825 Howard Hanson Dam Operations with Forecast Informed Reservoir Operations and Climate Change

Wednesday, 15 January 2025
Hall C (New Orleans Ernest N. Morial Convention Center)
Michael Warner, US Army Corps of Engineers - Seattle District, Tacoma, WA; and S. Marxen, K. Brettmann, K. Comanor, A. McCall, R. Aoki, C. Talbot, and F. M. Ralph

The Howard Hanson Dam and reservoir are situated on the Green River in the central Cascade Mountains in Western Washington and are operated by the US Army Corps of Engineers (USACE) in Seattle. The reservoir behind the dam has several functions: It is the main water supply for the city of Tacoma and is filled during the summer months, and it is empty during winter months when it is used for flood risk management during AR events, protecting billions of dollars of infrastructure downstream. It is often necessary to balance those functions with environmental concerns related to salmon. The reservoir is maintained by the Cascade Mountains’ abundant winter snowpack and precipitation and is currently refilled every year.

Since the reservoir behind Howard Hanson Dam must be empty before the flood season starts and is reliant on snowpack and precipitation to fill in late spring, impacts in precipitation and temperature due to climate change are important for how the USACE operates and manages flood risk and water supply in the future. Future concerns of insufficient flows for fish, timing of runoff for refill, potential for decreased winter snowpack, the possibility of adding storage space to the reservoir, and future changes in extreme precipitation, makes climate change an essential piece of the FIRO equation.

This study uses ResSim model designed to manipulate outflows at Howard Hanson Dam to keep the reservoir at the guide curve as closely as possible throughout the historical period of record. The model contains a combination of built-in and scripted rules to account for minimum flows, maximum flows, and refill operations. During flood events, this model follows these rules according to priority status in the and bring the reservoir back to the guide curve once the event has passed. This ResSim model was completed in January of 2024 and was updated to incorporate FIRO processes.

Reservoir operations under the current rule curve has the project storing no water in flood season, which ends in February. The first objective was to determine how much water could be stored at the end of flood season without increasing flood risk. Determining when stored water presents an increased risk of flooding accounts for time it takes to return to an empty pool before a significant Atmospheric River inflow event. The rule will not allow volumes of stored water larger than can be evacuated before potential flooding begins. A variant of this rule limited project outflows to 5 kcfs due to scour concerns. Both variants must account for inflows to the project during the x-day forecast period (where x = 1, 2, or 3, for instance) prior to potential flood events. Results were compared to current operations.

Redd scour downstream of the project is a concern during flood season. Outflows are limited to 5 kcfs or less anytime inflows are below 12 kcfs unless flood and/or dam safety issues require larger releases (Source: NMFS Bi-Op, 2019). FIRO would potentially add the flexibility to limit outflows under higher inflow conditions. The metric was the amount of scour avoided, both in terms of scour events and in hours of scour flows avoided. Scenarios using each x-day forecast were compared to a no-skill forecast, a.k.a., a “water on the ground” scenario.

The annual operation of the project depends on reaching a full pool in June and maintaining minimum in-stream flows at two downstream locations: USGS gages at Palmer and Auburn. Current operations have been in effect since 2007, and this analysis considered historical streamflow from 2008 to 2019. These scenarios will be especially important when considering the impacts of climate change on reservoir operations. As in the scour case above, scenarios using each x-day forecast were compared to a no-skill forecast, a.k.a. a “water on the ground” scenario.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner