Robot vacuum battery replacement: when it is worth it
The number that decides it is how much floor the robot cleans. Nobody publishes that number.
What it comes down to
- What actually decides the lifespan
- How much floor the robot cleans per charge. A studio flat and a two-storey house put the same cell through very different daily loads, and the same nominal “2–4 years” applies to both.
- Why the label misleads
- mAh is not capacity. The same 5,200 mAh is 74.9 Wh at 14.4 V and 56.2 Wh at 10.8 V — a third less energy behind an identical-looking number.
- What usually fails
- Rising internal resistance, not falling capacity. The robot heads back to its dock early while the app still shows charge left. That is a resistance symptom, and it looks nothing like a genuinely empty cell.
- The rule of thumb going around
- Replace the battery if it costs less than half a new robot. It is repeated widely, we could not find a source for it, and it compares the wrong two numbers.
Battery replacement is the decision that extends a robot vacuum’s life or ends it, and the advice available for it is mostly a rule of thumb with no origin. What follows is the part that can be checked: what the numbers on a replacement cell actually mean, which failure the symptoms point to, and the two chemistry families you cannot swap between.
How we approached this
We pulled together battery manufacturer technical datasheets for the cell chemistries used in these machines, the published specifications of replacement packs sold for the major robot brands, and the charging and storage guidance in manufacturer documentation. We then worked the energy arithmetic that the specifications imply but do not state.
We did not run a cell through a cycle-life test, and we have not measured the capacity of a used robot vacuum battery. Nothing on this page is a measurement we took. Where a figure comes from a manufacturer datasheet we name it; where it is our own arithmetic from published numbers, the text says so; and where we went looking for a source and found none, we say that too.
Why your floor plan decides your battery life
Ask how long a robot vacuum battery lasts and the answer you get is a range of years. That range is not a specification — it is the spread created by the variable nobody publishes, which is how much work each charge does.
A robot in a small flat may finish a run in twenty minutes on a single charge, using a fraction of the pack. A robot in a large house may run to the edge of its charge, return to the dock, top up, and go out again — two or three times over. Both owners describe their usage as “one clean a day”. Electrically they are not the same thing at all, and the second case can put several times the wear on the same cell per day.
This matters because a battery’s life is counted in charge throughput, not in calendar time. Two robots of the same model, bought the same month, can arrive at their first replacement years apart. The published lifespan figure is not wrong so much as unanchored: it is a range with no stated floor area behind it.
What this means for you: the useful question is not “how long do these batteries last” but “how much floor does my robot cover per charge, and how often does it top up mid-run”. If it finishes on one charge, you are at the gentle end. If it docks and redeploys, you are at the other end, and you should expect to buy a cell sooner than the box suggested.
mAh is not capacity
Replacement cells are advertised by milliamp-hours and by a nominal voltage, and buyers compare the first number across listings. That comparison is unreliable, because milliamp-hours only becomes energy when you multiply by voltage.
The arithmetic is one line: watt-hours equal amp-hours times volts. Two packs with the same mAh figure and different nominal voltages do not hold the same energy.
| Advertised pack | Nominal voltage | Energy (our arithmetic) |
|---|---|---|
| 2,600 mAh | 14.4 V | 37.4 Wh |
| 3,000 mAh | 14.4 V | 43.2 Wh |
| 5,200 mAh | 10.8 V | 56.2 Wh |
| 5,200 mAh | 14.4 V | 74.9 Wh |
The last two rows are the point. Identical mAh figures, and the 14.4 V pack carries a third more energy than the 10.8 V one. If you are replacing a cell and matching the original voltage, this does not bite you. If you are comparing two listings by their headline number, it does.
Nominal voltage is not arbitrary, either. It falls out of how many cells are stacked in series: nickel-metal hydride cells are nominally 1.2 V each, so a 12-cell pack is 14.4 V and a 10-cell pack is 12 V; lithium cells are nominally 3.6 to 3.7 V, so four in series land at 14.4 V. That is why the same machine family can be sold with packs whose numbers do not line up, and why the voltage on the label is the number to match.
The “50% rule” has no source
Search for advice on this decision and you will meet a rule of thumb within the first few results: replace the battery if it costs less than half the price of a new robot vacuum. It is stated with confidence, in roughly the same words, across a large number of pages.
We went looking for where that number comes from. It is not in any manufacturer documentation we could find, it is not attached to any published testing, and the pages that repeat it do not cite one another or anyone else. As far as we can establish it is a figure that has been repeated rather than derived — which does not make it wrong, but does mean it carries no more authority than the page you read it on.
The larger problem is that it compares the wrong two things. “Battery price against new robot price” treats the robot as a single object with a single remaining life. It is not. It is a battery, a fan motor, two or three drive motors, a brush motor, a wheel assembly, a charging dock and a mainboard, and the battery is the only one of those you are proposing to renew. The comparison that matters is battery price against the remaining life of everything else.
What actually fails first
The symptom owners describe is consistent: the robot starts going back to its dock earlier than it used to, and the run covers less floor than it did. The natural reading is “the battery is worn out”, and the natural expectation is that a worn battery means less runtime on a full charge.
There are two distinct failures behind that symptom, and they call for different responses.
| Failure | What happens electrically | What you observe |
|---|---|---|
| Rising internal resistance | The cell can still store charge, but its voltage sags under load. The robot’s controller sees the loaded voltage cross the low-battery threshold and stops the run while the pack still holds energy. | Runs get shorter while the app or indicator still shows charge remaining. The robot may report low battery almost as soon as the fan spins up. Charging may still complete normally and quickly. |
| Lost capacity | The cell genuinely stores less charge. It also takes less time to refill. | Shorter runs, and a charge cycle that finishes noticeably faster than it used to. |
Resistance is the one that catches people out, because it looks like a logic fault rather than a battery fault. A pack that measures a healthy resting voltage on a multimeter can still collapse the moment the fan draws current. That is why a clean resting reading is not evidence that the battery is healthy — a no-load measurement is the one test a failing cell will still pass.
Battery datasheets make this explicit. Energizer’s technical datasheet for its nickel-metal hydride cells lists internal resistance separately for the charged state and the discharged state — around 30 milliohms charged against 40 milliohms discharged, with a tolerance of plus or minus 20%, measured with an AC current set at 1.0 ampere. Energizer publishes resistance alongside capacity, because it moves with state of charge and it is not constant over the life of the cell.
NiMH and lithium are not interchangeable
Replacement packs for these machines are sold in both chemistries, and older robots were built around nickel-metal hydride while newer ones use lithium. The two are not drop-in substitutes, and the reason is the charging method rather than the shape of the pack.
Nickel-metal hydride cells are charged until the pack shows a characteristic voltage dip or a rise in temperature, and the charger stops on that signal. Lithium packs are charged at constant current to a voltage limit and then held at constant voltage until the current falls. A charger looking for a nickel-metal hydride end-of-charge signal will not find one on a lithium pack; a charger holding a lithium voltage limit will not fill a nickel-metal hydride pack properly. The same physical connector does not make them interchangeable.
There is a temperature dimension too. Energizer’s datasheet gives a charge range of 0 °C to 40 °C and a discharge range of 0 °C to 50 °C for its nickel-metal hydride cells, and states plainly that operating at extreme temperatures “will significantly impact battery cycle life”. A robot that charges on a dock against a wall in an unheated room, or in a garage, spends part of the year outside the window Energizer specifies — and its datasheet treats that as a cycle-life issue, not a temporary performance dip.
Match the chemistry, not just the connector. A pack advertised as fitting your model is not the same claim as a pack using the chemistry your machine was designed to charge. If the original was nickel-metal hydride and the replacement is lithium, the voltage label alone does not settle it — the charging behaviour is the part that has to match, and it is the part listings leave out.
When replacing beats replacing the machine
Strip the rule of thumb away and the decision reduces to three questions, in this order.
| Question | Why it comes first | What changes the answer |
|---|---|---|
| Is the battery replaceable at all? | If the pack is sealed into a module you cannot open, or soldered in, there is no decision to make — the machine is a disposable with a fixed lifespan. | Look for a documented user-replaceable battery, a serviceable cover, and a replacement part sold by the manufacturer rather than only by third parties. |
| Is everything else still working? | Renewing the cell only buys you the remaining life of the rest of the machine. | Navigation and mapping still accurate, both wheels driving, brushes turning, docking reliable, no repeated error codes. If the mapping has drifted or the robot keeps losing its dock, the fault is not the battery and a new cell will not fix it. |
| Has the category moved on? | A replacement cell restores the robot you already own — nothing more. | If your machine predates the current generation of navigation, obstacle handling or self-emptying, you are comparing a battery against a genuine capability jump, not against the same robot new. |
Note what is missing from that list: the ratio of battery price to robot price. It is missing because it is not a question about your situation — it is a question about the retailer’s pricing, and it tells you nothing about how much life the rest of your machine has left in it.
What to check before you buy the replacement
- Match the nominal voltage to the original, not to a similar listing. The mAh figure can differ within limits; the voltage should not.
- Confirm the chemistry. Nickel-metal hydride or lithium, matching what the machine was built to charge. If a listing will not state the chemistry, treat that as a missing specification rather than a detail.
- Check whether the capacity claim is a rating or an estimate. A capacity figure is only meaningful with the discharge rate it was measured at — Energizer’s datasheet, for example, rates its 2,300 mAh cell at a 460 mA (0.2C) discharge and shows the delivered capacity falling at higher rates. A robot vacuum draws far more than 0.2C in use, so a pack that meets its label on a bench test is not the same as a pack that meets it in the machine.
- Prefer a pack with a stated charge-temperature range. If the listing does not give one, you cannot tell whether it is specified for a dock against a cold wall.
- Plan for the reset, not just the swap. Several machines need the battery level recalibrated after a cell change — a full charge, then a full uninterrupted run down to the low-battery stop — or the charge indicator will keep reporting the old pack’s behaviour. Follow the machine’s documented procedure rather than assuming it is automatic.
The honest answer
There is no published lifespan figure for a robot vacuum battery that accounts for your home, and that is the actual finding here. The ranges quoted in buying advice are a spread produced by a variable nobody measures, and the decision rule most people apply to it has no traceable origin.
What you can do is narrower and more reliable: match the voltage and the chemistry, check whether the rest of the machine is healthy before you spend anything, and treat a robot that docks early with charge still showing as a resistance problem rather than a capacity one. For how the rest of the machine ages — and which failure actually ends a working robot — see how long robot vacuums actually last. For the same argument applied to a different filter, see air purifier filters, where the change indicator turns out to be counting hours rather than sensing anything.
Questions this page answers
- When is a robot vacuum battery replacement worth it?
- The widely repeated rule is: replace it if the battery costs less than half a new robot. We could not find a source for that rule, and it compares the wrong two numbers.
- Why is the mAh figure on a robot vacuum battery misleading?
- Because mAh is not capacity. The same 5,200 mAh is 74.9 Wh at 14.4 V and 56.2 Wh at 10.8 V — a third less energy behind an identical-looking label.
- What actually fails in a robot vacuum battery?
- Rising internal resistance, not falling capacity. The robot heads back to its dock early while the app still shows charge left.
- What decides how long a robot vacuum battery lasts?
- How much floor the robot cleans per charge. A studio flat and a two-storey house put the same cell through very different daily loads.
Where this comes from
The material behind this page is listed below, read on 1 October 2026. The cell figures come from Energizer’s product datasheet for its NH15-2300 nickel-metal hydride cell — the one primary manufacturer source here. The watt-hour and series-cell arithmetic is ours, shown so it can be checked. One entry is a negative finding: the “half the price of a new robot” rule is stated widely and cited nowhere, which is what our search found. The charging-mechanism comparison is described as a mechanism because we did not find a single document that stated it in this form.
- The internal resistance figures for a charged and a discharged cell; the ±20% tolerance; the 1.0 ampere AC test current; the nominal 1.2 V cell voltage; the 2,300 mAh rated capacity measured at a 460 mA (0.2C) discharge; the delivered-capacity curves at higher discharge rates; and the charge and discharge temperature ranges together with the statement that extreme temperatures significantly impact cycle life: the Energizer product datasheet for its NH15-2300 nickel-metal hydride cell (Form No. NH15-2300GL1118). This is the primary manufacturer source on the page, and the discharge-rate point in the checklist above rests on it.
- The watt-hour conversions in the capacity table: our arithmetic from the advertised mAh and nominal voltage of each pack configuration. Watt-hours equal amp-hours multiplied by volts. No source states these particular products as a table; the calculation is shown so it can be checked.
- That packs are sold at two nominal voltages for the same machines, and the series cell counts that produce 12 V, 14.4 V and 10.8 V packs: observed across published replacement-pack specifications and the 1.2 V nickel-metal hydride and 3.6–3.7 V lithium cell voltages. The series arithmetic is ours.
- That the “half the price of a new robot” rule could not be sourced: our own search across the pages that repeat it. We found the rule stated frequently and cited nowhere. This is a negative finding about the published advice, not a claim that the rule is incorrect.
- That the two charging methods differ — end-of-charge signal detection for nickel-metal hydride against constant-current, constant-voltage for lithium: general battery-charging practice as described in manufacturer technical documentation for both chemistries. We have described the mechanism rather than quoting a single source, because we did not find one document that stated the comparison in the form used here.
Two things we could not establish and are not going to imply: we found no published cycle-life rating with its test conditions stated for any robot vacuum replacement pack we looked at, and we found no manufacturer documentation that states a battery lifespan in terms of floor area cleaned. Both absences are the reason this page argues from the specification rather than from a lifespan number.
What this page deliberately does not do: naming a best replacement battery. The voltage, the chemistry, the rated discharge rate and the charge-temperature range are all checkable per pack, and a recommendation that skips them is the failure mode this page exists to argue against. See Cleaning for everything else we check in this category — and how we get paid if you want to know what the commission does and does not buy here.