Using Real Age As a Better Indicator of Predicting Asset Remaining Life

Abstract

In water and wastewater asset management, the Environmental Protection Agency has long emphasized the significant role of risk assessment to measure the criticality of an asset. Critical assets are defined as assets with a high probability of failure (old, poor condition, etc.) and high consequences if they do fail (major expense, system failure, safety concerns, etc.). In practice, the consequences of failure are typically estimated by experts through the Delphi method. However, the estimation of the probability of failure has been a challenging problem as it requires the analysis of the utility owned data such as condition assessment data, and repair and replacement records. The most common predictor of estimating the probability of failure for sewer pipes is by determining the remaining life. However, using the calendar age of a sewer pipe completely ignores different aging patterns of pipes under different operational and environmental conditions. This paper introduces a new method using the “real age” of a pipe in estimating the remaining life. Unlike calendar age, real age represents the adjusted age given the unique location and current operational conditions (grease condition, root problem, sludge problem, etc.). Depending on unique deterioration patterns of an individual pipe, the real age could be higher or lower than its calendar age. A variable selection process prior to a model development is conducted to ensure that the final model is lean but well performed. A case study using actual sewer pipe data collected from a sewer utility organization is discussed to show the differences of the estimated remaining life based on calendar age and real age.