Data centres get a forecast in state’s new transmission plan. Electric heating gets a footnote

VicGrid, the Victorian Government body now responsible for planning the state’s transmission network, has just consulted on the guidelines that will shape the 2027 Victorian Transmission Plan. Buried in the demand methodology is a curious asymmetry.

VicGrid starts with AEMO’s national demand forecasts for Victoria, then supplements them with Victorian-specific analysis for exactly one new source of demand: data centres.

The guidelines say so themselves. A footnote to the demand methodology explains that “VicGrid analysis” refers primarily to the development of Victorian-specific data centre demand forecasts.

That is a sensible thing to do. Data centre demand is uncertain, lumpy, and growing, and Victoria has its own policy for it, the Sustainable Data Centre Action Plan. A bespoke forecast is exactly what good planning looks like.

But Victoria has another new source of electricity demand, and it gets no equivalent treatment. The switch from gas to electric heating in homes and businesses, the load the Gas Substitution Roadmap is expressly designed to create, is left inside AEMO’s national demand traces.

Both policies sit side by side on VicGrid’s own list of Victorian government policy settings. One gets a bespoke Victorian forecast. The other does not.

That gap matters more in Victoria than it would anywhere else, for a simple reason: no other state in the National Electricity Market relies on gas for heating the way Victoria does.

Building electrification is the demand driver most specific to Victoria, and the one a national forecast, weighted toward warmer and summer-peaking regions, is least likely to capture well.

The winter evening is the moment that matters

For decades, the heaviest moment on Victoria’s grid has been a hot summer afternoon, when air conditioners run hardest. Rooftop solar has since reshaped that picture, meeting much of the midday load and trimming the peak.

Electrified heating introduces a different peak. A home that has swapped its gas ducted heater for a heat pump draws its heaviest load on a cold July evening, after dark, when the family is home, the heating is turned up, and the rooftop panels are producing nothing.

Figure 1. A winter day for a heavily electrified Victorian home. The household’s heaviest load falls in the evening, after rooftop solar has stopped generating.

This is no longer just a forecast. On 15 July 2024, Victoria set a new maximum winter electricity demand of 8,612 MW at 6pm, breaking a 17-year record, at the same time as the gas network was delivering its own winter peak.

The summer afternoon is still the system maximum, at 9,851 MW, but the winter evening is climbing toward it, and it is the peak that electrification will keep pushing up.

The planning consequence is plain. A network has to be built for the highest demand it must serve at a single moment, measured in megawatts.

Rooftop solar helps at the summer afternoon peak, because the sun is still up. It contributes nothing at the winter evening peak, because the sun has gone. As heating electrifies, the binding peak for parts of the Victorian system is likely to move from the summer afternoon to the winter evening.

Figure 2. Where the binding peak sits once heating is electrified. Rooftop solar trims the summer afternoon peak but is absent at the winter evening peak, which becomes the demand that sizes the network.

How big is the load?

The VEFN submission to VicGrid put a scale on it. Take Victoria’s daily residential gas consumption, isolate the weather-driven heating component, and convert it into the electricity a heat pump would need to deliver the same warmth.

Depending on the assumptions you make about heat pump performance and home insulation, electrifying today’s gas heating would add between about a fifth and a third to Victorian electricity demand on the coldest winter days, and close to nothing in summer.

Figure 3. The electricity equivalent of Victoria’s gas heating load, at heat-pump efficiency, as a share of current Victorian electricity demand, 2018 to 2025. It reaches between about a fifth and a third on the coldest winter days, and close to zero in summer. Source: VEFN analysis of AEMO demand and effective-degree-day data.

To be clear about what that number is and is not. It is not an argument against electrification.

Heat pumps are several times more efficient than gas heaters, households save money by switching, and the switch is already well under way: Victoria’s own electricity distributors are forecasting more than 2,600 GWh a year of new electrified heating load by 2031.

The question is not whether this load arrives. It is whether the network that must carry it, on a cold evening with no solar, is planned for it, in the right places and in time.

Nor is the energy share the whole story. The quantity that actually sizes a network is the coincident peak in megawatts on the coldest winter evening, when heating load concentrates, occupancy is high, and solar is absent.

Batteries help, and AEMO’s scenarios assume a lot of them in Victoria, but the case that sizes the network is a run of overcast, cold days, when there is little daytime solar to store between evenings. Storage softens the winter evening peak; on the days that matter, it does not remove it.

The national forecast has the same blind spot

AEMO’s final 2026 Integrated System Plan, published on 25 June, shows the same asymmetry at the national level. Data centres get their own demand category, a range across all three scenarios, and a dedicated Higher Demand sensitivity.

Building electrification gets none of this: it is folded into the aggregate residential and business forecasts, and reported as a net annual figure, with grid-supplied household consumption forecast to fall 44 per cent by 2050 as rooftop solar grows.

AEMO itself cautions that this net figure conceals much larger flows the network must still carry. A forecast that reports Victoria’s most peak-concentrated new load as a falling annual average is not one the state can rely on unsupplemented for a quantity defined by a single winter evening.

A modest fix

None of this requires VicGrid to change direction. The VEFN submission asked for two things. First, treat Victorian building electrification as a distinct demand component, supplemented with Victorian-specific data, in the same way the guidelines already treat data centre demand.

Second, confirm in the methodology how the winter evening coincident peak is preserved, so that the cold-evening heating peak is not understated by demand traces that net off rooftop solar which is present at midday and absent after dark.

The first brings one Victorian policy-driven load into line with the treatment already given to another. The second is largely a request for transparency: if the demand traces already preserve the winter evening peak, saying so would give everyone confidence that the moment which will increasingly size Victoria’s network is being planned for.

Victoria is going to electrify its heating. That is settled by economics as much as by policy. What is not yet settled is whether the electricity network will be ready to take over from gas, in the right places, in the right order, and in time. Getting the forecast right is where that starts.

John Godfrey is a retired chemical engineer and the facilitator of the Victorian Energy Future Network (vefn.au), an independent network of energy professionals working on the planning and regulation of Victoria’s gas-to-electricity transition. The VEFN submission to the Draft 2026 Victorian Transmission Plan Guidelines is available at vefn.au.