NIX // VANILLA CLIMATE LAB #01 — Autumn Outside. 21 Degrees Inside.

There comes a point in campervan ownership when most sensible people look at the temperature outside, decide whether they need a coat and get on with their day.

I appear to have taken a different path.

I now have sensors recording Vanilla and the outside world minute by minute. I have the Victron SmartShunt watching the Fogstar battery. I have the SmartSolar recording what comes down from the roof. I have the Webasto keeping everything civilised.

Which means Vanilla is quietly turning herself into a rolling laboratory.

And I had one question I really wanted answered.

Not how warm does she feel.

Not whether I can sit inside without a coat.

Not whether somebody on a campervan Facebook group reckons their diesel heater is “absolutely roasting”.

I wanted numbers.

Lots of them.

Temperature. Humidity. Dew point. Battery state of charge. Amp-hours. Solar harvest. Overnight behaviour. Warm-up speed. Outside temperature. Inside temperature.

The lot.

This is NIX // VANILLA CLIMATE LAB #01.

And the answer is rather magnificent.

Eight days. Two climates.

The temperature experiment covers roughly 10 October to 18 October 2026.

The inside and outside sensors were recording essentially once every minute, giving me 11,492 paired observations where Vanilla and the outside world can be directly compared.

That is what makes this interesting.

Every time Vanilla says 21 degrees, I can ask what autumn was doing on the other side of the bodywork at almost exactly the same moment.

And when the Webasto settles into its work, the two traces begin behaving like they belong to entirely different places.

That last pair is worth reading again.

Outside reached 6.9°C.

Inside Vanilla never went below 20.4°C.

There is a van door between those two climates.

The thermal bubble

I have started calling the difference between the cabin temperature and the outside temperature the Vanilla Thermal Bubble.

It is simply:

If Vanilla is 21°C and outside is 10°C, the bubble is +11°C.

If Vanilla stays at 21°C while outside falls to 7°C, the bubble grows to +14°C.

That is exactly what happened.

The wonderful part isn't simply that the Webasto can make the van hot.

Making a small enclosed space hot is easy.

The clever bit is that Vanilla settles into a narrow comfort zone while the world outside continues changing.

Autumn gets colder.

Vanilla essentially carries on at twenty-one-and-a-bit.

That is what a good heating system should look like in the data.

The 7°C morning test

The best part of the whole dataset arrives on the morning of 17 October.

Anyone can make a campervan comfortable at bedtime.

The proper test is the morning.

Six or seven o'clock.

The air outside has been losing heat for hours. The glass has cooled. The body has cooled. Everything has had all night to leak warmth into the outside world.

This is when a mediocre heating setup starts revealing itself.

Except Vanilla doesn't.

At around the coldest part of that first overnight run, outside gets down to:

Inside, the lowest recorded temperature across the same main overnight period is:

And for most of the night she sits higher than that.

Around 21°C.

Twenty-one degrees is not “warm for a campervan”.

Twenty-one degrees is a normal indoor room temperature.

That's an important distinction.

I'm not sleeping in an adequately heated vehicle.

I'm creating a little indoor environment that happens to be parked on four wheels.

And it barely moves

This may actually be more impressive than the temperature itself.

During the first night, Vanilla's internal temperature variation was tiny.

The standard deviation was only around half a degree Celsius.

The second main overnight period was similarly stable.

That means the graph isn't bouncing between seventeen and twenty-five degrees while the heater repeatedly cooks the cabin and lets it cool.

It settles.

The temperature trace looks almost suspiciously calm.

Outside is wandering around doing October things.

Inside is drawing a line across the graph.

Warm-up: from chilly to civilised

Maintaining warmth is one job.

Recovering a cold van is another.

A particularly clear event on 16 October gives us a nice look at that too.

Vanilla was down around 16.9°C.

Then the cabin starts climbing.

Twenty-two minutes.

That is the bit that matters when you actually use the van.

You've been out with the camera.

You've been walking.

You've been riding Torque.

Or, much more scientifically, you've been in a pub.

You get back into a slightly chilly Vanilla, fire the heater up, and within about twenty minutes the place has crossed into proper room-temperature territory.

No coat required.

No sleeping bag wrapped around your shoulders.

No heroic tales about how camping is supposed to be cold.

I have a diesel heater.

I'm using it.

The Webasto bit

The heater is central to this story, even though the data files don't contain a direct burner ON/OFF channel.

What they do contain is the heater's fingerprint.

Once sustained heating takes over, Vanilla moves into that extraordinarily narrow 20–23°C band while outside continues to move independently.

▶ Play film

Webasto Technical Services: Air Top 2000 STC installation overview.

The thermodynamics are beautifully simple.

Vanilla continuously loses heat through the windows, doors, roof, floor, body structure and ventilation.

The colder it gets outside, the greater that temperature difference becomes and the faster heat wants to escape.

The heater keeps replacing it.

Eventually the two rates balance.

Heat going in equals heat leaking out.

Cabin temperature stabilises.

And my graphs become remarkably boring in the best possible way.

Warm isn't enough. What about damp?

Anyone who has spent time in a campervan knows there is another enemy lurking in the background.

Moisture.

A warm van that feels clammy isn't properly cosy.

Human beings breathe.

Wet coats come inside.

Food gets cooked.

Tea gets made.

All of that moisture has to go somewhere.

And this is where the sensor data becomes especially reassuring.

Across the main heated periods, Vanilla averaged about:

Outside was considerably damper.

So inside we have roughly:

That is why Vanilla doesn't merely measure warm.

She feels comfortable.

The deeply geeky moisture bit

Relative humidity can be slightly deceptive because warm air can hold considerably more water vapour than cold air.

So the absolute amount of water in Vanilla can actually be greater than outside while the relative humidity is lower.

Which is exactly what the measurements show.

During the main heated-night periods, the inside absolute humidity was roughly around 9.4 grams per cubic metre, compared with about 6.3 g/m³ outside.

More actual moisture.

Lower relative humidity.

Temperature explains the apparent contradiction.

Warm air has much more capacity for water vapour.

This is one reason warming the cabin makes it feel substantially drier and more comfortable even when nobody has magically removed every molecule of water.

Dew point: where the condensation monster lives

Then there is dew point.

During the heated periods, the internal dew point sits at roughly 10–11°C.

That means any internal surface colder than that could potentially collect condensation.

The obvious suspects are:

• windscreen
• cab side glass
• door frames
• exposed metalwork
• thermal bridges
• cold corners

And this gives Climate Lab its next rabbit hole.

Air temperature tells us how warm Vanilla feels.

Surface sensors would tell us where the heat is actually escaping.

Stick probes on the windscreen, floor, sliding door, roof and bed area and suddenly we can start mapping the thermal anatomy of the van.

Which sounds excessive.

So obviously I want to do it.

Then I added the SmartShunt data

And this is where Climate Lab suddenly stopped being a temperature experiment.

It became an energy experiment.

The Fogstar battery is being watched by the Victron SmartShunt, and the export gives me hourly battery movement and state of charge.

That means I can align the overnight comfort data with what is happening electrically.

▶ Play film

Victron Energy: SmartShunt installation and monitoring overview.

One caveat is important here.

They are not a dedicated Webasto fuel or electricity meter.

The figures include the combined effect of whatever was connected to the Fogstar during those hours: heater electronics and fan, fridge, lighting and other 12V loads, minus any charging that may have occurred.

So I am deliberately not going to say:

“The Webasto used 26 Ah.”

That would be pretending the dataset knows something it doesn't.

What I can say is much more useful:

Comfort stayed flat. Battery paid the bill.

The first clean overnight power window runs approximately from 19:00 on 16 October to 07:00 on 17 October.

During it:

Using the Fogstar's nominal 12.8V voltage simply as a convenient energy conversion:

So while Vanilla is sitting there maintaining roughly 21.4°C against an outside average below ten degrees, the battery trace gently walks downhill.

Exactly what you would expect.

Temperature flat.

State of charge falling.

Comfort being purchased with stored energy.

The cost of cosy, hour by hour

This is probably my favourite new graph.

The temperature lines are almost boring.

Inside: steady.

Outside: cold.

Underneath them sit the hourly SmartShunt movements.

Little orange bites being taken out of the battery.

And suddenly “cosy” has an electrical shape.

The fascinating thing is that the van doesn't need some gigantic domestic-sized energy store to achieve this.

The main leisure battery is a 105Ah Fogstar LiFePO₄, roughly 1.34kWh nominal.

On that first clean overnight window, net depletion is about 26Ah.

That is substantial enough to measure clearly, but nowhere near catastrophic.

There is still useful battery left in the morning.

Which is rather the point.

Then it did almost the same thing again

This is where the data starts making me grin.

The next overnight selection records:

Night one:

26.23 Ah.

Night two:

26.45 Ah.

Difference:

Now, the selected windows are not identical in duration, so this is not a laboratory-grade appliance efficiency comparison.

Nor does it isolate the heater from the fridge and other 12V loads.

But the repeatability is still striking.

Two cold-weather overnight windows.

Two cabin averages around twenty-one-and-a-half degrees.

Two total net battery movements around twenty-six-and-a-bit amp-hours.

For real-world campervan data, that is a remarkably coherent story.

So what does cosy cost?

At this point I can give a sensible real-world answer.

I think that is a much more useful way of looking at it anyway.

I don't sleep inside a Webasto.

I sleep inside Vanilla.

What I really care about is:

And now I have the beginning of an answer.

Then the sun clocks in for the day shift

The roof has something to say about all this as well.

Vanilla's 200W solar panel feeds the Victron SmartSolar MPPT.

The SmartSolar history for the same period gives us the daytime half of the equation.

On 16 October:

On 17 October:

Across the two days:

Now, I need to be careful with the phrase “the sun paid it back”.

The 480Wh does not map directly onto one particular overnight discharge.

Solar occurs at different times. The van has other loads. Battery charging isn't perfectly lossless. Some energy may be consumed as it arrives rather than stored first.

But as an energy-budget comparison it is extremely useful.

A roughly 337Wh overnight net depletion is sitting in the same broad order of magnitude as a decent autumn solar day's harvest.

That changes how I think about cold-weather off-grid camping.

The real off-grid cycle

The old way I subconsciously thought about a campervan battery was:

That is far too simplistic now.

Vanilla behaves much more like a tiny energy system.

The actual cycle is:

Then add driving and the Orion DC-DC charger.

Then add the Blue Smart charger when deliberately charging from AC.

Then add the separate Jackery banks.

And suddenly I've stopped thinking of Vanilla as “a van with a leisure battery”.

She is a small mobile microgrid.

The best part: none of this felt marginal

This is probably the biggest lesson from all four datasets.

The temperatures don't show Vanilla struggling.

The cabin doesn't spend the night sliding slowly towards sixteen degrees.

The humidity doesn't disappear into some clammy wilderness.

The Fogstar doesn't wake up gasping on five percent.

The heating experiment feels controlled.

At the outside temperatures measured here, autumn isn't particularly testing the heater's ability to maintain comfort.

The interesting constraint is energy management rather than heating capability.

And even there, the numbers are encouraging.

The pub-stop equation

This is where the spreadsheets turn back into campervanning.

Picture the scene.

Late October.

Dark early.

Wet tarmac.

Country pub.

Vanilla sitting quietly in the car park.

Outside temperature around seven or eight degrees.

I come back after the pub.

Sliding door shut.

Webasto running.

Within twenty-ish minutes the cabin is around twenty-one degrees.

Humidity around fifty percent.

Fridge ticking away.

Lights on.

Maybe Vanilla's Cinema fired up.

Outside is miserable enough to make the whole thing feel worthwhile.

Inside is basically a tiny warm room.

By morning I've used roughly a quarter-ish of the Fogstar's nominal amp-hour capacity across the sort of overnight windows measured here.

And if the following day gives the solar panel something useful to work with, the recovery cycle has already started.

That is an entirely different psychological proposition from sitting there thinking:

“How long until the battery dies?”

The coldest moment is actually the best moment

There is something wonderfully satisfying about the point where the outside graph gets closest to seven degrees.

Because that's where the whole Vanilla project makes sense.

Outside:

Inside:

Battery:

Doing its job.

Heater:

Doing its job.

Solar:

Waiting for daylight.

Me:

Probably asleep.

That is exactly how the system should work.

But Climate Lab has only just started

This first experiment has accidentally created something much bigger.

Because now I have a baseline.

I know roughly what Vanilla does around seven to twelve degrees outside.

The next cold spell becomes another experiment.

Then another.

Eventually we can build proper temperature bands.

And because SmartShunt and SmartSolar are now part of Climate Lab, every future test can have an energy dimension as well.

I want Vanilla's thermal fingerprint

Every campervan must have one.

A characteristic thermal response created by its:

• insulation
• glass area
• floor
• roof
• furniture
• ventilation
• thermal bridges
• heater
• occupants
• outside weather

Feed enough real-world data into Climate Lab and I should eventually be able to describe Vanilla's own thermal fingerprint.

How quickly she warms.

How quickly she cools.

How large a temperature difference she can comfortably maintain.

How much battery that costs.

How much solar she gets back.

How humidity behaves through the night.

Where condensation begins.

What happens when the temperature outside goes below freezing.

Next: instrument everything

I already know where this is going.

More sensors.

Obviously.

I'd like temperature measurements:

• near the bed
• in the cab
• at floor level
• near the roof
• against the windscreen
• at the sliding door

That would let me measure vertical temperature stratification and surface temperatures.

It would answer questions I've previously only been able to guess at.

Is the cab the biggest heat leak?

How cold does the floor really get?

Does the bed area stay warmer than the cab?

What difference does a thermal curtain make?

What happens if the pop-top is raised?

How much heat disappears when I open the sliding door for thirty seconds?

How long does recovery take afterwards?

I can see this getting completely out of hand.

Excellent.

The NIX Cosiness Index

And because apparently it isn't enough simply to collect thousands of measurements, Climate Lab now needs its own score.

The NIX Cosiness Index considers:

• temperature
• temperature stability
• humidity
• warm-up performance
• thermal separation from outside
• energy sustainability

I'm keeping three tenths in reserve.

Winter hasn't arrived yet.

The ten numbers I'll remember

What did I actually learn?

Before doing this, cold-weather campervan use contained a little collection of vague questions.

Will the heater keep up?

Will the inside feel damp?

Will the battery disappear overnight?

Will I wake up cold?

How much difference does the solar actually make?

The data hasn't answered every possible winter question.

But it has answered the autumn ones very convincingly.

The heater holds roughly room temperature.

The humidity stays pleasant.

The temperature hardly moves.

The Fogstar takes a predictable hit.

The solar begins the recovery when daylight returns.

And instead of spending the night staring nervously at a battery percentage, I can actually use the van.

Final verdict

So there we are.

Outside can be seven degrees, dark and damp.

Inside Vanilla can be twenty-one-and-a-half, around fifty percent humidity, lights on, fridge running and the Fogstar quietly supplying the night shift.

Then morning arrives and the solar panel clocks back in.

That, for me, is the really exciting part of this experiment.

It isn't simply proof that my heater works.

It's proof that all the separate bits I've added to Vanilla are beginning to behave like one system.

Heat.

Battery.

Solar.

Monitoring.

Comfort.

One little mobile habitat.

NIX // VANILLA CLIMATE LAB
Measure it. Heat it. Sleep in it. Download the CSV in the morning.

Climate Lab #01 • October 2026 • Vanilla VW T6.1 • Webasto • Fogstar LiFePO₄ • Victron SmartShunt • Victron SmartSolar