question

Henrik avatar image

Solution: Three Phase Quattro System does not accept Generator on AC2 in while AC1 grid is present

I write this not as a question but as as a guide how to handle a three phase rotataion problem which occured at our three phase quattro system in Africa:

The system is based on 6 Quattros configured as "big three phase UPS", a large battery bank and a backup Diesel generator.

Normal Operation:

Occasionally the grid on AC1 IN fails, the systems switches over to the battery bank and as last action, a backup Diesel generator on AC2 IN is automatically started in case the grid failure is too long and the batteries are getting empty.

What happened?

Suddenly, the grid showed correct voltages on all three phases on AC1 inputs (!) but the inverters did not accept the input power and switched to inverter mode and kept draining the batteries.

Upon reaching the defined low level value in the GX the backup Diesel generator was correctly started but the Quattros did not measure any input voltages from the generator and thus did continue to drain the batteries until they were empty and all shutdown.

Solution:

Upon disconnecting the grid by opening the circuit breakers, the AC1 IN voltages went down to 0V and after restart of the whole system the Diesel generator was correctly detected on AC2 IN and system went up again.

Root Cause

After investing the case the following root cause was found: The grid operator has changed two phases by mistake during maintenance (=opposite three phase turn direction).

This leads to the situation that the Quattros "see" all three correct input voltages on all three phases on AC1 IN but all currents remain at zero because input not usable. System remains in inverter mode. At the same time, a started generator input power on AC2 is not detected by the Quattros because reasonable voltages from the grid on AC1 IN are present (but not usable!)


How to identify such a case:

If all three input voltages are present and in a reasonable range check first the voltages between all three phases. It must be in the 400V range, otherwise you don't have a correct three phase grid input.
If the voltages between all phases are correct then it is likely the wrong rotation direction not matching to your three phase system programming! --> Exchange two phases!

Emergeny Workaround

In order to enable the Quattro to use an alternative three phase power source like a Diesel generator on the other AC input while such a special grid rotation error exists, cut off the unaccepted power input source then restart the whole system and it will accept power on the alternative source.















grid3 phase
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

6 Answers
Alexandra avatar image
Alexandra answered ·

@Henrik

That is an interesting experience. Thank you for sharing it. Stuff like this keeps you on your toes (and sometimes scratching the head) for sure.

I do have a question though, did the GX not give a phase rotation error?

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Henrik avatar image
Henrik answered ·

@Alexandra

No, the Venus GX did not detect this error, there was neither the error highlighted in the GX console menu nor a respective alarm sent/logged. Because of this we searched several hours into various wrong directions.

I guess the rotation error detection does not work in the particular case for some reason. It is probably also quite unusual that during regular operation, one power input gets changed into opposite direction and the second remains in correct direction. Maybe somebody at Victron could try to reproduce this case.



2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Murray van Graan avatar image
Murray van Graan answered ·

@Henrik

Wiring the generator to AC 1 input would prevent this issue also. The Quattro will only accept the AC 2 input when no voltage is measured on AC input 1. Obviously, grid would then be wired to AC 2 input.

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

The manual actually states that the Generator should be on AC Input 1. There is another reason for this. You can only black start the system off AC Input 1, so if there is a grid failure and no DC connected PV, then you need to use the generator to start the system up and that can only be via AC Input 1. (Unless something has changed @Daniël Boekel (Victron Energy Staff) , @Guy Stewart (Victron Community Manager) ?)


One quirk about this is that Venus OS UI is worded such that it assumes the grid is on AC Input 1. So, in a roundabout way, the manual and Venus OS are contradicting one another.

Henrik avatar image
Henrik answered ·

We also noticed the contradiction in the VENUS manual which suggest to put the grid on AC1:

Option: "Do not run generator when AC1 is in use"

"This option is ideal for back-up systems where a Quattro has mains/grid electricity connected to its AC-in 1 terminal, and a Genset connected to its AC-in 2 terminal. With this option enabled, the Genset will only start after a mains failure."

This was the reason why we connected the grid to AC1.

So I propose to either change the option in the VENUS GX menu or solve it that way that such an error is somehow identified and clearly reported as an alarm.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

robert6569 avatar image
robert6569 answered ·

I have had similar a similar issue. If there is any voltage on the grid side of the master inverter, generator will not be accepted. I had to add a 3 phase window comparator, that also detects incorrect phase rotation, on the input.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

sa-victron-user avatar image
sa-victron-user answered ·

We just had a similar (but worse) problem. Two phases on Mains were fused together so we ended up with 2 phases coming down 3 lines. I.e. Our red and blue phases were identical in all respects and yellow was different. Another way of explaining this is that L1-L3 voltage was 0 instead on 380 volts despite all 3 phases having individual voltages of 230 volts. Victron did not accept the power as it was genuinely bad but also did not give an error (as an aside when phase rotation is reversed - which we have had in the past - Victron gives a perfect phase rotation error on our system - we did have to set this up as an alarm).

When the battery hit 25%, the generator turned on as expected but was not accepted on AC2 because there was still power on AC1 (even though it was not accepted power). Personally, I think this is a shortcoming in the current software.

We swapped the generator to AC1 and mains to AC2 as recommended and have figured out how to deal with a whole host of other challenges which we have (due to not all loads being on battery backup and an ATS between the generator and Victron. However, we are left with one challenge… When good quality mains comes back, the Generator does not turn off until it meets the requirements set (e.g. battery SOC returned to 80%) even though good quality mains is available. Does anyone have a clever solution for this?

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.