Application of derived probability and dynamic programming techniques to planning regional stormwater management systems

There exists a need on the part of land developers or municipalities responsible for stormwater servicing master planning to reduce the costs associated with the initial construction of stormwater facilities as well as with their operation and maintenance. Common integral components of stormwater control systems include stormwater management ponds for water quality control. These ponds may occupy valuable urban land and, hence, it is desirable to minimize the land coverage of these facilities while simultaneously satisfying water quality control objectives. The employment of optimization techniques in the planning and design process can thus play an important role by reducing the costs associated with the implementation of such facilities. This paper presents an optimization methodology for single catchments using a single stormwater quality control pond. This methodology is then further developed for a multiple parallel catchment (each with a single pond upstream of its outlet) optimization procedure employing principles of dynamic programming. The principal constraint of the problem formulation is to meet a specified pollution control level at the outfall to a receiving water body. The optimization technique employs analytical probabilistic models for stormwater management planning and analysis which are in a mathematically closed form and thus easily integrated into an optimization framework. The costs explicitly considered are land-associated costs and construction costs. Operation and maintenance costs can be incorporated into the framework if desired.