A reliable predictive model for the deterioration of water mains and sewers has been the quest of water industry asset managers for several decades. No doubt driven by finance managers, asset managers have been searching for that mythical formula that will tell them exactly when an asset will fail and allow them to accurately forecast replacement and renewal expenditure into the future.
Although it is imperative that water authorities act to avoid and prevent the catastrophic failure of major infrastructure assets where the consequence of failure is high, for most authorities these assets form a relatively small part of the asset base. Therefore, the detailed predictive modelling and analysis of pipe material properties employed to plan the replacement of these critical assets is not appropriate or practical to apply to reticulation assets that represent over 95% of the total infrastructure for most Authorities.
The application of predictive models based upon pipe material performance to develop replacement programs for extensive reticulation systems, does not adequately take into consideration the variations in construction methods, ground conditions, consumption patterns, topography and climatic conditions that exist in many networks. These variations are, however, implicit in historical performance data which also provides information of the rate of failure of assets which is the main driver for the replacement of reticulation assets. It is only when the rate of failure of an asset becomes unacceptable that replacement is necessary and this adopted level of service will have a greater influence on future asset replacement expenditure than the assessment of the deterioration of pipe materials. All the more reason to base asset replacement programs around historical data and go with what you know.