Exploring policy evaluation approaches addressing micro-economic decisions to improve energy efficiency in the residential sector

Abstract in Undetermined The objective of the report is to analyse simulation results coming from the use of a policy evaluation method that encompasses a wider set of determinants for technology choice. To that end, we use the EEB_Sweden model v1.0. Still in an early stage of development, the model offers an opportunity to capture or integrate critical modelling aspects that justify our research project. In particular, we address several aspects to capture a more decentralised micro-economic decision framework for efficient technologies, including: (i) equal level of importance between financial/technology choice decision criteria and non-financial aspects, (ii) use and inclusion of non-financial criteria (e.g. reliability, predictability and appearance); and (iii) development and introduction of owner-tenant relationship in order to capture the ‘principal-agent’ problem. We scrutinise the method by evaluating different policy instruments: (i) zero-net energy buildings combined with existing policy instruments (e.g. grants), (ii) increased in energy price combined with existing policy instruments, (iii) an integrated policy mix encompassing several policy instruments; including existing ones. Two alternative baselines for comparative analysis are developed. One baseline depicts business-as-usual scenario in which market-driven or autonomous energy efficiency improvements take place. The second baseline attempts to capture the long term implication of existing policy instruments targeting the Swedish residential sector. Simulation results suggest that: • The single- and two-dwelling building stock uses energy less intensively compared to baselines. The order of magnitude of efficiency improvements is related to the applied policy instruments. For instance, in the presence of financial and non-financial decision criteria, current policies combined with a zero-net energy building regulation drive a much less energy intensity pattern after 2020 compared to the baselines. Due to the enlargement of the building stock as such, primary energy consumption as a whole increases. In all scenarios, efficiency improvements reduce the growth rate of primary energy consumption. • The impacts of current policies, even if energy prices grow dramatically, are modest. When energy prices increase (by a factor of 10), our results are more modest than estimated by other modelling groups. In turn, this suggests that the impacts of current policies are likely to be overestimated. Our results seem to be consistent with ex-post assessments. The main reason for this is likely to be the behavioural components included in the technology choice decision framework we use for the policy assessment. • Small plugs (e.g. appliances) and large plugs (e.g. dish washing machines) deliver marginal improvements. This may suggest that even if efficient technologies in these segments are attractive from the financial point of view in the long-run (e.g. A+ refrigerator), non-financial aspects drive decision makers to adopt technologies that are more attractive in other criteria (e.g. appearance, reliability) which are not necessarily energy efficient. This pattern is consistent with empirical behavioural studies. • Space-heating represents the major supply of efficiency improvements on a per building basis under analysed scenarios. This result is also consistent with several studies addressing the Swedish residential sector. • Consistent with empirical (and not modelling) studies, our results show that decision-makers are less motivated to adopt efficient technologies based solely on (high) energy costs. • Results suggest that when non-financial aspects of technology choice are included, decision makers tend to adopt technologies more “rationally” when confronted with a more integrated policy scenario, one that does not focus on a single-technology but on buildings as a system. In our case, we simulated a first-cost subsidy which is given once the energy intensity of the building is reduced below a certain threshold (e.g. 50~75 kWh/m2/yr). This policy scenario triggers a radical transformation of the building stock. We conclude and stress that the entire modelling exercise must be taken as a departure point to further develop a more comprehensive and realistic micro-economic decision framework. Within this learning and experimental research context, modelling results were used for policy insights rather than the forecasting of specific of numbers – like most modelling studies. With due limitations, our modelling results show that that a larger representation of determinants in energy modelling tools is necessary to better evaluate and enhance our understanding of policy choices and their impacts. Whereas many aspects deserve much more analysis, results suggest that the modelling approach included in the EEB_Sweden model should be further developed and used for policy assessment. A critical first research step involves the development of a database that represents the current technology configurations and corresponding energy and carbon values of the Swedish residential sector