CS25C - Concurrent Session 25C: SQRA Methods Bringing Clarity to a Spillway Design Flow Decision

The Minnesota Department of Natural Resources (MnDNR) owns, operates, and regulates the Lake Bronson Dam in northwestern Minnesota. The gated primary spillway of this relatively remote, high-hazard dam is undersized and requires gate operation to safely pass large flood events. The MnDNR would like to address the undersized spillway, as well as known seepage issues at the dam. But, given the dam setting and hazard classification, how much bigger should the new primary spillway be?

We (Barr Engineering) teamed with Freese and Nichols to design a new labyrinth spillway structure, with Barr focusing on the hydrologic and hydraulic modeling work, and the determination of the design flow. To establish a risk-informed design flow for the primary spillway, we developed a 2D HEC-RAS model. We modeled a range of flood events, both with the dam breaching and without the dam breaching, and developed Python scripts to read the HEC-RAS HDF5 output, quantifying potential life loss impacts as a result of the floods and of the dam breaching. By integrating the consequences over the range of modeled probabilities, our team semi-quantitatively estimated risks for both the non-breach condition and the breach condition.

Comparing the non-breach condition risks with the breach condition risks, we were able to quantify the incremental risk associated with each primary spillway capacity that was tested. The results highlighted a clear design point for the primary spillway where the incremental risks were not further reduced substantially by building a larger spillway, and the overall risk magnitudes were in the acceptable range. In essence, there was no significant additional benefit to going bigger and spending more. The tools and methods that were used in this project managed the output data of a 2D HEC-RAS model covering over 110 square miles, modeling six flood events for each of six different potential spillway structures, in both breach and non-breach conditions, tracking over 1,200 residential structures, and synthesizing it all into one compelling chart that brought clarity to the question at hand.

The MnDNR was able to use the semi-quantitative risk information and make a well-informed decision on the design flow for the primary spillway. This method is transferrable to other dams in need of spillway rehabilitation and/or replacement, particularly where a risk assessment can support a design point less than the PMF.