How to Design a Gas Program Impact Evaluation
Jon Maxwell with Kathryn Parlin for Zondits, January 11, 2014The number and size of natural gas energy efficiency programs are increasing throughout the United States. Evaluation of such programs is increasing commensurately. There is less research history on gas programs than electric. The planning assumptions and measurement and verification techniques traditionally applied to electric measures need reconsideration for natural gas.
Specifically, gas program evaluators are finding that project-level evaluated savings varies considerably from programs’ reported savings, more than it does for similar electric projects. The error ratio, a measure of this characteristic used in sampling, commonly exceeds 1.0 for natural gas efficiency programs, compared to the typical range of 0.4 to 0.6 used in planning evaluations. The higher error ratio means that gas evaluation is consistently reporting results with less statistical precision than expected. The presentation addresses why this is happening and, perhaps contrary to expectations, recommends that for a given budget evaluators invest in more research per site and evaluate fewer sites, rather than spend less per site and increase sample sizes.
Gas measures can affect customer costs beyond their natural gas utility bills, especially when the projects are associated with equipment that uses or heats water. The value of non-energy benefits and costs are material ( up to $9.50 per MMBtu of natural gas saved) and worth considering during evaluation. Gas measures can also affect electricity, diesel fuel, steam, and other energy uses. Evaluation results illustrate the importance of accounting for them in natural gas impact evaluation and when assessing program cost-effectiveness. The presentation also summarizes natural gas realization rates and net-to-gross factors.