Case Study: RWE Uses SmartSignal to Achieve Value Based Maintenance

Operating a multinational power generation portfolio—including gas and biomass stations—means juggling complex demands. Yanquing Li, MIMechE, lead condition monitoring engineer for RWS , shared at a GE Vernova conference, “More renewables, less gas, but you’re having to be called on when there’s no wind, no solar... you’re having to come on so reliably. Of course, it’s very important. But at the same time, you’re two-shifting, sometimes double duty-shifting. So you’re starting and stopping the machines, which is a terrible thing for the gas plants and wasn’t what they’re designed for. But we’ve got to make do, right?”

The expectation from management is clear: “You’ve got to be totally reliable. It’s got to come on like that. Can’t fail.” Yet, the reality is far more challenging. Data needed for maintenance and risk assessment was scattered across different stations and systems. In this fragmented environment, if you need any information, you need to go to each individual station at time.

Critical information, from weather and commercial to work orders and asset performance was siloed in separate platforms, making comprehensive decision-making nearly impossible. This increases risk and inefficiencies.  Compounding this, there was no standardized failure mode list, and historical maintenance records were hard to track.  Li shared that when she started the Value-Based Maintenance (VBM) Journey, they didn’t have a failure mode list. This led to inconsistent work orders and made it difficult to extract and maintain reliable maintenance histories.

Faced with these issues, the team embarked on an ambitious VBM digitalization project, opting for an in-house solution to retain flexibility and fit their operational preferences.

Centralized Data and Custom Dashboards

The new platform integrated all relevant data—operational, maintenance, commercial—into interactive dashboards. These included the house dashboard (asset health scores by failure mode), dynamic risk dashboard (probability of failure and run-time analytics), work order planning dashboard, and the innovative event journal. A unified dashboard integrating the data and making it available in one place allows engineers, performance teams, and operators to see real-time asset health and prioritize interventions.

Building the Failure Mode Library
A key early step was compiling a comprehensive list of asset failure modes. The team drew from years of SmartSignal condition monitoring, historical records, and direct input from stations to identify and catalog failures. This enabled the creation of health models, where data from monitoring tools (like SmartSignal residuals and estimates) was translated into actionable health scores for each failure mode.

Dynamic Probability and Event Journal
Extracting and analyzing historical maintenance data was a challenge, given inconsistent record-keeping. To solve this, the team developed an event journal—a centralized log where all maintenance, inspection, root cause analysis, and corrective actions could be recorded and tracked for each asset and failure mode. This not only improved visibility and accountability, but also supported dynamic risk assessment, enabling engineers to understand the probability of future failures and prioritize actions accordingly.

Specialized Analytics and Advisory Management:
Performance models and analytic tools (such as SPC) were developed to monitor deviations and trends, supplementing SmartSignal’s core capabilities. Advisories from monitoring systems were aggregated and weighted to highlight areas needing attention and cross-linked with health scores and work order data for efficient follow-up.

Collaboration and Managed Services:
Given the scale and complexity, routine monitoring was supported by managed services, freeing up in-house experts to focus on actionable cases and system development. This partnership improved response times and ensured model maintenance and diagnostics were continually up to date. Routine monitoring is supported by managed services. “There was a time… where I’m the only person in the whole company, would you believe, in the UK to do SmartSignal stuff. So, if I was to do the routine monitoring as well, I might as well just not sleep the whole day. So luckily we have managed services to help me on that... I only report actionable cases to station because let’s not forget they don’t always get it right—but that’s okay. They get it right 90% of the time,”  Li shared.

Improved Reliability and Availability
Early detection and intervention have minimized unplanned outages. “Vibration issues on cooling water pumps were traced to loose plates and fixed before failure.” Performance models caught blocked condensers after abnormal temperature trends: “When we saw this... we saw it from higher winding temperatures. And when we went to trace back, we found that it was because the closed cooling water temperature was up. And why was the cold cooling water temperature up? Because our condensers were blocked. So we quickly unblocked that... Happy days.”

Lower Maintenance Costs and Optimized Scheduling
“Do it when I tell you to,” the speaker asserts—assets are serviced based on real condition and risk, not rigid intervals. Maintenance and outages can be timed to commercial needs, reducing unnecessary expenditures.

Enhanced Data Visibility and Team Collaboration
Dashboards and the event journal make asset health, maintenance activities, and failure tracking transparent and accessible. “Management will be asking everybody, what’s happening here? Why is it low? Has anything been done about it?... Now, if you put something in the event journal for that particular failure mode, they will know what was the last entry.”

Accelerated, Informed Decision-Making
Real-time analytics enable rapid, evidence-based decisions about resource allocation, outage timing, and prioritization. Even minor deviations—such as in winding temperatures—are identified and addressed before they escalate.

Continuous Improvement and Scalability
Every new failure mode, intervention, or missed detection feeds back into model refinement. “It’s a continuous improvement cycle. We start off with the criticality assessment, identify what strategy the stations want to deploy, then turn to the failure modes generating all of the failures that are connected to that station.”

Quantifiable Value
Major failures and lost generation have been avoided: “How much does it cost to replace a pump—a massive one like that? A million. How much in lost generation per day? 100 grand. Quickly your reason for buying SmartSignal justifies itself.”

The system is scalable and supports rapid rollout. “We start slow. But hopefully, as we now have all of the blueprints in place, we can roll out the next few plants out, you know, fairly quickly. That’s our aim—to be done by the looks of it, 2024.”

Conclusion
By unifying data, building robust failure mode models, and fostering a culture of collaboration and accountability, the value-based maintenance program has transformed asset management. What began as a fragmented, reactive process is now integrated, proactive, and data-driven—delivering higher reliability, lower costs, and a strong foundation for the future as the organization continues to grow and adapt.

The VBM initiative has delivered measurable gains across reliability, efficiency, and cost management.

This overview is based on a presentation by Yanquing Li, MIMechE, Lead Condition Monitoring Engineer in November 2022.