Since 2014, RTI has supported the Tennessee Valley Authority (TVA), a leader in implementing risk-informed decision making (RIDM), in developing models, datasets, and decision support tools for TVA’s dam safety program.
As an initial step, RTI worked with TVA to bring their hydrologic, reservoir, and hydraulic models into current software platforms that can efficiently generate results and outputs to be used in subsequent analysis. These updates were mainly completed under the Probable Flood Hazards Analysis, EAP Mapping Project, and Consequences Analysis.
Probable Flood Hazards Analysis
To determine the probability of flood loading at each dam in the TVA system, RTI has worked with TVA to build a system that calculates hydrologic hazard relationships for dams using the Stochastic Event Flood Model (SEFM) system. RTI’s teaming partners, MGS Engineering, MetStat, and Applied Climate Services, developed precipitation-frequency relationships for critical watersheds in the TVA system based on a regional frequency analysis. The precipitation-frequency relationships are used in combination with historical storm templates to generate feasible precipitation scenarios with stochastic sampling. A continuous long-term simulation is employed to generate a database of seasonally varying initial conditions (starting soil moisture, river and reservoir levels) that can be used to define starting conditions for each unique event simulation. Tens of thousands of combinations of storm depth, storm pattern, and initial conditions are drawn and run through a series of hydrologic and reservoir operations models to simulate the response of the system to extreme events. For each simulation, key statistics such as peak reservoir inflow, outflow, pool elevation or duration above critical thresholds are calculated. From this information, frequency curves (i.e., hydrologic hazard curves) can be generated for each variable.
The framework also allows for quantification of uncertainties associated with the resulting hydrologic hazard curves, sampling of gate reliability, and provides toolset to analyze and understand the scenarios and underlying simulations. This system is currently being used by TVA to quantify hydrologic loading for many of their dams to support their RIDM process.