I have an issue with the backup generator when charging with 6 x Multiplus 2’s 5000/70. When the multi’s connect to the generator (AC IN 1 - Generator) the system load is put onto the gen(~9kw) without a problem. When the multi dumps the charge load on, it sometimes stops charging within ~1sec and sometimes I can get it to settle but within 10sec it starts to oscillate until it stops charging again.
The system worked flawlessly when it was only 3 multi’s but after increases in load, the system was upgraded by adding another 3 and bringing storage up to 85kWh Lithium(BYD). The generator without the multi’s is happy to run the property. Barely flinches when starting a 20kw air compressor while carbon arc gouging(3-phase welder at 500A) as well as the rest of the loads.
The generator is 120kVA with upgraded digital Governor and AVR (I install and tune these for work). It is a fixed speed governor(no speed droop enabled)
So far I have tried.
Weak AC input. Didn’t change much, AC input is not weak, 125A supply to the Board. Too strong?
Engine speed PID’s - Gets slightly better when tuned over-responsive, cuts out ~30sec.
Other PID Modifiers - eg. Increase P and/or I at different speed deviations
AVR PID’s - Didn’t seem to make any change.
AVR Voltage Droop - taking it down from 7% to 0 made it more frequently make to to the 10sec.
AVR Bias PID’s(from engine control) - to lock the voltage in place. No real help either.
Combinations of all above - I have spent hours trying different settings. I can change settings on the go so it would be a few hundred different combinations. Change something, retune, try all the other combinations, repeat.
I cant seem to determine why the Multi’s reject the charge. Frequency stays within +/- 0.6Hz even with the oscillation. Voltage 235-247V same deal. On the short 1sec ones the kW sometimes is steady(20kW not sure why its sometimes lower) but the current can fluctuate 25A with 6 to -8kVAr. What is the Multi trying to do?
I am out of ideas and I’m getting a little anxious with a Cyclone headed directly for me. Hence the test run of the Backup Gen. The previous configuration worked without touching anything, I assumed this would be the same.
I will upload an overview of the main AC system layout and some graphs from the generator but please note the Current, kW and kVAr will be out by a factor of 10(divide the number by 10), It gives me better resolution and finer control, doesn’t change how it performs.
No I have not tried Dynamic current limiter. I wish to charge at full power so that never occurred to me. I did try limiting the charge current to 250A and it made no difference.
Only ESS in the assistants. I hope im correct in assuming that its not in use when the AC IN switches to generator
You have a multi, so single AC input. You can’t use ESS with a genset, it causes issues and is likely the source of your problem, or contributing to it, hence why it isn’t supported off-grid.
Would suggest you remove it and run some tests again.
On a quattro you have two inputs and the gennie can be setup to be ESS friendly, that is the only supported way of doing it.
The dynamic limiter just eases the load onto the generator, some times helps, and is worth a try.
Something else to check is that all your inverters are balanced, parallel installs can be tricky, and variations due to age and wiring can cause some weird behaviour, such as one unit pushing power backwards, it is good practice to get a clamp meter out and check each IN/OUT under load to see they are the same.
Obviously doubling your ability to charge and raising the CCL fairly dramatically will put a greater burden on the generator.
I do have it setup to togle the AC IN input type between Grid and Generator depending on the situation. I will try removing ESS anyway, i will not be impressed if I have to replicate its functionality externally but at least theres a path to a working system. I will give the dynamic limiter a go as well. Anything is worth a try.
I did run a clamp meter over the system when first installed. I ran matched length cables as well. From memory the worst deviation I could find was 0.8A. I cant remember the test situation exactly but I deemed it acceptable at the time.
Dynamic current limiter did not change anything at all. I have not been able to remove ESS yet something about a firmware mismatch even though I downloaded a fresh remote config file. Now I have that little gem to deal with. I have run out of time to run the generator without waking children. To add to my frustration.
I will update all the firmware’s and reconfigure from scratch and try again in the morning.
If ESS is causing the issue isn’t this something Victron should be fixing? If the input changes to generator shouldn’t ESS be disabled. Is there a use case where someone would want ESS with a generator? Without ESS all I have is a UPS. Without ESS I can’t even externally control charge/invert. I’m at a bit of a loss of what to do.
ESS works in a specific way that is not generator friendly - LOM detection and pushing/pulling power to/from grid, amongst others.
So, they are clear not to use it off-grid (switching a single input system from mains to generator is effectively off-grid), and that if genset backup is required, that the dual-input quattro is to be used, where the secondary input can be configured to suit the gennie, something that is more difficult to do on a single input that needs to be configured to be both grid and generator friendly.
You could also switch the inverters to charger only mode, where power will be passed through and the batteries can charge, but the inverter will be inoperable.
Nodered allows some ESS-like control without an assistant. Plenty of examples available.
The assistant itself primarily supplies the ability to export, to run schedules, enforce some limits etc.
Ultimately this comes down to system design, and your current architecture does not conform to best practice, nor the design guides.
It is strange you are getting firmware mismatches, so it is a good idea to start from a clean sheet, and step through the whole setup carefully and in the correct order to eliminate any funnies that may have crept in.
When I can, im going to try disabling LOM in the hope that works. I have external anti-islanding so I just need to feed back to an aux in to Ignore AC IN if the bus is hot when the grid is cold(i guess ESS will not be happy so not sure what that will do). And then remove ignore when the Gen is on the bus once AC IN is set to generator.
Unless there is anything I’m missing, I feel it should stand a chance. Do you know if ESS has some kind of LOM built into it?
As long as this cyclone isnt to wild tomorrow I’ll be generator testing.
ESS requires LOM and a suitable grid code.
Generators would usually not have LOM enabled and would be set to grid code of none, sometimes other.
None would effectively disable ESS.
Well thats disappointing. Is there any way to bypass that?
I actually have 3 levels of LOM for the generator. The generator has it itself but I set it quite wide. The grid feeder managment has it as well but the ufd.m33 is the standalone anit-islanding.
ESS works so well for normal use its worth every chance to keep it.
First step is confirming it is ESS, and that is best tested by disabling it.
You can disable LOM on the inverter, but this requires the grid password and we may not share that here.
Some fiddling with the config of the grid tab should help narrow down the issue.
If it is indeed ESS, then you are a bit stuck and would need to brush up on your nodered knowledge to recreate some of the functionality with dev work.
With immense happiness I have been able to get the system working.
With ESS removed the surging still remained as before but it no longer disconnected from the generator which made tuning immensely easier. Took some time but in the end it required a linier increase in Proportional and Integral with load, increasing the responsiveness as load increases. Once stable I tested with Dynamic Current Limiter and Weak Ac Input turned off and it remained stable. I re-added ESS and the system is stable at full charge power. Starting the big air compressor and a few pumps did not upset anything. Great Success.
The system also doesn’t mind the generator synchronizing and connecting the grid once the grid is available again. The multi does disconnect when the generator ramps down to 0kw over the grid and opens the grid contactor but I guess I cant win them all. I’m assuming the sudden voltage shift or LOM doing its job. Seems to be happy connecting to the grid then disconnecting the gen contactor which is good for a seamless transition when the grid is restored.
Thank you Nick for helping me work through this. I think the impeding storm got me a bit flustered and was trying to rush a fix.