EnergyPlus vs. DOE-2

Energy and Buildings 43(2011)1663–1675
Contents lists available at ScienceDirect
Energy and
Buildings
j o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /e n b u i l
d
EnergyPlus vs.DOE-2.1e :The effect of ground-coupling on energy use of a code house with basement in a hot-humid climate
Simge Andolsun a ,∗,Charles H.Culp a ,Jeff Haberl a ,Michael J.Witte b
a College of Architecture,Texas A&M University,053WERC,3137TAMU,College Station,TX 77843-3137,USA b
GARD Analytics Inc.,115S.Wilke Road,Ste 105,Arlington Heights,IL 60005,USA
a r t i c l e i n f o Article history:
Received 17November 2010
Received in revised form 27January 2011Accepted 14March 2011Keywords:BESTEST EnergyPlus DOE-2
Ground coupling Basement Crawlspace
a b s t r a c t
For low-rise buildings,the heat losses or gains through the ground coupled building envelope can be a sig-nificant load component.Studies have shown that current simulation tools give dissimilar results for the ground coupled heat transfer (GCHT)with basements.This paper quantifies and explains the differences between EnergyPlus and DOE-2.1e (DOE-2)basement GCHTs based on their results for an all electric code house in a hot and humid climate zone.The code house was modeled with two basement configurations i.e.a conditioned basement and an unconditioned crawlspace.DOE-2was used with Winkelmann’s base-ment model and EnergyPlus was used with its GCHT calculator utility,Basement.The results revealed that the ground isolated EnergyPlus houses used 3–23%more cooling,12–29%less heating and 3–7%lower overall HVAC electricity use when compared to the ground isolated DOE-2houses.Ground coupling added up to three times more heat loss in EnergyPlus than in DOE-2.This increased the overall energy use difference between these two programs from 3–7%(ground isolated)to 14–25%(ground coupled).These results showed that a truth standard is required for basement heat transfer calculations of low-rise buildings.
©2011Elsevier B.V.All rights reserved.
1.Introduction
Ground coupled heat transfer (GCHT)through underground concrete walls and floors can be a significant component of the total load for heating or cooling in low-rise residential buildings.It is estimated that an uninsulated basement can contribute up to 60%of heat loss in a house well-insulated above grade [1].Ground coupling is still considered a hard-to-model phenomenon in build-ing energy simulation since it involves three-dimensional thermal conduction,moisture transport,long time constants and heat stor-age properties of the ground [2].Comparative studies on ground coupled heat transfer models of current simulation tools showed remarkable variation for basements.The disagreement among the simulation tools with respect to the average values was estimated to be 87%for ground coupling heating load [3],138%for ground coupling cooling load [3]and 11–23%for the annual total heating load [4].
This study compared EnergyPlus and DOE-2.1e (DOE-2)ground coupled heat transfer for basements of low-rise residential build-ings.DOE-2has been used for more than three decades for building
∗Corresponding author.Tel.:+19798626415;fax:+19798622457.E-mail addresses:andolsun@ ,simgeandolsun@ (S.Andolsun).
design and code compliance studies,analysis of retrofit opportu-nities and developing and testing standards [5].In 1996,United States Department of Energy (DOE)initiated support for the devel-opment of EnergyPlus ,which was a new program based on the best features of DOE-2and BLAST [6].The shift from DOE-2to EnergyPlus raised questions in the simulation community on the differences between these two simulation programs [7–9].Ground coupled heat transfer is an area that EnergyPlus calculations significantly differ from those of DOE-2.EnergyPlus calculates z -transfer func-tion coefficients to compute the unsteady ground temperatures for underground surfaces [10];whereas DOE-2sets a single steady ground temperature for each month [11].DOE-2basement GCHT has been compared with that of BLAST-3.0,SERIRES/SUNCODE,SERIRES-1.2,ESP,S3PAS,TRNSYS,TASE,DEROB-LTH and CLIM2000in order to maintain consistency among the results of simulation tools for identical cases [3,4].EnergyPlus and DOE-2have been com-pared with each other based on thermal loads,HVAC systems and fuel fired furnaces [12]using the test cases defined in American National Standards Institute (ANSI)/American Society of Heating Refrigerating and Air-conditioning Engineers (ASHRAE)Standard 140-2007[13],which effectively isolates the test cases from the ground.This study extends the previous studies by quantifying the differences between the basement models of EnergyPlus and DOE-2based on the results obtained for a code house with a basement in Austin,Texas.
0378-7788/$–see front matter ©2011Elsevier B.V.All rights reserved.doi:10.1016/j.enbuild.2011.03.009
1664S.Andolsun et al./Energy and Buildings43(2011)1663–1675
Nomenclature
IGain daily internal gain(Btu/day per dwelling unit) CFA conditionedfloor area(ft2)
N br number of bedrooms
SLA specific leakage area(unitless)
L effective leakage area(ft2)
R eff effective resistance of the underground wall/floor
(h ft2F/Btu)
A area of the underground wall/floor(ft2)
F2perimeter conduction factor(Btu/h F ft)
P exp exposed perimeter(ft)
U eff effective U-value of the underground wall/floor (Btu/h ft2F)
R ub actual resistance of the underground wall/floor
(h ft2F/Btu)
R w resistance of8in.concrete wall(h ft2F/Btu)
Rfilm resistance of the inside airfilm(h ft2F/Btu)
R soil resistance of the soil(h ft2F/Btu)
Rfic resistance of thefictitious insulation layer
(h ft2F/Btu)
EPlus EnergyPlus
DOE-2DOE-2.1e
T z zone air temperatures(◦C)
T gf outside surface temperatures of the ground coupled floor(◦C)
T if inside surface temperatures of the ground coupled floor(◦C)
T gw outside surface temperatures of the ground coupled wall(◦C)
T iw inside surface temperatures of the ground coupled wall(◦C)
T
temperature difference between the outside and
the inside surface of the underground wall/floor:
T gw−T iw(◦C)for the wall and T gf−T if(◦C)for the
floor
2.Current state of research
2.1.Basement models of DOE-2and EnergyPlus
DOE-2and EnergyPlus select from multiple basement models. This study covers the models commonly used for compliance with International Energy Conservation Code(IECC).These models are:
1)Winkelmann’s[14]model
2)Basement
2.1.1.Winkelmann’s model
In1988,Huang et al.[15]calculated perimeter conductance per perimeter foot values for slab-on-gradefloors,basements and crawl spaces using a two-dimensionalfinite difference program and pre-sented theirfindings with their paper published in The American Society of Heating,Refrigerating and Air-Conditioning Engineers (ASHRAE)Transactions.In2002,Winkelmann[14]revised the work of Huang et al.[15]in the Building Energy Simulation User News and described how to use theirfindings in a DOE-2model.The base-ment model referred to as“Winkelmann’s basement model”in this paper is based on these descriptions from Winkelmann[14].
In Winkelmann’s[14]method,it is assumed that the heat trans-fer mainly occurs in the exposed perimeter of the underground surface since this region has relatively short heatflow paths to the outside air.To model this heatflow from the exposed perimeter, an effective resistance value(R eff)is assigned to the underground Fig.1.Underground construction layers of the Winkelmann’s[14]basement model. construction.This R eff value is calculated using the area to perime-ter ratio and the perimeter conduction factor of Huang et al.[15] for that specific location and amount of insulation.The inverse of the R eff value is also assigned as the effective U-value(U eff)of the underground construction.The construction of the underground surface is then redefined such that its overall resistance value will be equal to the calculated R eff value.The new undergroundfloor construction consists of three layers as shown in Fig.1.
This new construction accounts for the thermal mass of the neighboring ground when custom weighting factors are specified in DOE-2.Underneath thefictitious insulating layer,the system is exposed to the ground temperatures provided from the weather file.These temperatures are the monthly average outside air tem-peratures delayed by3months.
2.1.2.Basement model
Basement is a preprocessor program of EnergyPlus that calculates monthly ground temperatures for underground walls andfloors of basements using a3-D numerical analysis[16].Basement was origi-nally developed by Cogil[16]in1998based on the earlierfindings of Bahnfleth[17]in1989.Cogil’s[16]model was then further modified and integrated with EnergyPlus by Clements[18]in2004.
In1989,Bahnfleth[17]conducted a parametric study for slab-on-grade GCHT using a detailed three-dimensionalfinite difference model.He found that not the perimeter length(P exp)but the area to perimeter ratio(A/P)was a proper scaling factor to correlate the average heatflux for L-shaped and rectangularfloors.He also estimated that the perimeter heat loss method(F2method)which correlates GCHT with perimeter length(P exp)may be in error by50% due to this erroneous scaling.Bahnfleth’s study[17]also showed that the thermal conductivity of the soil,ground surface boundary conditions and shading of adjacent soil are important parameters for ground coupled heat transfer.Based on hisfindings,Bahnfleth developed a new GCHT model for slab-on-grade constructions.
The mathematical basis of this new model was a boundary value problem on the three-dimensional heat conduction equation[17]. The adopted boundaries were interior slab surface,far-field soil, deep ground and ground surface.This boundary value problem was solved in Cartesian coordinates by a Fortran program that imple-mented the Patankar-Spalding[19]finite difference technique.An irregular grid of10,000cells discretized the three-dimensional domain of the model.The minimum grid spacing was4in.(0.1m) near the ground surface and slab boundaries.The user inputs were domain dimensions and grid spacings,weather datafile(TMY), soil and slab properties,ground surface properties,slab shape and size,deep ground boundary condition,evaporative loss at ground surface and building height for shadowing calculations.
In1998,Cogil[16]developed a numerical model to predict basement heat loss based on the slab-on-grade model of Bahn-fleth[17],which formed the basis of the Basement preprocessor of EnergyPlus.Similar to the slab-on-grade model of Bahnfleth,Cogil’s basement model treated the GCHT calculation problem as a bound-ary value problem on three-dimensional heat conduction equation. This model,however,had a few significant differences from Bahn-fleth’s slab-on-grade model[16–18]:
S.Andolsun et al./Energy and Buildings43(2011)1663–16751665
•It had additional boundaries such as interior surfaces of the base-ment,basement ceiling and above-ground exterior basement surfaces.
•The transient foundation and soil temperature distributions were calculated for a1-h time step using the improved Alternating-Direction-Implicitfinite difference numerical technique of Chang et al.[20].Compared to the conventional Alternating-Direction-Implicit algorithms[21,22],the method of Chang et al.[20]had greater accuracy and required less computer ing an hourly time step,however,still required4–10h of execution time on PCs of varying capability.
•The heat transfer from the above-ground exterior surfaces of the basements was taken into account.Both convective and radiative effects were modeled on the above grade portion and interior surfaces of the basement wall.
•The interior and exterior below grade foundation insulation, foundation wall and under slab gravel drainage beds were mod-eled.
In2004,Clements[18]used Fortran90to update Cogil’s[16] basement heat transfer module.He proposed and tested a link between the results of this three-dimensional module and the one-dimensional calculations of EnergyPlus.Clements[18]also pro-posed and tested a foundation scaling parameter that simulated basements using an equivalent building geometry.He added an automated grid sizing function that provided gridflexibility.He introduced a variable time step that decreased the long execu-tion times of the basement model by reducing the superfluous calculations in the far-field.Clements also added a subroutine to automatically calculate the undisturbed ground temperature pro-file for initialization purposes.
In this study,the most current version of EnergyPlus program as of September2010(version5.0.0.031)has been used.In this ver-sion,EnergyPlus is integrated with its Basement preprocessor and automatically iterates with it once.We,however,ran EnergyPlus and Basement programs separately to control the iteration process. The current state of the calculations in the Basement program is still not well understood by the simulation community due to the limited resources.
parative studies on basement heat transfer calculation methods
Basement heat transfer has long been studied by many researchers and many methods have been developed during the years.The results of the newly developed methods have also been compared with those of the earlier ones simultaneously.The stud-ies conducted by Parker[23],McDonald et al.[24],Yuill and Way [25],Krarti[26],Sobotka et al.[27]and Amjad et al.[28]are exam-ples of these comparative studies.
Parker[23]found that the Mitalas method[29],which imple-ments2-D and3-D physical models of the basement,calculates uniformly greater annual heat loss when compared to the previ-ous methods that were entirely based on1-D and/or2-D modeling such as the conduction path length method of ASHRAE[30],the F-factor method[23]and the methods of Yard et al.[31]and Akridge and Poulos[32].McDonald et al.[24]compared two vari-ations of the Latta–Boileau method[33],which is the basis of the ASHRAE’s conduction path length method[30],with the methods of Mitalas[29],Yard et al.[31],Akridge and Poulos[32],Shipp [34]and Swinton and Platts[35]and obtained significant disagree-ments among all methods for the uninsulated walls andfloors.For the well-insulated basement walls,however,they obtained agree-ment between the methods of Yard et al.[31],Mitalas[29]and Shipp[34],which are all of a similar mathematical background based on two-dimensional numerical ter,Yuill and Wray[25]compared Krarti’s[36]semi-analytical two-dimensional interzone temperature profile estimation(ITPE)method with the Mitalas method and with thefinite difference heat conduction pro-gram(ESHD)developed in the Underground Space Center[37] for heavily insulated(R=3.5m2K/W)basement walls and uninsu-lated basementfloors.They obtained good agreement in the wall andfloor heat loss in all cases except for thefloor heat loss pre-dicted by the2-D ESHD program,which was significantly lower. Krarti[26]then used his ITPE method to determine the steady-state temperature distribution for basements with different insulation configurations,compared his results with the Mitalas method and obtained good agreement.Sobotka et al.[27]compared measured data with the conduction path length method of ASHRAE,the Mita-las method,the European Standard and the two-dimensionalfinite element method program.They found that the Mitalas method shows good agreement with the measured data due to its com-bined2-D/3-D solution.Their study also showed the shortcomings of one-dimensional modeling of deep basements which results in lower predicted heat loss,especially around the basement cor-ner.Amjad et al.[28]then developed a method which numerically solves two-dimensional heat conduction problems in large calcula-tion domains using the two-dimensional transfer functions method and the substructuration technique.They also compared the results of this method with those calculated with the alternative direc-tions implicit(ADI)method and obtained good agreement with considerably reduced computation time.
parative studies on basement models of DOE-2
The basement models of DOE-2have been compared with those of other building energy simulation programs by Judkoff and Ney-mark using two test suites:(1)International Energy Agency(IEA) Building Energy Simulation Test and Diagnostic Method(BESTEST) [3];(2)Home Energy Rating System(HERS)Building Energy Simu-lation Test and Diagnostic Method(BESTEST)[4].
Using the IEA BESTEST test suite,Judkoff and Neymark[3]com-pared DOE-2with BLAST-3.0,SERIRES/SUNCODE,SERIRES-1.2,ESP, S3PAS,TRNSYS,TASE,DEROB-LTH and CLIM2000.The basement test case of the IEA BESTEST(Case990)was a building1.35m sunk into the ground[3].To model this case,DOE-2was used with ASHRAE’s conduction path length method[30]that calculates average U-values for underground walls andfloors based on their conduction path lengths through the ground to the ambient air. For the basement case of IEA BESTEST,the DOE-2ground coupling heating load varied between−40%and+47%when compared to the results of other programs for the identical conditions.The DOE-2ground coupling cooling load was33–76%lower than the results of other programs for the same cases.Due to these unresolved dis-agreements between the tested programs,the basement test case (Case990)has been the only test case of IEA BESTEST excluded from ASHRAE Standard140[2].
Using the HERS BESTEST test suite,Judkoff and Neymark[4] compared DOE-2with BLAST-3.0and SERIRES/SUNCODE.The HERS BESTEST included two basement types:(1)uninsulated basement and(2)basement with internally insulated wall.These basements were modeled as one large zone including the upper mainfloor and then as two smaller zones where the mainfloor and the basement were two separate zones.These basements were modeled using two ASHRAE GCHT calculation methods[30].Thefirst method was the perimeter heat loss calculation method,which assumes that the primary heat loss occurs from the perimeter of the ground cou-pled construction.This method uses a heat loss coefficient together with the perimeter length of the underground wall/floor to simplify the GCHT into a steady-state thermal conduction[2].The second ASHRAE GCHT calculation method was the conduction path length method.This method accounted for the effects of mass and solar radiation incident on soil and eventually led to lower thermal loads
1666S.Andolsun et al./Energy and Buildings43(2011)1663–1675
when compared to the perimeter heat loss method of ASHRAE.The results showed that,for the same basement case,DOE-2calculated 4–14%lower heating load than BLAST and10–21%lower heat-ing load than SERIRES.The basement test cases of HERS BESTEST are currently being used by Residential Energy Services Network (RESNET)to test simulation tools in comparison with DOE-2,BLAST and SERIRES results for certification as a residential code compli-ance calculator[38].
parative studies on basement models of EnergyPlus
The current literature includes comparative studies on the slab-on-grade model of EnergyPlus[2,9,12,39].EnergyPlus slab-on-grade model has been compared with those of DOE-2[9],HOT3000, SUNREL,VA114[39],BASECALC,BASESIMP,EN ISO13370,TRN-SYS,SUNREL-GC[2]and ASHRAE1052-RP Toolkit[12].This study extends the previous studies by comparing the Basement model of EnergyPlus with Winkelmann’s[14]basement model that was developed for use with DOE-2.
2.2.
3.Studies that compared EnergyPlus with DOE-2
Previous studies compared EnergyPlus with DOE-2for ground isolated[9,12]and ground coupled[9]conditions of slab-on-grade buildings.This study extends these studies by examining the ground isolated and ground coupled conditions of a low-rise code compliant house with a basement.
3.Research method
In this study,an Austin code house was modeled using Energy-Plus version5.0.0.031and DOE-2.1e version119with two basement configurations:
1)a conditioned basement(CondBsmnt house)
2)an unconditioned crawl space(UncondCrwl house)
The study was conducted using three simulation sets.In thefirst simulation set(Simulation Set I),ground isolated base-case CondB-smnt and UncondCrwl houses were modeled in accordance with the building envelope properties described in the International Energy Conservation Code(IECC)without any load component except the exterior walls.In the second simulation set(Simulation Set II),load components were added to the base-case CondBsmnt and UncondCrwl houses as described in IECC.In the third simula-tion set(Simulation Set III),the Set I and Set II houses were coupled with the ground and the changes in their HVAC electricity con-sumption with ground coupling were identified.EnergyPlus and DOE-2results for each simulation set were compared with each other.
3.1.Simulation Set I
This section describes the modeling of the base-case CondBsmnt and UncondCrwl houses in terms of building envelope and building system.
3.1.1.Building envelope
The base-case CondBsmnt and UncondCrwl houses were modeled as square based rectangular prisms topped with an unconditioned attic(Fig.2).The base-case CondBsmnt house had the dimensions of20m×20m×6m and it was submerged into the ground up to2.4m height(Fig.2a).The base-case Uncond-Crwl house had the dimensions of20m×20m×4.2m and it was immersed into the ground up to0.6m height(Fig.2b).The base-ment ceilings of the base-case CondBsmnt and UncondCrwl houses were located0.6m above the ground level to separate the main living space from the basement(Fig.2
).Fig.2.The base-case Austin code house(a)with a conditioned basement(CondB-smnt house)and(b)with an unconditioned crawlspace(UncondCrwl house).
In the base-case CondBsmnt and UncondCrwl houses,the under-ground walls andfloors were modeled as adiabatic surfaces to block the heat transfer between the ground and the basement.The ceiling facing the unconditioned attic was modeled as an adiabatic surface to block the heat transfer between the unconditioned attic and the main living space.The base-case houses did not have any windows, doors,internal load or infiltration.The base-case CondBsmnt and UncondCrwl houses were identical above the ground.They differed only in their basements and basement ceilings.Table1shows the properties of the materials used in the constructions of the base-case CondBsmnt and UncondCrwl houses.
The underground walls of the base-case CondBsmnt and UncondCrwl houses were modeled as0.2m heavy-weight concrete (CC)walls.The basementfloor of the base-case CondBsmnt house was a0.1m concrete(CC)slab.The crawlspace of the UncondCrwl house did not have a slab and was directly seeing the soil.In order to model this condition,a soil(SL)floor was assigned to the crawlspace of the UncondCrwl house with a thickness of0.16ft(0.049m), which was the minimumfloor thickness that DOE-2allowed.The basement ceilings of the CondBsmnt and UncondCrwl houses were assigned different constructions,since they faced different base-ment air temperatures.From the basement towards the main living space,the basement ceiling of the CondBsmnt house had a0.013m gypsum board(GP),0.25m wood(WD)frame with25%framing ratio,0.013m plywood(PW)and a massless layer of carpet(CP), respectively.From the crawlspace towards the main living space, the crawlspace ceiling of the UncondCrwl house had0.25m wood (WD)that framed0.099m mineral wool(IN)with25%framing ratio,0.013m plywood(PW)and a massless layer of carpet(CP), respectively.
Above the ground,the only load component of the base-case houses was the exterior walls that had an overall U-value of0.47W/m2K(0.082Btu/h ft2F).From the outside towards the inside,the exterior walls had0.076m face brick(BK),0.013m plywood(PW),0.102m soft wood(WD)that framed0.112m min-eral wool(IN)with25%framing ratio and0.013m gypsum board
S.Andolsun et al./Energy and Buildings43(2011)1663–16751667
Table1
Properties of the materials used in the building envelope model.
Conductivity Density Specific heat Resistance
W/m K Btu/h ft F kg/m3Lb/ft3J/kg K Btu/Lb F m2K/W H ft2F/Btu
BK* 1.310.75720831309200.22––
PW*0.1150.0665453412130.29––
WD*0.1150.0665133213810.33––
IN*0.0430.025*******.20––
GP*0.160.0938********.20––
CC* 1.3100.75722431408370.20––
AR*––11217014640.350.0780.440
CP*––––––0.300 1.704
SL* 1.73118421154180.1––
*Materials are adopted from the DOE-2.1e Materials Library.
(GP)respectively.The ceiling of the main living space was insu-lated from the attic side leading to a U-value of0.199W/m2K (0.035Btu/h ft2F).From the living space towards the unconditioned attic,the ceiling had0.013m gypsum board(GP)and0.254m wood (WD)that framed0.3m mineral wool(IN)with25%framing ratio, respectively.The roof was uninsulated and consisted of a massless single layer(AR),0.013m plywood(PW)and0.254m wood(WD) frame,from the outside towards the inside respectively.
3.1.2.Building system
The base-case CondBsmnt and UncondCrwl houses had identical HVAC systems but they were zoned differently.They had a direct expansion(DX)cooling system with a heat pump.The base-case CondBsmnt house had two zones:(1)an unconditioned attic and (2)a conditioned main zone.The conditioned main zone included the main living space and the basement.The base-case UncondCrwl house had three zones:(1)an unconditioned attic,(2)a conditioned main living space and(3)an unconditioned crawlspace.
In EnergyPlus,a DX cooling coil,a DX heating coil,a supply fan and a supplementary heating coil were connected to build the main air loop side of the system.In DOE-2,the Residential System(RESYS) was assigned to the base-case houses as the equivalent of the sys-tem modeled in EnergyPlus.
Both in DOE-2and EnergyPlus,the systems were autosized for the peak heating and cooling days to maintain zone air temper-atures within20◦C(68F)heating set point and25.55◦C(78F) cooling set point.Since EnergyPlus assumes no throttling range[7], the throttling range in DOE-2was set to its minimum value(0.1◦C). The cooling(minimum)and heating(maximum)design supply air temperatures were12.8◦C and48.9◦C respectively in both pro-grams.There was no preheating or precooling in the systems.The supply fan had1.5in-H2O(373.6Pa)pressure rise and0.7overall fan efficiency.The efficiency of the fan motor was0.9and it was inside the air stream.The supplementary heating was electric and was available anytime when the outside dry bulb temperature fell below20◦C.
There were two critical differences between the system inputs of EnergyPlus and DOE-2due to the modeling differences between these two programs.Thefirst difference was in the performance inputs of the heating and cooling coils.In DOE-2,the performances of the heating and cooling coils are defined using a unique input parameter,energy input ratio(EIR),which is the ratio of energy input to the load handled[42].This EIR term is the multiplicative inverse of the coefficient of performance(COP),which is defined as the ratio of the load handled to the energy input.EnergyPlus uses the COP term directly as an input parameter to define the sys-tem performance;therefore,the performance input terms of DOE-2 and EnergyPlus happen to be reverse terms.In EnergyPlus,the rated coefficient of performance(COP)was used for the cooling and the heating coils and they were4.05and3.88respectively.In DOE-2,the electric input ratio(EIR)was used for the cooling and heating coils and they were0.25and0.26respectively.The second difference between the system inputs of EnergyPlus and DOE-2was in humid-ity control.The system modeled in EnergyPlus controlled humidity both at the system and zone level;whereas the RESYS system mod-eled in DOE-2did not allow for humidity control.In EnergyPlus,the zone cooling and heating design supply air humidity ratios,the cen-tral heating and cooling design supply air humidity ratios were all set to0.008kg-H2O/kg-air.
3.2.Simulation Set II
In the second simulation set(Simulation Set II),load compo-nents were added to the base-case CondBsmnt and UncondCrwl houses modeled in thefirst simulation set(Simulation Set I).These load components were:
1)windows,doors and shades
2)lights and equipment
3)infiltration
4)heat transfer through the ceiling
This section describes the modeling of these load components in EnergyPlus and DOE-2.
3.2.1.Windows,doors and shades
Windows and doors with U-value of4.5W/m2K(0.8Btu/h ft2F) were added to the main living space of the base-case CondBsmnt and UncondCrwl houses with25%window-to-wall ratio.The win-dows were designed in Window5.2.17a(Window5)and imported into DOE-2and EnergyPlus separately.The modeled windows had two panes with1.2mm argon layer between the panes.The overall solar heat gain coefficient of the windows was0.4and all windows had shades.From30th of April until31st of October,the shading ratio was set to70%as required by the IECC2009.At other times, the shading ratio was set to85%.
3.2.2.Lights and equipment
Lights and equipment were added to the main living spaces of the CondBsmnt and UncondCrwl houses and to the basement of the CondBsmnt house.In IECC2009,overall internal gain is calculated by Equation(1)given below:
IGain=17,900+23.8×CFA+4104×N br(1) It is assumed that there arefive bedrooms in the main liv-ing spaces of the CondBsmnt and UncondCrwl houses.It is also assumed that there are no bedrooms in the basement of the CondB-smnt house.In Equation(1),the N br value was,therefore,taken as “5”for both houses.The CondBsmnt house had4305.6ft2(400m2) conditionedfloor area above-ground with a full conditioned base-ment below yielding8611ft2overall conditioned space(CFA).The UncondCrwl house had4305.6ft2(400m2)overall conditioned。