Advanced Desktop Screening Techniques for Feeder Main Networks to Drive Condition Assessment Programs

Rationalizing and prioritizing condition assessment programs for large diameter water main networks are critical aspects of overall system asset management for virtually all water utility managers. Larger diameter mains are not feasible to be managed in response to failures, they require a dedicated condition assessment program to facilitate timely intervention before failure. As advanced condition assessment techniques can be very costly and inherently present their own risks associated with deployment, it is essential that the appropriate technology be selected in a staged manner with prioritization and schedule driven by true risk exposure. Winnipeg, MB, operates a water system that services a population of over 800,000. While the age of the system dates to the 1800s, the largest growth and network development occurred over the 1950s and 1960s. The majority of larger diameter water mains (>600 mm) were constructed of prestressed concrete cylinder pipe (PCCP) while larger mains up 600 mm are largely Asbestos Cement (AC). The native soils in Winnipeg are very corrosive to ferrous metals and some areas have extremely high soluble sulphate levels (>5,000 mm) which can be very damaging to cementitious materials. Being a cold climate environment, the City also uses de-icing salts extensively, which are known to elevate chloride levels in the soil and can be quite harmful to the corrosion protection properties of the exterior mortar of PCCP pipe. The age of much of the network and the exposure conditions highlights the need to better understand true physical condition. This case study presents the risk-based approach utilized to screen a feeder main network including some very innovative tools to assess risk comprehensively in a desktop model. An applied loads/deterioration tool was utilized that can analyze the combined effects of pressure, external loads, exposure conditions, and the unique design basis for each pipe segment in in the system in terms of its predicted factor of safety against failure. As the computational model can solve the entire network in near real time, it can be run in scenario analysis modes to provide increased clarity on which portion of the network should be assessed, when, in what order, and by what suite of CA technologies.