Between 2016 and 2026, US wind-farm operators will spend a whopping $400 billion just on maintenance activities. How can power generators leverage technology to optimize production, prevent downtimes and save costs? An IoT-based predictive maintenance strategy might just be the answer.
USD 28.24 billion – that’s the projected market size of the global predictive maintenance sector, which would account for 37% CAGR from 2019 to 2025. Power generators and utilities are looking at IoT technologies, AI and machine learning (ML) to reduce their operational and maintenance cost substantially.
Catching the faults before it is too late
Utility companies are all about their assets.
Between 2016 and 2026, US wind-farm operators will spend over $400 billion on maintenance activities.
Incidents such as wind turbines catching fire, biomass plants blowing up, are rare but not altogether ruled off. Many such incidents could be avoidable if there were predictive maintenance measures in place.
By inferring valuable data from its connected assets, they can minimize O&M costs through better connectivity, detect overloading assets and foster trust among all the key players.
Why Predictive Maintenance for Wind and Solar?
Predictive maintenance systems are gaining popularity with solar and wind asset management. Many Wind Farm operators have been using SCADA data for remote condition- monitoring, and failure detection, but there are two major limitations in this preventive measure.
Limited amount of data: Sensor data collected from the wind turbines is limited and failures may originate from previously untracked and unknown sources as well. Wind turbine failures both onshore and offshore were often hard to place, making it difficult to identify the root cause of the failure remotely.
Little or no turnaround time: Usually, SCADA-based technology alerts users only after a certain threshold is reached and the break has occurred. It is often then too late for preventive measures. In short, there is no long-term visibility and less turnaround time.
It is often then too late for preventive measures.
Solar PV sites around Europe also seem to have adopted preventive maintenance techniques. Yet it could benefit from the predictive check to contain undetected degradations in PV cells and increase overall efficiency, not to mention redacted costs. There needs to be a system-level instalment of predictive maintenance tools, however, it would vastly optimize return-on-investment and reduce the time to claim warranty in PV installations.
The Big move from Preventive to Predictive Maintenance
Switching over to the other side means that companies have to adapt to big data, big time. When executed correctly, predictive maintenance can ensure that assets are serviced only when required. With fewer, and scheduled or ad hoc services performed with preventive maintenance, there would be a dramatic increase in savings, time and effort.
Digital Twin technology has been making ripples in the world of remote condition-monitoring and predictive analysis.
Data-driven Digital Twins
Since Digital Twins technology is still in its nascent stages, definitions are fluid.
We could term it as a visual representation or a simulation of both the elements and the dynamics of how an IoT device operates and lives throughout its life cycle, through AI.
With a digital twin, you can easily access data right from the temperature of a wind turbine’s yaw motors to the strength or duration of wind flow. Other data points that one can remotely access with predictive maintenance routine are vibration, acoustic of the power generation or storage components, and so on. Of course, it is not only for wind energy.
GE counts over 2 million digital twins in production for a range of IoT products. With the help of digital twins by IBM, the port of Rotterdam, Europe’s largest seaport, has been predicting mooring and departure times more efficiently. EDF Energy has saved over $1m by preventing equipment damage and lost production with the help of Schneider Electric’s EcoStruxure Maintenance Advisor solution. Big names including Siemens, IBM, Microsoft, Oracle and SAP have launched solutions to cater to the growing need in this market.
Predictive Maintenance: Challenges to Overcome
Implementing Predictive Maintenance (PdM) systems is a capital intensive investment. There is significant costs associated with implementing the infrastructure which includes a wide array of IT and IoT infrastructure, processes, trained personnel. Technical advancements require extra care and learning to address targeted problems.
The Quality of Data:
Data fuels the predictive maintenance engine. The quantity and more importantly the quality of it will determine the efficiency in analyzing the root causes of failures. One of the inherent challenges is the training of systems to make sense of the huge volumes of the data collected from the sensors.
With increased R&D, there have been significant strides made in sensor technology, improving the quantity and accuracy of the data collected. Organizations are fast enhancing their capabilities for securely managing and analyzing big data.
This is the future for Energy Asset Management
We, at Prospero Events Group, are bringing together experts in Predictive Maintenance from leading clean energy operators to share insights and best practices on how they implement PdM for their assets. Join us online for the best-in-class knowledge sharing and networking experience with Energy Industry experts.