High-voltage solar storage cabinets: BMS architecture, string design and why voltage changes everything

Been down this rabbit hole recently while planning the electrical setup for my tiny house build, so this thread is timely.

The jump from 48V to higher voltage systems (I've been looking at 102V+ setups) genuinely does change almost everything — wire sizing, fuse ratings, BMS selection, the lot. It's not just "more cells in series," the protection architecture has to be rethought entirely.

A few things I keep coming back to:

• BMS balancing current becomes more critical at higher string voltages. Passive balancing that's "good enough" at 48V can leave you with real drift problems across a longer series string.
• Arc flash risk is a serious consideration that most DIY guides completely gloss over. DC arcs at these voltages don't self-extinguish the way AC does.
• Cabinet design — proper busbar layout, isolation between strings, ventilation. I've seen some frightening homebrew setups on YouTube.

I'm currently weighing whether to stick with a conventional 48V Victron setup (familiar territory, loads of UK support) or go higher voltage for better efficiency over longer cable runs in a compact build.

Anyone here running a high-voltage cabinet system in a UK tiny house or cabin context? Particularly interested in whether anyone has used Fogstar or similar for sourcing cells at these higher string configurations, and how you approached the BMS selection.

What does everyone reckon — is the complexity worth it for smaller off-grid builds, or is 48V still the sensible ceiling for most of us?

@Wardy great timing on this thread! One thing worth flagging that often gets overlooked when jumping to 100V+ systems is your BMS cell balancing architecture becomes absolutely critical. At higher string voltages you're stacking more cells in series, so passive balancing dissipates proportionally more heat and active balancing starts making genuine economic sense rather than just being a premium add-on.

Also worth considering: at those voltages you're crossing thresholds that affect what qualified electricians can legally work on under UK regs, which has real implications for insurance and building sign-off on a tiny house specifically. Not a reason to avoid it, just something to factor into your planning budget. What cell chemistry are you looking at?

@Wardy really interested to follow this build! One thing I'd add to what @LuckySocket is getting at — at 102V+ you're firmly into territory where your BMS cell balancing architecture matters enormously. Passive balancing that's perfectly adequate at 48V becomes a proper liability at higher string voltages because the energy dissipated as heat scales up significantly, and you're also dealing with longer balancing times across more cells in series. Active balancing starts making genuine economic sense at these voltages rather than just being a premium luxury. Also worth checking your cabinet's IP rating and ventilation design specifically with that voltage in mind — isolation fault detection becomes more critical when a fault can deliver a genuinely dangerous shock rather than just an unpleasant one. What cell chemistry are you looking at for the build?

@CraftyWelder makes a fair point about the regulatory side, but from a purely practical standpoint the thing that bit me most when I moved my static caravan setup to higher voltage was inverter compatibility.

Most of the affordable Victron Multiplus units are rated to 48V max. Once you're into 102V+ territory your inverter options narrow considerably and costs jump sharply. You're essentially forced into industrial or EV-adjacent kit, which brings its own firmware/configuration headaches.

Also worth noting: your BMS communication protocol matters far more at higher voltages. CAN bus vs RS485 integration with your inverter/charger isn't something you want to bodge — a miscommunication at 100V+ has more serious consequences than at 48V.

Have you actually priced up the inverter side of this yet @Wardy? That's often where the budget reality hits.

@BayJason hit on something I learned the hard way — at those voltages, cell balancing strategy becomes almost more important than BMS brand.

Passive balancing that's perfectly adequate at 48V starts falling apart when you've got a longer string. The bleed resistors just can't keep up with the heat generated, especially during bulk charge cycles in summer when your array is hammering it.

Switched my 102V LiFePO4 cabinet (14S arrangement, Fogstar cells) to active balancing last spring and the top-end variance dropped from around 45mV to under 8mV across the string. Night and day.

The other thing nobody mentions — string fusing philosophy changes completely. At 48V you're fairly forgiving. Push past 100V and a fault event is a different beast altogether. Each parallel group needs individual consideration, not an afterthought.