The usual threshold is 80 percent of original capacity, at which point the bank no longer covers what it was sized for. Measuring that requires a capacity test rather than a voltage reading, since voltage tells you very little about how much energy a battery still holds.
Most people notice the symptom before the number. Running out overnight when you never used to is the practical signal that a bank has crossed the line.
Quick Answer
Replace when usable capacity falls below what your loads require, commonly around 80 percent of original. Test by discharging a known load and measuring what you actually get. Sudden changes, swelling, or heat point at a fault rather than age.
Signs Worth Investigating
| Symptom | Likely meaning |
|---|---|
| Runs out earlier than it used to | Capacity loss, normal aging |
| Charges fully far faster than before | Less capacity to fill |
| Voltage drops sharply under load | Rising internal resistance |
| One battery differs from the rest | Single cell or unit fault |
| Swelling or heat | Fault, isolate rather than monitor |
| Sudden change over days | Fault rather than aging |
The distinction that genuinely matters here is gradual against sudden. A slow decline across several years is ordinary aging, where a change over days or weeks points at something specific and identifiable.
Why Voltage Is a Poor Test
Voltage indicates state of charge rather than total capacity, and those two are entirely different measurements.
A degraded battery still reaches full voltage. It simply holds far less energy at that voltage, which a voltmeter cannot see.
Lithium makes this worse, since LiFePO4 has a very flat discharge curve and voltage barely moves across most of its usable range.
Resting voltage after several hours disconnected is more informative than voltage under load, and it still does not measure capacity, covered in battery monitors for off grid solar.
How to Actually Test Capacity
Charge fully and rest
Bring the bank to full and let it sit disconnected for several hours so surface charge dissipates and readings settle.
Apply a known load
Something with a steady, measurable draw. A resistive load is ideal, and any consistent load with a known wattage works.
Measure to the cutoff
Record how long the load runs before the low voltage cutoff, then multiply by the load to get delivered watt hours.
Compare against rating
Delivered energy against nameplate capacity multiplied by rated depth of discharge gives you the percentage remaining, covered in depth of discharge.
Expected Lifespan by Chemistry
Flooded lead acid typically manages somewhere between three and seven years in solar service, depending very heavily on how deeply it was routinely cycled and how hot it was allowed to run.
AGM and gel sit in a similar range, sometimes slightly longer, with the same sensitivity to depth of discharge.
Lithium iron phosphate commonly lasts eight to fifteen years, with thousands of cycles rated at 80 percent depth of discharge.
Those ranges are wide because usage decides more than chemistry does, and a badly treated lithium bank can underperform a well-treated lead acid one, covered in why solar batteries degrade over time.
Replacing Part of a Bank
Replacing one battery in a bank of several rarely works well, since the new unit is mismatched against aged neighbors.
The bank continues behaving roughly like its weakest members, so the new battery delivers considerably less than its rating.
Where a single battery has failed early under warranty, replacing that one is reasonable as a repair rather than as an upgrade.
Otherwise, replacing the whole bank at once is the approach that actually restores capacity, covered in mixing battery brands in a bank.
Faults That Look Like Old Age
Before concluding a bank has aged out, several fixable problems produce the same symptoms and are worth ruling out.
Loose or corroded terminals raise resistance at the connection rather than in the battery, which shows up as voltage sag under load and reduced delivered energy. Cleaning and retorquing terminals is free and fixes it.
An undersized array is the common systemic cause, since a bank that never reaches full charge behaves like a smaller bank and, on lead acid, genuinely becomes one through sulfation.
Incorrect charge settings do the same thing quietly. A controller configured for the wrong chemistry or the wrong voltage will chronically undercharge, and the bank gets blamed.
Parasitic loads drain capacity between charges and are easy to miss, since inverters, controllers, and monitors all draw standby current continuously.
A single failed cell within one battery drags a whole string, and that presents as bank-wide capacity loss when the actual problem is one unit that could be replaced.
Checking those before buying a new bank is worth the afternoon, because replacing batteries does not fix a wiring or configuration problem and the new bank will decline the same way.
Deciding When It Is Worth It
The practical threshold is whether the bank still covers your requirement rather than a percentage on a datasheet.
A bank at 70 percent capacity that still gets you through the night is not urgent, where one at 85 percent that leaves you short in winter is.
Seasonal variation matters here, since capacity falls in cold weather and a bank that copes in summer may not in January.
Adding capacity rather than replacing is sometimes the answer, provided it goes in as a separate bank rather than mixed with the aged one.
Tracking Decline Before It Becomes a Problem
Capacity loss is gradual enough that people notice it only once it inconveniences them, and a small amount of record-keeping turns that into something you can see coming.
Run a capacity test when the bank is new and keep the number. Without a baseline measured on your own system, every later test is being compared against a datasheet figure rather than reality.
Repeat annually, at roughly the same time of year and the same ambient temperature, since testing a bank in July and comparing it to a January result measures the weather as much as the battery.
A shunt-based battery monitor does most of this passively, counting amp hours in and out so you can see delivered capacity trending across months without running a deliberate test.
Watch the trend rather than any single reading. Capacity fluctuates with temperature and recent usage, so a line drawn across several measurements is far more informative than one number.
Note charging behavior too, since a bank reaching full noticeably faster than it used to is telling you there is less to fill, and that frequently shows before the discharge side becomes obvious.
Having that history also makes a warranty claim considerably easier, since documented decline against a baseline is evidence where a subjective impression is not.
What Shortens Life Most
Heat, which roughly halves lifespan for every 10C above about 25C and is the largest single factor in most installations.
Deep cycling beyond the rated depth of discharge, which lead acid punishes far more severely than lithium does.
Sitting at low state of charge for extended periods, particularly for lead acid where it drives sulfation.
And chronic undercharging, where a bank never reaches full because the array is undersized for the load, covered in what drains a solar battery.
Related Reading
- why batteries degrade
- depth of discharge
- mixing brands
- battery monitors
- what drains a battery
- choosing a battery
Frequently Asked Questions
How do you know when a solar battery needs replacing?
When usable capacity no longer covers your loads, commonly around 80 percent of original. A capacity test measures that; a voltage reading does not.
Why can’t I just check the voltage?
Voltage shows state of charge rather than capacity. A degraded battery still reaches full voltage while holding far less energy at it.
How do I test capacity properly?
Charge fully, rest several hours, apply a known steady load, and measure how long it runs to cutoff. Multiply by the load for delivered watt hours.
How long should a battery last?
Three to seven years for lead acid, eight to fifteen for LiFePO4. Usage decides more than chemistry, so those ranges are wide for good reason.
Can I replace just one battery?
Rarely worth it. The bank behaves like its weakest members, so a new battery among aged ones delivers well below its rating.
What if capacity dropped suddenly?
That points at a fault rather than aging. Sudden changes over days, swelling, or heat all mean isolating and investigating rather than monitoring.
How do I track decline over time?
Run a capacity test when the bank is new and keep that baseline, then repeat annually at a similar ambient temperature. A shunt-based monitor does most of it passively.
What shortens battery life most?
Heat, followed by deep cycling beyond the rated depth and sitting at low state of charge. Heat alone roughly halves life for every 10C above 25C.
