Energy users big and small can tune their load to temper stress on the grid, writes Jeremy Chunn, but the benefits to big industry and networks may be larger than many participants anticipate.
Demand response is like Latin dance. You can understand the concept immediately, but that doesn’t mean it’s easy to take part. As part of a series of discussions on smart buildings hosted by EnergyLab on the campus of the University of Technology Sydney, a panel in October looked at the importance of demand response and how a market is slowly taking shape in Australia.
Demand response is one of the biggest ways businesses can cut their cost of electricity, says Marija Petkovic, managing director of Energy Synapse. For energy intensive businesses that have seen electricity costs double or triple in recent years demand response can be “incredibly valuable”. It can also offer relief to the grid during extreme peaks and may be enough to defer investment in peaking plants that might be expected only to run a couple of hours a year.
“If you use demand response you can avoid investing in that very capital-intensive asset,” she says. Beyond that, demand response is a useful tool to have at hand as the energy system transitions to renewables.
“As we move to more and more renewable energy that is variable, like wind and solar, we are losing control of the supply side of the equation. Harnessing the demand is going to be incredibly important in helping to integrate and facilitate the clean energy transition.”
The physical electricity system is not flexible. Networks are built to meet the absolute highest demand, says Chris Dunstan, research director at the Institute of Sustainable Futures. “If there is roughly $100 billion of assets in the electricity sector, about $10 billion of those assets only get a couple of days a year when there is peak demand.” If that peak demand can be reduced, he says, customers will save a lot of money. “Demand response was always a good idea, but as we move towards renewable energy it becomes absolutely essential if we’re going to have an affordable clean energy transition.”
At Ausgrid, manager demand management and forecasting Craig Tupper is working on a demand response solution for households around Maitland, NSW, that will guarantee reliable supply during the 40-60 hours a year when the network is stressed. The cheapest and easiest solution, the network found, was to seek to “moderate demand” from a subset of residential air-conditioners.
“With network investment, the reductions need to be in the right area,” Tupper says. “If they’re not, they don’t add value.”
About 2,800 customers were approached to take part and about 100 have registered, of which about half own appliances that are able to be managed. “We’ve done these trials before and about 95% of customers don’t really notice a difference.” Those who take part will be paid about $200.
The cons of air-con
Peak demand across the NEM is about 33GW, typically in the late afternoon on the third day of a heatwave, says Dunstan. The main contributor is residential air-conditioning, which is estimated to account for 10GW on such days. “If we can reduce the demand in air-conditioning, then we’ve got the potential to reduce the cost not just for those customers but for all customers across the grid,” he says.
Since Australia has one of the highest per-capita levels of rooftop solar in the world, it’s safe to assume many of the houses that turn to their air-con on those days will also have PV systems. “If we can shift demand from late afternoon to the middle of the day, when those houses are not just generating electricity but in most cases exporting it – if we could cool those homes in the middle of the day, then when people come home they’ll walk into a nice cool space,” Dunstan says. If that’s the case, air-con units can be switched off or set at reasonable levels to maintain comfort.
“If we can use some smart technology to know when the constraint is going to happen and hook that up to air-conditioner units to [turn on in the middle of the day], then we’re able to provide essentially free cooling and reducing the cost for all customers,” he says. The same logic would work for commercial spaces with PV systems, such as shopping centres and hotels.
Things aren’t as straight forward when it comes to industry, Petkovic says. “You’re affecting the revenue of the business with every shutdown and interruption,” she says. “It’s a much more complex problem for industrials. To get the best result for demand response you have to optimise the entire operation and supply chain.”
This might involve using product storage as a buffer, for example, if a high-energy-price event is anticipated. In that case customers won’t notice any interruption, even though a facility may be taken offline. “There are a few more moving parts in an industrial context.”
For big business, demand response needn’t only be an option when spot prices approach $14,000/MWh and there is danger of a blackout. Petkovic prefers a looser definition of demand response, where it is a temporary change in energy usage in response to some signal, which could be a higher spot price, negative spot prices, a frequency deviation or an effort to help manage grid congestion and peak demand. “It’s quite a broad concept,” she says.
“If you’re going to invest in equipment and the smart controls to enable you to supply demand response you then want to be able to do that in as many markets as you can to get the maximum return on your investment.”
Where have you been?
If demand management is so sorely needed in the grid, why does it not exist at the required scale? It’s a matter of looking back at recent history, says Dunstan. During the 1960s and ‘70s, electricity demand grew between 6-8% per annum. Coal plant was moved away from the cities and larger, more efficient units were established near coalfields. Since then, demand has flattened and environmental issues are at the fore. State’s supported the supply of energy, but it was up to consumers how they wanted to use it.
“What we missed was the opportunity to say, if we can help people to use energy more efficiently, and we can do that at a lower cost than building more infrastructure, why wouldn’t we do it?” Dunstan says.
It’s taken decades for that message to ring clear – but the transition to clean energy means more people are listening. The alternatives to not taking a smarter approach to how energy is used include expensive augmentations to the supply side, such as batteries, pumped hydro, gas-fired generation and interstate transmission.
In hot water
The controlled-load electric hot water system, introduced more than 70 years ago, is a great example of a simple solution to manage energy demand and household bills. “It is the largest or usually the second-largest load in the home and it is controlled very effectively in terms of insuring it does not drive costs,” says Tupper. “Air-conditioners have not been as well managed.”
In the Ausgrid network, Tupper estimates residential load makes up about 45% of maximum demand, and two-thirds of that, about 30%, is air-conditioners alone.
“Residential demand doubles between an average summer day and a peak simmer day,” he says. “Smart control of those air-conditioners can have a significant impact.”
Charging of electric vehicles will also need to be controlled in a similarly dynamic manner, he says. “We need to be smarter about how we can manage the system on behalf of customers.”
Conflicts of interest have muddied the demand response market so far. A retailer that owns generation assets will be loath to encourage a customer to cut load, for example, and a network that earns a guaranteed rate of return will benefit if its assets are upgraded because of high expected demand.
A rule change being considered by the AEMC will give consumers more choice about how they access demand response, Petkovic says. The wholesale demand response mechanism will create a new category of market participant, a third-party aggregator – neither the retailer nor customer – to supply demand response into the wholesale market.
“When this demand response is bid into the market it will be competing directly with generators in the price-setting process,” she says. “At the moment demand response is invisible – it responds to the price but it is not involved in setting the price. [With the rule change] … it can potentially put a lot more downward pressure on prices.”
In the US in 2016, Petkovic saw demand response around 4% of capacity result in price falls around 33%. “When you see numbers like that you can see why retailers have not been too thrilled about demand response being opened up.”