Adapting global technology to the Australian market, improving negative price compensation by 40%

One of the biggest challenges facing Australian renewable asset generators is economic curtailment. Curtailment is a deliberate reduction/enforced loss of potentially useful energy. With economic curtailment, a generator chooses to bid out of dispatch to avoid losing money due to negative pricing, this can significantly impact a generator’s revenue due to the wasted energy.

Financial risk from curtailment in Australia can be mitigated with technology already being used for overseas projects. At RES, we use Machine Learning (ML) to construct power curves. The trained power curves provide an appropriate representation of the ideal yield and efficiencies to expect, based on real site data.  

As a global company, RES draws upon the experiences from our international teams to develop and improve Australian processes. A data analyst from the Australian team shadowed the UK team for three months, learning the already established ML power curves system in Europe and tailoring the system to the local market in Australia. 

This innovative data-driven approach enables RES to conduct adaptive monitoring and performance benchmarking across our sites. Additionally, through the power curves, RES can provide technical reviews of the operational performance against modelled ideal yield to identify improvements financially. 

In one scenario, RES identified the expected output from a 100MW+ wind farm, challenged the current methodology and suggested adjustments with ML power curves. This improved negative price compensation by 40% in a quarter, resulting in significant savings for the client.