The PowerBill

Is it the rate, or is it you?

A bill going up says nothing on its own: the price per unit may have risen, or you may have used more, or both — and people routinely blame the wrong one. Enter two bills and this splits the change cleanly. It uses only the four numbers you type. No national averages, no assumptions, nothing uploaded.

Earlier bill
Later bill

Any currency — the split is proportional. Compare like with like where you can: the same month last year beats last month, because heating and cooling swamp everything else.


How the split is calculated

With earlier price p₀ and usage q₀, and later price p₁ and usage q₁, the change in the total is split three ways:

ComponentFormulaReads as
Rate effect(p₁ − p₀) × q₀What the new price would have cost you at your old usage
Usage effect(q₁ − q₀) × p₀What your new usage would have cost at the old price
Combined(p₁ − p₀) × (q₁ − q₀)The part that exists only because both moved together

The three add up to the whole change, exactly. Unit price here is your all-in price — total bill divided by kilowatt hours — so it includes standing charges, network fees and taxes. That is the number that matters to a household, and it is also why your all-in price can rise while the advertised energy rate does not.

Reading the result

If most of it is…Then
Rate effectThe decision that moved your bill was made somewhere — a tariff change, a rate case, a capacity cost pass-through. Find out who decided
Usage effectYour bill moved for reasons inside your own home — weather, occupancy, a new appliance. National headlines about data centres are not the explanation for your number
BothNormal, especially across seasons. Compare the same month a year apart to strip the weather out

This tool attributes nothing to AI or data centres, and it cannot: your two bills do not contain that information. It exists so the attribution question starts from an honest number. The evidence on the wider question is in Is AI raising your electricity bill?