Engineer News Network The ultimate online news and information resource for today’s engineer
Dr Benedikt Engel explains how physics-based, real-time monitoring of key components is key to understanding how long and how hard power stations can be driven to keep the lights on
Conflicts in Iran and Ukraine have reinforced the need for national energy security. And while the rapid increase in renewable energy has reduced the UK’s reliance on imported oil and gas, while also supporting Net Zero targets, it has also placed unprecedented pressure on critical national infrastructure.
The UK’s power infrastructure was never designed for the frequent cycling between gas, solar, wind, and nuclear energy that is now standard operating procedure. Plants designed to run at steady state are enduring significant asset damage as a result of effectively being turned on and off with escalating frequency. From stress to fatigue and creep, repeatedly starting and stopping to maintain grid stability is degrading assets at unforeseen rates across infrastructure that is, in many cases, already operating far beyond planned lifespans.
The UK has to extend the life of critical infrastructure, especially nuclear power. Operators have to find a way to manage the demands on energy and components created by inherently unpredictable, weather dependent energy sources and the associated frequent grid cycling.
Clean and secure energy
Local, clean power generation has become a priority in recent years as countries globally look to improve energy security and accelerate Net Zero targets. The UK is making progress towards the target for all electricity to come from 100% zero-carbon generation by 2035: during May 2026, 65% of electricity was generated by zero carbon sources. In addition to wind and solar generated power, the UK remains very reliant on legacy nuclear plants to insulate the nation from the impact of global conflict on oil and gas supplies and price rises.
However, mitigating one risk creates another: the new operating patterns demanded by frequent cycling to support grid flexibility and renewable integration create new stress regimes and accelerate asset damage. Not only was this infrastructure designed for a very different energy production and operating model but many of the power stations are running far beyond their original lifespan. Improving energy security is a critical aspect of energy production but the implications for infrastructure reliability, resilience and longevity are severe.
Unpredictable asset deterioration raises new safety concerns and can compromise operational performance. Operators need better visibility into how assets are aging. They need to quickly identify degradation such as fatigue, creep and stress. They need to recalibrate maintenance models based on accurate asset life consumption and remaining useful life.
Changing patterns of asset degradation
Throughout the power generating industry, traditional repair and replacement cycles are under the spotlight. The four-year manual inspection cycle remains set in stone: operators cannot afford to take power stations offline for eight weeks or more to undertake inspection, repair and replacement more frequently. Yet the simulations traditionally used to determine asset degradation no longer reflect this rapid cycling operating model. When plants ran unchanged for a long time, a subset of data used for simulation was representative of the entire production span. This is no longer the case when the unpredictability of power demand – from weather patterns onwards – creates enormous fluctuations.
While it is simply not feasible to run an engineering simulation across four years of data due to cost, the insight required to understand asset performance in the new operating model is available. Physics-Informed Neural Networks can achieve engineering-grade accuracy while delivering results fast enough for real-time decision support.
Critically, the industrial data required to support Physics AI is already available. Operators have an array of solutions to optimise operational performance. The temperature, pressure and operating history required to understand asset degradation is routinely collected. Converting this data into real-time engineering intelligence is transforming the way operators manage critical infrastructure.
Real-time insight
Combining engineering-grade physics with real-time operating information provides a continuous assessment of asset information, including stress, damage and life intelligence. It provides operators with immediate insight into trends in asset degradation, information that can be used to plan replacement and repair plans based on real, rather than surmised, asset performance.
In addition to reducing risk of failure, this insight will be key in managing costs, allowing operators to replace and repair based on real asset state, rather than a perceived risk of failure. Understanding the impact of frequent cycling on different asset types in real-time – potentially at not just one power station but across multiple sites – will enable operators to adopt an intelligence led approach to asset management that will avoid unnecessary repairs and replacements; reducing the total life cost of each asset whilst reinforcing safety protocols.
Continuous monitoring will enable operators to add asset performance and degradation into operational planning in real-time. It will provide insight to support replacement strategies, even inform future plant design. Critically, it will allow operators to determine how best to extend the life of the plant by combining real time asset performance information with the other operational factors to support grid cycling decisions.
Conclusion
Irrespective of global conflict and the implications for the cost and accessibility of oil and gas supplies, the ageing critical national infrastructure is under stress. With the push towards Net Zero in both the UK and Europe, thermal plants will be required to run ever more flexibly and the associated load cycling will lead to more asset damage. Existing plants, especially the nuclear sites, are already running long beyond their planned lifespan.
Operators need to manage the complex demands of resilience, efficiency and safety whilst also ensuring longevity for many years to come. To ensure safety, resilience and productivity, operators need to understand the new patterns of degradation and determine how best to manage the assets throughout their lifespan.
Combining physics AI with existing operational data is a powerful solution, providing real-time insight into the changing patterns of asset usage and, hence, degradation. It will allow operators to intelligently match repair and replacement strategies to actual rather than expected damage. Critically, continuous monitoring will provide the operational confidence required to keep the lights on for a nation looking to improve energy security, while also safely extending the life of national infrastructure.
Dr Benedikt Engel is CEO and Co-Founder, MatAlytics,
