Data, Analytics, and AI: Driving change in the wind industry

15 Sep 2018

Wind energy generation continues to gain ground on traditional fossil fuel production, but there is still much work to be done. Every turbine asset needs to perform at maximum capacity and maximum useful-life, to reach industry sustainability. Wind energy providers must meet cost and annual energy production (AEP) goals, while extending the life of every asset. If the goal is to create efficiencies, minimize downtime, and keep turbines running as long as possible, we must proactively manage them.  The solution is to capture data, and quickly compute that data to create actionable intelligence, so operators can use it to make wise business decisions, and meet productivity expectations.  

At almost any wind farm, among the rotating turbines are a handful of turbines that are at a standstill. To maintain productivity, it is essential to determine why some turbines may not be performing as well as the rest. To the naked eye, everything may appear normal. There may not be visible damage, but wind turbine efficiency is determined by many factors – including the blades. When blades are not performing at peak efficiency, owners and operators must decide whether to make the necessary repairs. This decision, to avoid downtime and extend blade life, should be informed by data. 

Capturing Data

The wind industry is driven by data. Whether it’s a 10-turbine farm, or a fleet numbering in the hundreds, keeping a finger on the pulse of every turbine is a necessary challenge. The key to managing inspection data - maximizing production - lies in a process that captures relevant data quickly, then derives value from that data through a convenient dashboard view. When the data is easy to consume, it can be used for both immediate and long-term lifecycle management of assets. 

Seemingly minor amounts of leading-edge erosion can result in 3-5 percent loss in AEP, so it’s critical to catch damage early. Blades bearing heavy wear can decrease AEP by up to 15 percent. In harsher environments that have a higher level of abrasive particles in the air, blade damage is even more pronounced. Early detection matters, so regular inspections make all the difference.

Until recently, all blade inspections were conducted by human teams; they either climbed down the blade, or used scopes to examine it from a distance. Once a turbine was shut down, an inspector would assess the blades in a one-time event. Inspecting all three blades could take a full day, with rope climbing. Using distant scopes was also time-consuming, and did not provide a complete picture of each blade. In the end, the inspector would hand the wind farm operator piece of paper or a hard drive full of images, leaving the management, review, and analysis to the internal team. 

Inspection has been transformed by the adoption of unmanned aerial vehicles (UAVs) to capture damage. What took teams of technicians days to complete, UAV pilots can accomplish in minutes. Operators began working with flight teams to collect inspection data. Each flight required three or more turbine stops to ensure a pilot could capture adequate imagery for each blade. 

Improvements to drone technology have streamlined flight capabilities dramatically. For the wind industry, this means faster, more accurate inspections that can capture all three blades in a single turbine stop. 

Using on-site tools with the newest software on the market ensures that the captured images cover every centimeter of each blade. The software can assess, categorize, and label each high-resolution image automatically in the field, creating a full mosaic image of the blade, and packaging the files for further analysis.

From Data to Action with AI-Assisted Analysis

Back at the office, the files are analyzed within a proprietary software portal, where artificial intelligence (AI)-driven algorithms evaluate thousands of images, looking for signs of wear and damage. In a matter of minutes, the software can automatically pinpoint areas of damage, wear, or erosion on each edge of an inspected blade. 

This technology is shaping the future of the wind industry. Instead of spending hours reviewing thousands of inspection images by hand, the always-learning AI can locate damage in minutes. The software stitches the images together, creating a high-resolution mosaic of each edge of the blade, and highlighting trouble spots. Quality data empowers owners and operators to make informed, proactive maintenance decisions on repairs. 

Data informs important decisions in wind energy production, and helps the industry to increase efficiencies and work smarter. Data helps us to know when to operate and when to conserve, when to repair or replace, and when to push through. Data from UAV inspections and AI-driven analytics are critical to successful blade management programs; they ultimately increase AEP, extend life, and reduce the total cost of repairs.

We’ve come a long way in just a few years. Industry leaders are moving beyond thoughts and ideas, to develop game-changing technology that is changing the face of wind energy production. As technology continues to improve, we will move toward a future of cleaner, more reliable, and less costly renewable energy.

 

Chris Shroyer is President and Co-Founder of BladeEdge. His background spans more than 25 years of experience in technology sales and service, from start-ups to Fortune 100 organizations. BladeEdge is a software portal that transforms raw data from blade-condition assessments and wind farm management systems, into actionable intelligence.

BladeEdge | bladeedge.net


Volume: 2018 September/October