A trial of a virtual power plant by SA Power Networks has shown the possibilities and limitations of storage on residential solar PV systems, as distribution networks prepare for the challenge of rising renewables.
SA Power Networks has won a national award from Energy Networks Australia for innovation for its Salisbury Battery Trial, which is providing important information for the energy industry on the impact of new technologies.
The Salisbury trial rolled out battery storage systems to 100 customers to create a 300kW Virtual Power Plant (VPP). Customers enrolling in the scheme received a battery at a discounted price in return for allowing SA Power Networks the opportunity, if needed, to access power stored in the batteries at times of peak demand.
The project was planned and implemented by the Network Strategy team at SA Power Networks, headed up by Mark Vincent, who also won an ENA award recognising his leadership in planning for transformation of the sector.
“The project has shown that centralised control of batteries can provide network support without reducing the primary customer benefit of bill reduction through solar shifting,” Vincent says.
“The scheme is successfully deferring a $2.9 million network upgrade and has demonstrated that significant wholesale market benefits are also available, based on an analysis of wholesale price variation through the trial period.”
The SA context
South Australia has one of the highest concentrations of solar PV in the world, with about 30% of the state’s 760,000 residential customers owning solar systems.
South Australian rooftop solar PV capacity is around 800MW, compared with a state peak demand of around 3,000MW. Energy generated was 548,000MWh in 2015-16, or 5.4% of the energy delivered to SA Power Networks’ customers.
The high penetration of solar was driven by a generous feed-in tariff of 44 cents/kWh, which ends in 2028, as well as high electricity prices and high levels of solar irradiation. While the state government subsidies are no longer available, reductions in production costs, coupled with further increases in electricity pricing, have maintained customer take-up of solar in South Australia.
SA Power Networks supports the trend to distributed generation, although it poses a significant challenge for network management and future network investment to maintain safety, reliability and quality of supply.
A key issue, says Vincent, is that the generation profile of solar PV is not well matched to demand on the electricity distribution network.
“During the South Australian summer, solar PV makes its major energy contribution between 9am and 6pm. The residential demand at air-conditioned premises is usually highest between 3.30pm and 9pm, but on hot summer days net peak demand on the network in residential areas occurs during the 6pm and 9pm period. Unfortunately, solar generation does little to support the network peak,” he says.
“However, the emergence of residential battery storage products has the potential to store and shift the usage of solar energy into the early evening, potentially flattening load profiles on the network and reducing network costs.”
Batteries also have the potential to be used in orchestrated ways, he says, to manage specific network issues and potentially defer large investments.
In 2016 an opportunity arose to explore how customer-side investment in solar PV and batteries could be an alternative to investing in a network upgrade in Salisbury in northern Adelaide, where the local network was reaching capacity at times of peak demand.
“We wanted to test this hypothesis in a real-world trial to see exactly what benefits could be derived for customers and other market participants, and to what extent pursuing these various benefit streams may be in conflict,” Vincent says.
Salisbury was selected for the trial because a $2.9 million network augmentation was planned for the area. The number of trial participants, 100, was selected to enable sufficient demand reduction to defer the network augmentation for three years. “If the trial outcomes are favourable, the potential for further deferral also exists.”
To manage system integration risks arising from the leading-edge products used, SA Power Networks worked with key technology suppliers Reposit Power, Tesla, Samsung, SolarEdge and Billcap, with extensive lab testing prior to deployment.
Mark said another important element of pre-work was undertaking customer focus groups to test indicative commercial offers.
“The customers’ primary investment driver was the reduction in their electricity bill, but a desire for energy independence was also a key factor, as well as having a source of back-up power.”
Customers were enrolled onto the trial using a combination of direct mail-outs, web-based promotion, advertising in local papers and outbound calling. Customers with relatively high energy consumption and existing solar systems were targeted first since modelling indicated they would receive the highest benefits. Customers registering interest were then screened to assess site suitability. By Christmas 2016 all 100 systems had been installed.
A review was conducted after the first full year of operation, and in the first 12 months, the VPP had: generated 531,639kWh, stored 140,109kWh and achieved $86,184 in system savings.
A customer survey after one year showed 80% were satisfied with the benefits of their system and 85% had changed the way they use energy as a result of their participation in the trial.
“The project has demonstrated that, with appropriate controls, batteries can provide benefits to customers, networks and other market participants, and moreover that these different benefit streams are complementary, not mutually exclusive,” Vincent said.
Most customers have achieved forecast savings of more than $600 a year from improved self-consumption of their solar output. Customers also highly value the availability of backup power.
“Consequently, SA Power Networks and Reposit developed an innovative scheme whereby we can remotely signal the battery to move to a fully-charged state if there are potentially damaging storm conditions forecast, to maximise the availability of backup power in the event of a blackout.”
But Vincent added that an early learning form the trial is that even with batteries, a typical solar system of about 5kW or bigger will still typically export significantly more energy than is imported since most customers’ batteries are fully charged by midday.
Salisbury battery trial load profile – mild weather
And in very hot weather, particularly if there is a level of cloud cover, the battery systems are not significantly reducing the network peak.
Salisbury battery trial load profile – very hot weather
“The way most battery management software is currently configured, a solar PV/battery combination doesn’t help us sufficiently in managing the peak, nor does it help reduce PV exports in the early afternoon,” he says. “We had hoped that batteries might have a much more positive impact on customers’ load profiles, but with a simple ‘solar-shifting’ algorithm they do little to reduce peak demand or peak generation, which are drivers of network costs.”
To maximise the benefits of solar PV/battery installations, smarter algorithms in battery management software are needed to slow down the rate of charging of the batteries and their rate of energy discharge, he says, so demand and generation peaks can be “lopped off”. “We are already working with our control system vendor, Reposit, to develop these sorts of algorithms,” he says. “We also need to make sure our tariffs are designed to encourage battery vendors to configure their systems in this way, and so that customers will also see a benefit. Without those changes to the configuration of batteries so that they charge and discharge in smarter ways, widespread uptake of batteries has the potential to lead to inefficiencies that will require a significant response from us as network managers.”
Provided there are appropriate controls in place, Vincent says battery storage devices are capable of providing benefits to customers and retail and network market participants.
“Although it doesn’t make financial sense for most customers to invest in batteries just yet, prices are reducing rapidly and we think it’s inevitable that customers will invest more and more in battery systems. Our challenge is to make sure that they operate these systems in ways that reduce and don’t increase network costs to all customers.”
OPTIMAL BATTERY OPERATION WITH FLAT NETWORK TARIFFS
In response to flat network and energy tariffs, the battery is usually fully charged before midday and fully discharged by early evening. The outcome is no reduction in peak demand or peak generation.
OPTIMAL BATTERY OPERATION WITH NETWORK DEMAND TARIFFS
An efficient network demand tariff provides an incentive to reduce the network peak demand. There is, however, still no incentive to reduce peak generation.
OPTIMAL BATTERY OPERATION REQUIRED TO MINIMISE NETWORK COSTS
The optimal charge and discharge operation would require some form of incentive to ensure peak generation is reduced as well as peak demand.