Analysing residential energy consumption using index decomposition analysis

Index decomposition analysis (IDA) has been a popular tool for studying changes in energy consumption over time in major energy consuming sectors. In the basic form, it allows such changes to be decomposed to give contributions associated with three different effects, namely activity, structure and intensity effects. In the literature, IDA studies on the residential sector, unlike those on industry and transport, show large variations in the choice of the activity indicator that drive energy consumption. Such variations greatly affect the decomposition results obtained and what these results capture. We investigate these issues and classify the conventional practices into two different decomposition models. We then propose a hybrid model which can better decompose changes of residential energy consumption and apply it to the data of Singapore. The relationships between the hybrid and the conventional models are analysed.