Author(s): Hector Bravo

Linked Author(s): Hector Bravo

Keywords: Real-time; Stochastic; Control; River heating;

Abstract: Temperature exerts a major influence on biological activity and growth in rivers and lakes. Temperature is also important because of its influence on water chemistry. Two papers co-authored by Forrest Holly demonstrated that it is possible to find optimal solutions to the problem of producing power while controlling river water heating by developing a real-time optimal stochastic control system for powerplant river heating. In the first paper Holly and co-authors presented a state space model that quantifies operational effects and deals with variability and uncertainty of natural processes. The physical processes modeled by the state-space model included advection, diffusion, and environmental heat exchange; the mathematical formulation was one dimensional. The deterministic formulation used a hybrid characteristics-finite differences numerical scheme for the solution of the governing equations. The stochastic formulation accounted for uncertainty due to model assumptions, errors in the model inputs and parameters, and river temperature measurements. In the second paper Holly and co-authors presented an example in which optimal operation of a once-through-cooling powerplant is affected by the uncertainty in the hydrometeorological conditions that control the heat exchange between the river water surface and the atmosphere. The general criterion for the control scheme minimized losses resulting from inability of the power plant to produce at the appropriate potential level, and due to incurring a penalty for exceeding the standard temperature at a control station. The study demonstrated that real-time optimal control, which accounted for natural uncertainties, is not only possible, but also outperformed other simpler schemes.

DOI: https://doi.org/10.3850/38WC092019-1574

Year: 2019