Optimized Risk-Aware Nomination Strategy in Demand Response Markets
Abstract
When wholesale prices are high and/or when the electricity system reliability is jeopardized, utilities have programs that encourage their customers to shed loads; these programs are called demand response (DR). Most large energy consumers, for example the buildings comprising Google’s Mountain View, CA campus, have some ability to control, and therefore reduce, their demand, and in this way contribute to stabilizing the grid. In this paper, we present a novel risk-aware nomination strategy in DR programs where the capacity commitments are made month ahead. The proposed methodology gives special treatment to the variability in load response to DR signals, variability that stems from the uncertainty in month-ahead weather forecasts and the inherent unpredictability of load profiles. The framework incorporates an efficient procedure for statistically characterizing the drop in demand associated with thermostatically controlled, heating, ventilation and air conditioning (HVAC) loads in commercial buildings, as well as applying the derived models to nominate revenue-optimal bids. The methodology is reasonably generic and adaptable to different kinds of DR markets and customer demand profiles. By way of example, we work out the capacity nomination for the Capacity Bidding Program (CBP) offered by Pacific Gas and Electric Company (PG&E) that Google participated in with a fraction of its buildings.
Research Areas
