1、460 ASHRAE TransactionsABSTRACT Two DX air cooling coil modeling methods, one used in EnergyPlus and the other termed as a generic rating-data-based (GRDB) DX coils modeling method, were summarized and their detailed calculation procedures were presented using one case study. Six rooftop models of t
2、wo manufacturers were used for this study to predict sensible cooling capacity. It demonstrated that the relative errors in sensible cooling capacity prediction ranged from -14.9% to 15.1% for the method used in EnergyPlus and from -7.4% to 4.2% for the GRDB method. In addition to higher accuracy an
3、d precision, the GRDB method is more robust against the variations in parameter selections, has a wider application range, requires less computation power, and is more straightforward.INTRODUCTIONThe DX air cooling coil model is a key module in model-ing or simulating an overall cooling system to fa
4、cilitate the systems design, operation, and maintenance. Generally, all engineering simulation or modeling programs include an element modeling module for DX cooling coils, but individual programs may use different modeling methods and thus have different performance. For example, EnergyPlus is one
5、of the most popular building energy simulation tools which was originally based on Blast and DOE-2 and has been improved during one decade. However, although it is often used as a benchmark for comparative analysis, its performance is still very sensitive to user inputs. For example, the selection o
6、f the rated air flow rate has a significant impact on the DX coil sensible cooling capacity. Within the wide range of rated air flow rate suggested by EnergyPlus, an individual user may select a different rated air flow rate so that the predication could be very different (Hand et. al, 2005). Recent
7、ly, Yang and Li (2009) developed a GRDB DX coils modeling method. This method was validated by extensive laboratory testing and demonstrated to have good accuracy and to be easy to use. The objective of this paper is to perform an extensive comparative analysis between this GRDB DX cooling coil mode
8、ling method and the DX cooling coil modeling method used in EnergyPlus. ENERGYPLUS DX AIR COOLING COIL MODELING METHODEnergyPlus Modeling ProcedureThe modeling method of the DX air cooling coils used in EnergyPlus (US Department of Energy, 2009; ASHRAE HVAC 2 Toolkit; Henderson et al. 1992; Henderso
9、n et al. 2000) is slightly different for different coil types, but their basic procedures and formula are the same. Thus, the Single-Speed Electric DX Air Cooling Coil is taken as an example, and only the key variables (i.e., , , and SHR) in the cooling model are discussed here. The basic logic is:
10、firstly, users define the rated condition for the cooling system, limitation range of air flow rate at operating conditions, and coefficients of two modifiers of total cooling capacity (i.e., entering air temperature and supply air flow rate); secondly, EnergyPlus program constructs the two modifier
11、s to calculate the total cooling capacity at operating conditions other than the rated condition; finally EnergyPlus program calculates the property parameters of air leaving and entering the coil (e.g., the enthalpy of the entering air) and the coils geometry parameters (e.g., the BF of the coil),
12、and then obtains SHR at the operating condition based on the fixed geometry parameter QtotalQsensibleA Comparative Analysis and Validation of Two DX Cooling Coil Modeling MethodsHuojun Yang Haorong Li, PhDMember ASHRAEHuojun Yang is a doctoral student and Haorong Li is a professor at the University
13、of Nebraska-Lincoln, Omaha, NE.AB-10-0122010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions (2010, Vol. 116, Part 2). For personal use only. Additional reproduction, distribution, or transmission in either print or
14、digital form is not permitted without ASHRAEs prior written permission.2010 ASHRAE 461and the physical model of the cooling coil. The detailed proce-dure is summarized in Appendix C. Potential Improvements of the EnergyPlus Modeling MethodFrom the calculation procedure in Appendix C, we can see:1. A
15、s a simulation program, EnergyPlus has some inputs that need to be determined by users. These inputs include and the coefficients of two modifiers. In terms of , users can choose any value when it can satisfy the requirement of /rated ton within 300-450. However, each has a specific , and , and the
16、whole calculation about the model may be different even for the same coil. In terms of the coefficients of two modi-fiers, users have to determine the regression equation order, a set of data as the regression base data, and the equations coefficients. EnergyPlus calculates the two modifiers through
17、 these coefficients to adjust the cooling capacity as a function of entering air tempera-tures and supply air flow rate when the operating condi-tion of air is not the rated condition. These input selections give rise to more uncertainties or sensitivity for the simulation result, and the models rob
18、ustness may be weak (one detailed calculation case is also shown in Appendix C). 2. There is one stated limited application range for operat-ing conditions, i.e, CFM/ rated ton at operating condi-tions should be between 200 and 500. This range will limit the wide application of the model, especially
19、 for cases simulating faulty operating conditions such as foul-ing or low air flow rates.3. From the perspective of EnergyPlus users, they need to be educated to perform the following tasks, and it is not easily accessible to common users because there is no clear procedure in some cases, e.g., the
20、EnergyPlus docu-ment doesnt clearly state how to get the coefficients of total capacity modifiers.define the rated condition for the cooling system and limitation range of air flow rate at operating conditions; choose a set of data as the regression base data and the regression equation order; obtai
21、n the coefficients of two modifiers of total cooling capacity. 4. In terms of calculation, the model calculates various physical parameters (e.g., BF and the psychrometric properties of air) and needs iterations for dry-coil condi-tions of each data point. The calculation characteristics may take En
22、ergyPlus more time to get results, especially when there are a lot of data points (especially data of dry-coil condition). GENERIC ALTERNATIVE MODEL OF DX AIR COOLING COIL BASED ON MANUFACTURERS DATAThe development and description of the generic alterna-tive model was detailed in Yang and Li (2009),
23、 and the follow-ing is the short summary of the main procedure.Development of the Generic Alternative ModelBased on the analysis of the free, public, and available manufacturers rating data, the cooling systems relationship of outputs with inputs can be presented by the following formor (1)(2)Combin
24、ing the above equations with the physical model, the generic alternative model further developed the common format of cooling system (seen in Eq. 3). (3)For any fixed ( , CFM, OAT), there is one critical point ( ) that divides the cooling coil condition into wet-coil condition and dry-coil condition
25、, as shown in Figure 1 where is normalized (i.e., the actual total cooling capacity ). In the figure, the vertical line ( ) through the critical point divides the whole graph area into two parts: the left one is the dry-coil condition (i.e., when real ), and the right one is the wet-coil condition (
26、i.e., when real ). In the dry coil area, both and SHR are constants (i.e., SHR = 1 and ), because has no effect on them at a fixed . In the wet coil area, both and SHR will vary with the increase of (SHR will decrease and will increase accordingly). Furthermore, for any operating driving input , the
27、 generic model gives the following cooling formula:CFMratedQtotalCFMratedCFMratedCFMratedQtotal rated,SHRratedQsensible rated,QtotalQTfETwbOAT CFM,()=QTfETwbETdbOAT CFM,()=SHR f ETwbETdbOAT CFM,=Cooling model formatWet-coil conditionDry-coil conditionQTfETwbCFM OAT,()=SHR f ETwbETdbOAT CFM, 1 SHR 0(
28、),=QSSHR QT=QTfETdbCFM OAT,()=SHR 1=QSSHR QT=ETdbETwb0QT0SHR, 1=QTQTrating capacityETwbETwb0=ETwbETwb0QTQTQT0= ETwbETdbCFM OAT,QTETwbQTETdbETwbCFM OAT,2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions (2010, Vol.
29、116, Part 2). For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.462 ASHRAE Transactions(4)To finalize the above formula, three items need to be determined: ETwb, the detailed functio
30、ns of = f(ETwb, CFM, OAT) and SHR = f(ETwb, ETdb, OAT, CFM), and their solutions will turn to the mathematic methods as shown in the equation (A-2) in Appendix D.Advantages of the Generic ModelFrom the above model development and calculation procedure in Appendix D, we can see:1. Basically, in terms
31、 of model methodology, the generic model is a grey box model that is constructed by a regres-sion method based on both the first principles and manual data. More specifically, a total cooling capacity wet curve, i.e., = f(ETwb, CFM, OAT), is obtained by using the manufacturers wet-coil condition dat
32、a; then a SHR wet curve, SHR = f(ETdb, ETwb, OAT, CFM) is derived from the wet coil condition data and the mixing condition data (i.e., the dry coil condition data close to the critical point); and finally the wet bulb temperature of the critical point (ETwb0) is solved by a simple and effective mat
33、he-matical equation. Therefore, the generic model will work well during the wet coil condition area because the above functions are regressively obtained from wet coil condi-tion data and the interpolation of a grey box model always has a good performance. However, the performance of the generic mod
34、el during the dry coil condition data, i.e., the extrapolation of the model, will depend on whether the determination of the critical point (or its wet bulb temperature ETwb0) is accurate. This is the reason why the regression base data for SHR incorporate the mixing condition data where the actual
35、critical point (or ETwb0) may exist.2. From the perspective of the generic models application range, there is no limitation to the operating condition, although there is an empirical parameter (i.e., 0.04) in the construction of SHR equation. In addition, the generic model is based on the freely ava
36、ilable manufacturers rating data, and then it can be easily implemented. 3. In terms of calculation, the generic model only relates with the parameters posted in the manual, and doesnt further calculate other physical parameters (e.g., BF) or air psychrometric properties (e.g., enthalpy, humidity ra
37、tio). However, it may still need several trial runs to tune up the right mixing condition data that are close to the critical point. 4. In terms of users, due to the freely available manual data, the simple calculation procedure and the less calculation burden, this modeling methodology is not only
38、more accessible to common users of the packaged HVAC units, but easier to HVAC engineers in terms of facilitating these units design, maintenance, and real-time automatic fault detection and diagnosis.COMPARATIVE ANALYSIS AND VALIDATIONFirst, the detailed calculations of the two methods are compared
39、 using Rooftop A17.5 (manufacturer A, rating ton 17.5), and its original manual data is attached in Appendix A, where CFM, OAT, and ETdbhave four different values and ETwb has three different values; that is, the total sets of data are 192 (4443). The original data of Rooftop B17.5 (manu-facturer B,
40、 rating ton 17.5) is also given in Appendix B for comparison. Then, the performance comparison and analysis for more system models follows. The calculation procedures of Rooftop A17.5 by EnergyPlus model and by the GRDB model are attached in Appendix C and Appendix D respec-tively, and all cases equ
41、ation coefficients obtained by the two models are presented in Appendix E and F respectively.Results Comparison The following shows the results of six rooftop machines (two manufacturers: A and B; and three types of tons for each manufacturer: 7.5 ton, 17.5 ton, 25 ton), including all of their possi
42、ble rated CFM scenarios in the EnergyPlus model for each rooftop, seen in Table 1. Each case is composed by the model scenarios (e.g., E1, G standing for the first EnergyPlus scenario and GRDB respectively) and rooftop name. For example, E1-A7.5 means manufacturer A, rating ton 7.5, rated Figure 1 T
43、he relationship of and SHR with ETwbat a fixed (ETdb, CFM, OAT).QTETdbETwbCFM OAT,()ETwbETwb0ETwbETwb0QTfETwbCFM OAT,()=SHR f ETwbETdbOAT CFM, 1 SHR 0(),=QSSHR QT=QTQT0fETwb0CFM OAT,()=SHR 1=QSQT0=QTQT2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.o
44、rg). Published in ASHRAE Transactions (2010, Vol. 116, Part 2). For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.2010 ASHRAE 463CFM 2700 in EnergyPlus, and G-A7.5 means manufacturer
45、 A, rating ton 7.5, GRDB.The relative error of sensible cooling capacity,is defined as the overall error of the two cooling models. In terms of wet coil condition, dry coil condition, and wet Qtotal) at the same (CFMrated, ETdb,rated) and to get a biquadratic regression equation, i.e.:TotCapTempModF
46、ac = f(ETwb, OATdb)b. Using a set of wet data (CFM; Qtotal) at the same (ETdb,rated, ETwb,rated, OATdb,rated) condition to get a quadratic or cubic regression equation, i.e.,TotCapFlowModFac = f(Actual CFM/CFMrated)c. Calculate Qtotalat operating conditions:(A-1)Thus, Qtotalis in fact the function o
47、f three inputs (ETdb,ETwb, CFM), i.e.5. Calculate SHR values at operating conditions (Carrier et al. 1959).a. Define the psychrometric properties of the air entering and leaving the cooling coil at the rated condition.Air entering the cooling coil:Air leaving the cooling coil:b. Calculate the geomet
48、ry parameter Aoof the bypass factor at the rated condition.Calculate SlopeRated at the rated condition.Get the psychrometric properties of the apparatus dewpoint (ADP) along the SlopeRated line and the saturation curve of the psychrometric chart, i.e.:QSQTETdb rated,80F 26.67C()=ETwb rated,67F 19.44
49、C()=OATdb rated,95F35C()=OATwb rated,75F 23.89C()=TotCapTempModFac = 2.21981236E+001.13591285E-03 OAT 4.86111108E-05 OAT23.76157413E-02 ETwb2.68683865E-04 ETwb2+1.02182541E-04 OAT ETwb+TotCapFlowModFac = 7.58809518E-013.59523822E-01CFM1.19047625E-01 CFM2+QtotalQtotal rated,=TotCapTempModFac()TotCapFlowModFac()Qtotalf CFM ETwbOATdb,()=Qtotal210 (2.21981236E+00 1.13591285E-03 OAT=4.86111108E-05 OAT23.76157413E-02 ETwb+ 2.68683865E-04 ETwb21.02182541E-04 OAT ETwb
