Assessing consumer choice projections in energy system models: A heat pump case study

Abstract

Consumer adoption of heat pumps is a critical component of uncertainty in load growth forecasts, yet projections of heat pump adoption are rarely a focus of deeper investigation. This paper conducts a case study assessment of the structure, outputs, and performance of a model of U.S. residential heat pump adoption that has been used in multiple buildings sector potential studies and cross-sectoral energy system modeling exercises. The assessment includes model sensitivity testing based on a review of key factors in the heat pump adoption literature. Model projections across a range of scenarios show substantial regional variation in projected heat pump sales shares and sensitivity to input adjustments, with sales shares generally higher and less sensitive to input changes in warmer regions with high existing heat pump penetration, and generally lower and more sensitive in colder regions with a large installed base of fossil- and/or resistance-based heating equipment. Model benchmarking shows good agreement between estimated vs. reported heat pump sales and fuel switching shares in 2020 at the national level, but a tendency to underestimate heat pump sales shares vs. historical benchmarks at the state and regional levels. Key limitations are noted in the coverage and depth of existing historical equipment sales datasets, particularly at the state level. The analysis illustrates the current state of the art in projecting U.S. heat pump adoption and highlights key opportunities for improving confidence in heat pump projections within larger load growth forecasts and energy system modeling efforts.