Date of Award
Summer 1993
Document Type
Dissertation - Restricted
Degree Name
Doctor of Philosophy (PhD)
Department
Civil, Construction, and Environmental Engineering
First Advisor
Novotny, Vladimir
Second Advisor
Brown, Ronald
Third Advisor
Feng, Xin
Abstract
One of the most cost effective methods for limiting sewer overflows and flooding is to actively control the sewer system, operating it to fully utilize available storage, conveyance and treatment capacity. Such requirements are now included in sewerage utility discharge permits. Installing regulators, such as moveable sluice gates and weirs, allows sewer operators to control flows, but can cause problems such as increased risk of flooding. These regulators require efficient control logic to meet operational objectives and constraints. Algorithms are developed to solve several typical control problems in inline and diversion gate control. One algorithm controls an inline gate to maintain a desired flow rate or upstream water level. Another algorithm determines the diversion flow required to achieve a desired flow rate or water level in the main sewer. Another set of algorithms determines required flow rate at a control structure to achieve a desired flow rate or water level at some downstream location where substantial delay exists between the control action and its effect. All algorithms are based on water level measurements and are implemented with a minimum of mathematical and controller complexity. Fuzzy control methods are proposed where conflicting operational objectives and qualitative information exist. Portions of the Milwaukee Metropolitan Sewerage District (MMSD) collection systems were modeled to assess algorithm performance. The few hydraulic models theoretically capable of simulating a sewer system subject to control are unstable and execute slowly. An improved model, based on the zero-inertia approximation, was developed to test control algorithms. The zero-inertia model assumes no inertial components in the governing equations for flow, the St. Venant equations, but can model backwater effects and flow reversals. The recommended algorithms are stable and can maintain elevated flow rates and water levels with little risk of flooding. Fuzzy control successfully balances conflicting objectives and manipulates qualitative information. The potential for substantial benefits from improved control of the MMSD collection system is demonstrated. The zero-inertia method is demonstrated to be a stable, accurate, and rapid method for simulating sewer systems subject to control.