1、 International Journal of Heating, Ventilating, Air-conditioning and Refrigerating Research Editor John W. Mitchell, Ph.D., P.E. Professor of Mechanical Engineering University of Wisconsin-Madison, USA Associate Editors James E. Braun, Ph.D., P.E., Associate Professor, Ray W. Herrick Laboratories, A
2、lberto Cavallini, Ph.D., Professor, Dipartmento di Fisicia Tecnica, University of Padova, Italy Arthur L. Dexter, D.Phil., C.Eng., Reader in Engineering Science, Department of Leon R. Glicksman, Ph.D., Professor, Departments of Architecture and Ralph Goldman, Ph.D., Chief Scientist, Comfort Technolo
3、gy, Inc., Framingham, Massachusetts, USA Anthony M. Jacobi, Ph.D. Associate Professor and Associate Director ACRC, Department of Mechanical and Industrial Engineering, University of Illinois, Urbana-Champaign, USA Jean J. Lebrun, Ph.D., Professor, Laboratoire de Thermodynamique, Universit de Lige, B
4、elgium Reinhard Radermacher, Ph.D., Professor and Director, Center for Environmental Energy Engineering, Department of Mechanical Engineering, University of Maryland, College Park, USA Keith E. Starner, P.E., Engineering Consultant, Architecture, Building Loads, Energy, and Weather, York, Pennsylvan
5、ia, U.S.A. Jean-Christophe Visier, Ph.D., Head, Centre Scientifique et Technique du Btiment Energy Management Automatic Controller Division, Marne La Valle, France School of Mechanical Engineering, Purdue University, West Lafayette, Indiana, USA Engineering Science, University of Oxford, United King
6、dom Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, USA Policy Committee Lee W. Burgett, chair Jack B. Chaddock Ken-Ichi Kimura John W. Mitchell Frank M. Coda W. Stephen Comstock Editorial Assistant Publisher ASHRAE Staff Jennifer A. Haukohl W. Stephen Comstock Robert A. Pa
7、rsons, Handbook Editor Scott A. Zeh, Publishing Services Manager Nancy F. Thysell, Typographer 02000 by the American Society of Heating, Refrigerating and Air-Con- ditionhg Engineers, Inc., 1791 Tullie Circle, Atlanta, Georgia 30329. All rights reserved. Periodicals postage paid at Atlanta, Georgia,
8、 and additional mailing offices. HVAC nor may any part of this book be reproduced, stored in a retrieval system, or transmitted in any form or by any means-electronic, photocopying, recording, or other-without permission in writing from ASHRAE. Abstracts-Abstracted and indexed by ASHRAE Abstract Cen
9、ter; Ei (Engineering Information, Inc.) Ei Compendex and Engineering index; IS1 (Institute for Scientific Information) Web Science and Research Alert; and BSRiA (Building Services Research :;) The complaint event process model described above is not in the standard form because the levels are not fi
10、xed. It can be converted into the standard form by the following changes of variables where pTHTB is the cross-correlation coefficient for TH and TB, and pTLTB is the cross-correla- tion coefficient for TL and TB. With these transformations, the mean number of hot complaints in a time period t is no
11、w the mean number of zero-level upcrossings of the variable zh, and the mean number of cold complaints in a time period t is now the mean number of zero-level upcrossings of the variable zp Most buildings are typically occupied only during the daytime. In this case, the number of complaints per day
12、per zone will depend on the level crossing frequencies, and on the probability VOLUME 6, NUMBER 4, OCTOBER 2000 293 that when someone arrives in the morning a complaint condition already exists. Mathematically, the expected number of complaints per zone per day is as follows Enh = Ph+Vht (5) Enl = P
13、,+vlt (6) where h = e -m JT;“ (7) and where t is the length of time each day that the building is occupied. The quantities Ph and P1 are the probabilities that a hot complaint condition and a cold complaint condition exist when the building is first occupied because a level-crossing may have occurre
14、d before the occupants arrived. In addition to being dependent on the mean and standard deviation of the three pro- cesses, the predicted complaint rate is dependent on the standard deviation of the rate of change of the three processes. This is evident from Equation (1 1) and Equation (12), which a
15、re similar to Equation (1). Equation (5) and Equation (6) may be converted to a complaint cost function if the mean (expected) times to handle complaints (denoted as ETh and ETJ for hot and cold complaints, respectively) are known and if the labor rate of the service technician, denoted as R, is kno
16、wn. ETh and ET are the average labor times associated with hot and cold complaints, respectively. If a technician is dispatched immediately, then there are the times from when the complaint 294 HVAC 2000000 E o O 3 million square feet or 280 O00 m2) cost of energy plus service calls resulting from c
17、omplaints during the summer (May through September). The energy costs in the figure include all energy costs, gas and electric, and not just HVAC costs. The energy cost is only a function of the mean building temperature. Figure 8 shows the relative magnitude of complaint costs to energy costs as a
18、function of the mean and standard deviation of the building temperature. The figure illustrates that energy costs are generally much higher than complaint costs. Although the magnitude of energy costs are higher, the sensitivity of the energy cost to the mean temperature is lower than the sensitivit
19、y of complaint cost to the mean temperature when the mean temperature becomes extreme. For a given standard deviation of the temperature, there is always a mean temperature that will mini- mize the sum of the energy and complaint cost. Figure 9 shows the cost effectiveness of the temperature control
20、s as a function of the mean temperature and the standard deviation of the temperature. The cost effectiveness is defined as the minimum cost with perfect control (oTB = O) divided by the actual cost. This figure illus- trates that even when the energy savings of raising the indoor temperature are co
21、nsidered, there is still a penalty associated with controlling building temperatures to the limit of the ASHRAE comfort zone. The magnitude of the penalty depends on the control performance. When oTB = 3.57“F (1.98“C), the cost effectiveness of controlling these fictitious buildings at the limit of
22、the ASHRAE comfort zone (Le., at 79“F, 26.1“C) is 89.7%. In other words, there is a cost avoidance potential of 10.3%. When oTB = 1.O“F (0.56“C), the cost effectiveness of con- trolling these fictitious buildings at the limit of the ASHRAE comfort zone (i.e., at 79F) is 96.6% (cost avoidance potenti
23、al of 3.4%). If weather data from climates milder than Houston had been used in the energy analysis then the cost effectiveness at the limit of the ASHRAE comfort zone would be less, and the cost avoidance potential more, because the energy costs would have been less dependent on the indoor temperat
24、ure. Figure 10 shows the optimal mean temperatures that minimize the energy plus service call cost and that minimize the service call cost during the summer. The independent variable is the standard deviation of the building temperature. Also shown in the figure is the point, marked A, at which Faci
25、lity A was estimated to operate. From the location of point A it is clear that the 302 HVAC their mass flow rates; and the fin height were predicted. The results were used to developed correlations to predict the heat and mass transfer coeffiecients (on the airside) between the air and the liquid de
26、siccant. These cor- relations can accurately be used to predict the outlet air and liquid desiccant conditions from the fin-tube arrangement. Correlations were developed to predict the rate of heat and mass transfer between the air and the desiccant film. Several numerical experiments showed agree-
27、ment with the available data in literature. INTRODUCTION In the last twenty years, packed beds have been successfully used for air dehumidification using liquid desiccant. Factor and Grossman (1980), Lf et al. (1994), and Elsayed (1994) studied the performance of packed beds using different types of
28、 liquid desiccant. It was found that packed beds have two major disadvantages: the pressure drop in the airside is relatively high and the mass flow rate of the desiccant is much greater than the airflow rate, which makes the pumping power of liquid desiccant per unit mass of air flow large. The fal
29、ling film of liquid desiccant was used to overcome these two disadvantages. Park et al. (1994) studied the heat and mass transfer between a cross airflow and a falling desiccant film over the fin surface of a heat exchanger. Conlisk (1995) used a configuration consisting of a film of a lith- ium bro
30、mide solution falling on a vertical tube and in contact with water vapor. The govern- ing equations were solved analytically with the assumption that the temperature in the film is conduction-dominated after a short distance down the tube and that the mass transfer occurs in a thin layer at the inte
31、rface. Ahmad S. Rahmah is a teaching assistant and Najem M. Al-Najem is a professor with the Mechanical and Industrial Engineering Department, Kuwait University, Kuwait. Moustafa M. Elsayed was a professor with the Mechanical and Industrial Engineering Department, Kuwait University, and is now with
32、ProService in Cairo, Egypt. 307 308 HVAC body force of air flow is negligible Heat conduction in both the x and z directions is negligible Species thermodiffusion and diffusion-thermo effects are negligible Effect of tube geometry on the heat and mass transfer process between the air and the desicca
33、nt film is negligible t The last assumption was validated experimentally by Park et al. (1994) and is also validated ana- lytically in this paper. The following are the pertinent equations that describe the conservation of mass, momentum and energy for both air and desiccant film based on the above
34、assumptions and using the defini- tions of the various symbols given in the nomenclature list. The x-momentum equations for air and desiccant with their boundary conditions are, respectively: 310 Ud = o atyd=O and the energy equations are: T,=T. ai at x = O The species diffusion equations are: W= Wi
35、 atx=O HVAC&R RESEARCH VOLUME 6, NUMBER 4. OCTOBER 2000 311 w = w* aty, = 6, (54 Integrating Equation (2a) subjected to boundary conditions Equations (2b), (2c), and (2d) gives the following profile for the desiccant velocity across the film thickness: from which the thickness 6d is obtained after u
36、sng the continuity equation, i.e. Similarly, the integration Equation (1) over the y direction gives the following distribution for air velocity: Again utilizing the continuity equation gives: Rahmah and Rahmah et al. (1997, 1998) solved the above equations under isothermal condi- tions on the fin s
37、urface, i.e. when the fin surface temperature T, was assumed to be constant. Therefore, the governing equations and their boundary conditions become independent of the z coordinate and a two dimensional solution in the x-y plane can be obtained. In that model, referred to as Model I, the temperature
38、 variation on the fin surface was neglected and this is not practically correct. Currently, a more reliable model (Model Ii) is considered to account for the variation of the fin surface temperature. This model requires the knowledge of Tw(x,z) over the fin surface and therefore the boundary conditi
39、ons become dependent on the z coordinate in addi- tion to the x and y coordinates. Here, we used two approaches to determine the surface tempera- ture Tw(x,z). In the first approach, the fin surface temperature was determined analytically for a circular fin of the same surface area as that of the pr
40、esent rectangular fin (Kern and Krause 312 HVAC&R RESEARCH 1972). The approach will be referred to as the analytical temperature distribution approach (ATDA). This approach has two approximations. First, it assumes constant film temperature and constant heat transfer coefficient between the film and
41、 the fin. Second, it assumes circular distri- bution of temperature which does not represent the actual case. In the second approach, the heat conduction equation was solved numerically over the x-z plane of the fin by finite difference method. We will call this approach the numerical temperature di
42、stribution approach (NTDA). This second approach is more accurate than the first approach, yet, it requires more computa- tional time (Ramah 1997). NUMERICAL PROCEDURES The coupled governing Equations (3), (4), (9, and (6) were descretized using the control vol- ume method explained by Patankar (198
43、0) and the resulting finite difference equations were solved numerically. The following is a brief description of the principle steps of the numerical algorithm procedure, using Model II. More details can be found in Rahmah et al. (1997). The solution starts in the x-y plane at z = O. The solution i
44、s then repeated for other x-y planes by marching in z direction to the end of the fin. Both the NTDA and ATDA are used to deter- mine the surface temperature of the fin. The following is a brief description of the solution pro- cedure using NTDA. 1. Specify the inlet conditions for both air and desi
45、ccant solution, the dimension of the fin, the diameter of the tube, and the base temperature of the fin (the temperature of the fin surface in contact with the tube). 2. As an initial guess, assume that the surface of the fin has a constant temperature, which is equal to the base temperature of the
46、fin. 3. Select the x-y plane at z = O, solve for the air and desiccant temperatures, humidity ratio, and water concentration in the desiccant. 4. Repeat Step 3 for other x-y planes by marching in the z direction by step Az until the back edge of the fin is reached. 5. Solve for the fin temperature d
47、istribution. Use the desiccant temperature distribution and the local heat transfer coefficient obtained in Step 3 for every node on the surface of the fin. 6. Calculate the error between the new value of the fin temperature and the assumed one for every node. If the maximum error for all the nodes
48、exceeds a prespecified value then go to Step 3, and use the new temperature distribution as the initial guess. 7. Repeat Steps 3 to 7 until the maximum temperature error of all nodes on the surface of the fin is within the required accuracy. Similarly, the solution using ATDA was done by assuming an
49、 initial guess for the values of the average heat transfer coefficient hd between the desiccant film and the average desiccant film temperature Td along the fin surface. These were used to get the temperature distribution for the fin surface using ATDA. Solution for Ta, Td, 6, and W for the whole domain was carried out following the same procedure “c” and “d” in the previous algorithm of NTDA. The average heat transfer coefficient hd and the average desiccant film temperature ?.d over the entire fin were calculated next, and if the err
