Victron 200Ah lithium showing 97% but dropping fast under load — BMS fault or something else?

Had exactly this with my Fogstar 200Ah last winter — turned out one cell was playing silly buggers and the BMS was reporting pack voltage rather than individual cell voltages, so the headline percentage looked rosy right up until it didn't.

Few things worth checking before you blame the BMS outright:

• Cell imbalance — if you've got a Victron BMV-712 or similar, it'll only show you pack-level data; you need something like a Bluetooth BMS app or a separate cell monitor to see if one rogue cell is dragging everything down
• Temperature — lithium hates the cold, and if your battery's in an unheated space right now it'll shed capacity faster than my motivation on a Monday morning
• High-resistance connection — a dodgy terminal can cause voltage sag under load that the BMS interprets as "oh crikey, we're empty"

What load are you pulling when it drops? And is the drop gradual or a sudden cliff edge? Cliff edge usually screams cell imbalance; gradual suggests capacity loss across the board — possibly from repeated partial charging cycles killing your top balance.

Mine recovered once I did a proper full charge cycle and let the BMS top-balance overnight. Worth trying before you start pointing fingers at expensive bits.

What's your setup — any chance it's a temperature issue if it's in a shed or garage?

@DaveMoore70 good shout on the cell-level monitoring point. Worth adding — on my boat setup I run a Victron SmartShunt alongside the BMS, and the two together tell a very different story than either alone. The SmartShunt tracks actual amp-hours consumed, so when it disagrees significantly with the BMS state-of-charge percentage, that's your first red flag that something's off internally.

If the OP has Bluetooth on the BMS, pull up the individual cell voltages under load rather than at rest. A weak cell will sag disproportionately — you might see 3.65V across five cells and 3.20V on one rogue. The BMS averages or uses pack voltage and masks it completely until the protection kicks in suddenly.

Also check your balance lead connections if it's a DIY pack — corroded connections read as artificially low cell voltage and trigger exactly this behaviour.

@DaveMoore70 @DougPearce — classic "lying dashboard" problem; worth checking if your Victron app is showing a suspiciously round coulomb-counted SOC vs the voltage-derived one, because when they diverge sharply under load that's usually your smoking gun for a weak cell dragging the pack down before the BMS finally throws its toys out.

Had identical grief on the narrowboat last February — turned out one cell had drifted so far out of balance that the BMS was cutting out at what looked like 97% but was actually one rogue cell already kissing its low-voltage threshold.

Run a full charge cycle, let it sit 30 minutes off load, then check individual cell voltages via Bluetooth if your BMS supports it — the spread between highest and lowest cell tells you everything.

Had this exact drama on my narrowboat last February — if the SmartShunt and the BMS are telling different stories, trust neither until you've pulled individual cell voltages via Bluetooth; my Fogstar's own app showed a rogue cell sitting 0.3V below the pack average which is basically your BMS screaming in a language the percentage bar doesn't speak.

Key things to check:

• Cell voltage spread — anything over ~0.05V delta under load is a red flag
• BMS temperature logs — cold cells lie worse than a dodgy car dealer
• Charge history — if it's never done a proper top-balance, one weak cell drags the whole pack down prematurely

Victron's VRM portal will show you the shunt data historically too, which helps spot if this is a new fault or a slow-burn problem you've been ignoring since summer. 🔍

@RetiredChef narrowboat in February, you brave soul — I had my Victron BMV-712 screaming at me mid-January in the motorhome like an angry smoke alarm while the BMS sat there looking perfectly smug at 97%.

Check your cell deviation in the BMS app — if one cell is sagging more than about 100mV under load you've found your culprit, and no amount of staring at the headline percentage will fix it.

Learnt this the hard way parked on a Welsh hillside with a failing Fogstar cell and a very cold flask of tea.

@DaveMoore70 good shout on the cell voltage issue. One thing worth adding — if you've got a Victron SmartShunt alongside the BMS, pop into the VictronConnect app and check the "history" tab. A rapid voltage collapse under load often leaves a telltale minimum voltage record that'll help pinpoint whether you're looking at a weak cell or a calibration drift on the state-of-charge readings. Also worth doing a proper capacity test when you get a chance — discharge at a consistent rate and log what you actually get out versus the rated 200Ah. If you're only pulling 140-150Ah before the BMS trips, that's your answer right there. @RetiredChef is spot on about trusting the SmartShunt over the BMS percentage display in most cases — the shunt is measuring real coulombs in and out rather than inferring from voltage.