Composite Modeling of the Success Dam Spillway; Lessons Learned

Success Dam located on the Tule River in the Sierra Nevada foothills in California is undergoing an extensive remediation study for seismic retrofitting and updating the spillway to handle the updated Probable Maximum Flood (PMF) event. The anticipated PMF ranges from approximately 150,000 cfs to 160,000 cfs depending on the geometry of the overflow structure and the flood routing calculations. Additionally, the geometry of the outflow channel requires the flow to make almost a 90 degree turn. The sharp channel curvature creates a super-elevation on the outside bank of the floodway. In 2003, a 1:25 Froude-scale physical model was conducted at Utah State University/Utah Water Research Laboratory (UWRL) that investigated different spillway configurations to successfully pass the PMF without exceeding the maximum allowable reservoir pool elevation. Although the study concluded with a curved ogee crest that met the design criteria, recent changes in the seismic/geotechnical analyses required additional modeling. To enhance the modeling effort, a composite model approach of combining a physical model with a numerical model was undertaken. The intent for utilizing composite modeling was to use the results from the numerical model to identify potential alternatives before they were constructed. However, because composite modeling is relatively new, the study also contained a strong verification and validation component of the numerical modeling. The new physical model was constructed at a 1:25 Froude-scale and Flow-3D was selected as the Computational Fluid Dynamics software for the numerical modeling. Multiple configurations and flow rates were tested in both the physical model and the numerical model. Additionally, numerical simulations were completed on the 2003 physical model. Verification and validation data such as flow discharge, water surface profiles and point velocities were compared between the models. There are many benefits to combining numerical with physical modeling. In addition, this study determined that there are also some limitations when using a numerical model of this kind. The paper summarizes and quantifies the limitations for this particular geometry. Specific lessons learned from the composite modeling technique are also discussed. 24 pp, 11 references.