Solar Absorption Cooling Feasibility

SOLAR ABSORPTION COOLING FEASIBILITY

DAN S. WARD ASSOCIATE DIRECTOR

Solar Energy Applications Laboratory Colorado State University SOLAR ABSORPTION COOLING FEASIBILITY

Dan S. Ward Associate Director Solar Energy Applications Laboratory Colorado State University Fort Collins, Colorado 80523

ABSTRACT

The feasibility of solar absorption cooling systems is dependent upon its technical and economically competitive position with respect to other cooling system alternatives. Technical feasibility can be shown by compari- sons of the thermodynamic efficiency of solar absorption cooling with conventional vapor-compression cooling equipment and by reference to numerous experimental evaluations. Economic feasibility is heavily dependent upon the financial parameters assumed (in particular the inflation rate of conventional fuel costs). In particular cases, i .e., particular assumptions of the financial parameters, economic feasibility of solar absorption cooling can be demonstrated. INTRODUCTION Solar space cooling of buildings may be accomplished by a number of alternative methods. One of the most promising is the use of an absorption refrigeration cycle. A principal advantage of this method is the small amount of mechanical work required. And, while a heat input many times greater than the work input of a mechanical vapor-compression cycle is required, the absorption cycle can be economically attractive if the heat is sufficiently cheap. An absorption refrigeration system could be labeled a vapor-compression system, where several of the components of the absorption refrigerati on cycle (notably the absorber and generator) are required to perform the function of the compressor in a mechanical vapor-compression system. But, while the coefficient of performance (COP) of a vapor-compression system is typically 2 to 4, the COP of absorption cycles is limited; the maximum attainable COP for an absorption system is equal to the coefficient of performance for a Carnot refrigerating cycle working between the evaporator temperature and the ambient temperature, multiplied by the efficiency of a Carnot engine working between the generator temperature and the ambient temperature. For a given ambient temperature, the COP will increase with an increase in the heating medium temperature input to the generator. The fluid in the generator consists of a solution of refrigerant and absorbent which have a strong chemical affinity for each other. In most corrunerci al systems the refrigerant/absorbent combinations are water/lithium bromide and arrmonia/water. The arrmonia/water absorption system may be applicable whenever ammonia is a suitable refrigerant. However, because the absorbent (water) is volati le, the refrigerant vapor leaving the generator will contain too much water, so that additional equipment to rectify the generator vapor and increase the ammonia concentration is required. 2

An outstanding feature of the water/lithium bromide absorption system is the non-volatility of the lithium bromide. In the generator only water vapor is driven off, eliminating the need for rectifying equipment. Com- pared to the ammonia/water system, the water/lithium bromide system is simpler and operates with a higher COP. The primary disadvantage of the water/LiBr system is its requirement for relatively high evaporating tempera tu res. Other differences between ammonia/water and water/LiBr systems include water-cooling versus air-cooling requirements, need for a solution pump, and use of direct-expansion evaporators. Ammonia/water systems, for example, can be air-cooled whenever generator inlet temperatures of 120 to 180°C are available. Water/LiBr units always require water-cooling, as do ammonia/water systems operating with generator inlet temperatures of less than l 20°C. Ammonia/water systems require mechanical solution pumps to pump the working fluid from the absorber pressure to the generator pressure; thus requiring additional parasitic electrical power. Because of the small pressure differential between the high and low pressure portions of the

11 11 water/LiBr unit, vapor-lift (or bu bble ) pumps may be utilized along with gravity return of solution from the absorber to the generator. Thus the parasitic electrical power requirements of water/LiBr units can be less, although in larger cooling capacity units, mechanical solution pumps are often utilized. Finally, ammonia is considered flammable and toxic, so that it is not utilized with direct-expansion evaporator coils, whenever the air to be conditioned will be in direct contact with the evaporator. Thus a separate chilled water loop is necessary. Carnot coefficients of performance for absorption cooling equipment could range from 0.5 to 2.5, when operated at temperature and heat input 3 rates applicable for solar applications [l]. Practical absorption cycles are, of course, non-ideal cycles due to hardware limitations, irreversible processes, working fluid properties, etc. Thus a practical COP of a commercially available, single-effect unit is in the range of 0.5 to 0.7 for ammonia/water units and 0.6 to 0.8 for water/lithium bromide units.

Solar Absorption Cooling Experiments The utilization of solar energy as the heat input to the generator of absorption cooling units has been reported by numerous researchers. These include efforts by Ltlf [2], Trombe, et al [3], Williams [4], Eisenstadt, et al [5], Chinnappa [6], Chung, et al [7], Duffie, et al [8], Swartman, et al [9,10] and Farber [ll]. The basic conclusion of these papers is the demonstrated ability of solar energy fla t-plate collectors to achieve the required temperatures necessary to provide the heat input to operate the absorption refrigeration units. In recent years the experimental incorporation of an absorption chiller into a solar heating and cooling system has also been accomplished. Note- worthy experiments have been conducted by Ward, et al [12, 13, 14, 15], Namkoong [16], San Martin,etal [17,18], and Jacobsen [19]. Numerous other experimen tal cooling systems have been designed and fabricated, but extensive performance data are not yet available. In addition to systems t esting, continuing efforts have been directed toward additional experimental testing of absorption cooling units as well as a variety of computer simulations. Recent efforts in the absorption refrigeration experiments on independent machines include Anderson [20], Simmons, et al [21], Dao, et al [22,23] and Merrick [24]. Considerable importance must be directed toward the experimental performance of solar cooling systems because it is these results which constitute the severest test of the feasibility of solar absorption cooling. 4

Computer simulations suffer from a severe lack of ability to predict and model actual operating conditions. And, while continued efforts at improving the individual units and their respective performance are in progress, such efforts have meaning only insofar as they improve the system performance. This fact has been emphasized by Newton [25].

Solar Absorption Cooling System Performance Because of their nearerterm commercial availability, water/lithium bromide absorption units have recei ved the bulk of cooling system performance testing. Therefore the below discussion is directed toward those systems using the water/LiBr units . The first noteworthy factors are the temperature requirements. The water/LiBr residential-sized units are designed for generator input temperatures of 88° to 90°C under condtiions of cooling water temperatures of 30°C, and in order to achieve chilled water or evaporator temperatures of 7° to 8°C. Lower generator input temperatures are possible when lower cooling capacity is acceptable (75 ° to 80°C input temperature to the generator allows for about 80 percent of the cooling capacity) or if lower cooling water temperatures are available (70° to 75°C generator inlet temperatures are possible for cooling water at 25°C and a cooling capacity of 80 percent is acceptable). It is important to realize that 88°C temperatures are easily obtained with existing, high quality, liquid-heating solar flat-plate collectors and at reasonable efficiencies. In fact, the summer operating efficiency can be expected to be better than the winter performance. This latter result can be seen if we assume some typical operating conditions [e.g., see reference 26]. In January the collector inlet temperature (Ti) might average 55°C with an ambient temperature (Ta) of -5°C. The January solar radiation on a tilted surface (HT) would be approximately 5

2 2 600 wa tt/m . This gives a value for (\ - Ta)/HT = 0.1 m ·°C/watt. In 2 July we could expect T. = 90°C, T = 30°C, and HT = 800 watts/m ; so that i a 2 0 (Ti - Ta)/HT = 0.075 m . c/watt. Under these circumstances we can therefore expect improved solar collector efficiency for the summer months. A critical assumption in the foregoing calculation is that thermal storage heat losses do not affect the results. In general, such heat losses cannot be neglected, and can substantially alter the results. This point is seen most clearly when we choose a hot water storage tank as the thermal storage medium. Note that a phase change storage subsystem is not feasible for storing the collected solar energy for both the summer and winter seasons because of the totally different seasonal temperature ranges (35° to 70°C winter; 75° to l00°C summer). This leaves hot water as the thermal storage medium. A critical problem with hot water is the heat losses from the storage tank. Jacobsen [19] has observed an actual heat loss coefficient of l .65 watts/m2. 0 c, which was approximately 50 percent greater than the predicted value of l .19 watts/m2·°C . For a 55°C temperature difference between storage and ambient the actual heat loss becomes 1200 watts! Such deviations from predicted values are apparently conmon [18,26,27]. For example, Ward [26] has reported heat losses from a hot water thermal storage unit of 880 watts (equivalent to two hours of operation of the installed chiller). Increased and heavier insulation reduced this to 330 watts. A common design in solar heating systems is to locate the thermal hot storage unit inside the heated space so that heat losses from the storage help to meet the heating demand. In this case, it is, of course, permissible to neglect the heat losses. But the addition of a cooling system implies that the heat losses not only degrade the ability of the solar system to meet the cooling load, but actually increase the cooling demand itself. Ward [26] has reported that the effect of heat losses 6 during a month of June actually reduced the percent of cooling load carried by solar to a negative number. One method of avoiding this increasing of the cooling load by thermal storage heat losses is to locate the thermal storage outside the conditioned space. Unfortunately, this increases the winter heat losses from storage (because of the greater temperature differential between storage and ambient), reduces the ability of the thermal storage to meet the heating load, and increases the chances of freezing the storage unit. A more preferable alternative would be to use a triple thermal storage system. Ward [28] has discussed the use of a "cool storage" system to reduce the normal operating temperature of the hot thermal storage unit to the minimum temperature that the absorption chiller can effectively utilize, and to allow operation of the chiller whenever solar energy is available, irregardless of the cooling demand. It is noteworthy that a cool storage will undergo some heat gains from the ambient and that, if the cool storage units are located within the conditioned space, this will constitute a heat removal method and assist the solar system to meet the cooling demand. In this respect it is similar to the concept of heat losses from a hot thermal storage unit contributing to the winter heating load. In order to take advantage of these aspects, an alternative storage system could be comprised of three water storage tanks. The first tank would be located exterior to the conditioned space and would be twice the volume of the other two identically-sized tanks, both of which would be located within the conditioned space. During the winter heating season the exterior tank would be empty and the two interior tanks used as a (slightly stratified) hot t hermal storage unit. During the summer cooling season, the exterior tank would be used for the thermal hot storage subsystem and the interior tanks would be used in the cooling subsystem (as described by Ward [28,29]). Thus heat losses from the interior tank in 7 winter would aid in meeting the heating load and in surrrner assist in meeting the cooling load. The heat losses of the exterior tank in summer would be less (due to a lower temperature difference, i.e., a higher ambient temperature) and would not add to the cooling load. And, because of the triple storage system, calculation of the higher summer collector efficiency described above can be justified and can be expected to result in an improved system efficiency.

Arrrnonia/Water Absorption Systems A principal motivation for the development of an ammonia/water solar absorption unit has been for the purpose of eliminating the use of a water cooling tower [21]. This is particularly important at the residential level, because regular preventative maintenance of the cooling tower is not always feasible and because of questions of local water quality. Dao [23] has reported recently on the results of a long-term research and development prog ram for ammonia/water absorption chillers. Past accomplishments include the modification of a 5-ton (17.58 kW), gas-fired unit to operate at a capacity of l .7-tons (6.00 kW) for conditions of generator temperatures of 77°C, condenser-absorber temperatures of 35°C, and evaporator temperatures of 6°C with a COP of 0.65. Work is continuing on units of 3 to 5 tons with a minimum COP of 0.65 and generator temperatures of 95° to ll5°C. The estimated costs of these units are $3,000 to $4,000, or about $1 ,000 per ton. The potential for ammonia/water chillers to replace water/LiBr chillers in the near future does not appear too great. The development of these smaller units are clearly in a less advanced stage than LiBr units; there have, for example, been no complete solar system experiments with ammonia/ water·.- , In, addition to systems considerations, there is also the safety hazard of using ammonia within the interior of the building. Consequently in 8 the remainder of this paper, the water/LiBr system will be emphasized because of its commercial availability and because of the more extensive experience with water/LiBr units in solar cooling systems.

COOLING SUBSYSTEM DESIGN In designing a solar heating and cooling system is it particularly important to consider the results of the solar cooling experiments briefly reviewed above. These include: (1) The near-term commercial availability of water/LiBr absorption machines (as compared to the research stage for ammonia/water units). (2) Experience in research and development of solar cooling with water/LiBr machines (both components and system) which demonstrate the technical feasibility. (3) The importance of adhering to design conditions (temperatures, flow rates, etc.) in the operation of an absorption cooling unit. (4) High quality, liquid-heating flat-plate solar collectors are adequate for providing solar heat to absorption units. (5) Thermal storage heat losses/gains are of vital importance in system design and must not be allowed to degrade the designed performance of the system. (6) Cool storage may provide for higher seasonal coefficients of performance of units, as well as allowing for smaller tonnages of units for the same cooling load. (7) Temperature stratification in the cool storage subsystem is critically important. (8) System designs should allow for safeguards against crystallization of the LiBr solution and other potential absorption unit failures. 9

Based on the above criteria, design schematics of the solar heating and cooling system may then be developed. If we consider the cooling capability of the solar heating and cooling system as an addition to a solar heating system, we might describe the additional complexity in terms of additional components or equipment. Thus, from an economic viewpoint, solar cooling adds the following equipment (and associated costs) to the solar heating system: (1) Water/LiBr absorption chiller (of appropriate tonnage) (2) Water cooling tower (of appropriate tonnage) (3) Two thermal storage tanks (cool storage) [OR specially designed, temperature stratified tanks for large commercial applications] (4) Two pumps (cooling tower pump, chilled water pump) (5) Four automatic valves (6) Liquid-to-air heat exchanger ("cooling coils"). The existing "heating coils" may be used (see Figure 1) (7) Additional piping, hand valves, vents, etc. (8) Additional. control instrumentation and complexity Figure l shows the arrangement of the absorption chiller in relation to three thermal storage units. Table l provides the operational modes of the system. As previously mentioned, the advantage of this system is to ensure that the heat losses/gains from hot/cool thermal storage always contribute to meeting the heating/cooling loads. The principal disadvantage is the requirement of shifting from heating to cooling modes in the spring and back to heating in the fall; thus assuming an accurate anticipation of the weather. This problem, however, can be alleviated by only a slight increase in the complexity of the system, and would reduce to a problem of ensuring that Tank #1 does not encounter freezing conditions when it is full. The importance of Figure l is to provide an i ndi cation of the necessary components for the addition of solar cooling to an existing solar heating Liquid-to-Air Heat E1chanoer

Return~ Supply~ Air Air

...

From coJlectors (vent)__. (winter) l _____+- _

Thermal Thermal Storaoe Storaoe Circulatory Tank#2 Tank# 3 Pump !

Evaporator : ._ ------'"1 0 Li Br Chiller Absorber I ~ ------.JI .,c Condensor I ~ CT Pum .. 1 7777777777 >pzzzz, azzzzzzmzz zzz >zzzzzzzzzzzzJ!J.. !W'7W!?J~Zzzz l Bulldlno Exterior

"//\~ Water Thtrmal +-From• Cool Ing Stora;• col l1ctor1 Tower Tank•I To• (1umm1rl

Figure I. Solar Heating and Coolln; Syatem ( Triple Storage Tank•) Table l. Operational Modes for the Solar Heating and Cooling System Shown in Figure 1

Automatically Pumps Operational Modes Actuated Valves Fan AVl AV2 AV3 Circ G c CT

WINTER OPERATION (Tank #1 drained, Valve Vl open, Valve V2 closed, Tanks 2 and 3 full) l. Collector delivering heat to storage -- A A -- Off Off Off -- (Tank 3, return via Tank 2) 2. Storage delivering heat to heating coils A A -- On Off Off Off On 3. Auxiliary delivering heat to heating coils B -- A On Off Off Off On

SUMMER OPERATION (Tank #1 full, Tanks 2 and 3 - variable level -half-full, Valve Vl closed, Valve V2 open)*

l. Collector delivering heat to storage A B ------(Tank l is hot storage) 2. Storage delivering heat to chiller A B A -- On On On -- 3. Auxiliary delivering heat to chiller A B B -- On On On -- 4. Chilled storage delivering cool to cooling coils A -- -- On ------On 5. Recycle chilled water storage A -- -- On Off Off Off Off (pump v1arm contents of Tank 3 back to Tank 2)

*Tank l is for hot storage (~90°C), Tank 2 is for warm storage (~l5°C) and Tank 3 is for cool storage (~9°C) l 2 system. These additional components are critical when we consider the thermodynamic efficiency and economic feasibility of solar cooling.

Thermodynamic Efficiency The essence of the feasibility of solar absorption cooling systems is its technical and economically competitive position with respect to other space cooling system alternatives, including in particular conventional vapor-compression cooling units . The critical factors here are the differences in seasonal coefficients of performance between the absorption and compression systems (including the differences in parasitic power requirements), a technical consideration; and the economic factors of the solar system capital costs (which are not subject to inflation over the life of the system) and conventional system fuel costs (which are strongly dependent upon the anticipated conventional energy inflation rates). In order to obtain a system efficiency for a cooling alternative, it is necessary to consider the efficiencies of all steps in the conversion of an energy source to the useful work performed, i.e., the extraction of heat from a building. For example, the overall conventional cooling system efficiency (nvc) would be defined as the amount of heat removed from the building (i.e., the amoun t of space cooling) by the conventional vapor- compression unit when operating with electrical energy (eve), divided by the fuel input required for producing the electrical energy to operate the

n ~ c E is just the vapor-compression cooling unit (Ee), i.e., =eve/Ee. c electrical energy input necessary to operate the vapor-compression system (Ee) divided by the efficiency of generating and delivering energy to the consumer (ne). Thus the expression for nv c can be written, nvc = ne e vc /E e . However, eve/Ee is just the coeff ici ent of performance of the vapor- compression unit, (eOP)vc · Thus:

nvc = (eOP)vc( ne) (1) 13

For a solar water/L i Br absorpti on system using an auxiliary fired by conventional fossil fuel s (natural gas, co al , fuel oil, propane, etc.), the various efficiencies which we must con side r include the efficiency of the auxiliary furnace in conve rting t he f uel input to useful heat energy for delivery to the generat or of the absorpti on unit (nA), the rated COP of the solar power un i t , and th e parasitic power requirements of the solar and auxiliary systems. Fo r a solar absorpti on cooling unit utilizing solar for a fractional percentage of t he cooling load , we obtain the overall system efficiency of the sol ar absorption cooli ng system of:

( 2) where f =Fraction of t he cooling load carried by solar energy being del i ve red to the generator of t he absorption unit, Cs= The amount of heat removed from the bu ilding (i.e., the amount of space cooling) by t he abso rption cooling unit when operati ng with solar energy, CA = The amount of heat removed from t he bu il di ng (i.e., the amo unt of space cooling) by t he ab sorption cooling unit whe n operati ng with auxiliary (conv ention al) energy, Gs= The amount of solar energy delivered t o the genera tor of the ab sorption cooling unit The amo unt of auxiliary energy deli vered to the generator of t he absorption cooling unit Es = The fu el input required for prod uci ng t he electrical energy used by the solar subsystem, EA = The fu el input requi r ed for produci ng the electrical energy use d by the auxiliary subsystem. We can simpl i fy equation (2) by defini ng ts and t A and the percentage of electrical energy i nput necessary to deli ver t he solar and auxiliary energy (respect ively) to t he cooling unit, so t hat:

ts G E = __s. ( 3) s ne ' 14

Utilizing equations (3) and the fact that Cs/Gs and CA/GA represent the coefficients of performance for the solar-driven (COP)s and auxiliary-driven (COP)A absorption units, respectively, we can modify equation (2) to obtain:

f(COP)s (1-f)(COP)A (4) ns l + 2s/ne + 1 + 2A/ne (nA)

Equations (1) and (4) now allow us to directly compare the coefficients of performance for solar absorption and conventional (electrically-driven) vapor-compression machines. For example, we may utilize some previous experimental data [14,28] to assume the values of some of the parameters in equation (4), i.e.,

(COP)s = (COP)A = 0.65

2s = 0.08

2A = 0.01

In addition we can assume an auxiliary furnace efficiency of nA = 0.75 and an overall average efficiency for different electrical-generating power plants of ne ~ 26 percentll [30]. Under these assumptions equation (4) becomes:

ns = 0.469 + 0.028 f (5)

An intermediate observation is that the fraction of the cooling load carried by solar has a minimal effect on ns; i.e., f has a minor effect on the thermodynamic efficiency of the solar/auxiliary system for the condition where the above assumption s are applicable. For an f = 0.5 to 0.8, ns = 0.48 to 0.49. For the conventional vapor-compression system we may use a seasonal coefficient of performance of (COP)vc = 2 to 3. In this case, equation (1) yields nvc = 0.52 to 0.78.

1Electrical generating efficiency (30%), transmission efficiency (91%), and distribution efficiency (95%); reference [30] 15

Thus the conventional vapor-compression system yields a thermodynamic improvement over the solar absorption system of:

l . 08 to l . 63 ( 6)

(avg = 1.35)

On the other hand, the absorption system has some practical advantages. It has, for example, fewer mechanical moving parts and is thus less subject to wear and should require less maintenance. The absorption system may operate at reduced evaporating pressures with little decrease in refrigerating capacity, and liquid carry-over from the evaporator does not cause difficulties as in the mechanical systems. The only practical disadvantages of absorption systems are the greater complexity and lower COP of the ammonia/water systems and the potential for crystallization of the absorbent and the maintenance of a strong vacuum (against production of hydrogen and mechanical leaks) with a lithium bromide absorption system.

IMPROVEMENTS IN SOLAR ABSORPTION EFFICIENCIES As previously mentioned, the COP of a single effect absorption unit using a conventional working fluid is limited and always less than 1.0. Some substantial improvements are possible, however, for higher generator inlet temperatures. For example, hot water at a temperature of 175° to 200°C could be employed in a double effect absorption cycle to improve the COP. For water cooled, double-effect, water/LiBr units, coefficients of performance could obtain values as high as 0.99, and the COP of the double-effect unit is no longer limited to 1.0. Unfortunately, the double-effect operation with air-cooled, amroonia/water units is not practical. See reference [l] for further discussion of the double-effect absorption cycles. Combination absorption-resorption cycles ·(CAR cycle) also have the potential for higher COPs (0.8 to l .0) but again involve high temperatures 16

(=150°C) in providing heat to the generator [l]. And, in this case, as well as for double-effect units, cooling capacities are normally in the range of 400 to 1100 tons. Thus for smaller tonnages, the apparent limitation of COP of absorption units is about 0.8. It is noteworthy that the efficiency of solar absorption cooling, defined in equation (4), would increase to a value of 60 percent for a COP of the absorption unit of 0.8 (instead of 0.65). This represents a 22 percent increase in performance and lowers the relative thermodynamics efficiencies of conventional vapor-compression systems to solar absorption systems to a range of: nvc/ns = 0.87 to 1.30. Effectively, therefore, the solar absorption system equivalent to the vapor-compression systems in terms of total system efficiency.

ECONOMIC CONSIDERATIONS A critical factor in the question of the utilization of solar absorption cooling systems is the economic feasibility. Generally an assumption is made to consider only the costs of the absorption chiller and the specific cooling subsystem costs as the capital cost of the solar cooling, and consider the cost of solar collectors, thermal storage, etc., as part of the solar heating system. This assumption favors solar cooling since the savings in the cost for cooling must offset only the cost of the chiller and related hardware, and not other portions of the solar system. (A listing of the required solar components is given above). Economically much of the thermodynamic advantage of a conventional system (see, e.g., equation 6) is lost in the costs of profit and overhead of the electrical utilities (in effect, the consumer pays for the capital cost of the power plant, plus administrative costs and profit). Therefore even in the event of natural gas prices increasing to the point of equal competition with other fuels, the cost of electricity over the cost of 1 7 natural gas will still be substantially higher to account for power plant efficiency and utilities' overhead and profit. In point of fact, 60 percent of the cost of electricity is "demand related" [30], which means that the rapidly escalating costs of capital construction of new and replacement power plants constitutes 60 percent of the total cost of electricity. The remaining 40 percent of the electrical cost is 11 base load-related", i.e., the cost of fuel and administrative overhead and profit constitutes only 40 percent of the electrical costs. For example, Public Service of Colorado pays about $0.75/mBtu for its energy (coal, gas, oil). When the efficiency of conversion, transmission and distribution are included, the fuel cost of the electric bill is about $3/mBtu. Electricity is sold at a price of $11/mBtu. Thus fuel costs constitute about 27 percent of the total price. It may be that, when rapidly escalating inflation costs of the conventional fuels are compared to the stable amortization of a solar system's capital costs, we can realize an economically competitive advantage of solar absorption cooling. This advantage would become even more pronounced when the environmental costs of electrical production are included. In evaluating the economic feasibility of a solar cooling system, two considerations are necessary. One is the determination of the fractional portion of the load which the solar cooling system can be expected to carry given the building cooling requirements, the location (site), and the size of the solar heating system. The second step is to evaluate the potential savings in life-cycle cost of the system using present worth costing. Bartlett [31] has performed an analysis of site-dependent factors which affect the economic feasibility of solar absorption cooling. These factors include the need for: (1) a high heating load relative to the cooling load; (2) a high collector efficiency during the cooling season relative to the heating load [a factor easily accomplished]; (3) a high insolation during 18 the sununer relative to the insolation in the winter; (4) a high absorption COP; (5) a high percent solar heating; (6) a high cost for conventional energy; and (7) a low cost for auxiliary energy. The results of Bartlett's analysis [31] indicated that, based on the assumptions made, residential applications of solar absorption cooling (i.e., low tonnage units , e.g .• 3 to 5 tons) are not currently economically attractive (1976 costs), and that commercial applications were found to be more cost-effective. In general, Bartlett found that the larger the chiller the more conomically feasible it would be. This latter factor is not particularly surprising, since the cost per ton of cooling capacity for absorption machines is reduced co ns iderably as we move to larger tonnage units. It is also noteworthy t ha t conventional absorption cooling units are much more feasible at larger tonnages, as evidenced by the large number of commercial absorption units now installed (25 tons and larger) and as compared to the relatively small number of residential sized units currently installed. It is noteworthy , however, t hat the Japanese firm, Yazaki Corporation, has found that by mass produc tion techniques, they can manufacture four 10-ton cooling units cheaper t han a si ngle 40-ton unit [32]. The smaller Yazaki units of l to 5 tons, however, are more expensive per ton and are generally competitive with prices of the American firm, Arkla Industries [33]. It should be po i nted out that Bartlett' s analysis [31] was based on the cost of utilities obtai ned in the fourth quarter of 1976. This means that solar heat is attempting to compete with natural gas, whose regulated cost is unrealistically low. Bartlett did use a fuel cost escalation rate of 14 percent, but a proposed ene rgy bi ll in Congress involves a first year increase in the price of natural gas of 30 to 100 percent or more. After some study it becomes evident that natural gas prices are an unrealistic price basis for conventional energy sources and, because of the 19 complete lack of ability to predict future pricing of gas, another source of energy should be considered in comparing solar and conventional costs. Because of its wide availability and since it incorporates within its rate structure the costs of coal, gas, hydroelectric and fuel oil, the cost of electricity can best be used for analysis . However, the economic feasibility of solar absorption cooling is basically related to the comparative costs of conventional electrically-driven vapor-compression units and solar absorption units with a non-electric auxiliary. Because natural gas can be expected to be used for the solar absorption system's auxiliary heat source, it is necessary to relate the current cost of electricity to some hypothetical, unregulated price of natural gas. This can be done by assuming that deregulation of natural gas prices would allow the price of natural gas to rise to the point where it is competitive on a dollar per Btu basis with coal and fuel oil. Because the cost of fuel represents only about 25 percent of the cost of electricity [30], we can then use a price of natural gas (and its associated cost inflation rate) of 25 percent of the respective values for electricity. The end result is to use current costs of electricity for our non-solar system cost of energy, and to use one-fourth this price for the cost of the auxiliary fuel for the solar system.

ECONOMIC ANALYSIS In general, the question of economic feasibility of solar energy systems is the balancing of the capital cost of a solar system against the savings in conventional fuel costs. The critical factor is whether or not the solar energy system will cost less over a specific life-cycle than the conventional system. In the economics of a solar heating system, procedures have been developed to determine feasibility. Kreith [34] has provided a method for analyzing the economics of heating and cooling for buildings and included 20 an overview of the current state of solar system design and optimization. Barley [35] and others [36] have performed similar analyses. A critical aspect of the economic feasibility of the solar system is the determination of the fraction of the load that the appropriately sized solar system can be expected to carry (f). Klein, et al [37] has provided a method for determining f for a solar cooling system. Using Barley's [35] economic analysis method, it is a straightforward calculation to determine the potential savings attainable by the use of solar absorption cooling. We will assume that the installed cost of a solar cooling system includes the equipment and installation costs of: (1) the chiller and cooling tower, Cc; (2) the cool storage units, Cs; (3) the associated equipment (pumps, exchangers, piping, etc.) necessary for interfacing the cooling subsystem with the solar heating system, CE; (4) the portion of the installed cost of the solar heating system which is chargeable to the cooling system, F; (5) the capital cost of the auxiliary cooling system, CA; and (6) the cost of installing the solar system, c1. Thus:

(7) where ACa is the cost of the installed solar cooling subsystem (as used by Barley [35,38]). Equation (7) may be simplified somewhat by assuming that the solar cooling system is an addition to an economically justifiable solar heating system. That is, we utilize Barley's [35] analysis to demonstrate the economic feasibility of a solar heating system (using ACa as the installed cost of the solar heating system) and then ask if the addition of a solar cooling system can also be economically justified. In one respect this favors the economics of solar cooling since the cooling system cost does not include charges for the installation of the solar collectors, thermal (hot) storage, etc. [i.e., F = 0], and thus the capital 21 cost of the solar cooling system is less. On the other hand, for applications where the cooling load is significantly greater than the heating load, the solar system is no longer economically optimized. In this latter case it may be preferable to add collector area (thus F > 0) in order to benefit the solar cooling system. An analagous consideration is the use of the auxiliary heating unit to supply conventional heat to run the chiller. In this case, CA= 0. We again obtain the advantage of dual use of equipment, but this assumes that the conversion of fuel to auxiliary heat (at some efficiency) to run an absorption chiller (at some COP) will provide cooling at a more economical rate than the use of a conventional mechanical-compression machine as the solar system auxiliary. Finally, we note that the additional cool storage tanks will require some building space and therefore some possible additional costs. But this is just the space requirements for the solar heating system, since for our triple tank design (see Figure 1), the interior tanks are used for winter heating and are therefore chargeable to the solar heating system costs.

ECONOMIC FEASIBILITY The large number of variables in the economic analysis make specific conclusions tentative and dependent upon the reliability of the assumptions used. Certain variables (particularly the inflation rates) allow for a variety of possible conclusions. Ward [39] has done an analysis of the effects on economic feasibility calculations of variations in the assumed economic parameters. Of course, many of the variables involved in the computations are relatively straightforward. For example, interest rates on a loan for the solar system capital costs can be estimated with some confidence and will, of course, remain constant over the period of the economic analysis. Similarly, the percent of downpayment is also easily 22 determined. Property taxes and insurance are also capable of realistic determination (although their adherence to general inflation rate is questionable). Income tax rates, deductions, investment credits, depreciation, etc. are strictly viable only for individual case studies, but can be estimated for typical situations. Even the general inflation rates over 20 years can be estimated with reasonable accuracy (about 6 percent over the last 20 years). Discount rates are also variable and depend to some extent on individual case studies. The fuel inflation rate is, however, far and away the more unreliable and important variable. One utility [30], for example, estimates an annual inflation rate on the cost of electricity of 11 percent per year for the next four years. The utility will not hazard a guess at inflation rates over a longer period of time. Despite these difficulties, it is nevertheless useful to consider some typical results of calculations which evaluate the economics of solar absorption cooling for varying rates of inflation, cooling capacities of the cooling units, and under conditions of different tax incentives. These are shown in Table 2. (Table 3 includes the technical and economic assumptions for the results shown in Table 2.) The values in Table 2 have been rounded off to the nearest hundred dollars even though the accuracy is probabJy less than two significant figures [39].

An obvious and expected conclusion is that the larger 25-ton unit for commercial applications is economically more competitive than the residential 3-ton unit. But perhaps more significant is the inflation rates necessary for an economic residential application. Based on the assumptions of Table 3, a fuel inflation rate of 13 percent is needed to break even on the solar installation. This can be compared to the predicted electrical cost inflation rate of one utility of 11 percent [30]. 23

Table 2. Results of Economic Analysis t

Cost of 25-Ton 3-Ton Application Savings ($) Electricity Application 25 % Ta x Rebate · 4% Loan Savings ($) Inflation No Tax on Solar Availability Rate, rel ec (no tax Incentives Capita 1 Equip. on Solar (%/year) incentives) Cao ital Cost

0 l, 000 (2,700) ( l , 600) (2,200) 3 2, 800 ( 2, 400 ) ( l , 300) ( l , 900) 6 5,000 (l ,900) (900) ( l , 500) 8 7, l 00 ( l , 500) (400) ( l , 000) 10 9,700 ( l , 000) l 00 (500) l 2 12 ,800 (300) 800 200 14 16, 700 600 l , 700 l , l 00 16 21 , 700 2,700 3,800 3,200 20 35,700 5, l 00 6,200 5,600

( ) indicates a negative value of savings, i.e., a loss t Based on the economic and technical assumptions listed in Table 3 and the assumption that

rgeneral = 112 relec

rgas = 114 relec (r is the energy inflation rate) 24

Table 3. Economic Assumptions

Variable 3-ton 25-ton AC a Capital cost of solar cooling subsystem $4,500 $22,000 CN Capital cost of non-solar cooling subsystem $1 ,500 $16,000 x Percentage of first year operating costs 0.08 0.05 Cmc-Cm Differential maintenance costs -$50/year 0 63.3xl0b 528 x lOb L Seasonal cooling load kJ/year kJ/vea r celec First year cost of electricity 5¢/kw-hr 5¢/kw-hr f Fraction of cooling load provided by solar 0.75 0.75 ro Inflation rate of operating costs relec relec g g r m Inflation rate of maintenance costs g General inflation rate 6%/year 6%/year y Tax factor for deductible expenses 1 1-t t Effective income tax rate of owner 25 % 25% a Downpayment fraction of first cost 20% 20 % s Investment tax credit 0 10% a Fractional salvage value at end of equipment 1 ife 0 20% d Annual discount rate 8% 12% n Period of economic analysis 20 years 20 years i Annual interest rate of loan 8%/year 8%/year h First year insurance rate on capital cost 0.5% 0.5% m Term of 1oan 20 years 20 years lr 1 rf Inflation rate for auxiliary fuel 4 elec -r4 elec r,. r f Inflation rate for non-solar fuel relec elec p First year property tax rate on solar system 0 0 Length of time for depreciation of solar k equipment (use straiaht line deoreciation) NA 20 years 25

Income tax incentives (presently under consideration) allow for lowering of the necessary inflation rate to 10 or 11 percent, which is in line with current prediction of electric inflation rates. A combination of the tax rebate and low cost loan places the economic breakeven point at an electricity inflation rate of about 8 percent.

CONCLUSIONS The thermodynamic efficiency of solar absorption cooling is very nearly equivalent to that of an electrically-driven, vapor-compression system with a high seasonal COP (on the order of COP= 3.0). In addition, water/lithium bromide absorption units have a history of demonstrated technical feasibility, particularly when integrated with a complete solar heating and cooling system. Economically, solar absorption cooling is marginal, but improves con- siderably with income tax incentives. For an electricity cost inflation rate of 11 percent (one electric power company's estimate [30 ]), the incorporation of a 25 percent tax rebate and the availability of a four percent interest loan on the solar equipment, an initial investment in a solar absorption cooling system of $4,500 would result in a savings of about $1,000 in electricity costs (including the capital cost of the electric cooling equipment) over the period of 20 years. It is noteworthy, however, that the owner of the building receives a return of $1,125 the first year and thereafter operates at virtually a breakeven point.

It must be emphasized that these conclusions are on a "best guess" basis. While the technical conclusions which demonstrate the relative thermodynamic efficiencies of conventional vapor-compression units and solar absorption cooling are not heavily dependent upon the quality of the assumptions made, the economic feasibility is critically dependent upon the assumed economic values. And such economic parameters allow only tentative and contingent conclusions. 26

ACKNOWLEDGEMENTS

Research supported in part by the Solar Heating and Cooling Branch, Conservation and Solar Applications, U.S. Department of Energy and the Thermal Conversion Branch, Solar Energy Research Institute, Golden, Colorado. Bibliography of References

l. Auh, P.C., "A Survey of Absorption Cooling Technology in Solar Applica- tions". Brookhaven National Laboratory Report No. BNL 50704, July 1977. 2. Uif, G.0.G., "Cooling with Solar Energy". Proceedings of the World Symposium on Applied Solar Energy, Phoenix, Arizona, 1955. 3. Trombe, R. and Foex, M., "Intermittent Ammonia-Water System with Solar Regeneration". Solar Energy Society and Engineering, Vol. l, p. 51, 1957. 4. Williams, D.A., "Cooling Systems Based on Solar Refrigeration". Refrigerating Engineering, Vol. 66, p. 33, 1958.

5. Eisenstadt, M., Flanigan, F. M., and Farber, E.A., "Tests Prove Feasibility of Solar Air Conditioning". Heating, Piping, and Air Conditioning, Vol. 32, p. 120, 1960. 6. Chinnappa, J.C . V. , "Experimental Study of the Intermittent Vapour Absorption Refri geration Cycle Employing the Refrigerant-Absorbent Systems of Arrmonia Water and Ammonia Lithium Nitrate" . Solar Energy, Vol . 5, p. l , 1961 . 7. Chung, R. and Duffie, J.A., "A Study of a Solar Air Conditioner". Mechanical Engineering, Vol. 85, p. 31, 1963.

8. Duffie, J.A. and Sheridan, N.R. , "Lithium Bromide-Water Refrigerators for Solar Operation". Mechanical and Chemical Engineering Trans. Inst. Engrs. Australia, MC-1, p. 79, 1965. 9. Swartman, R.K. and Alwa r d, R. , "Evaluation of an Experimental Inter- mittent Absorption Refrigerator Incorporating the Generator with the Flat-Plat e Collect or" . Proc . Solar Energy Society Annual Meeting, Palo Alto, Cal i fornia, 1968. 10. Swartman, R.K. and Swaminathan , C. , "Solar Powered Refrigeration". Mechanical Eng i neering, p. 22, 1971. 11. Farber, E.A., "Design and Performance of a Compact Solar Refrigeration System". Proc. Solar Energy Society Conf., Melbourne, 1970.

12. Ward , D.S. and Ltif, G.O.G., "Design and Construction of a Residential Solar Heating and Cooling System". Solar Energy, Vol. 17, p. 13, 1975. 13. Ward, D.S., Weiss, T.A . , and Ltif, G.O.G., "Preliminary Performance of CSU Solar House I Heating and Cooling System". Proc. International Solar Energy Society Congress, Los Angeles, Cal i fornia, 1975.

14. Ward, D.S., "Performance of the CSU Solar House I Cooling System". Proc . Second Workshop on Use of Solar Energy for the Cooling of Buildings, Los Angeles, California, p. 48, 1975. ·t 15. Ward, D.S., Smith, C.C . , and Ward, J.C., "Operational Modes of Solar Heating and Cooling Systems". Solar Energy, Vol. 19, 1977. 16. Namkoong, D., "Performance of a LiBr Water Chiller in a Laboratory- Scale, Experimental Solar System Test Loop". Proc. International Solar Energy Society Conf., Winnipeg, Vol. 3, 1976. 17. San Martin, R.L., Diamond, S., Packard, C., Shaw, H., and Stevens, W., 11 A Solar Heated and Cooled Office Buildings". Proc. International Solar Energy Society Conf., Winnipeg, Vol. 3, 1976. 18. San Martin, R.L., LaPlante, D., Packard, C., and Shaw, H., "Twenty Months of Operating Experience with a Solar Heated and Cooled Office Buildings". Proc. 1977 Annual Meeting of Amer. Sec. of International Solar Energy Society, Orlando, Florida, Vol. 1, 1977. 19. Jacobsen, A.S., "Solar Heating and Cooling of Mobile Homes, Test Results". Proc. 19 77 Annual Meeting of Amer. Sec. of International Solar Energy Soc i ety, Orlando, Florida, Vol. 1, 1977.

11 20. Anderson, P., "Progress Report on Solar Cooling at ARKLA • Proc. Second Workshop on the Use of Solar Energy for the Cooling of Buildings, Los Angeles, California, 1975.

21. Simmons, M. and Wahl ig, M., "Ammonia Water Absorption Air Conditioner". Proc. Second Workshop on the Use of Solar Energy for the Cooling of Buildings, Los Angeles, California, 1975 . 22. Dao, K., Simmons, M., Wolgast, R., and Wahlig, M., "Performance of an Air-Cooled Ammonia -Water Absorption Air Conditioner at Low Generator Temperatures". Proc. Internat'l Solar Energy Conf., Winnipeg, Vol. 3, 1976 .

23. Dao, K., Simmons, M., Wolgast, T., and Wahlig, M., "Development of Solar-Driven AITITionia-Water Absorption Air Conditioners and Heat Pumps". ERDA Solar Heating and Cooling Research and Development Contractors' Meeting, Reston, Virginia, 1977. 24. Merr ick, R., "Engineering Design, Construction, and Testing of a Salt- water Absorption Unit Optimized for Use with a Solar Collector Heat Source". ERDA Solar Heating and Cooling Research and Development Contractors' Mee ting, Reston , Virginia, 1977. 25. Newton, A.B., "Working Sessions on Absorption Systems". Proc. Second Workshop on the Use of Solar Energy for the Cooling of Buildings, Los Angeles, California, 1975. 26. Ward, D.S. and LBf, G.O .G., "Des ign, Construction and Testing of a Residential Solar Heating and Cooling System". Report to the Committee on the Challenges of Modern Society (CCMS) Solar Energy Pilot Study, July 1976. 27. San Martin, R.L., Private Communication, 1977. 28. Ward, D.S., Uesak i, T., and Ujf, G.O.G., "Cooling Subsystem Design in CSU So lar House III". Proc. Internat'l Solar Energy Society Conf., Winnipeg, Vol . 3, 19 76. 29. Ward, D.S. and Ward, J.C., "Design Considerations for Residential Solar Heating and Cooling Systems Utilizing Evacuated Tube Solar Collectors". Proc. Amer. Section of the Internat'l Solar Energy Society Conf., Orlando, Florida, 1977.

30. Public Service Company of Colorado, Private Communication, 1978.

31. Bartlett, J.C., "Site-Dependent Factors Affecting the Economic Feasibility of Solar Powered Absorption Cooling". Proc. Amer. Section of the Internat'l Solar Energy Society Conf., Orlando, Florida, 1977.

32. Ishibashi, T., Yazaki Corporation, Private Communication, 1978.

33. Anderson, P., ARKLA Industries, Private Communication, 1977.

34. Kreith, F. and Kreider, J.F., "Preliminary Design and Economic Analysis 11 of Solar-Energy Systems for Heating and Cooling of Buil di ngs • Energy, Vo 1 . 1 , pp. 6 3-76, 19 76.

35. Barley, C.D., "Relative Areas Analysis of Solar Heating Systems Perfor- mance". M.S. Thesis, Colorado State University, 1977.

36. Karaki, S., Ward, D.S., Winn, C.B., and LBf, G.O.G., "Solar Space and Water Heating for Residential Buildings". Training Manual for Design and Installation of Systems, Department of Housing and Urban Develop- ment, Washington, 1977. .. 37. Klein, S.A. and Beckman, W.A., 11 A General Design Method for Closed-Loop Solar Energy Systems". Proc. Amer. Section of the Internat'l Solar Energy Society Conf., Orlando, Florida, 1977.

38. Barley, C.D., Winn, C.B., and Huck, S.E., "Simplified Techniques for Sizing Residential Solar Heating Systems". Proc. of the ISES (U.S. Section) Conf., Orlando, Florida, 1977.

39. Ward, D.S., "Realistic Sizing of Residential Solar Heating and Cooling Systems". Submitted to Solar Energy Journal, April 1978. (Available from Solar Energy Applications Laboratory, Colorado State University, Fort Collins, CO 80523).