ESS grid setpoint calculation error (not factoring in system inefficiency)?

Been noticing something similar with my Victron ESS setup and it's been quietly annoying me for months.

The system is consistently pulling slightly more from the grid than the setpoint I've configured — we're talking maybe 30-50W of "unexplained" draw even when solar and battery state look perfectly healthy. Not a disaster, but when you're trying to run genuinely close to zero net import, it adds up.

My working theory is that the ESS algorithm is calculating the setpoint based on ideal power flows, without properly accounting for inverter inefficiency and conversion losses in the chain. So when I ask for 0W grid setpoint, the system is essentially promising the grid "I'll cover everything" — but the maths doesn't factor in the ~5-8% that gets eaten by the Multiplus doing its job.

A few things I've checked already:

• MPPT and battery figures look accurate in VRM
• Fronius AC-coupled PV readings seem consistent
• ESS assistant is on the latest version
• Fogstar lithium bank reporting SOC correctly via CAN

The discrepancy is small but consistent, which makes me think it's a calculation quirk rather than a sensor issue.

Has anyone else spotted this, particularly on larger systems where the inefficiency losses become more noticeable in absolute watts? I'd be curious whether tweaking the grid setpoint to something like -30W (telling the system to slightly export) effectively acts as a workaround — compensating for what the algorithm isn't accounting for.

Would love to know if this rings any bells or whether I'm barking up the wrong tree entirely. Anyone dug into the ESS assistant code or raised it with Victron directly?

@SolarRachel yes, this is a known quirk that catches people out. The ESS setpoint targets net AC power at the grid meter, but it doesn't account for inverter losses in the calculation — so if you're set to 0W, the Multiplus is still drawing maybe 30-50W internally just to run itself, and that has to come from somewhere.

Check your Minimum System Losses setting in the ESS assistant. Also worth looking at whether you've got grid setpoint vs feed-in excess configured correctly in VEConfigure.

On my narrowboat setup I bumped the setpoint to around -50W which effectively compensates and keeps me genuinely neutral. Not elegant but it works.

The Victron community forum has a decent thread on this — search "ESS self-consumption offset" and you'll find proper nerdy detail on it.

@MrBodge65 has nailed the core issue. Worth adding — the gap gets noticeably worse in colder weather when your battery internal resistance rises, so if you're seeing more drift now than over summer that's likely why.

I went through exactly this with my shepherd's hut setup last winter. Had my setpoint at -50W (slight export) and was actually sitting at grid neutral most of the time. Adjusted the setpoint down to around -120W and it started behaving as expected.

The other thing worth checking is whether your MPPT and inverter losses are being measured on the same AC bus — if anything's on a different leg the ESS algorithm simply can't account for it properly. Victron's own documentation is a bit vague on this, took me a while to piece it together from the community forums.

Has anyone found a practical workaround for this in a van setup? I'm running a fairly modest system — Fogstar Drift 100Ah and a Multiplus-C 12/1600 — so the inefficiency gap is probably smaller than a residential install, but I've still noticed the numbers don't quite add up when I'm plugged into hookup at a campsite.

Is there a way to offset the setpoint to compensate, or is it more a case of just accepting the discrepancy? I've had a poke around in VE.Configure but couldn't find anything obvious. Would tweaking the AC input current limit be a reasonable approach or would that just create different problems downstream?

@Thommo9 for a van/modest setup the simplest fix is just offsetting your setpoint manually — if you want 0W grid import, try setting it to -50W or so. Accounts for the inverter conversion losses without getting complicated.

On the Cerbo/GX you can also use the ESS scheduled charging combined with a small negative setpoint to keep things tight. Not perfect but works well in practice.

For what it's worth on my garden office system (Victron Multiplus-II, Fogstar cells) I run a -30W setpoint and grid import is pretty much spot on. Took a bit of tweaking to find the sweet spot — worth logging actual grid draw for a few days before settling on a number rather than guessing.

The loss percentage shifts with load size too, so heavier loads need less offset proportionally.

@Thommo9 one thing nobody's mentioned — if you're running a Fogstar Drift, check whether your BMS is reporting accurate SoC under load. A dodgy SoC reading will compound the setpoint error considerably because the ESS algorithm is partly making decisions based on that figure.

Had exactly this on my static caravan setup last winter. Multiplus was hunting all over the place. Turned out the Drift's internal BMS was reading 3-4% optimistic under any meaningful discharge current. Calibrated it properly via a full cycle and the grid setpoint behaviour tightened up noticeably.

Not saying that's definitely your issue, but it's worth ruling out before you start messing with manual offsets.