**VANILLA POWERED // THE LITTLE GREY VAN THAT LEARNED TO MAKE HER OWN WEATHER**

105Ah of lithium. 200W of sunlight. 30A from the road. Diesel fire under the floor. Thousands of measurements. One increasingly unreasonable fascination with electrons.

There is a point in every campervan build where the van stops being a collection of things you have fitted and starts becoming a system.

Vanilla has just crossed that line.

She began as a 2022 VW Transporter T6.1. Then came the conversion, pop-top, fridge, Webasto, cooking gear, awnings and all the bits that gradually transformed a commercial vehicle into my little grey metal house.

But underneath all that, another Vanilla has been developing.

A Vanilla made from volts, amps, watts, watt-hours, diesel, photons, battery chemistry and an increasingly ridiculous quantity of data.

After weeks of experiments, overnight tests, solar logging, Victron screenshots and battery detective work, I finally understand what I've built.

Vanilla isn't simply a campervan with electrics anymore.

She is a tiny mobile energy ecosystem.

And this is how it happened.

For a long time Vanilla's electrical heart was a 100Ah Exide EFB leisure battery.

Around it, the rest of the electrical system had quietly become rather sophisticated.

On the roof sat a 200W PV Logic solar panel, feeding a Victron SmartSolar MPPT 75/15.

While driving, a Victron Orion-Tr Smart 30A DC-DC charger took care of alternator charging.

On hookup, a Victron Blue Smart IP65 12/10 provided mains charging.

And watching almost every electron entering and leaving the battery was a Victron SmartShunt 500A.

Alongside the fixed system sat my Jackery power stations for the hungry 230V jobs.

It was already a good setup.

Then the battery started whispering that something wasn't quite right.

11.95V.

11.77V.

Then one particularly ugly morning:

10.62V.

At barely any load.

Meanwhile, the SmartShunt still thought there was useful capacity remaining.

And that was the clue.

A SmartShunt is a brilliant accountant. It knows what has gone in and what has come out.

What it doesn't know is whether an ageing battery can still genuinely store the capacity written on its label.

The accountant said there was money in the account.

The cash machine disagreed.

The SmartShunt history told quite a story.

Deepest discharge: -84Ah

Cumulative draw: -3437Ah

Discharged energy: 41.9kWh

Charged energy: 47.4kWh

Low-voltage alarms: 54

There was even an ancient minimum-voltage reading of 4.38V, which I am treating as some historic interruption or disconnection because a functioning 12V campervan at 4.38V isn't having a bad day. It is furniture.

But the wider message was obvious.

The Exide had worked hard.

Thousands of amp-hours had passed through it.

The system around it had evolved.

Storage had become the bottleneck.

The solar panel was telling a completely different story.

Across one of my main datasets, full-day solar harvest averaged about 289Wh per day.

Median harvest was around 310Wh.

Plenty of days crossed 300Wh.

Then came Dubbed Out.

26 September 2026.

Vanilla harvested:

570Wh IN ONE DAY.

From a 200W panel.

That is equivalent to:

570Wh ÷ 200W = 2.85 hours

or 2 hours 51 minutes of theoretical full-power output spread across a real day of changing sun, cloud and panel conditions.

Watts tell me how hard the tap is running.

Watt-hours tell me how much water ended up in the bucket.

That day Vanilla filled the bucket.

At roughly 14.2V charging voltage:

570Wh ÷ 14.2V ≈ 40Ah

About forty amp-hours of battery-side energy equivalent.

Remember that number.

It becomes important.

I wanted to answer a deceptively simple question.

How much electricity does keeping Vanilla properly warm all night actually use?

Not what a brochure says.

Not what somebody on a forum reckons.

Not "it seemed fine this morning".

Actual Vanilla data.

So I left the Webasto enabled all night.

The 50-litre compressor fridge stayed running.

The normal 12V systems stayed alive.

I placed one environmental logger inside Vanilla and another matching logger outside.

The SmartShunt monitored the battery.

The SmartSolar watched the roof.

VictronConnect watched the Victrons.

And then I did something very important.

I went indoors.

I did not sleep in the van.

No body heat.

No breathing moisture into the cabin.

No doors opening.

No kettle.

No human interference.

Just Vanilla sitting outside maintaining her own environment.

For anyone who wants to disappear down the Webasto rabbit hole:

The environmental sensors recorded roughly every minute.

Temperature.

Relative humidity.

Absolute humidity.

Dew point.

VPD.

So this wasn't:

"It felt pretty warm when I opened the door."

It was thousands of timestamped observations showing exactly what was happening inside and outside the van.

And the result was superb.

Average internal temperature:

22.45°C

Minimum inside:

21.3°C

Maximum inside:

23.7°C

Average outside:

15.84°C

Minimum outside:

13.9°C

Average inside-to-outside advantage:

+6.59°C

Maximum recorded difference:

+9.6°C

But one of my favourite numbers was the internal temperature standard deviation:

about 0.67°C

Translation?

Vanilla was gloriously boring.

And boring heating is excellent heating.

She wasn't bouncing between freezer and sauna.

The Webasto warmed the cabin, backed off, allowed the temperature to drift gently and then recovered it.

It behaved more like domestic central heating than a diesel furnace bolted beneath a Volkswagen.

That was exactly what I wanted to see.

Average relative humidity inside Vanilla:

58.2%

Outside:

74.8%

But relative humidity changes with temperature, so I also looked at absolute humidity.

Inside:

11.6g/m³

Outside:

10.1g/m³

No tropical greenhouse.

No instant condensation apocalypse.

No warm mobile swamp.

But remember, I wasn't sleeping inside.

That makes this test even more useful because it gives me an empty-van baseline.

Next time I repeat the experiment while actually sleeping in Vanilla, I can compare the datasets and measure exactly what one human adds to temperature, humidity and condensation risk.

Some people go camping.

I apparently conduct thermodynamic field research.

At about 21:00:

At around 09:00:

Total overnight electrical use:

21.96Ah

Call it:

22Ah.

That figure includes the complete overnight electrical ecosystem.

Webasto.

Compressor fridge.

Background 12V equipment.

Warm van.

Cold fridge.

Everything alive.

About twelve hours.

Twenty-two amp-hours.

That is one of the most useful numbers I've ever measured from Vanilla.

Previous nights without the Webasto running gave me an ordinary fridge-and-background baseline of roughly:

10Ah overnight

So the additional demand associated with the Webasto appears to be somewhere around:

10–12Ah

That makes perfect sense.

The Webasto isn't using electricity to make the heat.

Diesel makes the heat.

Electricity powers the fan, fuel pump, electronics and controls.

That means staying warm doesn't necessarily require a battery the size of a wardrobe.

It requires a battery that can reliably deliver its rated energy.

And that is where everything changed.

Out came the Exide.

In went the new:

FOGSTAR DRIFT 105Ah LiFePO4

And this time I have the actual 2025 V2.2 Fogstar manual sitting in front of me.

The specification is delicious.

Nominal voltage: 12.8V

Capacity: 105Ah

Stored energy: 1280Wh

Cells: Grade A EVE LF105

Configuration: 4S1P

Continuous discharge: 100A

Peak discharge: 200A

Maximum charging current: 100A

Bluetooth BMS.

Internal heating.

Approximately 10kg.

Approximate cycle specification: 3500 cycles at 80% depth of discharge.

On paper, replacing a 100Ah battery with a 105Ah battery sounds almost laughably pointless.

Five extra amp-hours.

Hold the front page.

Except that's not what happened.

This isn't really a capacity upgrade.

It is a chemistry upgrade.

EFB to LiFePO4.

And that changes almost everything.

Lead-acid capacity and comfortably usable capacity are not the same thing.

As a lead-acid battery discharges, voltage progressively falls and deep cycling punishes longevity.

Lithium behaves very differently.

Its voltage remains much flatter through most of the discharge.

Much more of the nominal capacity is practically usable.

Charging efficiency is higher.

Deep cycling is far less punitive.

So the real change isn't:

100Ah → 105Ah

It is:

And because I already measured a genuine heated night, I can now apply real-world numbers to the new battery.

Measured warm-night electrical requirement:

22Ah

Fogstar capacity:

105Ah

22 ÷ 105 = 20.95%

So the same entire heated night represents approximately:

Leaving roughly:

One full night.

Webasto.

Fridge.

Background systems.

Wake up warm.

About four-fifths of nominal battery capacity still available.

Two identical nights would use around 44Ah, leaving roughly 58%.

Three would use about 66Ah, leaving around 37%.

And that assumes no electricity comes back in.

Vanilla has three different ways of making sure that doesn't happen.

200W PV Logic → Victron SmartSolar MPPT 75/15 → up to 15A → Fogstar

VW alternator → Victron Orion-Tr Smart 30A → Fogstar

→ Victron Blue Smart 12/10 → Fogstar

Three completely different energy sources.

One storage battery.

That is when a campervan electrical installation becomes an actual energy system.

Fogstar's guidance puts normal charging somewhere around 20–30% of battery capacity.

For 105Ah:

20% = 21A

30% = 31.5A

My Orion:

30A

Which is:

28.6% of battery capacity

Almost exactly where I would put it if designing everything from scratch.

Driving in sunshine gets even more interesting.

Orion:

30A

Solar:

potentially up to 15A

Combined theoretical maximum:

45A

The Fogstar can accept up to 100A.

So even with both systems delivering heavily, there is loads of battery headroom.

Parked on EHU in sunshine?

Blue Smart:

10A

Solar:

potentially 15A

Total:

25A

Which is around 23.8% of capacity.

Again, right in the sweet spot.

The actual Fogstar manual gives the Victron targets:

So the new Vanilla lithium PIN code is:

14.2 / 13.6 / 30

Equalisation: OFF.

Lead-acid temperature compensation: OFF.

Low-temperature charging protection is handled by the battery itself.

The whole Victron network is essentially being retrained from:

"Hello, lead-acid."

to:

"Welcome to lithium."

The Fogstar has internal heating pads.

But they do not simply switch on and drain stored battery energy whenever it's cold.

They work from incoming charging current.

The manual gives:

Minimum incoming current to activate heating: about 4A

Heater power: 36W

So imagine a freezing morning.

Battery cells too cold to charge safely.

Solar arrives.

Or I start driving.

Or connect EHU.

Incoming electricity reaches the battery.

The BMS recognises the cells are too cold.

The incoming power warms them first.

Then normal charging begins.

The battery basically warms its hands before breakfast.

Wonderful.

The old SmartShunt settings belonged to the Exide.

100Ah.

Peukert 1.25.

95% charge efficiency.

The lithium starting point becomes:

Capacity: 105Ah

Charged-voltage threshold: 14.0V

Tail current: 4%

Charged detection: 3 minutes

Peukert exponent: 1.05

Charge efficiency: 99%

Current threshold: 0.10A

Discharge floor: 0% for monitoring purposes

Peukert is wonderfully nerdy.

A value closer to 1.00 means usable battery capacity changes less as discharge current changes.

Lithium behaves much closer to that ideal than traditional lead-acid.

So a tiny field buried in VictronConnect is actually describing the chemistry of the battery beneath my seat.

I love that.

Vanilla still has the Sargent EC155.

It remains the habitation distribution and control system.

But the philosophy is much cleaner now.

The dedicated Victron equipment manages charging specifically for the Fogstar.

That makes the whole setup easier to understand, easier to monitor and much easier to diagnose.

The Fogstar isn't trying to become a giant mains power station.

Its job is the van.

Webasto.

Fridge.

Lighting.

USB.

12V systems.

Control equipment.

The Jackerys handle the hungry 230V equipment.

Induction cooking.

Microwave.

Other big loads.

That separation matters.

It means I haven't distorted a beautifully sensible 12V system around occasional power-hungry appliances.

But there is still an emergency bridge between the two systems.

Jackery AC output → Blue Smart charger → Fogstar

Efficient?

Not particularly.

Useful?

Absolutely.

Sometimes resilience matters more than efficiency.

This is probably my favourite part.

A heated night costs:

about 22Ah

A decent solar day produces:

around 300Wh

At 14.2V:

300Wh ÷ 14.2V = 21.1Ah

Read those again.

Warm night:

~22Ah

Decent solar day:

~21Ah equivalent

That is astonishingly close.

One decent day's solar harvest is in roughly the same gross-energy territory as one entire measured heated night.

And the record day?

570Wh

570Wh ÷ 14.2V ≈ 40Ah

Nearly two heated nights' worth of electrical consumption.

One excellent day of sky.

That is when "off-grid" suddenly stops meaning:

How enormous a battery can I fit?

and starts meaning:

Can I replace tomorrow what I use tonight?

That is a far more interesting question.

Night arrives.

Webasto keeps Vanilla warm.

Fridge keeps everything cold.

Lights glow.

Phones charge.

The Fogstar supplies the electricity.

Morning arrives.

The 200W solar panel wakes up.

If I drive, the 30A Orion joins in.

If I'm on EHU, the Blue Smart adds another charging route.

The system doesn't simply drain until empty.

It breathes.

That is Vanilla Power 2.0.

Now I have something incredibly useful:

A complete Exide-era Night-Lab baseline.

So the obvious thing is to repeat it with the Fogstar.

Same Webasto.

Same fridge.

Same temperature target.

Same inside logger.

Same outside logger.

Same SmartShunt.

Then compare:

Voltage stability.

SOC.

Amp-hours consumed.

Minimum voltage.

Morning reserve.

Solar recovery.

And then I want to repeat it again.

Except next time I'll actually sleep inside Vanilla.

Then the experiment becomes:

versus:

At which point I will possess detailed scientific evidence quantifying my own contribution to the thermal and humidity characteristics of a Volkswagen Transporter.

Some people own campervans.

I apparently peer-review mine.

I could keep going.

Peukert exponents.

Coulomb counting.

MPPT curves.

Absorption voltages.

Battery-management logic.

Watt-hours.

Amp-hours.

And probably will.

But none of that is really why I built this system.

The point is somewhere much simpler.

Park somewhere beautiful.

Rain starts hammering outside.

Turn the heater on.

Fridge stays cold.

Lights stay on.

No hookup cable.

No generator.

No wondering whether the battery will survive until morning.

Wake up warm.

Look at VictronConnect.

See that the sun has already started putting yesterday's electricity back.

Make coffee.

Drive somewhere else.

That's what the graphs buy.

Not numbers.

Freedom.

Freedom measured in amp-hours.

Warmth measured in watt-hours.

Adventure measured in how long I can stay before I need anything from anybody.

🔋 Fogstar Drift: 105Ah / 1280Wh

☀️ Roof solar: 200W

🌞 Solar charging: up to 15A

🚐 Driving charge: up to 30A

🔌 EHU charge: 10A

🔥 Measured heated-night demand: ~22Ah

❄️ Normal overnight baseline: ~10Ah

🔥 Estimated additional Webasto demand: ~10–12Ah

🌡️ Heated interior average: 22.45°C

🌡️ Outside average: 15.84°C

📈 Average temperature advantage: +6.59°C

🚀 Maximum temperature advantage: +9.6°C

💧 Interior RH average: 58.2%

☀️ Solar record: 570Wh

📊 Desire to leave anything unmeasured: 0%

The old Exide wasn't a failure.

It did its job.

It powered trips.

Ran the fridge.

Supported the heater.

Carried thousands of amp-hours through Vanilla's adventures.

The rest of the system simply evolved beyond it.

Solar improved.

Charging improved.

Monitoring improved.

My expectations improved.

Eventually the battery became the limiting component.

So I changed the limiting component.

Now I measure again.

That is the satisfying part.

This isn't lithium because lithium sounds fashionable.

It's data finding the weak link.

Engineering replacing it.

Adventure testing the result.

That is very NIX.

105Ah UNDER THE SEAT.

22Ah FOR A WARM NIGHT. 570Wh FROM ONE GLORIOUS DAY OF SKY.

Vanilla doesn't just carry the adventure anymore.