Physics informed, data driven wind farm digital twins

The challenge

The transition to an energy mix dominated by renewable sources is critical to mitigate further and more substantial impacts of climate change. Wind power is a key pillar in this transition, expected to generate one quarter to one third of global energy by 2050, with many individual States targeting this level by 2030. Our research aims to make boost the productivity and profitability of wind farms through the development and implementation of high accuracy, physically informed, data driven modelling and optimisation of the whole wind farm.

 

Why this research is valuable

Our work targets an increase the power produced by a given wind farm by 5-20%, a reduction in costs associated with unscheduled maintenance which currently form 16% of the cost of wind power, fast identification of faulty wind turbines. When combined, these represent $250m of inefficiencies per year in Australian wind farms alone.

 

Research themes

Our research themes and associated research questions are:

  1. Data Driven Computational Engineering: how can near-real time, high fidelity, physics informed data driven models transform wind farm profitability and operations?
  2. Optimise while you learn: can statistical approaches to data driven high and low fidelity modelling deliver accurate faster-than real time digital twins needed to provide accurate forecasts?
  3. Power system and Structural: can power systems, aerodynamic, structural and predictive maintenance models be combined to both improve power output and reduce unscheduled maintenance?

 

Want to know more?

This collaboration aims to transform wind farm digital twins. We work with exceptional industry, government and academic researchers worldwide and if you would like to join our research network as a collaborator or student, click here to contact the team.