Trade-Off Between Cool Roofs and Attic Insulation in New Single-Family Residential Buildings

Size: px

Start display at page:

Download "Trade-Off Between Cool Roofs and Attic Insulation in New Single-Family Residential Buildings"

Alison James
6 years ago
Views:

Trade-Off Between Cool Roofs and Attic Insulation in New Single-Family Residential Buildings Steven Konopacki and Hashem Akbari, Heat Island Project, LBNL, Berkeley, CA. Danny Parker, FSEC, Cocoa, FL.

1 Trade-Off Between Cool Roofs and Attic Insulation in New Single-Family Residential Buildings Steven Konopacki and Hashem Akbari, Heat Island Project, LBNL, Berkeley, CA. Danny Parker, FSEC, Cocoa, FL. ABSTRACT This paper summarizes a comparative analysis of the impact of roof surface solar absorptance and attic insulation on simulated residential annual cooling and heating energy use in sixteen sunbelt climates. The buildings are single-story, single-family, new construction residences with either a gas furnace or an electric heat pump and with ducts in the attic or conditioned zone. Annual energy use is simulated with DOE- for dark and cool roofs and eleven attic insulation R-values ranging from 1 through 0. The simulations are regressed as a function of roof system conductance and roof absorptance for each heating system, duct location / insulation level, and climate. An equivalent change in conductance is calculated for a given change in absorptance from a dark to a cool roof. Equivalent attic insulation R-values are found from the conductance of the cool roof. Reductions in R-value are observed in all buildings and climates. The analysis demonstrates that a roof system with a cool roof and low attic insulation can be used as an alternative to the more conventional dark-colored roof with a high level of insulation, with a zero net change in the annual energy bill. Similar work was previously done in support of revisions to ASHRAE commercial building standard 0.1. Introduction Cool roofs have a lower solar absorptance (higher albedo 1, and contribute less to the cooling load of a building than dark-colored roofs, and thus, save energy and money by reducing cooling electricity use. In some climates, a heating energy penalty may occur due to the reduced solar load on the roof. Energy savings from cool roofs have been documented with computer simulations and measured data in residential and commercial buildings. The magnitude of the savings are dependent upon the reduction in roof absorptance, levels of attic and duct insulation, duct location, and climate. Computer simulations of residential and commercial building cooling and heating energy use in climatically diverse locations throughout the United States have shown the impact of cool roofs in reducing energy use (Akbari et al. 1; Gartland et al. 1; Konopacki et al. 1; Parker et al. 1). Cooling electricity savings have been measured in Florida from the application of white-roof coatings on several residences (Parker et al. 1) and on a strip mall (Parker et al. 1). Similarly, savings were measured in California in two medical office buildings, a retail store (Konopacki et al. 1), a residence, and two school bungalows (Akbari et al. 1). This paper summarizes a recent study conducted by Konopacki and Akbari (1) on a comparative analysis of the impact of roof solar absorptance and attic insulation on simulated residential annual cooling and heating energy use in sixteen sunbelt climates:, NM,, GA,, TX,, TX,, TX,, NV,, KY,, CA,, TN,, AZ,, NC,, CA,, UT,, VA (represents Washington, DC),, FL, and, AZ. These locations cover a wide range of climates where cool roofs are expected to save energy and money, and are areas with high growth rates in new residential construction. The residences are single-story, 1 Whensunlighthits a surfacesomeof the energyis reflected(this fractionis called the albedo= a) and the rest is absorbed(u = 1- a). Low-asurfacesbecomemuchhotterthan high-asurfaces. Trade-Off Between Cool Roofs and Atti c Insulation -1.1

2 single-family of new construction with either a gas furnace or an electric heat pump and with ducts in the attic or conditioned zone. Annual energy use is simulated with DOE- for dark and cool roofs and eleven attic insulation R-values ranging from 1 through 0. The results are then regressed as a function of roofs ystem conductance and roof absorptance for each heating system, duct location / insulation level, and climate. An equivalent change in conductance is calculated for a given change in absorptance from a dark to a cool roof. Equivalent attic insulation R-values are then found from the conductance of the cool roof. The Envelope Subcommittee of the ASHRAE Standing Standard Project Committee (SSPC) has recently voted for inclusion of reflective roofs in public review drafts for commercial building standard 0.1. Their decision was based on evidence of savings obtained from DOE- simulations as reported by Akbari et al. (1). The results presented in this paper can be used towards proposing modifications to building standard 0. for new residences, and in support of the US Environmental Protection Agency s (EPA) Energy Star@ Homes Program. Residential Building Model A single-story, single-family residential building was modeled with typical characteristics found in new constructions (Aktinson 1) and DOE national appliance energy standards (NAECA 1), as shown in Table 1. The model has two zones arranged in a non-directional floor plan: a conditioned living zone with a floor area of 100ft and an unconditioned attic zone above. The building characteristics were selected to be uniform for all locations, since the focus was on the effects of roof absorptance and attic insulation, and since, most roof, wall, and window parameters did not exhibit much local variation. The attic model was recently developed by Parker et al. (1). It calculates attic temperature, supply and return duct losses, and temperature-dependent heat conduction through the attic insulation. The roofs ystem is composed of asphalt shingles (infrared emittance 0.) attached to a 0 sloped roof deck, over a naturally ventilated unconditioned attic space (1% ceiling frame fraction ), with fiberglass insulation and 1/ drywall comprising the ceiling. The attic ventilation to floor area ratio was set at 1:00, and variable air infiltration was modeled by the Sherman-Grimsrud algorithm (Sherman 1). The cooling and heating system(s) were sized automatically by DOE- (including a sizing-ratio of 1,), and the equipment efficiencies were defined by the national energy standards as: air conditioner SEER= 10, gas furnace Tl=%, and heat pump HSPF=.. Modified part-load-ratio curves for a typical air conditioner, heat pump, and furnace were used in place of the standard DOE- curves, since they model low-load energy use more accurately (Henderson 1). DOE- Annual Cooling and Heating Energy Use Estimates Annual cooling and heating energy use were estimated with the DOE-. le building energy simulation program (BESG 10). The simulations were performed with the residential building model for both dark (ct=o.)and cool (a=o.) roofs, attic insulation R-values of 1,,,,, 1, 1,,0,, and 0, insulated (R-,,, and ) attic ducts, uninsulated attic ducts, and uninsulated conditioned zone ducts, and with Typical Meteorological Year (TMY) weather data for the sixteen previously listed locations. Local 1 residential average prices for electricity and gas (EIA 1) are shown in Table and were used to calculate total annual The ceilingframefractionaccountsforjoists, electricaljunction boxes,accessdoors,insulationvoids,etc. The albedosselectedfor these simulationscover a widerange of materials,both fresh and aged. An on-linedatabasecharacterizing someof thesematerialscanbe foundat Konopacki, et. al.

3 Table 1. Typical construction, equipment, and interior load characteristics for a new residence. construction floors single-story zones living: conditioned, attic: unconditioned floor area 100ft: conditioned orientation non-directional floor plan roof construction 1/ asphalt shingle, / plywood decking: 0 slope ceiling construction x studded frame (1%), variable fiberglass insulation, 1/ drywall insulation R-value = 1,,,,, 1, 1,, 0,, 0 wall construction brick, x studded frame (1%), R- fiberglass insulation, 1/ drywall foundation slab-on-grade windows 0ft: clear with double glazing and operable shades equipment sizing-ratio 1. cooling direct expansion: SEER = 10 heating gas furnace: q = 0., heat pump: HSPF =. distribution constant-volume forced air system with ducts located in attic (R-1,,,,) and living zone (no duct insulation): q = 0. W/cfm supply duct area = 0ft, return duct area = ft, duct leakage = 10% thermostat cooling setpoint = F, heating setpoint = 0 F (am - 10pm), setback= C F natural ventilation window operation available interior load infiltration Sherman-Grimsrud: fla = (living) fla = (attic) lighting & equipment 0. Wlft & 0. Wlft occupants Table. Local 1 residential average prices for electricity and gas (EIA 1). location,, GA, TX, TX, TX, NV NM, KY, CA, TN,, AZ NC, CA, UT, VA, FL, AZ electricity [$/lcwh] gas [$/therm] Trade-O fbetween Cool Roojl and Attic Insulation -1.

4 energy use in dollars. Annual cooling and heating electricity, gas, and total energy use for dark roofs and savings resulting from cool roofs are displayed in Table for residences with R-1 attic insulation, R- attic ducts, and both gas furnaces and electric heat pumps. The effect of the cool roof was greatest in the gas heated residence with art annual total energy savings followed by. (1),. (1),.0 (),. (),. (),. (),. (1),. (), 1. (), 1. (), 1. (), 0. (), 0. (), 0. (l), and 0. (). Although the savings were estimated with a single-story residential model, they can be applied to the top story of multi-story residences. Energy use in gas heated,, and residences with R- attic ducts are plotted in Figure 1 as a function of roof albedo and attic insulation R-value. The dark roof has an albedo of 0.1 and the savings (penalties) were estimated for a cool roof with an albedo of 0., a net change of 0.. Linear interpolation can be used to estimate savings (penalties) for other net changes in albedo regardless of the initial or final albedo for a given residence (Konopacki et al. 1: Attachment ). Simply adjust the value given by the ratio Aa/O.. As an example, the savings (penalties) associated with a cool roof with an albedo of 0. would be?h of those with an albedo of 0.. Regression Analysis The simulated annual cooling and heating energy use was expressed as a function of the overall roof system conductance and roof surface solar absorptance of the form in equation 1, E= CO+ C1U+CZU+CU(X [1] where, E is the annual cooling and heating energy use: electricity [kwh/ft], gas [therms/ft], and total [$/ft], U is the overall roof system conductance including outdoor air film [Btu/h.ft.0F], u is the roof surface solar absorptance, and Cn are regression parameter estimates. The conductance is related to the attic insulation R- value through equation, U=o.oil R where, 0.0 is frame resistive element Ginverted [Btu/hftz OFl, 0. is the cavity fraction,. is the cavity resistive element T[hft.0F/Btu] excluding insulation, and R is the attic insulation R-value [h.ft.0f/btu]. Linear regressions using eq. 1 were completed for sets of electricity, gas, and total energy use by heating system, duct location / insulation level, and climate. The analysis of variance (standard deviation of the error and R) and parameter estimates (C, C ~, C, and C) for new residences with R- attic ducts are shown for 0 total ($) energy in Table. The R values for this curve-fit ranged from 0.- for all data sets analyzed except for the uninsulated attic duct case, where R ranged from 0.1- (Konopacki & Akbari 1). Fig. 1 compares the simulated estimates with those of eq. 1 for electricity, gas, and total energy use in,, and for a new residence with gas heating and R- attic ducts. Electricresistancesupplementalheatingwasavailablefor the heatpump. U = I /(A RJ where,rtot= (I / Rfime+ 1/ RCavi(Y)-l total resistanceof the roof systemmodeledwith a parallelthermalcircuit. R (1%):outdoor air film 0. (ASHRAE 1: table 1: summer. mph), 1/ asphalt shingle 0., / plywood 0., x $%., naturallyventilatedattic.1 (ASHRAE1:table: reducedfor roof deck), x joist.,drywall0.,andindoor air film0. (ASHRAE1:table 1: averageof coolingand heating). R (%): outdoor air film 0., 1/ asphaltshingle0., / plywood0., naturallyventilatedattic.1, fiberglassinsulation (R~f drywall0., andindoorair film0.. [] 1.1- Konopacki, et. al.

5 Table. Simulated impact of roof albedo on annual cooling and heating energy use in a single-family new residence with R-1 attic insulation and R- attic ducts. Dark roof albedo is 10%. Savings (penalties) are calculated for increasing roof albedo from 0.1 to 0. (Aa = 0.). To estimate savings (penalties) for other Aa multiply savings (penalties) in this table by the ratio of Aa/O.. Total energy use is expressed in dollars and was calculated with local electricity and gas prices. electricity [kwh/ftj] gas [therms/ft ] total [$/ftl] location dark savings dark savings dark savings roof A % roof A % roof A % gas furnace heat pump , , Trade-O fibetween Cool Rooji and Attic Insulation -1.1

6 , AZ - electricity, FL electricity, VA - electricity.0..0 ~ dark: simulation IL ~, cool: simulation.0 o dark: regression 1. cool: regression. Q Q 0 u < 1.0 * oo...qg...g...g... ) u Q.Q...Q....JI.....y <>.0 ufwuq...!g-g...q...g... i- + I I I I I I I I I I I I I I I natural gas natural gas natural gas g 0.0 t i 0. g...j n... Q I I I I I total total IIIIIIIIIII,II!!!II,I total Q ) g b w Oo....g. g.....g.....g I I I I I o ? attic insulation R value attic insulation R value () attic insulation R value Figure 1. Simulation and regression estimates of annual cooling and heating energy use for a new residence with a gas furnace and R attic ducts in,, and Steriing. Electricity, gas and total energy use per sqft are presented for dark (aibedo 0.1) and cooi (aibedo 0.) roofs vs. attic insulation R-vaiues of 1,,,,, 1, 1,, 0,, and 0.

7 Table. Analysis of variance and parameter estimates from regressions of annual total ($) cooling and heating energy use (E) versus roof system conductance (U) and roof surface solar absorptance (cx)for a new residence with R- attic ducts. E($)= CO+ CIU+c. L U+CJJcx J location gas furnace heat pump variance 0. 1,00 1,00 co parameter estimates c c C Trade-~Between Cool Roofs and Attic Insulation -1.1

8 Equivalent Attic Insulation R-Values The objective of this work was to correlate energy savings from cool roofs to an equivalent reduction in the level of attic insulation. First, the total derivative (de) of eq. 1 was found. Next, the condition of a zero net change in annual cooling and heating energy use (de=o) was applied. Then, an equivalent change in roofsystem conductance (AU) from a dark to cool roof was found from equation, AU= K Au [] with decreasing absorptance (dark to cool roof Acx<O)the equivalent conductance will increase (AU>O)proportionally by the constant K defined in equation (for K>O), Cg u, K= c~+c~u,+ctx~ where, K is dependent upon initial values of U and U. The constant residence with R- attic ducts and RI of,, 1, 0,, and 0. [] K is identified in Table for a new Finally, by rearranging eq. the equivalent attic insulation R-values (R) were determined and are reported together with K in Tbl.. Reductions in R-value were observed for all climates, duct locations / insulation levels, and in both heating systems (Konopacki & Akbari 1). The highest impact for a residence with R- attic ducts was in with gas heating, where the equivalent R-value dropped from to 1 (1 w/ heat pump), followed by 1 (1), 1 (0), 1 (0), 1 (0), 1 (), 1 (), (), (), (), (1), (), 0 (0), (), (), and (). Example 1. A gas heated new residence in with R- ducts located in the attic is planned for construction with a dark roofs of ctl=0. and RI = attic insulation. This residence could alternately incorporate a cool roofg of cx=0.and a lower level of attic insulation with a zero net change in the annual energy bill. step l: Table - R=1 Example. The same residence in instead incorporates a cool roof 10 of cx=0. (Acx=-O. U1=0.01). step 1: Table s K = 0.0, step : Equation - AU= 0.01, step : Equation = R = 1 Example. The same residence in is planned with a dark roof of cxl=0. and a cool roof of cx=0. (Acx=-O.U1=0.01). step 1: Substitute parameter estimates from Table into Equation a K = 0.0, step : Equation - AU = 0.01, step : Equation - R = 1 The effect of duct location and insulation on equivalent R-values can be observed in Table. These R were based on total ($) energy use. The residence with uninsulated (R-1) attic ducts had the largest reduction in R and the most energy savings, where the smallest reduction and least savings were in the residence with conditioned zone ducts (Konopacki & Akbari 1). By comparing these two cases across climates, R decreased g A typicalnew blackasphaltshinglehas a measuredalbedoof 0,0 anddark brown0.0 (Berdahl& Bretz 1). (CRMD1). ]OSeveralrecentlydevelopedwhite-roofcoatinghavemeasuredalbedosof (Konopackiet al. 1:Att. 1). A typical new green asphaltshinglehasa measuredalbedoof 0,1 ~d white().(berdahl& Bretz 1), 1.1- Konopacki, et. al.

9 Table. The constant K and equivalent attic insulation R-values (R) based on annual total ($) cooling and heating energy use for a new residence with R- attic ducts and initial attic insulation R-values of,, 1, 0,, and 0. dark roof al = cool roof a = 0. location K R. R gas furnace heat pump , Trade-O ffbetween Cool Roofi and Attic Insulation -1.1

10 Table, Equivalent attic insulation R-values (R) based on annual total ($) cooling and heating energy use for a new residence with initial attic insulation R-values (Rl ) of 1 and, and attic duct R-values of 1,,,, 1 and and conditioned zone duct R-value of 1. dark roof a, = cool roof a, = 0. 1 L location R-1 R- R- R- R- R-1 cond R gas furnace heat pump Konopacki, et. al.

11 slightly (ARmm= at Rl=l and ~max at RI=) as wo~ld be expected with a decrease in energy savings. An exception was the electric heated residence in where AR increased by (R 1=) due to a high demand during the heating season. Conclusion This paper summarized a comparative analysis of the impact of roof surface solar absorptance and attic insulation on simulated residential annual cooling and heating energy use in sixteen sunbelt climates. The residences were single-story, single-family of new construction with either a gas furnace or an electric heat pump and with ducts in the attic or conditioned zone. Annual energy use was simulated with DOE- for dark and cool roofs and eleven attic insulation R-values ranging from 1 through 0. The simulations were regressed as a function of roof system conductance and roof absorptance for each heating system, duct location / insulation level, and climate. An equivalent change in conductance was calculated for a given change in absorptance from a dark to a cool roof. Equivalent attic insulation R-values were found from the conductance of the cool roof. The analysis demonstrated that a roof system with a cool roof and low attic insulation can be used as an alternative to the more conventional dark-colored roof with a high level of insulation, with a zero net change in the annual energy bill. Reductions in R-value were observed for all climates, duct locations / insulation levels, and in both heating systems. The highest impact for a residence with R- attic ducts was in with gas heating, where the equivalent R-value dropped from to 1 (1 w/ heat pump), followed by 1 (1), 1 (0), 1 (0), 1 (0), 1 (), 1 (), (), (), (), (1), (), 0 (0), (), (), and (). The Envelope Subcommittee of the ASHRAE Standing Standard Project Committee (SSPC) has recently voted for inclusion of reflective roofs in public review drafts for commercial building standard 0.1. The results presented in this paper can be used towards proposing modifications to building standard 0. for new residences, and in support of the US Environmental Protection Agency s (EPA) Energy Star@ Homes Program. Acknowledgement This work was supported by the U.S. Environmental Protection Agency (EPA) and the Assistant Secretary for Conservation and Renewable Energy, Office of Building Technologies of the U.S. Department of Energy, under contract No. DE-AC0SFOO0. References Akbari, H., Bretz, S., Kurn, D., and Hanford, J. 1. Peak Power and Cooling Energy Savings of High- Albedo Roofs. Energy and Buildings :-1. Lawrence Berkeley National Laboratory Report LBL-. Berkeley, CA. Akbari, H., Konopacki, S., Eley, C., Wilcox, B. Van Geem, M., and Parker, D. 1. Calculations for Reflective Roofs in Support of Standard 0.1. In ASHRAE Transactions vol. 10, pt. 1. Lawrence Berkeley National Laboratory Report LBNL-00. Berkeley, CA. Aktinson, C. 1. Lawrence Berkeley National Laboratory. Personal communication to author. New Residential Construction Characteristics. Trade-OfBetween Cool Rooj dnd Attic Instddtion -1.1

12 American Society of Heating, Refrigerating, and Air Conditioning Engineers (ASHRAE). 1. ASHRAE Handbook 1 Fundamentals. chapter., GA. Berdahl, P. and Bretz, S. 1. Preliminary Survey of the Solar Reflectance of Cool Roofing Materials. Energy and Buildings :1-1. Lawrence Berkeley National Laboratory Report LBL-00. Berkeley, CA. Building Energy Simulation Group (BESG). 10. Overview of the DOE- Building Energy Analysis Program, Version.1D. Lawrence Berkeley National Laboratory Report LBL- 1, Rev. 1. Berkeley, CA. Cool Roofing Materials Database (CRMD) Laboratory. Berkeley, CA. Lawrence Berkeley National Energy Information Administration (EIA). 1. Electric Sales and Revenue 1. DOE/EIA- 00(). Web version. Washington, DC. Gartland, L., Konopacki, S., and Akbari, H. 1. Modeling the Effects of Reflective Roofing. h Proceedings of the 1A CEEE Summer Study on Energy Efficiency in Buildings :. Lawrence Berkeley National Laboratory Report LBL-0. Berkeley, CA. Henderson, H. 1. Part Load Curves for Use in DOE- Draft report prepared for Lawrence Berkeley National Laboratory and Florida Solar Energy Center. CDH Energy Corp., Cazenovia, NY. January 1, 1. Konopacki, S. and Akbari, H. 1. Simulated Impact of Roof Solar Absorptance, Attic and Duct Insulation, and Climate on Cooling and Heating Energy Use in New Construction Single- Family Residential Buildings. Lawrence Berkeley National Laboratory Report LBNL- 1. Berkeley, CA. Konopacki, S., Akbari, H., Gartland, L., and Rainer, L. 1. Demonstration of Energy Savings of Cool Roofs. Lawrence Berkeley National Laboratory Report LBNL-0. Berkeley, CA. Konopacki, S., Akbari, H., Pomerantz, M., Gabersek, S., and Gartland, L. 1. Cooling Energy Savings Potential of Light-Colored Roofs for Residential and Commercial Buildings in U.S. Metropolitan Areas. Lawrence Berkeley National Laboratory Report LBNL-. Berkeley, CA. National Appliance Energy Conservation Act of 1 (NAECA). 1. Parker, D., Huang, J., Konopacki, S., Gartland, L., Sherwin, J., and Gu, L. 1. Measured and Simulated Performance of Reflective Roofing Systems in Residential Buildings. in ASHRAE Transactions vol. 10, pt. 1., GA. Parker, D., Sonne, J., and Sherwin, J. 1. Demonstration of Cooling Savings of Light Colored Roof Surfacing in Florida Commercial Buildings: Retail Strip Mall. Florida Solar Energy Center Report FSEC-CR- -. Cocoa, F1. Sherman, M., Wilson, D., and Kiel, D. 1. Variability in Residential Air Leakage. Measured Air Leakage in Buildings ASTM STP-0. Philadelphia, PA Konopacki, et. ul.

Lawrence Berkeley National Laboratory Lawrence Berkeley National Laboratory

Lawrence Berkeley National Laboratory Lawrence Berkeley National Laboratory Title Energy impacts of heat island reduction strategies in the Greater Toronto Area, Canada Permalink https://escholarship.org/uc/item/4w2091fk