1、630 2009 ASHRAEABSTRACTAn analysis of electrical billing data for single family customers of Austin Energy is performed with intent of iden-tifying home electrical use impact of the energy codes adopted by the City of Austin. Information from multiple sources is combined to classify homes by house s
2、ize, period of construc-tion, energy efficiency program participation and owner or rental property. Electrical billing information from Austin Energy is used to analyze changes in energy use by the prede-termined building characteristics for the time periods associ-ated with building construction un
3、der adopted energy codes. Analysis of the data identifies that based on energy consump-tion in 2008, adoption of the 1986 Model Energy Code (MEC) reduced energy use for homes constructed between 1985 and 1995 by 9.1 GWh, a mean of 930 kWh per house or 7.3 percent of annual energy use. Homes built be
4、tween 1996 and 2001, identified with the 1993 MEC, did not demonstrate a change in energy use when compared with homes built under the previ-ous energy code. Adoption of the 2000 International Energy Conservation Code resulted in a total energy reduction of 3.5 GWh, a mean of 790 kWh per home or 5.8
5、 percent of annual home energy use as compared to homes built under the 1993 MEC.INTRODUCTIONThe City of Austin has considered energy efficiency for new construction an important issue since the adoption of its first building energy code, the 1983 Model Energy Code (MEC), in late 1983 (Austin 2008a)
6、. Austin, as well as all Texas municipalities, is allowed to pursue more stringent building regulation than would exist at the state level as a result of home rule legislation. In the case of Texas, the state did not have an energy code prior to 2001 while Austin continued to adopt, and supplement,
7、updated energy codes such as the 1986, 1989 and 1993 MEC followed by the 2000 International Energy Conservation Code (IECC) in late 2001. Although the city used the MEC as a foundation for energy efficient construction, it incorporated more aggres-sive conservation features not covered in the MEC th
8、rough local admendments. One such example of an Austin amend-ment was an effective shading coefficient (SC) for fenestra-tion when the MEC prescriptive codes required none (CABO 1983, 1986, 1989).Due to the long history of energy efficiency ordinances by the city, combined with the availability of e
9、lectrical energy data from Austin Energy, this city is well suited to examine the energy impacts resulting from the implementation of residen-tial energy codes. It is the goal of this study leverage energy data, along with other information, to investigate the persis-tence and realization of energy
10、impacts from refinements in the residential building energy code.It is also the intent of this paper to examine the energy impacts as a potential for improvements in residential energy efficiency across the entire population. There are several cities currently exploring options in improving home ene
11、rgy effi-ciency through ordinances or consumer education at the point of sale of an existing home. These communities considering such initiatives include San Francisco, Seattle, Denver, Port-land and Berkeley as well as Austin itself. Although the administrative details for these initiatives are sti
12、ll being fleshed out, the question arises about the potential impact of minimum energy efficient standards. Improvements under consideration include attic insulation, repair or replacement of duct work and installation of solar screens; predominantly the Electrical Energy Impacts of Residential Buil
13、ding Codes for Homes in Austin, TexasJohn Trowbridge, PEAssociate Member ASHRAEJohn Trowbridge is a measurement and verification engineer at Austin Energy, a municipal utility in Austin, Texas, USA.LO-09-059 2009, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.a
14、shrae.org). Published in ASHRAE Transactions 2009, vol. 115, part 2. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.ASHRAE Transactions 631measures that form the foundation of the
15、 residential energy codes for the City of Austin.METHODOLOGYThe first step in the analysis is to group homes sharing similar building characteristics that affect energy use, spread across different periods of building code enforcement. The energy impact of the code under consideration is the differe
16、nce in the mean annualized energy use per conditioned square foot of pre-code and post code adoption housing stock. Statistical measurements of the difference in the energy use of housing stocks are employed to determine the validity of the impacts. To generate the distinct categories of housing sto
17、ck that effec-tively represent changes to the building code, data from several different sources are combined. These sources of information include: Travis Central Appraisal District (TCAD) (Travis 2008), City of Austin jurisdictional database (Austin 2008b), and three interdepartmental program reco
18、rds; the residential energy conservation, solar photovoltaic (PV) and residential Green Building databases.TCAD InformationTCAD information is combined with utility accounts through address matching to provide a means of supplement-ing billing data with construction characteristics. Pertinent inform
19、ation from TCAD includes:Year of home constructionTotal conditioned area of structureProperty description (Single Family, Duplex, Fourplex, Mobile Home)Central ACPool installationHomestead exemption (owner occupied or rental prop-erty)Market value of homeSince the focus of the study is on energy eff
20、iciency in a cooling dominated climate, homes having records indicating no central air conditioning system are eliminated from the analysis. Homes with pools, and thereby pool pumps, are also removed from the study owing to the impact in energy use a pool pump will have on the energy bill combined w
21、ith unknown parameters, such as pool size or heating type, for which to normalize energy use. The year of construction is used as the metric for determining the energy code under which the structure is built. The Texas homestead exemption allows for owner occupied residential homes to reduce the app
22、raisal value for some property taxes by filing paperwork with the tax assessors office. Rental properties are assumed to be homes without a homestead exemption in contrast to owner occupied homes that have filed for exemption. Since there exists a subset of the owner occupied population that have ne
23、glected to file for homestead exemption, these properties are further identified by comparing the last names between prop-erty owners and electrical account data where a two similar names, i.e. matching of first four letters, are considered to be an owner occupied residence.Utility Program Informati
24、onTo obtain a population of buildings that would represent a majority of the housing stock that is not significantly retro-fitted or built to a higher energy standard than minimum code requirements, three additional building types were identified from utility program databases. The first category is
25、 the resi-dences retrofitted under the Home Performance with Energy Star (HPwES) program. The utilitys HPwES program has been providing incentives for energy efficient upgrades of existing residences since 1982. The next category of high performance homes are participants in the solar PV program. Al
26、though a small base of participants, comprehensive records from the PV program make it easy to identify them. Participant information from the Austin Energy Green Builder program, a service providing consulting, resource and educa-tion for sustainable design of new and existing construction, are als
27、o identified. Removing known high performance homes from consideration allows a reasonable assumption that a majority of remaining residential homes in the population represent the quality of energy efficient construction under the respective building code.Building CodeA review of archived City of A
28、ustin ordinances is performed to identify the building code adopted and the dates of enforcement (Austin, 2008a). Although the energy codes do have performance criteria, the minimum prescriptive require-ments are used as the metric to identify significant changes in energy standards for analysis, li
29、sted in Table 1. Noted below, some wall insulation U-values are based on an overall thermal performance, including doors and windows, by mandate of local ordinance.Despite the fact the first reference to an energy code is in Chapter 52, Appendix A of the 1979 Uniform Building Code (ICBO 1979), it wa
30、s not until November of 1983 that the city council formally adopted an energy code for new construction, the 1983 MEC. Therefore, the first category of housing stock is defined by the period of 1975 to 1983, buildings with no required, minimum standard of energy efficient construction. Following the
31、 citation of the 1983 MEC in city ordinances on November 1983, a more stringent energy code is enacted less than 18 months later closely resembling the 1986 MEC (because this untitled energy code is an amendment to the existing City of Austin building code, the table cites the 1982 UBC while the pre
32、scriptive values shown are from amend-ment 800911-B). Due to this short period of time and small population, 1984 is not included in the study. Beginning March 1985, the second category of homes is identified by the 1986 MEC (CABO 1986) and 1989 MEC (CABO 1989), amended by local ordinance (Austin 88
33、0128-N, Austin 921112-B) and remains in force to the end of 1995. Adoption 632 ASHRAE Transactionsof the 1993 MEC (CABO 1993), with local amendments (Austin 950824-G), is used for the third category of homes built from 1996 to 2001 followed by the fourth group of homes built from 2002 to 2007, const
34、ructed under the requirements of the IECC 2000 (ICC 2000) with local amendment number 011129-78. References to sealed ductwork and low infiltration homes are consistently emphasized, but these measures are not tabulated since they are not quantifiable for this study.City of Austin JurisdictionTo con
35、trol for homes built under the jurisdiction of the City of Austins building code, data provided by Communi-cation and Technology Management (Austin, 2008b) is used. Austins Communication and Technology Management hosts an ftp site that houses Geographic Information System (GIS) data on properties fo
36、r the greater Austin area. Included with GIS data are specific details on the full or limited purpose annexation information for addresses. It is the annexation information that is used to create a population of properties that fall within the City of Austins building code jurisdiction through addre
37、ss matching.BILLING DATAMeter read dates for this study are from October 15, 2007 to October 14, 2008. Energy use for the residence is cumula-tive until the last month where the last bill is normalized to energy use per day. Any remaining days necessary to reach 366 days (2008 is a leap year) is the
38、n multiplied by the energy use per day of the final month to obtain an adjusted annual energy use for each residence. October is chosen as the cutoff due to the mild conditions of Austins fall thereby reducing impacts associated with heating and cooling energy on a bill-ing period adjustment. All bi
39、lling data used for analysis is based off of this annualized energy consumption and is normalized by the air conditioned area of the home.The billing records are minimally processed to check the quality of data. Simple checks are in place to ensure that extremely low energy users, homes that are uno
40、ccupied, are not included if the normalized energy use is less than 0.25 kWh per day. Since the data is based on metered information, there is no consideration of electrical transmission losses in the power lines for the final overall impacts.Home Area StrataThe final manipulation of housing populat
41、ion is deter-mined by creating three strata of residences using TCAD square foot data with the Dalenius-Hodges method (Bland 2007, EPA 2002). The Dalenius-Hodges method assigns strata boundaries based on the cumulative square root of a key parameters frequency. This is necessary to account for a tre
42、nd in large house sizes that began in 1986, see Figure 1. Energy use intensities (EUI) for larger homes, usually of multiple stories and multiple air conditioning units, built in larger quan-tities starting in the late 1980s, are found to have a lower EUI than smaller homes, consisting of primarily
43、single story and one air conditioning unit, built under the same energy code. A typical example of this discrepancy are homes built between 1996 and 2002 where the smaller homes in the first strata have a mean of 7.37 kWh/ft2(79 kWh/m2) annually while larger homes in the third strata use 5.44 kWh/ft
44、2(59 kWh/m2) annu-ally. It is therefore evident that the inclusion of house size strata is necessary to ensure that any reduction in energy use is not a result of comparing a lower EUI influenced by large homes to higher, pre code EUI of smaller homes. Although necessary for the first code compariso
45、n, the 1986 MEC, all remaining analysis is performed in the same manner for consistency.For this study, the strata parameters for the year preceding a change in building code, the base case, are imposed on the subsequent time period, the new code group, to maintain Table 1. Building Energy Code Mini
46、mum Prescriptive Requirements and Dates of EnforcementOrdinance CitationCeilingInsulation(U-value)WallInsulation(U-value)DuctInsulation(R-value)HVACEquipment EfficiencyWindowSolarPerformanceWindowU-valueEnforcement Date1973 UBC None None None None None None Nov761976 UBC None None None None None Non
47、e Jun781979 UBC None None None None None None Sep801983 MEC 0.05 0.265 5 6.5 SEER None None Nov831982 UBC 0.045 0.22 5 8.5 SEER 0.5 SC* None Mar851986 MEC 0.045 0.22 5 9 SEER 0.5 SC* None Feb881989 MEC 0.045 0.22 5 10 SEER 0.5 SC* None Feb931993 MEC 0.04 0.205* 5 10 SEER 0.5 SC* None Sep95IECC 2000
48、0.036 0.21 6.5 10 SEER 0.4 SHGC 0.75 1Dec*Local amendment requiring a window shading coefficient for fenestration facing directions from 15 North of East to 45 North of West.*Total wall thermal performance.ASHRAE Transactions 633consistency in building characteristics between the two groups. Tables
49、2, 3 and 4 show the quantity of single family homes analyzed in this study broken out by category and strata for the respective building code and the range of conditioned square feet.ANALYSISPresented in this section are the details relating to the statistical tests and methods to arrive at the final models used for determining the impacts. The most robust method for comparing differences in populations is an empirical distribu-tion function (EDF) (NIST 2006) such as the T-Test procedure (SAS 2004b, NIST 2006), but the populations evaluated using this must have a norma
