Conditional Demand Analysis: A cheap and effective approach to modeling residential energy usage

Incentivizing energy efficient upgrades is a common technique used by policymakers to offset energy costs and reduce consumption, but how do they decide what type of upgrade will maximize savings, given limited resources? Data on end-use energy consumption and behavior are crucial in order for policymakers to determine which upgrades to target, but acquiring this data from large enough samples can be costly and invasive.

In “A model of residential energy end-use in Canada: Using conditional demand analysis to suggest policy options for community energy planners,” Guy R. Newsham and Cara L. Donnelly from the National Research Council Canada use an approach called conditional demand analysis to estimate average annual energy use of electrical and natural gas appliances from 9,773 Canadian households. They find estimated energy savings for various appliance upgrades and behavioral changes, allowing local policymakers to determine their most cost-effective policy options.

Conditional demand analysis (CDA) is a statistical technique for estimating energy use that combines survey, consumption, and weather data. Energy appliance ownership and usage behavior is captured through a survey, and with the other information, a statistical regression is used to solve for the energy consumption of each appliance. The accuracy of self-reported data can vary, complicating the analysis, but when applied correctly CDA is widely accepted.

An alternative approach called direct sub-metering requires putting an energy meter on every appliance over an extended period of time. While effective, it is a much more expensive undertaking that requires recruitment and field personnel as well as incentives for households to offset the invasiveness. A third approach simply takes the engineering characteristics of the appliance and extrapolates consumption based on an assumed usage pattern, but this assumed usage may not actually be representative of the behavioral range of the population. Because it takes actual behavior data into account, and is not prohibitively expensive, CDA splits the difference between the two approaches and offers a cheap yet robust avenue for modeling end-use consumption.

In this study, Newsham and Donnelly acquired their initial data from the 2007 Households and Environment Survey, which was collected from 21,690 households and consisted of over 300 telephone survey questions on household, occupant, and environmental characteristics. A supplementary written questionnaire was then distributed with an additional 300 questions on more specific household, occupant, appliance, and heating/cooling characteristics. Combined with utility-provided gas and electricity data for each household, Newsham and Donnelly collected the requisite data needed to conduct CDA from 9,773 households across Canada, resulting in a 45 percent response rate.

Their analysis produced annual consumption estimates in kWh for each appliance. By comparing differences between newer and older appliances and between different behaviors, they were able to isolate the estimated energy savings of certain appliance replacements or behavior modifications that could be incentivized through policy actions. The following table summarizes these findings:

These kind of data make a policymaker’s task much more straightforward and may illuminate actions that otherwise would not be considered. For instance, it is common to rebate energy efficient light bulbs, but funds spent on those rebates might be better used toward an information and incentive campaign for programmable thermostats, since the energy savings are so much more pronounced.

For Canadian policymakers this study is extremely relevant, and the findings can be directly applied toward policy considerations. Given Canada’s unique climate, however, which factored into heating calculations and could potentially cause unique usage behavior, the data are less applicable to other policy markets. Despite the specificity of the findings, the broader application is a model of how to cheaply collect this valuable end-use information. As demonstrated, existing surveys could be supplemented and tweaked at very little additional cost. If implemented at the state or local level, community-specific consumption information could be used to minimize costs and maximize energy conservation.

Article Source: Guy R. Newsham and Cara L. Donnelly, “A model of residential energy end-use in Canada: Using conditional demand analysis to suggest policy options for community energy planners,” Energy Policy 59 (Aug 2013): 133-42.

Feature Photo: cc/(philippe leroyer)