Key takeaway: The dispatch model uses wholesale prices produced by the Capacity Expansion Model, which already reflect a specific battery fleet. A user's battery displaces capacity from that fleet, keeping prices consistent — unless the battery is larger than what the CEM built for that region, at which point revenues become overstated.
How prices reach the dispatch model
The forecast is produced in two stages. The Capacity Expansion Model (CEM) determines the optimal generation and storage buildout across the NEM, solving for the capacity mix that meets demand at lowest total system cost. As part of this optimisation, wholesale prices emerge for every node and time interval — these prices already reflect the arbitrage behaviour of the CEM’s battery fleet, which flattens price spreads as storage is added.
These prices are then passed to the dispatch model as fixed inputs. The dispatch model optimises a specific battery’s charge and discharge against these prices to produce a revenue forecast.
Displacement: why most battery sizes are consistent
When a user configures a battery in the dispatch model, that battery displaces an equivalent amount of storage from the CEM’s buildout. If the CEM determined that 400 MW of BESS is optimal for a region, and a user models a 200 MW battery, then 200 MW is attributed to the user’s asset and the remaining 200 MW stays in the CEM fleet. The total storage in the system is unchanged, so the wholesale prices remain consistent with the capacity mix.
This displacement means the dispatch model correctly represents the economics of a battery that takes the place of what the market would have built. At typical battery sizes, this assumption holds and forecast results are reliable.
Outscaling the CEM
The assumption breaks down when the user’s battery exceeds the total BESS capacity the CEM built for that region. If the CEM built 300 MW of storage in South Australia and the user models a 500 MW battery, the extra 200 MW represents storage beyond what the CEM priced in. The wholesale prices the dispatch model uses were computed assuming 300 MW of arbitrage flattening — not 500 MW. The additional storage would compress spreads further, but this is not reflected in the fixed input prices.
The result is that revenues are overstated. The dispatch model sees price spreads that would not exist if the full 500 MW were present in the market. The degree of overstatement grows with the gap between the modelled battery and the CEM fleet size.
Specific effects of this inconsistency include:
- Spread compression not captured: The model dispatches against spreads that the battery’s own activity would reduce.
- Concentrated dispatch timing: The model places all charging in the cheapest intervals and all discharging in the most expensive. A battery large enough to move prices would need to spread activity across more intervals to avoid shifting prices against itself.
- FCAS revenue inflation: NEM ancillary services markets are relatively thin. A battery exceeding the CEM fleet could dominate a service category, but the resulting price impact is not reflected.
- Calibration limits: The dispatch model calibration is derived from batteries in the 50–200 MW range. Results for larger batteries fall outside the validated range.
Soft limits by region
When the configured rated power exceeds the soft limit for a region, a confirmation prompt is displayed. The forecast can still be submitted — the limit is advisory, not a hard block.
| Region | Soft limit (MW) |
|---|---|
| NSW (Central, North, South) | 500 |
| QLD (Central, North, South) | 500 |
| VIC | 300 |
| SA | 300 |
| TAS | 200 |
These thresholds are a heuristic approximation of the CEM’s optimal BESS buildout per region. Larger regions like NSW and QLD (6–12 GW operational demand) support more storage before the CEM fleet is exceeded. Smaller regions like Tasmania (0.8–1.5 GW) reach that threshold sooner.
A hard cap of 1,000 MW applies across all regions regardless of confirmation.
Interpreting results above the soft limit
Forecasts submitted above the soft limit will still run and produce results. However, revenues should be treated as an upper bound rather than a central estimate, as the underlying prices do not account for the market impact of the additional storage. The gap between forecast and reality widens as the battery size increases beyond the CEM fleet for that region. For large battery assessments where price feedback is important, contact the Modo Energy team to discuss a CEM-integrated approach.