Industrial Reliability and Performance Improvement - Predictive Maintenance Application

There is a requirement for Sellafield ( to approach operations and maintenance with a manufacturing mindset where asset and process underperformance issues (which cause process bottlenecks) are identified prioritised based on value and solutions (both technology and operational) developed and implemented at pace All of the above steps require an ability to support each other with the core focus being to improve performance and reliability.

Across the site, there are several examples of equipment underperformance including reduced throughput, high levels of planned and unplanned maintenance, unstable operations, equipment availability.

Current methods of asset maintenance are either planned (based on elapsed time from installation) unplanned (reactive) These methods are as opposed to the approach of maintenance which is informed by actual asset use or its condition.

As a result, the root causes of asset underperformance are not being fully identified and addressed and can cause repeat events.

In response to this the Enterprise Asset Management (EAM) programme at SL has identified a need to develop predictive asset management problem solving capabilities that are underpinned by data and the application of data science methods to inform decisions.

This capability has been broadly termed Industrial Reliability and Performance Improvement (IRPI).

IRPI is at a relatively early stage of capability development currently defining its scope of work.

As maintenance is either planned or unplanned currently, there is no strategy to change to a predictive maintenance regime within the current constraints and processes.