Towards Standardization of Service Life Prediction of Roofing Membranes

A service life prediction approach for roofing membranes and systems is proposed based on a probabilistic modeling of the time-dependent performance. The membrane performance is modeled using a discrete Markov chain that evaluates the change of performance over time as a result of deterioration or repair. The transition probability matrix will be developed using in-field performance data collected during consecutive inspections for each roofing system, membrane type, age group, exposure conditions, and maintenance level. The use of in-field performance data captures the interaction and synergism of the various degradation factors, in addition to the impact of other roofing components, uncertainty and variability of material properties, environmental conditions, and quality of workmanship and maintenance. The service life is determined considering both the technical and economical performances, in addition to the risk of failure; this constitutes the first step towards the standardization of the service life prediction of roofing membranes and systems.