1、4768 (RP-1199) Development of the Residential Load Factor Method for Heating and Cooling Load Calculations Charles S. Barnaby Member ASHRAE ABSTRACT The recent ASHRAE project, “Updating the ASHRAEI ACCA Residential Heating and Cooling Load Calculation Procedures and Data” (RP-I 199), developed two n
2、ew resi- dential loads calculation procedures: residential heat balance (RHB), a detailed heat balance method that requires computer implementation, and residential load factor (RLF), a simpli- fed procedure that is hand tractable and suitable for spread- sheet implementation. This paper describes R
3、LF and its development. The form of RLF resembles prior methods. Howevel; the sensible cooling load procedure was derived using linear regression to jnd relationships between design conditions, building characteristics, and peak cooling load predicted by RHB. This eliminated the needfor semi-empiric
4、al adjustments, such as averaging, that have been used in the development of other methods. Results comparing RLF to RHB are presented. The RLF heating load calculation is also described; it uses the traditional UAAT formulation except for improvements to procedures for infiltration leakage rate and
5、 ground (slab and basement) losses. INTRODUCTION The research project, “Updating the ASHRAE/ACCA Residential Heating and Cooling Load Calculation Proce- dures and Data” (RF-1199), had two primary products. First, a new fundamental residential heating and cooling load calcu- lation method was develop
6、ed and tested. This procedure, called the residential heat balance (RHB) method, is based on heat balance first principles as described by Pedersen et al. (1997, 1998) and ASHRAE (2001). RHB is documented by Barnaby et al. (2005). It uses a computationally intensive 24- hour design-day simulation th
7、at is practical only when imple- Jeffrey D. Spitler, PhD, PE Fellow ASHRAE mented in software. Because of its fundamental approach, RHB can be applied with few restrictions to arbitrarily complex residential buildings, including those with large fenestration areas, novel construction features, or ha
8、ving non- summer peaks. The ResHB computer program, developed as part of RP- 1 199, implements the RHB method as described in Barnaby et al. (2004). ResHB is a batch-driven FORTRAN-90 applica- tion derived from the ASHRAE Loads Toolkit (Pedersen et al. 2001) that operates on Windows-based PCs. Sever
9、al key ResHB features are noted here. First, ResHB models room temperature swing: in addition to the standard fixed-setpoint capability, ResHB can find the sensible cooling extraction rate that results in a specified temperature swing above the ther- mostat setpoint. Second, ResHB incorporates the u
10、pdated models identified in W- 1 199 as appropriate for residential loads calculation. Third, ResHB is multi-room and multi- zone, allowing application to real buildings as well as simple test cases. Finally, ResHB can model typical residential master-slave control, where a thermostat in one room co
11、ntrols the cooling delivery in another, with resulting imperfect temperature control in the slave room. The second product of RP- 1 199 is a simpler procedure, designated “residential load factor” (RLF) method. RLF is tractable by hand or can be straightforwardly implemented using spreadsheet softwa
12、re. This simplification is achieved at the expense of generality-RLF is applicable only to conven- tionally constructed residences with typical space-condition- ing requirements. The procedures and data required to use RLF are presented in the “Residential Heating and Cooling Loads Calculation” chap
13、ter of the 2005 ASHRAE Handbook- Fundamentals (ASHRAE 2005). Charles S. Barnaby is vice president of research at Wnghtsoft Corporation, Lexington, Mass. Jeffrey D. Spitler is CM Leonard Professor in the School of Mechanical and Aerospace Engineering, Oklahoma State University, Stillwater, Okla. o200
14、5 ASHRAE. 291 This paper discusses the design of RLF and documents the methodology used in its development. Some testing results are also presented. While RLF includes both cooling and heat- ing load procedures, the heating calculations rely on the tradi- tional UAAT model that has proven satisfacto
15、ry for decades. Improvements have been introduced in relation infiltration leakage rate and to ground heat loss. The RLF cooling procedure resembles and builds upon prior methods but was developed using a linear regression approach that avoids some semi-empirical derivations used in the past. Prior
16、methods have been published by the Air-Condi- tioning Contractors of America (ACCA), including the widely used Manual J, seventh edition (ACCA 1986) and Manual J, eighth edition (ACCA 2003). The 1989-2001 editions of the ASHRAE Handbook-Fundamentals included a method based on 342-RP (McQuiston 1984)
17、. Canadian Standard CAN/CSA-F280-M90 (HRAI 1996; CSA 1990) specifies a cooling method also based on 342-RI and a heating procedure that includes enhanced ground loss calculations. RLF COOLING LOAD CALCULATION The RLF cooling load calculation is based on the idea of independent load components, as ar
18、e prior simplified meth- ods. The load contributions from various sources are sepa- rately evaluated and then summed. The following sections summarize the method, showing both sensible and latent components as applicable. Later sections document the deri- vation of the component models and coefficie
19、nts. In FUF, surfaces have associated load factors (LFs) or load contribution per unit area. These are designated CFs for cooling and HFs for heating. For the most part, HF values are simply UAT CF values depend on surface construction, climate, and, in some cases, surface orientation, solar absorp-
20、 tance, or other characteristics. Each unique LF needs to be evaluated once for a given set of site and construction condi- tions and then is applied repeatedly to building elements of the same type. This two-step process is convenient for hand or spreadsheet application. Note that LFs are the funct
21、ional equivalent of ManuaZSs heat transfer multipliers (HTMs) but are derived differently and in general do not have the same values. Surface Type Ceiling or knee wall adjacent to vented attic Ceilinghoof assembly Wall Floor over ambient Floor over crawlspace Total Cooling Load 9, = CAj. CF, + qvj,
22、+ qig,s 41 = 4ViJ + 9igJ where qs = sensible cooling load, W (BtUni) q1 = latent cooling load, W (Btuni) Ai = area of ith surface, m2 ( l/(Rcvr+ 0.12) W/m2.K or 1/(Rc, + 0.68) BtUni.F 292 ASHRAE Transactions: Research Table 2. Fenestration Coefficients Exposure N NE FFs 0.17 0.09 E I 0.17 I SE S 0.2
23、5 0.45 I sw I 0.54 I W NW 0.48 0.34 I Horiz I 0.66 I Fenestration Fenestration cooling factors are calculated as follows: CFyen = U.(AT-0.49.DR)+FFs.PXI.SHGC.IAC (5) where CFfen = u= DR = AT = FF, = PXI = SHGC = ZAC = Peak fenestration cooling factor, W/m2 (tih.) fenestration NFRC heating U-factor,
24、W/m2.K daily range of outdoor dry-bulb temperature, K (“F) cooling design temperature difference, K (OF) load factor (see Table 2) peak exterior irradiance, including shading modifications (see below), W/m2 (StUih.ft2) fenestration rated or estimated NFRC solar heat gain coefficient interior shading
25、 attenuation coefficient exterior irradiance (PXI) is the hourly maximum (tih.ft2.0) solar gain incident on the surface. PXI = TIE, (unshaded) (6) PXI = Tx(Ed + (1 - Fshd)ED) (shaded) (7) where PX - peak exterior irradiance for exposure, W/m2 (BtU/h.fP) peak total, difise, and direct irradiance for
26、exposure, W/m2 (t/h.ft2) transmission of exterior attachment (see Table 4) fraction of fenestration shaded by permanent overhangs, fins, or environmental obstacles For horizontal or vertical surfaces, irradiance values can be obtained from Table 3 for primary exposures or Algorithm 1 for any exposur
27、e. Skylights with slope less than 30“ from horizontal should be treated as horizontal. Steeper slopes, other than vertical, are not supported by the RLF method. Algorithm 1. Exterior irradiance Horizontal surfaces E, = 970 + 6.2L - 0.16L Ed = MIN(E, 124) ED = E, - Ed Vertical surfaces Y= x (normaliz
28、ed exposure) 114 E, = 462.2 + 1625“ - 6183Y3 + 3869“ + 32.38YL + 0.3237YL2 - 12.56L - 2 1.040LL 0.8959L +- Y+l 392.1 - l38.6Y + 2.107YL - -) i2i4 in most cases, they should not be considered shaded. m (fi) ASHRAE Transactions: Research 293 Table 3. Exterior Irradiance (W/m2) NEW ED 54 1 532 522 51 1
29、 501 490 480 470 46 1 Ed 163 154 147 140 135 130 I26 123 120 E* 704 686 668 652 636 62 1 606 593 580 EN ED 62 7 640 650 657 662 663 662 659 653 Ed 173 169 166 163 162 161 161 161 162 E, 800 809 816 82 1 824 825 823 820 815 SEISW ED 334 380 422 460 494 525 553 577 598 Ed 174 173 I74 175 177 180 I83 1
30、87 191 E+ 508 553 595 635 672 705 136 764 788 S ED O 65 146 223 297 368 43 6 50 1 563 Ed 149 171 175 180 186 192 198 205 212 E? 149 236 32 1 403 482 559 634 705 774 Ventilation and Infiltration Infiltration airflow is calculated as follows: Hor Qinr = infiltration airflow rate, L/s (cfm) AL = buildi
31、ng effective leakage area (including flue) at 4 Pa assuming C, = i, cm2 (in.2) Io - 1, = coefficients, as follows: ED 906 90 1 888 867 838 80 1 156 703 642 Ed 124 124 124 124 124 124 124 124 124 Et 1030 1025 1012 99 1 962 925 880 827 766 10 1, 1, H = building average stack height, m (fi) (approximat
32、ely AT = indoor-outdoor temperature difference, K (?F) A,-dflue = flue effective leakage area at 4 Pa assuming C, = 1, The ventilation airflow rate is determined according to the installed or planned ventilation equipment that is expected to be operating at design conditions. Generally, intermittent
33、ly 2.5 m 8 ft per story) cm2 (in.2) Cooling Heating Windspeed- Windspeed- 3.4 mls (7.5 mph) 6.7 m/s (15 mph) 25 (343) 51 (698) 0.38 (38) 0.35 (31) O. 12 (.28) 0.23 (S3) 294 Attachment Exterior insect screen ASHRAE Transactions: Research Tx 0.6 Shade screen Manufacturer SC value, typically 0.4 to 0.6
34、 Table 5. Shade Line Factors (SLF) GO G,.c Note: Shadow length below the overhang equals the shade line factor times the overhang depth Latent Sensible 136 (464) 20 (68) 2.2 (0.7) 0.22 (0.07) operated exhaust fans are not included. Overall supply and exhaust flow rates are determined and divided int
35、o “balanced” and “unbalanced” components and combined with infiltration. Qai = M”(Qsup,Qexh) (10) (1 1) (12) Qunai = MAX(Qsup9Qexh) - Qai Qvj = MAX(QunbalQjnf+ 0.5 . Qunbai) I Go, where Qbai = Qsup = Qexh = Qunai = Qvi = 12 (41) 22 (75) balanced ventilation airflow rate, L/s (cfin) total ventilation
36、 supply an-fiow rate, L/s (cfm) total ventilation exhaust airflow rate (including any combustion air requirements), Ws (cfin) unbalanced airflow rate, LIS (cfin) combined infiltrationhentilation flow rate (not including balanced component), L/s (cfm) Note that unbalanced duct leakage can produce add
37、itional pressurization or depressurization. This effect is included in distribution losses, discussed below. The cooling (or heating) load due to ventilation and infiltration is calculated as follows, taking into account the effects of heat/energy recovery ventilation (HRV/ERV) equip- ment: sensible
38、 ventilation/infiltration load, W (BtUni) air sensible heat factor, 1.23 W/(L/s).K (1.1 Btu/hNoc qig,i = GO,/ +Gcf,.Acf+Goc,i.Nac (18) where qig,s qig,i GX = coefficients, as follows: = sensible cooling load due to intemal gains, W (Btu/h) = latent cooling load due to internal gains, W (Btu/h) qdl =
39、 Fdl 4s where qd Fdl 9s = distribution loss, w (Btu/h) = distribution loss factor, from Table 6 = building sensible load, W (BtUni) ASHRAE Transactions: Research 295 Cooling Duct Unsealed Sealed Heating Furnace Heat Pump Unsealed Sealed Unsealed Sealed tightness Duct nsulationR (m2.W h.ft2.“F/ StUl)
40、 D EVE LOP M E N T OVE RVI E W The RLF formulation is conceptually transparent and hand-tractable: the “loads“ from each wall, window, and other gain sources are calculated and summed to get the total load. Unfortunately, however, it is not possible to find invariant models for each load component b
41、ecause of interactions among them. For example, a major interaction occurs between opaque surfaces and fenestration-the load resulting from solar gain is lagged and moderated by differing amounts depending on surface construction. Even simple convective gains, such as infiltration and ventilation, p
42、resent difficulty because they should be evaluated at the building-dependent peak hour. Note that RHB completely avoids these difficulties: 24-hour calculations allow gains to combine according to their case-specific profiles, and the heat balance procedure accu- rately represents component interact
43、ions. Development of a load-component method such as RLF requires that the significant interactions be identified and addressed (via configuration-specific load-component calcu- lations), eliminated (by restricting the configurations to which the method is applicable), or neglected if the effects ar
44、e deemed small. Addressing interactive effects introduces more complexity in the method, which defeats its purpose. Given the availability of RHB to handle essentially any configura- tion, RLF applicability is restricted to typical residential construction. Prior methods assumed the independence of
45、load compo- nents and developed models for each. The component models were in many cases the obvious choice (e.g., infiltration load derived directly from an air leakage rate). However, excessive loads are predicted by simply using maximum fenestration and opaque surface heat gain rates. For these c
46、omponents, semi-empirical factors or adjustments were invoked to make O 0.7 1.4 O 0.7 1.4 O 0.7 1.4 O 0.7 1.4 O 0.7 1.4 O 0.7 1.4 (4) (8) (4) (8) (4) (8) (4) (8) (4) (8) (4) (8) the results consistent with experience. In particular, a common strategy was to use factors equal to multi-hour averages o
47、f calculated instantaneous gains. While the averaging approach has some intuitive appeal, it has no rigorous basis, as is acknowledged in older editions of the ASHRAE Handbook- Fundamentals (ASHRAE 1972). Averaging in this manner was found to give results compatible with measured residential loads.
48、Hence, these are averages only in the sense that combining numbers in this manner results in accurate factors for calculating window loads of residential structures. Room temperature swing is one reason adjustment is required. Assuming a fixed indoor temperature, as is typically done in nonresidenti
49、al procedures, results in excessive loads for the residential case. Better overall system performance and cost-effectiveness results when equipment is sized to allow some temperature variation at design conditions. Averaging of gains derived assuming fixed room temperature mitigates their excessive peak. The RLF development procedure avoided adjustments by relying on RHB cooling loads calculated with temperature swing and deriving required factors using linear regression. Equation 1 was treated as a model for which submodels and coefficients were needed. Later sect
