Asset degradation is driven by many factors with a common root around the degradation of materials. While material degradation will vary across networks as a function of changes in environment, operation and construction etc, the behaviour should be predictable allowing the predictive mapping of asset condition to geographic location. As an example, the risks to concrete bridge foundations will not necessarily be the same across the network and there may be wide differences in remnant life. Predictive mapping overcomes the bias of case-by-case engineering and allows the network wide impact of policy change to be assessed.

This approach can replace the resource limitation of conventional planned inspection programmes with work targeted to asset need. An adaptation of Risk Based Inspection strategies applied to the Oil & Gas sector, the use of Degradation Models accommodates the much longer operating life and slower corrosion rates for most natural environments through the modelling of uncertainty. Uncertainty, expressed as probability distributions, allows the models to cope both with variations in environment and the lack of specific asset condition knowledge.

Key to its success is the integration of existing data covering the physical and operating environments of the assets, their geographical location, operational, maintenance and inspection data, as well as the tacit knowledge of operator personnel.

Degradation Modelling is particularly relevant to highly regulated industries such as those found in the utility sectors, predicting future operational performance to inform asset integrity management and maintenance planning and to support the business case for future investment.

Recent applications range from the network wide application to above ground power transmission infrastructure to the creation of a probabilistic model to predict the remaining service life of 70-year old cast iron water tanks that are an integral part of the water treatment process for a plant serving 250,000 customers.