Mystery Consumption in VRM - how to troubleshoot?

Had something eerily similar happen with my setup last winter — woke up to VRM showing a steady 800W draw at around 5am, and I was absolutely baffled. The shepherd's hut was locked up, nothing running, inverter barely ticking over.

Turns out it was a combination of things that took me a good few days to unpick:

1. The Victron MPPT was doing its cold-weather recalibration — apparently not unusual at low temperatures, though the numbers it throws into VRM can look alarming
2. A parasitic draw from the EV charge point controller — even in standby, mine was pulling a surprising amount. Easily missed if you're not logging at short intervals

The trick that cracked it for me was dropping the VRM logging interval right down to 1 minute during the investigation period. The pattern became obvious almost immediately — it was cycling, not constant, which pointed me away from a simple "something left on" scenario.

Also worth checking: if you're running a Multiplus or Quattro, the Advanced tab in VRM sometimes lags behind real-time values by a few minutes. Cross-referencing with the Cerbo GX's local console gave me a much cleaner picture.

Has anyone else found specific appliances that are serial offenders for this kind of phantom load? I'm half-convinced my Fogstar battery's BMS occasionally pulls weirdly during balancing cycles, but I've never managed to confirm that properly.

Would love to know what tools people are using beyond VRM itself — considering adding a dedicated clamp meter logger on each circuit just to nail these things down faster next time.

@DaleVicky 800W at 5am with nothing running? Either you've got a very ambitious ghost or your immersion heater decided to develop sentience and treat itself to a midnight bath 👻

Genuinely though — first thing I'd check is whether you've got any "vampire" loads hiding. My garden office was doing something similar and turned out to be

@DaleVicky that 5am draw rings a very specific bell for me. My shepherd's hut did something almost identical — turned out the Victron MPPT had kicked into a battery conditioning cycle and a small frost protection element I'd completely forgotten was wired into the circuit had triggered overnight. Two innocent culprits conspiring together.

Worth pulling your VRM consumption history for a few nights back — if it's recurring at roughly the same time, it's almost certainly something on a timer or a temperature threshold, rather than a fault as such.

Also check whether your Cerbo GX is logging individual loads if you've got current clamps fitted. Saved me hours of head-scratching when I finally bothered setting that up properly last February.

@DaleVicky had exactly this on my boat last November. Steady draw, nothing obviously on. Turned out to be a Victron MPPT that had developed a fault and was backfeeding through the array at night — not massive, but enough to show up weirdly on VRM.

Worth checking your DC loads separately from AC. People fixate on the inverter side and miss what's happening on the 12/24V rail entirely.

Also — is your VRM timestamp actually set to UK time or is it showing UTC? I've chased "mystery" overnight consumption before that was just my own afternoon kettle misattributed to 1am because the timezone was wrong. Sounds daft but it happens more than people admit.

What does the detailed view in VRM show for that period? Battery current vs inverter output should tell you immediately whether it's AC or DC side.

@DaleVicky this is giving me serious flashbacks to my shepherd's hut setup last February. Had a mystery draw at stupid o'clock that turned out to be... embarrassingly obvious in hindsight.

Few questions before diving in:

• Do you have a Cerbo GX logging device-level data, or just system-level?
• Any heating element connected to a timer you might've forgotten about?
• What does the draw look like — dead flat 800W or slightly variable?

A dead flat draw usually points to something resistive (immersion, panel heater). If it's slightly wobbly it's more likely a compressor or inverter doing something odd.

Can you pull the VRM graph for that period and share it? The shape of the curve is often more revealing than the number itself. Fogstar batteries will also show cell-level data if you've got comms set up — worth checking the BMS logs too.