Demand modeling and analysis for the management of underground infrastructure systems

Seunghyun Chung, Purdue University

Abstract

Factors such as rapidly growing population in metropolitan areas, stricter government regulations in wastewater infrastructure management, wastewater infrastructure assets nearing the end of its design life, along with limited budgets for maintenance and rehabilitation, are constantly challenging municipalities to keep sewer systems operational at their full potential. Sewer overflows have caused health and environmental hazards. Decrepit sewer pipes have led to sewer failures leading to costly repairs. Traditionally, municipalities have addressed the design, construction, maintenance, and rehabilitation of sewer systems on a “crisis-based” approach. Such management approaches have resulted in inefficient use of limited funds and often, costly repairs. In this study, a sanitary sewer management decision-making framework incorporating sewer demand forecasting and life cycle cost analysis is presented. The framework provides the asset managers with an alternative approach in sewer management. It is designed to allow asset managers to better allocate limited funds for maintenance and rehabilitation by identifying possible problematic sewers and devising a maintenance plan to prevent costly sewer failures. Sewer demand forecasting model is developed using an artificial neural network. The forecasted sewer demand is then used to identify “critical” areas (areas where the current hydraulic capacity is less than the forecasted sewer demand). In such areas, an optimal maintenance and rehabilitation strategy is developed through the application of probabilistic dynamic programming in conjunction with Markov chain modeling.

Degree

Ph.D.

Advisors

Abraham, Purdue University.

Subject Area

Civil engineering|Sanitation

COinS