Electric end-use efficiency is attracting more and more attentions, but it remains unclear what factors are driving state policy innovations to improve energy efficiency. Controversy also exists over the effectiveness of energy efficiency programs. Several critical problems are facing the policymakers: what factors drive the states taking distinct strategies in policy innovation? Have state policies being able to improve energy efficiency in the past? And, will state policies remain relevant to future efficiency improvements? This dissertation tries to answer these important questions and assumes that policy innovation is relevant to energy efficiency. It first explores the factors that influence the adoption of energy efficiency policies using Internal Determinants models. Results suggest that internal state factors affect policy innovation, including state socioeconomic factors, state fiscal capacity, ideology, and constituent pressure. Policy innovations are found to be correlated with each other. This dissertation also evaluates the impact of policy innovation on energy efficiency by decomposing electricity productivity into activity, structure, and efficiency effects. The findings suggest that financial incentives and building codes have significant impacts on state electricity productivity. Other regulations tend to have mixed effects. In addition, an estimation of the achievable potential of energy efficiency suggests that policies will cost-effectively drive significant electricity savings in the future. Overall, this dissertation offers an in-depth diagnosis of the relationship between policy innovation and energy efficiency. It provides a rigorous statistical analysis covering the most important energy efficiency policies. It represents the first attempt to evaluate policy impact by decomposing electricity productivity. However, the statistical models and energy models are subject to limitations and future research is needed to improve the models.

Energy efficiency of buildings is an important measure to obtain a reduction of emissions of greenhouse gases since the building sector currently emits 40% of the total emissions in the world. A modification of control systems within commercial buildings is shown in earlier research to have a possible energy saving. The conventional control is usually controlled by fixed schedules and might presumably result in unnecessary energy use, since it operates at full capacity during the set period. Thus, modification of such control systems in commercial buildings might have a significant energy efficiency potential. The aim was to investigate the overlooked effect of human behavior on buildings’ energy usage. The hypothesis was a significant energy saving could be achieved if control systems use occupancy-based (OB) control instead of a conventional control based on fixed schedules. Firstly, occupancy data acquired from motion sensors in offices in the natural science building at Umeå University was used to create average occupancy patterns. The occupancy patterns were used to create various schedules with different objectives while representing the average floor occupancy. Secondly, the created schedules were used in the building simulation program IDA ICE. The simulation program was used to investigate how the energy consumption changed if the control system was changed to operate as OB control. Lastly, the potential energy savings were calculated with values from the simulations. The results showed that an average energy saving of 14% was achievable if OB control was used instead of fixed-schedule control. It was also found that a potential energy saving largely depends on the behavior of the occupants. Depending on the occupancy pattern and intensity, a potential energy saving of 10% to 17% was obtained. The results also signify the importance of well-functioning sensors and accurate detection. Additionally, it was found that the summer months had less potential energy savings compared to the other months. The results show that human behavior can have a large effect on the energy consumption within a building. Thus, implementing occupancy information in control systems might yield a potential energy saving. The intensity of occupants’ affects the potential energy saving. It was found that a higher intensity yields a lower potential while a lower intensity yields a larger potential energy saving. An interesting theory regarding the sensors was found after studying the results. It was found that it might be possible to let six arbitrarily offices represent 40 offices in an office environment. An important measure when collecting personal occupancy information is to inform the occupants of the experiment, as to not have problems with perceived privacy. Including the occupants in the data-collecting project might yield a better understanding and thus better collaboration with the experiment.