A Probability-Driven Multilayer Framework for Scheduling Intermittent Renewable Energy

A probability-driven, multilayer framework is proposed in this paper for ISOs to schedule intermittent wind power and other renewables. The fundamental idea is to view the intermittent renewable energy as a product with a lower quality (i.e., the probability of energy availability in real time) than dispatchable power plants, such as thermal or hydro plants, from the operators' viewpoint. Multiple layers which consider the probability of delivery are proposed such that various loads (critical or non-essential controllable loads) may participate in different layers in the energy market. A layer with a lower expected probability of energy availability is generally anticipated to have a lower price. This is similar to having different prices for commodities of varying qualities. A methodology is proposed to gradually merge the multilayers in the day-ahead market to a single deterministic layer in real time. The merge is necessary because the market must be deterministic in real time, whether sources are available or not. This is also aligned with the higher accuracy of forecasts when the time frame moves closer to real time. Further, the proposed scheduling framework is extended to consider the transmission constraints with a case study based on a modified PJM 5-bus system.