Multi 'plant' operators (>1 AC Coupled 3.0)

I thought it might be useful for those Folk with multi battery / controller set-ups, to have a separate place to discuss issues & ideas - in the hope that this would prove valuable to new owners just starting out - as well as those planning or considering new systems.

I've recently had an installation of 2 X 8.2 batteries - each with it's own AC 3.0 Coupled.
The reason for 2 controllers is that I wanted to have plenty of available ESS, but also be able to run the 'Golden hour' as I call it, for the evening cooking session, with (as close as poss to) zero grid import. Bearing in mind that just switching on a kettle in the UK will practically gobble up the max output from one AC Coupled, I decided to go with one 8.2, each with it's own controller.
Altogether, I'm pretty pleased with the performance.
Here are some numbers from last Saturday for the half-hourly usage in Whrs (not kWhrs!) around the evening Golden hour. Numbers in the final column are the HH grid import from the SMETS2 meter (via Octopus), whilst HWSBO simultaneously used the kettle/oven/ hobs etc:

Sat-06-Nov 18:30 19:00 4
Sat-06-Nov 18:00 18:30 4
Sat-06-Nov 17:30 18:00 2
Sat-06-Nov 17:00 17:30 5
Sat-06-Nov 16:30 17:00 4
Sat-06-Nov 16:00 16:30 2
Sat-06-Nov 15:30 16:00 3

I'm pretty happy with that.

Downsides...
GE have said that the current firmware doesn't work particularly well in the 'Multi-plant' environment, but hope that new firmware, not due out until sometime next year, will be more robust.
Issues I've experienced so far have been:

1. One controller/battery discharging into the other
2. Un-demanded charge & discharge at max chat
3. Asymmetric charging / discharging

These seem to have been 'teething' problems, which appear to have been resolved by the deployment of a script onto the controllers, which monitors them, and re-starts them if they go AWOL.
Seemingly, one or both of the controllers loses comms with the EM115 meter, which sends them into a tail-spin.

There's one particularly annoying issue which still prevails though; in marginal weather, when the PV output is relatively low (2 - 3 kW ish) and erratic cloud cover means the PV output is pulsing up & down, one or both controllers are in a state of almost continuous flux, building up a charge, then 'tripping' out (presumably re-starting at the behest of the script) which affects both controllers, such that all of the surplus energy bleeds off to the grid. This seems to cycle over a period of about 5 mins or so.
Not entirely sure, but it seems worse when there's a split between the two battery SoCs (or should that be SsoC? !).
I've tried various mode changes, and splitting the modes, but to no avail.
Eventually, they sort themselves out - particularly if the PV surplus increases - but it's annoying to have spent so much spondula, only to see surplus energy weeing out into the grid for the DNO to sell onto an energy company, who in turn will flog it to my neighbours - at peak-price, no doubt!

Anyone else had similar issues?

To anyone considering a similar system, please don't let this put you off; all-in-all I really like the GE products, I'm very pleased with my purchases, and given the same situation, I'd choose exactly the same system again.
Despite it taking too long to get through to the support guys sometimes, when you do talk with them, they listen carefully to any issues, swiftly log-in to your system, and can talk with a good level of understanding about your issue, or query.

Clearly, a huge amount of work has gone (is going) into the new portal, and GE are constantly improving the end-user experience. Keep it up Folks!

Great to have a forum at last, too!

I've got 2 x 36 hybrid inverters, each with an 8.2kWh battery attached.
I am very happy with it so far. I've been charging at night before low sun days on cheaper Octopus Go electricity and this time shifting alone is a good money saver (5p vs 13.8p per kWh). I use a lot of electricity because I have a heat pump for heating the house and an electric car, so I can pretty much guarantee to use whatever I make.

However, I have also seen some oddities like uneven discharge - not sure I can definitely say one battery charging the other because it's only been in a couple of weeks - but I have suspected it. I've also seen the "waiting" status when there's grid import occurring. It was only for a few minutes, but as said wasn't necessary.

Just getting into the GivTCP protocol to better measure the real time stats so I'm hoping to better understand the way it works from that - the portal is just too slow to be at all definitive about anything other than performance over hours. The portal also adds everything together - I assume that's the same for everyone. So my import and export stats are doubled in the portal.

Hi Mike
Great to hear from someone with a similar system (albeit with different inverters).

I've been getting silly data back since installation. GE said it might improve when the new portal goes live, but it hasn't - for example, from my energy card, I've just selected 'generation' and am seeing this:
Generation to Home: 492.5 kWh
Generation to Battery: 5.4 kWh
Generation to Grid: 733.7 kWh
... now yes - we have a big roof with lost of PV on it (14.5kWP) but not THAT big!

MikeJ Just getting into the GivTCP protocol to better measure the real time stats

I'm interested in this, too - however, it'll be a while before I can start poking around in that area.

Like you, we have an EV (we use a zappi for this) and we use Octopus - the only difference being that we opted for a 5.5 hr window in the 'Go-faster' tariff options. The car has a 50kWhr battery, and I wanted to have the widest window possible, in case it came back needing a shed load of charge.

We're now seriously looking at an ASHP too; seems sensible with so much PV & storage; my only concern is what will happen if the cheap tariffs dry up!

Sykes
I have a single AC coupled 8.2 battery, so while I can't comment on the dual battery issues you're seeing, I have witnessed the same behaviour in fluctuating PV generation conditions. What I see is that the battery suddenly stops charging completely (sending excess to the grid) when PV generation ramps down because of a cloud - Instead of ramping down the battery charge in response. Your description of it "tripping out" is a good one. I've seen this looking at my Zappi interface and also on the EM112 meter, so I know it's happening in real-time. Also, I don't have the same scripts/dual set-up as you, so maybe it's not related to those.

TBH I've got 5p import (Go Faster) and 5p Export (So Energy), so the export is not exactly wasted, and I've come to accept it!

Have to agree though, having this forum is really useful and the progress being made at the moment on the product is great to see!

chrisdev
Hi Chris(?)
Thanks for that; interesting that it may not be the dual installation that's the issue. I'd be interested to know if you find out anything else about this behaviour - and particularly if you hear of a fix.

AND you also have a zappi! Very pleased with mine - especially now that ME seem to have sorted out the server migration. Response on the app seems pretty good again, recently.
I went for the 5 hr window on Go faster (5.5p)- just in case the EV came home needing a shed load of charge. Great tariff either way, and I really like Octopus, as a company. Kraken spits out billing in an odd way sometimes, but it all comes good in the end...

Are you using the beta portal, or sticking with the old one, for the time being?

Sykes

I have a similar setup, 2 x AC3.0 but with 5.2kWh batteries and not the 8s. Perhaps, on reflection, I should have "gone big" but the 6kW solar has yet to do much in these dark and damp days.

I also have a similar set of problems to you. The discharge at full chat was patched with some kind of script and the card data is pure fantasy but is apparently due to the data from plant systems adding the data from each previous day to the next. It's bonkers.

I worry about the asymmetric charge/discharge and also the thought of one battery discharging into another. This is clearly early days for GivEnergy and they seem to want to get it sorted but I feel the battery side of the cloud monitoring is sadly lacking. There is almost no information available and what is seems to show my batteries going from ₉₀% to 6% in the space of five minutes.

Very curious but this is the price of the early adopters.

Sykes
Thanks, yes I'm on the 5 hour Go Faster tariff too.
Fully invested into the new beta portal, that's where all the cool stuff is 😀

Hi JJB
Glad to hear it’s not just me!
My philosophy is that, thanks to GE, we can now communicate our gripes as well as our learning points. This can only be a good thing; I already don’t feel alone any more. 🙂
I still think that I’ve invested in good kit - and just like you, I applaud the efforts GE are making at the moment (in terms of products, the new portal, and the forum); I also can’t help feeling a little like a guinea pig, but that just seems to be the way of the world, nowadays.
I’m sure there’s more good things to come, and that the snags with multi plant systems will get ironed-out, as time goes on.

JJB
Have you seen the "battery data" tab on the "Inverter Card"?
That will show battery voltage, power, temperature, percent every 5 mins

I am guessing that as your SOC dropped within 1 data point window that may not be too much help, but just making sure you know that data is there

If you havn't already, rasie this issue with support@givenergy.co.uk
I will work with them from the portal perspective, they can help if there are any other technical issues

simon_swinburne Thanks Simon, I can see the battery information within the inverter card. I would prefer a separate card for battery stats, as requested on the beta site, since the power graph doesn't show SOC of the batteries. I'm also keen to view the two batteries side-by-side, rather than swiping between the two inverters. But I'm sure there are other priorities. I'm certainly keeping an eye on the stats and appreciate your responsiveness.

UPDATE: I see this has been done now - top job!

JJB UPDATE: I see this has been done now - top job!

I asked for a more accessible SoC readout, too - as you say, this has now been done. 😃
One question; mine seem stuck at 47% for both, when they're actually 30% & 64%.
Have you got the same issue, or are yours about correct?
Have fed this back via the portal.

JJB
battery card will come, thats just a bit more effort than the tweak recently pushed

Sykes
Mine also seems to show the average of 49%, as reported by the app, whereas they are (were) 63% and 35% respectively. I have asked Tech Service to take a look at this.

JJB Mine also seems to show the average of 49%, as reported by the app, whereas they are (were) 63% and 35% respectively. I have asked Tech Service to take a look at this.

All fixed, already! 😄

Sykes Wow, I just checked and it's workng like a charm. Thanks @simon_swinburne !

Battery card is getting closer to the top of the development pile.

Expect multiple iterations of that. I would prefer to get something out and then iterate, rather than sit on it whilst we polish it, When there is something to see I will share

Can anyone from GE comment on progress regarding plant installations and fixing the energy consumption charts? I'm still reviewing data that suggests I'm consuming megaWatts of power each day and that my 6kW solar plant is producing hundreds of kWh on a miserable day in December. I don't get what is so hard about fixing this and am disappointed that it doesn't appear to be any kind of priority - what's going on to fix it?

JJB Can anyone from GE comment on progress regarding plant installations and fixing the energy consumption charts? I'm still reviewing data that suggests I'm consuming megaWatts of power each day and that my 6kW solar plant is producing hundreds of kWh on a miserable day in December. I don't get what is so hard about fixing this and am disappointed that it doesn't appear to be any kind of priority - what's going on to fix it?

Agreed!
I'm in exactly the same boat; nonsense data, which means I don't trust any of it, and therefore am so far not benefitting from it.
It's so good to see so much progress on many fronts, but it seems to be at the cost of ignoring the many issues which still exist on some extant (and particularly, multi-controller) installations.
Some of the GE folk have told me that progress is being made in this area, but that doesn't make it any more palatable, when things don't work correctly (and why's it always during the weekends that things go wrong?!) or the data's unusable.

Does anyone have a hybrid and Ac coupled inverter in the same set up? Or know if this is possible? I have the 5kw hybrid inverter and 2*5kwh batteries, so was wondering if I wanted to add more battery storage it could be done in a way to get a higher discharge rate?

I have same set up of 2xAC3.0’s into 2x8.2kW batteries…I am charging on Octopus Go for 4hours at night at the moment. early days but this morning I see that batteries charged up to 100% by 03:30 and then, at04:30 discharged to grid?? I have 4% battery charge? Whats happened? Just given all my energy back to the DNO!.. any ideas? Can I call tech support?? Thanks

pauldavies Obviously something’s not right. Call Givenergy,they will look into your system 👍

pauldavies Whats happened? Just given all my energy back to the DNO!.. any ideas? Can I call tech support??

Hi Paul
Yes - I've had similar behaviour with my set-up since installation. Having had discussions with many of the GE tech support folk, the long-and-short of it seems to be that GE hadn't produced a system, where the multi-controller environment works seamlessly.
Apparently, there's a lot more effort going into the system design - especially on the firmware side, which is supposed to iron out a lot of these annoyances.
There's also talk of deployment of a 'box' which will apparently 'manage' multiple AC3s, but I don't know if or when this will begin.
Certainly yes - you can call tech support, who'll log in to your system, and discuss this issue with you.
There's often quite a wait, which is annoying in itself, but the guys will be both polite, and happy to discuss your system with you - including this morning's issue.
01377 252874.

Sykes @pauldavies

We had the same, the GE team will login to your system and apply a 'patch' that is a workaround. It hasn't happened to us since the patch was applied but I understand it involves an automated version of most IT support: "turn it off and back on again".

It's frustrating that the setup we have isn't being taken care of from a firmware perspective, I was not aware I had purchased something that was still very much in beta...

dcsh Your discharge rate will be limited by the battery and by the inverter itself. The GE website will tell you what the parameters are. Adding more batteries will mean you will be able to generate power for longer but the discharge rate will not increase, you would need to add an additional inverter for that.

JJB yes I understand the battery discharge rate limit with one inverter, that’s why I was wondering if a hybrid inverter and AC coupled inverter could be used together in one home each with different batteries. It sounds like some on here have 2 AC inverters each with their own battery and separate solar inverter.

dcsh You would expect that to work, wouldn't you. I'm sorry I don't have that setup and can't provide any more info.

Sykes Thanks for this - yes, the tech team were helpful (and quick 🙂) and as @JJB mentions, they restarted the system remotely - apparently due to a server issue their end not rebooting the system as part of the temp workaround until updated firmware available.. All seems stable for now... There's definitely a lot more going on with the two inverter system, impacting on how reports and dashboards are presented.. Fortunately I have an original MyEnergi system still measuring the import via the Harvey and I compare this to see actual import data for now. overall - despite the dashboards confusing outputs, I'm storing energy ok, and discharging during the day as planned..minimising daytime energy rates and daily £ have dropped by ₅₀% as a result...so some benefit coming back!. Looking forward to spring/summer sun charging the batteries rather than Octopus Go to reduce that even more!! 🙂

dcsh Certainly, I have separate PV inverters (2 - not hybrids) and separate battery AC Coupled Inverters (2) feeding 2x batteries in parallel...for the exact reason as many, to increase both charge input and discharge into system, to reduce grid import.. works fine..

pauldavies Fortunately I have an original MyEnergi system

Yes - I find that the ME app is by far the most-used app in my arsenal; it's my go-to place whenever I want to look at overall system performance. I don't use the GE app much at all, as it's increasingly unreliable as more & more changes are made to the new portal, along with only being able to access one AC3.0, which is next-to-useless, really.

Is there any update on progress to fix reports showing incorrect numbers for a multi plant set up? The dashboard can show me the generation figure for the 5 min snapshot (currently 750W on a grey day), but cant add up the days total generation?.. with 5kW inverter capacity (1 x3kw and 1 x 2kw) the dashboard is currently showing 129.8kW generated today ! Impressive but wrong!.. its more like 4kW.. ! My reported consumption of 2,237kWh today is even more unbelievable,(and frightening!) with the actual at 22.7kWh... from the ME app... Be good to get these figure fixed and make the app useable..

pauldavies
Paul
FWIW, I'm still getting silly numbers, too.
I've kind of given up on looking at them, and am just assuming that at some stage in the future, when GE roll-out new firmware designed properly for multi-inverter systems, that things will come straight.

Likewise... just bugs me !! Hopefully soon 😃

I too, got silly MW numbers a few days back - but only for an hour or so, it had corrected the next time I looked, and the historical data didn't reflect any erroneous readings.

Thats good if you have 2 inverters in the system and normally you’re getting correct readings on the dashboard? I never get sensible numbers!

I put in a bug report and it was closed saying thier figures are correct. I would have been using 120 amps from the inverter...

I shall raise the bug report again later.

Well at least I now know I can get a second AC fitted to our current 1ac 2x 8.2 system. Seeing the panels generate 8kw and 5 of it pissing away into the grid is quite maddening! Maybe we'll wait till they sort their firmware out.

To be fair what they have done with the firmware seems to be working well and I would say get the second inverter as soon as you can to take the opportunity of your generation. I am also topping up at night on the Octopus go tariff and with one inverter you would not be able to fill your batteries (3kW x 4 hrs) . The second inverter in parallel makes all the difference. The reporting is a minor issue overall and I am sure GE will sort soon!! 🙂

pauldavies No mistake - the second inverter is brilliant. Only you cant rely on the reporting. So if you forget the reporting and look at the DNO meter only - you are much better off.

anabanet “What they said”…forget the GE data, it’s nonsense. I recently installed Octopus Watch and it’s reassuring to look at accurate data, albeit retrospective.

pauldavies To be fair what they have done with the firmware seems to be working well and I would say get the second inverter as soon as you can to take the opportunity of your generation. I am also topping up at night on the Octopus go tariff and with one inverter you would not be able to fill your batteries (3kW x 4 hrs) . The second inverter in parallel makes all the difference.

pauldavies
You make another good point, which I failed to mention in my original post; with the restriction of only being able to draw circa 3kW through the one inverter, even a 5hr cheap-tariff window (which is what I chose with Go Faster) wouldn't be long enough to completely fill 2 X 8.2s from empty, and as I believe the replacement offering soon an the shelves is a 9.5 kWh job, two of these would need even longer to get topped-up.
As well as the zappi & ESS, we're also running a 16kW-output air source heat pump now, and I'm finding that, during the dark winter days when the cloud's too thick to provide any meaningful PV energy, I'm needing to fill both 8.2s from empty, during the overnight tariff window. As you rightly say, despite the firmware downsides, it's still definitely worth having an AC3 coupled on each battery to be able to charge them up fully on cheepo energy. As I've already mooted, the extra cost is relatively small, when compared to the potential benefits.

Anyone planning a new system would be wise to consider these points, and always bear in mind that, however modest your planned system is to begin with, the likelihood of it expanding (and more rapidly than you may have ever expected) is quite real - especially as we're driven more & more away from burning fossil fuels, and naturally, towards the wider use of electricity. Plan for expansion, wherever possible! 😀

While not strictly plant multi Invertor. I have one invertor and 3 batteries. The battery voltage, is this of one battery (Master) or the average of all 3? The battery temperature, is that of the master or again the average? How can we see the states of each separately? I guess as they are different addresses on the bus, they should be feeding their own BMS stats to the invertor??

TheDragon (GivEnergy) One inverter, multiple batteries, you see the average voltage. It looks at all the cells as if they were one super battery.

Sykes We may be going 3 inverter plant next month....

anabanet We may be going 3 inverter plant next month....

Wow! Good on you!
Must admit, I'm tempted to get a third set - particularly now that we've installed a big air source heat pump (16kW output, which, on cold day earlier this week, I saw was drawing circa 7kW!). We could do with that extra grunt in order to run everything (especially the kitchen & heat pump) during the period after sunset, and before the cheap tariff cuts-in.

The only downside would be that, because our DNO will only let us have an 80 amp cut-out fuse, I'd need to carefully time what bits of kit were drawing current, and when, during the cheap-tariff window. What with the heat pump, zappi and the two AC 3.0s, we're already bouncing off the limit at times...
Thank Heaven for the cheap Octopus Go faster tariff. Just hope it continues... 😃

Has anyone got a photo of the connections from the EM115 meter and in to the inverter. _want to make sure mine are wired correctly.

Is there any progress in the software improvements managing multiple inverters?. My 2 AC 3.0s still fight each other and currently one is charging the other when there is little surplus PV?. Its a shame as all this discharge and charging is using energy, reducing the efficiency of the system as a whole, so any spare PV we have is firstly used up by the inverter activity!. It makes the reporting figures quirky as well. Currently generating 2.1kW total, Consuming 2.3kW Total and charging 1.1kW total and exporting 5W!.. ...how does that work? ! Paul

pauldavies Sounds like a call to ask for support. Although I don't have AC-coupled, I do have two HY5, each with its own PV and battery and set to ECO mode. They are connected together on the 240Vac side but they aren't connected together otherwise for control purposes at all. These play together very nicely. They seem to load share 50/50 and when the PV generation falls off, they discharge roughly the same amount to the house. By morning their batteries are within 1-2% discharge of each other, so they have played nicely all evening and all night. The only time when they differ is when the PV output is different. One pair of PV strings, connected to one inverter catches the sun earlier in the day while the other carries on when the first tails off. It hadn't occurred to me that they wouldn't operate together smoothly until I saw your post.

Are your AC3s on the same Modbus with one as Master and the other as slave?

Thanks Tim
I will reach out to support thanks.. sounds like yours are working well. 😄

Only difference will be that the PV is fed into the AC board via two separate inverters but the AC 3.0s are like your HY's - wired independently to two separate 8.2kW batteries and both feed into the same AC board..

Regarding the Modbus?.. I have seen and heard of the concept of master/slave but have never found where this may be set or shown?? How do I check this?

pauldavies concept of master/slave

This links to the Givenergy Knowledge Base for configuring more than one AC3. The Modbus is the communications network between the inverter, EM115 meter, the wifi dongle, battery management system(s) and so on. For the HY5, the inverter has to be the master device and each Modbus can only have one master (although Modbus over Ethernet seems to be able to respond to more than one master). I think the schematic linked is intended to show how a single or multiple Givenergy AC3 plant can be retro-fitted to an existing (eg FIT) inverter/PV system with one of the Givenergy AC3s as master and one or more AC3 as slave devices, each with one or more batteries attached. HTH.

As Tim stated, each Modbus RTU (RS-485) network should only have 1 master device. This is often depicted with an ID of 0, with all the slave devices requiring a unique ID number. The linked schematic does not indicate what ID number is used for the 2nd and 3rd AC coupled inverters configured as a slave devices.

I work for a BMS company (Building Management System not Battery Management), and regularly have to configure Modbus communications to retrieve live data from a multitude of devices by other companies. Without delving in to too much detail, the master device polls each slave device for data over the shared network, so potentially the 2nd inverter can listen in for the meter data at the same time as the master receives it.

The complicated part for GE is how each inverter responds, especially to the grid meter data that indicates the current size of import, export or neither. Each inverter then needs to respond with a close match of charge or discharge of the battery. There will be change of value thresholds and time delays applied to prevent constant hunting of the inverter input/output. Of course with 2 inverters consideration also needs to be taken of the state of charge of each attached battery, and potentially different size batteries. I would not relish having to write that control logic.

The interactive control logic for the hybrid inverters may be less complex, as the battery charge cycle from the solar PV does not involve the ac component of the inverter. The ac charge cycle is only initiated by a timed charge command configured via the cloud portal or automatic API. This could explain why the hybrid inverters play better together.

pauldavies Is there any progress in the software improvements managing multiple inverters?. My 2 AC 3.0s still fight each other and currently one is charging the other when there is little surplus PV?. Its a shame as all this discharge and charging is using energy, reducing the efficiency of the system as a whole, so any spare PV we have is firstly used up by the inverter activity!. It makes the reporting figures quirky as well. Currently generating 2.1kW total, Consuming 2.3kW Total and charging 1.1kW total and exporting 5W!.. ...how does that work? ! Paul

Paul
I have very similar issues - along with others, to boot.
Pretty much same set-up as you.
I regularly have the 'loss-of-comms' issue, where the master inverter loses comms with the EM115. GE have deployed the software patch onto my system, which monitors for loss-of-comms, and re-boots the errant inverter where necessary. This causes all kinds of issues, including 'hunting' of battery charge, when recharging on the very valuable PV. It's really annoying to see all the PV energy being wasted to grid every 5 minutes or so, until the inverters decide to come back on line, and charging starts again - for another few, short minutes.
I, too, have had instances of one battery charging the other, which is a massive waste of energy - bearing in mind the round-trip efficiency of 80% or so.
I have some REALLY bonkers daily data - suggesting daily import/export etc in the MW range!

When I can, actually get through on the phone (!), GE always tell me that the long-awaited new firmware to cover the multi-plant environment is expected soon, but, like Dianna Ross, I'm still waiting...

To add insult to injury, my system's recently began behaving differently to the way it used to.
Since last September, I've been successfully charging the EV overnight on cheap Octopus energy (lovely jubbley!), whilst preserving ESS charge. I've manage this by setting the 'Enable AC Discharge' facility, and setting the two discharge periods to occur outside my Octopus cheap tariff window (midnight thirty to 05:30). Until a couple of weeks ago this worked fine - whether-or-not I had ECO mode (the old Mode1) selected simultaneously.
Then one night, I found both batteries had begun dumping at full-chat, into the car, during the time they previously wouldn't discharge. I've also had instances of one inverter (my master) just dumping to grid during the night. On that occasion, the only thing I could do was to physically turn off the battery, which subsequently gave me all manner of issues getting it back online again next day.
Either the server software's changed recently (possibly with the advent of the new app functionality) or I have a malfunctioning inverter. Either way, two phone calls to GE both yielded the same conclusion at their end - not that there could be an inverter problem, but that the system won't work in the way I wanted it to - despite (and seemingly ignoring the fact) that it's worked perfectly well that way for the past 8 months!

... to answer your question, yes - apparently there IS progress on the software managing multiple inverters, but of course, there's been 'progress' for the past 12 months or so...

Has anyone got a multi AC inverter (plant) that is working as expected? I am awaiting DNO approval for my 3 x AC plant setup to be installed and from what I read im getting more and more worried it won't work.

Sykes Thanks for the reply - apologies for the delay in getting back to you.. Really helpful to know and glad its not just me...The last conversation I have had with GE tech support was that the system was not correctly connected...(!).. the reason being that you can view the PV input from the two PV inverters we have, against the two battery inverters.. ie one array appears to be charging one battery and the other charges the second battery - even though they are connected completed separately and independently... This became apparent when I asked how to establish which inverter is the master... and from this discussion, the conclusion was that both were acting independently... and as a result the hunting and cycling of charge/discharge is possible... I am waiting for installer to confirm back the set up... and change as necessary.. 🙂

@pauldavies
Paul. I'm sure this will all get sorted out in time, but like you, I'm finding it very frustrating in the mean time.
I think we're suffering from Guinea Pig syndrome (just made that up!) in that there's so much fairly new technology out there that we're investing in - all of us for many different reasons - and the modern development strategy seems to be to get it working as well as possible, then get it on the market, before the competitors release theirs; the final testing being carried out by the fare-paying customers. Not saying it's just GE - it seems to be the case across the board.
Allied to that is the myriad of different equipment we all want to bolt-on to our own particular systems, in the hope it'll all play nicely together, and there's potential for a perfect storm.

The other thing I'm beginning to realise is, that the standard of the trades who supply & install our equipment is very often lower than we, as laymen, hope. We all like to think that there's a Big Brother watching over all of the trades, but in my experience (not just with solar/PV) the reality is that all trades are equal, but some are more equal than others (to paraphrase George Orwell)...

I thought I would just report and update on operation to date with the 2x AC 3.0's and 2x 8.2KWh batteries, as I feel I am just getting used to accepting some of the vagaries of multi-plant operation. Are iothewrs still having issues>..

1) I regularly (once a week) get a period where one inverter discharges to the second inverter for no reason - no house load etc..?

I have updated the BMS to 3007 on both inverters on the advice of GE support. One was stuck on a previous version which turned out to be because the onboard USB installed ( not the dongle) inside the unit was defective and not writing the new BMS and updating the battery. Replaced USB dongle to get this sorted.

State of charge definitely "falls off" dramatically in certain circumstances... and I have had a couple of recalibrations done remotely in the last month - annoying as it obviously dumps the charge into grid and recharges from grid if no PV - wasting money!

Reports and dashboards still seem to present random and unresolvable data - today for example the actual solar generation on the solar forecast widget is 2657kWh!! forecast is 16kWh which is realistic...

I still dont know if I have a "master" inverter set to manage the 2 as I continue to be able to see 2 sets of data for each inverter - pv generation, import and export figures.. even though GE support have said if they are correctly configured as master slave they should both report the same? Anyone have any experience/views or comments on that?..

I currently have one battery charged to 95% - most from dumping power from the other battery this am at 7am.. and the second one at 9%.?.. charging.. I just didn't expect this uneven management of the batteries by the inverters...rather a more balanced approach?.. obviously one inverter is "grabbing" or prioritising over the other...?/. Again.. is this right?.. I dont expect a lead to take power from the other battery as a priority?.. with all the inverter losses this is cost inefficient..?

P

Hi Paul,

Like you, we have two AC3.0 inverters, each with their own 8.2kW.h batteries - but ours has only been in for a couple of months. We have solar on the roof, which comes in via a Fronius Primo inverter. Broadly speaking, we seem to have similar experiences to you, as follows:

BMS
We have one on 3007 and one on 3008. I am told that 3008 is a special release version, for a 'specific issue', but I don't know what that issue is, or why the installer gave me that version.

Battery Discharge into 2nd Battery
We also get this once or twice a week. If I spot it happening, I usually restart both inverters via the web portal, which seems to stop it.

State of Charge
We sometimes get a jump up to 100% (from 80%, perhaps), or a similar jump down at the bottom end - I believe this is the same SOC calculation 'drift', which is being discussed on several other threads.

Dashboards
Many of the values we see are just crazy. The 'Home' view on the App is usually the most reliable for reporting. If I see anything which looks wrong, I ignore it. For example, the web portal reports that I have generated 1MW.h today! For any edits, I always use the web portal.

Master/Slave
My installer said he had set them up as Master and Slave, but I also get the two rows of information in the 'Inverters' section of the web portal. So I think this is normal (despite what you were told). But I have yet to take the covers off the inverters, to check the positions of the DIP switches (which define if it is a Master or a Slave).

Battery SOC Imbalance
Like you, we often have one battery which discharges much more than the other. This means that during solar charging we often end up exporting power, because one battery is full, and the second one can only consume 3kW from our solar. I am hoping that this will be addressed in the new software update, which I believe is close to being released (primarily targeting the SOC drift issue).

At the moment, it doesn't look like the 'Master and Slave' logic is actively trying to balance the SOC of the two batteries, which is frustrating. I suspect that slight differences in the two 'systems' means that one automatically becomes dominant - hence the big differences.

I have started to manually force more solar into the more empty battery, in the morning, to try to balance them out. But this relies on me being in front of my PC, so that I can adjust the charging values, as the solar power rises and falls.

I hope that gives you some minor comfort, that it is not just you!

Chris

Master and Slave
I have just been looking through some of the GE installation manuals, and I can only find references to Master and Slave in the context of multiple batteries, on the SAME inverter. So now I am questioning if there is, indeed, any kind of Master/Slave setup for multiple INVERTER systems like ours. If there is, I am not sure how it is achieved, assuming they do talk to each other (as opposed to both talking independently to the GE cloud).

Perhaps each system (i.e. one inverter plus battery) operates completely in isolation, and has no knowledge of the other???

Can any installers shed any light on this matter?

That might explain why you see batteries did charging into each other etc and that one or the other responds to the load rather than both equally..

ToothyChris The AC3 plant inverters are on the same Modbus and the protocol for Modbus RTU is that each device has its own unique address. By default, the inverters all have the same Modbus address (17, I think, but can't remember whether that's decimal or Hex). Since addresses have to be unique and there can only be one master, there must be something that configures slave AC3 inverters with different addresses. Don't know how it's done as it's not my area but if not by DIL/DIP switch as per battery, then by another means?

ToothyChris "Battery Discharge into 2nd Battery
We also get this once or twice a week. If I spot it happening, I usually restart both inverters via the web portal, which seems to stop it"

How do you see this behaviour is happening in the dashboard?

ivell Hi and thanks for your response.. How do I see it?.. Well on the App I check most mornings as I have an idea of how much charge should be in the batteries first thing..on this I still am not sure I am seeing the combined battery level, or just one, but usually I can spot high use at a time I dont expect... and then I look at the Web portal, where you can see each inverter, and invariably, one is at say 90% and the other at say 9%... I then can look at the individual charts and see when one inverter is charging the other, at the same time is discharging.. its like an exact mirror.. that's the only logic I use.. If I could attach a picture to this I would!..

ivell
I look at the graph in the Portal, and I check that both the blue areas (i.e. charging and discharging) are EITHER above the zero line, or below it - this is the normal/good situation. If you see blue area both below AND above the line, you know that one is discharging into the other.

The other clue is that your household consumption may appear to be higher than you expect.

See at 11:45 in here, and again right on the right-hand end:
https://iili.io/PV9wZB.png

I asked the electrician who installed my system, about the Master/Slave question, and he replied:

'The master and slave is something we determine when wiring it as it relates to the smart meters and them we confirm this during commissioning'

So I not sure how you would tell, if your system has been set up like this....or not.

When commissioning a "Plant" system there is the option to choose which inverter is the master and which is slave one and slave two. This is software based not hardware based unlike the dip switches on a multi battery setup.

ianhorton
Many thanks Ian,

Do you have any insight into what exactly happens in Master/Slave operation, for two AC3.0 inverters?

I ask, because the two parallel systems do not seem to operate in any obviously logical or co-ordinated way: They do not track each other in terms of charging power, or exporting power or battery SOC....so why do you need a Master and a Slave at all?

More.....
If it was a true Master/Slave relationship, surely it would be impossible for one system to be charging, whist the other is discharging (as we so often see). If it was Master/Slave, then the slave would follow the master.

This suggests to me that there isn't really much/any co-ordination between multiple AC3.0 inverters, and that they are not really operating in Master/Slave at all. Instead they are operating in isolation, which sometimes results in conflicts, and/or the unbalanced discharging/charging power and unbalanced SOC.

ToothyChris true Master/Slave relationship

This is just the way the devices on a Modbus are referred to. By convention, there can only be one master device which requests register data from the slave devices. All the Givenergy inverters are configured as master by default with the battery BMS and EM115 meter as slave devices. Hybrids aren't connected to each other over Modbus and just work fine (because they only feed the house load from PV/battery, they don't charge from surplus AC by default) whereas the AC-coupled are and therein lies the rub: dynamic control of multiple inverters to balance import/export.

Evidently, Givenergy are struggling to get the plant control of AC-coupled inverters sorted along with the battery BMS which keeps collapsing on some systems.

Tim sounds like there’s a new plan for this. Some kind of control box. Discussed elsewhere in the forum.

I have 2x3.0AC each with a 8.2 battery. Am I citron understanding that if one battery is finished before the other in charging, the battery that can still accept charge won’t charge faster that 3kw? (From it’s connected inverter) or will both inv (as it’s a plant) charge it at up to 6kw?

Smk82

Your understanding is correct. Each system can only charge at 3kW, so if one is full you end up exporting power, when you don't really want to. I end up manually trying to keep the batteries balanced, as they charge up, for this very reason.

Tim
Thanks Tim, so the use of the term 'Master/Slave' in this context (i.e. Multiple AC3.0) does NOT imply any kind of coordinated operation, between the multiple inverters.....That would certainly seem to fit the evidence, from our users' perspective!

ToothyChris Haha! I think Givenergy have AC3 firmware to enable the separate inverters operate as a single plant (otherwise, why bother connecting them) but evidently they've not got it right yet. I'm not pretending that getting the control algorithms right is an easy job but there are systems out there that do respond very quickly and reliably to changes. Most modern passenger planes have automatic landing systems which work for nearly all runways. I'm just glad that the airlines don't use AC3s to fly the planes!

ToothyChris
Thanks for the image
I have never had this - perhaps this is when someone has two ACs? I have two hybrid inv + 2 batts..

We would like to add a second AC3 to our setup but I'll just wait until they've got it working better I think.