Find the Weak Battery in a Series Bank (Jumbo UPS)

A large flat in Pitampura or an independent house in Faridabad has a Jumbo Home UPS that ran a fan, lights and the TV through long Delhi cuts for years. This summer the backup suddenly collapsed — from a couple of hours down to twenty or thirty minutes — and the inverter starts beeping battery low almost as soon as the power goes. The owner opens the battery cabinet, sees four (or two, or six) batteries sitting side by side, and reasonably concludes that one battery has gone weak. The next question is the hard one: which one, and do I replace just that one or the whole set?

The single most important idea in this guide: in a series-wired bank the weakest battery sets the backup time for the whole set. Four good batteries plus one tired one do not give you four-fifths of your backup — they behave like five tired batteries. A single failing unit really can sink the performance of the entire bank, which is exactly why your backup dropped so sharply when, on the face of it, only one battery is bad.

This guide resolves two decisions: how to find which battery is dying (with checks you can do yourself and the tests a technician runs), and whether to replace that one battery or the whole matched set (with the honest engineering reason behind each choice). Sometimes replacing all of them is the right call; sometimes it is overkill a fair technician would talk you out of. The aim is to give you the test logic so you can judge the evidence yourself before anyone quotes you for a new bank. Everything here applies to 24 V (2-battery), 36 V (3), 48 V (4) and 72 V (6) and larger banks in Jumbo and high-capacity Home UPS systems.

A single tubular battery is nominally 12 V. A small 800 VA home inverter is happy running off one such battery. But to build a 2,500 VA, 3,500 VA or higher Jumbo system, the inverter needs a much higher DC bus — 24 V, 48 V or 72 V — because pushing that many watts off a 12 V bus would mean enormous, impractical currents. So the batteries are wired in series: the positive terminal of one connects to the negative of the next, like links in a chain. Voltages add up while current stays common to every battery. Two 12 V batteries in series make 24 V; four make 48 V; six make 72 V.

The consequence that drives everything else: because the same current flows through every battery in a series string, they charge and discharge in lockstep, like rowers chained to one oar. The bank can only deliver as much usable energy as its weakest member allows. Worse, that weak battery gets driven harder than the others to keep up, so it degrades faster. It is a self-accelerating failure: the tired battery drags the bank down, and the bank wears the tired battery out. Ignoring an early-warning weak battery is the most expensive thing you can do.

One terminology correction: people often ask which cell is weak. In almost every Jumbo bank you do not have loose 2 V cells; you have several complete 12 V monobloc batteries (each internally made of six 2 V cells) wired in series. The right question is which battery is weak, and the answer is the whole 12 V unit you can lift out and replace. A 48 V bank is four batteries, not twenty-four loose cells.

Series is not the only way to wire batteries. Series adds voltage at the same capacity — what Jumbo systems use to reach 24 V to 72 V. Parallel wiring adds capacity (Ah, and therefore backup duration) at the same voltage. Some very large banks combine both. The FAQ goes deeper, but the headline is that series strings are the unforgiving ones: only as strong as their weakest link.

The field signals are fairly consistent. Backup time drops — either suddenly over a week or two, or steadily over months. The inverter trips to battery low far sooner after a cut than it used to. And the deceptive one: the system reaches full charge suspiciously fast. A weak battery surface-charges quickly, so the charger thinks the job is done while very little real energy has gone in. Fast charging is not good news — it is usually a red flag.

There are also physical signs you can check safely, with the system off and no tools. After a charging cycle, a battery running noticeably hotter than its neighbours is suspect — a failing battery wastes energy as heat. Look for a bulged or swollen casing, or heavy white or pale-green corrosion building up on one terminal. On flooded tubular batteries, if one battery needs distilled-water top-up far more often than its siblings, it is working harder and gassing more — another sign it is the weak link.

Crucially, not every backup problem is a dead battery. Before you condemn any battery, rule out the look-alikes: a loose or corroded inter-battery connector (adds resistance, starves the whole string), a failing charger inside the inverter (bank never gets properly full), sulphation from months of deep discharge (common after a long holiday or monsoon shutdown), or a bank that was always undersized for the load. The tell for an undersized bank: every battery tests healthy and similar, with no single outlier, yet the runtime is still short — the problem was never a bad battery but too little capacity. The table maps each symptom to its likely cause and, importantly, to the first thing to do before buying anything.

A safety word first. A 24 V to 72 V lead-acid bank can deliver a dangerous short-circuit current — enough to weld a spanner to a terminal or cause serious burns. Remove rings, watches and metal bracelets before going near the terminals, use insulated-handle tools, never bridge two terminals, and keep the area ventilated for flooded tubular batteries because charging releases flammable hydrogen. Anything beyond visual checks and a simple resting-voltage measurement is best left to a technician. With that understood, here is a safe self-inspection sequence.

The resting-voltage check is the most useful thing a non-technician can do. A healthy, fully-rested 12 V tubular battery sits at roughly 12.7-12.8 V open-circuit at room temperature; around 12.0-12.1 V means it is nearly flat. But absolute numbers drift with brand and temperature, so the real signal is the comparison between siblings. If three batteries read ~12.7 V and one reads 12.1 V after the bank has rested with the charger off, that one is your prime suspect. You are looking for the odd one out, not a textbook number.

Resting voltage is necessary but not sufficient. A battery can show an acceptable open-circuit voltage at rest yet collapse the moment it has to deliver current, because age raises its internal resistance and robs it of usable capacity. Voltage at rest tells you roughly how charged it is; it does not tell you how much real work it can still do. That is why the next step — testing under load — catches the batteries that looked fine on a resting reading.

The single most reliable field test for a series string is the under-load, mid-discharge voltage comparison. With a known load running off the bank, you measure each 12 V battery's voltage a few minutes in. The dying battery's voltage sags fastest and lowest while its siblings hold up. This is the test that exposes a tired battery that resting voltage flattered, and any proper service visit for a weak Jumbo bank should include it. The weak unit usually reveals itself within minutes.

For flooded tubular batteries you can add a specific-gravity (hydrometer) test. Specific gravity measures the strength of the electrolyte, which tracks state of charge in each cell. A fully charged cell reads roughly 1.260-1.280 and a half-discharged one around 1.200 (exact figures vary by brand and must be temperature-corrected). As with voltage, the diagnostic value is the comparison: if one battery's cells read markedly lower than every other battery even after a full charge, that is your weak link. Important caveat: this test does not apply to sealed SMF/VRLA batteries such as Exide Powersafe Plus — they are sealed and you cannot reach the electrolyte. For sealed batteries you rely on voltage, internal resistance and load testing instead.

The third tool is an internal-resistance reading from a battery analyzer. Internal resistance is how much the battery fights the flow of current. As a battery ages, this resistance climbs, and a high-resistance battery cannot deliver current even when its voltage looks acceptable at rest — which is why backup is poor. A service analyzer reports each battery's resistance in milliohms; in a bank installed together, the readings should cluster, and the outlier with markedly higher milliohms is the culprit.

The thread running through all three tests is the same: you are making a relative comparison across a set that was installed together and should behave alike. The battery that stands out is the weak link. But if several batteries have drifted together — multiple units sagging under load or reading high resistance — that is the bank telling you it is ageing out as a whole, and that shapes the replacement decision in the next section.

The diagnosis flow, start to finish:

A worked case from a 48 V bank (four 12 V batteries) where backup had fallen from ~3 hours to ~45 minutes. At rest: Battery 1 — 12.72 V, Battery 2 — 12.70 V, Battery 3 — 12.68 V, Battery 4 — 12.10 V. Three siblings agree closely; number four is the obvious outlier. Under a real load a few minutes in: the three good batteries held ~12.4-12.5 V each, while battery four sagged to ~11.0 V and kept falling. As it dropped it pulled the whole-bank voltage down toward the inverter's low-battery cutoff (~42 V on a 48 V system), which is exactly why the system tripped battery low and cut out so early. Verdict: battery four is the weak link. But note the other three are also three years into a hard Delhi life — precisely the situation where you must think carefully before just swapping the one, the subject of the next section.

To sanity-check what backup your bank should give once it is healthy, run your real load and battery rating through the calculator below before assuming the batteries are the problem. Sometimes the fault is simply that the bank was always a touch small for the load — the undersizing tell from the symptoms table, where every battery reads fine but the runtime is still short.

The rule, stated plainly: in a series bank more than roughly 12-18 months into service, replacing only the failed battery and leaving the aged ones is usually a false economy. It feels cheaper today and costs you more within a year. Here is exactly why.

A brand-new battery has full capacity and accepts charge eagerly. Its old siblings are tired, with reduced capacity and higher resistance. But the string discharges in lockstep — common current through all — so the fresh battery is throttled by the old ones and never delivers anything close to its rating; you have paid for capacity you cannot use. Meanwhile the old batteries, pushed to keep pace with one strong partner, are stressed harder and fail sooner. The usual outcome: the whole bank is back in trouble within 12-18 months and you are buying batteries again, having effectively wasted the new one you bought first.

There is a charging-side problem too. The charger sees the whole string as one unit and applies one charging profile to all of them. A mix of new and old batteries charges unevenly — the new battery and the worn old ones reach different states at different rates, so the new one ends up under-charged while the old ones get over-stressed. That uneven charging accelerates failure further, and mixing batteries can also conflict with battery warranty terms, so it is worth checking your brand's terms before mixing old and new.

In fairness, there is a narrow case where replacing just one is acceptable — and an honest shop will tell you so rather than blanket-selling a full set. If the bank is very young (well under a year), the failure is a clear manufacturing defect or warranty case, and the remaining batteries test near-identical to a fresh unit under load and on resistance, then a like-for-like replacement is reasonable: same brand, same Ah, same model, as close to the same manufacturing batch as you can get. Outside that narrow window, the matched-set principle wins.

The matched-set principle is the heart of it: all the batteries together, same brand, same Ah, same model, ideally the same manufacturing batch, so they age and cycle as one and no single unit is ever the odd one out. Mixing brands, capacities or ages in a series string is, in our experience across Delhi NCR, the most common single cause of repeat failures — owners who just changed one last year and are back this year. Use the decision table and the five-year cost example below to run your own numbers before you choose.

The money argument, with clearly illustrative figures for a ~150Ah-class tubular battery; treat these as ratios, not quotes. Suppose one battery costs ~Rs 14,000 and a matched set of four is ~Rs 56,000. Patch one battery into a three-year-old 48 V bank and you spend Rs 14,000 today. The throttled-new-plus-stressed-old dynamic typically brings a second failure in 12-18 months — another ~Rs 14,000 — and by around year 2 the remaining originals are done, pushing you toward a near-full replacement of ~Rs 42,000 more. The cheap path can run to ~Rs 70,000 over the period, with patchy backup throughout. Replace the matched set of four now for ~Rs 56,000 and it runs a full healthy service cycle with consistent backup, offset partly by buyback on the old lead. Lower upfront is not lower total.

The lowest-upfront, sensible path for most owners is a like-for-like matched tubular replacement. Matched is not a slogan: same brand, same Ah, same model, ideally same manufacturing batch, bought and installed together so the bank ages as one. If your load and runtime have not changed and your inverter is sound, a fresh matched tubular bank restores the system to how it performed when new. Browse matched tubular options on our inverter batteries page; if yours is a sealed SMF/VRLA bank, the UPS batteries range covers that side. Either way we will make sure all units in the set are genuinely alike.

A full-bank replacement is also the natural moment to ask a bigger question: should this Jumbo system move to lithium? A lithium (typically LiFePO4) bank carries its own battery-management system that balances the cells internally, sidestepping much of the series-mismatch problem that plagues lead-acid strings. The trade-offs are real: lithium offers far longer cycle life (often thousands of cycles versus hundreds for tubular) and a much smaller, lighter footprint — but at a noticeably higher upfront cost. Treat specific cycle-life and warranty numbers as things to verify for the exact pack you are considering. Our lithium-versus-tubular guide goes deeper on that comparison.

One caveat you must not skip: not every existing Jumbo inverter can run a lithium pack. The pack's voltage must match your inverter's DC bus, and the inverter ideally needs a charging profile suited to lithium (and, on some systems, communication with the pack's BMS). Forcing a lithium pack onto an inverter that only knows how to charge lead-acid is exactly how people damage an expensive battery. Confirm this with proper sizing and consulting before you buy; ask us to check whether your specific Jumbo inverter is lithium-ready first.

Whichever way you go, your old lead-acid bank has value, and getting rid of it responsibly matters. Under India's Battery Waste Management Rules, 2022, lead-acid must be routed back through authorised recycling — not dumped or sold to an informal scrap dealer. The scrap value of the old lead genuinely offsets the cost of the new bank, and we handle old-battery exchange and buyback across Delhi NCR so the lead goes to authorised recycling and the credit comes off your bill.

A few habits keep a series string healthy for its full life. Most cost nothing; avoiding deep discharge is the single biggest lever you have.

For the deeper how-to, we have dedicated guides on inverter-battery maintenance, on why a battery drains fast, and on how long power-backup equipment lasts. This article stays focused on the series-bank diagnosis and matched-replacement decision rather than repeating them.

To recap the logic: find the weak battery by comparison — starting with resting open-circuit voltage and confirming under load (add specific gravity for flooded tubular, or internal resistance for any type including sealed SMF/VRLA). Then decide by bank age and how many units have drifted. A very young bank with a clear warranty defect and near-identical siblings can take a single like-for-like swap. In every other typical case, a matched full-bank replacement — or a lithium upgrade if your inverter supports it — is the honest economic choice, because a lone new battery in an aged string gets throttled and the old units fail soon after. When you do replace, your old lead has buyback value that comes off the cost. And if every battery tests fine yet the backup is still short, the issue is sizing, not a dead battery. Before spending anything, insist on seeing the under-load reading of every battery in your bank, so the evidence — not a sales pitch — drives the decision.