The fluid dynamics of offshore wind farms: modelling, design and optimal control

This week’s seminar was given by Dr Georgios Deskos of the Department of Aeronautics. He focussed on his work around the fluid dynamics of offshore wind farms, a highly complex problem which has a massive impact on the design and control of what might become a big contributor to the UK’s energy mix in the future.

Evidence for the existence of wind energy converters can be traced back to Heron of Alexandria and his pioneering wind wheel used to operate a pneumatic organ, the Persian Panemones, and the windmills found in the Dutch countryside.

Dutch windmills at dusk

Over the last four decades however, wind energy has grown significantly, and its transformation has raised hopes that in the foreseeable future it may become a sustainable energy system. In 2017 the German government accepted the first subsidy-free offshore wind farm bid, while in 2018 wind power helped the UK go coal-free for a total of 1,000 hours. Moreover, the UK went three days in a row without needing to utilise coal for the first time since the industrial revolution.

Remaining challenges

Amongst the recognized remaining issues and challenges associated with wind energy systems are their low power density (energy produced per area), wind power intermittency and structural survivability under extreme loading, to name a few.

Aerial view of turbines showing air turbulence

To increase the energy density of the system, larger-scale wind farms (both in terms of the size of individual turbines and their number is concerned) have been built both on- and offshore. The key idea lies in capturing the larger wind energy resource available at higher altitudes, while increasing them in number to cover as much area as possible. In such systems, the operation of wind turbines results in the creation of wakes which interact with each other and with even larger-scale atmospheric turbulence flow phenomena.

Modelling the fluid dynamics of wind farms has attracted the interest of the research community as there is strong evidence that the complex turbulent flow phenomena can modulate the power output and mechanical loads on turbines. Subsequently, optimal farm layout designs (micro-siting) and farm-level control has become the main focus of the industry.

The common denominator in the three aspects (modelling, design and optimal control) is the wake-to-wake and wake-to-turbine interactions. A better understanding of such interactions can lead to better designs and more viable wind energy systems.

Modelling and HPC

Modelling the wind farm wakes can be achieved through both numerical and physical simulations. Numerical models which remain the cheapest option available to engineers can also vary from low-fidelity empirical models run on local machines, to high-fidelity numerical simulations run on high-performance computing (HPC) platforms using state-of-the-art numerical algorithms and thousands of computer processors.

Figure 2 High-fidelity wake data generated using the state-of-the-art wind farm simulator WInc3D/Incompact3D (Unpublished data Deskos et. al.).
High-fidelity wake data generated using the state-of-the-art wind farm simulator WInc3D/Incompact3D (Unpublished data Deskos et. al.).

For the latter, the recent developments of processors’ technologies (still following Moore’s law) have made numerical simulations a preferable option for research purposes. With such simulations, analysis of time-averaged wake effects and transient flow phenomena (including the interactions of the turbulent structures with the turbines) have shed light on the role of fluid dynamics in the operation of large-scale wind farms. In addition, thanks to our better understanding of the fluid dynamics, better wind farm designs and active-control strategies have been adopted by the industry to mitigate the adverse wake effects.

But, can high-fidelity models be used directly to drive optimizations?

While the research community has already used high-fidelity simulations within an adjoint-based optimisation algorithm to find an optimal control for an idealised wind farm1, the industry still prefers simpler analytical models with which to perform optimisation studies.

The challenge that the research community faces is the need to improve high-fidelity numerical models and their use within optimisation frameworks, while keeping an eye at the rapid advent of HPC systems.

Georgios (Yorgos) Deskos

Georgios (Yorgos) Deskos

Georgios (Yorgos) Deskos is a postdoctoral researcher in the Department of Aeronautics at Imperial College London where he works on the modelling, design and optimisation of offshore wind farms as part of the EPSRC funded project: FENGBO-WIND (Farming the ENvironment into the Grid: Big data in Offshore Wind).

He recently completed his PhD work on “Numerical simulations of wind turbine wakes” for which he won third place in the prestigious ERCOFTAC Osborne Reynolds Day 2018 competition.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.