Dispatch Model

Here, we go into the details of our battery dispatch model, specifically tailored for the NEM.

Our core strategy uses mixed-integer linear programming (MILP). This allows the model to maximise battery revenues by finding the optimal times to charge and discharge, considering both the wholesale energy market and ancillary service opportunities.

The model is designed as a forward-looking "look-ahead" optimisation that runs for each site scenario (e.g., a 2-hour battery, performing one cycle per day, with a specific degradation level).

We optimise revenue stacking across wholesale energy and FCAS

The NEM is an "energy-only" market where all electricity is traded on a real-time spot market with 5-minute settlement. A key feature of the NEM is the ability for batteries to earn revenue from multiple streams simultaneously. Our model is built to co-optimise these streams:

1. Wholesale energy arbitrage: Charging the battery when the spot price is low and discharging when the spot price is high.
2. Frequency Control Ancillary Services (FCAS): Providing services to help AEMO maintain grid stability. The model bids capacity into the ten distinct FCAS markets (contingency and regulation services, both raise and lower).

The model uses price forecasts for both energy and all FCAS markets to determine the most profitable combination of activities in each 5-minute interval.

Our modelling approach

The dispatch model maximises total revenue across all streams, while adhering to constraints that reflect the physical limitations of the battery and the rules of the NEM. Key constraints include:

• Physical Limits:
  • State of charge (SoC) boundaries
  • Maximum charge/discharge power
  • Round-trip efficiency
  • Degradation costs (SRMC)
• FCAS Enablement: The model ensures that if the battery is providing FCAS, it maintains sufficient headroom and footroom in its SoC. This guarantees it can deliver the contracted service if called upon by AEMO.