Home batteries are the fastest-growing storage fleet in the Australian National Electricity Market, but their energy is unscheduled and settles inside the retailer's book — there is no public 5-minute metered data. NEMPulse therefore models, rather than tracks, what a representative home battery could earn from wholesale energy arbitrage in each NEM region: a 5 kW / 13.5 kWh unit cycling once a day. The result is a benchmark for the value of the underlying price signal, not a forecast of what a particular household will earn.
The forecast tier plans once on AEMO's public D-1 23:30 PREDISPATCH run and is settled at the prices that actually occurred — a public-information benchmark, not an operator target. The perfect-foresight ceiling assumes every 5-minute price is known in advance. For the selected week, month or year, the inner band is a P10–P90 bootstrap of that window's daily values; the outer band is the union of the named scenarios' P10–P90 — a 2 kW export limit, a 10 c/kWh grid-charge adder, and a grid-only no-PV household — so the chart separates year-to-year variability from assumption risk.
Every figure is gross wholesale energy only. It excludes FCAS, network tariffs, retailer and VPP fees, degradation, backup reserve and household load. Midday charging is assumed to come from PV surplus valued at the forgone spot export, so the central case is optimistic for a roof without roughly 15.7 kWh of daily surplus. The no-PV forecast can be negative because it is the same frozen D-1 plan paying 10 c/kWh on every charge, not a rational household that would stop cycling when the spread did not cover the cost.
The fleet view multiplies the per-unit result by the CER's installed small-scale battery kWh for each region (CC BY 4.0, published monthly and revised upward for recent months). It is an upper bound on an upper bound: the aggregate fleet would move prices, and not every install is on a wholesale-exposed or VPP product.
No. Home battery energy is unscheduled and settles inside the retailer's book, so there is no public 5-minute metered data. NEMPulse models a representative unit from AEMO regional spot prices and publishes the result as an estimate, not tracked performance.
NEMPulse publishes a range, not a single number. The perfect-foresight ceiling is the most a representative 5 kW / 13.5 kWh unit could earn if it knew every 5-minute price in advance; the forecast tier plans once on AEMO's public D-1 price forecast and is settled at the prices that actually occurred. A real household sits between them at best, before retailer fees and network tariffs.
A 5 kW / 13.5 kWh unit, one full cycle per day, 86% round-trip efficiency, and state of charge starting and ending each day at 50%. Charging between 9am and 4pm is assumed to come from PV surplus valued at the forgone spot export; outside that window it is a grid import at the spot price. Settlement is at the regional wholesale price.
The forecast tier acts only on AEMO's public D-1 23:30 PREDISPATCH run, then is settled at actual prices — a public-information benchmark, not an operator target. The ceiling assumes every future price is known in advance and is an unreachable upper bound. Neither is a promise of household earnings.
No. The figures are gross wholesale energy only: no FCAS, network tariffs, retailer or VPP fees, degradation, backup reserve or household load. The fleet-scaled view is an upper bound on an upper bound — the aggregate fleet would move prices, CER data lags and revises, and not every install is on a wholesale-exposed product.
Method & data: Full home battery methodVirtual power plantsPublic API
Browse: Battery fleetActual vs optimalPerformance rankingsMarket economicsMarket eventsBid stackBidding strategyVirtual power plantsProject simulatorAskPulseEmail alerts
Loading live data…