c. Select a numerical procedure (coupling simulation and optimization) for simulation of
the impacts of the operational scenarios on temporal water quality. Although the
optimization process could be implemented by hand, using a numerical model for
simulation of operational scenarios, it is most often implemented as a part of the
numerical model, thereby allowing a computer routine to select the optimum scenario
based on the water quality objectives specified. A detailed description of an
optimization algorithm and its associated parameters is given in Poore and Loftis
(1983), and application is given in Dortch and Holland (1984). The numerical model
selected should accurately simulate the parameter(s) specified in the release water
quality objective(s). The model must be adjusted and verified to ensure that predictions
are accurate.
d. Conduct optimization simulations, using all viable operational scenarios to identify
the optimum operation. Since meteorological conditions will play an important role in
the reservoir water quality, extreme as well as average meteorological conditions should
also be simulated.
This general procedure can be used to develop operational procedures as well as used in a real-time
operational mode.
4.4.5.3 Implementation
The most extensive use of optimization techniques for water quality enhancement has been in
the area of selective withdrawal structure design. Numerous investigations using this technique have
proven effective for these purposes, particularly for an increase in reservoir storage for water supply
(Holland 1982, Schneider and Price 1988, Price and Holland 1989).
Wilhelms and Schneider (1986) describe an optimization technique using a one-dimensional
thermal model to operate a project to minimize temperature deviations over a long period of time. This
investigation assumed that small deviations over a long period of time were more acceptable than large
deviations over a short period of time.
Kaplan (1974) used a water quality index composed of several parameters linked to a reservoir
water quality model to optimize reservoir releases for both downstream and in-lake water quality. His
research concluded that the technique could be used to operate a reservoir to meet both flood control
and water quality objectives with a selective withdrawal structure.
Others have extended the optimization technique to assist in operation of multi-reservoir
systems. Fontane and Labadie (1982) developed a methodology for optimizing water storage
strategies in the West. This approach used an optimization simulation model to locate water supply
reservoirs while considering constraints such as water quality, flood control, and minimum flow needs.
4.4-17