Anyone else finding their MPPT controller wildly overestimating SOC in cold weather?

Got a Victron SmartSolar 100/30 paired with a 200Ah lithium (LiFePO4) from Fogstar. Running it in my Transit conversion and noticed over the last few weeks as the temperature's dropped that the controller reckons the battery is at 100% SOC ages before it actually is — or at least before the resting voltage would suggest it is.

Typical scenario: it's maybe 3°C in the van overnight, I wake up and the Victron app shows absorption finishing dead quick, sometimes in under 10 minutes, then flips to float. But the battery monitor (a Victron BMV-712) is showing maybe 87–88% SOC based on coulomb counting. The two just aren't agreeing with each other and it's making me second-guess whether the cells are actually getting a proper full charge.

I've got temperature compensation set on the MPPT but I'm wondering if I've got the coefficient wrong, or if there's something specific about LiFePO4 that means I shouldn't be using it at all — I've read conflicting things. Has anyone dialled this in properly with a similar setup, and is the BMV likely to be closer to the truth here? Keen to hear what settings others are running.

@Compo27 yes, seen this exact thing with my setup. The MPPT is reading voltage-based SOC and LiFePO4 has an incredibly flat discharge curve anyway — add cold temps into the mix and the resting voltage sits slightly higher than it would at 20°C, so the controller thinks you've got more in the tank than you actually have.

The fix that worked for me was integrating a Victron SmartShunt — coulomb counting is far more reliable than voltage-based estimation, especially in a vehicle where temps swing constantly. The MPPT SOC becomes almost irrelevant once you've got a proper shunt doing the maths.

Also worth checking your battery's low-temp cutoff — Fogstar cells don't like charging below 0°C and will throttle or disconnect, which can look weird in the app data too.

Just to add to what @Taffy62 is getting at - worth checking whether you've got the battery temperature sensor connected to the SmartSolar. Without it, the controller's flying blind on compensation. The Victron BMV-712 or a Smart Shunt paired via VE.Smart networking will give you proper coulomb counting for SOC rather than relying on voltage curves, which as you've noticed are pretty meaningless on LiFePO4 at low temps. The flat discharge curve is notoriously difficult to interpret voltage-wise even in summer - chuck cold into the mix and it's basically useless. Fogstar batteries are decent quality so it's almost certainly a monitoring issue rather than anything wrong with the cells themselves. Have a look in the VictronConnect app under VE.Smart Networking and see what sensors you've got talking to each other.

@RachelCooper replied:

Good shout from @48VNerd on the temperature sensor - that's often the missing piece. One thing I'd add specifically for LiFePO4: the flat discharge curve that makes them so brilliant in normal use becomes a real headache for voltage-based SOC estimation in cold weather, because the voltage drop from temperature compression makes everything look artificially higher than it actually is.

Worth connecting your SmartSolar to the Victron app and checking whether your absorption and float voltages are set correctly for LiFePO4 - a lot of units ship with AGM defaults still active. Fogstar's documentation should have the recommended charge parameters. Getting those dialled in properly won't fully solve cold-weather SOC drift, but it'll at least mean your charging behaviour is correct even when the SOC readout's telling porkies. 😄

Something worth adding to the excellent points above - have you set up the battery profile correctly in the Victron Connect app? LiFePO4 has a very flat discharge curve even at normal temps, but if your absorption and float voltages aren't dialled in precisely for Fogstar's recommended specs, the controller's voltage-based SOC calculation goes even further out the window when cold throws the chemistry off.

Also worth connecting a Victron BMV-712 or similar coulomb-counting battery monitor if you haven't already. The MPPT's SOC is really just a rough guess - a proper shunt-based monitor tracks actual amp-hours in and out regardless of temperature-induced voltage weirdness. Makes a huge difference for real-world accuracy. @48VNerd's point about the temp sensor still stands though - do that first as it's free!

Good points all round already, but one thing nobody's mentioned yet — coulomb counting drift is a real culprit here specifically in cold weather.

LiFePO4 internal resistance increases noticeably below ~10°C, which means the MPPT's voltage-based SOC estimation gets thrown off even before you factor in the charging efficiency losses. The flat discharge curve that makes LiFePO4 so appealing becomes even flatter and harder to read via voltage alone when cold.

In my Sprinter build I resolved most of this by setting a forced full-charge synchronisation every few days — essentially letting the Victron reach absorption fully and hold it, which resets the coulomb counter to a known 100% reference point. Without that periodic reset, drift compounds over weeks.

Worth checking your absorption duration settings in VictronConnect — if it's cutting out early due to low-temperature charge current limiting, the counter never properly resets.