On this page
UPS battery replacement is the most predictable expense of owning a UPS, and the least understood. Owners are often surprised that a battery "rated for 5 years" fails in 2, or that a unit that has never seen an outage still needs a new battery. Both make sense once you separate the three ideas below.
Design life, service life, and end of life
| Term | What it means | Where you see it |
|---|---|---|
| Design life | How long the battery lasts on float charge at 77 °F (25 °C) under ideal conditions before reaching end-of-life capacity | Battery datasheets, often a range such as 3 to 5 years or 10+ years |
| Service life | How long it actually lasts in your UPS, in your room, with your outages | Your experience, and the replacement reminder in some UPS software |
| End of life | Capacity has fallen to 80% of rated, the usual criterion in IEEE 1188 and battery maker literature | Test results; a failed self-test usually means a battery well past this point |
European battery makers commonly classify VRLA batteries by design life through the industry association EUROBAT, with categories ranging from general-purpose (roughly 3 to 5 years) to long-life (12 years or more). Most small UPS units use batteries from the lower end of that range. Larger systems often use long-life grades. The classification describes ideal conditions; it is not a promise.
What actually wears the battery out
Positive grid corrosion
Every lead-acid battery on float charge slowly corrodes the lead grid of its positive plates. Corrosion raises internal resistance and eventually reduces the active area that can deliver current. This is the dominant aging mechanism in batteries that are rarely discharged, and it runs faster at higher temperature and higher float voltage.
Dry-out
VRLA batteries recombine most of the gas produced on charge, but not all. A small amount of water escapes through the valve over years, more when hot or overcharged. Because the electrolyte is held in glass mats, there is little to spare; losing water raises internal resistance.
Sulfation
When a lead-acid battery sits partially discharged, lead sulfate crystals grow and harden on the plates and become hard to convert back. This is why a battery stored for many months without recharge, or one in a UPS with a weak charger, loses capacity permanently. See storing a UPS and batteries.
Cycling
Each discharge sheds a little active material from the plates. Small UPS batteries are not deep-cycle batteries; they typically tolerate only a few hundred deep cycles. A few outages a year barely matter, but daily brownouts or outages can make cycling the main cause of failure.
The heat penalty, worked out
Battery makers publish float-life curves showing life falling steeply above 77 °F (25 °C). A widely used rule of thumb summarizes those curves: life roughly halves for every 8 to 10 °C (15 to 18 °F) of sustained temperature above 25 °C. This follows from chemistry: corrosion is a reaction whose rate rises exponentially with temperature.
Applying it to a battery with a 5 year life at 25 °C, with both ends of the rule shown:
| Battery temperature | Life if halving per 8 °C | Life if halving per 10 °C |
|---|---|---|
| 77 °F (25 °C) | 5.0 years | 5.0 years |
| 86 °F (30 °C) | 3.2 years | 3.5 years |
| 91 °F (33 °C) | 2.5 years | 2.9 years |
| 95 °F (35 °C) | 2.1 years | 2.5 years |
| 104 °F (40 °C) | 1.4 years | 1.8 years |
Sample calculation for 33 °C with D = 8: the exponent is (33 - 25) / 8 = 1, so life is 5 x 0.5 = 2.5 years. For 40 °C with D = 10: the exponent is 15 / 10 = 1.5, and 0.51.5 is about 0.354, giving 5 x 0.354 = 1.8 years.
Two practical notes. First, what matters is the battery's temperature, not the thermostat's. A UPS inside a closed cabinet or next to a server's exhaust can run its battery several degrees above room temperature, and the UPS itself adds heat while charging. Second, cooler than 25 °C does not buy much extra life by this rule, and reduces available runtime, so there is no point refrigerating a UPS.
Our analysis: average temperature understates the damage
Because aging speeds up exponentially, a battery that spends half the year at 25 °C and half at 35 °C does not age like one held at 30 °C. Using the 8 °C rule, the summer half ages the battery about 2.4 times as fast as the winter half, for an average aging rate of about 1.69. Life works out to 5 / 1.69 = about 3.0 years, versus 3.2 years if you plugged in the 30 °C average. The gap grows with bigger swings. When estimating life for a space that gets hot seasonally, use the hot-season temperature for a conservative figure. The battery life estimator lets you enter seasonal temperatures.
Expected life by environment
| Environment | Typical service life | Main limiter |
|---|---|---|
| Air-conditioned office or living space, few outages | 3 to 5 years | Calendar aging |
| Home office in a warm climate without AC, or near a radiator or window | 2 to 4 years | Heat |
| Closed closet or cabinet with a NAS, switch or server | 1.5 to 3 years | Heat buildup |
| Unconditioned garage or attic-adjacent space in a hot summer climate | 1 to 2.5 years | Heat |
| Area with daily outages or brownouts | 1 to 3 years | Cycling plus recharge stress |
| Cool basement, stable power | 4 to 6 years | Calendar aging, slowly |
How outages and discharge depth add up
Cycle wear depends strongly on depth. A 30 second blip that takes 2% of capacity barely registers. A run to the low-battery cutoff is a full deep cycle. As a rough guide, a small VRLA battery might deliver a few hundred full cycles, and many more shallow ones, but figures vary a lot by design and are rarely published for UPS-grade batteries.
The consequence for planning: if outages are frequent where you live, configure computers to shut down early (after a couple of minutes on battery) rather than at low battery. You give up some ride-through but avoid a deep discharge on every event. See how much runtime you need.
How to get the full life out of a battery
- Keep the UPS cool. Give it open air around its vents, keep it out of closed cabinets and away from equipment exhaust, heaters and sunny windows. See where to place a UPS.
- Keep it plugged in and charged. Leaving a UPS unplugged for months lets the battery self-discharge toward sulfation.
- Avoid needless deep discharges. Use timed shutdowns and do full runtime calibrations only occasionally.
- Do not overload it. High discharge currents stress the battery and the inverter; keep load at or below about 80% of the watt rating.
- Buy fresh replacements. Check date codes; a battery that sat in a hot warehouse for a year has already used some of its life.
- Record the install date on the battery or UPS, and plan replacement by age rather than waiting for failure if the load is critical.
A hot battery is also a safety issue
Heat does more than shorten life. A failing battery in a hot environment can draw excess charge current, heat further and swell. If a battery case bulges, or the UPS smells hot or of sulfur, stop using it; see swollen UPS battery.
What about lithium?
LiFePO4 batteries in UPS units designed for them typically claim much longer service lives, often in the 8 to 15 year range, because they tolerate both cycling and moderate heat better. They still age faster when hot and their BMS and electronics have their own lifespans. If your UPS lives somewhere warm, or replacement visits are costly, compare the options in lithium UPS pros and cons.
When the time comes, when to replace a UPS battery covers the tests that tell you a battery is done.
Frequently asked questions
How long does a UPS battery last on average?
For consumer and small business UPS units with standard VRLA batteries, 3 to 5 years is a reasonable expectation in a climate-controlled room. Some last longer in cool locations with few outages; many last only 2 years in hot closets or places with frequent outages. Long-life VRLA and lithium batteries in units designed for them can last considerably longer.
Does a UPS battery wear out if there are no power outages?
Yes. Lead-acid batteries age on float charge even if they are never discharged, mainly through slow corrosion of the positive grid and drying of the electrolyte. That calendar aging, accelerated by heat, is what ends most UPS batteries, not the number of outages.
Does cold weather damage a UPS battery?
Cold reduces the capacity a VRLA battery can deliver at the moment, so runtime drops in an unheated space in winter. It does not accelerate aging the way heat does, as long as the battery stays charged. A discharged lead-acid battery can freeze at temperatures where a charged one will not, so keep it charged in cold locations.
Is the battery date on the label the install date?
Usually not. Labels often carry a manufacturing date code, and the battery may have sat in a warehouse for months. Many people write the install date on the battery or the UPS with a marker. Some UPS models let you record the battery install date in their menu or software so they can remind you.
Why did my new replacement battery die after one year?
Common causes are heat around the UPS, a battery that sat discharged for too long before sale, a mismatched or low-quality replacement, or a UPS charger fault that overcharges or undercharges. Check the location temperature first, then compare the UPS's charging voltage against the battery maker's float specification if you are comfortable doing so, or have it serviced.
Sources and further reading
- IEEE Std 1188, Recommended Practice for Maintenance, Testing, and Replacement of VRLA Batteries for Stationary Applications (80% capacity end-of-life criterion)
- IEEE Std 1184, Guide for Batteries for Uninterruptible Power Supply Systems
- EUROBAT, guide to VRLA battery design-life classifications
- Battery manufacturer VRLA technical manuals (float life vs temperature curves)