Example: marketing

Building Vintage and Electricity Use: Old Homes …

EI @ Haas WP 211 Building Vintage and Electricity Use: Old Homes Use Less Electricity in Hot Weather Howard Chong November 2010 Revised version published in European Economic Review, 56(5), 906-930 (2010) Energy Institute at Haas working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to review by any editorial board. 2010 by Howard Chong. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit is given to the source. Building Vintage and Electricity Use: OldHomes Use Less Electricity In Hot WeatherHoward G ChongDepartment of Agricultural and Resource EconomicsUniversity of California, BerkeleyNovember 22, 2010 This paper studies whether Electricity use in newer or older residential buildings risesmore in response to high temperature.

Building Vintage and Electricity Use: Old Homes Use Less Electricity In Hot Weather Howard G Chong Department of Agricultural and Resource Economics

Tags:

  Building, Electricity, Vintage, Building vintage and electricity use

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Transcription of Building Vintage and Electricity Use: Old Homes …

1 EI @ Haas WP 211 Building Vintage and Electricity Use: Old Homes Use Less Electricity in Hot Weather Howard Chong November 2010 Revised version published in European Economic Review, 56(5), 906-930 (2010) Energy Institute at Haas working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to review by any editorial board. 2010 by Howard Chong. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit is given to the source. Building Vintage and Electricity Use: OldHomes Use Less Electricity In Hot WeatherHoward G ChongDepartment of Agricultural and Resource EconomicsUniversity of California, BerkeleyNovember 22, 2010 This paper studies whether Electricity use in newer or older residential buildings risesmore in response to high temperature.

2 Peak Electricity demand occurs at the highesttemperatures which are predicted to increase due to climate change. Understand-ing how newer buildings differ from older buildings improves forecasts of how peakelectricity use will grow over time. Newer buildings are subject to stricter buildingenergy codes, but are larger and more likely to have air conditioning; hence, thecumulative effect is ambiguous. This paper combines four large datasets of build-ing and household characteristics, weather data, and utility data to estimate theelectricity-temperature response of different Building vintages. Estimation resultsshow that new buildings (1970-2000) have a statistically significantly higher temper-ature response ( , use more Electricity ) than old buildings (pre-1970). Auxiliaryregressions with controls for number of bedrooms, income, square footage, centralair conditioning, ownership, and type of residential structure partially decomposethe effect.

3 Though California has had extensive energy efficiency Building stan-dards that by themselves would lower temperature response for new buildings, thecumulative effect of new buildings is an increase in temperature response. As newbuildings are added, aggregate temperature response is predicted Codes: Q41, Q48keywords: Electricity , Temperature Response, Demand Forecast, Climate Change Impacts, Vintage -Differentiated Regulation, Building Standards, California, Load Factor, RosenfeldEffectUC Energy Institute provided graduate student support, access to utility data, and an invaluable intellectualenvironment to pursue this research. I thank Max Auffhammer, Peter Berck, Carl Blumstein, Severin Boren-stein, Lucas Davis, Koichiro Ito, Carla Peterman, Sofia Villas-Boas, and David Zilberman for many excellentdiscussions. Excellent feedback came from participants at UCEI Camp, UCEI Seminar, ARE DepartmentalSeminar, NBER Summer Institute (EEE), PG&E, USAEE/IAEE 2010 Conference, Australia National Uni-versity Microeconomics Brown Bag, and a University of Queensland seminar.

4 Financial support from theCalifornia Energy Commission is gratefully acknowledged. Any remaining errors are mine the relationship between Electricity usage and temperature, response, is important for climate change policy and long-range Electricity infrastruc-ture planning. Residential buildings are a substantial contributor to CO2 emissions. In theUS, residential buildings account for 21% of 2008 CO2 emissions (Environmental ProtectionAgency 2010), with about 50% of residential energy going to space heating and air condi-tioning (Energy Information Administration 2009). Furthermore, temperature increases fromCO2 emissions will affect Electricity demand through increased cooling loads, , air condi-tioning use. Electric power plant construction and infrastructure decisions are strongly drivenby peak Electricity demand which in California occurs during periods of highest new buildings have higher temperature response1, then the average temperature responsewill increase as new buildings are added.

5 Peak demand per household will also to reduce greenhouse gas emissions or reduce energy use often aim to decrease peakand total Electricity response is better than total Electricity use as a measure of the performance ofbuildings. As the component of Electricity usage that varies with temperature, temperatureresponse isolates factors such as the thermal performance of the Building , the size of thebuilding, and the thermostat preferences of occupants. In contrast, total Electricity useconflates these factors with appliance ownership ( , more televisions) and other factorsthat don t depend on the newer or older residential buildings in California have higher temperature responsehas not been studied using field data. California has had the most extensive energy efficiencystandards in the United States applied to new buildings. Engineering models ( , Marshalland Gorin (2007); Abrishami, Bender, Lewis, Movassagh, Puglia, Sharp, Sullivan, Tian,Valencia and Videvar (2005)), predict strong reductions in energy use (both peak and total1In this paper, temperature response is defined as the percentage increase (relative to usage on a 65 Fday)in Electricity use due to a 1 Fincrease in temperature.)

6 Higher temperature response means more incrementalelectricity I focus on temperature response, I also present comparisons of the total Electricity use across vintagein Appendix A. Unsurprisingly, new Homes use more Electricity , principally because they are ) due to these standards,ceteris paribus, but other factors can offset these increases. Thesign of the cumulative effect, measured as the difference between new and old buildings, isambiguous. I use field data to estimate the temperature response across houses of paper uses (household, monthly) field panel data on Electricity use linked to build-ing Vintage and other Building and household characteristics. Household Electricity usage(quantity) data in Riverside County, California, USA, is regressed on time series variation intemperature to estimate temperature response. Cross sectional variation in Building vintageand other characteristics at the Zip9-level or census block group-level identifies the temper-ature response by main finding is that each successive decade since 1970 has statistically significantlyincreasedtemperature response compared to older buildings (built prior to 1970).

7 Hence,average peak load is expected to increase due to population growth and ensuing new construc-tion. This exacerbates the impact of climate change on Electricity use. Auxiliary regressionsadd controls for bedrooms, income, sqft, central air conditioning ownership, and type ofresidential structure. These differ across Vintage and partially explain the increase in tem-perature response for newer buildings. With controls, 1990s Homes are estimated to have atemperature response of 8% less to 6% more than pre1970s Homes in the most organization of the paper is as follows. Section 2 presents existing related 3 presents a description of the data. Section 4 presents an econometric 5 estimates the model. Section 6 discusses results and potential mechanisms. Section7 Response and Building Vintage in Field Evidence and papers have focused on temperature response of buildings using field evidencebut ignore how buildings have changed across Vintage .

8 Aroonruengsawat and Auffhammer3(2009) examine the variation in the non-linear relationship between temperature and elec-tricity use by sixteen climate zones in California, showing that the strongest relationships arein hotter inland areas. Earlier work on temperature response with (annual, state)-level databy Desch nes and Greenstone (2008) predicts that climate change scenarios generate a 33%increase in residential energy consumption nationwidewith the current set of buildings. Newbuildings, if they perform worse than older buildings, may exacerbate this predicted ignoring Vintage effects, such studies would underestimate the impact of new and Calandri (1992) use an engineering model to estimate the impact of a Ctemperature increase, finding a 2-4% increase in Electricity use, but the study holds thebuilding stock fixed. More recent work suggests that newer buildings are more temperatureresponsive.

9 Every two years, the California Energy Commission runs a detailed simulationmodel to construct its demand forecast that includes a large mix of econometrically estimatedparameters and engineering estimates. In a recent revision, they find that air conditioningsaturation for newer buildings increased unexpectedly for both hotter (inland) and cooler(coastal) areas (Marshall and Gorin 2007).3A limitation of engineering studies is uncertainty about whether engineering parametersrepresent actual field performance. Joskow and Marron (1992) describe many factors thatcontribute to overstatement of program effectiveness. In particular, a rebound effect mayexist where occupants demand more services by responding to a decrease in the price due toefficiency (Greening, Greene and Difiglio 2000), interventions may imperfectly translate tothe field, or unexpected confounding effects could diminish or accentuate savings.

10 Althoughonly a small portion of their broader critique, they highlight the difficulty of extrapolatingfrom the laboratory to the field. In Joskow and Marron (1993), they find that the ratioof measured to estimated savings are for two 1980s retrofit programs; that is,engineering predictions overstated savings by a factor of 2 to 3. As more current evidence thatfield measurements and engineering estimates differ, Larsen and Nesbakken (2004) compare3 Their large simulation model reports aggregate result but does not explain how they model differencesbetween new and old Homes . They use an alternate but related concept of load factor, which is definedas average demand relative to peak demand. Load factor and average temperature response are inverselyrelated. They project that load factor will econometric decomposition approach to the predictions of engineering models in find that the two approaches decompose end uses quite differently.


Related search queries