A battery quote's payback is two savings and two costs added together. Each is set by a document the factory can read. Here they are, line by line.
A battery quote usually arrives with one number on the front page: the payback, in years. Three,
perhaps four.
That number is a sum. Behind it are two savings the battery makes on the factory’s electricity bill
and two costs it adds to the same bill, and the size of every one of them is set by a document the
factory can read for itself. A quote that shows the sum and not the parts is asking to be believed
rather than checked.
This is the bill, line by line.
The first saving: energy at the peak
Since 2023 the law has required a factory’s electricity to cost different amounts at different times of
day. Rule
8A of the Electricity (Rights of Consumers) Rules, added by the Ministry of Power on 14 June 2023,
required every commercial and industrial consumer with a maximum demand above 10 kW to be on a
time-of-day tariff by 1 April 2024.
The rule sets floors, not prices. At peak hours the energy rate must be at least 1.2 times the normal
rate. In the eight solar hours named by the state commission it must be at least 20 per cent below
it. Peak hours cannot run longer than solar hours. And the time-of-day rate applies to the energy
charge only.
A battery earns on the gap. It charges in the cheaper hours, discharges in the expensive ones, and the
factory buys less power at the peak rate.
How wide is the gap? That is the state’s decision, not the rule’s. Each state commission sets its own
hours and its own rates in its tariff orders, and it changes them. So the first thing a quote’s energy
saving rests on is a specific order, with a date. If the quote does not name it, you cannot know
whether its rates are this year’s.
The second saving: the demand charge
The second line is the one most factory owners know by its sting. An HT factory pays a demand charge
every month, in rupees for each kVA of its billing demand. In Karnataka, the HT industrial demand
charge was ₹340 per kVA a month in 2024-25, beside an energy charge of 690 paise a unit, as the
distribution companies’ own tariff proposals for 2025 to 2028 list it.
The demand charge is set by the highest demand the factory draws in the month. That peak is often
short: a furnace starting, two large machines running together. A battery that discharges into those
minutes lowers the peak the meter records, and the charge with it.
Two things decide how much it saves. The first is how the state bills demand. Some states bill the
month’s actual peak; some set a floor from the contract demand, or from earlier months. The second is
the shape of the factory’s own peaks, which a monthly bill does not show. It shows the highest
number, not when it came, how long it lasted or how often it returned. That is why a battery should
be sized from the meter’s fifteen-minute data, twelve months of it, and not from a bill.
The first cost: what the battery loses
Every battery gives back less energy than it takes in. The ratio is called round-trip efficiency, and
it is never 100 per cent.
On a factory, the loss is paid for every day. The battery charges on power bought at the cheaper rate
and returns less of it at the peak. The energy saving in the first section is therefore the gap
between the rates, less what the battery loses crossing it. A narrow gap and a lossy battery can
leave very little.
A quote should say what efficiency its saving was worked out on, and whether that figure is the cell
maker’s rated value or one measured on a running system. They are different claims, and only one has
been tested on a site.
The second cost: the battery’s own power
A battery’s cells have to stay within a temperature range, so the system runs its own cooling and
controls, and draws power to do it. It draws most in the hottest months.
Grid storage contracts treat this as a line of its own. In SJVN’s grid storage tender, the system’s
own consumption is metered separately and is entirely the developer’s cost. On a factory there is
rarely such a line: the cooling runs off the factory’s own supply, at the factory’s own tariff. If the
saving was worked out before that load, it is larger on paper than it will be on the bill.
Banking: when the grid does part of the job
One more thing changes what a battery is worth, and it is not on the bill at all until it is used.
A factory with rooftop or open-access solar can, in some states, put surplus daytime power into the
grid and draw it back later, for a charge. That is banking. The national green open access rules of
June 2022 require it to be permitted at least once a month for consumers of 100 kW and above, and
leave how much, when and at what charge to each state.
Where banking is generous, the grid is already doing part of what a battery would do, and the battery
is worth less. Where it is limited — where banked power cannot be drawn in the peak hours, for
instance — a battery becomes the only way to move daytime solar into the evening. The same battery
can be worth one amount in one state and a different amount in the next.
Why the rough number is not good enough
The fair objection is that all this is detail, and a rough payback is enough to decide whether to
look further.
It is not, for two reasons. The first is what happens to the rough number. It goes into a
board paper, and it becomes the figure the battery is judged against for years. When the bill does
not match it, the conclusion is rarely that the arithmetic was rough. It is that batteries do not
work.
The second is that a rough number is not wrong at random. A payback built on a national average gets
exactly the lines in this piece wrong, because they are the lines that differ by state and by
factory: the hours, the rates, how demand is billed, whether banking is allowed. The error is not
noise around the right answer. It is the answer for a factory that does not exist.
Five questions for the quote on your desk
- Which tariff order, and which date, did the peak and off-peak rates come from?
- How much of the saving is energy, and how much is demand charge, in rupees a month?
- Was the battery sized from twelve months of fifteen-minute data, or from monthly bills?
- What round-trip efficiency, and what power for its own cooling in the hottest month, is the saving
net of — and are those figures rated or measured?
- Did the quote assume you can bank power, in which hours, at what charge?
A quote that can answer all five is a quote you can check. One that cannot has not yet been worked
out for your factory.
Achala Energy’s page
Nandita Rao’s profile