PLT - Using the Simplified Stochastic Event Flood Modeling Approach to Support the Risk Informed Decision Making Method to Evaluate Dam Safety

One of the objectives and supporting strategies in FERC’s Strategic Plan for fiscal years 2014-2018 is to minimize risk to the public by using Risk-Informed Decision-Making (RIDM) for evaluating dam safety in parallel to traditional dam safety methods. Resulting risk estimates can be used, along with standards-based analyses, to decide if dam safety investments are justified. Consumers Energy Company (CEC) identified a concern at their Alcona Dam in Michigan regarding potential erosion of the unlined, earthen auxiliary spillway, and the potential subsequent failure of the dam during flood events more frequent than the inflow design flood (the PMF). However, given the possible consequences downstream, the estimated dam fragility, and the proposed and completed risk reduction measures, the risk may be low enough such that modifications to the auxiliary spillway are not warranted. Therefore, in 2017 CEC began a RIDM study of the Alcona Dam auxiliary spillway for submission to FERC.

RIDM requires a set of hydrologic hazard curves (HHCs) to estimate the overall risk. A Simplified Stochastic Event Flood Modeling (SSEFM) approach was used to develop the HHCs for Alcona Dam. The SSEFM method is a compromise between a purely deterministic approach which tends to be conservative and a fully stochastic, Monte-Carlo approach. The resulting HHCs are estimates of peak inflow rates for a range of annual exceedance probabilities (AEPs) from 0.01 to less than 1x10 7 for both cool-season (rain on snow) and warm-season (rain only) events. The SSEFM approach is the cornerstone of this RIDM study, allowing all other aspects of the study to relate important loading characteristics (peak water level, hydrostatic pressure, auxiliary spillway flow duration, etc.) to AEPs and therefore, a proper estimate of the risk. We will present the steps of the approach, data sources, the results of the SSEFM, subsequent RIDM steps using these results, and the lessons learned throughout the process.