Batteries carry an emotional weight that distorts the arithmetic. Everyone wants one, because everyone has sat through an outage. But storage is expensive, and a battery bought for the wrong reason spends most of its cycle life doing nothing while depreciating quietly. It is worth being cold about when it genuinely pays.
Three different jobs
A battery can do three quite different things, and the business case for each is unrelated. It can shift your own solar generation into the evening. It can shave a demand peak that your tariff punishes. Or it can hold you up through an outage. Most proposals blur these together; the sizing for each is different, and a battery sized for one may be badly wrong for another.
Job one: time-shifting
If you generate more than you use during the day and your evening load is significant, a battery moves free electricity into the hours you would otherwise buy it. The arithmetic is simple: the value is the units shifted, times the tariff, times the days per year. Divide the battery cost by that and you have a payback. For households with a heavy evening peak, this often works. For a business that closes at six and consumes nothing after dark, it never does — the array alone is the answer.
Job two: peak shaving
This is the one most operators miss, and it is frequently the largest number on a commercial bill. Many tariffs charge for your highest demand in a period, not just total consumption — so fifteen minutes of compressors starting together can set a charge that applies for the whole month. A battery that discharges into that peak addresses a cost no generator ever touches. If your bill has a demand charge and nobody has modelled it, that is where your money is.
The test: if your demand charge is a large share of your bill, storage may pay for itself on that alone — before it stores a single unit of solar.
Job three: resilience
Here the arithmetic changes entirely, because you are not buying savings, you are buying the avoidance of a loss. That is a legitimate purchase — but it must be priced against what an interruption actually costs you. For a diagnostic lab losing a batch of samples, or a warehouse stopping a packing line, the cost of downtime dwarfs the battery and the decision is trivial. For a household, the honest answer is that the value is comfort and it is worth what comfort is worth to you. Both answers are fine. Confusing them is not.
The number that decides it
Cost per usable cycle, not cost per kWh. A battery is a consumable with a very long life: divide the price by usable capacity, then by rated cycles, and you have the real cost of every unit it will ever store. On that measure the cheap chemistry frequently loses badly to LFP, and the argument stops being about the sticker.
- Heavy evening load + daytime surplus → time-shifting probably pays
- Large demand charge on the bill → peak shaving may pay on its own
- Interruption costs real money → resilience is the case, price it honestly
- Closes at six, no demand charge, rare outages → you may not need one yet
The best thing we can sometimes tell a client is to buy the array now and the battery in three years, when the tariff has moved and the storage price has not. Nobody enjoys hearing it. It is usually correct.
Written by Engineering team. This article is general guidance, not a substitute for a site-specific assessment.