Xbatt BatteryBattery Bank SizingDepth of DischargeVRLAOff-GridAutonomy

Your Battery Bank Is Not the Size on the Invoice

A 48 V 200 Ah bank rated 9.6 kWh gave an off-grid cabin roughly 3 kWh on a hot August night. Depth of discharge, discharge rate, temperature, ageing and inverter losses — five multiplications, and every one is a place to round in the supplier's favour.

Your Battery Bank Is Not the Size on the Invoice

A 48 V, 200 Ah bank went into an off-grid cabin in the spring. Nominal capacity 9.6 kWh. The load was honest — a fridge, lights, a pump that ran twenty minutes a day, about 3.2 kWh between sunrise and sunrise. Two days of autonomy, the installer said. The bank lasted fourteen months.

When it came back to us for testing, the plates were not worn out. They were sulphated, evenly, on all four blocks. The owner had done nothing wrong except believe the number on the invoice.

Here is what the numbers actually were. The cabinet sat against a west wall and reached 41 °C in July. The bank was cycling to about 70% depth most nights, not 50%. The inverter pulled 1.4 A at idle, around the clock, which nobody had counted. And the 200 Ah figure was measured at the 20-hour rate, while the evening load drew closer to the 5-hour rate.

Multiply those four things together and the usable capacity on a hot August night was not 9.6 kWh. It was closer to 3.1.

Two quotes, identical on paper

This is the part that costs buyers money. Two suppliers quote the same 48 V 200 Ah bank. One is sized at 25 °C, 50% depth, C20, and quietly assumes the inverter is 95% efficient. The other is sized at the temperatures and loads the site will actually see. The first quote is smaller and cheaper. It is also the one that fails in year two and gets replaced at the buyer's expense, in a country where the replacement has to clear customs again.

Sizing a bank is not one multiplication. It is five, and every one of them is a place where a supplier can quietly round in their own favour.

1. Depth of discharge is a design choice, not a spec

A lead-acid block rated 200 Ah does not deliver 200 Ah for long if you take all of it every night.

Chemistry Sensible daily DoD Cycles at that DoD What happens past it
Flooded deep cycle 50% 500–800 Positive grid corrosion, water loss
AGM deep cycle 50% (60% occasional) 400–700 Sulphation on the lower plate area
GEL deep cycle 60% 500–900 Less tolerant of high charge current
LiFePO4 80–90% 3,000–6,000 BMS cuts off; capacity fade, not damage

Those cycle counts are typical figures for traction-grade product, not a warranty statement, and they assume the bank gets back to full charge regularly. A bank that lives at 60% state of charge and only sees a full charge twice a month will not reach them no matter what the datasheet says. That is a separate failure mode and it is the one that killed the cabin bank — [partial state of charge is its own article](/news/psoc-sulfation-deep-cycle-lead-acid-battery-solar).

The practical rule: if a supplier quotes you autonomy without stating the depth of discharge, you do not have a number yet.

2. A 200 Ah battery is not 200 Ah at your load

Battery capacity is published at a discharge rate. Change the rate, change the capacity. For VRLA the difference is not small.

Discharge rate Typical capacity vs C20 (AGM) What it looks like in practice
C20 (10 A on a 200 Ah block) 100% The datasheet number
C10 ~92% A telecom load
C5 ~85% A UPS load, a pump start
C1 ~60–65% Engine cranking, inverter surge

This is the Peukert effect, and it is why a 200 Ah block that runs a 10 A load for twenty hours will not run a 40 A load for five. If your load profile has a big evening peak — and almost every off-grid house does — size against the peak rate, not the average.

3. The cabinet is not at 25 °C

Nameplate capacity assumes 25 °C electrolyte. Put the same block in a metal cabinet in the sun and two things happen at once: it holds less, and it ages faster.

Ambient around the battery Available capacity Expected life vs 25 °C
25 °C 100% Baseline
35 °C ~103% ~60%
40 °C ~105% ~45%
0 °C ~80% Charge acceptance drops sharply
−10 °C ~70% Do not charge fast; risk of freezing a discharged cell

Capacity goes up with heat. Life goes down much faster. This is the trade that catches people: a hot bank looks fine for a year because it is delivering slightly more than rated, then falls off a cliff. The usual rule of thumb is that lead-acid life halves for every 8–10 °C above 25 °C. If your cabinet hits 45 °C in summer, you have bought roughly a third of the life you thought you bought.

If the site is hot, the fix is cheaper than the battery: shade the cabinet, ventilate it, or move the bank to a north wall. We have seen a $40 sunshade add two years.

4. The inverter takes the first 12%

Load calculations usually start at the appliance and stop there. The bank does not see appliance watts. It sees appliance watts divided by every loss between the battery terminal and the appliance, plus whatever the inverter burns doing nothing.

A modern off-grid inverter runs 88–94% at moderate load and considerably worse at 10% load. Add 2–4% for cabling and terminations if the run is long or the cable is undersized. Then add the idle draw — 1 to 2 A is common on a 48 V unit, and on a small system that alone can be 10–15% of daily energy.

A 3.2 kWh daily load at 90% inverter efficiency and 1.4 A idle on 48 V needs about 3.6 kWh from the bank before any derating at all.

5. Design for year five, not day one

A lead-acid bank is considered end of life at 80% of rated capacity. That is not a fault, it is the definition. So the bank you install on day one has to still meet the load on the day it is 80% of itself, in the coldest month, at the highest load.

If you size to exactly 100% of day-one demand, you have designed a bank that fails its requirement somewhere in year three and gets blamed on the battery.

Putting it together

Same load, three applications. Usable kWh is what the bank can actually deliver on a bad day, after depth of discharge, rate, temperature and losses.

Off-grid cabin Telecom site, 8 h backup UPS, 15 min
Load 3.6 kWh/day 2.0 kW for 8 h = 16 kWh 10 kW for 0.25 h = 2.5 kWh
Design DoD 50% 80% (rare cycling) 80% (rare cycling)
Rate derating 0.9 0.85 0.7
Temperature derating 0.97 0.95 1.0
Ageing margin ÷ 0.8 ÷ 0.85 ÷ 0.9
Bank energy required ~10.3 kWh ~24.8 kWh ~5.6 kWh

That last column is why UPS batteries look oversized for the load. A 10 kW load for fifteen minutes sounds like 2.5 kWh, and the string that gets specified is often 5–6 kWh, because at that discharge rate the battery delivers a fraction of its twenty-hour rating.

What to put on the RFQ

Six lines. Suppliers who can answer them are sizing. Suppliers who cannot are quoting.

  1. Daily energy in kWh, and the peak load in kW with its duration
  2. Required autonomy in hours, and at what end-of-life capacity
  3. Expected ambient temperature range around the battery, summer maximum included
  4. Chemistry and depth of discharge you are designing to
  5. Inverter efficiency and idle draw
  6. Whether the site will ever see a full recharge, and how often

If you want to run the arithmetic before talking to anyone, our [battery bank sizing tool](/tools/battery-sizing) takes the same six inputs and returns amp-hours at your chosen voltage. It is a calculator, not a quotation — treat the output as a starting point to argue with.

What we would have specified for that cabin

Four 12 V 200 Ah AGM blocks was the right shape and the wrong size. At 41 °C, 70% DoD and a C5 evening peak, the honest number was 10–11 kWh, which is six blocks, or four blocks of a larger case — plus a vented, shaded cabinet.

The battery was not the problem. The size was. That is a much cheaper conversation to have before the container ships than after.

Looking for a reliable battery supplier?

Send us your requirements — model, quantity and target market. Quote within 24 hours.

Get a Quote