How Hydropower Operations Mitigate Flow Forecast Uncertainties to Maintain Grid Services in the Western US

Hydropower facilities in the United States (US) most often have non-powered objectives, for example storage and release of water for water supply or environmental benefit, or flood control. These objectives can limit the flexibility available to hydropower operations to generate power to provide maximum benefit to the power grid. There does exist however flexibility within a week to optimize hydropower generation while still ensuring non-powered objectives are met. We examine the flexibility available to optimize generation and the value of medium-range inflow forecasts using a dynamic programing reservoir optimization model applied at ~250 hydropower facilities over the US Western Interconnection. Optimization is performed using day-ahead hourly scheduling to reflect existing electricity markets, using Locational Marginal Prices (LMPs) provided by a production cost model, and using three flavors of medium-range inflow forecasts – perfect forecasts representing an upper limit on performance, persistence forecasts representing a lower benchmark, and synthetic forecasts as a surrogate for operational streamflow forecast products. Measures of direct performance and flexibility are examined at the grid-scale for Balancing Authorities within the Western Interconnection. This study highlights where and under what conditions medium-range forecasts influence flexibility in hydropower operations which will be increasingly valuable under an evolving grid with increased renewable penetration.