Aug 27, 2026Solar Knowledge

Why a Three-Phase Inverter Still Exports on One Phase When Zero Export Is Enabled

Learn why a three-phase inverter may still export on one phase when zero export is enabled, and how to check CTs, anti-backflow meter compatibility, and firmware.

zero export meter

Why One Phase Still Exports When Zero Export Is Enabled: Meter, CTs and Firmware

This comes up a lot during three-phase inverter commissioning.
Zero export is already turned on, and the inverter's screen shows total grid power sitting right around 0W. But when you check L1, L2, and L3 individually, one phase may still be quietly feeding power back to the grid.
The instinct for a lot of installers is to blame the zero export meter or the anti-backflow meter. In our experience, though, the meter is often only one part of the problem — the real answer usually lives somewhere in the inverter's control logic, the CT wiring, meter compatibility, or firmware, and it's worth walking through each of those before touching any hardware.

Total zero export and phase-by-phase zero export are not the same thing

This is the part that trips people up most often.
Take a system where L1 exports 300W, L2 imports 600W, and L3 exports 300W. Add those up and the total grid power comes out to 0W. If the inverter is only watching the combined three-phase figure, it may happily call that "zero export" — but a utility meter that reads each phase separately will still see L1 and L3 sending power out.
So the first thing worth confirming, before any troubleshooting starts, is what the site actually needs. Some projects only care about the combined total staying at zero; others require every single phase to individually stay out of the grid, which is a stricter and different requirement. If it's the latter, checking the total power reading isn't enough — L1, L2, and L3 each need to be verified on their own.

Don't rush to blame the meter

When one phase is exporting, the anti-backflow meter is usually the first suspect. That's understandable, but the meter only measures power and passes the data along — the inverter is what decides what to do with it and actually controls the output.
So rather than asking "is the meter broken," it's more useful to work through a few sharper questions:
  • Is the inverter reading the data correctly?
  • Is the CT phase sequence right, and is CT direction correct?
  • Are the CTs actually sitting at the main grid connection point?
  • Is the external meter compatible with this specific inverter?
  • Is the firmware fast enough to react to load changes?
A meter can measure perfectly and the system can still misbehave if any one of those other pieces is off.

Check the CT wiring before anything else

If the system relies on CTs for zero export control, this is usually where the real answer is hiding. CT1 needs to match L1, CT2 needs to match L2, and CT3 needs to match L3 — it sounds obvious, but a mismatched phase happens more often than people expect, and when it does, the inverter ends up working from the wrong data even though everything may look normal on the surface.
CT direction matters just as much. Reverse it, and the inverter can read import as export, or the other way around, which usually shows up as strange control behavior whenever the load shifts. Placement matters too — CTs need to sit at the main grid connection point, where they can see the full power flow between the whole system and the grid, not just on the inverter output cable or on one load branch. If a CT isn't seeing the full picture, there's no way for the inverter to control zero export properly, no matter how correct the rest of the setup is.

External meters need to be confirmed compatible, not just functional

An external zero export meter has its place, especially when CT cable runs are too long or the CT signal isn't reliable enough on its own. But "the meter works" and "the meter works with this inverter" are two different claims.
Eastron is a common example. Their three-phase meters can measure import and export on L1, L2, and L3 without issue — that's not where problems usually show up. The real question is whether the inverter can actually talk to that specific meter. With Deye inverters, a standard Eastron unit bought off the shelf may not communicate properly, because the Deye-supplied version often runs a specific protocol or configuration that a generic retail unit doesn't match.
The meter can still show correct numbers on its own display while the inverter reads it wrong, or doesn't read it at all — exactly the kind of mismatch that makes an installer suspect hardware failure when it's really a compatibility gap. We saw this play out in a recent Deye zero export case, where the final improvement came from a firmware update rather than a meter replacement.

Be careful with balancing functions

When a three-phase load is unbalanced, it's tempting to reach for a function like Asymmetric Phase Feeding as a quick fix. It's not really built for that, though — it's a power-balancing tool, not a direct solution for single-phase export issues, and turning it on early tends to muddy the troubleshooting rather than clarify it.
It's usually better to keep the initial test simple: set the inverter to Zero Export to CT, check the CT wiring and direction, confirm meter compatibility, check the firmware version, and only then start testing load changes phase by phase. Enabling several functions at once makes it much harder to tell which one actually fixed — or caused — the behavior you're seeing.

Firmware plays a bigger role than people expect

Zero export is a real-time control loop — the inverter has to read CT or meter data continuously and adjust output fast enough to keep pace with load changes. Older firmware that reacts slowly can let brief export spikes through right when a load suddenly shifts on one phase, even if every other part of the setup is correct.
If the inverter supports remote monitoring, it's worth keeping it online so the manufacturer can check the firmware version directly rather than guessing from a description of symptoms.

A troubleshooting sequence worth following

When a three-phase inverter is exporting on one phase despite zero export being enabled, this is roughly the order we'd work through:
  1. Confirm whether the requirement is total zero export or phase-by-phase zero export.
  1. Identify whether the system relies on CTs, an external meter, or both.
  1. If there's an external meter, confirm it's actually compatible with the inverter.
  1. If the meter's source is uncertain, disconnect it and test CT control alone first.
  1. Set the inverter to Zero Export to CT.
  1. Verify CT1, CT2, and CT3 line up correctly with L1, L2, and L3.
  1. Check CT direction.
  1. Confirm the CTs are installed at the main grid connection point.
  1. Leave unrelated balancing functions off during the initial test.
  1. Keep the inverter online so the manufacturer can check the firmware.
  1. Add load separately on each phase and watch whether the others still export.

The takeaway

A total reading near 0W doesn't guarantee every phase is clean — that's the single most common thing people miss. Getting to stable phase-by-phase zero export usually comes down to a handful of things working correctly at the same time: CT wiring, CT direction, CT placement, a genuinely compatible zero export meter, the right inverter settings, and firmware that's actually up to date.
The meter measures the power, the inverter decides what to do with it, and a total that looks like zero export doesn't necessarily mean every phase actually is.