Dispatch Lab · runs in your browser

Solve a day, your way

Choose a market day between 1 Jan 2018 and 31 Dec 2025, tune the battery, and an exact optimiser (the same model as the brief's linear program, checked against HiGHS) finds the schedule that earns the most. Four no-foresight policies run beside it for comparison.

01 · Market day

Pick a day

VIC1 · market day

Fri 17 Jul 2020

min $39.75mean $71.48max $163.47

Perfect foresight

Knows every price of the day in advance. The upper bound (mandatory task).

Solving the linear program…

Explain this scheduleoptional · AI · your own key

Sends only this schedule's numbers (price summary, battery settings, charging and selling blocks, revenues) to Anthropic, from your browser. What is sent · How it is evaluated

02 · Battery

Technical settings

Power300 MW

Energy capacity580 MWh

1.93 h of storage at full power

Charge efficiency90%

Discharge efficiency90%

Round trip 81%

Marginal loss factor0.991

Degradation cost$0/MWh

Charged per MWh delivered to the grid. The brief assumes $0.

Cycle limit

Unlimited1 / day2 / day3 / day

Caps the energy discharged in the day at cycles × capacity. Deliverable 1 asks for a single 580 MWh discharge on 17 Jul 2020.

Rolling persistence plans

To midnight24 hours

Each day here is scored on its own, so by default rolling persistence plans only to midnight. The 24-hour horizon is the original policy from the Benchmark, which carries charge overnight.