Hi, I have attempted to get to the bottom of this multiple times on this forum and other but I am still having the exact same issue.
System: 5000VA Fronius inverter with 6kw of panels, connected at the house. 40cable meters away there are 2 x 48/5000 MultiplusII in parallel, and 2 x 450/100 MPPT’s with 75kw of battery and Cerbo GX. Completely off grid. typical house load avg 1200w to 3500w during the day, afternoon evening can see upto 7kw for a couple of hours ( 2 x EV’s, Electric oven, deducted AC’s etc)
When we have nice clear sky days there are no issues, when the battery is charged and I look at the system I often find the MPPT’s all at zero output and the Fronius throttling nicely on frequency control matching the load. The MPII’s just idling along.
The issue: on sunny but partially cloudy days, the system will do its thing cloud rolls over all the solar drops down in output, the Multiplus take up the load, then the sun comes out from behind the cloud, so all the solar suddenly springs to life, around 40% of the time I will get an alarm on the second Multiplus stating “High AC voltage on output 1” and it trips the system, no power to the house at all. The current solution (which is not a good one) is to go isolate the Fronius on cloudy days and let the MPPT’s do all the charging. Not ideal, plus it always happens when I am not home, so I have 15 minutes to do something about it before my internet connection is lost.
I used to have 8000w AC coupled (2 x Fronius a 5000 and a 3000) it fits with the Factor 1 rule as I have two x 5000va MPII in parallel. I figured this was the issue, so I have dropped out the 3000 for about 4 months now. I am still getting the same issues on cloudy days. The AC cables out of the MPII are identical length (6m) 4mm2q TPS The battery cables are 600mm long both identical length 50mm2q to the lynx distributor. From the Lynx its 2 x 90mmsq cables on both legs (4 cables) around 2m long to the Lynx class T and then 2m cables to each Battery Bank (5 of)
I need some expert assistance with this please. I have had several attempts at getting this to work. The system is not ESS, I do have the Solar frequency control assistant in both Multies, and I do use the cerbo GX to reduce the charge current (down from 250amps) after the battery Soc hits 85% to slow down 25 amps every 5% increase in SOC (125amps in the last 5%) of the charge. I charge 2x EV’s and run AC’s etc in the house mostly evening loads after 1500hrs the batteries do get down to around 50% some nights and we need all the solar we can get to recover in winter. Summer is not a problem. Batteries are CAN bus and Cerbo Uses this for control. I do also have a smart shunt in series as a bit of a Checks and balances device.
Thanks for the reply, yes I can confirm it is set up as per the document. I have been through that doc several times in the past making sure I haven’t missed anything.
In general it all works on clear sky days the MPII’s increase the Freq which causes the Fronius to throttle back. Often the Fronius is supplying the load with just a couple hundred watts coming from the MPII’s. So typically it works fine.
But on cloudy days where essentially it is a nice sunny day, but intermittent big fluffy clouds that block 70% of the solar is the days I have issues on.If I have 3kw of load batteries are at 100%, the fronius inverter is fereq throttled to provide for the load, Cloud rolls over for a few minutes the MPII’s pick up the load, freq drops back to 50hz. Cloud rolls on and suddenly the Fronius ramps back to 100% output in a few seconds and the MPII’s just don’t throttle / raise the freq quick enough..
In the Assistant the Freq shift starts at 50.2 on the Fronius its 51 (I think from memory,) So I have even elevated the lower shift point to 50.8 in the Assistant and bought the top point down from 53 to 52.7 which is the cut off for the Fronius in an attempt to reduce the span and thus make it react quicker, but this still doesn’t work on those intermittent cloudy days. Works fine on full cloud cover days as the solar output is less over all but more consistent. Works perfectly on full sun days where typically we are starting to see Freq shift by 9:30 to 10 am generally fully charged by 10:30 / 11 am on a clear sky sunny day in the Spring / Summer /Autumn months. (Sept through April)
Frustrating, to the point I have removed the 3kw AC coupled solar and its sitting on back veranda, contemplating what to do with the 5kw. I dont have any roof space close enough to the batteries to put this solar on with MPPT controllers.
Have you tried running the Fronius only and MPPT only? To prove if fault is associated with the Fronius or the MPPT or the combination of both causing the problem.
Did the system ever run without this issue? Did it commission all good then months (lots of cloudy days) then later break? Did this problem occur the first cloudy day post commissioning?
Have you been able to measure the voltage during the high voltage alarm? Eg prove there is a high voltage is real or not on the system (not a false indication). Caution this is dangerous voltage to measure; please ensure you know and manage the hazards when working on live electrical system.
Yes, so the way I manage the issue is to turn off the fronius and use only the MPPT. The intial install for the first 18m was fronius only. First just the 5000VA unit, then I added the second 3000va for capacity and I was suffering the exact same issues. So in more recent times I have added 4 x MPPT arrays, removed the 3000w Fronius Array and looks like I might have to remove the 5000 as well for reliability. The problem is during the day I use this extra AC capacity to run house and charge cars.
No I dont believe it has ever run properly. Initially it was only 1 48/5000 and the one 5000va fronius, 2 batteries totaling 18kw. I was having the problems then. I anticipated it was a factor 1 rule, so I added the second Inverter which I needed as the EV arrived about then. This didnt fix it the issue, so I added an extra battery taking capacity to 30kw, this didnt help either.
No I havnt measured the voltage as yet, it always seems to happen when I am not home or not ready waiting for it.
Hi Nick, when I last tried to solve this in September last year, I made an adjustment lowered the voltage to 55.2v and then drop to 53.5 float. so there should be plenty of capacity across 5 x 15kw batteries. I have also set up Node Red to start reducing the charge amps as we get over 85% in an effort to early throttle the MPPT’s and Fronius to try limit the event, but also not be charging the batteries at 240Amps plus when it hits 100% SOC
Therefore the fault is coming from the Fronius. Have you tried one Fronius versus the other to see if the fault is in a specific fronius unit or common to with fronious?
Hi Sean, Cable is 16mm SQ TPS, switchboard on both ends, one end is the Inverters. Other end is the first house sub board, where the AC’s, EV’s and Fronius connect.
Correct No set grid code, and PV assistant set to start raising frequency at 50.8hz
I will have a look at the restart voltage, I was unaware this effected the Assistant interaction.
The connected BMS are JK inverter style BMS with Eve 280LFk cells, I dont normally get an over charge current limit, its normaly the slave inverter states Over voltage limit. I used to get the OCCL error, thats when I instigated Node red to reduce the charge current as the battery neared 100%
The slave Victron is throwing an “Over voltage” on output1 and trips, taking the system down. A reset or inverter on off command from VRM fixes the issue and every things as it should the Fronius Ramps as it should and all is well post the event.
Correct MPPT only no Issues, its only a problem when the 500va fronius is in service, and only on days when we have typically good sunlight, but the big white fluffly clouds pass over temporarily blocking the sun significantly, so all the solar production will drop down to maybe 2000w then springs back to a value that can be as high as 14000w after the cloud moves on (typically all happens in a couple of minutes) until the system gets hold of it and shuts off the MPPT’s and adjusts the freq to reduce the output of the fronius.
I have 3 Fronius Solar inverts in total 2 x 5000va and one 3000va. One of the 5000’s is in the shed as a spare. It was the duty unit unitl this started happening every few days in September last year. I swapped it out late Sept, thinking it may be the Fronius it didnt help, so I decommissioned the 3000va and moved the panels to the shed and onto MPPT. So the shed has 4 x 4kw MPPT arrays, house has 1 x 6kw AC coupled array.
Problem went away as the weather cleared up mostly, it has happened a couple of times since during cloudy days, but in more recent weeks we are starting to get the cloudy days again as our summer tapers off and the issue is re-occurring. Today is a cloudy day and before my wife went to work she shut off the Fronius inverter as no one would be home to reset the system or know it had tripped.
I am able to access my Ve config files from work, so I just took a look at the settings, in “PV inverter support” assistant I dont get asked the voltage it skips that screen and goes directly to the Frequency table.
If I open a new config and add that assassinate I get asked for the cold temp restart and its defaulted to 40v. My assumption then would be that the assistant in my case is set to 40v as I dont get asked, thus it would be using the default values.
I have used VRM, done a remote download of the settings in the master, found the assistant is set at 53v for this value, have adjusted it down to 40v as suggested and re-uploaded to the MPII’s. Hopefully this makes the change I am looking for, will know as soon as we get some more cloudy weather. Currently system is running at 4000w AC load and the Fronius is throttled at 51.7hz allowing 2300w which is pretty normal batteries are 100%
over the next week we are expecting fine days, so may not see the results for a week a more of this change.
Thank you very much for your suggested fix, Still not sure how a values described as “Cold temp restart” causes this issue, but it appears to be fixed, we have had a few mildly cloudy days and it appears to be controlling perfectly. We are due some bad weather this weekend, so Sunday Monday coming out of it will likely confirm the issue is solved. Thanks again.