An empirically grounded model for simulating normative energy use feedback interventions. (1st July 2016)

• Record Type: Journal Article
• Authors: Anderson, Kyle; Lee, SangHyun
• Is Part Of: Applied energy. Volume 173(2016)
• Page Start: 272
• Page End: 282
• Publication Date: 2016-07-01
• Subjects: Agent-based modeling -- Normative feedback -- Residential energy use -- Occupant behavior -- Behavior diffusion; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2016.04.063
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: An empirically grounded behavior model for normative energy feedback interventions. Only sending normative feedback to individuals above the mean energy works. This alternative technique can be easily applied on a large scale with little cost. Abstract: Researchers have begun developing simulation models as a cost-effective and expeditious means to explore and enhance our understanding of energy use behavior interventions. These models have provided unique insights into potential energy savings as a result of improved occupant behavior, but have not yet reached the capability to be used for predictive modeling purposes. Therefore, this paper attempts to build on previous modeling efforts and develops an empirically and conceptually grounded occupant behavior model for simulating normative feedback interventions based on literature from the social science and field data. This model is then used to conduct "what if" analyses testing three novel normative feedback intervention strategies and the effect of social network structure on intervention outcomes. The most successful and immediately applicable strategy consists of sending normative feedback only to individuals who use more energy than the group norm. This strategy resulted in a mean energy use reduction of 1.4 kW h per week per occupant, 2.2%, relative to traditional individual and normative feedback strategies used today. Lastly, it was found that the social network structure in which the interventions took place affected the absolute outcomes (i.e., net change) of the simulations but not the relative outcomes (i.e., strategy ranking).