An alliance of researchers and distribution networks putting Victorian smart meter data to work has vowed to share its resources, writes Jeremy Chunn.
To build an intelligent energy system you need intelligence. Where successful delivery of that energy relies on the coordination of multiple generation sources and optimal utilisation of transmission and distribution networks — with a satisfying experience for the bill-paying consumer — that intelligence needs to be shared.
To open the door a little wider to a national approach, the Centre for New Energy Technologies, or C4NET, has vowed to provide advanced metering infrastructure data to anyone working towards the transition to clean energy. This includes aggregated consumption data, information about the penetration of localised renewable energy sources and certain network operational information. The callout is intended for anyone, but a few examples might be councils who want to understand greenhouse reporting or where best to locate community renewable projects.
“We’re there for anybody, as a single point of contact for them to come to,” says C4NET chief executive James Seymour.
An independent not-for-profit company, C4NET’s members include all the Victorian distribution networks, six universities in the state and associate members, including retailers, who may participate in individual projects and activities. “It’s collaborated research,” Seymour says. “We coordinate between government, industry and the universities as to what are the challenges and then agree on research programs or projects.”
The Australian Energy Market Operator contributes technical advice and has participated in some projects.
The centre’s Victorian emphasis is a result of two things: initial grant funding from the state government and, probably more importantly, the ability to base its work on a deep and fast-flowing stream of data from smart meters installed in every home in Victoria. With all five DNSPs in the state involved in C4NET’s work, that makes up 60 per cent of all smart meter data available in Australia.
The clean energy transition is a national ambition but it is inconveniently broken up into the organised connection of assets in the east coast National Electricity Market — which includes five interconnected states — and Western Australia’s South West Interconnected System. This gives politicians and the media plenty to get confused about. That’s why research conducted in a state that is knee-deep in data will be valuable to all its fellows within the Commonwealth.
“These are national problems,” Seymour says. “It’s very rare the challenge is so localised to one state; there might be nuances but the underlying challenge is generally the same. We’re happy to use the Victorian data to inform issues elsewhere in the country; or we can use data from other areas in the country to inform Victoria.”
Victoria may be well ahead with wind, at about 17.5 per cent of demand, but it trails many states when it comes to rooftop solar, at 6.9 per cent compared with 15.6 per cent for South Australia and 13 per cent for Western Australia. Could there be better coordination of research around the country? Seymour pauses. For a start, research relies on data, and data is not always easy to get. It takes some caution to navigate compliance and privacy rules, he says, “for the purposes of research — it’s not about giving out private data”. The C4NET alliance is available to work with data from other states, he emphasises.
As most solar energy generated by the three million rooftop PV systems around the country is exported around midday, the distribution networks are beginning to strain. “There is a real challenge managing the physical flows and we can see it with data really clearly,” Seymour says. Most houses with solar today, he says, are exporting around 70 per cent of the energy they generate. This is pushing distribution infrastructure in parts of some networks to their capacity limits, and the owners of those networks are being asked to deal with it.
That raises all sorts of questions, he says, such as the balance of social equity between those who have solar and those who don’t; equitable access to the grid from those who already have solar to those who want to put on solar, and; how to develop policies and market structures in a way that suits all consumers.
“A lot of that is predictable, and it can be modelled and anticipated,” he says. “We’re going to see this in a heightened state as we move to electrification of other elements of the economy. That will all put pressure on that physical infrastructure, but it’s reasonably predictable from a data-fed analysis where you have really good data.” For policy, programs and network planning to succeed, they must be based on evidence.
The more connection points that send out real-time data, the easier the task to plan for a grid with accelerating levels of PV and battery storage. For Seymour the five-to-10-year outlook is pretty clear, but it’s less clear whether enough forward planning is in progress. High feed-in tariffs are a thing of the past, linked to a falling cost of wholesale electricity. The game will be to connect the storage that will accommodate a shift in demand. “That will change as the cost of storage comes down,” he says. “If storage was free, we would need no change of demand.”
The next wave of storage is rolling towards us in the form of electric vehicles, or “storage on wheels”, in Seymour’s estimation. As a shiny new fleet of EVs slowly pushes Australia’s guzzlers into the gutters this aggregate form of storage will offer a “huge opportunity” to the grid. “If we don’t have storage, we don’t have any choices,” he says. “When we do have storage available, we have choices.”
Will EV owners want to share their charge around? Research out of UNSW has shown battery-owners can be reticent to sign up to virtual power plants if they feel a powerful, distant orchestrator is getting better value out of the relationship. Seymour reckons it may take a big shift in attitudes, but it is possible.
“In the absence of trust, no-one is going to engage. Trust has not been high in the electricity industry, so that’s the opportunity. Consumers are going to rule this, and unless they have trust they won’t engage.” Whether the incumbent providers or new entrants turn out to be the trusted ones remains to be seen.
It’s worth acknowledging the storage that’s already there, of course. With 20 per cent of domestic hot water systems in Victoria running on grid power overnight, there’s an obvious benefit if they were timed to work during the day. “Many people can shift those to solar; the amount of solar generation would largely cover that need,” says Seymour, estimating total capacity of those heaters as equivalent to a 3GW battery.
Batteries, so long at the sidelines, are beginning to pop up all over the place. Melbourne networks CitiPower, Powercor and United Energy have added storage up power poles and on street corners to stabilise voltage levels affected by high penetration of solar. On top of that, the state government is handing out thousands in subsidies to homeowners who want to buy residential batteries.
C4NET’s research into the impact of household batteries on networks found they can reduce exports in cool weather by 20 per cent and imports by 60 per cent, but battery-owners export significantly more on hot days than non-battery households with PV. “Now what we need to do is understand why that is,” Seymour says. “Within those numbers there were some quite specific cohorts of activity around settings of battery management systems. We can’t see those.”
Some batteries were charging overnight, some during the day, and others did a bit of both. Were owners running the default battery management settings, had they customised them or were they trusting a third party? “That will be the next area of research.”
The study allowed the researchers to detect hyper-localised voltage impact challenging the hosting capacity of circuits to transformers on networks. “The distance a solar system and battery is from the transformer has quite a large impact,” he says. “The balancing need is quite complex. That’s a challenge for distributors to provide a seamless, ubiquitous service to consumers. Consumers don’t understand that and shouldn’t be expected to.”
Happily, the tools that can help iron things out have already been identified. They are front-of-meter neighbourhood batteries to complement behind-the-meter residential batteries. “That is a keen area of interest.”
Another interesting piece of ongoing work is the mapping of low-voltage networks, and it’s “a cool project” in Seymour’s words. In the old days of energy, networks never needed to know much detail about the final connection points. It would have cost a lot to do so, for a start, and everyone would have had to pay for that. “You never needed to know that, when you had a centralised distribution system.”
Times have changed, and the hosting challenge is complicated by generation and storage distributed throughout the grid. The supply-demand balancing act, therefore, is vastly more complex. It all comes down to the distance between a connection point and the nearest transformer, which sounds simple enough until you realise how much is not known about the nation’s suburban networks. With smart meters supplying data that can be parsed by machine-learning algorithms, the picture becomes much clearer, Seymour says.
“When it can be done algorithmically, that’s a scalable solution that can be applied right across the state — without having to invest in more hardware, more monitoring, more assessments, more people mapping these things.”
It’s fair to say the distribution networks have limited ability to say one way or the other how much solar should be connected at a localised level, he says. Also, any solutions that allow far keener visibility along networks’ poles and wires will be able to pick up faults that may be serious safety issues.
“That’s the real power in the data.”