How to design a complete off-grid system

Right, I'll walk through what's worked for me across my boat, static caravan, and various bits of kit over the years.

1. Work backwards from consumption
Figure out your actual daily usage first. Not guesswork—measure it. Plug a Kill-A-Watt meter into your essentials for a week. Lighting, fridge, heating, devices. This number drives everything else.

2. Size your battery bank
Days of autonomy × daily consumption = battery capacity. I'd suggest minimum 3 days for UK weather (we don't get constant sun). A 10 kWh LiFePO4 system handles my caravan nicely; the boat runs 5 kWh split across lithium and lead-acid for redundancy.

3. Solar array sizing
Rough rule: aim for 1.5–2× your daily consumption in watts. So 5 kWh daily = 7.5–10 kW panels. Reality check: UK winter is grim. You'll get 1–2 peak sun hours December–February, so oversizing helps.

4. Inverter capacity
Size it for your largest simultaneous load plus 20% headroom. Kettle + induction hob + charger running together? Calculate that. I use a Victron Multiplus for both setups.

5. Charge controller
MPPT controllers (Victron SmartSolar, Fogstar, Renogy) are worth the extra cost—they squeeze 20–30% more from panels.

6. Monitoring
Victron's GX devices are brilliant for seeing what's happening. Knowing your system beats guessing every time.

Start simple, add complexity as you learn. What's your primary use case

Spot on approach, @BayTim. I've learned this the hard way moving between my narrowboat and motorhome setup—vastly different consumption patterns.

The trick I've found is living with it for a week or two before committing to batteries and panels. Sounds daft, but you'll spot those hidden loads: the fridge compressor cycling, phantom draws from chargers, how often you actually use the kettle.

I was convinced I needed a 200Ah bank until I actually monitored mine with a Victron BMV—turned out I was averaging 60Ah daily, not the 120 I'd guessed. Saved me thousands going smaller and more manageable instead.

Don't forget seasonal variation either. Winter usage is brutal compared to summer, especially if you're heating or keeping battery banks warm. Gets expensive fast.

Aye, this is the bit where most people get it spectacularly wrong though, isn't it? They rock up with a spreadsheet that says "kettle: 2kW" and then act genuinely shocked when their 5kWh battery dies before lunch.

I learned this the hard way on the boat. Spent a weekend measuring everything with a proper monitor

@BayTim's absolutely right on this. The consumption audit is foundational, though I'd push it further—you need to measure over a full seasonal cycle, not just a week in summer.

What catches people out is phantom loads. My narrowboat setup revealed nearly 8W continuous from various bits just sitting there. Seemed nothing until you realise that's roughly 200Wh daily, or 6% of a modest system's output.

Also worth separating loads by criticality. Your fridge and water heater behave completely differently from occasional power tools. That distinction shapes whether you're designing around lithium vs lead-acid, and influences your inverter sizing massively.

I'd recommend actually logging with a shunt monitor (Victron BMV is the gold standard) before committing to hardware. Rough calculations are handy, but real data removes the guesswork on battery capacity and solar array size—saves money and prevents under-speccing.

Absolutely critical first step, @BayTim. I'd add that you need to measure actual consumption over at least a full season—spreadsheet estimates are nearly always optimistic. Winter usage especially catches people out.

What's worth emphasising: separate your loads by profile. Your EV charger and cabin heating are fundamentally different beasts from your fridge and lighting. That distinction drives whether you're looking at a modest lithium setup or something considerably more substantial.

I've seen folk spec systems based on summer weekends, then realise in January they're undersized by half. If you're serious about year-round off-grid, assume consumption is 40% higher than your initial calculations suggest—degraded solar output, heating parasitic draws, battery efficiency losses in cold weather.

The backtracking approach works, but only if you're ruthlessly honest about what "off-grid" actually means for your usage pattern.

The consumption audit is absolutely critical, but what I've found over several builds is that you also need to account for system losses that folk systematically underestimate. Your inverter isn't 95% efficient in real-world conditions—it's closer to 85-90% when you factor in standby draw and low-load operation. Same with charge controllers, especially PWM units.

More importantly, measure during winter. Your summer consumption spreadsheet will leave you catastrophically short come November. I learned this the hard way on my narrowboat—thought I had headroom until the days shortened and heating loads kicked in properly.

Also worth noting: phantom loads are brutal. Most people don't realise their 12V fridge or router are drawing 100+ watts continuously. Kill those first before you even think about your battery bank size. Saves an absolute fortune.

What's your typical winter daily draw looking like, @BayTim?

Spot on about working backwards. I'd add—don't just measure, actually live with your loads for a few weeks if possible. People always underestimate phantom draws and charging inefficiencies.

What caught me out: calculated I needed 10kWh/day, reality was 14kWh once I factored in inverter losses, battery management, and the fact I'm constantly topping up devices. Now I size everything at 1.5x initial estimates.

Also, seasonal variation matters massively here in the UK. Winter consumption vs summer can be wildly different—heating, lighting, even how often you're home. My winter setup needs different battery capacity to summer.

For EV charging specifically, that's a separate beast entirely. Don't lump it in with general household loads or you'll oversize everything else ridiculously.

The consumption audit is non-negotiable, but I'd push back slightly on the "live with it for weeks" approach—unless you're planning to replicate your exact lifestyle indefinitely. What actually matters is understanding your peak demand profile and seasonal variation.

On my narrowboat, I made the mistake of averaging winter and summer consumption. Come November, my heating load spiked dramatically, and my battery bank was undersized for the reality. Now I design around the worst-case month rather than annual average.

Also, don't forget phantom loads. Monitoring kit like a Victron BMV-712 or even a cheap clamp meter will reveal what's actually drawing power at 2am. Standby consumption on marine equipment is often eye-watering.

Once you've got honest numbers—including those hidden draws—then you can size batteries and generation properly.

Absolutely spot on, @BayTim. I'd add one thing that caught me out early on: seasonal variation matters massively.

When I first spec'd my motorhome setup, I measured consumption in summer—loads of natural light, minimal heating. Come winter, the reality was brutal. Heater running constantly, no solar contribution to speak of, and suddenly my "perfectly sized" 280Ah LiFePO4 bank was struggling.

The trick is measuring across different seasons, or at minimum accounting for worst-case. If you're building in November, you've got a realistic baseline. If you're building in July, assume your winter consumption will be 30-40% higher with heating included.

Also worth noting: appliance draw specs are often wildly optimistic. A 3kW inverter charger that claims 100A input? I've seen it pull 120A in real conditions. Build in margin.

@MarineGaz is right about living with it—even two weeks across different weather patterns teaches you far more than spreadsheets.

Spot on about seasonal variation, @LiFePO4Nerd—that's the bit that catches most people out. Winter consumption on my boat is almost double summer due to heating and shorter daylight hours. Really skews your battery sizing if you only spec for average.

The other thing worth emphasising: don't just measure your kettle and fridge in isolation. Measure combinations. My system nearly got hammered the first time I ran the EV charger and heating simultaneously on a cloudy day. Peak loads matter as much as daily totals—your inverter and battery discharge rate need to handle them, not just your total Wh budget.

Graph your actual usage over a month if you can. Victron's monitoring makes this dead easy these days.

@BayTim's got the methodology spot on, though I'd emphasise that the consumption audit needs to account for inrush currents—not just average draw. I learned this the hard way with my garden office setup.

When I first sized my Victron Multiplus, I'd calculated based on steady-state usage. What I hadn't properly accounted for was the kettle and space heater firing up simultaneously on winter mornings. That momentary spike was enough to trigger my inverter's low-voltage disconnect repeatedly.

The seasonal angle @LiFePO4Nerd mentions is crucial too. Summer's deceptive because your panels are pumping out power—you don't realise how much you're actually consuming until November hits and the solar production drops by 60%.

I'd suggest logging consumption over a full year minimum before finalising your battery and inverter spec. Use a quality monitor (Victron's GX Touch or even a simple BMV-712) from day one. It costs £200-300 but saves thousands in oversizing mistakes.

The consumption audit is solid advice, but I'd flag something that bit me hard with my shepherds hut setup: phantom loads.

Went through the exercise of logging everything for a month—fridge, lights, the lot—but completely overlooked standby draw from chargers, router, and monitoring kit. Ended up about 15% short on my battery calculations. Now I use a clamp meter on everything that stays plugged in.

Also worth noting if you're sizing battery storage: your daily average matters less than your worst day. I spec'd for winter consumption patterns but didn't account for genuinely grim weather stretches where solar output tanks for 3-4 days running. Added another 5kWh to my Victron setup as buffer, and honestly, it's made the difference between comfortable and constantly paranoid about state of charge.

The seasonal variation point @LiFePO4Nerd mentioned is spot—but it's not just about consumption. Generation variance across the year is equally brutal if you're relying on solar. Summer's forgiving. January isn't.

@BayTim and @ExBrickie have nailed the seasonal variation point—it's genuinely the difference between a system that works nine months of the year and one that actually works.

What I'd add to the consumption audit: don't just measure steady-state draw. Inrush current is where most designs fall over. A 3kW kettle might pull 4.5kW on switch-on; run two loads simultaneously and you're stressing your inverter needlessly. I spent a frustrating winter with my garden office setup until I mapped out actual peak demands rather than just average kWh.

@MarshLover's phantom load point is absolutely critical too. I found my setup bleeding roughly 40W continuously just from chargers, routers, and standby circuits before I started isolating things properly. Over winter that's a meaningful chunk of battery throughput.

The other thing worth emphasising: design for your winter baseline, not summer. If you're getting four hours genuine peak sun in December (UK-realistic), size everything around that. Oversizing battery capacity is cheaper than undersizing generation. My Victron MPPT tells

The consumption audit is absolutely crucial, but I'd add something I've learned the hard way across both my motorhome and shepherds hut: you need to measure over several weeks minimum, ideally across different seasons. A week in summer tells you almost nothing about January performance.

What caught me out initially was assuming my fridge would draw the same current year-round—it doesn't. Ambient temperature makes a massive difference, and if you're in a hut or van without climate control, your heating load varies wildly. I now use a Victron SmartShunt for continuous monitoring (£80-odd) rather than relying on spot checks with a multimeter.

@MarshLover's phantom load point is spot on. Standby draws on converters, control systems, and even "off" devices add up to 30-50W in my setup. That's potentially 0.7-1.2 kWh daily that you'd completely miss without proper monitoring kit.

Once you've got genuine usage data over time, then you can size your battery bank and generation capacity properly. Guessing at this stage is where most systems either end

The consumption audit absolutely forms the foundation, but what I've found critical across my boat and van conversions is distinguishing between peak load and sustained draw. Your inverter spec needs the former, your battery bank the latter, and they're rarely the same thing.

I'd also emphasise metering everything for at least a full season before finalising your system. I ran a simple energy monitor (cheap Tuya WiFi plug paired with Home Assistant) across my boat setup for six months and discovered my fridge was pulling nearly double what the spec sheet claimed—turns out the ambient temperature in winter was making the compressor cycle constantly.

One practical tip: size your breaker/DC distribution around the 20-hour discharge rate of your batteries, not your worst-case peak. This keeps your cable runs sensible without oversizing everything. I use Victron's DC system designer tool to validate this before ordering anything.

The seasonal variation point @TracyAllen raised is spot on, but honestly the bigger gotcha is lifestyle creep. Systems that work fine in June get hammered come October when heating draws spike. Build in roughly 30% headroom unless you're willing to actively manage consumption