Demand response evaluation and forecasting - Methods and results from the EcoGrid EU experiment

Understanding electricity consumers participating in new demand response schemes is important for investment decisions and the design and operation of electricity markets. Important metrics include peak response, time to peak response, energy delivered, ramping, and how the response changes with respect to external conditions. Such characteristics dictate the services DR is capable of offering, like primary frequency reserves, peak load shaving, and system balancing. In this paper, we develop methods to characterise price-responsive demand from the EcoGrid EU demonstration in a way that was bid into a real-time market. EcoGrid EU is a smart grid experiment with 1900 residential customers who are equipped with smart meters and automated devices reacting to five-minute electricity pricing. Customers are grouped and analysed according to the manufacturer that controlled devices. A number of advanced statistical models are used to show significant flexibility in the load, peaking at 27% for the best performing groups.