Effective demand response for smart grids: Evidence from a real-world pilot

We show how an electricity customer decision support system (DSS) can be used to design effective demand response programs. Designing an effective demand response (DR) program requires a deep understanding of energy consumer behavior and a precise estimation of the expected outcome. Excessive demand shifting or a high price responsiveness might create new peaks during low-demand periods. We combine insights from a real-world pilot with simulations and investigate how we can design effective DR schemes. We evaluate our pricing recommendations against existing economic approaches in the literature and show that targeted recommendations are more beneficial for customers and for the grid. Furthermore, we conduct robustness tests in which we apply our methods on two independent datasets and observe differences in peak demand and electricity cost reduction, dependent on individual characteristics. In addition, we examine the role of energy policy, as it varies across countries, and we find that the presence of competition in the electricity market creates lower prices and more cost savings for individuals. Finally, we measure the economic value of our DSS and show that our DSS can result in up to 38% savings on household electricity bills. Our results exhibit how the design of effective DR can be achieved and provide insights to energy policymakers with regard to understanding consumers' behavior and setting regulatory constraints. A data-driven approach to designing successful DR schemes is presented.Price sensitivity and awareness influence DR effectiveness.High price sensitivity yields electricity cost savings but not always peak reduction.Based on behavioral characteristics, personalized DR recommendations are outlined.Segmenting consumers and providing different prices is effective for the grid.