1、288 2008 ASHRAE ABSTRACTThis paper explains the up-to-date stage of developmentof an air-cooled H2O-LiBr absorption chiller. The methodol-ogy of development is based on the systematic application ofadvanced numerical and experimental techniques, taking intoconsideration economical criteria. In a pre
2、vious developmentof a laboratory prototype of this concept of absorption chiller,two levels of mathematical modelling of different degree ofdetail, implemented for design/prediction purposes, wereexperimentally validated: (1) modelling of the absorptioncycle, based on overall balances of mass and en
3、ergy in thedifferent elements; (2) modelling of the heat exchangers of theabsorption cycle, that take into account the specific heat andmass transfer processes implied by means of semi-empiricalmodels. The same models have been applied for a new pre-industrial prototype.INTRODUCTIONIn the last decad
4、es a significant increase in electricityconsumption has been produced due to the growth in coolingdemand. In order to save energy in cooling systems it is neces-sary to develop new technologies to take advantage of alter-native energies, e.g. solar energy or waste heat.In the case of solar cooling i
5、nstallations, the main obstaclethat impedes an extended use of absorption systems is the largeinitial investment necessary for both solar collectors andabsorption machine. For low capacity installations (less than15 kW), the price of the chiller is the most limiting factor.Therefore, innovative desi
6、gns of both solar collectors andabsorption chillers (Ziegler 2002, Zogg and Westphalen 2006)to reduce prices are necessary. Moreover, a higher simplicityof the absorption cooling installation would reduce the cost ofthe investment significantly. Therefore, air-cooling for theabsorber and condenser c
7、an be an important issue to avoidusing a cooling tower. For solar air-conditioning applications,the refrigerant-absorbent pair H2O-LiBr is competitive withrespect the most universal pair NH3-H2O due to its higherperformance, but the use of LiBr as absorbent implies the riskof crystallization, more i
8、mportant in air-cooled systems.Different prototypes of air-cooled H2O-LiBr machines havebeen developed in the past decade for air-conditioning inbuildings (Ohuchi et al. 1994, Tongu et al. 1993, Enjoji 1998,Ishino and Kawasaki 1998, Kawakami et al. 1998). However,those machines were gas-fired. There
9、fore, the type of cycleused in those cases was double effect in order to get the maxi-mum advantage of the input energy. LiBr is sometimes used asabsorbent together with LiI to overcome the problem of crys-tallization (Tongu et al. 1993, Ishino and Kawasaki 1998). Themain problem for all those machi
10、nes relies on the high elec-trical consumption of the fans to create adequate cooling effecton the absorber and the condenser, because reduced andcompact designs were the main premises. Izquierdo et al.2001, presented an air-cooled absorber suitable for being usedin absorption systems for public tra
11、nsport, which took advan-tage of the engines waste heat.In the present research, an air-cooled absorptionmachine is being developed, in this case driven by hot waterbelow 100C (suitable for solar cooling applications) with asingle effect cycle configuration (Figure 1). Although thefinal performance
12、of the system is slightly reduced due to theair-cooling, in both the absorber and condenser the mainpremise of design has been to keep the electrical fan con-sumption in these elements within moderate limits. Themachine developed has a mechanic solution pump, whichRecent Developments in the Design o
13、f a New Air-Cooled, Hot-Water-Driven H2O-LiBr Absorption ChillerJess Castro, PhD Assensi Oliva, PhDCarlos David Prez-Segarra, PhD C. Oliet, PhDJess Castro, Assensi Oliva, Carlos David Prez-Segarra, and C. Oliet are mechanical engineers with the Centre Tecnolgic de Transfer-ncia de Calor (CTTC), Univ
14、ersitat Politcnica de Catalunya (UPC), Barcelona, Spain.NY-08-0352008, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 114, Part 1. For personal use only. Additional reproduction, distribution, or transmission
15、 in either print or digital form is not permitted without ASHRAEs prior written permission.ASHRAE Transactions 289leads to a different behavior of the machine compared to theheat powered pumps used in the commercially availablewater fired chillers. These chillers experiment a drasticdecrease of the
16、COP due to the decrease of the circulatedmass flow at low driving temperatures (Yazaki 1996). Themachine developed consists in horizontal tube, falling filmgenerator and evaporator, and air-cooled absorber and con-denser that are tube-fin heat exchangers (see Figure 1) wherethe falling film absorpti
17、on and condensation processes is pro-duced in the interior of the tubes.The main novelty from the point of view of final productis the implementation of an air-cooled absorber and condenserin a hot water driven H2O-LiBr machine. From the point ofview of the methodology, the main novelty is the syste
18、maticapplication of mathematical models for the design and predic-tion of the thermal behavior of the new prototype. With thisstrategy of simulation, it will be possible to make an accuratedesign of the cycle and heat exchangers, avoiding situationswith risk of crystallization from the thermal desig
19、n and spec-ifying accurately the suitable conditions of operation. Theselimited operating conditions have been implemented bymeans of a control strategy. This methodology saves develop-ment costs by means of reduction of experiments.RESEARCH APPROACH AND METHODOLOGYThis research will complete the in
20、dustrialization of theabsorption cooling machine that was the result of the Euro-pean funded project CRAFT ACABMA (Oliva et al. 2002),with the following final objectives:1. Introduction into the market of a new air-cooled absorp-tion chiller of low capacity using H2O-LiBr as fluid pair.This machine
21、should have a final price lower than theones existent in the market, all of them water-cooled, thatneed cooling tower.2. To dispose of the necessary infrastructures (demonstra-tion plants) that assure the maximum diffusion of the newabsorption chiller, facilitating its introduction to themarket.Meth
22、odology of AnalysisThe methodology adopted in the development of theabsorption chiller is general for any thermal system:Exhaustive analysis of the physical phenomena andgeometry. This work was already performed in the firststage of the ACABMA project, first laboratory prototype.Proposal of possible
23、 mathematical formulations. Studyof the diverse possibilities of use of levels of modelisa-tion of different degree of detail.Numerical resolution of the governing equations. In thispoint all the numerical solutions must be checked in orderto achieve results with no programming errors and notdepende
24、nt of the calculation mesh used (verification ofnumerical errors).Experimental validation. Comparison of the numericalpredictions with the results obtained in an experimentalfacility, in this case the laboratory prototype developed inwithin the European project.Figure 1 Single-effect, air-cooled abs
25、orption machine.290 ASHRAE TransactionsUse of the simulations as virtual experimental units.Once the code has been fully verified and the mathemat-ical formulation experimentally validated, it is possibleto use the codes as virtual experimental units for designor prediction purposes of pre-industria
26、l prototypes. Thisis the present status of the research.MATHEMATICAL MODELLINGIn this section the model employed for the design andsimulation of the whole absorption cooling system isdescribed. The simulation of the whole absorption cycle isbased on two levels of modelling: (1) cycle modelling level
27、,where the main overall values of the cycle are calculated(temperature, pressure, LiBr concentration and mass flow inthe most significant points of the cycle) together with the UA(overall heat transfer coefficient multiplied by the area ofeach heat exchanger), the effectiveness or heat exchangedthat
28、 are calculated according the calculation mode (one ofthem must be data); (2) heat exchanger detailed modellinglevel, where the UA and the subcooling values at the outlet ofthe heat exchangers, needed by the above mentioned cyclemodelling level, are recalculated using the input data(temperature, pre
29、ssure, LiBr concentration and mass flow)given by the previous level of modelisation (design phase) oreventually by the experimentation in order to validatedirectly the detailed models. Figure 2 shows schematicallythe calculation process and the exchange of data between thelevels of simulation:The pr
30、ocedure of calculation in design mode is as follows:1. Cycle simulation: calculation of the main data of thecycle (temperatures, pressures, LiBr concentrations andmass flows) and UAs.2. Detailed simulation: calculation of the degrees ofsubcooling or overheating at the outlet of the heatexchangers an
31、d recalculation of the heat transfer coeffi-cients as close as possible to ones given in the previousoverall simulation. Determination of the true size of theheat exchangers.3. Cycle simulation: recalculation of the main data of thecycle. In this case the UA of the heat exchangers are data.4. Detail
32、ed simulation: check of the validity of the previousdesign with the new cycle data. If it is valid, finish thecalculation, if not, return to point 3.Cycle SimulationIn order to study the type of air-cooled absorption systemdescribed in the introduction, a mathematical model tosimulate the whole syst
33、em (see Figure 1) has been developed.For each element, overall mass and energy balances areimposed. At the design mode, the nominal cooling capacity(heat exchanged at the evaporator) and the heat transfer effi-ciencies of the other heat exchangers are data, the UAs, the heatexchanged in each element
34、, and the main data of the absorp-tion cycle (enthalpies, temperatures, pressures, LiBr massfractions and mass flows) are obtained. At the rating mode,both the area and the overall heat transfer coefficients of theheat exchangers are data. Thus, the effectiveness and heatexchanged in each element ar
35、e obtained in addition to the maindata of the absorption cycle. Negligible heat losses and pres-sure drops in the connections between elements are assumed.Under steady-state conditions, and neglecting both kinetic andpotential energy, the equations considered in the analysis forthe different element
36、s are the global and the LiBr conservationof mass and the energy balance:(1)(2)(3)together with the equilibrium relations of LiBr aqueous solu-tion and H2O, enthalpy expressions for the LiBr aqueous solu-tion, H2O and air:T = Tsat,sol( p,c)(4)T = Tsat,wat( p 5h hsol(T,c)(6)h hwat(T,xg7h hair(T,w)(8)
37、It is assumed that the enthalpy of compressed liquid (forLiBr solution and pure H2O), and H2O vapor only depend onthe temperature. The equilibrium relations for the H2O-LiBrsolution and enthalpy expressions have been obtained fromASHRAE (1997). For the case of pure H2O, the data have beenobtained fr
38、om Furukawa (1991). For the air, the source of theenthalpy expressions has been ASHRAE (1997). All the UAsare defined according to UA = /LMTD, where LMTD is thelogarithmic mean temperature difference. The UA and sub-cooling values are recalculated in an accurate way by meansof the detailed models pr
39、esented in the next subsection. In allthe simulations, the mass flow is fixed in the pumps as theinput conditions in the secondary streams. Additional hypoth-eses are assumed in order to close the non-linear system ofequations generated. The equation system is solved using anFigure 2 Exchange of inf
40、ormation between the levels ofsimulation.moutmin0=mc()outmc()in0=mh()outmh()in QW+=QASHRAE Transactions 291iterative procedure (Gauss-Seidel method), where in eachfunction the most updated values of the variables are usedaccording to a previously determined order.As illustrative example, the equatio
41、ns to be solved for theabsorber in the laboratory prototype developed (Oliva et al.2002) are the following (see Figures 3 and 6 for the subin-dexes):(9)(10)(11)(12)(13)(14)(15)Detailed Simulation of the Heat ExchangersFor the design of the heat exchangers of the air-cooledabsorption chiller numerica
42、l detailed codes have been devel-oped. These codes are also used for prediction purposes. Theyare capable of simulating the heat and mass transfer processesimplied under certain hypotheses or using the appropriateempirical information if necessary. The solution heatexchanger has been calculated usin
43、g standard methods in heatexchangers.For this detailed modelling, the thermophysical proper-ties of H2O-LiBr have been obtained from different sources.Density, viscosity from Lee et al. (1990), thermal conductivityfrom DiGuilio et al. (1990), heat capacity from ASHRAE(1997), mass diffusivity from Re
44、id et al. (1977), and surfacetension from Kim et al. (1994). In the case of pure H2O, thesource for the thermophysical properties has been Furukawa(1991). For the air, the source of the enthalpy expressions hasbeen also ASHRAE (1997).Air-Cooled Heat Exchangers (Absorber, Condenser):Air Side. For the
45、 simulation of the air-side of the air-cooledelements (absorber and condenser), a code developed forsimulating the thermal and fluid dynamic behavior of fin andtube heat exchangers (Oliet et al. 2007, Prez-Segarra et al.2007) has been used. This code adopts a strategy of resolutionbased on discretiz
46、ation of the heat exchanger into macrocontrol volumes around the tubes (see Figure 4). Over thesemacro volumes, the equations of conservation of mass,momentum and energy are applied for the air, and the energyequation for the solid elements (tubes and fins). The inner flowis a specific subroutine. I
47、n this case absorption of water vaporin a liquid falling film LiBr aqueous solution or pure fluidcondensation, which provides of the necessary boundaryconditions for the calculation of the solid elements. In order tokeep the CPU time consumption within reasonable limits, themathematical formulation
48、requires the knowledge of someempirical information, specifically the local heat transfercoefficients, the local friction factors and the mass transfercoefficients for the absorption process. These fundamental orelementary empirical coefficients do not depend on a specificheat exchanger but on the k
49、ind of heat transfer surfaces andlocal flow structure. This information is obtained from theliterature (Kim et al. 1997).Serpentine Heat Exchangers (Generator, Evapora-tor): Water Side. In the case of the generator and the evapo-Figure 3 Scheme of the installation of the laboratory prototype (Oliva et al. 2002).m1m6 m13 m140=m1c1m6c60=m1h1m6h6 m13h13 m14h14 Qab=m15h16h15()Qab=p1p6p13p14= =T1Tsat sol,p1c1,()SUB=161156161156ln)()(TTTTTTTTAUQabab/=Castro et al. 2007a, 2007b). The heat exchangeris divided into contro
