4 November 2026 — Alastair McDowell
Demand-responsive water heaters and the case for an incentive scheme: the slides
Demand response and demand flexibility were a huge topic at All Energy conference last week in Melbourne. But very little was said about the role water heaters can play in keeping their electrical loads off the grid’s peak periods.
That’s where my talk came in: how demand-responsive water heaters can help reduce peak demand on the electricity grid.
Peak demand and the cost of energy
Australia’s grid is stretched most in the evening. Rooftop solar covers much of the midday load, so demand on the grid dips at noon and climbs after sunset, when the sun is gone and households are cooking, cooling and charging vehicles.
Wholesale prices follow the same shape. On an average NSW summer day in 2025 the spot price was about $17/MWh in the middle of the day and reached about $207/MWh in the evening. The evening gap is filled by coal, gas, hydro, batteries and imports, and the cost of building and running the plant, batteries and network that serve those few hours is paid for in everyone’s bill.
Emissions
The same pattern drives emissions. At midday a large share of the grid’s supply is wind and solar, and some of it is switched off because there is nobody to use it. In the evening and early morning the supply that replaces solar is more emissions-intensive. 1 kWh used at midday therefore carries lower emissions than the same kWh used at 6pm.
A water heater uses energy in proportion to the hot water drawn, and most hot water is drawn in the morning and evening. A heat pump on a thermostat reheats after those draws, so a large part of its electricity falls in the hours when the grid is dearest and dirtiest. It is very efficient, but it is not well timed.
Water heaters as thermal batteries
A 250 to 315 L storage tank holds 10 to 15 kWh of heat. If the heater is allowed to heat in the middle of the day, it can fill the tank from surplus solar, then coast through the evening peak and the next morning’s showers without switching on. The household gets the same hot water at the same temperature. Only the time of heating changes.
This works for resistance elements and for heat pumps. It does not need a change in how people use hot water, and it uses the tank, the wiring and the poles and wires that are already installed. A water heater can shift about 5 to 6 kWh a day, which is its own load only, so it complements a home battery rather than competing with it. The battery stays free for the rest of the house.
Why today’s incentives do not reward water heaters’ time of use
Federal STCs under the SRES, and the state schemes built on the same method, incentivise total energy a product saves over a year, compared to a standard electric water heater. 1 kWh saved at 2am earns the same as one saved at 6pm, because the annual STC calculation has no time-of-day dimension.
That has a design effect. To maximise certificates means maximising COP: a manufacturer may lower set points and tighten thermostat deadbands to improve the COP of the heat pump’s cycles. While this marginally improves energy efficiency, it also reduces heat stored in the tank, and hot water availability, and limits demand response potential.
Meanwhile, a bigger tank or a fast element, which could be useful features for demand response, adds to modelled energy use under current incentive structures and is therefore disincentivised.
A product that adds demand response hardware can lose certificates and can fail the minimum energy saving, even though the same hardware helps the grid. And because STC modelling must include any default timer and can give it only zero or negative credit, a manufacturer has no reason to set a grid-friendly control strategy as the factory default setting.
What the modelling shows
We simulated 14 water heaters for a full year in TRNSYS: electric tanks, R290 and CO2 heat pumps, in 180 L and 320 L sizes, with and without a 3.6 kW element. Each ran on six control tiers, from a plain thermostat, through a simple timer, to one-way price signals and two-way control that reads the state of the tank.
Simulation inputs: in addition to AS/NZS 4234 load profiles and climate data, the simulation uses AEMO spot prices & total grid demand, as well as state-specific emissions intensity profiles, to help understand how altering water heater timing can impact on energy costs, peak demand and grid emissions.
The results were clear: grid-friendly control strategies can reduce grid emissions by up to 47% and reduce wholesale energy costs by 27 to 44%, all while only consuming 1 to 2% more electricity for solar timers, or 9 to 12% more for 65°C pre-heating, compared with standard thermostat controls.
These come from separate modelling runs. The cost figures are from the 14-heater NSW study. The emissions and electricity figures are from modelling of a typical 270 L R290 heat pump against an hourly emissions profile.
And the controls still deliver the same hot water ensuring the end user’s comfort, as well as maintaining safety - legionella control is never sacrificed. Every simulation result must ensure minimum delivery 45°C and daily or weekly legionella sanitation is passed.
But households only see the cost savings on a tariff that passes wholesale prices through, which is one of the barriers the scheme is meant to address.
The case for an incentive scheme
The benefits of demand response are well evidenced in theory and in early field trials, and uptake is still minimal. Households do not know it can save them money. Few aggregators can control water heaters at scale, and few tariffs pass on five-minute wholesale prices. Almost nothing rewards a household for deferring a network upgrade, apart from overnight controlled load. And manufacturers have little reason to build smarter controllers when the extra cost is not paid back.
Incentives change what the industry builds, and quickly. The Cheaper Home Batteries Program pays about 30% of the price of a home battery through STCs, and installations rose from a few thousand a quarter to more than 500,000 in about 13 months. Water heater installations followed the same pattern wherever a scheme paid for them. Manufacturers and installers respond to what a scheme pays for.
At the moment no scheme pays for when a water heater runs. A demand response certificate (DRC) would, and it can be calculated the same way as an STC: the electricity a product uses in the hours when the grid has surplus (low prices and low demand), over 10 years, compared with the same heater on a thermostat, and counted only if the household is no worse off. The white paper tested one version of this using spot prices, and the exact design is still being refined.
If the scheme is well designed, the benefits reach beyond the households that buy the products. Lower peak demand defers network and generation spending. More midday load means less solar curtailment. Lower wholesale prices in the evening peak could ultimately help lower energy costs for all consumers.
Open design questions
The method is still being refined, and three questions will shape it. Which hours should count: the lowest-price hours, the hours of both low price and low demand, or a fixed daytime window that a simple timer could meet? What should a product be measured against: a standard thermostat, or a reference control for its type? And who funds the certificate, and at what value: a figure tied to the network cost it avoids, as the white paper suggests?
If you work on tariffs, aggregation, controls or scheme design, I would like to hear your view. Get in touch.
The slides
The talk is based on the Demand Response Certificate white paper by Dr Dean Clift and me. Dean works part-time for Rheem Australia, and EnergyAE models water heaters for many manufacturers; no product is promoted. The slides below are from my presentation at All Energy Conference 2026. They show the method as presented on 29 October, and the design is still being refined with industry and regulators.
