Anyone else using Victron's VRM portal to track seasonal patterns in a shepherd's hut setup?

Been logging data through VRM for about 14 months now on the hut and it's genuinely fascinating watching how the system behaves across the seasons. Running a 400W array into a Victron SmartSolar 100/30, feeding a pair of Fogstar Drift 100Ah lithium batteries. The difference between a decent February day and peak July is obviously expected, but what I didn't anticipate was how consistent the morning absorption curves became once I'd dialled in the charge profile properly.

What's caught my attention lately is the VRM "consumption" graph showing a baseline draw I can't fully account for — sitting around 8-10W overnight even with everything I'd consciously switch off. I suspect it's the SmartSolar itself plus the Cerbo GX ticking away, but the pattern has this odd little spike around 2am most nights that's got me puzzled.

Has anyone used the VRM advanced dashboard to correlate those kind of ghost loads with specific kit? I know you can overlay MPPT data and battery state, but I haven't quite cracked building a custom widget that makes the overnight picture really readable. Would love to know if others have built anything useful there, or whether I'm better off adding a dedicated energy monitor like a Shelly EM on the AC side just to get cleaner data.

@OllieRoss74 Similar story on the boat — VRM has been running for going on two years now and the seasonal swing is genuinely eye-opening. What really got me was comparing October to June side by side; the difference in peak absorption time is almost comical.

The thing I found most useful wasn't the daily graphs but the longer trend lines — you start to see your actual consumption habits creeping up in winter (more lighting, longer evenings) at exactly the same time your harvest is tanking.

On the boat I've got a 600W array and a Victron MultiPlus, and that crossover point in late November where demand overtakes generation capacity... it's a proper gut-punch moment every year, even when you know it's coming.

Really interested in this thread — I've got 14 months of VRM data on a static caravan setup and one thing I started doing recently was exporting the CSV data and plotting it in a spreadsheet alongside local weather station records. The correlation between prolonged low-pressure systems and battery state-of-charge drift is genuinely eye-opening.

What I'd suggest @OllieRoss74 is setting up custom widgets on your VRM dashboard tracking your yield vs consumption ratio week-on-week — gives you a much cleaner picture of where your margins are tightest. For me it's consistently the six-week window either side of the winter solstice where things get precarious, regardless of panel orientation adjustments.

Does your hut have any significant thermal load? I'm curious whether heating demands are masking what's actually a decent winter yield on your array.

@OllieRoss74 this is exactly the kind of thread I live for honestly.

One thing I'd add that nobody's mentioned yet — VRM's custom widgets are where it really gets useful for spotting patterns. I've set up a dashboard that shows battery state-of-charge at the same time each morning, which gives you a much cleaner picture of overnight load vs. previous day's harvest than the default views do.

On the narrowboat I can clearly see the point in late October where the system shifts from "effortlessly topped up" to "actually have to think about this" — it's almost the same calendar week each year.

Worth also cross-referencing your VRM data against pvgis.jrc.ec.europa.eu for your location. When my measured output diverges significantly from PVGIS predictions, that's usually my cue that panels need a clean.

Great thread @OllieRoss74. One thing worth doing with your 14 months of VRM data is pulling the state of charge at sunrise as a separate metric rather than just looking at daily yield. In a shepherd's hut context it tells you a lot about overnight load behaviour across the seasons — you'll likely notice winter SoC at dawn dropping quite consistently even on days when the previous afternoon looked reasonable. I cross-referenced mine against the Met Office historical radiation data for my grid square and it helped me distinguish genuine weather events from underlying consumption creep. Also worth setting up the VRM alarm thresholds seasonally rather than leaving them static year-round — what counts as a "low battery warning" in June is quite different to what you'd want flagged in December.

@CopperWarden that sunrise SOC metric is a solid one — I've been doing similar with my setup and it's almost a better health indicator than anything else VRM gives you.

One thing I haven't seen mentioned: if you export the raw CSV from VRM and plot it against Met Office historical radiation data for your grid square, you can see where your array is underperforming relative to available irradiance. Found a shading issue that way in early March that I'd completely missed just eyeballing the VRM dashboard. Panel output looked "fine" until I overlaid the expected curve.

Worth doing even once — takes an afternoon but the insight is genuinely useful for deciding whether to reposition panels or add capacity before next winter bites.

@T5Wanderer agreed on sunrise SOC being underrated as a health indicator — I'd add that if you plot it weekly rather than daily you filter out the noise from single overcast days and the seasonal curve becomes really clean.

@OllieRoss74 one thing VRM lets you do that I don't see many people using is the custom widget dashboard — you can overlay your battery voltage against ambient temp if you've got a BMV-712 with the temp sensor. Revealed a pretty clear correlation in my setup between cold nights and apparent SOC drop that wasn't actually real discharge, just voltage compression. Took me a while to realise my Fogstar cells were behaving normally and I'd been worrying for nothing.

Worth exporting the raw CSV occasionally too — VRM's own graphs are good but having the data locally means you can do your own analysis without being tied to their interface.

Great points all round. One thing I'd add to complement the sunrise SOC discussion — have a look at your absorption time duration across the seasons in VRM. In summer mine's often wrapping up absorption in under an hour, but come November/December it's dragging out to 3+ hours even on decent days. That creeping extension can be an early warning of battery degradation before you'd notice anything in SOC figures alone. @OllieRoss74 with 14 months of data you should have a really clean before/after comparison straddling two winters now, which is ideal for spotting any drift.