structure for two or more of these purposes. For example, during the stratification season, the
operational purpose is to provide cool water temperatures to the downstream, but degradation of the
hypolimnetic oxygen content may result in low-dissolved oxygen (DO) water being released. In another
example, hypolimnetic releases during the early portion of the summer exhaust the supply of cold
hypolimnetic water, resulting in warm water releases that may be detrimental to the downstream fishery.
Guidance can be developed with numerical optimization techniques to avoid or minimize these conflicts
and deliver the desired water quality to the river and still meet other project purposes.
This technique relies on numerical water quality modeling to predict the release water quality
under a variety of operational scenarios. With a multilevel withdrawal structure and a release water
quality objective stated as a numerical function over time, a water quality model is used to predict in-
reservoir and release quality over time. From these predictions, a measure of achieving the release
objective for each operation (different release or withdrawal location) can be computed. Several
operational scenarios are simulated iteratively and systematically using optimization techniques to
compute how well each operational scenario meet the water quality objective. The operational
scenario that most closely matches the water quality objective is the optimum operation for the period
of concern. Figure 4.4.5 shows the release temperatures from a selective withdrawal tower based
upon a "best daily" operation and then a "seasonal" operation. Allowing a small violation of the
temperature objective in the early part of the year prevents a major violation in the late summer.
188.8.131.52 Design Criteria
Optimization techniques have been used in a variety of modeling applications, particularly in
water supply investigations and design and operation of selective withdrawal structures. Although the
techniques can become complicated because of the number of water quality objectives, the general
approach remains the same as for a one-parameter optimization process.
This approach for the optimization technique is outlined below.
a. Identify release water quality concern(s). This must be definable in scalar terms,
such that an objective function can be formulated for use in a numerical model. For
example, if the objective is to increase the release temperature during the fall period, the
temperature objective must be defined in specific increments over time.
b. Identify release operation concerns. This will include the number of ports, their
center-line elevations, heights, widths, and minimum and maximum discharges. If the
reservoir is operated under a guide curve, this must also be considered prior to