Detroit Dam Operational Scenarios


Detroit Dam Reference Conditions


To compare the operational and structural model scenarios, specific reference conditions (entitled “base”) were used to represent current operational guidelines and structures in place at Detroit Dam. These conditions then were applied with the model using a set of temperature targets to show the extent to which current operations and structures at Detroit Dam were able to meet those temperature targets under the three environmental scenarios.


base Operational Scenario


This operational scenario was intended to provide simulations that reflect the guidelines, timelines, rules, and understandings currently in place for the operation of Detroit Dam under the environmental scenarios developed for this study (table 2). The current operational rules for Detroit Dam were developed for the existing usable outlet structures (spillways, power penstocks, and upper ROs). As mentioned in the section “Detroit Dam Scenarios and Naming Convention”, these rules are based on a combination of current minimum outflows (mandated by the BiOP), maximum outflows, downstream irrigation withdrawals, minimum power production requirements, and a schedule of power-peaking operations (base scenario in table 4 and table 5). The majority of operational scenarios required a minimum of 40 percent of the total release rate to be routed through the power penstocks to allow a minimum amount of power generation. This is consistent with a current agreement between USACE and the Bonneville Power Administration, which distributes and markets hydropower from Detroit Dam and many other facilities across the Pacific Northwest. All operational scenarios discussed in this report assigned a higher priority to the power penstocks whenever possible. 


Before comparing modeled outflow temperatures, it is helpful to compare the modeled forebay elevations in each of the operational scenarios, as the timing of the rule curve can contribute greatly to the resulting temperature regime in the lake. The base operational scenarios generally led to modeled lake levels that closely matched the USACE rule curve during spring and early summer. As the summer progressed into the low-flow months, however, minimum flow requirements typically led to outflows exceeding inflows and a gradual decrease in lake level during mid-July through mid-October (fig. 7). 


Modeled temperatures from the base operations and existing structural scenarios serve as a basis to compare other structural and operational scenario outcomes. In this report, figures of the max temperature target scenario results show both the minimum and maximum temperature targets currently used by USACE for the North Santiam River (and adapted from targets used for the McKenzie River, Oregon), but only the maximum temperature target was used to drive the blending algorithm within CE-QUAL-W2 for these scenarios. In many of the following figures, “percent spill” is defined as the percentage of total flow that was directed to outlets other than the power penstocks; thus, spill includes the total releases from the ROs and the spillway. Outflow temperatures from base operational scenarios (fig. 8) did not meet the max temperature target during summer months (June–August), whereas omitting the minimum power generation constraint in the noppmin scenarios (fig. 9) generally allowed the model to be more successful at meeting the max temperature target. By allowing the “percent spill” to exceed 60 percent during the summer months (scenarios c2, n2,and h2), the model was allowed to release more of the warmer outflows from the spillway during summer, which retains some of the cooler, deeper water for release in autumn from the upper RO outlets (fig. 9). 


Without-Dams Temperature Target


To assess the potential for base operations and existing structures at Detroit Dam to match temperatures that might exist in the absence of Detroit Dam, calculated temperatures from the without-dams analysis were smoothed using a 7-day moving average of the daily maximum, then used as a temperature target in the wo_dams7dADM scenarios (table 7). The primary difference between the two sets of temperature targets used in this study (max, wo_dams7dADM) is evident in the allowable summer temperatures. Summer dam operations and release temperatures then help to determine the availability of cool water at the elevation of the available outlets later in autumn and the resulting autumn release temperatures (figs. 8 and 10). 


As was noted for the base and noppmin scenarios with max temperature targets (c1, n1, h1, c2, n2, and h2), the removal of minimum flow requirements to the power penstocks resulted in more warm surface water released in midsummer and cooler outflows from the upper ROs in autumn (compare figs. 8 and 9). The same is true when the wo_dams7dADM temperature targets were applied (compare figs. 10 and 11). The max temperature targets in June, early July, and late August are slightly higher than the wo_dams7dADM targets. As a result, midsummer release temperatures from the cwod2, nwod2, and hwod2 scenarios were slightly lower than temperatures from the c2, n2, and h2 scenarios during the same time frame, thus resulting in comparatively warmer outflow temperatures in autumn. 


Fixed Lake Level at Minimum Conservation Pool (1,450 feet)


One hypothetical operational scenario included a specification that the year-round Detroit Lake water-surface elevation be held at “minimum conservation pool” the current guideline used during the winter months (fig. 12). This fixed_elevation scenario is extreme in that it would greatly affect summer recreational activities on the lake, but might possibly have some advantages for passing fish downstream; this scenario might never be pursued, but resource managers felt it was important to assess the implications of such an action. This scenario also is referred to as “run-of-river” because outflows are approximately equivalent to total inflows throughout the year. Exceptions to this general rule are that outflows cannot fall below the minima or exceed the maxima shown in table 4. Outflow temperatures under fixed_elevation scenarios exceeded the max temperature target during the months of September–November (fig. 13), mainly because any cool water deep in the lake was below the elevation of the power penstocks and the upper ROs.


Delayed Drawdown


Reduced Minimum Outflows


When minimum outflows were decreased in summer under operational scenario delay_dd1 (tables 4, 5), the lake remained closer to full in mid- and late-summer until the rule curve dictated that the lake be drafted (lowered) to make room for potential flood storage. In this scenario, drawdown typically began in mid- to late-September under all three environmental scenarios (fig. 14), but the rule curve was not modified. Decreased outflows and a higher lake level in midsummer meant that the spillway crest was slightly deeper and accessing slightly cooler water, resulting in cooler releases at that time compared to base operations (figs. 8, and 15). In contrast, keeping a higher lake level in early to mid-September meant that the spillway could be used later in the summer compared to base operations (figs. 7, and 14), allowing more warm water to be expelled in early September and saving some cooler water for release later in autumn. This generally led to lower outflow temperatures in autumn when compared to base operations (figs. 8, and 15). 


When reduced minimum outflow operations were adjusted to free the model from the rule directing a minimum 40 percent outflow to the power penstocks (scenario delay_dd1_noppmin), the modeled release temperatures from Detroit Dam generally met the max temperature target in autumn (fig. 16). Given this scenario allowing the outlets to access both warm water near the lake surface and cool water at depth with few restrictions on minimum flows, downstream temperature targets generally can be met.


Reduced Minimum Outflows and Modified Rule Curve


Another way to keep more water in the lake and retain the use of the spillway in late summer is to modify the rule curve, keeping the target lake level higher late in the season and delaying the drawdown that is done to make room for potential flood storage. To examine the effects of delaying drawdown by about two months while minimizing reductions to summertime base operational scenario minimum releases, the “delay_dd2” operational scenario was developed (tables 4, 5). In this scenario, minimum releases were not decreased in midsummer as in the delay_dd1 operations; minimum releases were only modified after September 1. Drawdown under delay_dd2 typically began in late-October under all environmental scenarios except for hot/dry (fig. 17), in which minimum summertime outflow rules did not allow for the lake to remain above the spillway any later than it did under base operations. This scenario led to temperature management in autumn that generally was more successful than under base operations (compare figs. 8, and 18). During October and November, the cool/wet and normal scenarios under delay_dd2 operations resulted in simulated outflow temperatures that generally did not exceed the max temperature target (fig. 18). Results were not necessarily better than those using delay_dd1 operations (fig. 15).


Combining the delayed drawdown with no minimum outflow to the power penstocks (delay_dd2_noppmin in table 5) resulted in modeled outflow temperatures from Detroit Dam that generally met the max temperature target during autumn, aside from a period in November under the hot/dry environmental scenario that exceeded the max temperature target (fig. 19). As was noted for other scenarios with no minimum power generation, such a scenario allowed for more heat at the lake surface to be discharged during midsummer, thus meeting the max temperature target at that time and saving the deeper cold water for release in autumn when cooler releases were required by the temperature targets.


Although delaying the drawdown of the lake may be beneficial for downstream temperature management under certain conditions, such advantages must be balanced against the need to provide protection against potential flood damages. The delayed drawdown simulated in these scenarios specifies a drawdown of the lake that occurs primarily in November, which historically is one of the months with the greatest precipitation, leading to potentially large lake inflows. Clearly, if the drawdown of the lake is delayed, accurate and precise forecasting techniques must be used to balance the need for temperature management against impending needs to make room in the reservoir to capture high-inflow events. 


First posted October 30, 2012

Revised June 11, 2013