​The July 2026 version of Modo Energy's battery energy storage revenue forecast for Australia's NEM is now live.

This is the biggest structural upgrade to the forecast since launch. The model now solves the NEM as a nodal network of over 200 nodes, rather than nine zones, so prices and battery revenues reflect where an asset actually connects.

Modo Energy forecast subscribers: create your own asset-specific forecast now with the latest update.

View the changelog for a breakdown of everything that has changed in this release. To find out more about the forecast, check out our methodology or book in a demo.

Key updates:

• Nodal price forecasting: the model now clears over 200 network nodes, capturing intra-state congestion and locational differences in battery revenues for the first time.
• Inputs rebuilt on AEMO's 2026 ISP: near-term build costs rise, but projected large-scale solar capex falls 26% by 2035-36.
• Gas futures down approximately $1/GJ in the near term.
• Data-centre demand now follows Modo Energy's own forecast, growing from 5.5 TWh in 2026 to 36.8 TWh by 2050.
• Household battery capacity is raised 59% by 2030, to 28.7 GWh, reflecting Modo Energy's updated consumer energy resources forecast.

Modelling changes

The forecast is now nodal

Previous releases modelled the NEM as nine zones. The July 2026 release solves a transmission network with over 200 nodes, with prices settled at each region's reference node: Sydney West, South Pine, Thomastown, Torrens Island, and George Town.

Location now carries a price. The nodal model produces forecast marginal loss factors (MLFs) for every connection point, and these scale every dollar of market revenue an asset earns. In 2026, forecast solar MLFs span 0.82 at the weakest node to 1.05 at the strongest, and by 2040 a tenth of nodes sit below 0.81 as renewable buildout deepens local losses. Asset-specific forecasts in the terminal now reflect the specific node at the connection point.

Input changes

Inputs rebuilt to incorporate the 2026 ISP

The forecast includes updates from AEMO's 2026 Forecasting Assumptions Update and the technology costs underpinning it, from CSIRO's 2025-26 GenCost. Near-term build costs rise, with gas peaker capex for 2026 around 32% higher than assumed in April. Further out, the picture flips: projected costs decline more steeply than in the April assumptions, leaving large-scale solar capex 26% lower by 2035-36 and small gas peakers 27% lower. BESS capex assumptions continue to be set by Modo Energy's own market intelligence.

Gas down since the April release

Near-term Wallumbilla gas futures have fallen around $1/GJ since the April release: 2027 averages $11.35/GJ, down from $12.01/GJ.

Data centres: now Modo Energy's own forecast

Data centre load is now based on Modo Energy's own forecast, replacing the AEMO assumption used previously. It remains a dedicated, flat 24/7 demand block concentrated in New South Wales and Victoria, reflecting the largely inflexible nature of Australian data centre load, cloud computing and AI inference workloads that run continuously rather than shifting with price. It grows from 5.5 TWh in 2026 to 14.2 TWh by 2030. Then, to 36.8 TWh by 2050, around 15% of today's NEM consumption.

Flat load is valuable for batteries late in the horizon. It deepens overnight demand while solar continues to depress midday prices, supporting spreads in the 2040s.

Household batteries revised up again

Reflecting the continued momentum in home battery installations and sustained longer durations, Modo Energy's consumer energy resources forecast now reaches 28.7 GWh of household battery capacity by 2030. This is up 59% from 18.1 GWh in the April release, and reaches 47.2 GWh by 2050. Coordinated household batteries remain explicit dispatchable storage in the model, with passive units netted off demand.

Output changes

Capacity buildout shifts towards solar

The capacity expansion now favours solar over onshore wind: 77 GW of solar by 2050, up from 69 GW in the April release, against 31 GW of wind, down from 42 GW. Lower GenCost solar capex and the nodal network's better representation of renewable energy zones drive the shift. Utility-scale BESS reaches around 47 GW by 2050, in line with April.

Near-term battery buildout strengthens. The pipeline reaches 22.8 GW by 2030, around 3 GW more than the April outlook, on continued CIS and LTESA progression.

Prices and spreads: lower to the mid-2030s, higher beyond

Near-term wholesale prices are lower than the April release. Cheaper gas, more household batteries, and a stronger BESS pipeline all compress prices and spreads through the early 2030s. Trading spreads in 2030 are roughly 40% below the April outlook.

From the late 2030s, the picture reverses. A solar-heavier buildout, growing data-centre load, and a higher market price cap lift both prices and spreads above the April trajectory through the 2040s and into the 2050s.

What this means for forecast users

Location matters now more than ever. Two batteries in the same state can carry different revenue outlooks, and site selection becomes something the forecast can quantify rather than a qualitative risk.

The revenue story is a trade-off in timing. Merchant revenues to the mid-2030s sit below our April outlook, while the long-run case strengthens on data-centre load and a solar-heavy mix, conditions that also favour co-locating storage with solar.