top of page

Future sources of grid demand

Do you find the discussion on electricity dry or revolving around the same central themes?  Well in this piece we take a step back from looking at the ‘how’ we generate electricity and go to the ‘why’.  By why I mean – what will our society look like in the future, what does ‘electrication’ of our economy mean, will it change our relative competitive or comparative advantages?   This question, when thought through, goes far beyond whether you’ll have a shower every morning in 2046 (I hope you do), but rather what device will heat and deliver the water and how will it be powered – maybe ‘ice baths’ will move beyond a passing trend and this is irrelevant (but you still need to cool the water!).   Seriously though, as an investor in this space, I’m far more interested in how the energy will be used in the future – this is the really transformative element of the energy transition for our economy.  This article starts an investigation into this theme by looking at AEMO’s Integrated System Plan modelling, however there’s far more to this topic than can be explored in a single article, so I’m sure there will be more on this topic in the future.

First things first, we need to start with the basics and the status quo.  Grid demand can be broken down into operational demand and underlying demand. Operational demand is the demand for electricity produced by generators sold through the wholesale energy market and is the “official” grid level demand. Rooftop solar output is treated as an offset against grid demand because it replaces electricity that would otherwise by supplied by large generators. Underlying demand is gross demand before accounting for behind the meter solar. It is estimated by adding back the estimated generation from rooftop solar (and home batteries).

Underlying demand has largely been stagnant over the last 20 years, with the increase in population and GDP growth offset by the increase in energy efficiency of household appliances (eg we moved from Plasma TVs to LED TVs) and improved building standards. More recently over the last decade, the increasing penetration of rooftop solar has seen a significant contraction in operational demand.

In the last calendar quarter of 2023, year on year operational demand, that is demand net of rooftop solar, increased for the first time since 2015! The increase was by 315 MW or 1.6% over the previous calendar quarter. This was driven by an increase in underlying (gross) demand of 820 MW or 3.7% and despite rooftop solar increasing 505 MW or 17%.

Is this a turning point in both underlying and operational demand from the grid?

It is difficult to breakdown where demand is coming from, however it does appear that a large part of the increase last quarter could be attributed to the warmer weather in Queensland and New South Wales. In the future, warmer and more volatile weather caused by climate change will be a driver of energy demand in of itself.

Total grid demand in the National Energy Market was 188 TWh in 2023 and is the baseline we will be comparing to. AEMO is forecasting underlying demand to double by the 2040s in their central case with a large part of the increase in demand serviced by rooftop solar.

Electric vehicles

Electric vehicles made up 7.3% of all 2023 new car sales in Australia. Globally one in five new car sales were electric in 2023 and in China it was one in three. The Australian Bureau of Statistics conducts five yearly surveys on vehicle usage. The last survey was conducted in 2020 with some of the data presented below.

Type of vehicle

km travelled (millions)

Megalitres of fuel

Litres per 100 km

% of total fuel

Passenger vehicles










Light commercial





Rigid trucks





Articulated trucks





Non-freight truck
















Electrifying the entire vehicle fleet would present an opportunity to displace 33,000 megalitres of fossil fuel.  To provide an estimate of how much additional grid demand that would result from this adoption, we have created a hypothetical scenario where all passenger and light commercial vehicles are displaced by the most popular electric vehicle in Australia, the Tesla Model Y. Tesla does not officially disclose battery capacity but quick googling reveals that the long range Model Y has a battery capacity of 81 kWh and a range of 533 km. This would imply that the model Y achieves about 6.5km/kWh of battery charge or 15 kWh per 100 km. To cover the 162,983 million kilometres this would require an additional 25 TWh of grid demand. Should the range be overstated by 20% (that is, real world efficiency doesn’t quite match the theory of standardised mileage tests) the required energy usage to electrify the entire passenger fleet would be 30 TWh or 16% of current grid demand.

The electricity usage assumptions would be higher for commercial vans/trucks and buses. Assuming all the other categories of vehicles (excluding motorcycles) would require twice the energy per kilometre  of a passenger electric vehicle this would be another 30 TWh of grid demand for vans, trucks and buses which would take total grid demand to electrify vehicle transport to around 30% of current grid demand.

Thus, while electric vehicles wont “blow up the grid” as some commentators suggest, they will result in a meaningful boost to electricity usage over a 10-20 year.

Electrification of heating and cooling

Heat pumps and induction cooktops are significantly more efficient that their gas counterparts. The coefficient of performance (COP) is a measure of the ratio between the energy output to the energy input of a heating or cooling unit. For a heat pump the COP is usually around 4. There are some models such as the Sanden heat pump system that can go up to 6. A gas boiler has a coefficient of performance of 0.9-1.0. For our assumptions we will assume a COP of 1 for a gas boiler. On average electrification of household appliances will require 4 times less energy compared to their gas equivalent. The efficiency of an induction cooktop has a similar COP to a heat pump but requires significantly less cooking time.

The main drawback of a heat pump or induction cooktop is that the upfront cost which is significantly higher than the equivalent gas system. However, the operating costs of a heat pump are next to nothing with a higher COP and “free” electricity from a rooftop solar system. The counterfactual operating cost of a gas system is high.  It is currently a no brainer to electrify your heating/cooling/cooking requirements if building a new home or if you plan to live in your house for an extended period of time.

Based on the 2023 AEMO Gas Statement of Opportunities current household gas consumption is 175 PJ per annum. If all of gas demand at the residential level were to be electrified this would be equivalent to 48.6 TWh of electricity (for the non-engineers, this is based on a conversion factor of 3.6 MJ to MWh). With the higher COP of heat pumps and induction cooktops the load would be reduced to 12 TWh (assuming a COP of 4). This is 6-7% of current operational demand.

Utility scale storage

Storage is key to achieving a grid with higher levels of renewable energy supply. It allows for the mismatch between intermittent renewable energy and consumer load to be matched. For coal plants to close, it will require building a significant amount of dispatchable capacity. AEMO is forecasting that 33 GW / 514 GWh of storage capacity will be required by 2035. Short duration storage of less than 4 hours is forecast to be 10 GW/20-40GWh. This will be served by batteries which have a round trip efficiency of about 85%. The round trip efficiency losses mean a battery is a net user of electricity from the grid. Assuming the AEMO short duration forecast is met, there would be an annual draw from the grid of 1-2 TWh (assuming one cycle per day) from round trip efficiency losses or 0.5-1.0% of existing operational demand of the grid.

Rooftop solar

As of the end of February 2024 there is 22.6 GW of rooftop solar installed across 3.745 million sites. Based on the last census, there are 9.275 million households in Australia of which 70% are detached dwellings or 6.5 million households. It is likely that 80% of households will eventually have solar, which means there is potentially another 1.3 million households to install solar systems. If the trend of larger system sizes continues there is potentially another 15-20 GW of potential rooftop solar or 25-35 TWh per annum of additional solar generation (negative demand).

Data centres

Over the last decade data storage has migrated from onsite company owned facilities and to remote locations on the cloud. This has led to the steady growth of data centre demand. More recently this has accelerated with streaming, gaming, and AI/machine learning requirements becoming more relevant to a variety of different facets in our society. The main operators in Australia are AirTrunk, NextDC, Equinix, Canberra Data Centres, and Macquarie Telecom. Data centres are essentially a property play with landlords selling a margin over electricity to operate and cool the data centre. Data centres are capital-intensive, volume-based businesses and require sophisticated cooling, security, and energy backup systems for clients to access data any time.

Currently, there are 307 data centres in Australia. According to IBIS World, the revenue has grown by 5.6% to $5.2 billion over the last 5 years and is forecast to expend by 8% in the next 5 years to $7.7 billion. There is around 900 MW of data centre capacity with 1.5 GW of potential data centre capacity in the pipeline. Utilisation rates vary among data centre providers but based on NextDC public disclosures their centres are running at 80-90% utilisation rates. Assuming 80% utilisation of the existing capacity, and making the very simplified assumption that the servers are running 24/7, this would equate to an electrical load of 17 TWh per annum on existing data centres increasing to 30 TWh in the next 5-10 years as capacity increases to 2.5+ GW. This could add 8-10% of additional grid demand.


Hydrogen is a transportable and storable alternative fuel source to fossil fuels, which produces no carbon emissions when used. It is a platform to allow the global trade of clean energy and a path forward for emissions reduction in hard-to-abate sectors such as steelmaking, mining, chemicals, cement, aviation, shipping, and heavy road transport.

If/when Australia’s Hydrogen industry gains traction, Green Hydrogen is likely to be a significant portion of energy demand. Therefore, the growth of the Hydrogen industry may become a significant driver of increased NEM load. Nevertheless, there are significant uncertainties surrounding Hydrogen’s future. The AEMO recognises this, and have created three Hydrogen forecasts within the Integrated System Plan’s ‘Green Energy Exports’, ‘Step Change’ and ‘Progressive Change’ scenarios:

2040 Hydrogen related electricity consumption (TWh)


Green Energy Exports (‘Hydrogen Superpower’)

Step Change

Progressive Change













In a ‘progressive change’ scenario, at 15TWh, the AEMO are predicting hydrogen production will consume the same amount of electricity as 3 million average Australian households in a year (or just over that of the State of Victoria). This is not insignificant, let alone a ‘Step Change’ or ‘Green Energy Exports’ scenario demanding a load of 2.3x or 15.5x the ‘Progressive Change’ scenario respectively.


It is conceivable we could have an additional 50-100 TWh of underlying demand from electric vehicles and datacentres over the next 20 years and increase grid demand to 250-300 TWh. Getting to 400 TWh would seem like a stretch but plausible if there was a hydrogen industry in Australia. The rooftop solar boom will continue and operational demand will grow at a slower rate than underlying demand.

Of course, this analysis really only assumes one new industry – Hydrogen.  But if we’re really to become an ‘energy superpower’ it means that we must have the lowest cost energy (comparative advantage) and so this begs the question what else could we do – which we’ll explore in future articles.


bottom of page