Trading with a plug-in battery: it works, just not where your salesperson thinks

A viewer, Willem, wrote under the previous video:

"I'm getting a bit dumb from salespeople who claim a plug-in battery can't trade, but I don't know if that's true."

That's the real problem. Not that the answer is complicated — it's not. The problem is that you're standing in a showroom, someone says something that sounds technical, and you have no way to check what's actually true.

So I looked into it. The video is the accessible version. This is the full version: all sources, the brand table, the backtest, and the part that didn't fit.

Key points

• "Trading" means three different things. One is straightforward, one is through your supplier, and one is genuinely impossible. Your salesperson uses the conclusion from the strictest case for the other two.
• The EMS is nowhere the bottleneck. Of the brands that matter, there isn't one you can't control. With Zendure and Anker, that's officially supported by the manufacturer itself.
• Your energy supplier is the bottleneck. Not the battery, not the controller. The question is whether someone will do it for you.
• There's community software that controls five brands simultaneously from a single Home Assistant installation. But it buys smart and doesn't sell — selling on your own initiative is listed as a bug report.
• "Charge when it's cheap" isn't wrong, it's too short. On 590 days of Dutch electricity prices, a fixed price threshold yielded 25 days where charging cost money. With a gate that first checks whether a peak follows: zero.
• That 800 watts is a setting, not a feature. Zendure ships its own device at 800 W and documents how to bump it to 2,400 W.

"Trading" is three things, and that's the whole argument

The confusion lives in one word that covers three shades of meaning. Put them side by side and the argument resolves itself.

When a salesperson says "a plug-in battery can't trade itself on the imbalance market," they're right. But that applies to every home battery, including the expensive fixed model they're trying to sell you. Your battery isn't the problem.

The hinge isn't whether trading happens, but who holds the knob. With arbitrage, you do. With a VPP, your supplier.

Even participating isn't trading — and that's what the trade press says, not me

How does a home battery get onto that market at all? Through an intermediary. An aggregator or energy supplier bundles thousands of batteries into one large flexible capacity — a virtual power plant, VPP. That party has the certification and trades on your behalf.

Trade journal Energeia wrote a piece about it with a title that leaves little room for doubt: "No, home battery owner, you don't trade on the imbalance market." Their point: even if you participate in such a program, you don't trade on that market. What actually happens is

"buying from and selling to the exclusive energy supplier at prevailing imbalance prices."

So you have a contract with one party, at prices tied to the imbalance market. You can't go to a better price elsewhere — for that, you'd have to switch suppliers.

So even the parties selling you "imbalance trading" stretch that word.

The claim is eight months old, and was already eleven months out of date when made

I looked into when that claim was actually made. That's the sequence the video shows.

The second row is the interesting one. That statement from April 2026 was, at the time of writing, already eleven months out of date: in May 2025, a plug-in battery was already doing it. The ZinVolt, on the imbalance market, via Frank Energie.

Not at 800 watts. At 2,000, and on its own circuit — precisely the condition this whole topic hinges on.

And that word in the Green Wave statement, "of the well-known brands," saves him. Barely.

The EMS isn't the bottleneck

Willem asked specifically about the difference in EMS — energy management system — between a fixed battery and a plug-in.

An EMS is a controller: software that, together with the hardware, decides when to charge and discharge. Both types of batteries have one. That's not where the difference lies.

What it's about instead is whether you get to participate.

For the last two: the fact that I couldn't find local control doesn't mean it doesn't exist. It means I didn't find it. If you know otherwise — let me know, and I'll correct this.

The conclusion is clear: of the brands that matter, there isn't one you can't control, and two are officially supported by the manufacturer.

Your energy supplier, yes

If it's not the battery and not the EMS, what is?

There's your detour. Not in the battery, not in the controller — in the question of whether someone will do it for you. If your supplier won't, you automate it yourself: Home Assistant, your brand's API, and you trade against your own contract's prices.

That's also what I do myself.

The software that buys smart and never sells

This is the part that got thirty seconds in the video, and it deserves more.

That self-automation is no longer a build project. There's community software — Omnibattery — that controls five brands simultaneously from a single Home Assistant installation: Marstek, Zendure, Anker Solarbank Max AC, Sessy (experimental) and Hoymiles. GPL-3.0, over a thousand commits, daily releases.

With Anker, you flip one switch in the app: third-party control. No installer, no firmware request.

But look at what that software does and doesn't do.

Does: a zero-export controller that drives your grid exchange to zero, and predictive grid charging with three modes — time slot, dynamic price, and real-time price.

Doesn't: selling. Searching sell to grid across the entire issue tracker yields zero hits. Selling on your own initiative is even listed as a bug report (issue #9: "Predictive Charging discharges to grid during PV surplus"). The only conscious discharge function exists to make room before negative prices — prevent damage, not sell on a peak.

⚠️ Fair to add: this project is a few weeks old, runs on daily beta releases, and the license says literally "as is, without warranty of any kind." Wrong Modbus settings can damage hardware. It's community software on a thousand-euro device.

"Charge when it's cheap" isn't wrong, it's too short

The advice is always the same. And it's true, until you recalculate.

A charge cycle costs yield. Electricity goes in, less comes out. On a Marstek Venus E, that's measured: 13.4 % loss, round trip — marginal yield 0.866, over thirty days with a correlation of 0.985.

So you charge on a cheap hour but no expensive peak follows, you've paid to move electricity you could have just bought directly.

And that's exactly what happens in winter. The price curve flattens: cheap hours stay cheap, but the peak behind them disappears.

The backtest that makes that case

This wasn't in the video, and it's the best evidence I've found.

The developer of Omnibattery built a minimum arbitrage margin gate: charge only if the price difference exceeds the round-trip loss of a cycle. His backing, literally:

"A fixed charge ceiling says whether a price is low, not whether it is low enough to be worth a cycle. In winter the curve flattens: cheap hours still sit under a 0.25 ceiling, but there is no spread behind them, so the battery cycles for less than the round-trip losses cost."

He tested that on 590 days of Dutch EPEX prices, all-in. Fixed threshold of €0.25 versus the spread gate:

His summary: "Better at every window length, no losing days. Winter margin roughly doubles on 29% fewer charge days."

What you take from this: a battery can lose money on a cheap hour. Ask any salesperson's yield figure which season it comes from.

What a cycle costs

There's another cost under that which appears on no yield sheet. A battery has a finite number of cycles, and every time you fill it to grab a few cents, you use up a piece of its life.

Ten thousand cycles is a lot. At one full cycle per day, that's over twenty-seven years — your warranty runs out before your cycles do. At three cycles per day, you're at nine.

That's not a prohibition. If you want to sell back to reduce your costs: fine. But factor in that wear, because then marginal trading looks different.

That 800 watts is a setting

The other argument against a plug-in battery is power: too little to earn anything from it.

Every plug-in battery in the Netherlands starts at 800 watts, and almost everyone thinks that's a feature of the device.

It's not.

The strongest evidence comes from a manufacturer. Zendure writes about its own SolarFlow 2400 AC: "a single SolarFlow 2400 AC unit has a default output power limit of 800W", and then documents how to bump it to 2,400 W. The 800 W is therefore not only demonstrably a setting — the manufacturer says so itself.

Why that limit exists is worth the effort, because it's not arbitrary. If your battery is on a circuit that also has other devices, the breaker in your consumer unit measures what flows through the grid. If your battery feeds power back onto that same circuit, that comes off the measurement. The wiring carries sixteen amperes, the breaker sees ten, and then it trips. Your battery makes its own protection blind — Anker calls that the cancellation effect.

On its own circuit, where only that battery sits, the breaker measures exactly what the battery does. No blinding. And then the full power comes off.

What didn't fit in the video

Two things are above and not in the video, because they broke the pace:

1. The 590-day backtest behind the spread gate. The video says cheap isn't the same as cheap enough; here are the numbers underneath.
2. The full Omnibattery story — which five brands, which three charge modes, and the evidence that the selling half isn't in it.

And two things I keep deliberately back, because they deserve their own piece rather than a footnote here:

• What your consumer unit and grid connection determine. One phase, three phases, 1×25A to 1×40A, internal net metering, and what it costs to upgrade. That's a completely different subject — it's about how much power your house can handle, not whether a plug-in battery can trade. That will be its own video and its own article.
• Whether this keeps working when everyone does it. A second viewer, Strandje, asked that question, and the answer isn't easy. The flex market saturates, and on 1 January 2027 the calculation changes once more. I have numbers for that that demand their own episode.

Frequently asked questions

The video

The accessible version of this story is on Jay's Desk: Handelen met een stekkerbatterij. Twelve minutes, with the brand table and the cancellation effect in view.

This article is the full version: the backtest, the sources, and the part that a video's pace can't handle.

Transparency

The Zendure units shown in the video I received as a review model during my time at Thuisbatterij Nederland. I have no current agreement with them, and Zendure doesn't know this video or article exists. No sponsor, no affiliate links, no brand preference.

The technical answers about the Smart Load Port of the Anker Solarbank Max AC come from Anker's product manager, at my request, and appear nowhere publicly documented. That's manufacturer claim, not independent measurement — and it's stated as such here.

Sources

All sources consulted on 7 August 2026. If you find an error, or know of local control for a brand listed here as "nothing found" — let me know, and this article will be updated.