Better home energy audit modelling: incorporating inhabitant behaviours

Abstract

Building energy modelling is often used in US home energy audits to assess a home's energy performance and to determine energy-efficiency retrofit recommendations. These models promise quantitative, engineering-based, defensible information on a home's energy retrofit opportunities. Modelling is based on assumed standard use behaviours, despite highly variable energy use practices. This research reports on tests that incorporate household behaviour in home energy audit modelling, based on a sample of single-family households that received a utility-sponsored home energy audit in Seattle, Washington, a US city with a cool temperate climate. The use of a compact set of self-reported behaviours in place of standardized behavioural assumptions improved the match between actual home energy consumption and model estimates, and shifted retrofit savings predictions. These were modest improvements over the initially poor match, but highlight the opportunity for better customizing home energy audit modelling by using simple information on household behaviours. A comparison of modelled savings of heating-related conservation actions shows that energy savings from moderate behavioural changes are on par with retrofits for many homes. These steps provide a gateway to modelling household behavioural changes alongside retrofits, and a means to bring behaviour into conversations with homeowners and into the technically oriented efficiency paradigm in general.

Keywords::

• asset models
• energy
• energy models
• home energy audits
• homeowners
• housing
• inhabitant behaviour
• low carbon
• renovation
• retrofit
• United States

Acknowledgements

This research was based on a study completed by researchers from Portland State University, Research Into Action and Earth Advantage Institute, working together with Seattle City Light, Lawrence Berkeley National Laboratory and Home Performance Collaborative.

Funding

The data for this study were collected as part of a research effort supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Building Technologies Program, of the US Department of Energy [Contract No. DE-AC02-05CH11231].

Notes

1 The ‘home energy audit’ term is used here broadly to refer to various approaches used to translate the characteristics of a home into guidance for energy efficiency improvements. These have also been referred to as reviews or assessments. The focus here is on audits that involve a site visit by a home energy professional, but audits may also be self-driven (often internet-based) or remote.

2 The scope of this paper is limited to audits that use computational energy modelling and simulation to estimate energy end uses and to design retrofits.

3 The first three models: Home Energy Scoring Tool (http://homeenergyscore.gov/), Home Energy Saver Pro (http://hespro.lbl.gov/), and EPS Auditor, recently rebranded CakeSystems (http://cakesystems.com), were included in the analyses informing this paper. The Building America simulation protocols (Hendron & Engebrecht, 2010) provide the nearest thing to a standard set of assumptions in US home energy modelling, while the California Home Energy Rating System (HERS) is the official approach for existing homes in California (California Energy Commission, 2013).

4 Home Energy Saver Pro (http://hespro.lbl.gov/) is an auditor-focused tool; a consumer-focused tool (Home Energy Saver; http://homeenergysaver.lbl.gov/) is also available, as well as an asset-focused tool (Home Energy Scoring Tool; http://www.homeenergyscore.govwww.homeenergyscore.gov/) which was in beta testing at the time of these analyses.

5 Utility bills were normalized to account for the difference between actual and typical weather. To do this, model-estimated energy use using a typical year weather strip was compared with an estimate from a weather strip constructed from actual local weather station measurements for the utility billing period of interest. Each home's historical natural gas and electricity usage were each then multiplied by the proportion of typical weather-year estimated use to actual historical weather-year estimated use.

6 Cost-effectiveness comparisons considered only differences in estimated energy savings, and assumed fixed retrofit costs estimated in Home Energy Saver Pro. Actual retrofit costs may be much different from audit assumptions, also affecting the accuracy of cost-effectiveness estimates.

7 Behaviour- and technology-based measures interact, so the savings are not additive. Rather than decide whether behavioural measures or technical measures are queued first, the results are simply expressed as a percentage of total of behaviour- and technology-based savings computed individually.