VANILLA SOLAR

There comes a point with campervan solar where you stop asking, “Is the panel working?” and start asking something much more interesting:

That is where I am with Vanilla now.

Since the new PV Logic 200W thin rigid solar panel went onto the roof and started feeding the Victron SmartSolar MPPT 75/15, I’ve watched watts, volts, amps, charging stages, daily harvests, battery deficits, overnight loads and enough Victron screens to qualify for some sort of electrical counselling.

But now I’ve got something better than impressions.

I’ve got 24 genuinely useful full-day solar results.

And once the partial days are stripped away, the picture is remarkably clear.

Vanilla’s new solar system isn’t merely maintaining the leisure battery.

It has become one of the main engines of her off-grid capability.

I wanted this to be about full solar days only, because partial days can badly distort averages.

The SmartSolar export covers 27 August through 19 September. The 19 September entry was incomplete at only 90Wh, so I’ve excluded it from the full-day statistics.

That leaves 23 complete exported days from 27 August through 18 September.

Then we have 20 September, which finished on 310Wh.

That gives us:

The current day is excluded too, obviously.

Sunrise is not a full day’s solar harvest.

Across those 24 complete days, Vanilla’s roof produced:

6,930Wh

6.93kWh

Nearly seven kilowatt-hours of electricity generated by one flat 200W panel on the roof of a VW Transporter in Britain.

No generator.

No fuel.

No engine idling.

No campsite bollard.

Just daylight.

The 23 complete exported days account for 6.62kWh.

The completed 20 September day adds another 310Wh.

At a rough 12.5V equivalent, 6.93kWh represents around:

That does not mean 554Ah went directly into the battery.

Vanilla was using electricity while solar was generating it, charging voltage changes throughout the day and there are efficiency losses.

But as a way of visualising the scale, it’s remarkable.

The roof has generated the energy equivalent of more than five nominal 110Ah leisure batteries.

Across all 24 complete days:

The median is perhaps the more interesting number.

Half the days were at or above roughly 310Wh, while a handful of poor days pull the mathematical average down.

So if someone asked me what a normal late-August-to-September day has looked like for Vanilla, I’d now answer:

And that is far more useful than saying I have a 200W panel.

A panel rating tells you what hardware might produce under standard conditions.

Daily watt-hours tell you what you actually got.

| Date | Harvest | |---|---:| | 27 Aug | 120Wh | | 28 Aug | 370Wh | | 29 Aug | 130Wh | | 30 Aug | 400Wh | | 31 Aug | 230Wh | | 1 Sep | 340Wh | | 2 Sep | 160Wh | | 3 Sep | 200Wh | | 4 Sep | 320Wh | | 5 Sep | 310Wh | | 6 Sep | 310Wh | | 7 Sep | 450Wh | | 8 Sep | 140Wh | | 9 Sep | 320Wh | | 10 Sep | 340Wh | | 11 Sep | 410Wh | | 12 Sep | 200Wh | | 13 Sep | 320Wh | | 14 Sep | 250Wh | | 15 Sep | 300Wh | | 16 Sep | 400Wh | | 17 Sep | 300Wh | | 18 Sep | 300Wh | | 20 Sep | 310Wh |

That table tells the story better than any single peak number.

There are poor days.

There are fantastic days.

But there are remarkably few useless days.

Out of 24 full days:

That is exactly two-thirds of the sample.

Five days produced 350Wh or more.

Four reached 400Wh or more.

Only four failed to reach 200Wh.

The four 400Wh-plus days were:

That isn’t one lucky afternoon.

It’s a pattern.

The current champion remains:

A 450Wh day is equivalent to the panel running at its full 200W rating for about:

At around 12.5V, that is roughly:

Now compare that with Vanilla’s recent lighter overnight usage of around:

One excellent solar day can theoretically harvest the gross energy equivalent of roughly three such nights.

That is why the off-grid calculation has changed so dramatically.

The lowest full day in the dataset was:

on 27 August.

At 12.5V equivalent, that’s roughly:

And that number is interesting because one of the most recent clean overnight tests used around:

Would 120Wh automatically restore everything?

No.

Vanilla continues using energy during daylight and charging isn’t perfectly efficient.

But it means even the weakest complete solar day in this dataset still harvested a genuinely useful amount of energy.

Across the 23 complete exported days, the average of each day’s maximum PV power was about:

The median daily peak was around:

The record remains:

on 30 August.

That is about:

For a flat-mounted roof panel on a campervan, that’s a cracking real-world figure.

This is where solar gets interesting.

28 August produced:

11 September produced:

16 September produced:

All impressive.

Daily watt-hours tell you what happened across the day.

That’s why I now care more about:

than:

Peak watts are fun.

Daily energy is freedom.

The first seven complete days averaged around:

The latest seven complete days are closer to:

I am not suggesting the panel has somehow become better at being a solar panel.

Weather, cloud, daylight, load and starting battery condition all matter.

But it does show something useful.

The strong results have continued well into September.

This isn’t just an August sunshine story.

I think 300Wh has become Vanilla’s most useful benchmark.

At around 12.5V, that represents roughly:

before losses and daytime consumption.

Now place that between our observed overnight ranges.

A lighter night can be around:

A deliberate heavy Webasto/fridge/lights night can be around:

So a 300Wh solar day has ample gross energy to replace a light night.

It may not completely undo an especially heavy night once daytime loads and charge losses are included, but it can make a huge dent in it.

And that’s exactly what we’ve seen.

Some morning deficits have been substantial:

and roughly:

during one of the harder experiments.

But this is where the solar system started proving its worth.

On one day Vanilla went from around:

to:

while the panel generated roughly:

That’s around:

despite Vanilla continuing to consume electricity throughout the day.

That’s real-world performance.

The latest overnight cycle was much gentler.

At 19:22 Vanilla was at:

By 06:48 she was at:

That means roughly:

across about 11½ hours.

Average net draw:

That’s tiny compared with the energy available from a 300Wh day.

This is the point where the whole off-grid philosophy changes.

Instead of asking:

“How many nights until the battery runs out?”

I’m now asking:

And quite often recently, the answer has been yes.

The SmartSolar history also records how long the controller spends in each charging stage.

Across the 23 complete exported days, Vanilla reached:

So the controller reached Float on roughly:

But some of the best solar days never reached Float.

30 August produced 400Wh and spent 819 minutes in Bulk.

7 September produced 450Wh and spent 780 minutes in Bulk.

11 September produced 410Wh and spent 784 minutes in Bulk.

That sounds strange until you remember what Float actually means.

Float does not mean:

“Amazing sunshine today.”

It means:

“The battery has effectively finished charging.”

A battery that starts hungry can absorb hundreds of watt-hours without ever reaching Float.

A battery that starts almost full can reach Float on a mediocre solar day.

That distinction matters.

Across those 23 complete exported days, the controller logged roughly:

That makes perfect sense.

Vanilla is being used.

The fridge cycles.

The heater runs.

Lights go on.

The battery gets discharged.

Solar wakes up and spends much of the day trying to refill what was consumed.

That’s exactly what Bulk is for.

The SmartSolar export includes a Consumption(Wh) column, usually showing tiny numbers such as 0, 10 or 20Wh.

That is not Vanilla’s actual total consumption.

It relates to the MPPT’s dedicated load output.

The main campervan electrical loads are elsewhere in the system, which is why the SmartShunt is much more useful for understanding actual battery consumption.

Otherwise someone could look at:

and assume Vanilla used almost nothing.

The fridge would like a word.

The solar performance is overwhelmingly positive.

The battery itself remains the one unresolved part.

Across the 23 complete exported days, the median daily minimum battery voltage is around:

There are some particularly low recorded minima, including:

Those figures need context because the export does not tell us what load was present at the instant of the minimum.

A heater start or another heavy load can create significant voltage sag.

The more interesting readings are the mornings where we have seen:

while only a few tenths of an amp were flowing.

Those are the readings that make me think the current 110Ah lead-acid battery’s real usable capacity deserves a proper test.

There is an amusing irony here.

The better the solar gets at refilling the battery every day, the easier it becomes for a tired battery to keep doing its job.

The panel keeps rescuing it.

The Orion can hit it with large charging currents when driving.

EHU can refill it when necessary.

So even if the battery has lost a meaningful amount of its original 110Ah capacity, the rest of the system may be good enough that I barely notice.

That teaches another important lesson:

The Jackery 1000 started at around:

I plugged it into Vanilla via the Anderson connection.

At one point the roof was producing:

while the SmartShunt was still showing a small positive current at the leisure battery.

In effect:

A few hours later:

Then solar kept working.

Vanilla eventually reached:

and the roof finished the day at around:

That one experiment transformed how I think about the system.

Vanilla isn’t just storing electricity anymore.

She can move it around.

The available routes are now:

And the Jackery can then run the hungry 230V equipment.

That gives the system redundancy and flexibility.

No single battery has to do everything.

The solar results are impressive by themselves, but the touring environment adds a:

We’ve seen the Orion produce charging currents approaching its rated output.

One morning Vanilla began around:

After driving, the SmartShunt showed the battery climbing rapidly back towards full.

This means solar doesn’t have to solve every deficit alone.

If I drive, the Orion joins in.

If I stay put, the roof keeps harvesting.

If I hook up to EHU, another charging route appears.

That’s why the system is much stronger than the raw panel rating suggests.

This is where 24 days of solar data suddenly becomes incredibly practical.

I now know a recent normal overnight deficit can be around:

I know the median full solar day has been:

That is roughly:

before losses.

I know the roof has generated at least:

I know the Jackery can take the big 230V jobs.

I know the Orion can provide serious recovery if I drive.

That doesn’t guarantee infinite off-grid power.

But it removes a huge amount of guesswork.

Suppose the weather is rotten and I get only:

That is roughly:

If the previous night cost around 9Ah, even a poor solar day may get surprisingly close to keeping the system balanced.

If I’ve used 30Ah overnight because the Webasto has worked hard, then no, a 120Wh day will not erase it.

But the Jackery and Orion give me options.

The system bends rather than breaks.

If conditions are decent and the roof produces:

that equates to around:

A normal night becomes easy territory.

Even a fairly heavy night can be substantially recovered.

And if the leisure battery becomes full while useful daylight remains, that creates an opportunity to divert energy into a Jackery.

That is completely different from rationing every light because I’m frightened of flattening the battery.

Before the new solar installation, off-grid planning was mostly about capacity.

How much battery do I have?

How long will it last?

Should I fit lithium?

Those questions still matter.

But now the most important question is:

If I use 10Ah overnight and solar replaces more than that the next day, I’m winning.

If I use 30Ah, solar replaces 20Ah and the Orion adds the rest while driving, I’m balanced.

That is what these full days have taught me.

The 450Wh record is brilliant.

The 158W peak is brilliant.

The 6.93kWh cumulative total is brilliant.

But the figure I probably like most is:

Because that isn’t a record.

It is the middle of the dataset.

It tells me solar has become dependable rather than merely exciting.

And dependable electricity is far more useful.

One small line in the data deserves mentioning.

Every recorded SmartSolar error field across the export is:

That doesn’t prove every wire and battery in the van can never have a problem.

But as far as the solar controller is concerned, the history is pleasingly boring.

And boring is excellent electrical engineering.

| Metric | Result | |---|---:| | Full days analysed | 24 | | Total harvest | 6.93kWh | | Average full day | 289Wh | | Median full day | 310Wh | | Best full day | 450Wh | | Worst full day | 120Wh | | Days ≥300Wh | 16 / 24 | | Days ≥400Wh | 4 / 24 | | Peak PV | 158W | | Average exported daily peak | 112W | | Median exported daily peak | 120W | | Days reaching Absorption | 18 / 23 | | Days reaching Float | 14 / 23 | | Recorded controller errors | 0 |

That is enough data to stop calling the system promising.

Not for every season.

Not for December.

Not for a week of permanent cloud.

But proven for the real-world conditions actually tested through late August and September.

Winter will be fascinating.

Daylight will shorten.

Sun angle will fall.

Cloud will become more frequent.

A flat roof panel will spend less time near those lovely 130 to 150W peaks.

So I absolutely expect the numbers to decline.

But now I’ve got a baseline.

When December arrives, I won’t have to guess how much performance has disappeared.

I’ll be able to measure it.

The grid value of 6.93kWh isn’t the point.

The real value is not needing EHU.

Not having to move because the battery is getting low.

Being able to leave the fridge running.

Being able to wake up with a deficit and know daylight has a realistic chance of repairing it.

Being able to refill the Jackery.

Being able to head into a festival knowing what a normal night actually costs.

That is what this panel has bought.

Twenty-four complete solar days have now answered the original question.

The new panel isn’t simply working.

It is changing how Vanilla can be used.

I consume power.

The sun returns.

The panel harvests.

The battery recovers.

Then I do it again.

Three weeks ago I had a new solar panel.

Now I have nearly seven kilowatt-hours of evidence showing what that panel actually means.

NIX // VANILLA POWERED

☀️ 24 FULL DAYS • ⚡ 6.93kWh HARVESTED • 🔋 310Wh MEDIAN • 🚐 450Wh RECORD • 🎪 READY FOR THE FIELD