Automatic Control Strategies for Urban Stormwater

A number of United States cities have applied, or plan to study, automatic computer control technology to urban stormwater management. Several control strategies are compared, from simple set point or reactive control, to stochastic optimal control with full storm forecasting carried out adaptively in real-time. A simulated real-time experiment was conducted which generated storm inflow forecast errors that document the sensitivity of control strategy performance to forecast error. The qualified conclusions of this study are that strategies using limited lead-time forecasts are safer in that they guard against untreated overflows occurring. If the size of the storm indicates that overflows cannot be avoided, then more complete forecasting and sophisticated stochastic control is needed to manage the time and space distribution overflows oveflows so as to minimize adverse impacts on the receiving waters.