Full text loading...
We developed a multi-level workflow for short-term optimization of daily controls for wind-powered reservoir management. The motivation for the workflow is to minimize emissions on a daily basis by taking advantage of the day-to-day variations of available wind power while still staying relatively close to a monthly target. The monthly target is based on optimizing the economic output (net present value) of the full model. The short-term optimization is performed with a coarsened version of the full-scale model. Prior to each short-term optimization run, the coarsened model is re-calibrated to the full model. One loop of the workflow is demonstrated on the Drogon test reservoir. In the experiment, the workflow was able to reduce the emissions by 14.75\% from the target strategy, with only a 2.9\% decrease in net present value.