Econometric Models For Major Residential Energy End-Uses

CREEDAC
Canadian Residential Energy End-use Data and Analysis Centre
Econometric Models for Major Residential Energy End-uses

(CREEDAC-1999-04-05)
Submitted to: Cristobal Miller Policy Development and Analysis Division Efficiency and Alternative Energy Branch Natural Resources Canada 580 Booth Street Ottawa, Ontario Canada K1A 0E4
CREEDAC Project Team: Alan S. Fung Merih Aydinalp V. Ismet Ugursal
April 1999
CREEDAC, Dalhousie University 5269 Morris Street, Halifax, Nova Scotia, Canada B3J 1B6 Tel: (902) 494 6183 Fax: (902) 494 3165 E-mail:creedac@dal.ca http://www.dal.ca/daltech/creedac
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table of Contents
TABLE OF CONTENTS ................................................................................................ 2 LIST OF TABLES .......................................................................................................... 3 LIST OF FIGURES ........................................................................................................ 5 EXECUTIVE SUMMARY ............................................................................................. 6 1. 2. INTRODUCTION ................................................................................................... 9 REGRESSION MODEL ........................................................................................ 11 2.1 OVERVIEW ...................................................................................................... 11 2.2 GENERAL FORM OF THE MODEL EQUATIONS ........................................................ 12 2.3 DETERMINATION OF THE ELASTICITIES .................................................................. 14 2.4 MODEL EQUATIONS FOR DIFFERENT END-USES..................................................... 15 2.4.1 Household Appliance model Equation ......................................................... 15 2.4.2 Space Heating Model equation .................................................................... 16 2.4.3 DHW Heating Model equation .................................................................... 16 2.4.4 Combined Space and DHW Heating model equation .................................. 17 3. DATA SET............................................................................................................ 18 3.1 OVERVIEW .......................................................................................................... 18 3.2 PHYSICAL AND DEMOGRAPHIC CHARACTERISTICS OF HOUSEHOLDS ....................... 18 3.3 HOUSEHOLD APPLIANCE CHARACTERISTICS......................................................... 19 3.4 WEATHER CHARACTERSITICS ............................................................................... 20 3.5 HOUSEHOLD ENERGY CONSUMPTION DATA ......................................................... 21 3.5.1 Disaggregation of Energy Consumption data ............................................... 22 3.5.1.1. Disaggregation of Energy Consumption in Natural Gas Heated Houses . 23 3.5.1.2. Disaggregation of Energy Consumption in All-Electric Houses.............. 24 3.6 AVERAGE PROVINCIAL FUEL PRICE ..................................................................... 25 4. RESULTS AND DISCUSSION............................................................................. 27 4.1 4.2 4.3 4.4 4.5 5. 6. OVERVIEW .......................................................................................................... 27 HOUSEHOLD APPLIANCE ENERGY CONSUMPTION MODEL .................................... 29 SPACE HEATING ENERGY CONSUMPTION MODEL ................................................. 32 DHW HEATING ENERGY CONSUMPTION MODEL ................................................. 34 COMBINED SPACE AND DHW HEATING ENERGY CONSUMPTION MODEL .............. 37
CONCLUSION...................................................................................................... 64 REFERENCES AND BIBLIOGRAPHY ............................................................... 66
PAGE 2
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
List of Tables
Table 1: Household End-use Energy Consumption Elasticities calculated using 20% significance criterion ....................................................................................................... 8 Table 2: List of model equations ................................................................................... 12 Table 3: List of model equations developed................................................................... 12 Table 4: Average Current and Lagged Provincial Fuel Prices ........................................ 26 Table 5: Data Statistics for Appliance Energy Consumption Model for Natural Gas Heated Households........................................................................................................ 40 Table 6: Data Statistics for Appliance Energy Consumption Model for Electric Heated Households........................................................................................................ 41 Table 7: Appliance Energy Consumption Model Summary for Fossil Fueled Households ................................................................................................................... 42 Table 8: Appliance Energy Consumption Model Summary for Electric Heated Households ................................................................................................................... 43 Table 9: Appliance Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households ................................................................................................. 44 Table 10: Appliance Energy Consumption Model Elasticities Estimates for Electric Heated Households........................................................................................................ 45 Table 11: Data Statistics for Natural Gas Space Heating Energy Consumption Model in Natural Gas Heated Households ................................................................................ 46 Table 12: Data Statistics for Electric Space Heating Energy Consumption Model in Electric Heated Households........................................................................................... 47 Table 13: Natural Gas Space Heating Energy Consumption Model Summary for Natural Gas Fueled Households..................................................................................... 48 Table 14: Electric Space Heating Energy Consumption Model Summary for Electric Heated Households........................................................................................................ 49 Table 15: Natural Gas Space Heating Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households ............................................................... 50 Table 16: Electric Heating Energy Consumption Model Elasticities Estimates for Electric Heated Households........................................................................................... 51
PAGE 3
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 17: Data Statistics for Natural Gas DHW Heating Energy Consumption Model in Natural Gas Heated Households ................................................................................ 52 Table 18: Data Statistics for Electric DHW Heating Energy Consumption Model in Electric Heated Households........................................................................................... 53 Table 19: Natural Gas DHW Heating Energy Consumption Model Summary for Natural Gas Fueled Households..................................................................................... 54 Table 20: Electric DHW Heating Energy Consumption Model Summary for Electric Heated Households........................................................................................................ 55 Table 21: Natural Gas DHW Heating Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households ............................................................... 56 Table 22: Electric DHW Heating Energy Consumption Model Elasticities Estimates for Electric Heated Households ..................................................................................... 57 Table 23: Data Statistics for Natural Gas Space and DHW Heating Energy Consumption Model in Natural Gas Heated Households ............................................... 58 Table 24: Data Statistics for Electric Space and DHW Heating Energy Consumption Model in Electric Heated Households............................................................................ 59 Table 25: Natural Gas Space and DHW Heating Energy Consumption Model Summary for Natural Gas Fueled Households ............................................................... 60 Table 26: Electric Space and DHW Heating Energy Consumption Model Summary for Electric Heated Households ..................................................................................... 61 Table 27: Natural Gas Space and DHW Heating Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households ............................................. 62 Table 28: Electric Space and DHW Heating Energy Consumption Model Elasticities Estimates for Electric Heated Households ..................................................................... 63
PAGE 4
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
List of Figures
Figure 1: Typical monthly natural gas consumption in natural gas heated households.... 23 Figure 2: Typical monthly electricity consumption in air-conditioned electric heated houses ........................................................................................................................... 24
PAGE 5
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
EXECUTIVE SUMMARY
A comprehensive econometric regression analysis was carried out on the three major components of the residential end-use energy consumption, namely space heating energy consumption, domestic hot water (DHW) heating energy consumption and appliance energy consumption, to determine their long- and short-term price and income elasticities. The analysis was carried out on the 1993 Canadian housing stock. Three main data sources were utilized in the analysis: (1) Survey of Household Energy Use (SHEU, 1993), (2) Energy billing records for a subset of SHEU data, and (3) Energy Statistics Handbook (Statistics Canada, 1998), Electric Power Statistics (Statistics Canada, 1992), and Canadian Economic Observer (Statistics Canada, 1998). SHEU provided most of the demographic and physical housing characteristics while its subset of billing records provided the total annual household fuel consumption data for both natural gas and electricity. Both current and lagged average provincial fuel prices were obtained and/or estimated using Statistics Canadas Energy Statistics Handbook, Electric Power Statistics, and Canadian Economic Observer.
Log-linear variable price and variable income elasticity energy demand model, which was first proposed by Betancourt (1981) and later used by Donnelly and Diesendorf (1985) and Douthitt (1989), was adopted and modified to determine the price and income elasticities of the major components of the residential end-use energy consumption in natural gas and electric heated households. Model equations that incorporate demographics, weather and equipment characteristics as explanatory variables were developed, and price and income elasticities were derived from the model equations. The price, income and expenditure elasticities thus derived are presented in Table 1.
Both the long- and the short-term own fuel price elasticities were found to have the expected negative sign, implying that fuel price has a negative effect on energy consumption (i.e. as fuel price increases, energy consumption decreases) for all types of
PAGE 6
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
end-use energy consumption and fuel types. The magnitude of the short-term price elasticities were found to be significantly less than unity, indicating that the impact of price increase on end-use energy consumption was quite inelastic. The magnitude of the long-term price elasticities were found to be negative and above unity, indicating that in the long-term consumers would respond to a one percent increase in fuel price by reducing energy consumption by more than one percent.
Income elasticity on end-use energy consumption was found to be statistically insignificant in most cases, and the ones that proved to be significant were very inelastic, i.e. the absolute magnitudes of the elasticities were very close zero. Household energy expenditure elasticities for all end-uses and fuel types proved to be positive indicating that the energy consumption for all three major residential end-uses (appliance, space heating and domestic hot water heating) behave like normal goods. All cross price effects were proved to be statistically significant with negative sign, except that for electric DHW heating which was insignificant. This indicated that consumers would have a similar response on energy demand to substitute price change as to own price change.
PAGE 7
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 1: Household End-use Energy Consumption Elasticities calculated using 20% significance criterion
End-use Energy Consumption Elasticities Appliance Energy Consumption Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity Space Heating Energy Consumption Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity Cross Price Effect DHW Heating Energy Consumption Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity Cross Price Effect Space + DHW Heating Energy Consumption Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity Cross Price Effect All Natural Gas Heated Household -0.341 -1.746 0.009 1.115 -0.246 -0.904 -0.005 1.791 -0.793 -0.534 -2.510 N/A 1.197 -0.366 -0.303 -1.036 N/A 1.733 -0.701 All Electric Heated Household -0.594 -1.208 N/A 0.830 -0.467 -2.804 N/A 1.755 -0.606 -0.548 -1.213 0.022 0.860 N/A -0.419 -1.694 N/A 1.292 -0.257
Note: N/A indicates the variables are statistically not significant at 20% level
PAGE 8
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
1.
INTRODUCTION
In 1996 the end-use energy consumption in Canada was about 7630 PJ (NRCan, 1998), making Canada one of the highest per capita energy consumers in the world. The carbon dioxide emission associated with this energy consumption was about 416 Mt. Mostly owing to its northerly location and the prevalence of single family housing, residential energy consumption was about 1450 PJ, representing 19% of the total energy consumption in Canada. The carbon dioxide emission associated with the residential energy consumption was about 72 Mt.
With the Kyoto Protocol signed in December 1997, developed countries committed to legally binding greenhouse gas (GHG) emission reductions of at least five percent by 2008 to 2012. Canada promised a six percent reduction below 1990 emission levels by 2010. Canada's commitment represents a reduction in greenhouse gas emissions of at least 19 percent below what they would be without the agreement. To meet this challenging commitment, Canada has to evaluate and exploit every feasible measure to reduce the energy consumption and GHG emissions while maintaining its economic growth.
Residential energy consumption in Canada is primarily for space heating (61%), followed by domestic hot water heating (21%), appliances and lights (17%), and finally, cooling, which is negligibly small (0.4%) (NRCan, 1998). End-use energy consumption in the residential sector could be reduced by improving the building envelope characteristics (better insulation, better windows and doors, tighter buildings), and by using higher efficiency lighting, appliances, domestic hot water (DHW) heating and space heating, ventilating and air-conditioning (HVAC) equipment. Reducing the end-use energy consumption and switching to less carbon-intensive fuels for space and domestic hot water heating would result in reduced carbon dioxide emissions from the residential sector. However, the energy consumption pattern of consumers is a result of many complex and interrelated socioeconomic and technical factors as well as consumers energy conservation consciousness. Thus, an effective and efficient energy conservation
PAGE 9
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
policy requires accurate and comprehensive estimates of residential energy demand parameters. These parameter estimates are among the most important inputs into informed policy decisions. In turn, accurate estimation of energy demand parameters requires realistic modeling of the consumers energy demand behavior, detailed data on energy consumption, and careful treatment of any econometric problems created by the model and the data used.
In this report, econometric residential end-use energy models are developed to identify income and price elasticities of major residential energy end-uses. Using the models, the impact of various socioeconomic and physical housing characteristics on the residential energy consumption in Canada are also evaluated. In addition, some conclusions are made regarding possible strategies for reducing energy consumption from the residential sector.
P A G E 10
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
2.
2.1 OVERVIEW
REGRESSION MODEL
The objective of this study is to develop price and income elasticities for energy demand in the Canadian residential sector. The analytical approach is based primarily on the work done by Betancourt (1981), Donnelly and Diesendorf (1985) and Douthitt (1989). Ideally, the analysis should provide both price and income elasticities of energy demand for each of the three major residential energy end-uses; namely, space heating, domestic hot water heating, and appliance energy end-uses.
It is expected that each one of the three major residential energy end-uses poses a distinct
consumption profile for different fuel types and equipment characteristics. Consequently, each different end-use is expected to have a distinct demand equation based on the fuel type and equipment characteristic. Thus, three sets of model equations are developed here: (i) appliance, (ii) space heating, (iii) DHW heating. Additionally, a fourth set of model equations are developed for combined space and DHW heating since these two end-uses generally utilize the same fuel.
The major fuel types considered in this study are natural gas, electricity and oil. Thus, a set of four equations (one for each end use; i.e. appliance, space heating, DHW heating, combined space and DHW heating) were to be developed for each of the three fuel types as shown in Table 2. However, as it is explained in Section 3.5, the number of oil heated households in the database is not sufficient to conduct a statistical analysis; therefore, oil is not included in the analysis. On the other hand, two sets of equations would be needed for households heated with electricity: (i) all-electric households in regions where natural gas is available, and (ii) all-electric households in regions where natural gas is not available. However, it is not possible to categorize the households in the database according to this criterion because the exact locations of the houses are not known, consequently, availability of natural gas is not known either. Therefore, a single set of equations were developed for all of the electric heated households.
P A G E 11
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
To facilitate the evaluation of long-term price elasticity as suggested by Poyer and Williams (1993), a set of model equations were developed by excluding the lagged fuel price from the variable list used in the regressions. The long-term price elasticities can be estimated by dividing the short-term estimates by unit minus the estimated coefficient of the lagged variable. In conclusion, a total of 16 model equations were developed as shown in Table 3.
Table 2: List of model equations

End-use Energy Consumption Model Appliances DHW Heating Space Heating Combined Space & DHW Heating Electric Heated Households Natural Gas Heated Households
Table 3: List of model equations developed

End-use Energy Consumption Model Appliances DHW Heating Space Heating Combined Space & DHW Heating Electric Heated Natural Gas Heated House holds House holds Model I Model II Model I Model II (No Lag) (with Lag) (No Lag) (with Lag)
2.2
GENERAL FORM OF THE MODEL EQUATIONS
The general stochastic econometric variable elasticity demand model, first proposed by Betancourt (1981) and later modified by Donnelly and Diesendorf (1985) to yield unitless measures of elasticity, is given in Equation (1):
n Q i = a 0 Pia m X m Y a 2 E a 3 Pja 4 Pka 5 X a n
(1)
P A G E 12
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
where
Qi = Total end-use energy consumption (GJ or kWh) a0, a1, . an = Variable coefficients Y = Household income ($) E = Total expenditure on fuel or energy ($) Pi and Pi,-1 = Average current and lagged chosen fuel prices ($/GJ) Pj, Pk = Substitute fuel prices ($/GJ) Xn = Other socioeconomic and engineering variables Xm = Other socioeconomic and engineering variables
Vector for other variables Xn and Xm includes all other pertinent variables. These are given below categorized in groups of related parameters (units given in parentheses): 1) Structural Variables: Number of floors (-): STOREY Total Heated floor area (m2): AREA Basement area (m2): BASE Number of windows (-): WIN Number of doors (-): DOOR Number of skylights (-): SKY 2) Weather and Temperature for space heating: Average indoor temperature (C): TEMP Average heating degree days (C-day): HDD 3) Human Capital Factors: Household size (no. of people): HHS Overall heating equipment efficiency (%): EFF Total appliance energy consumption (GJ): APP 4) Domestic Hot Water Heating: Number of aerators and low-flow shower heads (-): AERATOR Average ground temperature (C): GT Hot water tank size (liters): TANK Presence of tank insulation (0 or 1): INSUL 5) Cooling:
P A G E 13
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Central air-conditioning usage in the cooling season (% of time): AC_USAGE Average cooling degree days (C-day): CDD 6) Appliances: Dishwasher load (no. of loads/year): DW Clothes washer load (no. of loads/year): WA Clothes dryer load (no. of loads/year): DRY Cooking range/oven age (years): RANGE_A Presence of cooktop (0 or 1): COOKTOP Freezer age (years): FREEZER_A Freezer size (ft3): FREEZER_S First frost-free fridge age (years): FF1_A First frost-free fridge size (ft3): FF1_S First non frost-free fridge age (years): NFF1_A First non frost-free fridge size (ft3): NFF1_S Second frost-free fridge age (years): FF2_A Second frost-free fridge size (ft3): FF2_S Second non frost-free fridge age (years): NFF2_A Second non frost-free fridge size (ft3): NFF2_S Number of light bulbs (-): LIGHT Presence of furnace fan (0 or 1): FFAN
2.3
DETERMINATION OF THE ELASTICITIES
Variable coefficients in the general model equation can be determined by multi-variate linear regression conducted on the available data once Equation (1) is transformed by taking the natural logarithm of both sides: ln(Q i ) = A 0 + a m X m ln( Pi ) + a 2 ln( Y ) + a 3 ln( E) + a 4 ln( Pj ) + a 5 ln( Pk ) + a n ln( X n ) (2)
Once the variable coefficients are determined, the price and income (and other parameters) elasticities on energy demand can be determined using the classical
P A G E 14
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
economic elasticity definition. The general and price elasticities derived from Equation (1) can be expressed as: Q Q Q Q X Q X = X X Q P Q P = = amXm P P Q
EX =
(3)
EP =
(3a)
Since the general elasticity, shown in equation (3), provides the measure of percent change in Q for one percent change in X, it is a unitless measure of elasticity. On the other hand, Equation (3a) does not provide such a unitless measure of elasticity as in Equation (3), since it has the unit of parameter Xm. Thus, it is necessary to transform the exponent variable, Xm, in Equation (3a) into a unitless variable. The method of transforming the exponent variables by dividing them by their corresponding standard deviations, as proposed by Donnelly and Diesendorf (1985), is used in this study.
2.4
MODEL EQUATIONS FOR DIFFERENT END-USES
To identify the variables to be included in the model equations and the final form of the model equations, various model equations were tested. As a result of these tests, the variables and the form of the equations were determined. The resulting equations are presented below.
2.4.1 HOUSEHOLD APPLIANCE MODEL EQUATION The model equation for appliance end-use energy consumption is given in Equation (4). There is no substitute price in the equation because electricity is predominantly the only fuel choice for appliances.
P A G E 15
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Qi = a 0 Pi
( a1Pi , 1 +a 2HDD+a 3GT)
Y ( a 4HHS) E a 5 EFFa 6 TEMPa7 STOREYa8

(4)
n AREAa 9 BASEa10 X a n
where variables Xn for appliance energy consumption model are:

a n1 n Xa WA a n 2 DRY a n 3 FREEZER _ Sa n 4FREEZER _ A n = DW
FF1 _ S a n 5FF1 _ A FF2 _ Sa n 6FF 2 _ A NFF1 _ Sa n 7 NFF1 _ A NFF2 _ S a n 8 NFF2 _ A RANGE _ A a n 9 AC _ USAGE a n 10CDD LIGHTa n11 e a n12COOKTOP e a n13FFAN 2.4.2 SPACE HEATING MODEL EQUATION
(4a)
The model equation for space heating end-use energy consumption is given in Equation (5). The appliance energy consumption variable, APP is a proxy for internal heat gain.
( a 1Pi , 1 + a 2 HDD + a 3GT )
Q i = a 0 Pi
Y (a 4 HHS ) E a 5 EFF a 6 TEMP a 7 STOREY a 8 AREA a 9
n BASE a10 Psa11 APP a12 X a n
(5)
where variables Xn for space heating energy consumption model are:

a n1 n Xa SKY a n 2 WIN a n 3 n = DOOR
(5a)
2.4.3 DHW HEATING MODEL EQUATION The model equation for space heating end-use energy consumption is given in Equation (6): Q i = a 0 Pi
( a 1Pi , 1 + a 2 HDD + a 3GT )
Y (a 4 HHS ) E a 5 EFF a 6 TEMP a 7 STOREY a 8 AREA a 9
n BASE a10 Psa11 X a n
(6)
where variables Xn for DHW heating energy consumption model are:
P A G E 16
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
a n1 n Xa AERATOR a n 2 DW a n 3 WA a n 4 e a n 5INSUL n = TANK
(6a)
2.4.4 COMBINED SPACE AND DHW HEATING MODEL EQUATION The model equation for combined space heating and DHW heating end-use energy consumption is given in Equation (7)
( a 1Pi , 1 + a 2 HDD + a 3GT )
Q i = a 0 Pi
Y (a 4HHS ) E a 5 EFF a 6 TEMP a 7 STOREY a 8
n AREA a 9 BASE a10 Psa11 APP a12 X a n
(7)
where variables Xn for combined space and DHW heating energy consumption model are:
a n1 n Xa SKY a n 2 WIN a n 3 TANK a n 4 AERATOR a n 5 DW a n 6 n = DOOR
WA a n 7 e a n 8 INSUL
(7a)
The equation contains only one substitute price and several building and DHW heating equipment specific parameters. There is only one substitute price in the model equation because most homeowners have only one alternative fuel choice. 1
Only 31 households out of a total of 320 electric heated households have natural gas available in their areas. Thus, the only alternative fuel choice for these electric heated households is fuel oil. Most of the natural gas heated households are located in regions where oil is not a major player in residential home heating market due to significantly lower price of natural gas compared to oil.
P A G E 17 CREEDAC A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
3.
3.1 OVERVIEW
DATA SET
To construct the necessary data set for the development of the econometric model, three main sources of data were used in this study: (i) Survey of Household Energy Use (SHEU, 1993), (ii) Energy billing records for a subset of SHEU data, and (iii) Energy Statistics Handbook (Statistics Canada, 1998), Electric Power Statistics (Statistics Canada, 1992), and Canadian Economic Observer (Statistics Canada, 1998). SHEU (1993) contains comprehensive data physical and demographic data on 10,982 households across Canada. 8,767 of these 10,982 records are for low-rise single family dwellings. These include single detached and attached dwellings. The energy billing data for about 3,000 of the 8,767 single family dwellings are also available from a SHUE data file. Since this study focuses on low-rise dwellings, physical, demographic and energy billing data from these 3000 households were used here. Both current and lagged average provincial fuel prices were obtained and/or estimated using the Energy Statistics Handbook, Electric Power Statistics, and Canadian Economic Observer published by Statistics Canada.
3.2
PHYSICAL AND DEMOGRAPHIC CHARACTERISTICS OF HOUSEHOLDS
The values for most of the variables used in the parameter vector Xn of the models given in Equations (4) - (7) were obtained directly or indirectly from the SHEU data as given below (the column numbers refer to the SHEU micro-data file): 1) Structural Variables: Number of floors/storeys: Column 181. Total heated area: Column 189 (mean value of each category is use). Basement area: Column 192 (mean value of each category is used). Number of windows: Column 231, 232, 234, 235, 237, 238, 240, 241. Number of doors: Column 210, 211, 213, 214, 216, 218.
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Number of skylights: Column 227, 228, 229. 2) Weather and Temperature: Average indoor temperature: Column 178, 179, 180. 3) Human Capital Factors: Household size: Column 8. Household income: Column 355 (mean value of each category is used). Overall heating equipment efficiency: Column 142: (mean value of each category is used. 100%, 150%, and 250% are assigned for electric resistance, air-source heat pump, and water/ground source heat pump heated households, respectively.) 4) Domestic Hot Water Heating: Number of aerators and low-flow shower heads: Column 327, 329. Hot water tank size: Column 323. Presence of DHW tank insulation: Column 324.
3.3
HOUSEHOLD APPLIANCE CHARACTERISTICS
In order to capture the effect of appliance stock holding, its characteristics and usage on energy demand, data on major household appliances were included in the appliance energy demand equation. The following data from SHEU were used: 1) Dishwasher: Number of loads: Column 75 (value is multiplied by 52 to obtain number of loads/year). 2) Clothes washer: Number of loads: Column 100, 101 (values are multiplied by 26 and added to obtain number of loads/year). 3) Clothes dryer: Number of loads: Column 114, 115 (values are multiplied by 26 and added to obtain number of loads/year). 4) Freezer: Age of freezer: Column 79 (mean value of each category is used).
P A G E 19
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Size of freezer: Column 83 (mean value of each category is used). 5) First Refrigerator: Age of Refrigerator: Column 42 (mean value of each category is used). Size of Refrigerator: Column 46 (mean value of each category is used). Frost free or manual defrosted: Column 48. 6) Second Refrigerator: Age of Refrigerator: Column 43 (mean value of each category is used). Size of Refrigerator: Column 47 (mean value of each category is used). Frost free or manual defrosted: Column 49. 7) Cooking equipment: Electric range/oven age: Column 56 (mean value of each category is used). Presence of electric cooktop: Column 53. 8) Lighting: Number of light bulbs: Column 331, 332, 334, 335, 345, 346, 347. 9) Furnace fan: Presence of furnace fan: Column 138. 10) Central air conditioning: Central air-conditioning usage: Column 300: assign 0, 25, 50, 75, and 100 percent usage to response 1 to 5, respectively.
3.4
WEATHER CHARACTERSITICS
Weather plays a major role on overall residential energy consumption. Ideally, both longterm average and temporal heating degree-days (HDD) should be included in the model development. Long-term average heating degree-days would be expected to have an influence on price elasticity since consumers are likely to be energy conscious in colder climates due to high energy expenditures. On the other hand, temporal heating degreedays mainly influence the temporal space heating energy consumption because of its direct influence on the amount of heat loss through the building envelope.
P A G E 20
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
It was decided not to include the temporal heating degree-days in this study because of the time and cost associated in obtaining temporal degree-day information. Cooling degree-days have an effect on air-conditioning energy consumption, therefore long-term average cooling degree-days were used in the model.
3.5
HOUSEHOLD ENERGY CONSUMPTION DATA
To formulate the energy demand model equations, actual data on overall as well as disaggregated end-use energy consumption for houses are needed. A subset of SHEU database contains one year of energy consumption billing records from utility companies. There are a total of 133, 972, and 1424 households with oil, natural gas, and electricity billing records, respectively. To ensure data quality, a number of data screenings were performed. The following procedures were employed to screen the data: No household was accepted if supplemental space heating is different from the main space heating fuel. For example, if wood or propane was indicated as the supplemental space heating fuel in a natural gas/electric/oil heated house, then this household was rejected. Only households that use the same fuel for both space and DHW heating were accepted because the number of households in the database that use different fuels for space and DHW heating is 18, which is not sufficient to conduct a statistical analysis. Only households with complete twelve month fuel billing records were accepted. This was essential for disaggregating the space heating energy consumption from DHW heating energy consumption for fossil fuel heated houses, and for disaagregating the space heating energy consumption from other end-uses in electrically heated houses. After applying this screening process, only 22 oil, 249 natural gas, and 320 electric heated households were left to be used in the analysis.
P A G E 21
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
The data set of 22 samples for oil heated houses is unacceptably low to carry out a meaningful statistical analysis with a high number of explanatory variables. Moreover, an inspection of these 22 household files revealed that all are from one province. This renders the fuel price invariant in all of these 22 households, making the regression process on oil heated households even more pointless. Thus, oil heated households are not included in this work.
3.5.1 DISAGGREGATION OF ENERGY CONSUMPTION DATA To formulate the energy demand model equations, it is necessary to disaggregate the energy consumption data obtained from energy bills into three major components, i.e. space heating, DHW heating and appliances energy consumption. To accomplish this, it is commonly assumed that that there is no space heating requirement during the summer months of July and August. Thus, the energy consumption in these two months can be assumed to be for DHW heating and appliances (including lighting) only (see Figure 1). Furthermore, it is assumed that the energy consumption for DHW heating and appliances is independent of the time of the year (i.e. constant over the whole year). This second assumption is clearly questionable since water supply temperature, daylight hours, indoor temperature, etc. all have impact on DHW and appliance energy consumption. However, in the absence of disaggregated data, the inaccuracy involved with this assumption is tolerated. The disaggregation of the energy consumption data was accomplished as described below.
P A G E 22
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Typical Monthly Household Natural Gas Consumption Profile
Jan
Feb
Mar
Apr
May
June
July
Aug
Sept
Oct
Nov
Dec
Month
Figure 1: Typical monthly natural gas consumption in natural gas heated households
3.5.1.1. Disaggregation of Energy Consumption in Natural Gas Heated Houses Appliance energy consumption: The total annual appliance energy consumption is available from the total annual electricity billing. DHW heating energy consumption: The average monthly natural gas consumption for DHW heating was estimated from the natural gas billings for July and August. Space heating energy consumption: The annual space heating energy consumption is estimated by subtracting the estimated DHW heating energy consumption from the total annual natural gas billing record.
P A G E 23
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
3.5.1.2. Disaggregation of Energy Consumption in All-Electric Houses Appliance and DHW heating energy consumption: The average monthly electrical consumption for appliances and DHW heating (combined) was estimated from the electricity billings for July and August. In air-conditioned households, July and August energy consumption includes electricity consumption for air-conditioning (see Figure 2). Therefore, in air-conditioned houses, data from shoulder months of June and September were used to represent the average monthly electrical consumption for appliances and DHW heating.
Typical Monthly Household Electricity Consumption Profile

3000
Electricity Consumption (kWh)
2500
2000
1500
1000
500
0 Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec Month
Figure 2: Typical monthly electricity consumption in air-conditioned electric heated houses To disaggregate the electrical energy consumption for DHW heating and appliances, the following approach was used: It was assumed that the average ratio of appliance energy consumption to that of DHW heating energy load is constant in all houses. The average efficiency of natural gas heated DHW heaters is 55%. The ratio of appliance energy consumption to DHW heating energy
P A G E 24
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
load in natural gas heated houses was calculated using this average efficiency and the estimated appliance and DHW heating consumption values as follows: R APP DHWload = ApplianceE nergyConsu mption 0.55DHWconsumptio n (8)
This ratio was found to be 1.51. The average efficiency of electrically heated DHW heaters is 83%. Thus, using the ratio calculated using Equation 8 and this average efficiency, the ratio of electricity consumption of appliances to the electricity consumption for DHW heating was calculated as follows: ApplianceE nergyConsu mption = (0.83)(1.51) = 1.25 DHWconsumptio n (9)
Using this ratio, the total energy consumption for appliances and DHW heating was disaggregated into its two components. Space heating energy consumption: The annual space heating energy consumption is estimated by subtracting the estimated appliance and DHW heating energy consumption from the total annual electricity billing records.
3.6
AVERAGE PROVINCIAL FUEL PRICE
It was not possible to obtain the actual fuel price, tax, rate structure, and basic service charge information for each household in the database since the exact location of the households is not available; therefore, average provincial fuel prices were used in this study. Average provincial natural gas prices for residential uses were obtained directly from Statistics Canadas Energy Statistics Handbook (1998). Since electricity prices are not given in Energy Statistics Handbook, electricity price data were estimated using data obtained from Electric Power Statistics (1992) multiplied by the consumers price index from Canadian Economic Observer (1998). The average current and lagged provincial
P A G E 25
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
fuel prices used in this study are presented in Table 4. The total household energy expenditure for each household was calculated using the billing records and the average provincial fuel prices. Table 4: Average Current and Lagged Provincial Fuel Prices
Province NFLD PEI NS NB QUE ON MAN SAS AB BC Lagged Price Current Price Electricity Natural gas Oil Electricity Natural gas Oil (cents/kWh) (cents/m^3) (cents/liter) (cents/kWh) (cents/m^3) (cents/liter) 4.70 N/A 39.70 4.71 N/A 39.60 11.70 N/A 35.50 11.69 N/A 36.00 7.70 N/A 36.50 7.69 N/A 36.50 5.60 N/A 38.20 5.79 N/A 38.70 4.80 28.07 37.70 4.87 28.12 37.00 7.00 20.60 36.40 7.47 20.99 38.00 4.70 20.04 41.00 4.73 20.70 42.50 5.90 15.90 36.10 6.24 16.49 35.70 5.20 12.67 N/A 5.37 14.59 N/A 4.60 19.06 40.40 4.78 19.75 41.40
P A G E 26
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
4.
4.1 OVERVIEW
RESULTS AND DISCUSSION
Eight end-use energy consumption models were developed and analyzed. These are: - For households heated with natural gas: . Space heating energy consumption, . Appliance energy consumption, . DHW heating energy consumption, . Combined space heating and DHW heating energy consumption; - For households heated with electricity: . Space heating energy consumption, . Appliance energy consumption, . DHW heating energy consumption, . Combined space heating and DHW heating energy consumption.
To facilitate the evaluation of long-term price elasticity suggested by Poyer and Williams (1993), another eight sets of model regressions without lagged price (Pi,-1) variables were also performed. These eight sets of model equations were denoted as Model I while the set of model equations with lagged fuel prices are denoted as Model II. All of the regression analyses were performed using Version 9 of SPSS (1998) computer statistical software package. For each model equation analysis, two statistical analyses were run using SPSS: (i) "As Entered": The first scenario was run using the As Entered feature of SPSS to determine the coefficients of all of the entered variables regardless of their statistical significance. The results of this scenario are denoted as As Entered in the summary tables 7, 8, 13, 14, 19, 20, 25, and 26. (ii) "Backward Elimination": The second scenario was run using Backward elimination" feature of SPSS to determine the coefficients of the variables with the 20 percent significance level option. This relatively high significance level option of 20 percent was used for the backward elimination method to ensure that
P A G E 27
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
most of the important variables (such as the lagged price, substitute price, and income) were included in the final model equations so that price and income elasticities as well as cross-price effects on end-use energy consumption can be identified. It was however found that most of these variables are statistically significant at or below 5 percent significance level. The results of this second scenario were denoted as Up to 20 Percent in the summary tables 7, 8, 13, 14, 19, 20, 25, and 26.
The statistics on the variables used in each one of the energy end-use model equation are presented in Tables 5, 6, 11, 12, 17, 18, 23, and 24. The coefficients of the variables, the t-statistics and the R2 values are presented in Tables 7, 8, 13, 14, 19, 20, 25, and 26. The estimated short-term and long-term price elasticities, income elasticities, and cross-price effects are presented in Tables 9, 10, 15, 16, 21, 22, 27, and 28. To represent price elasticities for households that face average, higher than average, and lower than average initial fuel prices, price elasticities were estimated at (i) sample mean, (ii) sample mean plus its standard deviation, and (iii) sample mean minus its standard deviation, respectively. At up to 20 percent significance level, the adjusted R2 values (i.e. goodness of fit) range from 0.376 to 0.987. All R2 values are above 0.64 except that for natural gas DHW heating, which has an adjusted R2 of 0.376. The adjusted R2 values are surprisingly high for combined space and DHW heating energy consumption models. The adjusted R2 values are 0.949 and 0.987 for natural gas and electric heated households, respectively. These adjusted R2 values are much higher than those reported by Douthitt (1989) for a similar study of the Canadian housing stock. This could be attributed to the fact that estimated, instead of actual, household space heating energy consumption data were used in his analyses.
Some interesting observations can be made from comparisons of the data sets of natural gas heated and electric heated households:
P A G E 28
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
- Natural gas heated households, in general, consume more energy for all end-uses, including electric appliances (higher energy consumption for both space and DHW heating energy end-uses is to be expected since natural gas has lower overall fuel utilization efficiency), - On average, natural gas heated households have higher household income and bigger home. - Natural gas heated households experience colder climate (higher heating degree-days) in winter and warmer climate (higher cooling degree-days) in summer. - Indoor temperature setpoint in natural gas heated households is higher than those of electric heated households. - Natural gas heated households have a greater number of light bulbs.
4.2
HOUSEHOLD APPLIANCE ENERGY CONSUMPTION MODEL
Results from the appliance end-use energy consumption model are presented in Tables 5 to 10. The discussion of the results here is limited to the results for the "up to 20%" significance level. All economic factors used in the models proved to be statistically significant except the lagged price in the electricity heated household model. Appliance energy consumption proved to be positively related to household energy expenditure and income, and negatively related to price and heating degree-days. Other appliance variables that proved to be significant are dishwasher load, clothes dryer load, refrigerators, and air conditioning usage. As expected, all of these variables, except the first non frost-free refrigerator, are positively related to energy consumption. This could be attributed to the fact that non frost-free refrigerators consume less energy than their frost-free counterparts since they do not employ defrost-cycles. As expected, both estimated short- and long-term price elasticities were negative. The short- and long-term elasticities, evaluated at the sample mean, were found to be 0.341 and 1.746 for natural gas heated households, and 0 .594 and 1.208 for electric heated households, respectively. Natural gas heated households respond to price changes less
P A G E 29
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
than the electric heated households in the short-term. This trend is reversed in the longterm price responses. Since electric heated households face higher fuel prices and have higher household energy expenditures, in the short- term they are expected to respond to price changes more than natural gas heated households. The elastic nature of the longterm price elasticities indicate that consumers tend to become more energy conscious if they are faced with long-term price increases. The higher long-term price elasticity for the natural gas heated households may indicate that these households would consider switching to natural gas fueled appliances (such as gas cooktop/oven/range and dryer) if they faced long-term electricity price increases.
Household energy expenditure elasticities were found to be significant and positive, indicting that appliance energy consumption behaves as a normal good. However, the magnitude was greater for natural gas heated households (1.115 versus 0.83.)
Income elasticities were found to be significant only for natural gas heated households. The income elasticity for this group of households was positive and very inelastic at 0.009.
The final econometric model equations for appliance end-use energy consumption for both natural gas and electric heated households at 20 percent significance level are presented in Equations 10 to 13.
Natural gas heated household without lagged fuel price: Q i = 40.98Pi( 1.12 E 1HHD 8.75 E 2GT ) Y (5.75 E 3HHS ) E 9.68 E 1
n TEMP 3.73 E 1STOREY 1.93 E 1 BASE 9.81E 2 X a n
(10)

1.26 E 2 n Xa DRY 3.40 E 2 NFF1 _ S 1.59 E 3NFF1 _ A n = DW
AC _ USAGE 7.33 E 5CDD LIGHT1.18E 1e 3.39 E 1FFAN
(10a)
P A G E 30
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Natural gas heated household with lagged fuel price:

( 5.76 E 2 Pi , 1 2.05 E 2 HHD )
Q i = 7.92Pi
Y ( 3.54 E 3HHS ) E 1.12 E
n STOREY 1.64 E 1BASE 8.68 E 2 X a n
(11)

1.00 E 2 n Xa DRY 2.70 E 2 FF2 _ S1.45E 3FF 2 _ A n = DW
NFF1 _ S 2.32 E 3 NFF1 _ A AC _ USAGE 5.33E 5CDD LIGHT 7.91E 2
(11a)
Electric heated household without lagged fuel price: Q i = 8.23Pi( 4.73E 2 HHD 4.28 E 2 GT ) Y (9.27 E 3HHS ) E 8.30 E 1
n EFF 4.12 E 1STOREY 1.29 E 1 BASE 6.86 E 2 X a n
(12)

1.53 E 2 n Xa DRY 2.13E 2 NFF1 _ S 2.53E 3NFF1 _ A n = DW
NFF2 _ S1.68 E 3 NFF 2 _ A
(12a)
Electric heated household with lagged fuel price: Q i = 8.23Pi( 4.73E 2 HHD 4.28 E 2 GT ) Y (9.27 E 3HHS ) E 8.30 E 1
n EFF 4.12 E 1STOREY 1.29 E 1 BASE 6.86 E 2 X a n
(13)

1.53 E 2 n Xa DRY 2.13E 2 NFF1 _ S 2.53E 3NFF1 _ A n = DW
NFF2 _ S1.68 E 3 NFF 2 _ A
(13a)
4.3
SPACE HEATING ENERGY CONSUMPTION MODEL
The results for the space heating end-use energy consumption model are presented in Tables 11 to 16. Major economic factors, such as substitute and lagged prices, income,
P A G E 31
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
and household energy expenditure, proved to be statistically significant. However, substitute fuel price for electric heated households was not significant. This could partially be explained by the fact that oil is not truly an alternative fuel for electricity due to its higher capital requirement, marginally lower price (in most provinces), and inconvenience. Space heating energy consumption is positively related to household energy expenditure, house area, and heating degree-days, and negatively related to lagged price, income, and total household appliance consumption. Other housing variables proved to be significant are number of storeys, total heated area, total basement area, number of doors, windows, and skylights. As to be expected, variables such as number of doors, windows, and skylights, are positively related to space heating energy consumption since these house features provide less thermal resistance against heat loss than solid wall assemblies. Also as expected, both average ground temperature and total appliance energy consumption variables proved to be negatively related to space heating energy consumption. This can be explained by the fact that higher ground temperature reduces heat loss through basement while higher total appliance energy consumption increases the amount of internal heat gain in the house, lowering its overall space heating requirement. Space heating equipment efficiency variables in both cases are found to be not significant statistically.
As expected, both estimated short- and long-term price elasticities are negative. The short- and long-term elasticities, evaluated at sample mean, were found to be 0.246 and 0.904 for natural gas heated households, and 0.467 and 2.804 for electric heated households, respectively. Natural gas heated households respond to fuel price changes less than electric heated households in both the short- and the long-term. Since electric heated households face higher fuel prices and have higher household energy expenditure, they were expected to respond to price changes more than natural gas heated households in the short-term. The very elastic nature of the long-term price elasticities in electric heated households indicate that consumers tend to become extremely energy conscious if they are faced with long-term price increases; or, they may supplement their space heating energy requirements with a less expensive fuel such as wood.
P A G E 32
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Household energy expenditure elasticities are found to be significant and positive, indicting that appliance energy consumption behaves like normal goods. However, the magnitude is larger for natural gas heated households (1.791 versus 1.755).
Income elasticities are found only to be significant for natural gas heated households. The income elasticity for this group of households is positive and very inelastic at 0.005.
Cross-price effects for both natural gas and electricity space heating energy consumption are statistically significant and negative. The cross-price effect coefficient values are estimated to be 0.793 and 0.606 for natural gas and electricity, respectively. For every percent increase in electricity price, there would be a 0.793 percent decrease in natural gas space heating energy consumption. Similarly, for every percent increase in oil price, there would be a 0.606 percent decrease in electricity space heating energy consumption.
The final econometric model equations for space heating end-use energy consumption for both natural gas and electric heated households at 20 percent significance level are presented in Equations 14 to 17.
Natural gas heated household without lagged fuel price: Q i = 0.15Pi( 3.5 E 2 HHD 5.1E 2 GT ) Y ( 1.6E 3HHS ) E1.8E STOREY 8.3E 2
n AREA 9.5 E 2 BASE 3.2 E 2 Ps8.2 E 1 APP 9.4 E 1 X a n
(14)

3.5 E 2 n Xa n = WIN
(14a)
Natural gas heated household with lagged fuel price: Q i = 0.069Pi

( 5.7 E 2 Pi , 1 +1.1E 2 HHD )
Y ( 2.0 E 3HHS ) E 1.8E
n STOREY 8.2 E 2 AREA 9.0 E 2 BASE 3.1E 2 Ps7.9 E 1 APP 9.2E 1X a n
(15)
P A G E 33
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES

4.7 E 2 n Xa n = WIN
(15a)
Electric heated household without lagged fuel price: Q i = 0.0011Pi( 1.3 E 1HHD 1.5 E 1GT ) E 1.7 E TEMP1.1E 1
n STOREY 6.0 E 2 AREA 4.4 E 2 Ps7.1E 1 APP 7.7 E 1X a n
(16)

1.3 E 1 n Xa WIN 3.0 E 2 n = SKY
(16a)
Electric heated household with lagged fuel price: Q i = 0.019Pi

( 4.0 E 2 Pi , 1 2.5 E 2 GT )
E 1.8E TEMP1.0 E 1
n STOREY 6.2 E 2 Ps 6.1E 1 APP 7.9 E 1 X a n
(17)

3.0 E 2 n Xa SKY 9.7 E 2 n = DOOR
(17a)
4.4
DHW HEATING ENERGY CONSUMPTION MODEL
The results for DHW heating end-use energy consumption models are presented in Tables 17 to 22. Major economic factors, such as substitute and lagged prices, income, and household energy expenditure proved to be statistically significant. Similar to the space heating energy consumption model, the substitute fuel price for electric heated households is insignificant. DHW heating energy consumption is positively related to household energy expenditure and household income, and negatively related to lagged price. Other DHW related variables that proved to be significant are DHW tank size, ground temperature, and number of dishwasher loads. As expected, the ground temperature is negatively related to DHW heating energy consumption.
P A G E 34
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Again, both estimated short and long-term price elasticities proved to be negative. The short- and long-term elasticities, evaluated at sample mean, are 0.534 and -2.510 for natural gas heated households, and 0.548 and 1.213 for electric heated households, respectively. Natural gas heated households respond to fuel price changes less than electric heated households in the long-term, whereas for the short-term, for every percent increase in DHW fuel price, there would be a about half of a percent decrease in energy consumption for both natural gas and electric heated households. The very large magnitude of long-term fuel price elasticity for natural gas heated households may indicate that consumers are very sensitive to fuel price on DHW heating energy consumption.
Household energy expenditure elasticities are significant and positive, indicting that appliance energy consumption behaves like a normal good. However, the magnitude is larger for natural gas heated households (1.197 versus 0.860.)
Income elasticities only significant for electric heated households. The income elasticity for this group of households is positive and very inelastic at 0.022.
The cross-price effect for natural gas DHW heating energy consumption is statistically significant and negative. The cross-price effect coefficient value is 0.366 for natural gas heated DHW, which means that for every percent increase in electricity price (substituted price for natural gas), there would be a 0.366 percent decrease in natural gas DHW heating energy consumption.
The final econometric model equations for DHW heating end-use energy consumption for both natural gas and electric heated households at 20 percent significance level are presented in Equations 18 to 21.
Natural gas heated household without lagged fuel price:
P A G E 35
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Q i = 1.08Pi( 7.02 E 2 HHD 6.78 E 2 GT ) E 1.18 E TEMP 4.83E 1

n Ps 4.66 E 1 X a n
(18)

6.01E 1 n Xa n = TANK
(18a)
Natural gas heated household with lagged fuel price: Q i = 0.23Pi

( 7.91E 2 Pi , 1 )
n E1.20 E TEMP 4.82 E 1Ps3.66 E 1 X a n
(19)

5.65 E 1 n Xa n = TANK
(19a)
Electric heated household without lagged fuel price: Q i = 0.12Pi( 5.24 E 2 HHD 4.84 E 2GT ) Y (9.70 E 3HHS ) E 8.60 E 1
n STOREY 1.06 E 1 X a n
(20)

1.51E 2 n Xa n = DW
(20a)
Electric heated household with lagged fuel price: Q i = 0.12Pi( 5.24 E 2 HHD 4.84 E 2GT ) Y (9.70 E 3HHS ) E 8.60 E 1
n STOREY 1.06 E 1 X a n
(21)

1.51E 2 n Xa n = DW
(21a)
P A G E 36
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
4.5 COMBINED SPACE AND DHW HEATING ENERGY CONSUMPTION MODEL The results for combined space and DHW heating end-use energy consumption models are presented in Tables 23 to 28. Major economic factors, such as substitute and lagged prices, and household energy expenditure, proved to be statistically significant. However, income effects in both cases are not significant. For natural gas heated households, the combined space and DHW heating energy consumption is positively related to household energy expenditure, total heated area, heating degree-days, number of windows, DHW tank size, and number of clothes washer loads are negatively related to lagged price, substitute price, total household appliance consumption, and extra DHW tank insulation. On the other hand, the combined space and DHW heating energy consumption is positively related to household energy expenditure, indoor temperature, and number of storeys, and negatively related to lagged price, substitute price, number of doors, number of windows, number of clothes washer loads, and total household appliance energy consumption.
As expected, both estimated short- and long-term price elasticities are negative. The short and long-term elasticities, evaluated at sample mean, were found to be 0.303 and 1.036 for natural gas heated households, and 0.419 and 1 .694 for electric heated households, respectively. Natural gas heated households respond to fuel price changes less than electric heated households in both short- and long-term. Similar to the space heating energy consumption case, electric heated households face higher fuel prices and have higher household energy expenditures; therefore, they are expected to respond to price changes more than natural gas heated households in the short-term. The very elastic nature of the long-term price elasticities in the electric heated households indicate that consumers would tend to become extremely energy conscious if they are faced with longterm price increases.
Household energy expenditure elasticities are significant and positive, indicting that appliance energy consumption behaves like normal goods. The magnitude for natural gas heated households is larger than that for electric heated households (1.733 versus 1.292).
P A G E 37
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Income elasticities are statistically insignificant for both natural gas and electric heated households.
Cross-price effects for both natural gas and electricity combined space and DHW heating energy consumption are statistically significant and negative. The cross-price effect coefficients are 0.701 and 0.257 for natural gas and electricity, respectively. This indicates that for every percent increase in electricity price, there would be a 0.701 percent decrease in natural gas consumption. Similarly, for every percent increase in oil price, there would be a 0.257 percent decrease in electricity consumption. The adjusted R2 values for combined space and DHW energy consumption are quite highat 0.987 and 0.949 for natural gas and electricity, respectively.
The final econometric model equations for combined space heating end-use energy consumption for both natural gas and electric heated households at 20 percent significance level are presented in Equations 22 to 25.
Natural gas heated household without lagged fuel price: Q i = 0.25Pi( 6.02 E 2 HHD 7.44 E 2GT ) Y (1.13E 3HHS ) E1.73E TEMP 1.03E 1STOREY 3.87 E 2
n AREA 2.93E 2 Ps7.42 E 1 APP 8.41E 1 X a n
(22)
where variables Xn for space and DHW heating energy consumption model are:
1.07 E 1 n Xa TANK 5.38 E 2 n = SKY
(22a)
Natural gas heated household with lagged fuel price:

( 5.48 E 2 Pi , 1 +1.04 E 2 HHD )
Q i = 0.1Pi
E1.73E TEMP 1.04 E 1
n STOREY 3.23 E 2 AREA 3.06 E 2 Ps7.01E 1APP 8.30 E 1 X a n
(23)
P A G E 38
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
1.08 E 1 n Xa WIN 1.61E 2 TANK 6.13 E 2 n = SKY
WA 7.01E 3 e 3.03 E 2 INSUL
(23a)
Electric heated household without lagged fuel price: Q i = 0.6Pi( 8.33 E 2 HHD 9.84 E 2 GT ) Y ( 8.82 E 4 HHS ) E1.29 E TEMP 7.59 E 2
n STOREY 4.20 E 2 AREA 3.26 E 2 Ps 4.85E 1APP 2.96 E X a n
(24)
1.88 E 2 n Xa DW 1.65E 2 n = AERATOR
(24a)
Electric heated household with lagged fuel price: Q i = 0.078Pi

( 3.05 E 2 Pi , 1 9.11E 3 HHD 2.08 E 2 GT )
E 1.29 E TEMP 4.64 E 2
n STOREY 3.33E 2 Ps2.57 E 1 APP 2.98 E 1 X a n
(25)
1.40 E 2 n Xa WIN 1.50 E 2 WA 3.68 E 3 n = DOOR
(25a)
P A G E 39
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 5: Data Statistics for Appliance Energy Consumption Model for Natural Gas Heated Households
Variables App E Con (kWh) Energy Exp. ($) Income ($) HHSize Old price ($/GJ) Price ($/GJ) HDD (C-day) GdTemp (C) SHEff (%) Temp (C) No. of Storey Hted Area (m^2) BsmtSize (m^2) DW-Total Wash-Total Dryer-Total Age-free Size-free (ft^3) FF-1-Age FF-1-Size (ft^3) NFF-1-Age NFF-1-Size (ft^3) FF-2-Age FF-2-Size (ft^3) NFF-2-Age NFF-2-Size (ft^3) Range Age Cooktop (0/1) CDD (C) AC usage (%) Light Ffan (0/1)
Name Q EXP IN HHS P-1 P HDD GT EFF TEMP STOREY AREA BASE DW WA DRY FREEZER-A FREEZER-S FF1-A FF1-S NFF1-A NFF1-S FF2-A FF2-S NFF2-A NFF2-S RANGE-A COOKTOP CDD AC-USAGE LIGHT FFAN
Count 271 271 230 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271 271
Max 39688.0 2864.6 100000.0 8.0 32.5 32.5 6562.0 12.3 77.5 23.3 3.0 278.7 223.0 780.0 780.0 780.0 28.0 25.0 23.0 22.0 23.0 19.0 23.0 22.0 23.0 19.0 23.0 1.0 238.0 100.0 132.0 1.0
Min 3901.0 699.3 6000.0 1.0 12.8 13.1 3007.0 4.8 57.5 15.0 1.0 46.5 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 19.0 0.0 10.0 0.0
Average 10207.4 1282.4 47467.4 3.1 16.2 16.7 5513.4 6.4 63.3 19.8 1.3 124.0 96.3 174.6 308.9 268.1 12.4 13.4 8.2 14.8 1.9 1.6 2.0 1.9 3.9 2.7 10.4 0.0 140.8 17.4 40.1 1.0
Median 8956.0 1199.8 45000.0 3.0 15.6 16.1 5889.0 6.1 64.0 20.0 1.0 116.1 97.5 156.0 260.0 208.0 13.0 16.0 9.0 15.0 0.0 0.0 0.0 0.0 0.0 0.0 9.0 0.0 155.0 0.0 37.0 1.0
StDev 5145.3 398.3 25198.9 1.3 5.0 5.0 734.1 1.8 6.6 1.5 0.4 45.4 32.5 189.0 192.3 190.5 9.1 6.9 6.4 6.0 5.7 4.8 5.6 5.2 7.8 5.4 6.7 0.2 57.2 25.8 19.0 0.1
25 Percentile 6716.0 1008.2 27500.0 2.0 13.1 13.1 5345.0 4.8 57.5 19.0 1.0 116.1 77.4 0.0 156.0 130.0 4.0 11.0 2.0 15.0 0.0 0.0 0.0 0.0 0.0 0.0 5.0 0.0 94.0 0.0 27.5 1.0
75 Percentile 11645.5 1461.0 70000.0 4.0 16.4 17.3 5920.0 6.6 64.0 21.0 1.5 118.4 111.5 312.0 455.0 364.0 18.0 16.0 13.0 19.0 0.0 0.0 0.0 0.0 0.0 0.0 13.0 0.0 189.0 25.0 49.0 1.0
P A G E 40
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 6: Data Statistics for Appliance Energy Consumption Model for Electric Heated Households
Variable Name App E Con (kWh) Energy Exp. ($) Income ($) HHSize Old price ($/GJ) Price ($/GJ) HDD (C-day) GdTemp (C) SHEff (%) Temp (C) No. of Storey Hted Area (m^2) BsmtSize (m^2) DW-Total Wash-Total Dryer-Total Age-free Size-free (ft^3) FF-1-Age FF-1-Size (ft^3) NFF-1-Age NFF-1-Size (ft^3) FF-2-Age FF-2-Size (ft^3) NFF-2-Age NFF-2-Size (ft^3) Range Age Cooktop (0/1) CDD (C) AC usage (%) Light Ffan (0/1)
Q EXP IN HHS P-1 P HDD GT EFF TEMP STOREY AREA BASE DW WA DRY FREEZER-A FREEZER-S FF1-A FF1-S NFF1-A NFF1-S FF2-A FF2-S NFF2-A NFF2-S RANGE-A COOKTOP CDD AC-USAGE LIGHT FFAN
Count 320 320 263 320 320 320 320 320 305 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320 320
Max 19423.7 3246.8 100000.0 9.0 32.5 32.5 7930.0 11.3 250.0 25.0 3.0 278.7 278.7 780.0 780.0 780.0 28.0 25.0 23.0 22.0 23.0 19.0 23.0 19.0 23.0 15.0 23.0 1.0 189.0 100.0 109.0 1.0
Min 1599.7 490.6 6000.0 1.0 12.8 13.1 3007.0 2.7 100.0 15.0 1.0 46.5 13.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 19.0 0.0 5.0 0.0
Average 6783.6 1535.9 40667.3 3.0 15.1 15.5 5077.2 7.3 101.1 19.1 1.3 116.4 94.9 102.1 340.4 248.4 9.4 12.6 7.5 14.4 1.6 1.9 0.8 0.7 1.8 1.3 9.0 0.1 108.6 1.0 32.7 0.9
Median 6403.3 1487.2 37500.0 3.0 15.6 16.1 4771.0 7.7 100.0 19.7 1.0 116.1 98.0 0.0 312.0 195.0 9.0 16.0 6.5 15.0 0.0 0.0 0.0 0.0 0.0 0.0 9.0 0.0 125.0 0.0 30.0 1.0
StDev 2699.0 447.4 24397.7 1.3 1.8 1.9 904.4 1.4 10.3 2.1 0.4 46.2 33.5 148.7 218.2 183.5 7.9 7.7 6.0 6.2 4.6 5.1 3.6 3.2 5.5 3.9 7.0 0.3 39.1 7.8 16.5 0.3
25 Percentile 4809.1 1244.0 22500.0 2.0 15.6 16.1 4709.0 7.6 100.0 17.7 1.0 74.3 74.3 0.0 156.0 104.0 1.0 11.0 2.0 15.0 0.0 0.0 0.0 0.0 0.0 0.0 3.8 0.0 97.0 0.0 20.0 1.0
75 Percentile 8291.5 1806.1 55000.0 4.0 15.6 16.1 4884.0 7.7 100.0 20.7 1.5 118.4 116.1 208.0 546.0 364.0 13.0 16.0 10.8 19.0 0.0 0.0 0.0 0.0 0.0 0.0 13.0 0.0 138.0 0.0 43.0 1.0
P A G E 41
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 7: Appliance Energy Consumption Model Summary for Fossil Fueled Households
Variable Name LN_CON LN_EXP LN_EFF LN_TEMP LN_STORE LN_AREA LN_BASE LN_DW_L LN_WA_L LN_DRY_L LN_LIGHT HDD_LN_P GT_LN_P HHS_LN_I A_N_S_FR ALN_S_F1 ALN_SN1 ALN_S_F2 ALN_SN2 CDD_LN_U CT_I FAN_I LN_RAN_A OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (kWh) ($) (-) (C) (-) (m^3) (m^3) (-) (-) (-) ($) ($/GJ) ($/GJ) ($) (ft^3) (ft^3) (ft^3) (ft^3) (ft^3) (%) (-) (-) (-) ($/GJ)
Variable Description Ln (Appliance Con.) Ln (Household energy exp) Ln (Heating equipment efficiency) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (Basement area) Ln (No. of dish washer load) Ln (No. of clothes washer load) Ln (No. of clothes dryer load) Ln (No of light blubs) HDD*Ln (Price) Ground Temp*LN (Price) Household size*Ln (Income) Freezer Age*Ln (Size) First Frostfree Fridge Age* Ln(Size) First non-Frostfree Fridge Age* Ln(Size) Second Frostfree Fridge Age* Ln(Size) Second non-Frostfree Fridge Age* Ln(Size) CDD*Ln (A/C Usage) Cooktop Indicator (0 or 1) Furance Fan Indicator (0 or 1) Ln (Range/oven age) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t 9.77E-01 -1.13E-01 2.64E-01 -1.79E-01 -2.58E-02 -9.69E-02 -1.10E-02 -1.62E-02 4.23E-02 1.12E-01 -1.08E-01 -8.30E-02 5.65E-03 -4.47E-05 -1.47E-03 -2.70E-03 1.94E-03 8.51E-04 6.59E-05 9.95E-02 3.51E-01 3.32E-02 4.452 0.706 0.675 13.81 -0.63 1.25 -2.66 -0.38 -2.37 -1.48 -0.67 2.11 2.52 -12.13 -9.49 3.07 -0.06 -1.28 -1.97 1.67 0.93 3.19 1.07 1.39 1.40 3.797 9.68E-01 3.73E-01 -1.93E-01 -9.81E-02 -1.26E-02 3.40E-02 1.18E-01 -1.12E-01 -8.75E-02 5.75E-03 14.62 1.13E+00 -2.00E-02 1.84 4.72E-02 -3.15 -1.47E-01 -4.86E-02 -2.60 -7.43E-02 -1.75 -8.39E-03 -5.88E-03 2.16 2.88E-02 2.80 8.12E-02 -12.81 -1.79E-02 10.37 4.05E-03 3.31 3.72E-03 -4.45E-04 -9.35E-04 -1.44 -3.19E-03 1.92E-03 8.15E-04 3.70 4.99E-05 1.18E-01 1.36 1.25E-01 3.10E-02 -5.75E-02 4.525 1.828 0.753 0.725 16.24 -0.12 0.24 -2.35 -0.78 -1.96 -1.23 -0.26 1.55 1.97 -1.07 0.25 2.17 -0.69 -0.88 -2.52 1.79 0.96 2.60 1.38 0.53 1.42 -6.21 1.577 1.12E+00 17.39
-1.64E-01 -8.68E-02 -1.00E-02 2.70E-02 7.91E-02 -2.05E-02 3.54E-03
-2.94 -2.52 -1.55 1.87 2.06 -3.49 2.22
-1.59E-03
-2.32E-03 1.45E-03 5.33E-05
-2.31 1.46 2.90
7.33E-05 3.39E-01
3.713 0.696 0.678
-5.76E-02 2.070 0.746 0.732
-13.39 4.537
P A G E 42
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 8: Appliance Energy Consumption Model Summary for Electric Heated Households
Variable Name LN_CON LN_EXP LN_EFF LN_TEMP LN_STORE LN_AREA LN_BASE LN_DW_L LN_WA_L LN_DRY_L LN_LIGHT HDD_LN_P GT_LN_P HHS_LN_I A_N_S_FR ALN_S_F1 ALN_SN1 ALN_S_F2 ALN_SN2 CDD_LN_U CT_I FAN_I LN_RAN_A OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (kWh) ($) (-) (C) (-) (m^3) (m^3) (-) (-) (-) ($) ($/GJ) ($/GJ) ($) (ft^3) (ft^3) (ft^3) (ft^3) (ft^3) (%) (-) (-) (-) ($/GJ)
Variable Description Ln (Appliance Con.) Ln (Household energy exp) Ln (Heating equipment efficiency) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (Basement area) Ln (No. of dish washer load) Ln (No. of clothes washer load) Ln (No. of clothes dryer load) Ln (No of light blubs) HDD*Ln (Price) Ground Temp*LN (Price) Household size*Ln (Income) Freezer Age*Ln (Size) First Frostfree Fridge Age* Ln(Size) First non-Frostfree Fridge Age* Ln(Size) Second Frostfree Fridge Age* Ln(Size) Second non-Frostfree Fridge Age* Ln(Size) CDD*Ln (A/C Usage) Cooktop Indicator (0 or 1) Furance Fan Indicator (0 or 1) Ln (Range/oven age) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t 7.97E-01 3.65E-01 1.63E-01 -1.34E-01 4.01E-02 -8.99E-02 1.27E-02 8.49E-03 1.94E-02 5.28E-03 -4.04E-02 -3.69E-02 9.23E-03 -1.76E-04 6.00E-04 -2.28E-03 1.59E-04 1.59E-03 3.42E-05 9.81E-02 3.79E-03 1.55E-02 1.710 0.671 0.641 12.65 1.08 1.16 -2.00 0.71 -1.74 1.70 0.51 1.37 0.14 -2.18 -2.46 5.22 -0.24 0.53 -1.47 0.10 1.22 0.23 1.48 0.08 0.73 1.037 8.30E-01 4.12E-01 -1.29E-01 -6.86E-02 1.53E-02 2.13E-02 -4.73E-02 -4.28E-02 9.27E-03 14.54 7.96E-01 1.77 3.62E-01 1.62E-01 -2.19 -1.34E-01 4.16E-02 -1.63 -8.82E-02 2.34 1.23E-02 7.80E-03 1.80 1.93E-02 5.59E-03 -2.72 -3.26E-02 -3.03 -2.95E-02 5.78 9.28E-03 -1.78E-04 5.79E-04 -1.90 -2.27E-03 2.09E-04 1.47 1.60E-03 3.48E-05 9.71E-02 3.14E-03 1.53E-02 -2.41E-03 1.794 1.551 0.672 0.640 12.61 1.07 1.15 -2.00 0.73 -1.70 1.64 0.46 1.36 0.14 -1.18 -1.20 5.23 -0.24 0.51 -1.47 0.13 1.22 0.23 1.46 0.07 0.72 -0.38 0.910 8.30E-01 4.12E-01 -1.29E-01 -6.86E-02 1.53E-02 2.13E-02 -4.73E-02 -4.28E-02 9.27E-03 14.54 1.77 -2.19 -1.63 2.34 1.80 -2.72 -3.03 5.78
-2.53E-03 1.68E-03
-2.53E-03 1.68E-03
-1.90 1.47
2.108 0.665 0.650
2.108 0.665 0.650
1.794
P A G E 43
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 9: Appliance Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households
Model I (w/o lagged) As Entered 20% Variable Coeff Lagged Price HDD GT HHS Variable mean Lagged Price (mean) Lagged Price (mean+StDev) Lagged Price (mean-StDev) HDD GT HHS EXP CON Own Price Elasticity at Mean Short-run Price Elasticity Long-run Price Elasticity Own Price Elasticity at Mean+StDev Short-run Price Elasticity Long-run Price Elasticity Own Price Elasticity at Mean-StDev Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity
Model II (w/ lagged) As Entered 20% -5.747E-02 -1.794E-02 4.048E-03 3.720E-03 3.255 4.255 2.255 7.511 3.527 2.433 1282.418 10207.354 -0.308 -1.594 -0.365 -1.739 -0.250 -1.472 0.009 1.130 -5.756E-02 -2.050E-02 0.000E+00 3.539E-03 3.255 4.255 2.255 7.511 3.527 2.433 1282.418 10207.354 -0.341 -1.746 -0.399 -1.913 -0.284 -1.605 0.009 1.115
-1.080E-01 -8.303E-02 5.645E-03
-1.120E-01 -8.749E-02 5.746E-03
7.511 3.527 2.433 1282.418 10207.354 -1.104
7.511 3.527 2.433 1282.418 10207.354 -1.150
-1.104
-1.150
-1.104
-1.150
0.977
0.968
P A G E 44
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 10: Appliance Energy Consumption Model Elasticities Estimates for Electric Heated Households
Model I (w/o lagged) As Entered 20% Variable Coeff Lagged Price*P HDD*P GT*P HHS*IN Variable mean Lagged Price (mean) Lagged Price (mean+StDev) Lagged Price (mean-StDev) HDD GT HHS EXP CON Own Price Elasticity at Mean Short-run Price Elasticity Long-run Price Elasticity Own Price Elasticity at Mean+StDev Short-run Price Elasticity Long-run Price Elasticity Own Price Elasticity at Mean-StDev Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity
Model II (w/ lagged) As Entered 20% -2.413E-03 -3.261E-02 -2.950E-02 9.282E-03 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 6783.606 -0.430 -0.739 -0.433 -0.742 -0.250 -0.561 0.021 0.796 0.000E+00 -4.729E-02 -4.278E-02 9.274E-03 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 6783.606 -0.594 -1.208 -0.594 -1.208 -0.594 -1.208 0.830
-4.041E-02 -3.692E-02 9.228E-03
-4.729E-02 -4.278E-02 9.274E-03
5.614 5.253 2.253 1535.903 6783.606 -0.421
5.614 5.253 2.253 1535.903 6783.606 -0.490
-0.421
-0.490
-0.421
-0.490
0.797
0.830
P A G E 45
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 11: Data Statistics for Natural Gas Space Heating Energy Consumption Model in Natural Gas Heated Households
Variables SpHt+DHW E Con Income HHSize Old Price ($/GJ) Price ($/GJ) Price-s ($/GJ) Energy Exp. ($) HDD (C) GdTemp (C) SHEff (%) Temp (C) No.Storey Hted Area (m^2) BsmtSize (m^2) Total Door Total Sky Total Win App E Con.
Name Q IN HHS P-1 P Ps EXP HDD GT EFF TEMP STOREY AREA BASE DOOR SKY WIN APP
Count 226.0 190.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 226.0 8.0 224.0 226.0
Max 181.4 100000.0 7.0 5.5 5.6 20.8 2499.0 6562.0 11.1 77.5 23.3 2.5 278.7 223.0 7.0 4.0 38.0 98.5
Min 20.6 6000.0 1.0 3.4 3.9 13.1 699.3 3962.0 4.8 57.5 15.0 1.0 46.5 1.3 1.0 1.0 3.0 14.0
Average 92.8 45144.7 3.2 4.7 4.9 15.7 1190.3 5740.8 5.9 62.5 19.9 1.2 117.5 93.9 2.6 1.6 11.2 35.1
Median 87.4 45000.0 3.0 4.3 4.4 17.3 1160.3 5889.0 6.1 64.0 20.0 1.0 116.1 93.6 2.0 1.0 10.0 32.3
StDev 27.8 24379.3 1.2 0.7 0.6 2.3 288.8 476.8 1.2 6.3 1.5 0.4 40.1 31.3 0.9 1.1 5.3 14.4
25 Percentile 75 Percentile 73.6 108.0 27500.0 55000.0 2.0 4.0 4.3 5.4 4.4 5.6 13.1 17.3 984.0 1336.5 5889.0 5920.0 4.8 6.1 57.5 64.0 19.0 21.0 1.0 1.5 74.3 116.1 74.3 111.5 2.0 3.0 1.0 2.0 7.8 14.0 25.1 39.8
P A G E 46
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 12: Data Statistics for Electric Space Heating Energy Consumption Model in Electric Heated Households
Variables SpHt+DHW E Con Income HHSize Old Price ($/GJ) Price ($/GJ) Price-s ($/GJ) Energy Exp. ($) HDD (C) GdTemp (C) SHEff (%) Temp (C) No.Storey Hted Area (m^2) BsmtSize (m^2) Total Door Total Sky Total Win App E Con.
Name Q IN HHS P-1 P Ps EXP HDD GT EFF TEMP STOREY AREA BASE DOOR SKY WIN APP
Count 320.0 263.0 320.0 320.0 320.0 320.0 320.0 320.0 320.0 305.0 320.0 320.0 320.0 320.0 310.0 7.0 311.0 320.0
Max 148.1 100000.0 9.0 32.5 32.5 11.0 3246.8 7930.0 11.3 250.0 25.0 3.0 278.7 278.7 11.0 4.0 40.0 69.9
Min 8.7 6000.0 1.0 12.8 13.1 7.3 490.6 3007.0 2.7 100.0 15.0 1.0 46.5 13.4 1.0 1.0 2.0 5.8
Average 56.0 40667.3 3.0 15.1 15.5 10.1 1535.9 5077.2 7.3 101.1 19.1 1.3 116.4 94.9 2.7 2.1 10.7 24.4
Median 53.2 37500.0 3.0 15.6 16.1 10.0 1487.2 4771.0 7.7 100.0 19.7 1.0 116.1 98.0 2.0 2.0 10.0 23.1
StDev 21.0 24397.7 1.3 1.8 1.9 0.7 447.4 904.4 1.4 10.3 2.1 0.4 46.2 33.5 1.1 0.9 4.3 9.7
25 Percentile 75 Percentile 41.8 67.0 22500.0 55000.0 2.0 4.0 15.6 15.6 16.1 16.1 10.0 10.0 1244.0 1806.1 4709.0 4884.0 7.6 7.7 100.0 100.0 17.7 20.7 1.0 1.5 74.3 118.4 74.3 116.1 2.0 3.0 2.0 2.0 8.0 13.0 17.3 29.8
P A G E 47
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 13: Natural Gas Space Heating Energy Consumption Model Summary for Natural Gas Fueled Households
Variable Name LN_CON LN_S_P LN_EXP LN_EFF LN_TEMP LN_STORE LN_AREA LN_BASE LN_DOOR LN_SKY LN_WIN LN_APP DD_LN_P HHS_LN_I GT_LN_P OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (GJ) ($/GJ) ($) (-) (C) (-) (m^3) (m^3) (-) (-) (-) (GJ) ($/GJ) ($) ($/GJ) ($/GJ)
Variable Description Ln (Total space heating energy) Ln (Substitute price) Ln (Household energy exp) Ln (Heating equipment efficiency) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (Basement area) Ln ( No. of doors) Ln (No. of skylights) Ln (No. of windows) Ln (Total appliance energy consumption) HDD*Ln (Price) Household size*Ln (Income) Ground Temp*LN (Price) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t -8.1E-01 1.8E+00 -3.4E-02 -1.3E-01 -8.5E-02 8.6E-02 -3.1E-02 2.4E-02 3.3E-02 -9.4E-01 -3.3E-02 -1.7E-03 -5.1E-02 -1.404 0.806 0.791 -7.719 -8.2E-01 21.751 1.8E+00 -0.296 -0.907 -1.901 -8.3E-02 1.907 9.5E-02 -1.304 -3.2E-02 0.597 1.184 -18.710 -4.750 -1.535 -9.465 -1.695 3.5E-02 -9.4E-01 -3.5E-02 -1.6E-03 -5.1E-02 -1.884 0.804 0.793 -8.10 -7.6E-01 22.152 1.8E+00 -3.6E-02 -1.3E-01 -1.886 -8.3E-02 2.148 8.0E-02 -1.343 -3.0E-02 8.8E-03 1.302 5.1E-02 -18.831 -9.2E-01 -5.048 2.4E-02 -1.460 -2.2E-03 -9.733 1.5E-02 -7.2E-02 -3.757 -2.421 0.832 0.808 -7.596 -7.9E-01 22.806 1.8E+00 -0.328 -0.953 -1.938 -8.2E-02 1.834 9.0E-02 -1.329 -3.1E-02 0.233 1.915 4.7E-02 -19.192 -9.2E-01 1.550 1.1E-02 -2.076 -2.0E-03 0.880 -4.081 -5.7E-02 -2.910 -2.676 0.821 0.811 -8.187 23.207
-1.964 2.127 -1.391
1.864 -19.408 1.564 1.946 -10.954 -5.671
P A G E 48
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 14: Electric Space Heating Energy Consumption Model Summary for Electric Heated Households
Variable Name LN_CON LN_S_P LN_EXP LN_EFF LN_TEMP LN_STORE LN_AREA LN_BASE LN_DOOR LN_SKY LN_WIN LN_APP DD_LN_P HHS_LN_I GT_LN_P OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (GJ) ($/GJ) ($) (-) (C) (-) (m^3) (m^3) (-) (-) (-) (GJ) ($/GJ) ($) ($/GJ) ($/GJ)
Variable Description Ln (Total space heating energy) Ln (Substitute price) Ln (Household energy exp) Ln (Heating equipment efficiency) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (Basement area) Ln ( No. of doors) Ln (No. of skylights) Ln (No. of windows) Ln (Total appliance energy consumption) HDD*Ln (Price) Household size*Ln (Income) Ground Temp*LN (Price) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t -7.0E-01 1.7E+00 -2.3E-02 1.0E-01 6.2E-02 -4.4E-02 6.5E-04 1.9E-03 -1.3E-01 3.0E-02 -7.6E-01 -1.3E-01 -6.0E-04 -1.5E-01 -0.610 0.888 0.882 -3.617 34.819 -0.173 1.317 1.631 -1.363 0.022 0.072 -1.908 1.565 -21.301 -11.193 -0.627 -17.158 -0.756 -7.1E-01 1.7E+00 1.1E-01 6.0E-02 -4.4E-02 -5.9E-01 1.8E+00 -5.5E-03 9.5E-02 7.8E-02 -2.0E-02 2.3E-02 -2.8E-02 -9.5E-02 -1.0E-02 -7.9E-01 -1.4E-02 -2.9E-04 -3.6E-02 -3.8E-02 -3.713 0.938 0.934 -4.138 47.820 -0.056 1.602 2.779 -0.842 1.052 -1.378 -1.819 -0.703 -29.706 -1.232 -0.404 -3.578 -14.107 -5.792 -6.1E-01 1.8E+00 1.0E-01 6.2E-02 -4.650 52.299 1.766 2.680
35.414 1.482 1.869 -1.717
-1.3E-01 3.0E-02 -7.7E-01 -1.3E-01 -1.5E-01 -6.820 0.888 0.883
-1.980 1.616 -23.815 -11.351 -17.431 -1.283
-3.0E-02 -9.7E-02 -7.9E-01
-1.569 -1.857 -33.298
-2.5E-02 -4.0E-02 -3.989 0.937 0.935
-7.041 -22.055 -10.141
P A G E 49
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 15: Natural Gas Space Heating Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households
Model I (w/o lagged) As Entered 20% Variable Coefficients Lagged Price HDD GT HHS Variable mean Lagged Price (mean) Lagged Price (mean+StDev) Lagged Price (mean-StDev) HDD GT HHS EXP CON Own Price Elasticity at Mean Short-run Price Elasticity Long-run Price Elasticity Own Price Elasticity at Mean+StDev Short-run Price Elasticity Long-run Price Elasticity Own Price Elasticity at Mean-StDev Short-run Price Elasticity Long-run Price Elasticity Income Elasticity Expenditure Elasticity Cross Price Effect
Model II (w/ lagged) As Entered 20% -7.152E-02 2.424E-02 1.477E-02 -2.214E-03 6.753 7.753 5.753 12.351 4.952 2.430 1190.261 92.830 -0.110 -0.745 -0.182 -0.810 -0.039 -0.689 -0.005 1.794 -0.763 -5.709E-02 1.133E-02 0.000E+00 -2.033E-03 6.753 7.753 5.753 12.351 4.952 2.430 1190.261 92.830 -0.246 -0.904 -0.303 -0.978 -0.189 -0.841 -0.005 1.791 -0.793
-3.338E-02 -5.050E-02 -1.696E-03
-3.478E-02 -5.103E-02 -1.591E-03
12.351 4.952 2.430 1190.261 92.830 -0.662
12.351 4.952 2.430 1190.261 92.830 -0.682
-0.662
-0.682
-0.662 -0.004 1.785 -0.805
-0.682 -0.004 1.787 -0.824
P A G E 50
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 16: Electric Heating Energy Consumption Model Elasticities Estimates for Electric Heated Households
Model II (w/ lagged) As Entered 20% -3.780E-02 -1.442E-02 -3.644E-02 -2.884E-04 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 55.974 -0.593 -3.678 -0.631 -4.055 -0.555 -3.366 -0.001 1.769 -0.593 -3.966E-02 0.000E+00 -2.471E-02 0.000E+00 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 55.974 -0.467 -2.804 -0.506 -3.029 -0.427 -2.610 0.000 1.755 -0.606
-1.280E-01 -1.480E-01 -6.002E-04
-1.280E-01 -1.480E-01 0.000E+00
5.614 5.253 2.253 1535.903 55.974 -1.496
5.614 5.253 2.253 1535.903 55.974 -1.496
-1.496
-1.496
-1.496 -0.001 1.720 -0.695
-1.496 0.000 1.721 -0.713
P A G E 51
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 17: Data Statistics for Natural Gas DHW Heating Energy Consumption Model in Natural Gas Heated Households
Variables DHW Con. (GJ) Income ($) HHSize Old Price ($/GJ) Price ($/GJ) Price-s ($/GJ) Energy Exp. ($) App Con. (GJ) HDD (C) GdTemp (C) SHEff (%) Temp (C) No.Storey Hted Area (m^2) BsmtSize (m^2) Total Door Total Sky Total Win TankSize (l) TankInsulation (0/1) Total Aerator DW-Total Wash-Total
Name Q IN HHS P-1 P Ps EXP APP HDD GT EFF TEMP STOREY AREA BASE DOOR SKY WIN TANK INSUL AERATOR DW WA
Count 226 190 226 226 226 226 226 226 226 226 226 226 226 226 226 226 8 224 226 221 106 226 226
Max 117.1 100000.0 7.0 5.5 5.6 20.8 2499.0 98.5 6562.0 11.1 77.5 23.3 2.5 278.7 223.0 7.0 4.0 38.0 270.0 1.0 6.0 780.0 780.0
Min 0.4 6000.0 1.0 3.4 3.9 13.1 699.3 14.0 3962.0 4.8 57.5 15.0 1.0 46.5 1.3 1.0 1.0 3.0 140.0 0.0 1.0 0.0 0.0
Average 41.8 45144.7 3.2 4.7 4.9 15.7 1190.3 35.1 5740.8 5.9 62.5 19.9 1.2 117.5 93.9 2.6 1.6 11.2 174.9 0.1 1.6 177.5 319.9
Median 40.5 45000.0 3.0 4.3 4.4 17.3 1160.3 32.3 5889.0 6.1 64.0 20.0 1.0 116.1 93.6 2.0 1.0 10.0 180.0 0.0 1.0 156.0 286.0
StDev 14.6 24379.3 1.2 0.7 0.6 2.3 288.8 14.4 476.8 1.2 6.3 1.5 0.4 40.1 31.3 0.9 1.1 5.3 25.4 0.3 1.0 188.8 194.6
25 Percentile 75 Percentile 32.3 49.0 27500.0 55000.0 2.0 4.0 4.3 5.4 4.4 5.6 13.1 17.3 984.0 1336.5 25.1 39.8 5889.0 5920.0 4.8 6.1 57.5 64.0 19.0 21.0 1.0 1.5 74.3 116.1 74.3 111.5 2.0 3.0 1.0 2.0 7.8 14.0 180.0 180.0 0.0 0.0 1.0 2.0 0.0 312.0 156.0 455.0
P A G E 52
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 18: Data Statistics for Electric DHW Heating Energy Consumption Model in Electric Heated Households
Variables DHW Con. (GJ) Income ($) HHSize Old Price ($/GJ) Price ($/GJ) Price-s ($/GJ) Energy Exp. ($) App Con. (GJ) HDD (C) GdTemp (C) SHEff (%) Temp (C) No.Storey Hted Area (m^2) BsmtSize (m^2) Total Door Total Sky Total Win TankSize (l) TankInsulation (0/1) Total Aerator DW-Total Wash-Total
Name Q IN HHS P-1 P Ps EXP APP HDD GT EFF TEMP STOREY AREA BASE DOOR SKY WIN TANK INSUL AERATOR DW WA
Count 320 263 320 320 320 320 320 320 320 320 305 320 320 320 320 310 7 311 320 305 138 320 320
Max 55.6 100000.0 9.0 18.3 32.5 11.0 3246.8 69.9 7930.0 11.3 250.0 25.0 3.0 278.7 278.7 11.0 4.0 40.0 270.0 1.0 6.0 780.0 780.0
Min 4.6 6000.0 1.0 7.2 13.1 7.3 490.6 5.8 3007.0 2.7 100.0 15.0 1.0 46.5 13.4 1.0 1.0 2.0 140.0 0.0 1.0 0.0 0.0
Average 19.4 40667.3 3.0 8.5 15.5 10.1 1535.9 24.4 5077.2 7.3 101.1 19.1 1.3 116.4 94.9 2.7 2.1 10.7 195.2 0.2 1.7 102.1 340.4
Median 18.3 37500.0 3.0 8.8 16.1 10.0 1487.2 23.1 4771.0 7.7 100.0 19.7 1.0 116.1 98.0 2.0 2.0 10.0 180.0 0.0 1.0 0.0 312.0
StDev 7.7 24397.7 1.3 1.0 1.9 0.7 447.4 9.7 904.4 1.4 10.3 2.1 0.4 46.2 33.5 1.1 0.9 4.3 33.6 0.4 1.2 148.7 218.2
P A G E 53
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 19: Natural Gas DHW Heating Energy Consumption Model Summary for Natural Gas Fueled Households
Variable Name LN_CON LN_S_P LN_EXP LN_TEMP LN_STORE LN_AREA LN_TANK LN_AERAT LN_DW_L LN_WA_L INSUL GT_LN_P HHS_LN_I HDD_LN_P OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (GJ) ($/GJ) ($) (C) (-) (m^3) (liter) (-) (-) (-) (-) ($/GJ) ($) ($/GJ) ($/GJ)
Variable Description Ln (Total DHW heating energy) Ln (Substitute price) Ln (Household energy exp) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (DHW tank size) Ln (No. of low-flow aerators) Ln (No. of dish washer load) Ln (No. of clothes washer load) DHW tank insulation indicator (0 or 1) Ground Temp*LN (Price) Household size*Ln (Income) HDD*Ln (Price) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t -4.66E-01 1.19E+00 4.67E-01 1.17E-01 -4.06E-02 -6.12E-01 -3.03E-02 4.06E-03 8.20E-03 7.23E-02 -6.73E-02 -8.52E-04 -6.93E-02 0.254 0.388 0.343 -1.62 7.69 1.20 1.03 -0.33 -2.86 -0.37 0.32 0.23 0.60 -4.47 -0.28 -3.52 -4.66E-01 1.18E+00 4.83E-01 -1.69 -4.17E-01 9.34 1.22E+00 1.30 4.74E-01 1.15E-01 -4.26E-02 -2.88 -5.95E-01 -1.95E-02 4.59E-03 5.04E-03 6.59E-02 -4.75 1.38E-02 -1.34E-03 -3.74 2.19E-03 -8.92E-02 0.041 -1.200 0.399 0.351 -1.45 7.90 1.23 1.02 -0.35 -2.80 -0.24 0.37 0.14 0.55 0.29 -0.44 0.05 -1.80 -0.559 -3.66E-01 1.20E+00 4.82E-01 -1.62 9.62 1.33
-6.01E-01
-5.65E-01
-2.79
-6.78E-02 -7.02E-02
0.127 7.891E-02 0.382 0.362
-7.91E-02 -1.462 0.393 0.376
-5.96 -0.822
P A G E 54
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 20: Electric DHW Heating Energy Consumption Model Summary for Electric Heated Households
Variable Name LN_CON LN_S_P LN_EXP LN_TEMP LN_STORE LN_AREA LN_TANK LN_AERAT LN_DW_L LN_WA_L INSUL GT_LN_P HHS_LN_I HDD_LN_P OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (GJ) ($/GJ) ($) (C) (-) (m^3) (liter) (-) (-) (-) (-) ($/GJ) ($) ($/GJ) ($/GJ)
Variable Description Ln (Total DHW heating energy) Ln (Substitute price) Ln (Household energy exp) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (DHW tank size) Ln (No. of low-flow aerators) Ln (No. of dish washer load) Ln (No. of clothes washer load) DHW tank insulation indicator (0 or 1) Ground Temp*LN (Price) Household size*Ln (Income) HDD*Ln (Price) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t -5.01E-01 8.03E-01 1.70E-01 -8.62E-02 1.29E-02 1.05E-01 -1.76E-02 1.58E-02 1.85E-02 1.96E-02 -4.49E-02 9.12E-03 -3.83E-02 -2.023 0.657 0.639 -1.57 12.84 1.21 -1.49 0.27 1.00 -0.42 2.33 1.27 0.46 -3.11 5.20 -2.03 -1.928 -4.91E-01 8.02E-01 1.68E-01 -8.69E-02 1.68E-02 1.03E-01 -1.94E-02 1.54E-02 1.74E-02 1.85E-02 -3.40E-02 9.20E-03 -2.70E-02 -3.52E-03 -2.281 0.657 0.638 -1.53 12.81 1.20 -1.50 0.35 0.97 -0.46 2.25 1.19 0.43 -1.40 5.22 -0.97 -0.56 -1.987
8.60E-01 -1.06E-01
15.20 -1.93
8.60E-01 -1.06E-01
15.20 -1.93
1.51E-02
2.38
1.51E-02
2.38
-4.84E-02 9.70E-03 -5.24E-02 -2.123 0.649 0.640
-3.43 6.30 -2.99 -3.738
-4.84E-02 9.70E-03 -5.24E-02 -2.123 0.649 0.640
-3.43 6.30 -2.99 -3.378
P A G E 55
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 21: Natural Gas DHW Heating Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households
Model II (w/ lagged) As Entered 20% -8.919E-02 2.191E-03 1.379E-02 -1.342E-03 6.753 7.753 5.753 12.040 4.769 2.552 1190.261 41.770 -0.510 -2.358 -0.599 -2.883 -0.421 -1.995 -0.003 1.224 -0.417 -7.909E-02 0.000E+00 0.000E+00 0.000E+00 6.753 7.753 5.753 12.040 4.769 2.552 1190.261 41.770 -0.534 -2.510 -0.613 -3.023 -0.455 -2.145 0.000 1.197 -0.366
-6.926E-02 -6.730E-02 -8.517E-04
-7.024E-02 -6.783E-02 0.000E+00
12.040 4.769 2.552 1190.261 41.770 -1.155
12.040 4.769 2.552 1190.261 41.770 -1.169
-1.155
-1.169
-1.155 -0.002 1.189 -0.466
-1.169 0.000 1.184 -0.466
P A G E 56
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 22: Electric DHW Heating Energy Consumption Model Elasticities Estimates for Electric Heated Households
Model II (w/ lagged) As Entered 20% -3.515E-03 -2.700E-02 -3.403E-02 9.196E-03 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 19.423 -0.360 -0.705 -0.364 -0.709 -0.357 -0.701 0.021 0.802 -0.491 0.000E+00 -5.237E-02 -4.837E-02 9.696E-03 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 19.423 -0.548 -1.213 -0.548 -1.213 -0.548 -1.213 0.022 0.860 0.000
-3.831E-02 -4.492E-02 9.115E-03
-5.237E-02 -4.837E-02 9.696E-03
5.614 5.253 2.253 1535.903 19.423 -0.451
5.614 5.253 2.253 1535.903 19.423 -0.548
-0.451
-0.548
-0.451 0.021 0.803 -0.501
-0.548 0.022 0.860 0.000
P A G E 57
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 23: Data Statistics for Natural Gas Space and DHW Heating Energy Consumption Model in Natural Gas Heated Households
Variables SpHt+DHW E Con Income HHSize Old Price ($/GJ) Price ($/GJ) Price-s ($/GJ) Energy Exp. ($) HDD (C) GdTemp (C) SHEff (%) Temp (C) No.Storey Hted Area (m^2) BsmtSize (m^2) Total Door Total Sky Total Win App E Con. TankSize (l) TankInsulation (0/1) Total Aerator DW-Total Wash-Total
Name Q IN HHS P-1 P Ps EXP HDD GT EFF TEMP STOREY AREA BASE DOOR SKY WIN APP TANK INSUL AERATOR DW WA
Count 249.0 211.0 249.0 249.0 249.0 249.0 249.0 249.0 249.0 249.0 249.0 249.0 249.0 249.0 248.0 11.0 247.0 249.0 249.0 243.0 115.0 249.0 249.0
Max 255.2 100000.0 7.0 5.5 5.6 20.8 2572.6 6562.0 12.3 77.5 23.3 3.0 278.7 223.0 7.0 4.0 74.0 142.9 270.0 1.0 7.0 780.0 780.0
Min 62.0 6000.0 1.0 3.4 3.9 13.1 699.3 3007.0 4.8 57.5 15.0 1.0 46.5 1.3 1.0 1.0 3.0 14.0 140.0 0.0 1.0 0.0 0.0
Average 133.5 47191.9 3.1 4.7 4.9 15.4 1218.5 5590.7 6.2 62.8 19.8 1.3 121.7 95.1 2.7 1.7 11.5 37.8 176.0 0.1 1.7 176.8 309.5
Median 126.7 45000.0 3.0 4.3 4.4 14.9 1165.5 5889.0 6.1 64.0 20.0 1.0 116.1 96.6 2.0 1.0 10.0 33.4 180.0 0.0 1.0 156.0 260.0
StDev 37.3 25436.4 1.3 0.7 0.6 2.3 328.0 715.9 1.8 6.3 1.5 0.4 44.9 32.3 1.0 1.0 6.6 18.9 26.2 0.2 1.1 187.2 195.3
P A G E 58
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 24: Data Statistics for Electric Space and DHW Heating Energy Consumption Model in Electric Heated Households
Variables SpHt+DHW E Con Income HHSize Old Price ($/GJ) Price ($/GJ) Price-s ($/GJ) Energy Exp. ($) HDD (C) GdTemp (C) SHEff (%) Temp (C) No.Storey Hted Area (m^2) BsmtSize (m^2) Total Door Total Sky Total Win App E Con. TankSize (l) TankInsulation (0/1) Total Aerator DW-Total Wash-Total
Name Q IN HHS P-1 P Ps EXP HDD GT EFF TEMP STOREY AREA BASE DOOR SKY WIN APP TANK INSUL AERATOR DW WA
Count 320.0 263.0 320.0 320.0 320.0 320.0 320.0 320.0 320.0 305.0 320.0 320.0 320.0 320.0 310.0 7.0 311.0 320.0 320.0 305.0 138.0 320.0 320.0
Max 171.9 100000.0 9.0 32.5 32.5 11.0 3246.8 7930.0 11.3 250.0 25.0 3.0 278.7 278.7 11.0 4.0 40.0 69.9 270.0 1.0 6.0 780.0 780.0
Min 27.4 6000.0 1.0 12.8 13.1 7.3 490.6 3007.0 2.7 100.0 15.0 1.0 46.5 13.4 1.0 1.0 2.0 5.8 140.0 0.0 1.0 0.0 0.0
Average 75.4 40667.3 3.0 15.1 15.5 10.1 1535.9 5077.2 7.3 101.1 19.1 1.3 116.4 94.9 2.7 2.1 10.7 24.4 195.2 0.2 1.7 340.4 248.4
Median 73.9 37500.0 3.0 15.6 16.1 10.0 1487.2 4771.0 7.7 100.0 19.7 1.0 116.1 98.0 2.0 2.0 10.0 23.1 180.0 0.0 1.0 312.0 195.0
StDev 23.7 24397.7 1.3 1.8 1.9 0.7 447.4 904.4 1.4 10.3 2.1 0.4 46.2 33.5 1.1 0.9 4.3 9.7 33.6 0.4 1.2 218.2 183.5
P A G E 59
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 25: Natural Gas Space and DHW Heating Energy Consumption Model Summary for Natural Gas Fueled Households
Variable Name LN_CON LN_S_P LN_EXP LN_EFF LN_TEMP LN_STORE LN_AREA LN_BASE LN_DOOR LN_SKY LN_WIN LN_TANK LN_AERAT LN_DW_L LN_WA_L LN_APP INSUL DD_LN_P GT_LN_P HHS_LN_I OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (GJ) ($/GJ) ($) (-) (C) (-) (m^3) (m^3) (-) (-) (-) (liter) (-) (-) (-) (GJ) (-) ($/GJ) ($/GJ) ($) ($/GJ)
Variable Description Ln (Total space + DHW heating energy) Ln (Substitute price) Ln (Household energy exp) Ln (Heating equipment efficiency) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (Basement area) Ln ( No. of doors) Ln (No. of skylights) Ln (No. of windows) Ln (DHW tank size) Ln (No. of low-flow aerators) Ln (No. of dish washer load) Ln (No. of clothes washer load) Ln (Total appliance energy consumption) DHW tank insulation indicator (0 or 1) HDD*Ln (Price) Ground Temp*LN (Price) Household size*Ln (Income) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t -7.37E-01 1.72E+00 -4.10E-02 -1.10E-01 -3.77E-02 2.78E-02 -2.24E-05 5.31E-03 -1.06E-01 -2.79E-03 5.91E-02 -2.54E-03 1.91E-03 6.87E-03 -8.42E-01 -1.96E-02 -5.88E-02 -7.30E-02 8.00E-04 -1.223 0.937 0.931 -14.87 43.38 -0.72 -1.61 -1.75 1.32 0.00 0.27 -1.68 -0.21 1.57 -0.18 0.83 1.13 -35.41 -0.87 -13.46 -18.86 1.43 -2.788 -7.42E-01 1.73E+00 -1.03E-01 -3.87E-02 2.93E-02 -15.86 -6.93E-01 46.20 1.73E+00 -3.53E-02 -1.58 -1.17E-01 -1.97 -3.14E-02 1.52 2.70E-02 1.09E-03 -5.69E-03 -1.77 -1.11E-01 1.88E-02 1.48 6.19E-02 -1.24E-03 8.41E-04 5.59E-03 -36.34 -8.30E-01 -2.92E-02 -14.52 1.90E-02 -20.24 1.00E-02 2.20 2.17E-04 -6.10E-02 -4.018 -2.197 0.953 0.948 -16.05 50.39 -0.72 -1.97 -1.69 1.48 0.10 -0.33 -2.04 1.57 1.90 -0.10 0.42 1.07 -40.30 -1.49 1.83 0.93 0.44 -8.06 -5.521 -7.01E-01 1.73E+00 -1.04E-01 -3.23E-02 3.06E-02 -17.73 52.21 -1.84 -1.94 1.80
-1.07E-01 5.38E-02
-1.08E-01 1.61E-02 6.13E-02
-2.06 1.47 1.93
-8.41E-01 -6.02E-02 -7.44E-02 1.13E-03 -1.381 0.936 0.932
7.01E-03 -8.30E-01 -3.03E-02 1.04E-02
1.45 -41.66 -1.58 3.30
-5.48E-02 -2.281 0.953 0.949
-24.43 -7.637
P A G E 60
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 26: Electric Space and DHW Heating Energy Consumption Model Summary for Electric Heated Households
Variable Name LN_CON LN_S_P LN_EXP LN_EFF LN_TEMP LN_STORE LN_AREA LN_BASE LN_DOOR LN_SKY LN_WIN LN_TANK LN_AERAT LN_DW_L LN_WA_L LN_APP INSUL DD_LN_P GT_LN_P HHS_LN_I OP_LN_P CONSTANT R-SQUARE ADJ. R-SQUARE
Variable Unit (GJ) ($/GJ) ($) (-) (C) (-) (m^3) (m^3) (-) (-) (-) (liter) (-) (-) (-) (GJ) (-) ($/GJ) ($/GJ) ($) ($/GJ)
Variable Description Ln (Total space + DHW heating energy) Ln (Substitute price) Ln (Household energy exp) Ln (Heating equipment efficiency) Ln (Indoor temp.) Ln (No. of storey) Ln (Heated area) Ln (Basement area) Ln ( No. of doors) Ln (No. of skylights) Ln (No. of windows) Ln (DHW tank size) Ln (No. of low-flow aerators) Ln (No. of dish washer load) Ln (No. of clothes washer load) Ln (Total appliance energy consumption) DHW tank insulation indicator (0 or 1) HDD*Ln (Price) Ground Temp*LN (Price) Household size*Ln (Income) Lag. Price * Ln (Price)
Model Coefficients and t Values Without Lagged Fuel Price With Lagged Fuel Price As "ENTER" Upto 20% As "ENTER" Upto 20% Coeff t Coeff t Coeff t Coeff t -4.93E-01 1.27E+00 -1.15E-02 7.44E-02 2.95E-02 -2.42E-02 -1.47E-02 6.10E-03 -4.09E-02 1.58E-02 2.54E-02 1.65E-02 1.47E-02 4.38E-04 -2.95E-01 8.00E-03 -8.10E-02 -9.68E-02 -7.96E-04 -0.527 0.952 0.948 -4.77 46.87 -0.16 1.75 1.41 -1.40 -0.94 0.41 -1.02 1.25 0.79 1.32 2.57 0.10 -15.17 0.61 -12.24 -19.57 -1.43 -1.152 -4.85E-01 1.29E+00 7.59E-02 4.20E-02 -3.26E-02 -5.05 -2.66E-01 51.72 1.29E+00 -1.99E-02 1.81 4.56E-02 2.39 3.68E-02 -2.35 -3.55E-03 4.95E-03 -1.22E-02 -2.38E-02 -1.50E-02 1.99E-03 1.55 3.15E-03 3.41 -2.05E-03 -2.76E-03 -15.57 -2.95E-01 4.71E-03 -13.08 -9.07E-03 -20.40 -2.08E-02 -1.65 -7.88E-05 -3.05E-02 -1.821 -2.444 0.987 0.986 -4.93 92.42 -0.55 2.09 3.43 -0.40 0.61 -1.58 -1.14 -2.27 0.12 0.49 -0.68 -1.20 -29.45 0.70 -2.05 -5.29 -0.27 -25.36 -9.879 -2.57E-01 1.29E+00 4.64E-02 3.33E-02 -5.17 95.29 2.21 3.75
-1.40E-02 -1.50E-02
-1.95 -2.35
1.88E-02 1.65E-02 -2.96E+00 -8.33E-02 -9.84E-02 -8.82E-04 -0.516 0.951 0.949
-3.68E-03 -2.98E-01 -9.11E-03 -2.08E-02 -3.05E-02 -2.557 0.987 0.987
-1.90 -32.80 -2.14 -5.50 -27.02 -15.668
P A G E 61
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 27: Natural Gas Space and DHW Heating Energy Consumption Model Elasticities Estimates for Natural Gas Fueled Households
Model II (w/ lagged) As Entered 20% -6.096E-02 1.895E-02 1.002E-02 2.174E-04 7.003 8.003 6.003 7.809 3.389 2.475 1218.548 133.527 -0.245 -0.936 -0.306 -1.018 -0.184 -0.866 0.001 1.732 -0.693 -5.481E-02 1.037E-02 0.000E+00 0.000E+00 7.003 8.003 6.003 7.809 3.389 2.475 1218.548 133.527 -0.303 -1.036 -0.358 -1.124 -0.248 -0.960 0.000 1.733 -0.701
-5.882E-02 -7.297E-02 7.999E-04
-6.019E-02 -7.435E-02 1.134E-03
7.809 3.389 2.475 1218.548 133.527 -0.707
7.809 3.389 2.475 1218.548 133.527 -0.722
-0.707
-0.722
-0.707 0.002 1.724 -0.737
-0.722 0.003 1.732 -0.742
P A G E 62
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Table 28: Electric Space and DHW Heating Energy Consumption Model Elasticities Estimates for Electric Heated Households
Model II (w/ lagged) As Entered 20% -3.048E-02 -9.068E-03 -2.079E-02 -7.875E-05 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 75.397 -0.419 -1.657 -0.449 -1.749 -0.388 -1.575 0.000 1.290 -0.266 -3.046E-02 -9.111E-03 -2.077E-02 0.000E+00 8.490 9.490 7.490 5.614 5.253 2.253 1535.903 75.397 -0.419 -1.694 -0.449 -1.788 -0.388 -1.610 0.000 1.292 -0.257
-8.098E-02 -9.677E-02 -7.957E-04
-8.329E-02 -9.840E-02 -8.822E-04
5.614 5.253 2.253 1535.903 75.397 -0.963
5.614 5.253 2.253 1535.903 75.397 -0.984
-0.963
-0.984
-0.963 -0.002 1.269 -0.493
-0.984 -0.002 1.285 -0.485
P A G E 63
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
5.
CONCLUSION
A comprehensive econometric regression analysis was carried out on the three major components of the residential end-use energy consumption in natural gas and electric heated households, namely space heating energy consumption, domestic hot water (DHW) heating energy consumption and appliance energy consumption, to determine long- and short-term price and income elasticities. The analysis was carried out on the 1993 Canadian housing stock. Three main data sources were utilized in the analysis: (1) Survey of Household Energy Use (SHEU, 1993), (2) Energy billing records for a subset of SHEU data, and (3) Energy Statistics Handbook (Statistics Canada, 1998), Electric Power Statistics (Statistics Canada, 1992), and Canadian Economic Observer (Statistics Canada, 1998). SHEU provided most of the demographic and physical housing characteristics while its subset of billing records provided the total annual household fuel consumption data for both natural gas and electricity. Both current and lagged average provincial fuel prices were obtained and/or estimated using Statistics Canadas Energy Statistics Handbook, Electric Power Statistics, and Canadian Economic Observer.
Log-linear variable price and variable income elasticity energy demand model, which was first proposed by Betancourt (1981) and later used by Donnelly and Diesendorf (1985) and Douthitt (1989), was adopted and modified to determine the price and income elasticities of the major components of the residential end-use energy consumption in natural gas and electric heated households. Model equations that incorporate demographics, weather and equipment characteristics as explanatory variables were developed, and price and income elasticities were derived from the model equations. A total of sixteen separate model equations were developed and analyzed. The equations are for four end-use categories (appliance, space heating, DHW heating, combined space and DHW heating), two household types according to heating fuel (all electric and natural gas) and two terms (long- and short-term). It was not possible to develop models for oil heated households due to the small sample size of oil heated households with complete energy billing records in the database.
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Major economic variables, such as household energy expenditure and current, lagged, and substitute fuel prices were found to be statistically significant, and had the expected signs in most models. The effects of income variables on these major residential end-use energy consumption categories proved to be statistically not significant in most cases, and the magnitude of their derived elasticities on energy consumption was extremely small (i.e. very inelastic). Both the estimated short- and long-term price elasticities of space heating end-use were lower in natural gas heated households compared to those of electric heated households. The reversed was true for DHW heating energy consumption. The magnitude of substitute price (cross-price) effects was determined to be larger for natural gas heated households. Some, but not all, physical house and equipment characteristics were determined to be significant, and they usually had the expected sign of relationship with energy consumption. Furthermore, it was also found that consumers tended to have greater responses to both short- and long-term fuel price changes on energy demand if they were facing higher than average fuel prices. The overall levels of model fit were quite good for all model equations, with the adjusted R2 values ranging from 0.376 to 0.987. The combined electric space and DHW heating energy consumption model had the highest adjusted R2 value while the lowest value was found in natural gas DHW heating energy consumption model. It was found that consumers were, in general, very price inelastic (with elasticity less than unity) on energy consumption in the short-term, while somewhat price elastic (with elasticity at or above unity) in the long-term. These results are to be expected since residential energy consumption is a "normal service"; i.e. a necessity. Also, residential energy consumption is derived and influenced by existing household physical characteristics and equipment stock. Consequently, there are very few quick and easy opportunities for consumers to reduce their energy consumption without capital investment (consumers cannot change such variables as insulation level, furnace/boiler, efficiency, etc. to reduce their energy demands without major capital investment).
P A G E 65
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
6.
REFERENCES and BIBLIOGRAPHY
Baugham, M., and Joskow, P., The Effects of Fuel Prices on Residential Appliance Choice in the United States, Land Economics, Vol. 51, No. 1, 1975, pp. 41-49. Betancourt, R., R., An Econometric Analysis of Peak Electricity Demand in The Short Run, Energy Economics, Vol. 3, 1981, pp.14-29. Bohi, D. R., and Zimmerman, M B., An Update on Econometric Studies of Energy Demand Behavior, Ann. Rev. Energy, Vol. 9, 1984, pp.105-154. Canadian Economic Observer, Statistic Canada Cat. No: 62-010XPB, 1998. Capper, G., and Scott, A., The Economics of House Heating: Further Findings, Energy Economics, Vol. 4, 1982, pp. 134-138. Donelly, W .A. and Diesendorf, M., Variable Elasticity Models For electric Demand, Vol. 7, July 1985, pp.159-162. Douthitt, R. A., The Demand for Residential Space And Water Heating Fuel by Energy Conserving Households, The Journal of Consumer Affairs, Vol. 20, No. 2, 1986, pp. 231248. Douthitt, R. A., An Economic Analysis of The Demand For Residential Space Heating Fuel in Canada, Energy, Vol. 14, No. 4, 1989, pp. 187-197. Dubin, J. A. and McFadden. D. L., An Econometric Analysis of Residential Electric Appliance Holdings And Consumption Econometrica, Vol. 52, No. 2, 1984, pp. 345-362. Dubin, J. A., and Henson, S. E., The Returns to Insulation Upgrades: Results from a Mixed Engineering/Econometric Model California Institute of Technology Social Science Working Paper: No, 582, September 1985. Electric Power Statistics, Statistics Canada Cat. No: 57-202, 1992. Energy Statistics Handbook, Statistics Canada, Cat. No: 57-601, June 1998. Griffin, J., Methodological Advances in Energy Modeling, The Energy Journal, Vol. 14, No. 1, 1993, pp. 111-123. Griliches, Z. and Intriligator, M.D., Handbooks in Econometrics, Vol. 2., Elsevier Science Publishers B.V., The Netherlands, 1984. Griliches, Z., Distributed lags: a survey, Econometrica, 1967, pp. 16-49.
P A G E 66
CREEDAC
A P R I L 1999
ECONOMETRIC M ODEL
FOR
RESIDENTIAL ENERGY
END-USES
Haas, Reinhard, Biermayr, Peter, Zoechling, Josef, and Auer, Hans, Impact on electricity consumption of household appliances in Austria: a comparison of time series and crosssection analyses, Energy Policy, Vol. 26, No. 13, 1998, pp. 1031-1040. Heslop, L. A., Moran, L., and Cousineau, A., Consciousness in Energy Conservation Behavior: An Exploratory Study, Journal of Consumer Research, Vol. 8, Dec. 1981, pp. 299-305. Hunt, L., and Manning, N., Energy Price- And Income- Elasticities of Demand: Some Estimates for The UK Using The Cointegration Procedure, Scottish Journal of Political Economy, Vol. 36, No. 2, May 1989, pp. 183-193. Hutton, S., Domestic Fuel Expenditure: An Analysis of Three National Survey, Energy Economics, Vol. 6, Jan 1984, pp.52-58. Johnston, J. and DiNardo, J., Econometrics Methods, Fourth Edition, McGraw-Hill, New York. Kahn, E., Sathaye, J., and Robbins, D., An Engineering-economic Approach To Estimate the Price Elasticity of Residential Electricity Demand, Energy Economics, Vol. 8, April 1986, pp. 118-126. Koyck, L. M., Distributed Lags and Investment Analysis, Amsterdam, 1954. Madlener, R. and Alt, R., Residential Energy Demand Analysis: An Empirical Application of the Closure Test Principle, Empirical Economics, Vol. 21, No. 2, 1996, pp. 203-220. Natural Resources Canada (NRCan), Energy Efficiency Trends in Canada, Energy Publications, Ottawa, 1996. Poyer, D. A., and Williams, M., Residential Energy Demand: Additional Empirical Evidence by Minority Household Type, Energy Economics, Vol. 15, No. 3, April 1993, pp. 93-100. Ryan, D. L., Wang, Y., and Plourde, A., Asymmetric Price Responses of Residential Energy Demand in Ontario, Canadian Journal of Economics, Vol. 29, Special Issue Part 1, April 1996, pp. 317-323. SPSS, (1998), SPSS v9 .0 Statistics, SPSS Science Marketing Department, Chicago, Illinois. Survey of Household Energy Use (SHEU), Statistics Canada, February 1993. Taylor, L. D., The Demand for Electricity: A Survey, Bell Journal of Economics, Vol. 6, 1975, pp. 74-110.
P A G E 67
CREEDAC
A P R I L 1999

Econometric Models For Major Residential Energy End-Uses

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Econometric Models For Major Residential Energy End-Uses

Diunggah oleh

Hak Cipta:

Format Tersedia

CREEDAC

Canadian Residential Energy End-use Data and Analysis Centre

Econometric Models for Major Residential Energy End-uses

CREEDAC Project Team: Alan S. Fung Merih Aydinalp V. Ismet Ugursal

CONCLUSION...................................................................................................... 64 REFERENCES AND BIBLIOGRAPHY ............................................................... 66

Table 2: List of model equations

Table 3: List of model equations developed

GENERAL FORM OF THE MODEL EQUATIONS

DETERMINATION OF THE ELASTICITIES

MODEL EQUATIONS FOR DIFFERENT END-USES

( a1Pi , 1 +a 2HDD+a 3GT)

Y ( a 4HHS) E a 5 EFFa 6 TEMPa7 STOREYa8

where variables Xn for appliance energy consumption model are:

Y (a 4 HHS ) E a 5 EFF a 6 TEMP a 7 STOREY a 8 AREA a 9

n BASE a10 Psa11 APP a12 X a n

where variables Xn for space heating energy consumption model are:

Y (a 4 HHS ) E a 5 EFF a 6 TEMP a 7 STOREY a 8 AREA a 9

n BASE a10 Psa11 X a n

where variables Xn for DHW heating energy consumption model are:

a n1 n Xa AERATOR a n 2 DW a n 3 WA a n 4 e a n 5INSUL n = TANK

Y (a 4HHS ) E a 5 EFF a 6 TEMP a 7 STOREY a 8

n AREA a 9 BASE a10 Psa11 APP a12 X a n

PHYSICAL AND DEMOGRAPHIC CHARACTERISTICS OF HOUSEHOLDS

HOUSEHOLD APPLIANCE CHARACTERISTICS

HOUSEHOLD ENERGY CONSUMPTION DATA

Typical Monthly Household Natural Gas Consumption Profile

Typical Monthly Household Electricity Consumption Profile

Electricity Consumption (kWh)

AVERAGE PROVINCIAL FUEL PRICE

RESULTS AND DISCUSSION

HOUSEHOLD APPLIANCE ENERGY CONSUMPTION MODEL

where variables Xn for appliance energy consumption model are:

AC _ USAGE 7.33 E 5CDD LIGHT1.18E 1e 3.39 E 1FFAN

Natural gas heated household with lagged fuel price:

Y ( 3.54 E 3HHS ) E 1.12 E

n STOREY 1.64 E 1BASE 8.68 E 2 X a n

where variables Xn for appliance energy consumption model are:

NFF1 _ S 2.32 E 3 NFF1 _ A AC _ USAGE 5.33E 5CDD LIGHT 7.91E 2

where variables Xn for appliance energy consumption model are:

NFF2 _ S1.68 E 3 NFF 2 _ A

where variables Xn for appliance energy consumption model are:

NFF2 _ S1.68 E 3 NFF 2 _ A

SPACE HEATING ENERGY CONSUMPTION MODEL

where variables Xn for space heating energy consumption model are:

Natural gas heated household with lagged fuel price: Q i = 0.069Pi

Y ( 2.0 E 3HHS ) E 1.8E

n STOREY 8.2 E 2 AREA 9.0 E 2 BASE 3.1E 2 Ps7.9 E 1 APP 9.2E 1X a n

where variables Xn for space heating energy consumption model are:

where variables Xn for space heating energy consumption model are:

Electric heated household with lagged fuel price: Q i = 0.019Pi

n STOREY 6.2 E 2 Ps 6.1E 1 APP 7.9 E 1 X a n

where variables Xn for space heating energy consumption model are:

DHW HEATING ENERGY CONSUMPTION MODEL

Natural gas heated household without lagged fuel price:

Q i = 1.08Pi( 7.02 E 2 HHD 6.78 E 2 GT ) E 1.18 E TEMP 4.83E 1

where variables Xn for DHW heating energy consumption model are:

Natural gas heated household with lagged fuel price: Q i = 0.23Pi

where variables Xn for DHW heating energy consumption model are:

where variables Xn for DHW heating energy consumption model are:

where variables Xn for DHW heating energy consumption model are:

Natural gas heated household with lagged fuel price:

E1.73E TEMP 1.04 E 1

n STOREY 3.23 E 2 AREA 3.06 E 2 Ps7.01E 1APP 8.30 E 1 X a n