Uncertainty Analysis in Reservoir Sedimentation

The capability o accurately predicting sedimentation rates in a reservoir is of critical importance when projecting future reservoir capacity, both during the design and operational phases of a reservoir. The useful life estimated during the design stage is based on projected sediment inflow and trap efficiency typically estimated from measured historical data. Such deterministic approach has a great disadvantage in terms of flexibility in coping with changes in the condition from that considered during the design stage. This paper presents an introduction of incorporating probability design into the process of estimating reservoir deposition. Uncertainty analysis of reservoir sedimentation problems is not new. However these studies are normally conducted by statisticians whose interests are mostly related to the accuracy of regression models or the uncertainty of the model itself and not so much to the actual problem i.e. reservoir deposition. A review of the common empirical methods of estimating reservoir sedimentation was conducted and the notion of uncertainty analysis of the parameters affecting deposition in reservoirs is presented. In order to appreciate the importance and the benefit of uncertainty analysis in reservoir engineering, a case study of Ringlet Reservoir has been carried out to develop uncertainty characterization using probabilistic methods.