1、4769 (RP-1199) The Residential Heat Balance Method for Heating and Cooling Load Calculations Charles S. Barnaby Member ASHRAE Jeffrey D. Spitler, PhD, PE Fellow ASHRAE Dongyi Xiao Student Member ASHRAE ABSTRACT The recent ASHRAE project, ?Updating the ASHRAE/ ACCA Residential Heating and Cooling Loa
2、d Calculation Procedures and Data? (1199-Rp), developed two new resi- dential load calculation procedures: residential heat balance (RHB), a detailed heat balance method that requires computer implementation, and residential load factor (RLF), a simpli- jedprocedure that is hand tractable and suitab
3、le for spread- sheet implementation. This paper describes RHB and its development. FOY calculation of sensible cooling load, RHB applies the general approach of the ASHRAE heat balance (HB) method, based on room-by-room 24-hour design-day simulation. The 24-hour procedure eliminates issues of gain d
4、iversi intermodel, analytical, and empirical test results will be reported in future publications (Xiao et al. 2005). Charles S. Barnaby is vice president of research at Wrightsoft Corporation, Lexington, Massachusetts. Jeffrey D. Spitler is C.M. Leonard Professor and Dongyi Xiao is a graduate stude
5、nt in the School of Mechanical and Aerospace Engineering, Oklahoma State University, Still- water, Oklahoma. 308 02005 ASHRAE. The sections below describe RHB and its implementation in ResHB. References to RHB and ResHB are made somewhat interchangeably because, in many ways, ResHB is RHB. Detailed
6、equation-based model descriptions are not included here; readers are referred to cited sources and the ResHB source code. BACKGROUND AND RHB DESCRIPTION Residential heating and cooling load calculations produce information needed for equipment selection and distribution system design. These results
7、include design values for heating, sensible cooling, and latent cooling equipment capacity plus room-by-room heating and sensible cooling loads. Experience has shown that simple procedures are sufficient for heating and latent cooling load calculations. Sensible cooling load calculations are more pr
8、oblematic. Sensible load results from the combination of several load components having building- and climate-dependent profiles. Excess sensible capacity increases first cost and results in performance problems, including poor humidity control, excessive power demand, and noisy operation. Thus, usi
9、ng conservative estimates of load components is not acceptable, and the overwhelming focus of the 1199-RP research project was on calculation of sensible cooling loads. Prior Methods Prior residential load calculation methods have been published by the Air-conditioning Contractors of America (ACCA),
10、 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-Funda- mentals include a method based on 342-Rp (McQuiston 1984). Canadian Standard CAN/CSA-F280-M90 (HRAI 1996; CSA 1990) specifies a cooling met
11、hod also based on 342-RP and a heating procedure that includes enhanced ground-loss calculations. These methods share many features. Their heating load procedures differ only in details; all ignore solar and internal gains and are based on summing surface UAAT heat losses, infiltration loss, ventila
12、tion loss, and distribution loss. Sensi- ble cooling loads are similarly derived by summing compo- nent contributions calculated using tabulated or formula- based factors incorporating temperature and solar effects as appropriate. With the exception of Manual J, eighth edition, all perform a single
13、design-condition calculation, implicitly making assumptions about relative timing of various gains and the zone response that transforms the gains into load. Recent addenda to Manual J (eighth edition) have added an adjust- ment that involves evaluation of the full-day room and zone fenestration gai
14、n profiles. The single design-condition calculation of sensible cool- ing load has long been problematic. Using the sum of peak component gains as the design load usually produces an exces- sive result because the gains generally occur at different times over the day. To account for gain diversity,
15、factors used in pnor methods were derived using semi-empirical adjustments such Temperahimswing (“0 O 15 3 45 1.100 m $ 0.800 i 0.700 L O .- - : 0.600 o. 500 1- I- : i I t ! 0.400 1 + O00 0.83 1.67 2.50 Tempetaaire swing (“c) Figure 1 Sensible cooling load reduction due to temperature swing. as mult
16、i-hour averaging. However, for situations with limited exposure (e.g., apartments), the dominant fenestration gains peak simultaneously and the sum-of-peaks estimate is more appropriate. To handle such configurations, prior methods have included alternative factors and/or adjustments variously calle
17、d “multi-family or “peak” (as opposed to “single- family” or “average”). User judgment is required to select the applicable condition. A multi-hour calculation eliminates the averagepeak distinction-the design load is simply the peak of the hourly profile. The only motivation for using a single desi
18、gn condi- tion is hand tractability. Implementers of past methods made the decision that an approximate method that would actually be used was preferable to a more accurate but impractically complex alternative. Given that personal computers are now ubiquitous, it is reasonable to use a 24-hour calc
19、ulation for an updated procedure. MasterlSlave Control, Temperature Swing, and Cooling Load Residential air-conditioning applications rely on multi- room, constant volume systems controlled by a single thermo- stat in one room (masterhlave control). Assuming sufficient capacity, good temperature con
20、trol occurs in the master (ther- mostat) room. The slave rooms maintain reasonable tempera- tures to the extent they have load profiles similar to that of the master and/or are conditioned by air mixing with adjacent rooms. In general, their temperatures will not be held at the setpoint even when th
21、e system is operating. The resulting temperature variation, or swing, has the effect of reducing the required capacity. This has long been recognized as a major consideration in residential cooling load calculations. Its importance is confirmed by this work. Temperature swing generally occurs in sla
22、ve rooms. However, with reduced cooling capacity, a thermostat room will experience temperature swing as well. Figure 1 shows the results of 192 ResHB sensible cooling load calculations for a ASHRAE Transactions: Research 309 single room with one exterior exposure in a variety of climates and in fou
23、r primary orientations. At a 1.67”C (3F) swing, the load reduction ranges from 13% to 50%. Some of this reduc- tion is due to the higher average room temperature when swing is allowed. However, most of the effect results from a portion ofpeak gains being absorbed in building mass as room temper- atu
24、re rises. This energy is “carried forward” and is removed at a later time when gains have moderated and the system has adequate capacity to bring the room back to the setpoint, re- cooling the building mass. Permitting a small, short-duration temperature excursion at design conditions usually result
25、s in a significant reduction in required sensible capacity, with asso- ciated cost reduction, moisture removal improvement (as a result of longer run times), and electrical demand reduction. Note that these are primarily capacity effects-approximately the same amount of energy is removed over the da
26、y with or without temperature swing. Second-order considerations, such as higher average part-load ratio, may result in some energy savings when swing is allowed, but their relative magnitude is much smaller than the capacity savings. Heat Balance for Residential Applications Over the last ten years
27、, ASHRAE nonresidential cooling load calculation procedures have moved to the heat balance (HB) method as the fundamental procedure (Pedersen et al. 1997) and the radiant time series (RTS) method as a simplified procedure derived from HB (Spitler et al. 1997). HB and RTS were evaluated regarding the
28、ir suitability as the basis for an updated residential procedure. Both are 24-hour methods. HB was selected because it can readily calculate either load at a known space temperature or a space temperature given a known extraction rate. The latter capability makes HB well suited to the residential ap
29、plication where room temperature swing is so important. To handle these floating temperature cases, RTS would have to be modified. An RTS extension, designated “Period Space Air Response Factor” (PSARF), was explored during 1199-RP. PSARFs relate extraction rate to air temperature deviation from a n
30、ominal setpoint and are analogous to the space air temperature weighting factors in the transfer func- tion method (TFM) (McQuiston and Spitler 1992). The PSARF approach was not ultimately pursued because an extended RTS method would be in essence a re-invention of TFM, which has been superseded by
31、HB. Given that direct application of HB is now computationally practical, there is no reason to resort to simplifications. An important goal for the updated residential procedure was simplicity of required input, preferably comparable to prior methods. Specifically, a detailed geometric building des
32、cription was deemed impractical. Using a simplified geometric model, where surface areas and orientations are known but their positions are not, implies (1) exact surface-to- surface view factors are not available and (2) room adjacencies are not known. The heat balance method can model longwave rad
33、iant exchange with good accuracy without exact view factors using an MRT formulation (Liesen and Pedersen 1997). Room adjacencies are unnecessary if room loads are calculated independently. Thus, there is a good match between heat balance and the requirement of simple input. RHB Definition The resid
34、ential heat balance method is a specialized application of the ASHRAE heat balance method. The follow- ing HB changes and extensions define RHB: Multi-room, multi-zone, and multi-system. The funda- mental RHB modeling unit is the room. Independent heat balances are performed for each room. Zones and
35、 systems are accounting structures to which loads are accumulated to provide overall results. Specialized algorithms. Temperature swing and master/ slave control can be modeled to produce realistic sensi- ble cooling load estimates. Residential models and assumptions. Component mod- els and assumpti
36、ons used for RHB are appropriate for the residential application. Simple heating and latent cooling procedures. As dis- cussed above, the simple UAAT model has proved satis- factory for heating load calculations. Similarly, latent load can be estimated from moisture gain from infiltra- tion, ventila
37、tion, duct leakage, and occupants. These simple approaches are retained in RHB. It should be noted that RHB is not a fully elaborated cool- ing system design procedure. In particular, RHB does not specifj how temperature swing and masterhlave control should be considered during the design process. R
38、HB can model rooms with or without swing, allowing choice on the much-debated question as to whether systems should be sized to allow swing at the thermostat on the design-day. Slave room temperature results from a case-specific combination of limited capacity and control profile mismatch, so its de
39、sign implications are more complex. It may be that RHB master/ slave capabilities should be used for investigation of zoning options only after primary load calculations are done on an independent room-by-room basis (with or without tempera- ture swing). The remaining sections of this paper provide
40、details about the above aspects of RHB in its current form. One major advantage of a heat balance formulation is that it can be tested and refined via direct comparison to empirical data. It is expected that RHB will evolve as additional research results become available. CALCULATION ALGORITHMS The
41、HB method is a design day procedure that requires iteration to find the steady-periodic solution at which all heat flows correctly balance. RHB adds the additional requirement of finding loads under floating temperature conditions in order 31 O ASHRAE Transactions: Research to handle temperature swi
42、ng and masterlslave control, as described here. Calculation Sequence and Convergence Criteria The fundamental RHB load calculation sequence is: repeat swing repeat day for hour = 1 to 24 for all rooms repeat for all surfaces end for surfaces perform air heat balance perform surface heat balance unti
43、l room convergence for current hour end for rooms until day convergence determine room supply airflow rates for next swing iter ation end for hours until swing convergence The convergence criteria are discussed below. The sequence was modified several times during development and its logic is worth
44、examining: The outer loop handles temperature swing (discussed below). Temperature swing occurs when cooling capac- ity is less than required to hold a room at the setpoint. The swing search algorithm adjusts each room supply air flow rate and repeats the entire calculation until the specified swing
45、 is achieved. The hour loop is outside the room loop. This means that current hour conditions are available for all rooms (either from the current day iteration or, at worst, from the prior day iteration), allowing inter-room references. One of the issues with a design-day heat balance proce- dure i
46、s determining when the solution has converged. A common technique is to continue iteration until calculated temperatures change a very small amount between iterations. The difficulty is to determine a “small amount” that truly represents convergence. Unfortunately, there are cases that change very l
47、ittle, iteration to iteration, but will continue to change, resulting in significant drift in results. Various conver- gence criteria were attempted for ResHB and the following are the best found to date: Hour. For each room, the current hour calculations are repeated until the sum of the absolute c
48、hange in surface temperatures plus air temperature is less than 0.0005 K (0.0009”F), indicating that a fully simultaneous solution has been closely achieved. Day. The day calculations are repeated until a22 rooms meet (a) the fractional difference between daily total inside and outside surface flux
49、is less than 0.005 and (b) the area-weighted total absolute temperature change for all surfaces plus air is less than 0.0002 K (0.00036”F). Note that the all-room requirement means that some rooms will be iterated beyond this point. Swing. The swing search is continued until swings for all rooms are within 0.01 K (O.O1SoF) of specified. Each room has a specified swing. This allows different swings in master and slave rooms, for example. Again, the all- room requirement means extra iteration for some rooms. In addition to these basic criteria, there are various safety checks that detect osc
