SizingKitElectronics
Battery runtime calculator
A 200 Ah 12 V bank taken from full down to 50% has 1,200 Wh to spend, and a 400 W appliance behind a 90% inverter pulls 444 W off the terminals — so 2 hours 42 minutes. This page runs that division from wherever your monitor says the bank is now rather than from full, takes the load in watts or in DC amps on the bus, and prints what ten more points of discharge would buy you. It then does the part the division misses entirely: an inverter stops on volts rather than on charge, and two meter readings are enough to show how much earlier that happens.
- 100% free
- No signup
- Watts or bus amps
- Sag measured, not tabulated
- 6 loads, printable
The bank you have, and what is pulling on it
Start from where the monitor says you are, not from full. A bank at 70% has 30 points to spend before it reaches a 40% stop, and that is the number this page divides by the load.
The whole bank. Parallel strings add amp-hours; series strings do not.
Nominal, and the same figure the disconnect below is set against.
From a shunt-based monitor if you have one. A voltmeter under load will not tell you this.
Where you intend the discharge to end, which is not necessarily where the inverter will end it.
Amps entered here are DC amps at the battery, not the AC amps an appliance draws.
Default 90% from the efficiency curve on the inverter's own datasheet, read at the load you are running. Efficiency is a curve rather than a number: most units peak between a quarter and three quarters of rating and fall away at both ends, so a 3 kW inverter running a 40 W load can be under 70%.
Clear it for a load wired to the DC bus — those pay no conversion loss at all.
Runtime
RUNTIME
2 h 42 min
AS A DECIMAL
2.7 h
- Spendable charge
- 100 Ah
- Spendable energy
- 1,200 Wh
- Off the battery terminals
- 444 W · 37 A
If you relax the floor: taking the stop point ten points lower buys 0 h 32 min more, for a total of 3 h 14 min. Whether that is worth the cycle life it costs is a question for the chemistry rather than for this arithmetic.
Where the discharge actually ends
An inverter does not measure charge — it measures volts, and it drops out when they fall past a setting. Because terminal voltage sags under load, it will reach that setting while there is still charge in the bank. Two meter readings, thirty seconds apart, are enough to measure how much.
Default 10.5 V — the low-voltage disconnect setting on the inverter itself — 10.5 V is the usual factory value on a 12 V unit, and the manual is the authority.
- Sag at this load
- 0.6 V
- Measured internal resistance
- 16.2 mΩ
- Headroom before dropout
- 1.6 V
That sag means the disconnect sees 0.6 V less than the bank's own resting voltage, so it trips while the resting figure is still 11.1 V. Nothing on this site converts a resting voltage into a state of charge: the relationship is specific to the chemistry, moves with temperature, and needs hours of rest to settle, so a shunt-based monitor is the honest instrument for that question.
The same bank against six loads
1,200 Wh spendable, divided by each load in turn. This table is exactly proportional by construction, and reality is not: see the note under it for the two effects that bend it and where each is quantified.
| Of your load | Watts | Bus amps | Runtime | Decimal |
|---|---|---|---|---|
| 25% | 111 W | 9.26 A | 10 h 48 min | 10.8 h |
| 50% | 222 W | 18.5 A | 5 h 24 min | 5.4 h |
| 75% | 333 W | 27.8 A | 3 h 36 min | 3.6 h |
| 100% | 444 W | 37 A | 2 h 42 min | 2.7 h |
| 150% | 667 W | 55.6 A | 1 h 48 min | 1.8 h |
| 200% | 889 W | 74.1 A | 1 h 21 min | 1.35 h |
Two things bend this table downward at the heavy end and neither is applied here. A battery delivers less capacity the harder it is drawn, which the amp hour calculator quantifies with a Peukert exponent; and a conversion stage burns a fixed number of watts just being switched on, which the UPS runtime calculator carries as an idle draw. Applying either one here would mean this page owning a figure that already has a home.
How to work out how long a bank has left
Two figures give you the arithmetic. The third pair of readings tells you whether the arithmetic will be allowed to finish.
Say where the bank is now and where you want it to stop
Both ends, not just the bottom one. A bank showing 70% that you intend to stop at 40% has thirty points to give, and entering it as though it were full overstates the answer by more than twice. Read the starting figure off a shunt-based monitor if one is fitted, because a voltmeter under load will not tell you.
Enter the load the way you know it
Watts for an appliance nameplate, DC amps if you are reading a clamp meter on the battery cable. Amps go straight into the division because they are already past the inverter; watts get the conversion loss added first, and only if the load actually runs through an inverter rather than off the DC bus.
Measure the sag and compare it with the disconnect
Read the terminal voltage with the load off, switch the load on and read it again, then enter the setting your inverter drops out at. The difference between those two readings is what the disconnect sees on top of the discharge itself, and it is the reason a bank with charge left in it stops working.
Technical specifications
| Division performed | Spendable energy over the power at the battery terminals. 200 Ah at 12 V spent from 100% to 50% is 1,200 Wh; a 400 W appliance behind a 90% inverter is 444 W off the terminals, giving 2 h 42 min. |
|---|---|
| Load entry | Watts, any power unit written into the field, or DC amps on the bus. Amps bypass the efficiency figure entirely, because a current measured at the battery is already past the conversion. |
| Starting point | Any state of charge from 1% to 100%, paired with a stop point from 0% to 99%. Runtime from 70% down to 40% spends thirty points, not the seventy a full-bank calculator assumes. |
| Sag measurement | Internal resistance derived from your own resting and loaded readings over the bus current: 0.6 V of sag at 37 A is 16 mΩ. No resistance table is published here, because a meter measures the real figure in about thirty seconds. |
| Disconnect default | 10.5 V, the usual factory low-voltage setting on a 12 V inverter, which belongs to that unit's manual rather than to this page. Editable — a 24 V unit sits near 21 V and a 48 V unit near 42 V. |
| State of charge from voltage | Not offered anywhere on this site. The resting-voltage relationship is specific to the chemistry, shifts with temperature and needs hours of rest to settle, so a shunt-based monitor answers that question honestly and a lookup table does not. |
| Printed table | 6 loads from a quarter to double the one entered, in watts, bus amps, hours and minutes. Exactly proportional by construction, with both effects that bend it named and linked rather than silently applied. |
| What is transmitted | Nothing at all. Bank size, meter readings and disconnect settings are read in the page and discarded when the tab closes. |
Frequently asked questions
Why did my inverter shut down when the monitor still said 60%?
Because the inverter was watching volts and the monitor was counting amp-hours, and under load those two disagree. Current through the bank's own internal resistance drops the terminal voltage below the resting figure — a 16 mΩ bank at 37 A loses 0.6 V — so a disconnect set at 10.5 V is reached while the resting voltage is still 11.1 V and there is real charge behind it. Heavier loads, colder batteries and older banks all widen that gap, which is why the shutdown seems to arrive earlier every winter.
Can I tell the state of charge from the battery voltage?
Not while anything is drawing from it, and only roughly even at rest. The voltage under load is the resting voltage minus the sag, so it says as much about the load as about the charge; and the resting curve itself has to settle for hours, moves with temperature, and is nearly flat across the middle of a lithium iron phosphate discharge, where a hundredth of a volt covers a large fraction of the pack. A shunt that counts charge in and out is the instrument that answers this.
Should I enter the load in watts or in amps?
Enter whichever one you actually measured, because the page treats them differently on purpose. Watts describe the appliance and have to be divided by the inverter's efficiency to find what leaves the battery; amps read on a clamp meter around the battery cable are already the answer to that question and go in untouched. Entering an appliance's AC amps as though they were bus amps is the common error, and on a 12 V system it understates the draw by roughly a factor of ten.
Does this include what the inverter burns doing nothing?
No — the efficiency figure here is a multiplier on the load, not a fixed overhead. A conversion stage also consumes a constant number of watts simply by being switched on, which matters enormously for a small load left running overnight and hardly at all for a large one; that fixed draw is carried on the UPS runtime calculator, where it is the difference between the marketing runtime and the real one.
The table says twice the load is half the time. Is that actually true?
It is true of the arithmetic and slightly optimistic about the battery. Two effects bend the real curve down at the heavy end: capacity itself falls as the discharge current rises, which is the Peukert effect and is quantified on the amp hour calculator, and the fixed overhead above does not shrink when the load does. Neither is applied to the table here, because a page that quietly folds in another page's assumptions is a page you cannot check.
My bank is six years old. Does this still apply?
The method does, the capacity figure does not. Enter the capacity the bank actually has rather than the one printed on it — a capacity test, or a monitor's history of a full discharge, will tell you — because plates lose active material and lithium cells lose lithium inventory, and a bank that has faded to 70% of rating will do exactly 70% of the runtime this page reports from its nameplate. The sag measurement is a useful second opinion: internal resistance climbs as a battery ages, so a bank that now sags twice as far as it used to is telling you something the amp-hour figure has not caught up with.
Is it harmful to run the bank all the way down?
For lead-acid, yes, and repeatedly so; for lithium, the pack usually stops you first. A lead-acid battery left deeply discharged sulphates, and the damage is cumulative and permanent rather than a single event, which is why the habit of stopping at half exists at all. A lithium pack has a battery management system that disconnects at a cell voltage of its own choosing, so the practical answer there is that you will not get the chance — but repeated full discharges still cost cycle life, and the cost is on the pack's own cycle-life curve.
About discharges that end early
Runtime is stored energy divided by the load, and the reason it is worth doing in watt-hours rather than amp-hours is that an inverter sits between the two. Amp-hours are a count of what leaves the battery; watts are what the appliance consumes, and the conversion between them costs a tenth or so of everything that passes through. Doing the division in energy keeps that loss visible, and it also lets a DC load and an AC load share one calculation without one of them silently paying the other's toll. The other half of doing it honestly is starting from where the bank is rather than from where it was bought: a monitor reading of 70% is thirty usable points against a 40% floor, and that distinction moves the answer more than any efficiency figure ever will.
The interesting failure, though, is not arithmetic. An inverter has no idea how much charge is in the bank; it has a voltage comparator and a setting, and when the terminals fall past that setting it drops the load whether or not there is energy behind them. Current through the bank's own internal resistance is what gets it there early, and that resistance is not a number worth looking up — it varies with the make, the age, the temperature and how many cells are in the string, and you can measure the real one in half a minute with a meter and the load switch. Resting reading, loaded reading, divide the difference by the current: that is the resistance, and multiplying it back by any other current tells you what the disconnect will see. It is also a diagnostic, because a bank whose sag has doubled since it was new has aged in a way its amp-hour rating does not show.
What this page deliberately leaves out has homes elsewhere. The battery's own rate dependence — the reason a heavy discharge yields fewer amp-hours than a gentle one — is on the amp hour calculator; the fixed idle consumption of a conversion stage is on the UPS runtime calculator; and the question of how large the bank should have been in the first place is the battery capacity calculator. Turning an appliance nameplate into the watts this page asks for, power factor and all, is watts to amps read backwards. And a runtime figure is only ever as good as the load list behind it — the same discipline a Manual J load calculation imposes on heating and cooling, where a room-by-room list beats a rule of thumb by enough to change the equipment.
Where your meter readings go
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type — loads, lengths, nameplate ratings, the rates your utility charges you — is uploaded, logged or kept, which is also why the calculators carry on working in a mechanical room with no signal.
The resting and loaded voltages you type are a small diagnostic record of somebody's installation, and they are treated as one: nothing is sent anywhere, so the readings stay between you and the meter.