Improving energy self-sufficiency of a renovated residential neighborhood with heat pumps by analyzing smart meter data

Abstract In the energy renovation process, usually, buildings are upgraded to become energy-neutral annually with installed photovoltaic systems and heat pumps. However, the energy self-sufficiency of these buildings is surprisingly low. Therefore, the rapid deployment of heat pump based heating systems creates a shift of natural-gas consumption from the previously consumed building side (boilers) towards the electricity production side (power-plants). Fortunately, the development of information and communication technology enables access to consumption/generation data of building-related energy systems. Thus, there is an opportunity to strategically use this data and improve energy self-sufficiency and accommodate heat pump based heating systems. In this study, the improvement of self-sufficiency is discussed using a renovated neighborhood. The presented method incorporates a smart-grid application with a data-driven clustering, prediction, and an energy management strategy. First, clustering of similar demand-profiled dwellings with the k-means algorithm, and demand-prediction using the random-forest technique was performed. Afterwards, electric energy storage was introduced and multi-objective optimization reducing annualized costs and carbon emissions have been performed. For the carbon-dioxide optimal case, when aimed at the entire neighborhood, an annual self-sufficiency increment of more than 25% can be achieved, while four months out of the twelve being 100% energy self-sufficient.