1、OR-05-1 1-4 Run-Around Heat Recovery System Using Cross-Flow Flat-Plate Heat Exchangers with Aqueous Ethylene Glycol as the Coupling Fluid Haisheng Fan Student Member ASHRAE Robert W. Besant, PEng Fellow ASHRAE ABSTRACT A two-dimensional steady-state mathematical model is developedfrom physical prin
2、ciples to study heat transport of a run-around heat recovery system for air-to-air heat recovery in HVAC applications. A jnite diference method is employed to solve the governing equations of the cross-flow flat-plate heat exchanger, giving the outlet air properties for any inlet operating condition
3、s. The accuracy of the model is verified by comparisons with known theoretical solutions for individual cross-flow exchangers and run-around systems. The efective- ness of each exchanger and the overall run-around heat recov- ery system is shown to be dependent on the dimensionless area to the therm
4、al capacity rate, N, and the thermal capacity ratio, Cr, of the heat exchanger. Designers are presented with a method to select the best operatingflow rate of the coupling fluid and the dimensionless design parameters of the exchang- ers that will maximize the overall efectiveness and heat rate of t
5、he run-around system and allow for its part-load control. INTRODUCTION Providing a comfortable and healthy indoor environment for building occupants is the primary concern of HVAC engi- neers. An adequate outdoor air ventilation rate is one of the key factors for a comfortable and healthy indoor env
6、ironment, especially for commercial and institutional buildings. The minimum requirement for the outdoor air ventilation rate in buildings has been modified over the years. ASHRAE Stan- dard 62- 1989 (ASHRAE 1989) recommended 9.4L/s (20cfm) of outdoor air per person in office buildings, whereas only
7、 2.4LIs (5 cfm) per person was recommended previously in ASHRAE Standard 62-1 98 1. The new ASHRAE Standard 62- 2003 now requires modified calculations of ventilation air Carey J. Simonson, PhD, PEng Associate Member ASHRAE Wei Shang, PhD flow rates that result in values somewhat similar to Standard
8、 62-1989. Ventilation airflow results in increased heating, ventilating, and air-conditioning equipment capacities and building operating costs. One ofthe ways to reduce these costs for ventilation is to transfer heat between exhaust and supply airstreams when it is cost-effective (ASHRAE 2001). Run
9、-around heat recovery systems, as shown in Figure 1, are an economically attractive way of increasing ventilation rates in buildings, especially in retrofit applications. Unlike other air-to-air heat or energy recovery facilities, such as air- to-air cross-flow plate heat exchangers, heat pipes, and
10、 rotary energy or enthalpy wheels, the run-around heat recovery system doesnt require the supply and exhaust air ducts to be side by side within a building. This gives the run-around system an advantage over other systems when cross-contam- I 7 Auxiliary pumu Exhaust Air Flow Supply Air Flow Y Speed
11、 Controlled Pump Figure I Schematic diagram of a run-around heat recovery system. - Haisheng Fan is a research assistant, Carey J. Simonson is an associate professor, Robert W. Besant is professor emeritus, and Wei Shang is a research associate in the Department of Mechanical Engineering, University
12、 of Saskatchewan, Saskatoon, SK, Canada. 02005 ASHRAE. 901 ination is a serious concern or for retrofitting where the ducts have already been installed. Many studies have been carried out on run-around systems in recent years. Zeng et al. (1992) studied the effect of temperature-dependent properties
13、 on a two-coil run-around system. Bennett et al. (1994a, 1994b) included wavy fins and liquid bypassing in their simulation and studied the life-cycle cost (LLC) savings of the run-around systems. Dhital et al. (1995) studied the maximum outdoor air ventilation rate and the energy performance of off
14、ice buildings with and without run-around heat recovery systems. Johnson et al. (1995a, 1995b) studied multi-coil run-around heat exchanger systems and performed a life-cycle cost analysis. Studies on the run-around heat recovery system with cross-flow flat-plate exchangers has not been reported in
15、the literature perhaps because bypass-control valves have been used to manage part-load conditions in pressurized systems circulating the aqueous glycol coupling liquid through finned- tube coils. Now that it may be practical to use speed- controlled pumps for part-load flow control and low-cost pla
16、s- tic cross-flow exchangers, run-around heat recovery systems operating at lower liquid pressures should be reconsidered for some ventilation air heating or cooling applications. A run- around system using plate heat exchangers may be cost-effec- tive because plate heat exchangers are typically les
17、s expensive to manufacture than coil-type heat exchangers. Furthermore, the pressure drop across plate heat exchangers can be lower than coil-type heat exchangers. These factors could reduce the payback period and increase the life-cycle savings of flat-plate run-around systems compared to coil run-
18、around system. Objective It is the purpose of this paper to develop the theoretical/ numerical model for run-around heat recovery systems with cross-flow flat-plate heat exchangers, verify its accuracy as far as possible, and investigate the overall effectiveness of this run-around heat recovery sys
19、tem for a range of operating conditions. MATHEMATICAL MODEL Flow and Exchanger Configuration The geometry of one pair of flow channels for a cross- flow flat plate heat exchanger and the coordinate system for the mathematical model are shown in Figure 2. We propose using low-cost polyethylene as the
20、 plate material in each exchanger. Major assumptions in the formulation of the math- ematical model are: 1. Heat transfer is in steady state and only in the z direction normal to each plate, and the heat transfer process is fully developed. The airflow and liquid flow are fully developed and unmixed
21、 when they flow through the channels of the heat exchanger. The assumptions can be shown to be accurate for most typical geometries and operating conditions using sensitivity studies. 2. Govern i ng Eq uat i onc The governing energy balance equations for analyzing the heat transfer through the plate
22、 at any point (x, y) on the surface are: Air Side. At any point ,y) in the exchanger, the heat flux through the plate is balanced by the heat gaidloss in the air: 2U.x0 8TA (TA - TL) = - CA ay where TA and TL are the bulk mean temperatures at (x,y) in the air and water-glycol solution, respectively,
23、 and Air flow ait Plates /“I Liquid Liquid flow ait flow in Air flow in Yo (a) One pair of flow channels IZ Air flow kG Plate X ) The coordinate system Figure 2 Schematic of a cross-ow flat-plate heat exchanger showing three membranes separating one liquid and one air channel and one liquid and one
24、airflow. 902 ASHRAE Transactions: Symposia CA pn, = mass flow rate ofthe dry airthrough asingle channel Convective Heat Transfer Coefficients = heat capacity rate of the air (i.e., CA = mA .CpA), Determination of the per unit width of channel. is the overall heat transfer coefficient between the air
25、 and liquid. The convective heat transfer coefficients in the air stream, hA, and water-glycol solution, h, are assumed to be constant; therefore U will be independent of position in the exchanger. For many operating conditions, the effect of entry length will be negligible. Liquid Side. At any poin
26、t (x,y) in the exchanger, the heat flux through the plate surface is balanced by the net heat gain/ loss in the liquid: where CL mL = = heat capacity rate of the liquid (i.e., C, = m, . Cp,) mass flow rate of the liquid through a single channel per unit width of channel. Boundary Conditions Air side
27、 (inlet) (x,y = O): Liquid side (inlet) (n = 0,y): (4) (5) A set of equations can be developed by rewriting the above governing equations to include the dimensionless groups for heat transfer. They are: Air side: Liquid side: where (7) For an effective design with cross-flow exchangers, NA and NL wi
28、ll be typically between 4 and 10. Convective heat transfer coefficients can be obtained experimentally or from empirical theoretical equations from the literature (Incropera and DeWitt 1996; Kays and Crawford 1990; Kays and London 1984). An empirical correlation of the heat transfer for turbulent fl
29、ow in a channel with Reynolds number larger than 10,000 can be represented by an algebraic expression of the form: Nu = C. RemPrn (10) For lower Reynolds numbers (3,000 ( Re ( 5 x lo6): h.D, (11) cf/S) . (Re - 1000). Pr 1 + 12.7. cf/8)/2. (P,2/3 - i) - - k Nu = where f = 0.791n(Re)-1.64-2 for smooth
30、 walled surfaces. with Re I 2,000, For the fully developed laminar flow (ASHRAE 2001) In this study, Nu = 8.24 is used, which is the case for fully developed heat transfer between infinite rectangular plates with uniform surface heat flux (Kays and Crawford 1990). In the current study, the air and l
31、iquid flows through the exchang- ers are laminar because this is the most likely operating condi- tions for the liquid and air sides of these flat plate exchangers. Properties of the Aqueous Ethylene Glycol Fluid Aqueous ethylene glycol is used in the coupling fluid circuit to prevent fieezing durin
32、g cold weather. The physical properties (i.e., density, specific heat, thermal conductivity, and viscosity) of the aqueous-glycol solution depend on its temperature and concentration. Increasing the concentration of ethylene glycol in the solution above 60% by mass will increase the freezing point o
33、f the solution (ASHRAE 2001). To prevent equipment damage during idle periods in cold weather in HVAC applications, 30% ethylene glycol is suffi- cient (ASHRAE 2001). Therefore, 30% ethylene glycol water solution is used as the coupling fluid in the run-around system in this paper. Its properties ca
34、n be calculated from quadratic curves fit to data points (ASHRAE 2001): Spec heat: 42 CpL = 3588.554+2.8441 .t-1.135(10 ).t (-10“CSt (1 8) For no heat and mass gain or loss in the connecting pipes, the inlet property conditions are: TL,in,S= TL,out,E (19) 2 pL = 1/(0.2409 + 8.8667( t + 9.3209( i5) f
35、 ) (-IOOC 5 i CL: It is this latter case (Equation 31), which is of greatest interest for heat recovery in this study because for part-load control CA C, (i.e., the heat transfer will be controlled by reducing the flow rate of the coupling fluid). With Equations 29-3 1 and correlation Equation 28, w
36、e can predict the overall effectiveness of the run-around system and compare these results with simulations. The overall effectiveness results are presented in Figure 9 and Figure 1 O for the case of equal exhaust and supply air mass flow rate and Ns = NE. From Figure 9, we can see that the overall
37、effectiveness of the run-around heat recovery system with cross-flow flat-plate heat exchangers increases with the increase of N for the same Cr. For both cases, when N 3 the overall effectiveness of the run-around system depends strongly on Cr and is a maximum when Cr approaches 1 .O. In contrast,
38、for a single cross-flow flat-plate heat exchanger, its effectiveness is minimum when Cr = 1.0 for the same N as shown in Figure 6. The overall effectiveness of the system varies dramatically with N for N 6. This indicates that we should keep N large when designing a run-around system with cross-flow
39、 flat-plate heat exchangers. The numerical results agree well with the theoretical-correlation results, except when N 2 1 O. This slight difference is thought to be caused by the correlation Equation (28), which is less accurate when N is larger. The total effectiveness of the run-around system is m
40、ore sensitive to Cr for the case of CA C, than the case of From Figure 10, which shows the overall effectiveness cg versus Cr, we can see that for any N the overall effectiveness CA Ck of the run-around system reaches its maximum at approxi- mately Cr = 1 for the case of CA C,. But for the case of C
41、A i. For 3 IN 5 10, the maximum effectiveness occurs at CLICA 1, but the effec- tiveness doesnt change SO much for 0.8 I CLIC, I 1.2. That means the heat capacity of the glycol solution C, should be kept in the range of (0.8 . CA C, 1.2 . CA) for the designing peak load condition of a run-around sys
42、tem for maximum heat rate. It is noted that these findings for the optimum heat rate when C, I CA = 1 .O differ from the findings of Bennett et al. (1994a, 1994b) and Johnson et al. (1995a) for run-around . Numerical -Correlation 0.9 0.8 0.7 O. 6 0.5 0.4 0:; - : - o. 1 of , O 12 3 4 5 6 7 8 9 10 11
43、CUCA Figure 11 Variation of the overall efectiveness of the run- aroundsystem as afunction of CL/CA with Nas a parameter: systems using finned-tube coils because they were dealing with designs where both the air and the liquid sides of the exchangers had turbulent flows. In this study, using low-cos
44、t, plastic, cross-flow heat exchangers, laminar flow is used on both the air and liquid sides-so the best design conditions are quite different. Zeng (1990) showed that for a run-around heat recovery system with counterflow finned coil exchangers, N must be at least 5 to achieve an overall effective
45、ness of 70%. In our model, N must be at least 1 O to achieve an effectiveness of 70%. Although this indicates a doubling of area for cross- flow exchangers made of plastic, the production cost of the plastic exchangers is expected to be lower than metal finned- tube coils. When the heat rate of this
46、 run-around system exceeds that which is required (i.e., during moderate outdoor air tempera- ture conditions), the system heat rate can be controlled for part-load conditions by reducing the value of CLICA. This can be realized by reducing the mass flow of liquid (i.e., the pump- ing rate of the co
47、upling fluid). Of course, as CLIC, is decreased, Nwill increase-so the trajectory of this control on Figure 11 will not be on a constant N line. CONCLUSIONS A theoreticallnumerical model of a run-around heat exchanger system for heat exchange between ventilation and exhaust air has been presented. T
48、he numerical results are shown to agree well with the well-known analytical results for heat transfer in cross-flow flat-plate heat exchangers. It is shown that the number oftrans- fer units Nand thermal capacity ratio Cr need to be carefully selected in order to achieve a high overall effectiveness
49、. To achieve a high overall effectiveness ofthe system for a specific design, N should be high (e.g., N 2 5) and CLIC, in the range This optimum design value of .so will result in the highest design heat transfer rate between the exhaust and supply air at the design conditions. For part-load control, C, can be reduced of 0.8 2 CLIC, 1.2. ASHRAE Transactions: Symposia 909 by reducing the pumping rate to obtain any heat rate between the design heat transfer rate and no heat transfer. This numerical method is the most effective way to inves- tigate the overall effectiveness
