Eight of Australia’s large-scale solar farms will partner with global solar forecasting company Solcast to utilise the latest in predictive technologies. This follows the Australian Energy Market Operator’s (AEMO) announcement that solar farm operators can ‘self-forecast’ the amount of electricity they generate.
The market change, first announced last year, paves the way for solar farm operators to save on costs by reducing the volume of frequency control services and to better manage supply and demand across the National Electricity Market (NEM). Improvements in both will ultimately save on how these costs are borne by the consumer.
Historically, AEMO has been responsible for forecasting how much electricity will be generated by renewable generators, however these in-house technologies have resulted in higher than expected costs for solar farm operators.
For solar forecasting vendor, Solcast, the timing proved opportune. The company had already completed a demonstration pilot project at the 53MW Moree solar farm in mid 2017, that yielded 10-15 per cent improvement over the AEMO models. Solcast operates a global network of satellite ‘nowcasting’ technologies, which generate up to 600 million unique forecasts every hour, technology it has been able to apply for large solar sites in Australia. Further refinements in the technology over the past year have introduced the use of real-time solar farm power data feeds and ground based ‘sky-imagers’ which record images of the local cloud cover.
Now with nearly $800k of ARENA funding, this forecasting solution is being operationalised on the National Electricity Market, where 5-minute ahead forecasts will be submitted to AEMO operationally by project partner Dispatch Solutions.
“Ultimately, this is about both energy and financial security,” Solcast CTO Dr Nick Engerer said.
“It is an energy security issue at the market level, as the utility-scale solar segment is adding multiple gigawatts of generating capacity each year. With each new solar farm, the importance of accurate forecasts grows.”
But as Dr Engerer alludes, it is not just energy security at stake, the ongoing costs of inaccurate forecasts can negatively impact the expected return on investment in investment portfolios with solar exposure.
“Accurate forecasting is a key enabler for high-penetrations of renewable energy in electricity grids,” Nick Morley, Project Development Engineer at First Solar, said.
The Manildra solar farm, developed and operated by First Solar, features in the project, and will begin submitting self-forecasts to AEMO through the project by the middle of this year.
“We are excited to be participating in this trial, installing technology which has the capability to perform both near and long-term forecasting,” Mr Morley said.
First Solar also believes site level forecasting will help decrease frequency control costs, as well as help optimise the use of dispatchable generation and energy storage. It also believes site level forecasting will improve system reliability and bring down electricity prices.
Tom Geiser, Senior Market Manager at Neoen, agrees.
“Forecast inaccuracy results in penalties, so it’s in our interest to provide the best evaluation we can. Ultimately, the improved forecasts will drive better efficiencies for AEMO and create stronger alignment amongst everyone involved, allowing power producers like Neoen to more effectively deliver sustainable, reliable and competitive energy to Australians,” Mr Geiser said.
Neoen and FirstSolar are not alone in their support of the project, and are joined by a broad consortium that includes Wirsol, Foresight Group, Edify Energy, CWP Renewables, Canadian Solar and New Energy Solar.
“This is about building the solar future,” Dr Engerer said.
“We’ve aimed to bring as many partners onside as possible, so that we can quickly get the best forecasting technologies operating in the NEM.”