Our Research on Optimal Reservoir Operation Under Uncertainty

As the demands of water for various uses increase, it becomes almost imperative that regulated river systems be optimally operated to meet the challenges posed by growing populations, a changing climate and shifting hydrologic conditions. Within the area of optimal control of reservoir systems, Arturo’s group developed a novel and computationally efficient model for the near real‐time and optimal control of multi‐objective and multi‐reservoir systems under uncertainty. This model was developed for enabling near real-time optimal decisions of water use in the Columbia River system that involves four States (OR, ID, WA, MT) in the Pacific Northwest. For more details see the link

Sketch of the big ten-reservoir system in the Columbia River

Schematic of our framework ketch of the ten-reservoir system in the Columbia River

Clustering of simulated variables for three reservoirs in the Columbia River System. This particular plot shows the clustering of the total volume ratio (TVR) and weekly volume ratio (WVR) based on forebay elevation groups for reservoirs Grand Coulee (left), Lower Granite (middle) and McNary (right)