Other papers
Note: The original document contains many mathematical formulas that could not be translated to this web page. For critical reading, we suggest downloading the Acrobat file.
ABSTRACT
Temperature boosting of waste heat from spacecraft by means of heat pumps makes it possible, under some conditions, to achieve considerable savings in radiator area and mass. This study considers several possibilities for employing work-actuated and heat-actuated heat pumps (WAHP and HAHP, respectively) for this purpose. In the former case, the spacecraft power source is required to generate extra power to operate the heat pump; in the latter, use is made of the heat rejected from the power source to energize the heat pump. The mass and area savings are calculated for a range of operating parameters, including the temperatures of the waste heat from the power source and payload and that of the effective heat sink. The dimensionless parameters governing the behavior are determined. It is shown that for given operating conditions, a proper choice of the temperature boost and the associated heat pump coefficient of performance leads to an optimum in radiator area and system mass savings. A detailed cycle calculation is presented for an absorption heat pump - a particularly promising HAHP with few moving parts. Design considerations are given for space-based WAHP and HAHP systems.
1.0 INTRODUCTION
The trend toward increasing spacecraft power has created a growing need for effective methods to reject the waste heat produced by power generating and power-consuming equipment aboard. In particular, the current SDI and Space Station plans for lo-100 kW of steady electric power represent an order of magnitude increase in waste heat compared to present and past missions. Conventional radiators for such power rates become very large, and revolutionary advances in thermal management may be required [l]. One promising option for enhancing heat rejection is the use of heat pumps to boost the radiator temperature, thereby reducing its area and mass.
Figure 1 describes schematically a typical spacecraft thermal management system [2]. The power source, equipped with a radiator for rejecting its own waste heat at the required temperature, produces electric power W distributed to one or several 7 payloads (e.g., electronic equipment, sensors, etc.). Each of the latter is equipped with a heat acquisition unit (also known as "cold plate") which keeps it at a temperature low enough,for its required operation. A heat transport loop, (also known as "thermal bus") transfers the waste heat from the payloads to a common radiator for rejection. The transport loop may employ a forced-circulation fluid or a heat pipe. This method of collective payload heat rejection requires that the transport loop be at a temperature lower than any of the cold plates it serves. The payload radiator must accordingly be kept at as low or lower temperature and therefore requires a large area (and mass) per unit heat rejected, generally more so than the power radiator. Therefore, payload waste heat radiators tend to be the heaviest and bulkiest components in a thermal management system. Since the radiated heat rate is approximately proportional to the fourth power of the absolute temperature, an increase in the radiator temperature would help reduce its size. The key idea in this work is to show that such an increase may be effectively achieved by a heat pump which may be energized by electric power or by waste heat.
One of the first studies on the use of heat pumps in space was performed by Dexter and Haskin. They considered the option of an electrically-driven vapor compression heat pump, using a compressor to transfer a working fluid from an evaporator at the low source temperature to a condenser at a higher (radiator) temperature. They carried out a thorough calculation of the possible mass savings with the heat pump system compared to a pumped fluid loop, for electric power ranging from 0 to 100 kWe. Assumptions were made, based on available experience, of the power mass penalty, compressor and motor mass, cold plate and fluid loop lines and their variation with the power rating. A sensitivity analysis considered the effects of varying the source and sink temperatures, the coefficient of performance and the extra mass of the power plant. The analysis showed the heat pump system to have lower mass and lower radiator area in all cases, with significant savings under some conditions.
The electrically-driven compressor heat pump is only one of several heat pump options and not necessarily the best one. Its use requires extra electric 'power from the spacecraft's power source; there are some practical and reliability problems regarding the operation of compressors in space, on which there is little or no experience. A study of wider scope was carried out by Drolen [4], who considered both work-actuated and heat-actuated heat pumps (WAHP and HAHP. respectively). as well as their hybrids. Recognizing that in most cases waste heat is available from both the power source and the payload, with the former at a considerably higher temperature, one can utilize the former as a source of energy for a HAHP to boost the temperature of the latter, thus achieving overall radiator savings. In evaluating different -heat pump options, Drolen [4] calculated the available (or affordable) heat pump mass based on radiator area savings and data. on specific mass of the radiator and power source. Using this approach, rather than calculating the actual mass savings, made it possible to perform the evaluation in the absence of reliable information on the mass of heat pumps designed for space applications. Drolen [4] carried out an extensive study of the effect of different parameters on the available heat pump mass, including radiator hardening level and geometry, payload rejection temperature, effective space temperature and heat pump efficiency. His optimization was, however, limited. No attempt was made, for example, to find the optimum heat pump COP and temperature boost which would yield maximum radiator savings under given operating conditions. As will be shown later, in the case of a HAHP there are two choice intermediate temperatures (between those of the power and payload waste heats) which affect the overall performance. Also, some of the savings calculated in [4] came from the assumption of different types of radiators for the power and payload, where the heat pump's role amounted basically to the transfer of heat from a heavy to a light radiator. This would result in radiator mass savings even with no heat pumping at all.
Kerrebrock [5] considered the payload radiator area savings achievable by a heat engine-driven heat pump, which is essentially a HAHP. He did calculate an optimum value for radiator temperatures, intermediate between the power and payload rejection temperatures, at which the heat engine and heat pump dispose of their waste heat. Kerrebrock [5] assumed a zero sink temperature which represents radiation into deep space, but is unrealistic for low earth orbits. As a result, he found significant area savings possible only for large temperature ratios of the power to payload waste heats. 'Merrigan and Reid f61 calculated the mass savings for a similar heat &iine/heat pump combination with a non-zero sink temperature, using a common radiator temperature for the heat engine and the heat pump. Neither [5,6] considered the associated savings in power radiator area and mass resulting from the diversion of part of this radiator heat toward the heat engine.
In this study, the potential radiator area and mass savinos are calculated using both WAHP and HAHP. without the limitations which hid been outlined for the earlier studies. It is shown that reasonably simple, yet descriptive equations may be derived without too many simplifying assumptions. A non-dimensional approach is taken where the characteristic dimensionless parameters governing the behavior are developed. Optimum conditions for operation are calculated.
The main purpose of this study has been to consider some practical approaches to the design of space-based heat pumps. The limitations on reliability of different rotating and reciprocating machinery are well known. It is shown that an absorption heat pump may be used which combines several favorable features, including high reliability, low mass, few moving parts and good performance.
2.0 AREA AND MASS SAVINGS WITH WORK-ACTUATED HEAT PUMP (WAHP)
The first option we investigate for boosting the temperature of the payload radiator is the use of a work-actuated heat pump of the type most commonly used in terrestrial applications. It employs a vapor compression cycle where the low-temperature heat input causes the evaporation of a working fluid in a low-pressure evaporator; an electrically driven compressor transfers the vapor to a higher-pressure condenser, where it gives up its heat of condensation at a correspondingly higher temperature; the condensate is returned to the evaporator to complete the cycle [7]. The vapor compression cycle had been considered by Dexter and Haskin [3]; we will derive expressions for the area and mass savings and compare them, on the same basis, with a case using a heat-actuated heat pump.
Figure 2 describes schematically a WAHP incorporated in the thermal management system of the spacecraft. The individual sub-units of the system have been numbered consistently with Figure I, the base case with no heat pump. Temperatures and heat quantities pertaining to the present case are distinguished from those of the base case by a prime. The power loads, cold plates and heat transport loop shown separately in Figure 1 have been combined into a single box in Figure 2. The payload is assumed to be maintained at the same temperature (T3' = T3) and to consume the same amount of electric power, W, in both cases. It is further assumed that this entire power is converted into waste heat, so that Q3' = Q3 = w. the temperature of the power radiator is also the same for both cases (T2' = T2). Finally, it is assumed that the power and payload radiators are both of the same type, with the maximum efficiency and minimum specific mass allowed by the mission. Assuming otherwise a hardened, one-sided and therefore heavy radiator for one and an unhardened, two sided and light radiator for the other, as in 1419 would result in the obvious conclusion that considerable mass saving may be achieved by transferring all the heat from the former to the latter.
2.1 Area Saving
For the case with no heat pump, the payload equals that of its radiator, Q4 = Q3 = W. ' heat pump added, the payload radiator must reject extra heat due to the electric power input to the heat pump. The coefficient of performance (COP) of the latter is defined as the ratio of the heat input to the work input. Denoting the coefficient of performance by 0, it is easy to show that for a heat input Q'3 = W the required work input is W/o. Thus, the heat rejected by the payload radiator (Q4') as well as the total power required from the power source are both equal to W (1 + /3)/fl. The power source efficiency is defined as the ratio of the electric power output to the input, and the heat rejected from it is the difference, between the two.
3.0 AREA AND MASS SAVINGS WITH HEAT-Actuated HEAT PUMP (HAHP)
The heat-actuated heat pump is an important alternative to the WAHP discussed in the previous section. Figure 5 describes schematically a HAHP incorporated in the thermal management system of a spacecraft. The same sub-unit numbering system has been used consistently with Figure 1, the base case with no heat pump, and with Figure 2, the WAHP case. Heat quantities and temperatures pertaining to this case have been designated by a double prime. Again, the payload is maintained at the base case temperature (T3" = T,J and consumes the base amount of electric power, W. All the power is converted into waste heat (Q3" = U3 = W) and the power radiator temperature remains unchanged (T2" = T ).
The HAHP is described conceptually as a work-actuated heat pump driven by a heat engine, which uses part of the power source waste heat for its operation. We say 'conceptually' since there are several options for building a HAHP, including some in which no mechanical energy is being transmitted from ;I "driving" to a "driven' part of the system. One such option is based on the absorption cycle, which will be discus-:d in detail later. The feature of no moving parts (except for small, auxiliary equipment) makes the absorption heat pump highly The HAHP has two radiators for heat rejection, which may operate at two different temperatures, although in many cases they are united and made to work at the same temperature.
The use of a HAH? at the optimum operating point is subject to a limitation on the availability of waste heat from the power source. It is easy to see that a solution of (24), for any ee, is T4e' = T The optimization drives the heat engine radiator to operate at the highest possible temperature, equal to that of its source; this makes the heat engine very inefficient and requires very large input heat in order to generate the motive power for the heat pump. However, the heat available for this purpose from the power source is limited to W (1 - e)/e, which can generate an amount of work eIW (1 - e)/e. Matching this quantity with the heat pump requirement (W/PI)
Figure 3b describes the area savings under optimum conditions as a function of the dimensionless space sink to payload temperature ratio, for different values of the power to payload waste heat temperatures.
3.2 Mass Saving
Unlike the case of the WAHP, there is no power mass penalty associated with the use of a HAHP; no extra electric power is required since the heat pump makes use of the available waste heat. The system mass saving in the case of a HAHP for the system excluding the HAHP is therefore directly proportional to the area saving and given by:
4.0 THE ABSORPTION HEAT PUMP (AHP)
A particular type of HAHP holding considerable promise for space applications is the absorption heat pump. Its major advantage is in the almost total absence of moving parts, which provides for high reliability. Unlike the heat engine-driven compression cycle discussed in the previous section and there is no mechanical energy involved. The cycle.employs two working fluids -- a refrigerant and an absorbent - instead of the single fluid used in the vapor compression cycle. The temperature lift occurring when one substance is absorbed in the other provides the heat pumping action without compression.
Absorption systems-may be built with several working fluid combinations and in a variety of cycle configurations; they may be staged to provide higher COP or higher temperature lifts [9]. A complete discussion of the different options is outside the scope of this article. A description of the basic, single-stage cycle may be found in [10]. Unlike the heat engine driven vapor compression system, the absorption system may not be assumed to operate at a fixed portion of the Carnot efficiency for the same operating temperatures; it has a different operating curve, approaching Carnot for some temperatures and quite different from it for others [ll]. To calculate its performance, a complete cycle analysis is required for the given working fluids and operating conditions.
In order to evaluate the capability of the AHP, a specific application was selected in the present study - the rejection of waste heat from steady state, 100 kWe electronic equipment aboard a spacecraft in low earth orbit (LEO), to be kept at 3OO°K. A conceptual design of an absorption heat pump for this purpose was carried out and the resulting radiator area and mass savings calculated. The system prior to adding the heat pump is as described schematically by Figure 1, with Q3 = W = 100 kW, T3 = T4 = 300°K, T = ZSO'K. The system with the AHP added is described in Figure 6. The AHP has six main components -absorber, condenser, evaporator, generator, recuperator and precooler - with the first two serving for heat rejection. An absorbent/refrigerant combination commonly used for absorption cooling in terrestrial applications - lithium bromide/water has been selected. Other working fluids which may be more suitable for space application are discussed later.
Liquid refrigerant (water) enters the evaporator at state 1 and evaporates at 300°K, thereby removing heat from the payload. The vapor at state 2 passes through the precooler and enters the absorber where it is absorbed by a highly hygroscopic lithium bromide/water solution, entering concentrated at state 3, and leaving more dilute at state 4. The solution must be cooled during the process to reject the heat of absorption, in order to keep a low vapor pressure and high hygroscopic capability. The rejection temperature is higher than that of the evaporator by the temperature lift of the working fluid pair. The dilute solution at state 4 passes through the recuperative heat exchanger and enters at state-5 into the generator, where waste heat from the Dower source at 533k is added to it. This heat causes the desorption of water from the solution, reconcentrating it to state 6. The reconcentrated solution returns to the absorber via the recuperator. The desorbed water vapor at state 7 condenses to state 8 in the condenser, where the heat of condensation is rejected to space. The condensate exchanges heat with the vapor at state 1 'before returning through the expansion valve to the evaporator. A small pumping device is required to transfer the dilute solution from the low pressure absorber to the higher pressure generator.
Cycle calculations were performed to define the operating temperatures and concentrations at the different state points, and the heat quantities in the sub-units. A modular computer program was employed, capable of simulating absorption systems in different cycle configurations and with different working fluids [lZ]. The rejection temperatures in the absorber and condenser may be optimized, as explained in Section 3. In the present case study with the LiBr/H 0 working fluid, crystallization sets a limit on the $.lgh solution concentration which was selected to be 67%. The results of the cycle calculations are given in Table 2. Note that in the absorber the solution temperature varies, and an average value is taken. Thus, 233 m of radiator area and a corresponding mass of 1165 kg have been saved. The available mass for the generator, evaporator, recuperator and the auxiliary parts of the AHP is 11.65 kg/kW.
The above example demonstrates the possibility ' for reducing radiator area and mass. using an absorption heat pump. The values calculated for the savings provide only an indication of what may be achieved in practice. Considerably greater potential exists with higher power radiator temperatures, lower payload temperatures and higher sink temperatures.
For given temperatures, operating conditions of the AHP such as the solution circulation rate may be optimized. A system design suitable for space applications must be developed, which would differ in some respects from the common AHP technology. Some practical considerations are discussed next.
5.0 DESIGN CONSIDERATIONS
A great deal of experience is available on the design and operation of heat pumps, both WAHP and HAHP, for terrestrial applications. An extensive literature search has revealed no information on specific designs for space applications. The primary concern for earth systems is usually cost and often physical dimensions, where the heat pump has to be shipped a certain way or fit into a designated space. The two predominant considerations guiding the design of space-based heat pumps are mass and reliability. Volume, once deployed, is of no major concern. Thus, these particular issues must be addressed.
The main problem in the design of a space-based WAHP is the compressor. Usually, rotating or reciprocating machinery operating continuously at many cycles per second poses serious reliability problems. On the other hand, compressors for low pressure, high volume vapors such as water, which are problematic on earth could have an advantage in space. A study of an inflatable/retractable radiator [13] has come up with a design of a system serving the functions of a compressor and a radiator/condenser at the same time, thus comprising the major part of a WAHP. The above system was developed for a different application - handling burst power. It consists of an inflatable fabric bag, which is extended and filled with steam produced by the waste heat released during the short power burst. As the steam condenses, the bag is retracted and the remaining vapor compressed to maintain a fixed saturation pressure. A special drive mechanism is available for this purpose as well as a sponge system for recovering the liquid condensate from the walls of the bag. A slightly modified version of this inflatable/retractable radiator is capable of serving as a compressor for steam. combined with a condenser. The compression process is gentle in comparison with conventional technology; it is also performed under isothermal, rather than adiabatic conditions, thereby eliminating the possible need for intercooling and saving work.
Similar problems related to rotating machinery are involved in HAHP's with heat engine-driven compressors. The absorption heat pump therefore seems more suitable for space applications, as it avoids most of the problems due to moving parts. The only device involving work - the solution pump - is very small compared to other components, as evident from the example in Table 2. Lightweight fabric radiators of the type mentioned earlier may be of advantage in an AHP design. The two sub units involving heat rejection - the absorber and condenser - may be made out of coated fabric. Under the inflatable radiator study [13] a sponge system was proposed to recover liquid from the walls of the condenser. A similar system may be employed to deliver and recover absorbent solution to and from the walls of a fabric absorber.
Other AHP considerations are associated with working fluids. For terrestrial applications, a refrigerant with sub-atmospheric vapor pressure is usually not desirable, due to the risk of leaks and the need for continuous purging of non-condensables. In space, non-condensables do not constitute a problem and a low vapor pressure refrigerant would be desirable. There is a need for an absorbent/ refrigerant pair with high temperature lift. Water, with its high latent heat is a promising refrigerant with a number of possible absorbents.
6.0 CONCLUSIONS
The radiator area and system mass savings achievable through the use of heat pumps have been calculated for WAHP and HAHP. For given power and payload waste heat temperatures and a riven space sink temperature, there is generally a- degree of freedom in the choice of the temperature boost delivered by the heat pump, traded off against the coefficient of performance. An optimum condition which maximizes the savings may therefore be found. The results have been expressed in dimensionless form and the governing parameters identified. The fraction of radiator area saved under optimum conditions is a function of two temperature ratios: To/T3 (space sink to payload) and T2/T3 (power source to payload waste heat). The available (or affordable) heat pump mass per unit power, normalized with respect to radiator mass per unit power, is also a function of these two temperature ratios, but in the case of WAHP it depends additionally on - the dimensionless power mass penalty.
Cases with ideal (Carnot) and non-ideal heat pumps have been analyzed. Both the area and mass savings under optimal conditions increase with To/T3 and with T2/T3. Best results are achieved where the payload temperature is close to that of the available sink. For the same operating temperatures, the fraction of area saved is about the same for the WAHP and HAHP, with some advantage to the former. The mass saving is considerably in favor of the HAHP, due to the power mass penalty associated with the WAHP. The absorption heat pump, a particular form of the HAHP, holds considerable promise for space applications due to few moving parts and possible lightweight design. Further development of the AHP
is needed with the specific space application in mind, particularly in working fluids and system design.
REFERENCES