Calibration to index

The Modo forecast is calibrated to real-world numbers using the Modo Energy benchmark for the Australian NEM.

To capture the realities of operating a battery in the NEM, we use historical data, both from our forecast model and the real world, and apply a calibration to the forecast battery revenues going forward.

Modo Energy's BESS Index tracks each battery energy storage asset above 5 MW across the NEM. It calculates their earnings during each five-minute dispatch interval across the wholesale energy and Frequency Control Ancillary Services (FCAS) markets. More information on accessing this data can be found on our Indices and Benchmarks pages.

This gives us a robust, real-world dataset to calibrate our forward-looking numbers, taking our revenue forecast from a fundamentals-backed model to one cemented in reality.

We compare fleet revenues to the model backtest

In 2025, the NEM BESS fleet consists of batteries with a wide range of operational dates, technologies, and configurations. The average asset duration varies significantly by region:

• New South Wales: ~1.8 hours
• Queensland: ~1.9 hours
• South Australia: ~1.4 hours
• Victoria: ~1.7 hours

Newer assets benefit from enhanced protection against degradation, higher cycling allowances, and better thermal management. This means they can have more advanced optimisation across markets and greater availability compared to older assets.

To keep our revenue calibration relevant, we compare the total historical revenue of the fleet in each region, as captured by the ME BESS AUS NEM Index, to our dispatch model's backtest. Backtest revenues are calculated by running our dispatch model using historical prices for energy and all eight FCAS markets.

By comparing the two for a recent period, we calculate a median 'capture rate' for each region. These regional values are then averaged to produce a single, market-wide calibration factor.

Calibration that captures the realities of asset operation

There are several reasons why our fundamentals-backed model will have higher revenues than actually observed in the market, which is why this calibration is necessary.

• Battery availability is less than 100%. This could be due to distribution or transmission network downtime, transformer outages, or cell degradation.
• Degradation from nameplate MWh capacity occurs as the battery cycles.
• Bidding and dispatch complexity. The model assumes a perfect bidding strategy to maximise revenue from the centrally-dispatched NEM. Real-world bidding strategies are far more complex, balancing risk and reacting to uncertain forecasts. Participation and success in the eight different FCAS markets can vary significantly between assets and is difficult to perfectly replicate.
• The dispatch model assumes perfect foresight. This is the most significant factor. Our model optimises an asset's dispatch over a 24-hour horizon with perfect knowledge of energy and FCAS prices. In reality, operators work with imperfect forecasts and have a much shorter, more uncertain view of future prices. This perfect foresight allows the model to capture ideal arbitrage opportunities that are not fully achievable in the real world. The impact of perfect 24-hour foresight versus real-world forecasting is the largest driver of the difference between model outputs and actual revenues, accounting for over 20% of the variance.

By comparing the output of our revenue forecast to reality, we can capture the impact of all these effects (and more), making the revenue forecast more realistic.

A single factor brings the forecast in line with reality

Our central case uses a single calibration factor derived from the process above. By comparing the backtest to the index, we calculate a robust calibration factor that encapsulates all the real-world effects described, such as availability, degradation, bidding complexity, and imperfect foresight.

This calibration factor is then applied to all future revenue streams in our forecast. This ensures our central case is grounded in and aligned with historical, real-world asset performance across the NEM.