A trial of a 1,000-battery virtual power plant on the South Australian network has boosted solar exports, supplied much-needed services to the grid and earned tidy revenue.

For the past five years or so virtual power plants have been held up as an ideal way of leveraging the potential in distributed energy resources, especially residential solar and battery storage. If a legion of batteries can be connected and instructed to act as one generator or as an aggregated load, this “virtual” power plant can be called upon to leap to action when the grid needs help. Financial benefits will follow, the theory goes, as the entrance of renewables sees the frequency control ancillary services (FCAS) markets become more volatile. As VPPs soothe the market, energy prices will ease lower.

But will it happen in practice? Research has shown battery-owners can be reluctant to share their stored energy with an aggregator and in some cases architects of VPP trials have found it takes a long time to get people interested. Also, let’s face it, batteries are expensive.

A VPP of 1,000 Tesla batteries installed in social housing in South Australia provides almost ideal conditions for a trial, where participants are on a low energy rate and have no control over the aggregator’s actions.

EcoGeneration spoke with SA Power Networks future network strategy manager Dr Bryn Williams, who is in no doubt about the dynamic role a VPP can fulfil.

When has the VPP proved most useful in the grid?

Bryn Williams is convinced the value of virtual power plants makes them worth the effort.

The VPP trades in the wholesale market and also provides frequency support services. I think the frequency support function – FCAS – has been the most valuable. Over the period of the trial we had a number of incidents, for example the separation of South Australia from the rest of the NEM, which has increased the demand for those services and the value those services can provide.

Tesla and Energy Locals are the VPP operator, so this trial is about us supporting them trading in the market. VPPs and batteries in general can respond super quickly. The fastest response market we have today is the six-second market and that is where the VPP has been trading most effectively, because they can respond much faster than that. The faster you can respond, the higher the value of the service.

There must be some anticipation of the two-second FCAS market opening up.

There is a lot of talk about new markets for fast frequency response, two seconds or less. In future we’ll have things like synthetic inertia, which Tesla is trialling with its big battery [the Hornsdale Power Reserve]. These more high-value services will take advantage of the fact that batteries can respond in milliseconds, not just in seconds.

How has the VPP helped you as a network owner?

It is showcasing how VPPs and networks can integrate and operate together effectively in the future. That’s what this whole trial is about. We are helping the VPP on the one hand because we are able to increase the solar export limits for VPP customers most of the time – we have a dynamic export limit, not a fixed one. We only need to reduce [exports] at those times when we have capacity constraints and in areas where we have capacity constraints. The rest of the time we can open up more capacity [using the VPP]. That’s how we can help the VPP to do more in the market. In this trial what would have been a 5MW generator effectively can turn into a 6-8MW generator most of the time throughout the year by using this dynamic export limit.

In terms of how the VPP is helping us as a network, primarily it is because the VPP understands our new time-of-use network tariff and that gives a general signal as to when are beneficial times to use the network. For example, the VPP will charge from the network in the early hours of the morning and use that energy to supply the customer’s home in the morning peak, which helps take load off the distribution network and flatten out those peaks. Similarly in the afternoon it’s shifting energy from the middle of the day to the afternoon.

The VPP was able to take advantage of spikes in the spot price.

You’re saying batteries are sometimes charging before the sun has risen? The VPP will sometimes charge from the grid, not solar?

Yes, it’s because we’ve introduced this time-of-use network tariff that provides that incentive. This is a way we can provide incentives for customers to help the network without complicated contacts or network support arrangements. It’s a general means by which we can incentivise good behaviour and share the value of that.

As VPPs grow there will be opportunities for us to engage and provide specific point solutions. If we have a particular substation that’s overloaded, for example, we can engage in a specific contract for that and do some peak shaving – a classic demand response sort of contract. Given the size of VPPs at the moment and nature of constraints in our network we haven’t been testing those sorts of arrangements through the trial because we don’t have those opportunities now.

There have been times when parts of your network have run in reverse and you’ve had to increase voltage settings. Will a VPP have much effect on voltage levels?

That’s an example of where we’ve got lots and lots of rooftop solar feeding into the system and not a lot of demand. That energy is flowing up through our network and being consumed by commercial and industrial customers connected higher up in the network and non-solar customers, and ultimately some of it will be exported interstate. The system is operating OK in that configuration.

But during those times because there is very little spinning generation running – and it’s spinning generation which traditionally keeps the system stable – the system is at its least stable. That’s when frequency support is most valuable. Batteries can respond really quickly and help iron out ripples in frequency as supply and demand rise and fall.

The report shows some participants’ batteries are far more active in the VPP than others.

It’s the ones who have most benefited from having a raised export limit. Depending on the configuration of [the home system and load] and where the house is connected in the network, that will help determine the benefit.

Are you happy with how the VPP has performed in the trial?

Solar exports were significantly higher for some households.

We’re really pleased with it. There’s an awful lot of debate in the industry at the moment about the future of the market and how we need to evolve to make sure customers’ resources and virtual power plants can get the most value. We see this project as showcasing a vision of how that can work.

It’s a VPP that is trading really actively in the market, it’s benefiting customers and it’s doing it with an understanding of the distribution network. This is the first time we’ve really put together these elements of the jigsaw puzzle in a real-world field trial with real customers and real market participation. It’s a template for the future, we think.

What’s your feeling about AEMO’s expectations for VPPs as defined within the Integrated System Plan?

The energy transformation is happening even more quickly than AEMO had forecast even in its most aggressive forecast, so they are updating their ISP at the moment. In general, we see batteries have a really, really important role to play in the future market because they are super good at providing this very fast balancing service, which is really important when you have less and less spinning machinery and more solar and wind farms. Grid-connected and small batteries will both play a part.

In South Australia the government has been encouraging through subsidies the uptake of home batteries and requiring them to be VPP-capable. In our view, VPPs are going to be a crucial part of the energy mix in the future. They are going to be actively working behind the scenes to keep the system stable, making microscopic corrections all the time to frequency, which will be a key to retiring our old-school generators and bringing more and more renewables online.