What one cycle is actually worth

Charging costs C cents. The pack returns a fraction E of what it takes. So delivering one kilowatt-hour into the evening costs C divided by E, and the cycle only gains if the expensive hour is priced above that. On a pack rated 90 percent on the AC side ^[8], charging at twenty-four cents needs the peak to clear about twenty-seven before a single cent of margin exists.

That is the whole test, and it takes ten seconds against your own bill. Two prices and one division decide whether any battery on the market can work for you, before size, brand, warranty or price enters the conversation. If the answer is no, the fix is a different rate plan; hardware cannot repair a spread that is too narrow.

A note on which efficiency figure to use: the same packs publish 96 percent on the DC side ^[9] and 90 percent on the AC side. Your meter sees the AC figure, because the conversion losses sit between the cells and the house. Using the DC number makes the arithmetic look six points better than your bill will. The manufacturer prints both, on one line, without saying which one your meter sees:

These are three residential plans published by one utility, transcribed with their own period boundaries. Nothing about the house changes between the rows — only the plan it is billed on.

Read the last two columns together with the third. The widest spread on the list — 74¢/kWh ^[1] against 34¢/kWh ^[2] — also has the shortest window to earn it in. A three-hour peak means the pack must deliver its whole useful charge into three hours, so a household that only draws two kilowatt-hours in that span cannot use most of a large battery no matter how attractive the price per kilowatt-hour looks.

The middle plan trades a narrower gap for a longer window: 58¢/kWh ^[3] against 34¢/kWh ^[4] over five hours. And the plan named for battery households sits at 59¢/kWh ^[5] against a notably cheaper 26¢/kWh ^[6] — the cheapest charging of the three, which is what a battery actually wants.

Winter is a separate calculation and usually a worse one. Two of these plans open a daytime block cheaper than the overnight rate — as low as 32¢/kWh ^[7] — which is good for charging, while the evening price falls, which is bad for selling. The worksheet weights the two seasons by the months each plan assigns to them rather than quoting a summer figure and calling it a year. Both rails, as the utility publishes them:

Multiplying a good summer day by 365 is the most common way these numbers get oversold. Three separate subtractions stand in the way. The peak on most plans runs weekdays only, which removes about a hundred days. Winter prices are lower, so the days that remain are not all worth the same. And a household that travels, or that deliberately holds the pack full ahead of a storm, gives up cycles it could have monetised.

Our starting figure of 230 useful cycle days ^[10] is an assumption, not a measurement, and it is labelled as ours everywhere it appears. It scales the annual result linearly, so if you know your own pattern, put it in the box — that single change is worth more than any refinement to the price data.