Machine Health | Case Study

Generator system health determination for maintenance optimisation

Our client, the UK’s largest electricity producer, sought out a new and modern method for optimising their generator maintenance using plant data. This would dramatically reduce maintenance costs, reduce outage duration and provide greater insight to system performance.


In most modern-day power stations, the maintenance for large, base-load power generators is centred around traditional, planned routines. These routines are developed by the original equipment manufacturer upon installation, leaving them prone to becoming outdated, without consideration for more agile technologies. They encourage operators to invest in rigid maintenance schedules regardless of system performance or risk condition. Occasionally, these planned routines can even induce costly faults or defects into the system.

Our client maintained their large power generators on a planned schedule resulting in high costs, extended downtime and maintenance induced faults. Although performance was monitored on a day to day basis, the overall system health was not being reviewed ahead of the shutdown. They had a huge amount of plant data at their disposal but weren’t using it as effectively as possible.

Our client sought a solution that would present a view of generator system health based on plant data. This would provide station engineers with a detailed view of historic performance and plant condition such that a condition based maintenance strategy cold be adopted.


Ada Mode developed an alternative way to interrogating system condition by utilising the high resolution historic data derived from plant instrumentation, walkdowns and chemical sampling. Data was extracted from multiple sources, over a decade of operation, encompassing a wide range of different operating conditions. A large base-load generator comprises over 15 sub-systems each with multiple key health parameters. The technology had to process and analyse a large volume of numeric, unstructured and time-series data from disparate silos into one user-friendly interface.

The analysis was presented on a customised reporting interface which enabled engineers to make well informed maintenance decisions based on an accurate and reliable view of system health. The interactive report made use of our time series exploration technology, which analyses trends and identifies patterns to determine any potential anomalies.

Ada Mode examined the client’s current maintenance planning process and integrated the output from the analysis into the a revised outage scope. This allowed plant managers to optimise maintenance work, significantly reduce costs and keep the duration of planned shutdowns to a minimum.


Ada Mode developed an approach that enabled a 30% reduction in generator maintenance. This greatly reduces annual maintenance costs (>£500,000 per shutdown), reduces maintenance induced defects and prioritises effort based on system health and risk. The methodology can be scaled across multiple sites and systems.

Other case studies