How the model optimises battery operation to maximise revenues.
How does the dispatch model work?
The dispatch model simulates how a battery would operate in each market to maximise total revenue. It decides when to charge, discharge, or provide grid services based on price forecasts.
Key inputs
- Power capacity (MW): Maximum rate of charge/discharge
- Energy capacity (MWh): AC useable energy capacity at point of connection — the energy the battery can actually export to the grid, measured at the meter (determines duration)
- Round-trip efficiency: Measured at point of connection, including all losses in the round-trip (~88% default). Applied during the charging phase to calculate correct import volumes
- Cycling limits: Maximum cycles per day to protect warranty
What the model optimises
- Wholesale arbitrage: Buy low, sell high across day-ahead and intraday markets
- Ancillary services: Reserve capacity for frequency response and other grid services
- Co-located renewables: When paired with solar/wind, optimise combined export
Key assumptions
- Price foresight within each window: The model optimises against forecast prices over the optimisation window for that battery duration. How much foresight the optimiser is given, and how results are adjusted to reflect real-world uncertainty, varies by region — see Price foresight and calibration vary by region
- Physical constraints: Respects all battery physics including ramp rates and state of charge limits
- No simultaneous charge/discharge: The battery cannot charge and discharge at the same time
Price foresight and calibration vary by region
A dispatch model with full knowledge of future prices earns more than a real trader can, so every region accounts for real-world uncertainty in one of two ways, often combined:
- Imperfect foresight in the optimisation: The model schedules against smoothed, blended or rolling-window prices, so it sees the shape of each day but not the exact level of every interval
- Revenue calibration: The model optimises with full price knowledge, and modelled revenues are then scaled against observed battery performance, such as a Modo Energy BESS Index
The table summarises each region’s approach. The regional dispatch model pages are the source of truth for the details.
| Region | Price foresight | Calibration |
|---|---|---|
| Great Britain | Perfect in day-ahead; 2-hour rolling imperfect foresight intraday | Scaled to the Modo Energy GB BESS Index |
| Germany | Perfect in day-ahead and intraday auction; imperfect in real time | No revenue scaling; realism comes from the imperfect real-time foresight |
| Italy | Known day-ahead, intraday auction and balancing prices; imperfect continuous intraday | Day-ahead revenue scaled per scenario |
| Poland | Known across the delivery day in day-ahead; 2-hour rolling imperfect foresight intraday | All revenues scaled per scenario |
| ERCOT | Imperfect foresight built into the optimisation | Tuned against the Modo Energy ERCOT BESS Index |
| WECC (CAISO) | Full price knowledge within the optimisation | Scaled to the Modo Energy CAISO BESS Index |
| SPP | Schedules against a three-hour smoothed price, settled at actual prices | Flat 80% factor on battery revenues |
| Eastern Interconnection | Perfect foresight over the day-ahead horizon | Flat 80% factor on energy arbitrage |
| NEM | Accurate price forecasts ahead of each day-ahead and real-time step | Two-tier calibration to the Modo Energy NEM BESS Index |
Optimisation window length depends on battery duration
The dispatch model does not use a single rolling horizon for every battery. Longer-duration batteries need a longer look-ahead so the optimiser can plan full charge and discharge cycles across multiple days. Site forecasts support durations up to 12 hours.
| Battery duration | Days per window | Hours |
|---|---|---|
| Up to 4 hours | 1 | 24 |
| Above 4 hours and up to 6 hours | 2 | 48 |
| Above 6 hours | 3 | 72 |
For example, a 2-hour battery is optimised in 1-day windows, while an 8-hour or 12-hour battery uses a 3-day window. The same rules apply across European market regions that use the global dispatch model.