Sponsor: GWRI
Start Date: 1992-04-01; Completion Date: 1992-04-01;
Keywords: Reservoir Operation, Electric Power, Water Supply, Competing Uses, River Forecasting
Description:
A major complicating factor in water resources systems management is handling unknown inputs. Stochastic optimization provides a sound mathematical framework but requires that enough data exist to develop statistical input representations. In cases where data records are insufficient (e.g., extreme events) or atypical of future input realizations, stochastic methods are inadequate. This article presents a control approach where input variables are only expected to belong in certain sets. The objective is to determine sets of admissible control actions guaranteeing that the system will remain within desirable bounds. This approach is applied to the operation of Lake Lanier, a U.S. Army Corps of Engineers multiobjective reservoir and is shown to be an effective management tool. This investigation led to the following conclusions: (1) Increases of dependable energy generation are possible, albeit at the expense of wider reservoir level fluctuations and more frequent water shortages. Thus, recreation and water supply reliability are in conflict with energy generation. (2) Good quality forecasts can minimize lake level fluctuations and, at the same time, meet higher dependable energy commitments and water supply requirements. (3) Forecasting also facilitates reservoir operation with maximum outflow and generation hour restrictions. Such constraints can be used to minimize off-peak energy generation, thus maximizing overall energy value.
Technical Report
Principal Investigator: Aris Georgakakos (GWRI/Georgia Tech)
Principal Investigator: Huaming Yao (GWRI/Georgia Tech)