1 Phase Parallel Quattro system frequently disconnecting from Grid & Geny running not detected

Hi All, new here and looking for some tips or advise on a very strange issue on a fair size commercial site.

Site had the following errors when I arrived:

• Grid Lost
• MPPT Error #27
• Ve.Bus Error #17
• Low Battery
• Ve.Bus Error #10
• Ve.Bus Error #19
• Generator Not Detected at AC Input
• Ve.Bus Error #3 Not all devices found
• Slave does not have AC Input
• MPPT Error #67 – No BMS

System comprises of 4x Quattro 15000 in parallel on single phase supply, 1x master + 3 slaves, system includes a large number of BYD LVL 15.4 units with BMU looking after them, a Cerbo GX with colour screen, 4x Victron MPPT RS 450/200 Solar charge controller connected to 120x 380w solar panels, a common DC Bus between the MPPT’s, Quattro’s and batteries, 1x 63KVA single phase Generator supplying all 4x Quattro Inverters on AC in 2 terminals. Grid supply is connected to AC in 1 terminals on Quattro’s. The site had a perfect storm event (apparently soon after the original installer did a service visit), the client has also advised the Grid supply has been poor for a long time. The event included the following: MPPT’s were not charging batteries (the cause of which I’ve since rectified), the Quattro’s constantly disconnect from mains (Grid) supply + one of the slaves had a low battery fault (which I’ve also since rectified), the generator kicked in and ran at set battery SOC % but also didn’t charge the batteries as the cerbo and Quattro don’t recognise the generator when it is running and supplying the Quattro’s. This caused the batteries to deeply discharge to SOC of 4.8%. I was asked to come in to assist the site as the installer has done a runner. So I replaced the RJ45 Ve.Can cable between the Cerbo and the first of the MPPT’s which got the solar running again (quick win) as they weren’t identified devices and missing. The batteries started charging but barely and constantly going to idle state and yet Quattro Absorption LED is on while at 55% SOC, this remained the same on next day which was caused by bad battery within the bank, one BMS had a voltage lower than the rest, that battery must have taken the hit when deeply discharged, it’s been disconnected, bypassed on the CAN and BYD BMU software all been re-configured with one less battery so that problem is now sorted, the other batteries were all at 95% SOC which made sense why they kept going into idle state, the BMU was healthy without faults etc. but it was essentially confused as well as the Cerbo. Batteries now all fine at 99% SOC and charged up by the solar. Final issues remaining are the Grid mains disconnecting at Quattros constantly and the Geny when running not being detected at the Quattro, Geny start signal works fine from Relay 1 from the Cerbo using SOC %, voltage is measures fine at the AC in 2 terminals but it’s doesn’t reflect that on the Cerbo Overview page. Have done extensive live and dead testing of all supplies, Grid is unstable with poor Hz at AC in 1 across master and slaves, drops to 49Hz which I can see dropping and jumping around when the Quattro’s disconnect from mains (you can also hear the unstable supply while they’re ON), Voltage – Line to N = 228.9V ; Line to Earth = 232.5V ; Neutral to Earth = Ranging from 4.7 to 5.4V (Not ideal) Without taking apart the Quattro’s have measured the AC in 1 and 2 resistors all okay at 219KOhm, measured the AC in 1 and AC out 1 terminals and all similar: Carried out Dead testing on INV units: (To ascertain if any INV units differ greatly, possible hardware failure on relays, resistors or internal transformer winding)

INV1 AC input 1 terminals = 368.7 Ohms (Questionable perhaps)
INV2 AC input 1 terminals = 390.5 Ohms
INV3 AC input 1 terminals = 384.9 Ohms
INV4 AC input 1 terminals = 389.7 Ohms

INV internal transformer winding measurements:

INV1 AC Out 1 terminals = 0.15 Ohms
INV2 AC Out 1 terminals = 0.14 Ohms
INV3 AC Out 1 terminals = 0.13 Ohms
INV4 AC Out 1 terminals = 0.10 Ohms

Tested for Neutral continuity between AC In 1 and AC out 1 which was okay on all.

Measured AC input resistors: All 219 KOhm ±5% (Red, Red, Yellow, Gold)

INV1 – AC In 1 = 368 Ohm ; AC In 2 = 219 KOhm

INV2 – AC In 1 = 389 Ohm ; AC In 2 = 220 KOhm

INV3 – AC In 1 = 384 Ohm ; AC In 2 = 221 KOhm

INV4 – AC In 1 = 389 Ohm ; AC In 2 = 219 KOhm

Grid LOM settings are:
AC in 1 (UK G99) plus LOM (Grid Input)
AC in 2 (UK G99) no LOM (Geny Input)

Weak AC input selected seems to help only slightly but issue persists.

Quattro’s are on latest firmware 2656556 (Updated during last visit just before issues started), considering reverting to previous which version which seems to be 2656552 on Victron firmware online. Yet to try this but this is one of my next steps when returning to site.

System has ESS with assistants. All settings of which I’ve saved copies of etc.

Updated Quattro settings battery capacity setting to match new battery config of 2700AH instead of 3000AH, so that’s okay now.

Reverted Victron Cerbo GX firmware from version 3.60 to version 3.55 after researching and finding on the Victron website, a public statement saying v3.6 has known bugs and false error codes that Victron is aware of causing issues on many sites and they recommend reverting to previous firmware – reverted firmware as a further elimination step to avoid existing or more issues and alarms and any possible communication issues with INV and MPPT units etc. made no difference.

Disabled firmware auto updates on Cerbo GX.

When checking all the Quattro settings via Ve.bus and MK3 (with Cerbo cable disconnected), I saved photos and text files for each unit Master and Slaves settings in full and saved the actual settings file for backup purposes and evidence of findings for client.

I then tried a quick elimination tactic making one of the slaves the new master and visa versa to try rule out a hardware or measuring card/circuit issue within the master inverter (leaving firmware alone), re-configured the devices to suite, then settings and re-configured successfully using Ve.Configurator, test ran the inverters with mains/grid on and still experiencing the same issue with them disconnecting from grid all the time at least every 5min or less sometimes twice a minute or more.

What is confusing is the Quattro’s seem to disconnect from mains/grid well before the reaching the LOM thresholds for under voltage and under frequency which are 184VAC and 47Hz, however I’m using a standard fluke multi-meter so thinking of bringing a Fluke Power Meter (which I have access to) to site to get accurate evaluation of the Grid supply with a scope function and graph logger. Will test the external loop impedance (Ze) of grid supply incoming cable as did not get around to that yet, to see if it’s at an acceptable measurement from the DNO. Based on lots of reading other discussions the fact that the client site is at an end of line and experiencing such bad Grid quality over time it would be good to get an accurate assessment of the supply.

Before test running the Geny I did turn all AC In circuit breakers off to simulate a grid loss.

The only alarm/fault the system gets now is “Grid Lost” when the Quattros disconnect themselves from Grid. I’ve managed to resolve all other issues successfully. Solar and Batteries are running sweet. System AC Outputs circuit breakers all off until I can sort this last issue so client has changover switches to bypass system enitrely thankfully.

While planning next attack, I’ve noticed the Cerbo setting is enable for "Stop Generator when AC Input is available”. I could deselect this and try simulate a grid loss, then run the force run the geny and see if the cerbo see’s it as running or not to rule that out.

I could try temporarily adjusting the LOM settings to see if they have any effect on the behaviour on the inverters.

I could revert the firmware version on Quattro’s by a version or two previous and re-test system. I’m told by client the issue began on the same day not long after the installer had left site after finishing their “system service”, when I’ve asked the client the believe the installer had updated the Quattro’s firmware which is why I suspect that being a potential culprit. I have no idea if they followed the correct steps to do the firmware update or loaded previous settings after the firmware update which may have been an issue with the new firmware registries..? Or they have re-entered all settings manually following firmware update like I would do but have selected the wrong combination of settings to previous but I’m doubting that they would have taken the long correct route and the former is more likely what they in fact did.

The setting was found selected for “use AUX 1 as disable feed in signal” which in the case of the INV master unit this AUX port is not wired up or in use so this could also be de-selected and monitored with a test run of Generator?

Following reviewing of INV settings, I would like to temporarily adjust the settings for the Frequency Start and Start Delay time which is set at 50.4 Hz at 0.08 seconds?

If none of the above identifies the cause by changing behaviour I thought of doing this:

Swap the Grid and Generator supply cable connection points locally at each inverter and make the Generator supply as priority instead of Grid, so move Geny to AC in 1 and grid to Ac in 2, changing the settings within inverters to suite this configuration. I imagine the Generator will still be set to only start at a set battery SOC % regardless of its input terminals location once these are changed within settings and the Grid will be the backup supply, albeit an unreliable one. It would be a good test to see how the inverters behave as the Generator supply is consistent, if the behaviour changes for the better this would prove the INV’s are functioning fine and simply reacting to a poor Grid supply, this will eliminate the suspicion of any hardware issues with the inverters and avoid removing and stripping them apart. If the issue persist after that then maybe the Quattro’s have damaged measuring circuit boards or failing internal relay/contactors on the AC in 1 & 2 terminals due to bad Grid quality over a period of time. (The grid AC in 1 did go down to 120VAC for a few hours and briefly even lower before the system experienced the perfect storm event, according to VRM graph data, but the Inverters would have disconnected from the Grid and protected themselves when this occurred), I am leaning more towards a firmware or settings issue at the moment)

The client wants to be as off grid as possible so this might be a way forward if it does work.

Has anyone experienced an issue like this and what steps did you take to resolve the issue?

Going to attempt a warranty claim and investigation with BYD as I’ve exported all the BYD BMS logs as required with screen shots of all findings. Uncertain of if the BYD BMU has a built in Battery deep discharge prevention SOC% as couldn’t see anything of the sort.

Any advise would be well received.