The relative importance of risk factors influencing PRV failure is poorly understood. This project will take both engineering and statistical approaches to understand the probability of PRV failure to help optimise investment and maintenance frequency.
Many water utilities schedule the maintenance of PRVs based on asset criticality or consequence of failure; others are loosely based on RCM (reliability-centred maintenance) principles. As a result, it is likely that many PRV maintenance programmes are overly cautious meaning that too many inspections are carried out on PRVs each year, thereby inflating Totex on this asset class.
Best practice risk management can provide better understanding of the risk of asset failure from PRVs. Whilst the consequences of PRV failure are better understood, the influencing factors on probability of failure are less well quantified. An engineering and statistical study into the causal factors influencing PRV failure is therefore required.
Many water utilities currently base their PRV routine maintenance activities on a criticality assessment or time-based scheduling. If the probability of failure of different PRV cohorts is known - and explained through causal relationships with e.g. environmental factors - then water companies can optimise the maintenance frequencies to avoid over-maintaining healthy assets.
This could enable water companies to increase their PRV stock without the need to train more maintenance staff, and/or allow resources to be more efficiently utilised in the field.
The project will determine the key factors that influence PRV failures. These can then be used to help drive business decisions on where and how to utilise PRVs, as well as informing their recommended frequency of maintenance. It will also help determine the expected PRV investment needs for water infrastructure investment for PR19 and beyond.
This project will rely on sources of data on PRV failure and associated data such as size, make, model of PRV, pressure differential, ground conditions, source water quality and existing maintenance schedules.
Some of these data will be requested from project participants to maximise the value of the project from the pooling of industry data.
This data pool will then be used to investigate and identify common characteristics of PRV failure using predictive analytics methods.
Engineering knowledge of PRVs will be used to sense-check these statistical associations and develop physically meaningful reliability models for the different PRV cohorts. Where possible, probability of failure curves over the expected lifetime of the asset will be provided for different cohorts to provide evidence of deterioration rates for investment planning. Recommendations on maintenance frequency by asset cohort will also be provided.
WRc is an independent and trusted consultant to the water industry. By pooling data from project participants, WRc will apply our combination of engineering knowledge and predictive analytics expertise to provide you with additional insights into PRV failure. We will then work with you to determine how this could be used to help shape maintenance schedules and plan PRV asset investment.
PRVs play a vital role in optimising service from water networks. This project is an important step in maximising their utility and value.