Connecting batteries in parallel — do's and don'ts

Ah, the classic "let's just bung them together" approach — works until it spectacularly doesn't. Been there, nearly burned the shed down.

The actual dos:

Match your batteries properly. Same chemistry, voltage, capacity, and ideally the same age. Mixing a 10-year-old lead-acid with a shiny new one? That's asking for trouble. The knackered one will drag the good one down and charge unevenly.

Use proper cabling. Both positive and negative terminals, same gauge wire, kept as short and equal length as possible. Unequal lengths = unequal current distribution = one battery works harder and dies faster.

Install a BMS on each battery (or one per parallel string if you're running multiple groups). A Victron Smart BMS or similar means each battery gets monitored properly rather than hoping for the best.

Balance them before connecting. Get them all to the same voltage first. Connecting a fully charged battery to a half-dead one causes massive current spikes that'll make your connectors glow.

The don'ts:

Don't mix different capacities thinking they'll "share the load fairly" — they won't. Don't assume old cabling will do. Don't skip the BMS to save £50 when you're protecting thousands in batteries.

Most importantly: don't parallel different battery types. LiFePO4 and lead-acid in the same system? That's how you end up with an expensive heap of scrap.

Done it right for three years now without incident. Worth getting it proper from the start.

Spot on about the matching bit. I learned this the hard way with my shepherd's hut setup — tried pairing two different capacity LiFePO4s and the BMS went mental. The smaller one was constantly balancing, killing efficiency.

What actually matters:

• Same chemistry, voltage, and ideally same amp-hour rating
• Balanced charge/discharge curves or you'll get parasitic drain
• Quality interconnects — I use proper battery cables with proper lugs, not dodgy crimp connectors that'll fail when you need them most

The real gotcha is mixing old with new. Even if specs match on paper, aged batteries have different internal resistance. Your charger won't know whether to prioritise voltage or amp-hours, and something's getting fried.

If you've got mismatched kit, isolate them with independent circuits and separate charge controllers. Costs more upfront but saves the shed.

@ExBrickie94's right about the matching — can't stress it enough. I've seen the chaos when someone thinks "close enough" works with parallel banks.

The bit that catches people out though: even matched batteries will have slightly different internal resistances and voltages. That's why I always fit a battery isolator or diode per string on my narrowboat setup. Keeps one dodgy cell from dragging the whole parallel array down.

Also worth considering: if you're mixing older and newer batteries, the older one'll charge slower and discharge faster. You'll end up with a perpetual voltage imbalance that stresses your BMS. I learned that one the hard way.

What's your actual setup? Lithium, LiFePO₄, lead-acid? The rules shift a bit depending on chemistry.

Dead right about the matching. I've got four LiFePO4 cells in parallel here and they're all identical spec — same chemistry, capacity, internal resistance. Even then, I've got them on individual monitoring.

The bit that gets overlooked though: cable routing. Everyone focuses on matching cells but forgets that unequal cable lengths create voltage drops and current imbalances. Each connection should be dead equal in length and cross-section. I use the same gauge throughout.

Also worth noting — if you're mixing older with newer, the older one will drag performance down. Better to retire it properly than risk a thermal runaway situation. @ExBrickie94's shed-nearly-burning anecdote isn't uncommon.

BMS worth its salt should catch issues, but it's not a substitute for proper setup. Victron's monitoring kit helped me dial mine in properly.

Matching's crucial but don't sleep on the wiring either. I've got two 200Ah LiFePO4s in parallel on my static caravan and used proper 35mm² cable with equal lengths to each cell — even a few centimetres difference in cable length can cause current imbalances.

Also worth fitting individual breakers on each battery string if you can. Sounds overkill but means one dodgy cell won't tank the whole bank. Seen it go wrong where one battery started dropping voltage and dragged everything down with it.

BMS integration's the other bit — make sure they're talking to each other properly, not just assuming they'll sort it out themselves.

Spot on about the matching, but I'd add something I learned the hard way on my motorhome setup: balancing isn't just about specs, it's about state of charge (SOC) when you connect them.

I made the mistake of paralleling two identical 280Ah LiFePO4 cells when one was at 100% and the other at 40%. The BMS screamed, current flowed like a fire hose, and the wiring got properly warm. Took three hours to equalise them before safely connecting.

Now I always pre-charge to within 5% of each other using a separate charger first. Sounds faffy, but it's the difference between a smooth handshake and a violent current spike that stresses everything — especially the BMS contacts.

Also, use proper busbars if you're paralleling more than two units. Those dodgy crimped connections people bodge together are asking for trouble. A quality Victron busbar costs a tenner and could save your entire system.

@ExBrickie94's point about identical chemistry is dead right though. Don't even think about mixing LiFePO4 with lead-acid or different LiFePO4 brands.

Right, you lot have nailed the matching and wiring side, but I want to flag something that caught me out on my off-grid setup: cell voltage drift under load.

I've got two identical 48V 100Ah LiFePO4 banks paralleled through a Victron Lynx distributor, and even with matched cells, they'll drift slightly differently under heavy draw — especially if one's slightly warmer or has marginally different internal resistance. The BMS handles it, but that's precisely where people get complacent.

What saved me was fitting individual shunts to monitor each bank's current draw. You realise pretty quickly if one's shouldering more of the load. If the drift gets significant, you're accelerating degradation on the struggling unit.

Also — and this is daft but I've seen it — make sure your interconnecting cables are genuinely the same length and gauge. People bodge it with whatever's lying around, introduce voltage drops that exacerbate the drift.

The matching spec is non-negotiable, but active monitoring of how they're actually behaving in your system is what keeps you out of the

Has anyone actually run into issues with battery age mismatch specifically? I've got a pair of Victron LiFePO4s I'm looking to parallel up on my array setup, but one's about 18 months older than the other. They're both the same capacity and spec, but I'm wondering if degradation on the older one could cause problems — like, would it just charge/discharge at different rates and potentially damage the newer battery, or is that overthinking it?

Also, @IanHenderson79, when you mention wiring — are you talking gauge, or more about keeping the runs equal length? I've read conflicting stuff about whether unequal wire lengths actually matter in practice or if that's forum folklore at this point.

Cheers for flagging the balancing side, @LiFePO4Nerd. I'm assuming that's more critical with lead-acid or older chemistries?

Question on the balancing side — @LiFePO4Nerd, when you say balancing isn't just a formality, are you talking about active balancing during the parallel connection, or making sure the cells are balanced before you wire them together?

I've got two Fogstar 48V LiFePO4 units sat waiting in my shepherd's hut setup, and I've been putting off the parallel run because I'm not entirely confident about the pre-connection checks. The spec sheets mention cell voltage variance, but they're a bit woolly on what's actually acceptable tolerance before you should balance them out first.

Also curious whether people are running individual BMS units on each battery, or daisy-chaining them? I've seen folk do both and I'm trying to work out which gives you better protection against the catastrophic runaway scenario @ExBrickie94 mentioned.

Had a proper nightmare with this on my narrowboat a few years back. Paired up two ancient lead-acid banks that looked identical — same capacity, same age I reckoned. Turns out one was shot internally and the other was pulling all the charge, got so hot I could barely touch the cables. Nearly set the whole thing alight.

@ExBrickie94's spot on about matching, but I'd add: get the actual specs written down. Not "looks the same," not "the bloke at the marina reckoned they're similar." Voltage, chemistry, internal resistance if you can measure it. With LiFePO4 it's less forgiving — they'll happily overcharge a weak cell if the BMS isn't dialled in properly.

On the balancing front, it's not just passive top-up nonsense. If your cells drift, parallel banks will fight each other trying to equalise internally. Costs you capacity and lifespan. I switched to Victron kit specifically because you can actually see what each bank's doing.

The narrowboat runs four matched Renogy panels now with a proper Fogstar controller. Night

Matched my two Fogstar LiFePO4s in the shepherd's hut last year — same voltage, same capacity, same manufacturing date because I'm paranoid — and the BMS absolutely hates you when you don't bother. The balancing isn't just the battery being a bit moody, it's actual current flowing between cells at wildly different states of charge, which is basically asking for a thermal event you didn't budget for. @SmartSolarNerd, age mismatch is the sneaky killer because an older cell might be at 95% capacity and the new one at 100%, sounds trivial until you're running high discharge rates and the old one drops like a stone while the new one holds steady. Seen it turn a perfectly good battery bank into an expensive paperweight. Just accept you're replacing them as a set, not individually — worked out cheaper long-term than any drama.

I learned this the hard way in the motorhome, actually. Started with a single Victron LiFePO4 100Ah, then got ambitious and added a second one six months later thinking "job done, just wire them up."

The voltage matching bit is obvious enough, but what caught me out was the state of charge difference. My original battery had gone through maybe 300 cycles by then, whilst the new one was fresh. They spent the first week bickering with each other through the BMS — loads of current sloshing between them trying to equalise, which absolutely murdered efficiency and generated heat I didn't need in a confined space.

Should've fully cycled the new one on its own first, then brought them in at matched SOC. Now I treat it like introducing new livestock — you don't just chuck them in the field together.

@Jess1989's narrowboat saga doesn't surprise me one bit. Lead-acid especially will punish you for matching on appearance alone. The internal resistance degrades so differently between old and new that you might as well be connecting batteries from different centuries.

Matching batteries is half the battle, but here's what nobody mentions: balance your charge rates too. Stick dissimilar chargers on identical cells and you'll watch one age faster than the other — it's like feeding one teenager and starving the other, except your batteries actually explode.

In my motorhome setup, I learned this running a Victron MPPT with a dumb 240V charger on parallel banks — nightmare. Now everything feeds through a single Victron Multiplus and the difference is night and day.

@MarineVicky's got it right about matching dates; I'd add — check internal resistance if you can. Two cells that look the same at 12.8V might have wildly different impedance, and that's where parallels get spicy.

The real tell? If one battery's warming up noticeably faster during charge, you've already made a mistake. Don't learn this in a motorhome like I did.

The balance point @FormerCop raises is spot on, though I'd push it further — it's not just charge rates, it's the BMS handshake that matters.

I've got two Victron 48V setups on the boat (separate systems, different purposes) and learned the hard way that even matched batteries will drift if your charger doesn't communicate properly with both BMSs. If one pack sees 51.2V and cuts out while the other's still accepting charge, you've created a phantom load scenario that'll age one faster.

For parallel config, you genuinely need:

• Same chemistry (obviously)
• Same capacity and state of health (capacity degradation kills this)
• Individual BMS units that can actually talk to your charge controller

The garden office setup uses Fogstar 5kWh LiFePO4 — single unit, problem solved. But if I were paralleling, I'd be running them through a Victron Multiplus with CAN comms, not a dumb MPPT.

Unmatched parallel configs are why people end up with one battery doing 80

The charge rate point @FormerCop mentioned is crucial, but there's a practical layer most folk miss: cell voltage variance under load.

I ran two mismatched LiFePO4 banks in parallel on my cabin setup for about six months before the penny dropped. One was a Victron 48/5000, the other a Fogstar equivalent with slightly different internal resistance. Under heavy load (inverter draw), the voltage sag was different enough that one battery was doing most of the work whilst the other sat relatively idle. Over time, that stressed the harder-working unit's BMS.

What actually sorted it: I installed individual DC breakers on each battery string with a small series resistance (proper marine-grade isolator, not bodged with wire). Meant a tiny voltage drop, but it forced load-sharing proportionally to their charge state rather than their internal resistance characteristics.

The real safety consideration isn't just the BMS circuitry talking to itself—it's ensuring fault current paths are managed. Parallel batteries without proper isolation mean if one BMS trips, you've got current flowing through the other's protection circuits in ways they weren't designed for