Energy Revenues

Forecasting revenues in the NEM

The dispatch model generates a forecast of battery revenues by simulating an optimal dispatch strategy against our forecast prices for both the energy market and the Frequency Control Ancillary Services (FCAS) markets in the NEM.

The model optimises the battery's charge and discharge schedule across all available revenue streams to maximise profitability, while respecting the battery's physical and operational constraints. The resulting revenue forecast is broken down into energy and FCAS components.

Two-stage approach to modelling battery dispatch

The NEM operates as a real-time market with prices and dispatch instructions updated every five minutes. To realistically model how a battery asset would be traded in this environment, our dispatch model employs a two-stage optimisation process that mirrors the strategy of an energy trader.

1. Initial dispatch plan: First, the model creates a comprehensive dispatch plan for the entire day. This is based on day-ahead pre-dispatch price forecasts, establishing an initial strategy for energy arbitrage. This stage determines a baseline charge and discharge schedule to capture predictable price spreads.

2. Real-time re-optimisation: Trading in the NEM is a continuous process. To capture the more volatile, and often more lucrative, price movements that occur closer to real-time, the model continuously refines its initial plan. Using a rolling window approach, it re-optimises the battery's schedule as updated, more accurate short-term price forecasts become available. This ensures the model can react to near-term market dynamics, adjusting its dispatch to seise opportunities missed by the initial, longer-term plan.

Co-optimising energy and FCAS

A key feature of the NEM is the ability for generators and batteries to participate in ten distinct Frequency Control Ancillary Services (FCAS) markets simultaneously with the energy market.

Our dispatch model fully co-optimises across these markets. In each five-minute interval, it determines the most profitable allocation of the battery's capacity between charging/discharging for energy arbitrage and providing any of the FCAS products. This integrated approach ensures that the revenue forecast accurately reflects the complex trade-offs an operator must make between selling energy and providing valuable grid stability services.

FCAS Revenues

FCAS market saturation

TLDR; FCAS markets are highly competitive. We account for this in the model by restricting the bidding capacity of each battery.

Frequency Control Ancillary Services (FCAS) contracts will get harder to come by. As battery deployment accelerates, saturation in Frequency Control Ancillary Services (FCAS) markets will increase. Consequently, individual batteries will likely secure a smaller share of FCAS contracts and engage more frequently in the wholesale energy market.

We quantify this effect using projected battery buildout and FCAS requirements:

% of battery fleet accepted in FCAS market = Total service requirement / Total battery buildout

We then calculate the average MW that an individual battery can bid for each FCAS market by multiplying this acceptance percentage by the battery's maximum power (MW). For example, if 50% of the fleet is accepted for Raise 6-second services, our model would permit a 10 MW battery to bid a maximum of 5 MW for that service.

The NEM has distinct requirements across its various FCAS markets (e.g., contingency services like Raise/Lower 6-second, and regulation services like Raise/Lower Regulation). Therefore, saturation levels may differ between services. Our dispatch model accommodates this by allowing different bidding limits for each FCAS product, reflecting their unique market dynamics (see an example dispatch here).

Modelling realistic FCAS activation and battery throughput

Winning a contract in an FCAS market does not mean a battery will be dispatched at its full bid capacity for the entire duration of the service. Regulation services, in particular, involve continuous, small adjustments to charge and discharge state, rather than sustained full-power activation. To capture this reality, our dispatch model incorporates an Expected Throughput Factor for each ancillary service.

This factor is a percentage representing the average activation level we anticipate for a service based on historical data. It allows the model to distinguish between the full capacity offered to the market and the actual energy that is likely to be cycled through the battery.

The expected energy throughput is calculated as follows:

Expected Energy Cycled (MWh) = Bid Capacity (MW) * Expected Throughput Factor (%) * Duration (h)

For instance, if a 10 MW battery provides 'Raise Regulation' services for one hour with an expected throughput of 10%, the model will account for only 1 MWh of energy being discharged. This prevents the model from incorrectly reserving the battery's full 10 MWh of energy capacity for that hour, enabling a more realistic and economically optimal dispatch by allowing the remaining capacity to be bid into the wholesale market or other services.