Evoenergy is exploring numerous ways to manage demand through virtual power plants.
Australian utility Evoenergy is carrying out one of the most comprehensive demand management trials ever in Australia, using a combination of batteries and traditional demand response.
One of them, a virtual power plant used to avoid a substation upgrade, could save Evoenergy around AUD $2 million (USD $1.6 million).
Evoenergy, an electricity and gas provider in the Australian Capital Territory (ACT), has undertaken seven demand-management tests in the last six months.
The largest of these, in February, involved three types of participants: residential customers, contracted commercial customers, and people with on-site battery systems.
During the trial, coordinated from the energy company’s Canberra control room, Evoenergy sent participating residents text messages asking them to cut their electricity use. Commercial customers were asked to curtail loads or switch to alternative power.
The residential battery systems, meanwhile, acted as a VPP dispatching renewable electricity into the grid.
Residents who cut their energy consumption in response to a text message got no formal incentive; customers offering up their battery systems for use in the VPP “were offered a payment,” said Glenn Pallesen, acting general manager at Evoenergy.
In its multi-demand response test, Evoenergy signed up 402 batteries and linked them into the VPP through control systems belonging to Reposit Power, Evergen and ActewAGL.
During the test, 367 batteries responded to a signal from Evoenergy and dispatched energy into the grid.
The batteries delivered 570 kilowatts of power, which added to 2.1 megawatts of curtailment from large customers that delivered 2.6 megawatts of demand response. Evoenergy said the trials had no negative impact on the safety or reliability of the grid.
“At scale,” said Evoenergy, “we would be able to strategically and quickly reduce enough load on the ACT electricity network to avoid load-shedding in almost all circumstances.”
However, an analysis of trial results shows the impact of curtailment by residents was minimal. In two tests carried out in September 2017, only 6 percent of customers in the trial area responded to a text message from Evoenergy.
“We only expected to see a small impact,” Evoenergy admitted.
A further test last month revealed potential limitations in the VPP concept.
Working through Reposit Power smart controllers, Evoenergy requested 333 battery systems to start discharging for an hour from 8 p.m. and 325 of them responded, delivering almost a megawatt to the grid. After 30 minutes, though, the power began to drop, ending at 600 kilowatts by the end of the hour.
Evoenergy said the test yielded “the clearest result of the VPP trials” because there was no PV generation at the time.
Nevertheless, once it is fully operational Evoenergy expects the VPP to deliver power for one to two hours, Pallesen said. The experience so far has given the power company the confidence to submit a substation deferral plan to the regulator.
“In our most recent regulatory determination submission, Evoenergy has included a demand response proposal to defer a multimillion-dollar zone substation for two years, until 2025-26, through the use of non-network demand management,” said Pallesen.
“This will involve the use of residential and network-scale batteries in a geographically specific VPP for a new residential estate development.”
By 2020, Evoenergy predicts more than 5,000 ACT dwellings will have solar systems paired with batteries, capable of generating up to 36 megawatts or 5 percent of the territory’s peak load.
“This would be double the amount of load we curtailed, with the support of the community, which prevented load-shedding during the heatwave crisis in summer 2017,” said Leylann Hinch, Evoenergy branch manager, asset and network performance, in a press note.
During the February 2017 heatwave, Evoenergy asked customers to curtail 18 megawatts of power as peak demand topped 637 megawatts. It came close to reaching 650 megawatts, which would have led to load-shedding.
“We avoided load-shedding then by going out to our major customers and the public, requesting them to reduce consumption,” said Hinch, “but we don’t believe it’s best practice to do this reactively during an emergency, as the result is uncertain.”