A nuclear utility reduced operations and maintenance costs by replacing time-based maintenance with a predictive maintenance strategy using sensors and advanced analytic models.
Integral Analytics facilitated this transition by conducting a Failure Modes and Effects Analysis (FMEA), selecting optimal sensors, and deploying analytic models to detect equipment degradation in real time. As a result, the utility achieved an estimated savings of over $4.2 million, enhanced operational efficiency, and maintained strict safety and reliability standards.
The Situation
A nuclear utility was facing significant operations and maintenance (O&M) costs due to its rigorous planned maintenance schedule. With stringent safety and reliability standards, this utility conducted frequent visual inspections and preventive maintenance tasks, often performing time-based maintenance to replace wearable long before expected failure. While this approach ensured compliance with site maintenance procedures, it resulted in decreased capacity factor and excessive unnecessary parts and labor costs. The utility sought a more data-driven, predictive approach to asset management that would maintain safety standards while reducing O&M costs and unneeded maintenance activities.
“Ensure your stakeholders receive the right information at the right time to make the right decisions.”
-Aaron Hussey, PE, Principal & Founder
The Task
To transition from a traditional planned maintenance model to a predictive maintenance approach, the nuclear utility engaged Integral Analytics. The objective was to leverage existing models to monitor equipment conditions in real-time, avoiding unnecessary maintenance and improving operational efficiency. This was achieved by using Integral Analytics’ Knowledge Engineering approach:
-
Conduct a comprehensive FMEA to determine the most common degradation mechanisms for each identified failure mode.
-
Match sensors and analytic methods to provide the shortest time between the onset of incipient failure and detection.
-
Translate the FMEA into an automated, real-time diagnostic model that alerts the analyst or SME of the failure.
-
Establish alert thresholds and communication protocols based on the severity of the failure.
The Resolution
Integral Analytics initiated the process by performing an in-depth analysis of high-risk components in balance of plant (BOP) systems, while keeping specific component history and previous costs to the utility in mind. Then, an FMEA was performed on each of the identified components using EPRI’s Continuous Online Monitoring Guidance where available. Based on these insights, Integral Analytics recommended a network of advanced sensors and analytic methods capable of real-time data collection and evaluation of the identified components.
Once the sensor plan was in place, the team designed and deployed advanced analytic models to process the data and predict the identified degradation methods and failure modes. These models leveraged machine learning and statistical analysis to provide actionable insights, allowing the utility to transform their maintenance approach to a condition-based maintenance strategy. Integral Analytics also developed a comprehensive communication plan to ensure that key stakeholders—maintenance teams, engineers, and executives—receive the right information at the right time to make the right decision.
“It takes an army, but it’s not impossible. You need buy-in from the CEO to your engineering and operations staff from the very beginning or you won’t be successful.
-Michael Taylor, CMO and Senior Consultant
The Result
By transitioning to this predictive maintenance strategy, the utility significantly reduced planned maintenance costs while improving capacity factor and maintaining reliability and safety standards. Key outcomes included:
-
Net reduction in O&M costs for the identified components due to the elimination of unnecessary inspections and wearable replacement parts, resulting in an estimated $4.2 million over the life of the plant.
-
Streamlined resources allocation that allowed maintenance teams to focus only on those assets requiring attention, improving efficiency.
Lessons Learned
-
Data-driven decisions yield cost savings: Moving away from rigid planned maintenance schedules in favor of predictive analytics can yield significant financial benefits without compromising safety.
-
Failure Modes and Effects Analyses are essential: Conducting a FMEA at the outset ensures that the most critical and common failure points are identified and addressed effectively.
-
Effective communication is key: Successful implementation requires a robust communication plan to ensure that all stakeholders can act on predictive insights in a timely manner.
-
Technology integration is a continuous process: Ongoing monitoring and model refinements ensure that the system continues to improve over time, adapting to evolving operational needs.
By leveraging and adopting advanced analytics and sensor technology, this nuclear utility successfully modernized its maintenance strategy, setting a precedent for efficiency and innovation in the industry.
Integral Analytics’ Take
Nuclear utilities have what may be the most stringent maintenance philosophy, and for good reason. Safety is of utmost importance. But the nuclear industry can be equally or more safe by a data-based approach to maintenance. It takes an army, but it’s not impossible. You need buy-in from the CEO to your engineering and operations staff from the very beginning or you won’t be successful.
The staff at Integral Analytics combines over 100 years of operations, engineering, and data analytics in the nuclear, non-nuclear, and renewables industries. If you’re experiencing a problem, call us. We’ve been there.
