In outward appearance, wind turbines still resemble technology from the future. To watch one up close, silently drawing invisible energy from the air, is to witness the stuff of science fiction. Of course, these gleaming carbon fiber monoliths have been around for decades—a fact that’s all too evident behind the scenes.
Every year, wind farm operators lose countless kilowatt-hours to unplanned downtime, operational inefficiency, and inaccurate forecasting. At the same time, schedule-based maintenance and an incomplete picture of asset health are shaving years off the lives of turbines valued at millions of dollars. With rising demand from consumers and a need to optimize the energy mix and costs, the industry is facing mounting pressure to resolve these challenges and drive greater operational efficiencies.
For most industrial companies, these are familiar problems. Whether your business is energy, chemicals, or aerospace, being competitive in the digital age depends increasingly on adopting a more sophisticated, data-driven approach to managing assets, measuring risk, and making informed business decisions.
Leaders in renewable energy are now applying learnings from these adjacent markets to advance their own digital industrial transformation. Renewables are prime for digital transformation, and will create new economic opportunities for wind operators and new levels of productivity.
When a wind turbine experiences an unexpected power decrease, operators must dispatch technicians to manually diagnose and resolve the problem. Anomalies often go undetected until they’re discovered in the course of scheduled maintenance. The absence of real-time information to perform preventative maintenance can lead to catastrophic failures that could easily be avoided. And a lack of data-driven forecasting leaves wind farms unable to adapt to changing weather patterns and grid demand.
Fortunately, these trends are beginning to shift as wind farm operators embrace the Industrial Internet of Things (IIoT) and digital solutions. Embedded sensor data, combined with robust analytics and artificial intelligence (AI) are making turbines intelligent and unlocking new levels of productivity. They operate more reliably, generate more energy, increased revenues and are cheaper to maintain.
Rather than sending technicians into hazardous or remote environments to identify a gearbox malfunction or drivetrain problem, connected sensors can detect issues automatically. AI-enabled software can then suggest a recommended course of action. Oftentimes, malfunctioning equipment can be reset remotely without the need for human intervention.
The digital wind farm solutions built on Predix Platform, a distributed, secure, edge-to-cloud platform that powers industrial applications, enable operators to more effectively grapple with the intermittency of wind energy. In order to capture maximum power, wind turbines need to react in real time to changing velocity and direction. At the same time, turbine operators need to balance fluctuating grid demand to avoid congestion or surplus. Fundamentally, this all boils down to a data analytics problem—one that Predix Platform is purpose built to address. Physics-based modeling combined with data science and AI, allows wind farm operators to mitigate unplanned downtime and increase asset life.
By building a digital twin that uses both historical data and real-time telemetry to predict asset failure, organizations can adopt a proactive maintenance model to extend life of an asset, and that keeps minor problems from turning into major ones.
The digital wind farm is being deployed globally with many customers, delivering a 10 % reduction in maintenance costs, up to 10 % AEP growth, and about a 3 % increase in revenue.
Digital transformation is seldom easy, but for companies hoping to remain competitive in an age of constant disruption, it is a necessity. To deliver on the promise of clean, renewable energy at scale, wind farms must become digital wind farms, and GE is uniquely suited to helping organizations through the toughest parts of the learning curve.