Purdue University Purdue e-pubs Internationa Refrigeration and Air Conditioning Conference Schoo of Mechanica Engineering 2016 Investigation of Air-Cooed Condensers for Ammonia-Water Absorption Chiers Subhrajit Chakraborty Georgia Institute of Technoogy, United States of America, schakraborty38@gatech.edu Victor C. Aieo Georgia Institute of Technoogy, United States of America, victor.aieo3@gmai.com Srinivas Garimea Georgia Institute of Technoogy, United States of America, sgarimea@gatech.edu Foow this and additiona works at: http://docs.ib.purdue.edu/iracc Chakraborty, Subhrajit; Aieo, Victor C.; and Garimea, Srinivas, "Investigation of Air-Cooed Condensers for Ammonia-Water Absorption Chiers" (2016). Internationa Refrigeration and Air Conditioning Conference. Paper 1676. http://docs.ib.purdue.edu/iracc/1676 This document has been made avaiabe through Purdue e-pubs, a service of the Purdue University Libraries. Pease contact epubs@purdue.edu for additiona information. Compete proceedings may be acquired in print and on CD-ROM directy from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/ Herrick/Events/orderit.htm
2270, Page 1 Investigation of air-cooed condensers for ammonia-water absorption chiers Subhrajit CHAKRABORTY, Victor C. AIELLO, Srinivas GARIMELLA * Sustainabe Therma Systems Laboratory George W. Woodruff Schoo of Mechanica Engineering Georgia Institute of Technoogy Atanta, GA 30032 Phone: (404) 894-7479; sgarimea@gatech.edu * Corresponding Author ABSTRACT This paper presents the resuts from experimenta and anaytica investigation of air-cooed condensers for use in sma-scae, direct-fired ammonia-water absorption chiers. Two nove muti-pass tube-array condensers, compatibe with ammonia-water refrigerant mixture, are designed, fabricated and evauated. The condensers are buit for an absorption system that deivers 2.71 kw cooing capacity in severe ambient temperatures as high as 52 o C with a design condenser heat rejection rate of 2.51 kw into ambient air. A singe-pressure ammonia-water test faciity is constructed and used in conjunction with a temperature- and humidity-controed air-handing unit to evauate the condensers at design and off-design operating conditions. Condenser performance is recorded over a range of air temperatures, refrigerant temperatures, air voumetric fow rates, and refrigerant mass fow rates. The maximum ambient temperatures and minimum air fow rates required to reject design heat duty are identified for both the condensers. A heat transfer mode is deveoped for the first condenser, which is determined to be air-side imited. The mode predictions are compared with the measured heat transfer rates at various ambient conditions and the deviations expained. Resuts from this investigation guide the deveopment of air-couped zeotropic mixture condensers for compact absorption heat pumps. 1. INTRODUCTION Residentia and commercia space conditioning is typicay provided by eectricay driven vapor-compression systems, which ead to high oads for eectric utiities during the hottest hours of the day. In addition, these systems have reied on the use of synthetic refrigerants that contribute to goba cimate change. Vapor absorption based heating, ventiation, and air-conditioning (HVAC) systems can utiize ow-grade or waste heat streams to provide heating and cooing. These systems provide an aternative to conventiona vapor compression systems, reducing the peak demand for eectricity (Zieger, 1999). These systems aso use environmentay benign working pairs, such as ammonia-water, with zero potentia for goba warming and ozone depetion (Lorentzen, 1995). Ammonia-water has been used as the fuid pair in absorption systems since the ate 1800s, and is receiving attention due to its environmentay benign aspects (Herod et a., 1996). The ow freezing point of ammonia ( 77 C) enabes ow system operating temperatures with appications in both heating and cooing (Srikhirin et a., 2001). In addition, the high operating pressure of ammonia-water systems aows for compact component design that is essentia for mobie and sma-capacity units. Mini- or microchanne heat exchangers have been shown to effectivey transfer heat in a compact voume due to high heat transfer coefficients and high surface area-to-voume ratios. Severa researchers (Ferreira et a. (1984); Meacham and Garimea (2004); Fernández-Seara et a. (2005)) have utiized these compact heat exchangers for the condenser or absorber by indirect couping to the ambient with an intermediate fuid oop. Whie this couping oop reduces ammonia-water fuid inventory, the eectrica input for pumping the intermediate fuid, and increased temperature difference between the ambient and the refrigerant coud ead to ower system coefficients of performance (COPs). The use of compact direct air-couped components that simiary reduce fuid
2270, Page 2 inventory woud eiminate the intermediate oop, reduce system size, and improve system COP. This is particuary important for the instaation and operation of these heat pumps in forward operating miitary bases at T ambient as high as 52 o C. In crossfow air-couped heat exchangers, mini and microscae geometries are impemented through the use of extruded rectanguar tubes. Garimea et a. (1997) deveoped a mode to compare the performance of fat-tube muti ouver fin heat exchangers with that of conventiona round tube heat exchangers with various fin geometries. It was concuded that the use of fat tubes and muti-ouvered fins enabed fexibiity in heat exchanger design whie providing substantia reduction in size and weight for the same heat duty. Ammonia-water soution is incompatibe with copper and auminum, which are commony used to form fat tubes. Materias that are compatibe, such as carbon or stainess stee, cannot readiy be extruded into fat tube geometries; therefore, other geometries must be expored for heat exchanger improvement. Garimea and Coeman (1998) conducted a modeing anaysis for air-cooed condensation of ammonia-water vapor focusing on round-tube heat exchangers of tube diameter 9.525 10-3 m with four commony used fin-types: fat, wavy, ouvered, and annuar. The materias compatibiity issue with ammoniawater was addressed by seecting carbon stee for the tube materia whie auminum was used for the fins. Variations in mixture concentration, saturation pressure, and air temperature were investigated to predict the performance of such condensers over a wide range of operating conditions. The study concuded that a round-tube wavy-fin heat exchanger provided the argest heat transfer within the aowabe pressure drop and size constraints. There is a need for further improvement in compact air-couped heat exchangers compatibe with ammonia-water for sma-scae energy recovery. The present study extends the anaysis of air-cooed heat exchangers to nove muti-pass tube-array designs with minichanne geometries. Experimenta investigations of air-couped condenser performance, which are imited in the iterature, are performed and compared with predictions from modeing and anaysis. 2. DESIGN AND FABRICATION Two nove air-couped condensers were designed to yied better heat transfer performance over conventiona designs. The condensers are buit to work in an absorption system of 2.71 kw cooing capacity with a design condenser heat rejection rate of 2.51 kw into the ambient air. A muti-pass horizonta tube-array design is adopted for the condensers consisting of mutipe tube banks joined in a singe header. Both the condensers are 0.406 m high by 0.508 m wide and consist of a tota of 120 tubes with an outer diameter of 3.175 10-3 m and wa thickness of 5.080 10-4 m. The first version of the air-couped condenser (Condenser A) is shown in Figure 1. The refrigerant vapor enters at the top of the condenser, spits into the first bank of 48 staggered parae tubes, and makes four passes in cross counter-fow through the air. The partiay condensed refrigerant then recombines in the header before spitting into the second tube bank of 36 parae tubes and makes another four passes in cross co-fow through the air. This pattern continues through the subsequent two tube banks before the refrigerant exits as a subcooed iquid at the bottom of the condenser. The sma tube diameters resut in high heat transfer coefficients and reativey high surface area on the air side. The number of parae tubes in each bank decreases as the refrigerant condenses to account for the decrease in specific voume. The condensers are designed in this manner to produce a reativey constant homogeneous veocity of the refrigerant through the condenser, which resuts in fow Figure 1: Schematic of refrigerant fow in tube-array mutipass condenser (Condenser A)
2270, Page 3 regimes that are optima for heat transfer. The second version, Condenser B, is made to improve performance further and is shown in Figure 3. During testing of Condenser A, it was observed that the pressure drop on both the air side and refrigerant side is ower than the budgeted vaue. Hence, pain fins are instaed to a density of 394 fins per meter, which resuted in a four-fod increase of air-side heat transfer area. The number of tube banks for this condenser is aso increased from four to six whie sti maintaining the same tota number of tubes as Condenser A. This decreases the number of parae tubes per tube bank, shown in Figure 2, and therefore increases the tube-side refrigerant mass fux. Higher mass fux yieds better heat transfer coefficients and aso typicay reduces madistribution of iquid and vapor in the headers. The headers, tubes, and end pates for both condensers are a made of 304 stainess stee. The cavities and tube hoes in the header were machined using a CNC mi. The tubes were inserted 3.175 10-3 m into the header and hed in Figure 2: Modification of header design from Condenser A pace by mid-pane and end pates. In the case of (eft) to Condenser B (right) Condenser B, the aser-cut fins were inserted over the tubes prior to insertion into the header. Combike spacers were instaed to propery distribute the fins and bock air fow in between the tube banks. A nicke vacuum brazing process was used to sea the tubes into the header. Nicke brazing aoy was aso appied to the joints between the tubes and fins of Condenser B to provide essentia tube-fin contact. Both condensers were pressure tested to 3,000 kpa to ensure a successfu braze sea between the condenser components. Figure 3: Modified tube-array muti-pass condenser (Condenser B) with fins and spacer (zoomed) 16th Internationa Refrigeration and Air Conditioning Conference at Purdue, Juy 11-14, 2016
2270, Page 4 3. EXPERIMENTAL APPROACH The two condensers are evauated in a temperature-controed air-handing unit for a variety of air inet temperatures and fow rates, and refrigerant inet temperatures and fow rates. A singe pressure ammonia-water test faciity is constructed to directy contro the condenser inet properties; the diagram of the test oop with state points is shown in Figure 4. A separate pressurized hot water oop is used to boi the refrigerant in the boier. The two-phase soution fows to the separator, where saturated vapor separates from the iquid and exits at state point 5. A sight gass is used to verify that no iquid is fowing into the condenser. The subcooed soution fowing out of the condenser mixes with the diute soution from the separator and enters the subcooer. The subcooer coos the soution further to avoid cavitation in the gear pump. The subcooed refrigerant is pumped into the boier, competing the oop. The air handing unit is used to condition the air entering the condenser to the desired temperature and voumetric fow rate. For air temperature measurements, 4 4 thermocoupe arrays are instaed immediatey upstream and downstream of the condenser. To prevent the disruption of air fow by the temperature measurements, 0.254-mm thermocoupe wire is used in the thermocoupe arrays. These sodered thermocoupe beads are hed in pace by four structura 0.51-mm stainess Tabe 1: Test matrix for evauation of condensers stee wires, stretched verticay across the face of the condenser. Test Number Variabe Unit The test matrix used to evauate this 1 2 3 4 5 condenser in the air-handing unit is shown in Tabe 1. A heat pump cyce mode T air,in C 51.7 35 52 51.7 51.7 deveoped by Forinash (2015) is used to determine the reevant vaues in the test T ref,in C 65.13 56.98 65.15 65.13 65.13 matrix. Test number [1] in the matrix represents the design conditions for the V air,in m condenser in the heat pump system under 3 s -1 0.35 0.35 0.35 0.33 0.46 consideration, but with a modified refrigerant inet temperature of 65.13 C. Test numbers [2] and [3] represent the upper and ower bounds of air inet temperature variation, which simuate the expected severe ambient operating temperature range for this system. The refrigerant inet temperature of the condenser is aso changed with the ambient air temperature based on the cyce simuation resuts. This is done to correate component eve testing with actua operating conditions of the heat pump. Test numbers [4] and [5] represent the ower and upper imits of voumetric fow rate of air controed by fans in the absorption system. Figure 4: Ammonia-water test faciity schematic
2270, Page 5 4. MODELING APPROACH A segmented heat transfer mode is deveoped on the Engineering Equation Sover (EES) software patform (Kein, 2014) to predict the performance of Condenser A. The modeing approach adopted here divides the heat exchanger into severa segments of equa engths to account for the variation in heat transfer processes as condensation proceeds. As ammonia-water is a binary mixture, three independent properties are required to determine the state point after every segment in the condenser. The ammonia concentration (X refr ) of the mixture is constant for a segments and is used as one of the independent properties. However, the ammonia concentration in the vapor and the iquid phases changes aong the ength of the condenser. Initiay, the iquid phase is at ow ammonia concentration and consists primariy of water. As the condensation process continues, the ammonia fraction in the vapor condenses, thereby increasing the concentration of the iquid phase. The other two independent properties are the enthapy (h refr ) and pressure (P refr ) of the soution, which are updated after every segment based on the pressure drop and heat transfer correations. We mixed soution conditions are assumed such that therma equiibrium exists between the vapor and the iquid phases in every segment of the condenser. The mode aso assumes a uniform temperature profie in the inet air across the face of the condenser. A therma resistance network is used to cacuate the heat transfer in each segment of the mode. The therma resistance between the air and the refrigerant is composed of interna therma resistance (with vapor and condensate fim components), conductive resistance (inner to outer tube wa), and externa resistance (outer tube wa to buk air). The tota therma resistance for a segment is cacuated using Equation (1). 1 1 n D D 1 UA h D L k L h A o i R tota R in R cond R out refr i sg 2 tube (1) sg air o,sg Inet refrigerant- and air- side properties are specified for the segment and the outet properties are cacuated using the Effectiveness-NTU method. For a particuar segment, the refrigerant inside the tube is mixed whereas the air is unmixed. Furthermore, the heat capacity rate of the refrigerant is greater than air, suggesting the use of Equation (2) for the cacuation of (Incropera et a., 2011). 1 1 exp Cr 1 exp Cr( UA/ Cmin ) (2) Finay, the heat transfer rate from the segment is cacuated using Equation (3). Subsequenty, the outet temperatures of the refrigerant and air are cacuated using the heat transfer rate for the segment. q ( q ) C T T sg max min refr,in air,in (3) Unike condensation of a pure fuid, in this case there is a drop in refrigerant temperature in the saturated region. Due to high ammonia concentration (> 99%) in the vapor entering the condenser, this temperature gide occurs near the inet of the condenser where the water fraction is reativey higher (Garimea and Coeman, 1998). Thus the heat capacity ratio ( C r ) approaches zero after the initia preferentia condensation of water near the inet of the condenser. 4.1. Air side For air-couped heat exchangers, the average heat transfer coefficient on the air side is typicay constant across a the passes. The correation by Žukauskas (1972), shown in Equation (4), is used here for modeing the average heat transfer coefficient of air fowing around the tube bank. The parameters C and m in this equation are functions of the maximum air Reynods number. Maximum Reynods number is cacuated from the maximum air veocity in the tube bank. Maximum air veocity in the tube bank is determined to be occurring on the A 1 pane (shown in Figure 5) and cacuated using Equation (5) where S T is the transverse Figure 5: Staggered tube bank (Incropera et a., 2011)
2270, Page 6 spacing of the tubes. As the condensers designed in this study consist of eight tube rows in staggered arrangement, the Nusset number shoud be corrected by a factor as shown in equation (6) (Žukauskas, 1972). The average heat transfer coefficient is from this Nusset number. The properties used in this correation are cacuated at the arithmetic mean of the inet and outet air temperatures except for Pr s, which is cacuated at the tube surface temperature. Nu C Re Pr Pr m 0.36 D( Rows20) D,max Prs u s u T air,max air, face st Do 0.25 (4) (5) Nu C Nu (6) DRows ( 8) 2 DRows ( 20) It is assumed in the mode that the air after every tube row gets mixed we before passing over the next tube row. The air temperature at the inet of a particuar segment is the outet air temperature of the segment in the previous tube row. This updated air temperature profie in the tube bank aows for accurate modeing of heat transfer from the refrigerant to the air. 4.2. Tube side The vapor entering the condenser is spit into parae tubes and goes through four passes in each tube before returning to the header on the other side. Thus the mass fux of refrigerant fowing through a segment depends on the number of parae tubes, which varies from one tube bank to another. The prevaiing two-phase fow regime inside the segment is important in determining the appropriate correation for tube-side heat transfer and pressure drop. Whie there are severa studies in the iterature on two-phase fow regimes, two studies that focused on fow regimes during condensation in horizonta tubes (Coeman and Garimea, 2003; E Haja et a., 2003) are used to estabish that stratified two-phase fow exists in the tubes throughout the condenser. This gravity driven stratification of the vapor and iquid phases is due to the ow mass fow fux of refrigerant typica of such a sma-scae absorption system. Among the studies on heat transfer in stratified fows, the comprehensive correation by Dobson and Chato (1998) for horizonta tubes takes into account heat transfer through fim condensation in the upper part of the tube and forced convection in the iquid poo at the bottom of the tube. The correation for stratified-wavy fow is shown in Equation (7). Nu D 0.12 0.25 v Ga Pr 1 0.58 tt Ja 0.23Re Nu 11.11X forced (7) Here, Pr and Ja represent iquid Prandt number and Jacob number, and their ratio is given by Equation (8). This correation requires the temperature of the inside surface of the tube ( T tube, in ), which is cacuated iterativey. Pr Cp / k h fg Ja Cp ( T T )/ h k ( T T ) refr tube, in fg refr tube, in (8) Here is the ange subtended from the top of the tube to the iquid eve at the bottom and is cacuated using the correation of Jaster and Kosky (1976). The Nusset number for forced convection at the bottom of the tube is given by Equation (9). Nu 0.0195Re Pr (9) 0.8 0.4 forced 2 4.172 5.48Fr 1.564Fr 1.376 (10) X 1.7730.169Fr tt
2270, Page 7 Depending on the ambient conditions, there is subcooing occurring in the condenser. The heat transfer coefficient for subcooing is usuay ower than that for condensation. The Churchi (1977) correation for singe-phase fow is used for segments in the subcooed region of the condenser. 5. RESULTS AND DISCUSSION 5.1. Data Anaysis The condensers are experimentay evauated over a range of conditions, with each data point taken after ensuring steady state operation. The data are anayzed on the Engineering Equation Sover (EES) (Kein, 2014) patform. For data anaysis, the refrigerant is assumed to exit the separator as a saturated vapor, which then enters the condenser. Air-side temperatures are cacuated by averaging the 16 thermocoupe measurements upstream and downstream of the condenser. Heat transfer rates are cacuated on both refrigerant and air sides using inet and outet conditions and fow rates. At steady state, the heat transfer rates on both sides are compared and verified to be within 10% of each other. 5.2. Experimenta resuts The measured heat transfer rate is given by Equation (11), where A is the heat transfer area of the condenser, U is the overa heat transfer coefficient, and LMTD is the ogarithmic mean temperature difference. Q U ALMTD (11) Experimenta resuts at various air temperatures aong with varying soution inet temperatures are shown in Figure 6 for both the condensers. The air fow rate and other parameters are kept constant at design vaues during this variation of air inet temperature to the condenser. (Improvements were made to the test faciity between the evauations of the two condensers that resuted in ower uncertainty vaues in test data point of Condenser B.) The rate of heat rejection from the condensers graduay decreases as the air temperature increases. This is due to a decrease in temperature difference between the two fuid streams, reducing the LMTD of the heat exchanger. As shown in Figure 6, heat transfer rate for Condenser A drops beow design (2.51 kw) at about 48 C; therefore, the condenser woud perform to capacity in a packaged system heat pump unit operating at ambient temperatures beow 48 C. Simiary, Condenser B woud be abe to perform better than design capacity at ambient temperatures beow 50.5 C. The air fow rate was aso varied and the resuts are shown in Figure 7 for both condensers. It can be seen from both Figure 6 and Figure 7 that Condenser B performs better than Condenser A at a ambient conditions. This is because of the higher UA in Condenser B, which is due to the addition of fins on the airside and the arger number of tube bank passes on the tube-side. Figure 6: Heat duty of condensers with variation in ambient temperature Athough there is about four-fod increase in air side area for Condenser B, ony about a 3.8% average increase in heat duty is observed compared to Condenser A. This is because
2270, Page 8 the condenser is unabe to reject heat as the air temperature approaches the saturated iquid temperature of the refrigerant, which is 64.8 C for the design air inet of 51.7 C. As the air fow rate is increased, the heat transfer rate increases for both condensers due to two factors. First, higher air veocities increase the airside heat transfer coefficient. Second, higher mass fow rate of air increases the therma capacity rate, eading to smaer increase in air temperature and a arger temperature difference across the fuids in the heat exchanger. This is further verified by the anaytica mode described in the next sub-section. As shown in Figure 7, Condenser B performs better than design (2.51 kw) at air fow rates of 0.40 m 3 s -1 or greater. Figure 7: Heat duty of condensers with variation in air fow rate 5.3. Mode comparison The segmented heat transfer mode discussed above was used to understand the performance of the two condensers. The segmented cacuations proceed aong the ength of the condenser in the direction of refrigerant fow. The therma resistances at design conditions, test number [1] (Tabe 1), are shown in Figure 8 for the entire ength of Condenser A. It can be seen that the externa therma resistance is about an order of magnitude higher than the interna resistance. Therefore, the Condenser A is heaviy imited on the air-side. This substantiates the rationae for the addition of fins to Condenser B. Mode predictions for the heat duty of Condenser A for various air inet temperatures are potted with the experimenta vaues in Figure 6. The mode heat transfer rate predictions foow the trend of the experimenta resuts and decrease with increasing air inet temperatures. However, the measured heat transfer rates are higher than the predicted vaues. One potentia reason for this discrepancy is the technique used for air temperature update in the mode. As the tubes are very cosey packed in the condenser, it is assumed that the air inet temperature to the second row of tubes is the air outet temperature from the first tube row. Due to the staggered arrangement of tubes, there is a possibiity that the second row receives cooer air, which woud increase the heat transfer rate. Furthermore, there coud be heat being ost from the header, which is not accounted for in the mode. Figure 7 compares the predicted and experimenta heat transfer rates for Condenser A at various air fow rates. The mode cosey predicts the experimenta resuts and both the mode Figure 8: Segmenta therma resistances for Condenser A in the direction of refrigerant fow
2270, Page 9 and the experimenta heat transfer rates increase with air fow rate. The mode predicts about 26% increase in heat transfer rates when the air fow rate is increased from 0.35 m 3 s -1 to 0.46 m 3 s -1 (31% increase). This increase in heat transfer rate is attributed to a 14% increase in air-side heat transfer coefficient, with the remainder being due to an increase in the LMTD between the air and refrigerant sides. 6. CONCLUSIONS Two muti-pass tube-array condensers for use in a sma-scae ammonia-water absorption system designed for severe ambient temperature (52 o C) operation were designed and experimentay evauated. Manufacturing chaenges associated with the deveopment of compact stainess stee heat exchangers were addressed during the fabrication of these two condensers. The condensers were tested in an air-handing unit at various ambient conditions. The modified tube-array condenser with fins performed better than the one without fins and was abe to reject the design heat duty at ambient temperatures of 50.5 C or ower. A heat transfer mode was deveoped for the first tube-array condenser and compared with experimenta vaues. The mode showed that the condenser is air-side imited. The heat transfer mode is currenty being extended to finned tube-array condensers with finer segments for accurate assessment of performance. Future work and further improvement of compact condensers for ammonia-water systems wi invove exporing various fin geometries for tube-array condensers. NOMENCLATURE -2-1 h Heat transfer coefficient W m K C -1 Heat transfer coefficient W K k Conductivity -1-1 W m K Viscosity 1 1 kg m s D Diameter of tube m L Length m Heat exchanger effectiveness - Nu Nusset number - Re Reynods number - Pr Prandt number - Subscript i o v sg inner outer iquid vapor segment
2270, Page 10 REFERENCES Churchi, S. W. (1977), "Comprehensive Correating Equations for Heat, Mass and Momentum Transfer in Fuy Deveoped Fow in Smooth Tubes," Industria & Engineering Chemistry Fundamentas Vo. 16(1) pp. 109-116. Coeman, J. W. and S. Garimea (2003), "Two-Phase Fow Regimes in Round, Square and Rectanguar Tubes During Condensation of Refrigerant R134a," Internationa Journa of Refrigeration Vo. 26(1) pp. 117-128. Dobson, M. and J. Chato (1998), "Condensation in Smooth Horizonta Tubes," Journa of Heat Transfer Vo. 120(1) pp. 193-213. E Haja, J., J. R. Thome and A. Cavaini (2003), "Condensation in Horizonta Tubes, Part 1: Two-Phase Fow Pattern Map," Internationa Journa of Heat and Mass Transfer Vo. 46(18) pp. 3349-3363. Fernández-Seara, J., J. Sieres, C. Rodríguez and M. Vázquez (2005), "Ammonia Water Absorption in Vertica Tubuar Absorbers," Internationa journa of therma Sciences Vo. 44(3) pp. 277-288. Ferreira, C. I., C. Keizer and C. Machiesen (1984), "Heat and Mass Transfer in Vertica Tubuar Bubbe Absorbers for Ammonia-Water Absorption Refrigeration Systems," Internationa Journa of Refrigeration Vo. 7(6) pp. 348-357. Forinash, D. M. (2015), "Nove Air-Couped Heat Exchangers for Waste Heat-Driven Absorption Heat Pumps." Garimea, S. and J. W. Coeman (1998). Design of Cross-Fow Condensers for Ammonia-Water Absorption Heat Pumps. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., Atanta, GA (United States). Garimea, S., J. W. Coeman and A. Wicht (1997), "Tube and Fin Geometry Aternatives for the Design of Absorption-Heat-Pump Heat Exchangers," Journa of Enhanced Heat Transfer Vo. 4(3). Herod, K., R. Radermacher and S. A. Kein (1996). Absorption Chiers and Heat Pumps, CRC press. Incropera, F. P., D. P. DeWitt, T. L. Bergman and A. S. Lavine (2011). Fundamentas of Heat and Mass Transfer, John Wiey & Sons. Jaster, H. and P. Kosky (1976), "Condensation Heat Transfer in a Mixed Fow Regime," Internationa Journa of Heat and Mass Transfer Vo. 19(1) pp. 95-99. Kein, S. A. (2014). Engineering Equation Sover, F-Chart Software. Lorentzen, G. (1995), "The Use of Natura Refrigerants: A Compete Soution to the Cfc/Hcfc Predicament," Internationa Journa of Refrigeration Vo. 18(3) pp. 190-197. Meacham, J. M. and S. Garimea (2004), "Ammonia-Water Absorption Heat and Mass Transfer in Microchanne Absorbers with Visua Confirmation," ASHRAE Transactions Vo. 110(1). Srikhirin, P., S. Aphornratana and S. Chungpaibupatana (2001), "A Review of Absorption Refrigeration Technoogies," Renewabe and sustainabe energy reviews Vo. 5(4) pp. 343-372. Zieger, F. (1999), "Recent Deveopments and Future Prospects of Sorption Heat Pump Systems," Internationa Journa of Therma Sciences Vo. 38(3) pp. 191-208. Žukauskas, A. (1972). Heat Transfer from Tubes in Crossfow. Advances in Heat Transfer. P. H. James and F. I. Thomas, Esevier, Vo. Voume 8 pp. 93-160. ACKNOWLEDGEMENT The authors gratefuy acknowedge the financia support for this research through Advanced Research Projects Agency Energy (ARPA-E), Department of Energy, USA Award # DE-AR0000370.