Key points are not available for this paper at this time.
Energy System Models (ESMs) that aim at describing and exploring pathways towards a decarbonized future energy system currently account insufficiently for the behavior of households and individuals. To address this shortcoming, this study evaluates models' existing approaches to incorporate behavior considering social science insights to advance the models' behavioral realism. A structured literature review and expert interviews were employed, selecting sixteen ESMs and two sectoral energy models for further investigation. Main data sources for the analysis were model descriptions and interview notes. The results show a predominant focus of models on financial aspects of adoption decisions and energy service use, while there is less consideration of non-economic behavioral drivers. Models also often rely on a weak empirical foundation for behavioral drivers. Based on these findings, advancing the representation of behavior in ESMs is needed to strengthen the realism of models’ explorative and descriptive insights. This analysis outlines concrete strategies to guide such an endeavor. It is recommended to consider relevant drivers of energy-related behavior, to employ a data-driven approach which relates behavioral outcomes to these drivers, and to define actor heterogeneity according to meaningful behavioral differences. In comparison to optimization approaches, the flexibility of simulation modelling provides a wider range of options for incorporating and analyzing behavioral aspects in ESMs. Future interdisciplinary research should further align social science insights with energy system modelling, building on the suggested strategies, to improve the accuracy of model predictions and to facilitate the consideration of behavioral aspects in the energy transition. • Behavioral drivers are needed to realistically explore energy transition pathways. • Energy System Models currently include behavioral drivers to a limited extent. • Simulation models show comparably more advanced examples for including behavior. • The use of empirical data is essential to improve models' behavioral realism. • Various feasible opportunities for advancing the modelling of behavior exist.
Galster et al. (Wed,) studied this question.