Storage profitability is usually discussed as a function of the price of electricity. The price is the least useful number in a tariff.
The assumption, stated so it can be tested
The claim is that expensive electricity makes storage profitable and cheap electricity does not. It is reasonable, because a high price per kilowatt-hour feels like a larger prize to capture.
It is wrong for a simple reason. Storage does not sell electricity at the average price. It moves consumption from one hour to another, and it removes a separately billed peak. Neither operation depends on the average rate; both depend on the structure around it, and the structure can change without moving the average at all.
One site, one battery, three tariffs
Worked example. The site consumes 2.4 GWh a year, runs 330 days, and carries a monthly peak of 560 kW that it wants reduced to 440 kW. The same 500 kW container is quoted under all three tariffs, so only the tariff moves.
Ruibit publishes the 500 kW container at 1.045 MWh with 940.5 kWh usable at 90 percent depth of discharge, and those two numbers set the duration the arbitrage leg can capture: 940.5 kWh at 500 kW is 1 hour 53 minutes.
Source: the structures in the middle column are the three tariffs compared; the annual figures are arithmetic on the published 1.045 MWh rating, 940.5 kWh usable energy and 500 kW power of the container, at an 88 percent round trip efficiency assumed for this example.
Why the cheapest tariff produces the most value
Read the average rates in order: tariff A is the most expensive electricity at EUR 0.16 per kWh, tariff B at EUR 0.14, and tariff C the cheapest at EUR 0.13.
Now read the value: EUR 0, EUR 12,960, and EUR 50,721. The ranking is inverted exactly.
The two things storage sells are missing from the first two tariffs. Tariff A has no demand charge, so there is no peak to remove, and no rate difference, so there is nothing to shift. The system can run and produce no saving at all, and no price rise changes that, because a flat tariff without a demand component has no storage value to capture.
Tariff B has a peak window, and the window is where intuition fails again: it exists to measure demand, not to price energy differently. The system can cut the monthly peak by 120 kW, worth EUR 12,960 a year, and do nothing else. Payback on EUR 420,000 installed is 32.4 years.
Tariff C prices those four hours differently as well as metering them, and that single addition creates the arbitrage leg, the larger of the two.
The two numbers that decide it
Both come from the tariff, and neither is the average rate.
The first is the spread after efficiency. Storage buys at the off-peak rate and sells at the peak rate, and the conversion is lossy. Charging at 93.8 percent one-way efficiency and discharging at the same figure costs EUR 0.117 to deliver EUR 0.216, so a peak rate of EUR 0.23 against EUR 0.11 off peak is worth EUR 0.0985 per kWh cycled, not the EUR 0.12 the spread suggests. That figure, not the headline spread, builds EUR 30,561 of the tariff C total.
The second is the duration the window has to cover. The container holds 1 hour 53 minutes at full power, so a peak-rate window of one hour captures 500 kWh of the 940.5 kWh it holds and loses EUR 14,314 of annual value without any rate changing. A peak window constrains capture, not only price.
Where the shape stops mattering, and what to ask
Two conditions blunt the structural argument, and both are worth checking before it is used.
If the site's own peak falls outside the peak-rate window, the two legs stop sharing one discharge. On the 60 days a year when a two-hour event sets the peak, the system holds 240 kWh back, which costs EUR 1,418 a year and moves payback from 8.3 to 8.5 years. That is the difference between one discharge serving two duties and two discharges consuming the same window.
A tariff with a large demand charge and no rate difference still justifies capacity, but only for the peak, so the unit should be sized for the event rather than the energy.
Ask for the tariff in the structure that matters: the demand charge per kW per month, the off-peak and peak rates, the length of the peak window, and whether it changes with the season. An answer giving the average rate, or the annual spend, leaves the storage value unstated.
Ask which hours count as peak in the month of highest demand, and whether that differs from the month of lowest. An answer that the tariff is the same all year is worth confirming in writing.
Ask what the rate difference is after the site's own efficiency, and what share of annual consumption currently falls inside the peak window. An answer that quotes the spread between the two rates without the efficiency adjustment overstates the value by about 18 percent in this example.
Three things are fixed once the project is built. The installed cost is settled at signature and does not move with the tariff. The duration is fixed at 1 hour 53 minutes, so a shorter peak window cannot be recovered by control. And the tariff is the one input that can change while everything else stays still, which is why a project underwritten at one cycle a day belongs against the tariff's next revision rather than its current table.
The formulation that replaces the assumption
Replace "is electricity expensive here" with two questions answered from the bill. Is there a demand charge, and how large is it per kW per month. Is there a rate difference between hours, and is the peak window longer than the duration of the unit.
A site that answers no to both is not a storage project at any electricity price. A site that answers yes to both is a storage project even where the average rate looks low, and in this example that difference is worth EUR 37,761 a year.
FAQs
1. What makes battery storage profitable under a tariff?
Two structural features, neither of which is the average price. A demand charge gives the peak something to remove, and a rate difference between hours gives the battery something to shift. A 500 kW system earning EUR 14.00 per kW per month on a 120 kW reduction collects EUR 20,160 a year before arbitrage.
2. Can storage be profitable where electricity is cheap?
Yes, and the cheapest tariff can pay best. In the comparison used here, the tariff with the lowest average rate, EUR 0.13 per kWh, produced EUR 50,721 of annual value, while a tariff averaging EUR 0.16 with no demand charge and no rate difference produced nothing at all.
3. Why is the average electricity price the wrong metric?
Because storage never sells at the average. It reduces a separately billed peak and moves energy between hours. A flat tariff contains neither a demand component nor a spread, so no rise in its price creates storage value.
4. How long does the peak window need to be?
At least as long as the unit can discharge at full power. A 1.045 MWh container rated 500 kW holds 1 hour 53 minutes, so a one-hour peak-rate window captures 500 kWh instead of 940.5 kWh and loses EUR 14,314 a year without any rate changing.
5. What should be checked on the bill first?
The demand charge per kW per month, the off-peak and peak rates, the length of the peak window, and whether the peak hours change by season. The spread should also be adjusted for round trip efficiency, which is worth about 18 percent in the example above.
