Skip to content

Solve

With perfect hindsight, when should a 100 MW / 200 MWh battery have charged and discharged on a given day to earn the most? Pick a region and a day, then solve. To see the infrastructure that runs the solve, visit the site architecture page.

Data

Region

South Australia, Sat, 24 Oct 2026: 288 five-minute spot prices, the only data the model needs besides the battery's limits. Times are market time (AEST).

Loading prices…

Formulation

A mixed-integer program: maximise revenue from buying and selling at the spot price, given the battery's limits. The binary utu_t stops it charging and discharging at once, which it would otherwise do on negative prices to burn energy through its losses.

max⁡∑t∈Tpt (dt−ct) Δ\max \sum_{t \in T} p_t \,(d_t - c_t)\,\Delta

subject to

st=st−1+η ctΔ−dtΔη∀t∈T,  s−1=s0ct≤P ut∀t∈Tdt≤P (1−ut)∀t∈Ts287≥s00≤ct,dt≤P,0≤st≤E,ut∈{0,1}∀t∈T\begin{aligned} s_t &= s_{t-1} + \eta\, c_t \Delta - \frac{d_t \Delta}{\eta} && \forall t \in T,\; s_{-1} = s_0 \\ c_t &\le P\, u_t && \forall t \in T \\ d_t &\le P\,(1 - u_t) && \forall t \in T \\ s_{287} &\ge s_0 \\ 0 \le c_t, d_t &\le P,\quad 0 \le s_t \le E,\quad u_t \in \{0, 1\} && \forall t \in T \end{aligned}
SymbolMeaningValue
TTIntervals of the trading day0 … 287 (5-minute)
ptp_tSpot price (RRP) in interval tFrom AEMO, $/MWh
ct, dtc_t,\ d_tCharge and discharge powerMW, chosen by the solver
sts_tEnergy stored at the end of t (state of charge)MWh, chosen by the solver
utu_tMode: 1 allows charging, 0 allows dischargingBinary, chosen by the solver
PPPower limit100 MW
EEEnergy capacity200 MWh
η\etaOne-way efficiency, the square root of round trip√0.88 ≈ 0.938
s0s_0State of charge at the start, and the minimum at the end100 MWh
Δ\DeltaInterval length5/60 h

Assumptions: the battery takes prices as given (it doesn't move them), knows the whole day's prices in advance, trades energy only, and ignores loss factors, ramp limits and degradation. HiGHS stops within 0.1% of the best possible revenue.