Portfolio and contract design for demand response resources

Electricity generation from renewable energy sources has seen significant growth in recent years. To improve grid integration of these volatile energy resources, demand side flexibility can be operationalized through intelligent scheduling of flexible electrical loads. Prior research has mainly focused on determining dispatch schedules for exogenously given portfolios of flexible electricity demand. We expand on this literature by considering both contracting and dispatch of flexible loads. To,this end, we formulate a demand aggregator's problem of designing and dispatching a portfolio of supply (volatile renewable and conventional generators) and demand (inflexible base, shiftable and curtailable load) assets as a stochastic program. To account for different contracting regimes, we consider three market scenarios: (1) a liquid flexibility market where the energy retailer is a price-taker, (2) an opaque market where non-strategic customers will accept any individually rational contract offer and (3) a bi-level model with strategic customers optimally self-selecting into available contracts. In the last case, we also explore the effect of price discrimination versus uniform pricing. We evaluate these models using empirical load and generation data. Contrary to conventional wisdom, we find that flexibility contracting is not monotone-increasing in renewable generation capacities. Our research also suggests that curtailable load is a closer substitute to supply flexibility than shiftable load. Finally, we find that contracting of curtailing flexibility will be harder to implement if the aggregator has limited market power. (C) 2017 Elsevier B.V. All rights reserved.