Table of Contents

A commercial battery wins when its kilowatts, its hours, and its controls match the site’s tariff, transformer, and protection rules. A lower price per kilowatt-hour loses when the duration is wrong, the transformer has no room to charge, or the system is not allowed to do the thing the savings model assumed. Fit is not a slogan beside the product. It is the list of limits that decide whether the battery is allowed to discharge on the afternoon that pays for it.

The tariff has a shape, and the battery has to match it

Demand charges bill the highest kilowatts in a defined window. Energy charges bill kilowatt-hours. A battery sized for a two-hour factory peak does nothing extra for you if the tariff’s demand window is fifteen minutes and the peak is a single compressor start you could have staggered. It also does nothing if the site is on a flat energy rate and the sales model assumed a demand ratchet.

Energy-Storage.News has described U.S. cases where demand charges are more than half the bill. Those cases are why C&I storage exists. They are not every contract. Ask for the tariff PDF and a year of interval data before you accept a savings sheet. If the interval file shows a flat daytime load, a short battery will be empty and a long battery may be buying hours you cannot value. The 2024 U.S. storage baseline shows pack cost per kilowatt-hour falling as duration grows, which tempts people to buy hours. Hours you cannot discharge are the expensive kind of cheap.

The grid connection can veto the brochure

Charging is a new load. If the transformer, the feeder, or the contracted import limit is already near its ceiling at night, the battery cannot refill, or it can refill only by paying for an upgrade that was not in the hardware price. Discharging can be vetoed the other way. Where the agreement forbids export, IEEE’s behind-the-meter guidance is that the controller must keep discharge inside the building’s consumption, with a reverse-power relay at the supply in case the controller fails. A savings model that exports the leftover noon solar is a different project from a non-export site. One of them will be rejected in the utility study.

Fault level, protection settings, and the point of connection are the same class of veto. A container that is electrically fine in a field can be the wrong fault contribution on a weak industrial bus. “The cells are certified” does not answer the protection study. The study is about this transformer and this relay, not about the cell.

Fire layout and the room are part of the power system

A battery that cannot be placed where the fire authority will sign will not be a cheaper battery. It will be a redesign. Indoor retrofits, shared walls, and missing access for removal show up as months, not as a small adder per kilowatt-hour. Outdoor containers have their own clearances. Either way, the energy management system still has to see the revenue meter. A controller that only sees the battery’s own sensors will shave a peak the utility is not measuring, or miss the one it is measuring.

Backup is a separate fit. Carrying the whole site through an outage is a different inverter rating, a different transfer switch, and often a different duration from shaving a demand charge. Sites that need both should say so at the start. A demand-charge battery wired only in parallel with a live grid does not become a UPS when the grid fails.

How to test fit before you test price

Put these in order. Stop at the first no.

  • The tariff actually contains the charge the battery is supposed to cut, and the interval data shows a peak the proposed kilowatts and hours can cover.
  • Import headroom exists for charging, or the upgrade is inside the price.
  • Export is either allowed and metered, or forbidden and blocked with a scheme that includes a relay, not only software.
  • The protection study and the fire layout have been named as deliverables, with a revision path if the utility or the fire authority marks them up.
  • The operating mode in a blackout is drawn, including which loads stay on.

Only then is a dollar per kilowatt-hour comparable. How those dollars break apart is in commercial storage cost. What the category means, if the proposal is mixing a home battery with a factory, is in what C&I energy storage is.

Frequently asked questions

Why does system fit matter more than the storage price?

Because a cheap battery that cannot charge, cannot legally discharge, or is sized for a demand charge you do not pay will not return the savings on the spreadsheet. The kilowatt-hour price is real money. It is the second question, after the tariff, the transformer, and the protection rules.

What does “fit” mean on a real grid?

The battery’s power and duration match the peak in the interval data. The site can import enough to recharge. Export is either permitted or properly blocked. Fault level and fire clearance are acceptable to the utility and the fire authority. The meter the controller watches is the meter that sends the bill.

Can software fix a transformer that is already full?

It can slow the charge so you stay under the limit. It cannot create headroom that is not there. If the only way to refill the battery is a new transformer, that transformer is part of the project price. Leaving it out makes the battery look cheaper than the job.

Is a longer-duration battery always a better fit?

No. Longer duration lowers the pack price per kilowatt-hour in the national cost models, and it helps when the peak really lasts those hours. A short demand spike does not need eight hours of cells. Unused hours are a fit failure even when the unit price looks better.

Does a grid-parallel battery provide backup?

Not by default. Backup needs a transfer scheme and a list of circuits that remain supplied when the grid is gone. A battery set only to shave peaks in parallel with a live grid stops being useful in the outage unless that second mode was designed and tested.

Sources