Statistical modelling of energy saving measures

Research output: Contribution to conferencePosterAcademic


Estimating guaranteed financial benefits due to households energy saving measures given their demographic and psychological characteristics

M. Namazkhan, C. Albers, L. Steg, R. Meijer
University of Groningen

To reduce households’ energy consumption, using cost-effective energy saving measures such as insulation, solar panels and heat recovery systems will be beneficial for consumers, the energy industry and the environment. Such a reduction in dependency to fossil energy consumption requires household behavioural change. Inducing the necessary changes in household behaviours, not only requires a precise understanding of the energy consumption patterns of households but also a deep insight on house characteristics, household demographics, psychological variables, such as attitudes and beliefs of inhabitants, as well as external factors influencing energy consumption such as the weather, energy prices, inflation, etc. The objective of this project is to enabling industry to offer consumers guaranteed savings and to assists consumers in making informed decisions on investing in energy saving measures given their demographic and psychological characteristics.

Mathematical-statistical model(s) incorporating a comprehensive list of variables will be developed to determine the probability distribution of the expected energy savings due to implementation of energy saving measures. The modelling framework will use several existing statistical, psychometric and econometric techniques such as regression, correlation, quality control and discriminant analysis as well as techniques from machine learning. The model(s) will be used to predict gas and/or electricity consumption of existing homes after a proposed energy saving measure, based on a wide range of relevant variables. The model will also be used in determining a minimum guaranteed saving. For households contemplating such an investment, this is important information.

In our project, we use data on household characteristics and daily energy consumption patterns from the Dutch and possibly other European households that will be gathered through a survey and by our industrial partners respectively. The model will be developed, calibrated and tested on data from about 15 thousand participating households. Expected results will provide information on the projected short and long-run financial and energy savings due to various energy saving measures, taking into account potential behaviour changes of the households. Implications for the society as a whole, from the effectiveness of various measures in reducing fossil energy consumption and lower energy prices will be analyzed and reported. Furthermore, we will investigate which psychological variables are key influencers in energy consumption.
Original languageEnglish
Publication statusPublished - 2017

Cite this