Reading 3 CT's and getting single CT result -- Possible?


We want to add battery storage. We have a problem in that we have a three phase supply.

We already have a single phase PV system feeding just one phase. This works without issue as our meter correctly sums import and export. We propose adding a single phase battery to one of the other phases. The problem is that apart from Tesla who can monitor all 3 phases with 3 CT’s on their single phase battery system all the other manufacturers seem to have a single CT

2 suppliers have advised that they have users who have a single phase battery on the same phase as the solar.

However I am not happy with that idea as our single phase usage is split across all three phases. We could have the distribution box rewired so that every thing is on the same phase as PV and battery but that would be expensive.

Our favoured system is the Lux AC with Pylontech batteries. This is much cheaper than Tesla and there are a number of Octopus Agile customers who are very pleased with the same setup.

The Lux has 2 CT’s. One monitors the PV generation and the second the mains supply.

I am thinking on the lines of somehow monitoring the three phases and producing a current output that is the sum of all three and feed that into the Lux.

Am I correct that if it were physically possible and I could have a very large CT with all 3 incoming cables passing through the result be what I am after? IE 2 phases importing and one exporting I would see the correct result.

Alternatively I did wonder if Robin Emleys 3 phase solar diverter could be used as a basis of what I can best describe as a single phase emulator using 3 CT’s on the individual incoming cables.

Any input much appreciated.

Does this mean that it’s a 3-phase meter and it records the nett energy flow across all three phases - as it should?

No, that will give the vector sum of the three currents, which is the same as the neutral current, which, if everything is accurately balanced, will be zero.

I’m still not totally clear what the problem you’re trying to solve is. Are you happy to use the grid as a means of transferring energy exported from the battery to import it on a different phase? That appears to conflict with

but it seems to be what you’re saying or implying here:

That would (if it works) enable the battery’s control system to respond to and balance your nett load across the three phases. The problem I see there is you need to measure the nett power in three phases - not a real problem, but the hard part will be synthesising a signal that looks like a current transformer as far as the battery’s inverter is concerned. That could be tricky, partly because it could be hard getting any sensible information about what it needs.

The other tack could be to ask if you can have 3 inverters fed from the same battery?
Where you charge the battery from is a separate but related question - probably off the phase that receives the PV infeed.

Thank you for responding.

Yes that is correct.

They are never in balance. Our loads are nearly all single phase and despite the best efforts of our installers the use on each phase varies wildly during the day and on sunny days the PV phase is often exporting while the others are importing. What we need to know is when the nett energy flow is export so we can charge the battery.

Yes quite happy with that. I was struggling to put the possible problem I can see into words. And some of this is conjecture.The issue I am trying to deal with is the situation where the PV is generating and there is an export surplus of say 500 watts on the PV phase. My understanding is that if the battery inverter CT is on the same phase as the PV it would start to charge. It will not be aware we may be importing 2kW on the other 2 phases which is not what we really want. If the battery inverter CT was on a different phase to the PV and we had a nett export but the battery CT phase was actually importing it would not charge, again not a situation we want.

Tesla can provide 3 CT’s so they can monitor the nett energy flow on all three phases and charge their single phase battery accordingly. Unfortunately the Tesla is nearly twice the price of the system we are considering. Even more important is that Tesla do not have an open API control mode.

We are moving onto the Octopus Agile TOU tariff and that is what makes a battery system viable. We have been monitoring our use as if we were on the TOU tariff and we can reduce our energy bill by nearly 50%. As we also get capital allowance for this type of installation so the pay back is very short.

It is quite likely that the TOU tariff battery charging would be viable without the PV element but it does it seems a pity to add power to the grid without recompense when we could store it and use it ourselves.

Our PV is grid tied without feed in tariff and frequently in the summer we are a nett exporter. Even so it has been viable as we self installed and had access to a batch of pre-used very low cost panels. The major expense was the new Inverter which is SMA and even that we managed to buy at trade price.

Thinking a little more about the “black box” that you’d need, do you actually need to emulate a c.t., or is the battery charger/inverter capable of accepting instructions via some other means? That’s got to be something to check.

If you are stuck with a black box emulating a c.t., then for best accuracy it would measure current & voltage, hence power, on all three phases. You’d need to combine those into a single value of nett power, and then use that to control the amplitude and phase of a voltage derived from the phase that the battery charger thinks it should be monitoring, so that it emulates a current transformer on that phase. That’s a quite hardware-intensive project, and would involve a significant amount of research to implement.

I knew that would inevitably be the case - that’s why I wrote it in the terms I did. Even so, the neutral/vector sum current doesn’t mean anything useful to you.

I could tell that. You didn’t do too badly. :+1:

Thank you for the advice.

Just as a follow up. I have been monitoring energy usage for a few years using emoncms. I have in the past monitored nett 3 phase energy flow. Thanks to the power of emoncms I added an individual feed to each phase. Within days it showed me that most consumption now (wintertime) is on one phase. As luck would have it this is the phase my PV is on. The second most used phase is the supply to one particular outbuilding. The third hardly used phase is the office probably because our IT is nearly all on a 24V DC hybrid solar/mains system we installed a few years ago.

So following input from our electrician we are moving the outbuilding phase to the PV phase and treating the installation as a solar/battery single phase system to start with. We are going to leave all installation tails very long so that we can easily switch phases if need be in the future.