseur se .[ait gt partir du condenseur et I'injeetion du condensat au compresseur est assur~e. Ce cyle Jbnetionne h une pression d'aspiration relativernent Jaible et, par consequent, ni'eessite un compresseur plus grand que le cycle de la Figure 1. La ldgende est la m6me que celle de la Figure 1

E

Figure 3 Hybrid vapour compression absorption chiller based on a double-effect LiBr H20 absorption cycle. Compressor suction is taken from the intermediate pressure condenser and this cycle requires a larger compressor than the cycle of Figure 1. Key as in Figure 1 Figure 3 Groupe refroidisseur h)'bride ~ compression de vapeur-absorption, fond~ sur un ~3,ele d absorption bromure de lithium-eau 2t double effet. L 'aspiration au eompresseur se Jait h partir du condenseur h pression intermkdiaire et ce o'ele n~cessite un eornpresseur plus grand que celui de la Figure 1. La I{'gende est la m~me que pour la Figure 1

power output for a particular torque and speed (rev rain-i). The assumptions involved in the model are described by referring to the cycle in Figure 1. This figure depicts the version of the cycle which showed the best overall

Hybrid compression-absorption cycle: K. E. Herold et al. Table I Hybrid cycle model inputs Tableau 1 Donnbes d'entrde dans le eyele hybride modble Input

Model value

Solution heat exchangers: effectiveness or approach temperature 0.85 Internal heat exchange: approach temperature 5°C Evaporator temperature 2°C Condenser temperature (set equal to absorber outlet temperature) 35.6°C Solution loop composition difference 3% Compressor location 3 (Figure 1) Number of compressor stages 1 Compressor inlet quality 0.78 Compressor isentropic efficiency 0.70 Engine energy distribution (percentages of each useful component) Work 28 Jacket heat 26 Exhaust heat 27

potential. Other cycles were modelled in a similar fashion. The general assumptions were as follows: minimum approach temperature for internal heat exchange is an input value; no pressure drop in the heat exchangers or pipes; isentropic efficiency of p u m p s = 0 . 7 ; isentropic efficiency of compressor=0.5, unless otherwise noted (according to manufacturer's specifications); and all fluids and phases are in thermodynamic equilibrium at all locations. The assumptions relating to Figure 1 were: saturated liquid at points 2, 9, 12, 16, 19 and 24; saturated water vapour at 1 and 22 - - vapour at 22 has been cooled, from 17, by passing it through condenser C2; saturated LiBr H20 vapour at 7 and 17 (can also be viewed as superheated H20 vapour); and composition difference between solution streams in each solution circuit is an input value.

Hybrid cycle model input requirements The model can be run in either a batch mode or interactively at a terminal. When the model is run interactively, the user is prompted for data to describe the cycle and operating conditions. The primary model inputs are summarized in Table 1. The user must define the internal heat-exchange processes in both of the solution heat exchangers and in the two components where internal heat exchange occurs between a condenser and a desorber. The evaporator temperature and condenser temperature must also be specified. The composition in each of the two solution loops is assumed to be identical. This is characteristic of a parallel double-effect absorption chiller where each desorber is fed from a common absorber and the desorbers are matched so that the concentration change across each is the same. This characteristic is assumed and the concentration change must be input by the user. The term 'compressor location' refers to the position (in a thermodynamic sense) of the compressor with respect to the absorption cycle. The default position, 3, is the preferred cycle illustrated in Figure 1. Compressor locations l and 2 are shown schematically in Figures 2 and 3, respectively. The inlet quality to the compressor must be input to specify the amount of condensate being re-injected for temperature control. Without condensate injection, the exit temperature from the compressor is

fur rated torque

30

=

75 %

50 % 20

i 1000

2000

3000

4000

Engine Speed ( rpm )

Figure 4 Engine efficiency (%) versus rotational speed for different loads. Taken from manufacturer's data 20 Figure 4 Efficacitb du moteur ( % ) en jbnction de la vitesse de rotation pour d(ff~;rentes charges. D 'apr~s des donn~es obtenues aupr~s de Jabric a n t s 2°

fairly high (frequently exceeding 260°C) and most compressor designs cannot withstand these conditions. The fraction of the primary energy input (to the gas engine) which is delivered to the cycle is specified by entering the percentages of mechanical work, jacket heat and exhaust heat. When these percentages are summed, the difference between the sum and 100 represents the losses of primary energy from the system. These losses are mainly flue losses with a minor component of heat transfer to the environment. The work and heat fractions were obtained from the engine model of Figure 4 in generating the results given in this paper.

Cycle model outputs The inputs indicated in Table 1 result in the output given as Figure 5. This output example is provided to illustrate the use of the model. A discussion of performance predictions is provided later in this paper. The input values to the model are echoed in the output. The primary inputs are shown in the lines labelled 'inputs'. Following the input echo is a table of state point data referring to the points defined on Figure 1. The table shows the pressure, temperature, enthalpy, composition and flowrate at each numbered point in the cycle. Note that numbers 20 and 21 are skipped in the list and on the figure. The heat transfer duty of each of the major components is indicated next. Positive values indicate heat transfer into the cycle. The magnitude of the numbers is based on 100 kW of primary energy (fuel) being delivered to the engine. Internal heat exchange can be determined by comparing the duty of the paired heat exchangers. For example, the energy that must be rejected from condenser C2 is 14 kW, but the energy requirement of desorber DI is 40 kW. Thus the external heat requirement to component D1 is the difference, 26 kW. Simi-

Rev. Int. Froid 1991 Vo114 Septembre

267

Hybrid compression-absorption cycle: K. E. Herold et al. Hybrid

Compression/Absorption

I n p u t s : T e v a p (C) : 2 Tcond Comp. elf.: .7 % w o r k : 28 Point 1 2 3 4 5 6 7

P(kPa) 0.71 0.71 5.81 5.81 5.81 5.81 5.81 0.71 5.81 5.81 57.03 5.81 57.03 57.03 57.03 57.03 57.03 5.81 57.03 57.03 381.23 381.23 57.03

8

9 i0 ll 12 13 14 15 16 17 18 19 22 23 24 25 All hea~ Component

and work

Cycle

(C) : 35.6 % jacket:

T(C) 2.00 35.60 35.60 69.68 42.21 79.66 73.17 40.58 35.60 73.17 129.47 73.17 73.17 122.41 82.73 136.90 129.47 79.66 84.66 84.66 141.90 141.90 84.66 quantities

heats:

E 155

- LiBr/H20

w~ight 0.0 57.4 57.4 57.4 60.4 60.4 0.0 60.4 0.0 57.4 57.4 57.4 57.4 57.4 60.4 60.4 0.0 60.4 0.0 0.0 0.0 0.0 0.0

%

2.363

COPc:

1.553

0.065908

1.072895 1.072895 1.072895 1.019563 1.019563 0.053332 0.053332 0.056919 0.072973 0.072973 0.016054

in k W

A -194

Heat

(C) : 5

flow(kg/s) 0.065908 1.325894 1.325894 1.325894 1.259986 1.259986 0.012576

C1 -42

D1 40

C2 -14

D2 168

Compressor work: total: 28 p e r kg: 383.71 p e r k g yap: D i s p l a c e m e n t (1/kJ) : 1 . 0 4 9 4 2 Suc. yap. v o l . : 2.86380 L i q u i d inj.: YES Q u a l i t y : inlet: 0 . 7 8 0 0 0 outlet: 0.90379 (kg l i q ) / ( k g v a p o r ) : inlet: 0 . 2 8 2 0 5 outlet: 0.10646 COPh:

3

location:

D e l t a X (%): 3 Delta T internal 26 % exahaust: 27 % loss: 19

H (kJ/kg) 2505.97 93 98 93 98 162 29 121 86 193 74 2637 95 I18.72 148.99 169.29 282.42 169.29 169.29 268.22 199.63 303.73 2739.96 193.74 354.31 2651.31 2529.68 597.01 354.31 are

- Compressor

hal:

-0.00

Pump

l:

C3 -141

491.93 T avg: i 1 3 . 2 8

0.000

Pump2:

0.001

Solution circuits: Input: effectiveness Circulation ratio: 20.1174 ist s t a g e sol. c i r c u i t : Elf: .85 D e l t a T: 6.612911 C H e a t s (kW): h e a t i n g : 9.285305 Cooling: 3.954145 HX: 90.56567 2nd s t a g e sol. c i r c u i t : Elf: .85 D e l t a T: 9.567604 C H e a t s (kW) : h e a t i n g : 15.23702 Cooling: 6.009831 HX: 106.1383

Figure 5 Hybrid cycle computer model output for the inputs of Table 1. Reproduced directly from program output. This example is presented to illustrate the use of the model. Detailed results of the performance predictions are presented in the text Figure 5 R~sultats du modOle informatique du cycle hybride pour les donnbes d'entrbe de la Tableau 1. Reproduits directement ~ partir des rbsultats du programme. Cet exemple illustre l'utilisation du modble. Les rbsultats dbtaillbs des pr~visions de la per[ormance sont prbsentbs clans le texte

larly, condenser C3 provides most of the energy required by desorber D2, leaving an external requirement of 27 kW. Note that these external heat requirements match the heat available in the engine cooling water and exhaust. The computer model matches the heat duty of the paired heat exchangers by adjusting the working fluid flow-rates. The next group of lines on the output is a summary of information about the compressor. The compressor power requirement is given first (in kW), then normalized (kJ kg -t) on the total mass flow-rate at the compressor outlet and, finally, normalized (kJ kg ') on the mass flow-rate at the compressor inlet. The compressor displacement is given in litres of suction gas per kilojoule of evaporator capacity (1 kJ-J), along with the specific volume of the refrigerant at the compressor suction conditions (m 3 kg 9, and the average compressor temperature, defined as the average of the temperatures at the inlet and outlet, is given in °C. Two lines describe the injection of the condensate. When condensate injection is used (i.e. the inlet quality is specified as less than 1.0),

268

Int. J. Refrig. 1991 Vo114 September

then the outlet quality may also be less than 1.0. The ratio of liquid to vapour at the inlet and outlet of the compressor is also given. This ratio, r~, is related to the quality, X, by rx = (1 - X ) / X . The coefficient of performance (COP) in heating and cooling is given next. For the LiBr-H20 cycles considered in this paper, normal heating applications are not practical because the refrigerant (water) will freeze at evaporator conditions. Thus the heating COP is calculated for reference purposes only. The energy balance on the cycle is calculated and should always be zero. This feature was used by the authors during program development. The pump work for each of the solution pumps is also calculated. The final group of lines of the output are a summary of the performance of the solution heat exchangers. The circulation ratio is the ratio of the liquid flow-rate through the pumped side of the solution loop to the vapour flow-rate for the upper stage. For each solution heat exchanger the effectiveness and minimum temperature difference are reported. In this example the effective-

Hybrid compression-absorption cycle: K. E. Herold et al. I Cooling tower ]

Exhaust --

high stage condenser (C2). The compressor discharge is fed back to a portion of the high stage desorber (D2) tubes, the inside of which forms condenser C3. The piping connections between the compressor and the absorption machine need to be as short as possible to avoid unnecessary heat loss from the surfaces.

,,~

absorption chiller

gas

÷1

Steam compressor

~ Pump ..¢ ,g P Flow meter T Temperature

Natural g a s

---.--(E3E3 ~'~

Gas engine

sensor

clutch

Screw

~

Valve

compressor

Figure 6 Schematic diagram of hybrid vapour compression absorption cycle equipment. Major components are the steam compressor, the double-effect chiller and the internal combustion engine Figure 6 Schbma de l'bquipement h o'cle hybride ~ compression de vapeur-absorption. Les principaux composants sont le compresseur de vapeur, le groupe refroidisseur ~ absorption ~ double effet et le moteur combustion interne

ness was input and the temperature difference is calculated. The entry labelled 'Heating' is the sensible heating (kW) needed to bring the solution stream entering the desorber up to the saturation temperature defined by the pressure and the concentration. On entry to the absorber cooling is required to achieve the saturation state (either 8 or 18) and is given as 'Cooling'. The heat transfer rate (kW) for each solution heat exchanger is given as 'HX'. Review of available hardware

The key hardware components required for the hybrid cycle are the steam compressor, the absorption chiller and the internal combustion engine. The project goal was to assemble the hybrid chiller from off-the-shelf components. The major problem encountered was to find an oil-free steam compressor with adequate performance. To focus the investigation of hardware, a design capacity of 100 tons (352 kW) was selected. A layout drawing of the equipment needed for the hybrid cycle is given as Figure 6. The interconnections of all the major components are shown. Both the heat and work output from the engine are utilized by the cycle to cool the load. The heat transfers from the exhaust and cooling water are used by the absorption side of the cycle. The exhaust gases deliver heat to the absorption chiller in the normal manner where the hot combustion gases are passed through tubes which are immersed in solution in the high temperature desorber. A difference here is that, in addition to the exhaust heat, the high temperature desorber is heated by internal heat exchange with condenser C3. The heat from the engine cooling water is delivered to the low temperature desorber by way of a closed liquid loop. The cooling water heat is delivered to the low temperature desorber because the available temperature range is not high enough to fire the high temperature stage of the chiller. Thus the low temperature stage is fired jointly by the engine jacket heat and by internal heat exchange from condenser C2. The suction line of the compressor is connected to the condensate end of the

As the absorption machine and the compressor share the same refrigerant, the question ofoil contamination of the heat transfer surfaces is important. The effect of oil on the heat transfer surfaces of the absorption machine is not known but it is expected that oil might interfere with the wetting of the absorber surface as well as adding additional heat transfer resistance to the other components. Another problem associated with oil in the cycle is oil accumulation in the solution circuits and consequent oil starvation of the compressor. With these significant problems in mind, an oil-free compressor was selected for the hybrid cycle. Both reciprocating (piston) and screw compressors are available in oil-free designs. Reciprocating compressors are made oil-free by introducing an additional shaft seal that separates the lubricated bearings from the compression chamber. Owing to the size of the proposed application (100 ton) and the relatively low rotational speed which large reciprocating compressors can tolerate, a reciprocating compressor for this application needs to be very large. A screw compressor seems to be a more natural choice. Oil-free screw compressors operate with close tolerances between the rotating screws and between the screws and the compressor casing. The main bearings of the compressor shaft are lubricated but oil seals isolate the oil from the compression chamber. Several companies were found 16:7 that supply oil-free screw compressors for steam service to the chemical industry. For the purpose of the design analysis, a compressor from Mycom ~6was chosen. A problem found during this study with all the oil-free compressors is that they exhibit reduced performance compared with standard compressor designs. The best isentropic efficiencies quoted for oil-free screw compressors are less than 50%. This poor performance results directly from the design implications of oil-free operation. In the case of the chosen screw compressor, the lack of an oil seal between the rotating screws and the housing leads to significant blow-by and consequent loss of efficiency. With such poor compressor performance, the potential advantages of the hybrid cycle are masked by low component performance.

Absorption chiller The hybrid cycle is based on a double-effect LiBr-H20 absorption chiller. Several manufacturers ~8,t9have chiller designs based on this cycle. A 100 ton machine from Hitachi t8 was chosen as the best match to the hybrid cycle design requirements. The design of the absorption machine is such that it would be possible to make the necessary connections with a minimum of alterations to the absorption machine. The compressor suction and discharge connections to the chiller require that the heat exchanger headers on the absorption machine are modified. The high temperature desorber needs to be modified Rev. Int. Froid 1991 Vo114 Septembre

269

Hybrid compression-absorption cycle. K. E. Herold et al. t.8

160

" 0

~

~ ~C)

~

Hybdd, ~=0.7

120

14

-

~

~

}-<

- O. 5

,?,

1.2

~._ _

~; 5 0.8 26

28

30

32

34

38

38

40

42

44

24

Figure 7

Thermal performance of the hybrid chiller, for two compressor efficiencies, plotted versus condenser temperature. Engine-driven R22 cycle and gas-fired absorption performance curves are included for comparison. Calculations based on an evaporator temperature of 2°C Figure 7 PerJormance thermique du groupe reJ~oidisseur hybride, pour deux ~ffficacitks de compresseur, calcul~e en fonction de la temp~'rature du condenseur. On inclut, pour les comparer, les courbes de la perJormance d'un cycle h absorption au gaz et celle d'un cycle au R22 avec moteur. Les cah'uls sont fondds sur une temperature de l'kvaporateur de 2°C

to accommodate heat from both the condenser C3 and from the engine exhaust gases. The low temperature desorber needs to be modified to be able to use heat from the engine cooling jacket.

Internal combustion engine Internal combustion engines fuelled by natural gas are the leading candidates for powering the hybrid cycle. Similar engine installations are currently in widespread use in both co-generation and engine-driven heat pump configurations. The engines in the size range of interest here are manufactured primarily by the automobile companies. Gasoline engines can normally be converted to run on natural gas with a simple conversion kit. For the purposes of this design, an engine from Ford Motor Co. 2° was chosen. This engine has a continuous power rating of 95 kW (159 HP) when fired by natural gas. A plot of the engine characteristics is given as Figure 4.

Hybrid cycle performance The performance of each of the three major subsystems (engine, absorption chiller and steam compressor) influences the performance of the hybrid cycle. The performance models of these components were described earlier. In particular, the compressor performance seems to be the limiting factor for the hybrid cycle as the need for an oil-free compressor limits the choice of hardware to compressors with degraded performance. The cooling performance (expressed as the COP) of the hybrid cycle is plotted in Figure 7 as a function of the heat rejection temperature. Cooling performance for the hybrid cycle is plotted for compressor efficiencies of 0.5 and 0.7. For comparison, the performance of an enginedriven R22 vapour compression cycle (with a compressor efficiency of 0.7) is also plotted along with the performance of a gas-fired double-effect L i B ~ H 2 0 chiller. The engine-driven R22 cycle model includes an engine model Int. J. Refrig. 1991 Vo114 September

26

28

-=-

30

Absocber

32

34

36

38

40

42

44

CONOENSERTEMPERATURE(C)

CONDENSERTEMPERATURE(C)

270

Low Desorber,D1

40

DoubleEffect Absorption

24

~--~ -

Figure 8

Key temperatures in the absorption components of the hybrid cycle versus condenser temperature. Calculations based on an evaporator temperature of 2°C Figure 8 Temp&atures cl~s des composants de I'absorption dans le (3,cle hybride en fonction de la temp&ature du condenseur. Les calculs sont fond~s sur une tempkrature de l'~vaporateur de 2°C

identical to that used in the hybrid cycle calculations. The gas-fired absorption cycle calculations include a burner efficiency of 0.85. Note that for the hybrid cycle the absorber exit temperature is assumed to be identical to the condenser temperature. Heat is rejected from the cycle to a sink from both the condenser C1 and the absorber A (for component designations, see Figure 1). As the heat rejection temperature increases, the performance of the hybrid cycle decreases. The performance of the R22 cycle falls off faster than the hybrid cycle due to the relative insensitivity of the absorption chiller performance to the sink temperature. It can be seen that for a compressor efficiency of 0.7 the performance of the hybrid cycle is better than the R22 cycle over the entire range of expected condenser conditions. However, for a compressor efficiency of 0.5, the performance of the hybrid cycle is better than the R22 cycle only when the condenser is above 35.5°C. Hybrid cycle performance advantages result from the more complete utilization of the thermal energy. The calculated temperatures in the absorption cycle components are plotted in Figure 8 versus condenser temperature. The lowest pair of curves represents the inlet (top curve) and outlet (bottom curve) temperatures in the absorber. The middle pair of curves represents the inlet and outlet temperatures in the low temperature desorber (D 1) and the upper pair of curves represents the high temperature desorber (D2). It is important to note that the high temperature in the absorption cycle increases at a faster rate than the condenser temperature. This can be seen by noting that the slopes of the uppermost curves in Figure 8 are greater than the slopes of the lowest curves. In fact, for each I°C rise in the condenser temperature, the high desorber temperature rises by approximately 4°C. This characteristic has important ramifications in sizing the compressor for the hybrid system as the compressor must be sized to handle the full range of planned operating conditions. Another way to look at the hybrid cycle is to plot the ratio of compressor power to the evaporator duty as in Figure 9. The two curves represent the hybrid cycle, for a compressor efficiency of 0.5, and an engine-driven R22 cycle. For the same size load, the hybrid cycle requires a

H y b r i d c o m p r e s s i o n - a b s o r p t i o n cycle." K. E. Herold et al. 0.3 14

Y

+

1.2

1

0.2

=~"~'~Hybrid

I--

Cycle, 1) = 0.5

0.8

idCycle. ~ =0.5

o,<

0.6

E3

0.4

0.1

EngineDriven R-22

0.2

o

,

~

,

;+

,

~

,

&

,

~

,

~

,

~

,

~

,

,

,

+,

0

i

24

i

l

l

28

+

i

i

30

32

i

i

i

34

~

i

36

+

i

38

J

+

40

~

+

42

CONDENSER TEMPERATURE(C)

CONDENSER TEMPERATURE(C)

Figure 9 Specific compressor power for the hybrid cycle and for the

Figure 10 Specific compressor displacement for the hybrid cycle and

reference R22 cycle versus condenser temperature. Calculations are based on an evaporator temperature of 2"C and a compressor efficiency of 0.5 for the hybrid and 0.7 for the R22 cycle. Specific compressor power is defined as the compressor power requirement divided by the evaporator load. Figure 9 Puissance spec(fique du compresseur pour le cycle hybride e/

the reference R22 cycle versus condenser temperature. Calculations are based on an evaporator temperature of 2°C and a compressor efficiency of 0.5 for the hybrid and 0:7 for the R22 cycle. Specific compressor displacement is defined as the volumetric flow-rate at the compressor inlet divided by the evaporator load Figure 10 D~placement sp&'(fique du compresseur pour/e cycle hybride

pour le cycle de r~f~rence au R22 en fonction de la tempkrature du condenseur. Les calculs son t Jbndks sur une temp&ature de l'~ vaporateur de 2°C et sur une £fficaciti" du compresseur de 0,5 pour le cycle hybride et de 0,7 pour le cycle au R22, La puissance spkcifique du compresseur est d~finie par le rapport: la puissance rappelke du compresseur ?t la charge de l'~vaporateur

et le cycle de r~J~rence au R22, en Jonction de la tempkrature du condenseur. Les cak'uls sontJbndks sur une tempkrature de l'kvaporateur de 2°C et sur une efficacitk du compresseur de 0,5 pour le o,cle hybride et de 0,7 pour le cycle au R22. Le dkplacement spi'c(fique du compresseur est d~fini par le rapport." dkbit volume h I'aspiration du compresseur h la charge de l'kvaporateur

smaller c o m p r e s s o r w o r k i n p u t t h a n the R22 cycle over the full range o f c o n d e n s e r t e m p e r a t u r e . The d a t a in F i g u r e 9 were used to size the engine for the cycle. F o r the worst case o f a c o n d e n s e r t e m p e r a t u r e o f 41.1°C, the c o m p r e s s o r requires 0.232 k W o f shaft w o r k per k W o f e v a p o r a t o r load. F o r the n o m i n a l e v a p o r a t o r design l o a d o f 352 k W (100 ton), the c o m p r e s s o r requires 82 k W . T o p r o v i d e a m o d e s t c a p a b i l i t y b e y o n d the rating point, the system was sized for a l o a d o f 437 k W at a c o n d e n s e r t e m p e r a t u r e o f 41.1°C. T h e engine was sized to p r o v i d e 120 k W to run the c o m p r e s s o r t h r o u g h the gear box a n d p r o v i d e a safety factor on p o w e r o f 20%. A n i n d i c a t i o n o f the c o m p r e s s o r size r e q u i r e m e n t is p l o t t e d in F i g u r e 10 versus c o n d e n s e r t e m p e r a t u r e . T h e c o r r e s p o n d i n g c o m p r e s s o r size for a c o n v e n t i o n a l R22 cycle is also p l o t t e d for reference. It can be seen t h a t the c o m p r e s s o r for the h y b r i d cycle is a p p r o x i m a t e l y f o u r times larger t h a n the R22 c o m p r e s s o r at low c o n d e n s e r t e m p e r a t u r e s . This is due p r i m a r i l y to the fact that the v a p o u r pressure o f w a t e r is fairly low at the t e m p e r a t u r e s o f interest. T h e r e f o r e the specific v o l u m e is large, leading to a large c o m p r e s s o r . Several o t h e r cycle v a r i a t i o n s ( F i g u r e s 2 a n d 3) which were c o n s i d e r e d early in the w o r k were a b a n d o n e d in f a v o u r o f this v a r i a t i o n because they required even larger c o m p r e s s o r s . The size o f the c o m p r e s s o r is g o v e r n e d p r i m a r i l y by the refrigerant a n d the o p e r a t i n g c o n d i t i o n s , b o t h o f which are c o n s t r a i n e d in this a p p l i c a t i o n . The effect o f c o n d e n s a t e injection on p e r f o r m a n c e is indicated in F i g u r e 11 where the h y b r i d cycle cooling C O P is p l o t t e d versus inlet q u a l i t y for three c o n d e n s e r t e m p e r a t u r e s . The simple m o d e l used here for c o n d e n sate injection indicates an i m p r o v e m e n t in p e r f o r m a n c e as the c o n d e n s a t e mass flow-rate increases from an inlet quality o f i.0 up to an inlet quality o f a p p r o x i m a t e l y 0.88. As the c o n d e n s a t e flow-rate is increased still further, the p e r f o r m a n c e d r o p s off slightly. The m o d e l

Condenser Temperature

=

30.0 °C ,35.8 o

1.5 o

o

o

o

o

41.1°C

0.5

o

o,

o e

oi+

olg

IqLETQUAUTY

Figure 11 Hybrid cycle cooling performance versus condensate injection rate (specified by the inlet quality) for three condenser temperatures. Calculations are based on an evaporator temperature of 2°C and a compressor efficiency of 0.5 Figure 11 Perfi)rmance de re/?oidissement du cycle hybride en.fonction du taux d'injection du condensat ( ddfini par la qualit~ ~t l'aspiration ) pour trois tempkratures du condenseur. Les cak+uls sont jbnd~,s sur une tempdrature de l'['vaporateur de 2°C et sur une efficacit¢; du compresseur de 0,5

indicates only a condensate mass densate injection p e r a t u r e s in the results presented o f 0.78.

mild d e p e n d e n c e o f p e r f o r m a n c e on flow-rate. This is e n c o u r a g i n g as conis required to m a i n t a i n r e a s o n a b l e temsteam c o m p r e s s o r . The h y b r i d cycle in this p a p e r are for a fixed inlet quality

Economic analysis F r o m a t h e r m o d y n a m i c perspective, the engine-driven h y b r i d cycle represents an interesting alternative for

Rev. Int. Froid 1991 Vol 14 Septembre

271

Hybrid compression-absorption cycle: K. E. Herold et al. combustion motivated cooling. However, in terms of current technology the hybrid cycle is not economically attractive for several reasons. The choice of an oil-free compressor implies poor compressor performance and poor hybrid cycle performance. The current state of the art in oil-free steam compressor design is such that isentropic efficiencies of 0.4-0.5 are considered good by manufacturers. However, these low efficiency values tend to bring the hybrid cycle performance down to a range comparable to a more conventional engine-driven R22 chiller which can operate with a high efficiency oil-lubricated compressor. Another factor which has a negative effect on the economies of the hybrid cycle is the high cost of the steam compressor. Although several companies give quotes on steam compressors, very few steam compressors are sold each year. As a result, the cost of these units carries a development cost burden which makes their use prohibitive in many promising applications. The authors received a quote from Mycom ~6for a screw compressor selected for our 352 kW application of $250 000. It should be noted that the quoted compressor has approximately 2.5 times more capacity than that required by this design. It is quoted here as it represents the smallest compressor of this type that is available. The compressor cost dwarfs the cost of the absorption chiller 18 of $60 000, and the gas engine 21 ($4000). Even if the compressor cost were to scale with the capacity, the compressor capital cost dominates the economic calculations. A comparably sized R22 chiller, with engine drive, would cost about $70 000 for a packaged unit2L Although the hybrid cycle has a slightly higher seasonal COP than an engine-driven R22 system, the capital cost of the system results in a prohibitive payback period for typical US economic conditions. It is unlikely that the cost of these steam compressors will decrease significantly unless the demand increases. Conversely, the manufacturers have little incentive to improve performance when the demand for their product is so low. For the hybrid cycle application described here, a compressor with an efficiency of 0.7 and a cost of $50 000 or less would probably be needed to bring the payback period down to a level which threatens the engine-driven R22 system market. According to the manufacturers, this is not expected to occur.

facturers have not resulted in a compressor choice that provides reasonable economics. However, the intrinsic potential of the hybrid cycle may be a motivation for those manufacturers to develop a more versatile oil-free steam compressor.

Acknowledgements The support of the Gas Research Institute for this work is gratefully acknowledged. The work was supported under Contract No. 5086-260-132122 . Acknowledgement is also made to the University of Maryland, Department of Mechanical Engineering, for providing computing resources.

References 1 2 3 4 5 6 7 8

9 10

11 12

Summary and conclusions A series of performance calculations were performed on several hybrid vapour compression-absorption cycles based on LiBr-H20. The cycles considered are truly hybrid where the refrigerant, in this case water, is common to both the compressor and the absorption chiller. The advantage of such cycles is the excellent match which can be obtained between the energy output of a gas engine and the energy needs of the cycle. Both the mechanical and thermal outputs of the engine can be used directly by the hybrid cycle to cool the load. In theory, this results in a higher COP for gas cooling as compared with either gas-fired absorption machines or gas engine-driven vapour compression machines. Performance calculations indicate that there is a real potential for these cycles. However, the implementation of the cycles in hardware presents several challenges. The main challenge is to find a low cost steam compressor with an adequate performance. Discussions with manu272

Int. J. Refrig. 1991 Vo114 September

13

14 15 16 17 18 19 20 21 22

Howe, L. A., Radermacher, R., Herold, K. E. Combined cycles for engine-driven heat pumps lnt J Refrig (1989) 12 21-28 Colosimo, D. D. Introduction to engine-driven heat pumps-concepts, approach and economics ASHRAE Trans (1987) 93 Pt. 2, 987-996 Wurm, J., Kinast, J. A. History and status of engine-driven heat pump developments in the U.S. ASHRAE Trans (1987) 93 Pt. 2, 997 1005 Morgan, J. R. Combined gas-engine-driven and absorption chillers, seminar presentation at ASHRAE Meeting on 28 June 1988 (report from Tecogen, Inc., Waltham, MA, USA) Alefeld, G., Ziegler, F. Advanced heat pump and air-conditioning cycles for the working pair H20/LiBr: domestic and commercial applications ASHRAE Trans (1985) 91 Pt. 2B, 2062-2071 Alefeld, G. U.S. Patent 4 531 374 (1985) .~,hlby,L. Compression/absorption cycles for large heat pumps system simulations Research Report, Chalmers University of Technology, G6teborg, Sweden (1987) Radermaeher, R., Zheng, Z., Herold, K. E. Vapor compression heat pump with two-stage solution circuit: proof of concept Proceedings of International Workshop on Absorption Heat Pumps, London, April 12 14 1988 Commission of the European Communities, Luxembourg Stokar, M. Kompressionswarmepumpe mit Losungskreislauf PhD Dissertation ETH Zurich, Switzerland, No. 8101 (1986) Sunye, R., Prevost, M., Bugarel, R. High temperature sorption cycle for heat pumping: compressor aided heat transformer Proceedings of International Workshop on Absorption Heat Pumps. London, April 1~14 1988 Commission of the European Communities, Luxembourg MeLinden, M. O., Klein, S. A. Steady-state modeling of absorption heat pumps with a comparison to experiments ASHRAE Trans (1985) 91 Pt. 2B, 1793 1807 Radermacher, R., Herold, K. E., Howe, L.A. Combined vapor compression/absorption cycles Proceedings of International Workshop on Absorption Heat Pumps, London, April 12 14 1988 Commission of the European Communities, Luxembourg Herold, K. E., Moran, M. J. A thermodynamic investigation of an absorption temperature boosting heat pump cycle Proceedings of Advanced Energy Systems Symposium at the ASME Winter Annual Meeting, 1985 AES, Vol. 1, 81 88 McNeely, L. A. Thermodynamic properties of aqueous solutions of lithium bromide ASHRAE Trans (1979) 85 Pt. 1,413 434 Keenan, J. P., Keyes, F. G., Hill, P. G., Moore, J. G. Steam Tables Wiley, New York (1969) Mycom Corp., 19475 Gramercy Place, Torrance, CA 90501, USA, Model STM200LL AC Compressor, 1126 S. 70th St., Milwaukee, Wi 53214, USA Hitachi, Gas Energy Inc., 166 Montague St., Brooklyn, NY 11201, USA, Model 10N Trane, 3600 Pammel Creek Rd., Lacrosse, WI 54601, USA Ford Motor Co., Engine Distributors Inc., 332 S. 17th St, Camden, NJ 08105, USA, Model LSG-875 Tecogen, 45 First Ave., P.O. Box 9046, Waltham, MA 02254. USA Radermacher,R., Herold, K.E., Howe, L. A. Combined gas-fired vapor compression/absorption cycles Final Report for GRI, Contract No. 5086-260-1321 (1988)