(1977), Fontane, Labadie, and Loftis (1982), and Smith et al. (1987) have performed research into
density-influenced flow and selective withdrawal. These references examine some of the governing
equations as well as theoretical, laboratory, and field work to document this phenomenon.
The US Army Engineer Waterways Experiment Station (1986) compiled a reference document
on selective withdrawal which presents design considerations for selective withdrawal structures as well
as operational experiences and guidance.
Results and descriptions of numerical modeling of the selective withdrawal process are
presented by Howington (1989) for operation of a selective withdrawal structure at Lost Creek
Reservoir in Oregon. Selective withdrawal was also used in the thermal analysis of Prompton
Reservoir (Price and Holland 1989). In this investigation, the number and location of selective
withdrawal ports were identified for a proposed modification to raise the pool elevation. Optimization
techniques included with the model that was used resulted in a minimum number of ports required to
maintain downstream release temperature.
Maynord, Loftis, and Fontane (1978) conducted a study of the proposed Tallahala Creek
Lake in Mississippi to determine an optimum design for the intake structure to supply the necessary
water quality. Davis et al. (1987) documented the one-dimensional numerical model SELECT and
described its operation and limitations.
Selective withdrawal is a technique used to identify the withdrawal zone for a given structure.
Using this knowledge, modification of the structure or operation to achieve a given release objective
may be possible. Selective withdrawal is used as part of several of the other techniques mentioned in
this report. The theory is based on identification of the density-impacted withdrawal pattern.
Numerous research and specific reports have been published on selective withdrawal. Most
computational methods for selective withdrawal have been included in the computer program SELECT.
This program provides predicted release characteristics based on input profiles and operating
conditions. Table 4.3.1 presents a summary of this technique.