New Marina Pedestal tripping, is it the inverter?

[Circledog]

Hello Everyone,
I have a magnum 2812 inverter/charger. With the new electrical code at Florida Marinas the ground fault is set to trip at 30 ma at the pedestal. I have gone through my system and does not trip when the inverter is disconnected. It will also not trip if I power the inverter, before activating the breaker that sends the power from the inverter to the panel. If I try to leave the breaker that allows the inverter back to the panel on when powering up from the pedestal, it trips the pedestal. I know others have said that if any circuit that shares a neutral bus bar from before and after the inverter will cause this. I am under the impression this should have caused a trip when the inverter was taken out of the equation.
With that, it has led me to believe that maybe the neutral to ground relay is slower than the pedestal when power is applied, and causing the trip. This is why when the inverter is first powered up, before the breaker is turned on to allow the power through, it has already activated the neutral to ground relay.
What are your thoughts? Am I on the right track, is it a shared neutral bus issue, or something else? Any help would be greatly appreciated.
Thanks,
Paul

 

[Bmarler]

These dang new rules on pedestals are tough. You must be absolutely perfect or those nuisance trips occur.
You’d need to provide more information about the wiring scheme for anyone here to be helpful.
Gfci trips on imbalance or power on the ground wire. Inverters do weird stuff when you power them up. They sync with the power company so you don’t miss a blink when power cuts off. There’s something about that syncing that bothers me. It makes my isolation transformers much more noisy than if the inverter is disabled.
Speaking of isolation transformers, I installed two in anticipation of the new dock power regs. Haven’t tripped a gfci since.

 

[SteveK]

Actually what I see happening is that when power passes through the inverter all neutrals are connected to the pedastal neutral. But when you turn on the inverter not all neutrals are behind the inverter, one or more neutral is still directly connected to the pedestal and causing the trip at pedestal.

 

[O C Diver]

The first thing I would check is the boat's ground and boat's bonding system. The bonding system should only be tied to the negative battery, not the inverter chassis. Basically the bonding system is usually tied to the negative buss which feeds the inverter. Check your owners manual for specific location of shore power ground and if there is a ground wire going back to the ground buss, and is the inverter ground buss separate from the fuse panel groud buss.

I had a 2812 on my boat that was fine with sensitive GFCI breakers, but everything has to be correct. Unfortunately I don't remember exactly how I wired it 8 years ago.

Ted

 

[SteveK]

The bonding system should only be tied to the negative battery, not the inverter chassis.

however ABYC wants the AC ground and battery negative are bonded. In effect whether done direct or not, it is still connected. From my 2012 Magnum

The AC output neutral conductor and the DC negative conductors are not connected (bonded)to the inverter chassis. Both the input and output conductors are isolated from the enclosureand each other.

DC Equipment Ground Terminal – this connection is used to tie the exposed chassisof the inverter to the DC grounding system.

Only an internal failure of the inverter would apply power to the chassis, thus the bonding system.

 

[Patemey]

Separate the neutrals that go thru the inverter to their own buss bar, and the neutrals the do not go thru the inverter to their own buss bar. That fixed mine from tripping the breaker.

 

[Bacchus]

Gfci trips on imbalance or power on the ground wire.

My understanding is the first part is true but GFCI /ELCI do not monitor ground. Imbalance between hot & neutral for 125V and imbalance between L1, L2 & neutral for 250V. If that's incorrect I'd appreciate an education from someone better educated than I am.

 

[Bmarler]

My understanding is the first part is true but GFCI /ELCI do not monitor ground. Imbalance between hot & neutral for 125V and imbalance between L1, L2 & neutral for 250V. If that's incorrect I'd appreciate an education from someone better educated than I am.

Yes,technically they don’t monitor the ground wire itself. But any leakage of power to the ground wire will create the imbalance needed to trip the gfci. A neutral to ground can trip them as well, even though those wires are bonded at the source. It’s pretty easy to make a 5ma imbalance.

 

[SteveK]

HERE is a good article from Panbo on topic
But the quoted part below has me scratching my head. The only thing in my mind an output breaker would do is isolate power to the outlets that can operate from inverter. If the breaker is ON (closed) inverter off, with shore power pass thru how did it solve the problem. It can only isolate when OFF (open).

The first step I suggested to allow the boat to draw power from an ELCI equipped pedestal was to add a breaker to the output side of the inverter. This allows isolating the inverter from the AC panel which obviates the ground to neutral issues when connected to shore power. It also allows a quick and definitive way to ensure no power is flowing from the inverter onto the panel.

 

[twistedtree]

Just to be clear, it's the shore breaker that's tripping, not a breaker on your boat, right?

I think you are on the right track WRT the grounding relay's response time. When the inverter is not passing through shore power, the bonding relay is closed to ensure that the inverter output neutral is grounded. When the inverter switches to pass through shore power, it opens the bonding relay and connects the shore neutral (which is grounded on shore) through to the load neutral. If the relay is too slow to open, you have a period of time where the neutral and ground are connected on the boat as well as on shore. This causes the neutral current to split between the neutral and ground because they are parallel paths back to shore. The RCD detects the missing neutral current and trips. I don't know what the trip delay is in an RCD, nor the expected open/close time of the bonding relay in the inverter, but I have heard of this being an issue.

The key clue that you picked up on is that the relay doesn't trip if you disconnect the inverter loads during the switchover to shore power, and turn them on after the switchover without a trip.

As a test, you could disable the grounding relay, assuming that is a configuration option. If things work correctly afterwards, then I think you have proven it's the cause. Magnum may also be able to tell you if this is a known problem, and whether there is a solution. It might be possible to use a faster external relay, for example.

I just checked and google says that an RCD trip time is 25 to 40ms, so the bonding relay needs to be faster than that. Say under 20mS to be reliable.

 

[twistedtree]

Oh, and do you have a galvanic isolator? I assume no isolation transformer, right?

 

[SteveK]

On my Magnum you have to physically separate the ground and neutral should you want that option. There is a connecter where the input/output wires are located, just pull apart.

 

[sunchaser]

We had the same exact issues when the new dock codes were implemented. First off we had the grounding system redone, likely similar to what OC mentioned.

Once pulling into a dock we shut off all the various 110 breakers including turning off the 2812 inverter. Then we plugged in, turned on shore power and one by one each breaker. Once all the breakers were on we then could usually have no trips when 2812 turned on.

A lot depends on how your incoming power is split with many different ways to do it. With an inverter bypass and a separate charger from the Magnum we had a few good work around options.

We intentionally didn’t want an IT and our systems rejuggling avoided that need. Good luck.

 

[Circledog]

Just to be clear, it's the shore breaker that's tripping, not a breaker on your boat, right?

I think you are on the right track WRT the grounding relay's response time. When the inverter is not passing through shore power, the bonding relay is closed to ensure that the inverter output neutral is grounded. When the inverter switches to pass through shore power, it opens the bonding relay and connects the shore neutral (which is grounded on shore) through to the load neutral. If the relay is too slow to open, you have a period of time where the neutral and ground are connected on the boat as well as on shore. This causes the neutral current to split between the neutral and ground because they are parallel paths back to shore. The RCD detects the missing neutral current and trips. I don't know what the trip delay is in an RCD, nor the expected open/close time of the bonding relay in the inverter, but I have heard of this being an issue.

The key clue that you picked up on is that the relay doesn't trip if you disconnect the inverter loads during the switchover to shore power, and turn them on after the switchover without a trip.

As a test, you could disable the grounding relay, assuming that is a configuration option. If things work correctly afterwards, then I think you have proven it's the cause. Magnum may also be able to tell you if this is a known problem, and whether there is a solution. It might be possible to use a faster external relay, for example.

I just checked and google says that an RCD trip time is 25 to 40ms, so the bonding relay needs to be faster than that. Say under 20mS to be reliable.

Thanks for the help.

 

[Circledog]

Oh, and do you have a galvanic isolator? I assume no isolation transformer, right?

Yes we have galvanic isolator(which I have checked and is fine), and no isolation transformer.

 

[Circledog]

HERE is a good article from Panbo on topic
But the quoted part below has me scratching my head. The only thing in my mind an output breaker would do is isolate power to the outlets that can operate from inverter. If the breaker is ON (closed) inverter off, with shore power pass thru how did it solve the problem. It can only isolate when OFF (open).

SteveK thank you for the article. Similar to the article I have breakers before the inverter gets to the panel. If I do the same as the article suggested, it powers up fine. It’s the power blips that are occurring after Milton that made this evident. Every time the power blips, the inverter is not quick enough to keep it from tripping when power returns.
I will check to make sure my pre inverter and post inverter circuits do not share the same neutral bus as another measure.
Thank you for your help.
Paul