Development of a Methodology to Predict the Failure of Large-Diameter Cast Iron Water Mains

Abstract

Many of the water mains in North America are legacy cast iron pipes. These are susceptible to corrosion, which accelerates the failure of these pipes as they age. A reactive approach to the rehabilitation of small-diameter water mains can be justified due to the relatively low consequences of their failure. However, a proactive approach should be taken for largediameter pipes, which tend to have a lower rate of failure but higher consequences of failure. The aim of this paper is to present a new mechanistic model to aid in predicting the failure of large-diameter, grey cast iron water mains. In this new model, failure is assumed to occur due to a combination of corrosion pitting and hoop stresses from external and internal loads on the pipe. A fracture mechanics approach is used to account for the loss in strength of the pipe due to corrosion pitting. To account for uncertainty in the data collection and modeling processes, model inputs are treated as stochastic variables and the model is applied within a Monte Carlo Simulation (MCS) framework. A deterministic sensitivity analysis was undertaken to determine the sensitivity of the factor of safety to key variables. The methodology was applied to a 24” nominal diameter grey cast iron water main for an exposure time of 300 years. MCS was used to generate 10,000 realizations of the water main factor of safety over the 300-year period. An empirical cumulative distribution function (CDF) was developed and the interval in which 80% of the resulting factor of safety values occurred was determined for each exposure time. The preliminary results suggested that a 24” cast iron main under the loads considered is not expected fail.