A joint winner of AGL’s 1,000MW battery storage tender is ready to ramp up its presence in Australia, bringing valuable learning from across the globe.
The world loves reliable power. To clean up our air, however, energy systems are being converted to unreliable renewables. Tomorrow’s grids will rely on massive amounts of storage if they are to be as dependable as today’s.
In Australia, work has begun on building the storage required to ease the way for wind and solar plants to come. The incumbent energy suppliers – owners of coal and gas plants – are part of the action. In January, AGL announced it had chosen Finland-headquartered Wärtsilä and US-based Fluence to supply up to 1,000MW of grid-scale battery storage.
In a grid where flighty renewables will replace predictable coal and gas, large batteries can yield revenue for their owners by trading the frequency control ancillary services (FCAS) markets, which includes six contingency markets (6-second, 60-second and 5-minute markets to lower or raise frequency); and two regulation markets (to lower or raise frequency).
Owners rely on software smarts to enable “revenue stacking”, where a battery can tap multiple revenue sources. Other than transacting in the FCAS markets it could mean managing a power purchase agreement with a solar plant so that a customer is assured of supply regardless of the weather.
“A battery system at rest, neither charging nor discharging, doesn’t create value for the owner,” says Wärtsilä vice-president of energy storage and optimisation Andy Tang. “It’s similar to an aeroplane that when it takes off any unused seat is not monetizable.”
Land of opportunities
Wärtsilä is active in Australia, Tang says, scoping opportunities for storage at microgrids to replace single-wire transmission lines and at existing solar and wind plants with offtake agreements that would benefit from firming. “There is this opportunity in Australia to move to the next generation PPA and sell pure renewable power, possibly for a premium,” Tang says. “A lot of the work we’re doing in island microgrids becomes very relevant for Australia.”
Tang is also watching for arbitrage opportunities within the eight FCAS markets. “We think there is a way to enhance markets but also enhance revenue for the asset owner by being able to manage supply and demand for each of these products.”
Neoen’s 150MW Hornsdale Power Reserve in South Australia, powered by Tesla technology, has harvested the FCAS markets over the past few years to great profit but margins are sure to slim as more storage assets trade in the same pool.
“The more successful storage is, the more flattened those markets will become,” Tang says, and trading arbitrage between the FCAS markets will further reduce volatility. “That’s exactly what South Australia has seen with the Hornsdale system.”
Given the characteristics of our grid, any actions that stabilise the system will be highly valued as Australia shifts away from coal. (Although based is California Tang has familiarised himself with the National Electricity Market through regular visits to Australia over the past four years, before covid-19 put an end to all that.) “The Australian energy market is really exciting,” he says, before listing weaknesses in the NEM: reasonably weak grid; some inadequate interconnection; localised connection issues, especially in Victoria; radial design in many parts, without alternative routes. “The world is interested in the Australian grid because it’s an example of the energy transition in action.”
Thermal assets will remain
Storage in whatever form (batteries, pumped hydro, hydrogen … one day) is a vital component of any energy system with medium to high levels of intermittent renewables, and Wärtsilä is seeing some fine results from projects around the world. Tang cites a job in the Azores, about 1,500km west of Portugal, where the addition of storage to a wind-solar-diesel hybrid plant saw the levelized cost of energy fall from about US26c/kWh to about US16c/kWh. Renewables penetration rose to 60% at the site.
“We were able to forecast the wind and the solar, forecast the load demand and optimise when you would be taking energy from the solar, when you would be taking it from the wind and when engines would be turned on and off,” says Tang, speaking to EcoGeneration from San Francisco.
The world is on the path to an energy transition where renewables will dominate generation. As more dispatchable supply is retired, providers of “flexibility services” will be called on to “somewhat instantly” offer backup when clouds come over or the wind drops. Batteries are the obvious tools for the task, but there are times they won’t be enough. “Sometimes the wind doesn’t blow for days on end,” Tang says, “and sometimes you will have a storm system that may linger.”
Long-duration battery storage is needed, yes, but thermal assets will still provide a vital, reliable service. “Do I think we should have 24 hours of battery storage,” Tang puts it, “or do I think we should be looking at the usage of thermal assets … and then migrate the fuel over time, so we get to carbon-free fuel to power those [thermal] power plants.” Hydrogen and biofuels could qualify, or any fuel or process that creates heat.
Modelling by Wärtsilä of a theoretical 100MW city in Germany to reach 100% renewables with an over-build of solar and over-build of storage resulted in an unreasonable cost per kilowatt for consumers – “three or four times what they’re paying today” – but backing that up with a thermal asset expected to run a few hundred hours a year reduced the energy cost to the equivalent of today.
“Once the politicians get over their soundbites and we actually have to deploy this stuff and pay for it, that’s when the rubber really hits the road on how do we do this,” he says.
In the meantime, storage proposals get bigger and bigger.
Energy storage was added to Wärtsilä’s offering less than four years ago when it bought Greensmith Energy, a company founded in 2009 with the ambition of building a technology-agnostic control layer for storage. “We discovered there weren’t any hardware programs out there, so we ended up becoming an integrator,” he says. “Along the way we built what we believe to be one of the most robust and state-of-the-art software control platforms for energy storage.”
Tang can’t supply much more detail about the AGL deal other than saying the Australian company’s 1,000MW commitment will be divided between Wärtsilä and Fluence, with each supplier working on distinct systems. More detail will be available as AGL and Wärtsilä work on the generator performance standards process with the Australian Energy Market Operator.
Wärtsilä worked with AGL on its 211MW Barker Inlet Power Station in South Australia, a flexible gas generator required to help integrate renewables into the grid.
AGL has committed itself to build 850MW of grid-scale battery storage by 2023-24, and last year announcing plans to build batteries at Loy Yang A power station in Victoria (200MW), Liddell power station (150MW) and Broken Hill (50MW) in NSW and Torrens Island (250 MW) in South Australia.
“Our grid-scale battery plans provide critical firming capacity to the market and will play a leading role in the energy industry’s transition over the coming decades,” said AGL chief operating officer Markus Brokhof. “Grid-scale batteries allow AGL to leverage excess solar and wind generation to provide capacity when renewable sources are not generating.”