Studying turbulence to optimise wind farm performance

Engineer sits atop of an enormous wind turbine (turbulence)
Image: Shutterstock

Turbulence is something we usually only think about when we’re jostling about mid-plane flight, but for University of Sydney Professor Ben Thornber, understanding and confidently predicting how turbulence behaves is critical for the design and operation of wind farms. 

When you consider that an estimated 25 per cent of the energy consumed by global industry is used to move fluids or move objects through fluids, a fuller understanding of turbulence would have profound economic benefits.

But even as we constantly hear about rapid advances in technology, they’re not rapid enough for turbulence researchers like Professor Thornber.

Related article: Gippsland named first offshore wind priority area

“Even using the world’s fastest supercomputers, we can only simulate the turbulence around a few centimetres of a civil aircraft wing,” he explains.

“It will be around 2050 before we have the computing power to simulate in full, the flow over a whole wing.”

Another example of the challenges is the turbulence generated by scramjets. Thornber was involved in what became the world’s largest calculation of turbulence relevant to these supersonic propulsion systems. It ran for the equivalent of 228 years on one computer. 

Rather than wait for technology to make these calculations happen faster, Thornber and the team in the University’s Fluid Dynamics Research Group have focused on developing software algorithms and models of turbulence that can allow current computers to squeeze out the sort of information that future computers should be capable of. 

The future figures prominently in another element of Thornber’s work, which is understanding how turbulence affects wind turbines. As climate change threatens the human liveability of our planet, wind turbines are a cornerstone of the renewable energy response, and Australia has some of the best locations in the world for using wind to generate power, particularly coastal Western Australia and southern Australia around Bass Strait.

Professor Thornber’s research team is working on visualisations to increase our understanding of turbulence.

“This is something we’re really excited about actually,” says Thornber.

“We’ve put an expert team together that covers data sciences, computer software and hardware, structural design and optimisation and people from the Australian Centre for Field Robotics for real-time asset management.”

Even if you remove natural wind turbulence from the equation, the turbulence generated in the wake of an upwind wind turbine enormously affects what happens to any turbines downstream. As wind passes through the forward turbine, it slows down and takes on strong turbulent eddies. 

For the downstream turbines this means a 5 to 20 per cent drop in efficiency from the slowdown and greater structural fatigue from the increased turbulence. Some wind farms will even turn off turbines in certain circumstances to prevent this fatigue damage. Add together the lost power production and fatigue damage repair and the cost to the wind farm industry is around a quarter of a billion dollars every year in Australia alone.

Related article: Italy’s Enel to build first wind farm in Australia for BHP

“We’re spending a lot of time on this, and applying what we’ve learned from helicopters,” says Thornber.

“The goal is to develop a digital twin of the wind farm and feed in real-time data from the actual wind farm. That way we can tune the model and simultaneously use it to optimise the performance of the real wind farm.”

The team is in place, the need for this approach is clear, and there is no shortage of ideas,” Thornber says.

“We are co-developing our digital twins with the world number one producer of wind power, Iberdrola. Australia is so well set up for a renewable energy future. It’s an area we have to go into. And I know the team here can make a big contribution.”