Guidelines for Energy Simulation of Commercial Buildings
Total Page:16
File Type:pdf, Size:1020Kb
Guidelines for Energy Simulati n of Commercial Buildings Michael B~ Kaplan e Kaplan Engineering oebe Caner 8 Seattle City Light Grant .. Vincent, Bonneville Power firiJr"!a"lliilniiC!'+lIl'~l'il·in,n Energy Edge is a large-scale research and demonstration project funded by the Bonneville Power Administration to test energy conservation in commercial buildings. The project involves intensive build ing computer simulation work using an hourly analysis program, DOE2.1C. We have written a set of guidelines that distills the experience we have gained from the building simulation work. The two primary purposes of the guidelines are to advise conservation program managers on the use of modeling, and to improve the accuracy of design-phase computer models. This paper provides an overview of the guidelines. We address at some length a few of the more impor tant issues raised in the guidelines. These include: (1) what program managers can reasonably expect from computer simulation, (2) what program managers can do to improve the reliability of simulation in their programs, (3) what are some of the main sources of discrepancies between the savings predictions of design-phase models and those of as-built or calibrated (4) which simulation have the J,;.,AVUa.Vli:71l.- impact on the results, and (5) what modelers can do to minimize error in their work. Introduction lbe Simulation of Commercial is a research and demonstration hereafter called the project funded by the Bonneville Power Administration liU.lae~lln.es'J distills the experience gained from intensive This was initiated to determine CornPllter building simulation work for the whether commercial buildings can be designed and The purpose of the Guidelines is twofold: to constructed to use at least 30% less energy than if they advise conservation program managers on the use of were and built to meet the regional model energy modeling, and to the accuracy of design-phase code, the Model Conservation Standards (MCS) developed models. To achieve these purposes, the by the Pacific Northwest Electric Power and Conservation Guidelines: Planning Council. Secondary objectives of the project are to determine the incremental costs and energy savings of a @ discusses the value of for energy wide of energy conservation measures (ECMs) and conservation programs to compare the accuracy of design-phase models with models that are calibrated with monitored ® discusses and weaknesses of OUJUOlln2 data. commercial buildings were models selected to participate in Energy Edge. All but two were new construction. In the area of research, Bonneville decided to pay special @ discusses that are un'usllalJly drivers attention to the methods for determining actual energy of energy use and model InaCCtlra(~v savings. Typically, design-phase computer analysis must deal with relatively little information about the reality of a n1"'r,,'u1,!"'IAoO R;UJlaeJlm~~S for developing baseline models specific building and its operation. Energy Edge attempted to expand the present limits of energy modeling by moni nll~r"'U1 ,!"'IAoO basic reQiuu~errlents for model documentation. toring the selected buildings in great detail and then using the monitored data to ground the models in reality. This most of our Edge is associated resulted in at least three distinct models for each moni with the DOE2 software, we have attempted to make the tored building--the design-phase model, the as-built Guidelines more broadly applicable. model, and the calibrated tuned) model. In this research, .Kaplan Engineering was responsible for @ Simulation input. This section covers the most critical the coordination of the as-built modelers and the review of aspects of loads, systems, and central plant input, the as-built building models as well as the calibration of including: zoning, infiltration, window and wall the models to monitored data. It is this experience that has U-values, shading and solar gain, daylighting, thermal led to the Guidelines for Simulation of Commer mass, unconditioned spaces, interior walls, above cial tlUHalLn~S. ceiling spaces, heat loss to ground, weather, general receptacle loads, computer rooms, lighting, commer Bonneville is now incorporating the Guidelines into the cial refrigeration, cooking equipment, HVAC equip and operation of its Energy Smart Design (ESD) ment capacity and part-load efficiency, HVAC program. The ESD program is Bonneville's primary vehi controls, multiple zone systems, ventilation, fans cle for energy from commercial build schedules and supply air volume, fan heat, and hidden in the Pacific Northwest energy users. Organization of the Guidelines $ Development of the baseline model (the building as modeled without This section presents guide lines for developing consistent and reasonable The Guidelines has two main The first part presents baselines. mOdelln2 from the pel~splect:lve of the conservation pro- gram manager. include: @ Three for model error input output check, and check energy use indices @ The and weaknesses of This compiled by others. section seeks to the program manager better decide when and when not to use mC~C1e_l1nj2;. @ Recommendations for contents for model documentation. ® Alternatives and COlnpJlerrlents to mc.aeun:f2:. 1be of the Guidelines several techni ® Basic control pr()CelC1Ul~es. cal aids for the modeler. These include a table of V-values for wans and roofs (corrected for Areas for further research. tables of Eill values end-use for various building recommended default occupancy schedules for The and ImJ:Hlc:atUJns various building types, and salnple controls sequences and errors. 'Ibis section has the n,::lo11"1"n-r"tn'!l1"llI"'·a. specifications for our Renaissance modeler. progranl managers a more realistic UnlC1el"stand:lng what can from mc~ae.unJg. Modeling From the Program ® in the context of the anager's Perspective In this section we discuss a few of the major modeling The second rn'~~p:nt~ detailed technical 2Ulldejllnt~S for of interest to conservation program managers$ The modelers. include: Guidelines expands upon these and other related topics. A for an role for the modeler--that nnW'nllll'''\W'lII'!:il''o Application of Simulation rnodelers take more in that recommended ECMs are Computer simulation of building energy performance can The Guidelines recommends that modelers write be an expensive endeavor. For this reason, program detailed np.~I"-tn·Mn'!JInr\p, C'<i"'llon... 1-1l,t"'HJlf"1I"'W'IlC' and check design dOCUlnents and submittals. managers want to know when this expense is warranted. Whole-building modeling is not necessarily the optimal tool for an energy analysis. It should not be used when a ® aimed at model more simple or less costly method of analysis can yield adequate results. Similarly, a complex software program Discussion of how modelers can interact actively with should not be used when a simple one can adequately their simulations$ address a specific building or ECM. Complex programs do not yield more accurate results. ® Common sources of simulation error$ 1~ 138 NO ''''''''''''.''''''''''''61 et a/" (Conversely, a simple program should not be used when a Sometimes the more complex one is needed to adequately address specific weaknesses or error. ECMs or buildings.) Can be used for program evaluation? More On the other hand, program managers may see some value research is needed on the statistical of model in a standardized approach to energy analysis. This predictions. The project case studies indicate permits easier review of modelers' work and eliminates a strong variability in the results from different models, discrepancies due to different software. It does not modelers, and model- es. This may tend to however eliminate discrepancies due to different lTIodelers disappear and the estimates become more reliable when or techniques. It also tends to force selection of software applied to a very number of it suited to the common denominator (Le. the most has to be determined if (with its demanding building or This can be expensive. is more accurate than other lower cost evaluation tech- when over a number of OUllIQllD.2;S. In general, modeling is best suited for the of ECMs that either interact with or directly affect HVAC M()a.e~lml2 is one of the technical tasks in a performance. An important exception to this is successful conservation program. Model estimates cannot light-di~g controls in day-lit buildings. is be reliable for many ECMs if the ECMs are not Install.eo, overkill for analysis of ECMs that yield to manual and maintained calculations. We also suggest that energy many of the failure of non- COlTIputer programs are poor tools for ""''"'''.U..I!..M:::Jl,i.I....lJl;:;.. commissioned ECMs. electrical demand or for HVAC eql1uprnel1lL uanty As a should not resort to ~~'~'JrYlII~~r."Y" method can do the Models are and vulnerable to error. Models are as weB. is the program selected to error as sponges are to water. simulation is should fit the it should not be selected for the often the best available method forECM convenience or comfort of the modeler. Sometimes model- But it is a method that needs can be with another form of control. We make several recommendations calcula~ control: tions~ Whenever should check the results of one form of .r:1l1"Hllil'!i!C!1IC! another method. We recommend that utilities ae,reiC)D minimum cations for their modelers and that enforce them. IV1()Oe~UIllg can be no more accurate than the aSS;Unl1PtloI1S vu.allllclCltU)nS should include in modeling a that lie behind both the and the baseline wide rangf' of and OUllldJlll2 models. Even the lnodel with several different software programs. If it is COIDP,Lex calculations on these ass:un:1Ptllons, the that programs win be result win be are l~he reQIUlr'ea~ then extensive in such programs Guidelines of some of should be <r~(,lnl1rP:('m of the modelers.