WSP12615 (2) - Laserfiche WebLink

Home Browse Search

002559 powerplant releases throughout the runoff period, a BHBF would be triggered to cause a "pre-emptive strike", Such an early BHBF would mitigate against the impacts of pending extremely high powerplant releases or actual spills which could occur later in the runoff period. Such a BHBF would not materially reduce the risk of future spills since the volumes ofBHBF's are relatively small, but would accomplish the goal of moving sediment from the main channel to the side channels and eddies. Subsequent high powerplant releases to control the spring runoff then would have a less detrimental effect on sediment resources. These BHBF' swould occur very early in the runoff forecast (mid-January through March), The value of 140 percent was chosen as the threshold value in an attempt to balance the number offalse alarms (BHBF's that after-the-fact were not actually required) with the number of missed spills (actual spring spills that were unforeseen earlier in the runoff season). This percentage value of 140 percent is near the level of the January forecast during 1986, an historic spill year. b - anytime an increasing forecast would require a powerplant monthly release greater than 1.5 MAF or use of the 0.5 MAF storage buffer, a BHBF could be released prior to increasing the releases above 25,000 cfs. Such conditions may occur with forecasts such as occurred in the years 1973, 1983, and 1995. These were years in which spring spills were undetectable earlier in the winter/spring and which would result in BHBF's or spills later in the runoff year. The timing of such releases may occur in the March or April time frame, but also could occur later in the runoff season, Previous presentations to the AMWO and the TWO have addressed the use of forecast risk curves as a means for triggering BHBF's. However, the subgroup felt this approach to be too complex. The substitution of the forecast percent ofnonnal and 1.5 MAF monthly release volume as thresholds are actually similar measures offorecast risk. The subgroup recognizes that these trigger mechanisms will initiate BHBF's in months other than the months of March and April which have been targeted previously for such flows. However, the long tenn frequency ofBHBF's is greatly dependent on an ability to recognize statistically extreme events such as much above nonnal snowpack conditions early in the year or the potential for large spring precipitation events that result in large increases in the forecast late in the runoff period, The proposed triggering mechanisms attempt to (I) comply with the Secretary of the Interior's comminnent made in the OCDEIS ROD, (2) minimize the number of actual spill years that were not recognized in time for a BHBF in March, and (3) minimize the number of years in which BHBF's were triggered but in which an actual spill did not materialize. 5 - For the following three paragraphs, the statistical results of the model runs apply to conditions only when the reservoir is at the target content of21.5 MAF on January I. As a long tenn average, this-mTI condirron is expected to occur about half the time in the future. Therefore, the lon~ tenn frequency of spills and BHBF's (I year in 6) is about half of the modeled frequency (I year in 3). Table I compares the proposed vs, current occurrences of spills and BHBF's. 7