Galvanic Isolator

[DotDun]

Originally Posted by fairbank56

That drawing no longer exists in ABYC standards. It was in a previous version of standard E-8 which is no longer used. E-8 was incorporated into E-11 7 years ago. The current version of E-11 does not have "method 2" in the Charles manual. I wouldn't worry about it. I don't see the need to have the shield connected either. The RCB's are a great idea and I hope they become standard in the U.S. but even with them, you can still have up to 30ma of leakage before they trip.

Eric

I don't have any experience with toroid transformer failures, is there a failure mode that demonstrates a primary to shield of <30ma?

I felt comfortable with the connection design as I measured ~400mv between the boat ground and shorepower ground. I believe that would support tripping the RCB is the event of a failure.

[osirissail]

Check to see exactly where and to what your dock power ground wire is connected to inside the "isolation" transformer - same with the ship's ground system wire. They may be inadvertently connected together electrically. That would allow a faulted/leaking shore power ground to pass currents through to your boats ground/zinc system.

[DotDun]

Originally Posted by osirissail

Check to see exactly where and to what your dock power ground wire is connected to inside the "isolation" transformer - same with the ship's ground system wire. They may be inadvertently connected together electrically. That would allow a faulted/leaking shore power ground to pass currents through to your boats ground/zinc system.

On my boat, the shorepower ground doesn't make it to the isolation transformer, it's not connected at the inlet. (neither is N, L1 & L2 are the only connections).

BTW, I also run my genset thru the isolation transformer (via a transfer switch) to help cleanup the generator noise.

[osirissail]

On a 120VAC or 220VAC 1:1 isolation transformer, L1 is the "hot" from shore power on the shore power side of the isolation transformer and L2 is the "neutral/return." Same thing on the ship's side of the isolation transformer.
- - On a "step up/down isolation transformer there will be a third terminal that is the mid-tap. It becomes the "hot" depending upon whether you are stepping up or down.
- - If your shore power "ground/green or green-yellow" wire is not connected to anything in your vessel you have an "ungrounded" shore-power system.
- - A galvanic isolator effectively does this in the shore-to-ship circuit but allows stray voltages in the ship to travel back to the shore power ground system.
- - Without a galvanic isolator a terminated shore power ground may or may not be a problem. If you have an AC grounding plate on the exterior surface of your hull dedicated to providing an "earth ground" for your ship's AC system there is no problem. However, it there is no "earth ground" for your internal AC system then you might have problems with GFCI's or other mandated ground fault interuptor devices.
- - Connecting your internal AC system "earth ground" to the DC system ground (e.g. the engine/prop shaft) can often allow stray AC voltage to impress itself into your DC system and cause havoc with some DC instruments/equipment. Not a good idea, but a very commonly found situation. So long as there are no stray AC voltages leaking to ground all will be fine.

[osirissail]

Here is a color coded example of a 1:1 isolation transformer and its wiring. Notice that the shore power ground (green) wire does go to the isolation transformer. This is necessary to provide protection should the shore power side winding develop a short. However, the ship's side ground is attached to the same terminal as the ship's side neutral (white). Ne'er the twain (grounds) shall meet.

[DotDun]

Originally Posted by osirissail

On a 120VAC or 220VAC 1:1 isolation transformer, L1 is the "hot" from shore power on the shore power side of the isolation transformer and L2 is the "neutral/return." Same thing on the ship's side of the isolation transformer.
- - On a "step up/down isolation transformer there will be a third terminal that is the mid-tap. It becomes the "hot" depending upon whether you are stepping up or down.
- - If your shore power "ground/green or green-yellow" wire is not connected to anything in your vessel you have an "ungrounded" shore-power system.
- - A galvanic isolator effectively does this in the shore-to-ship circuit but allows stray voltages in the ship to travel back to the shore power ground system.
- - Without a galvanic isolator a terminated shore power ground may or may not be a problem. If you have an AC grounding plate on the exterior surface of your hull dedicated to providing an "earth ground" for your ship's AC system there is no problem. However, it there is no "earth ground" for your internal AC system then you might have problems with GFCI's or other mandated ground fault interuptor devices.
- - Connecting your internal AC system "earth ground" to the DC system ground (e.g. the engine/prop shaft) can often allow stray AC voltage to impress itself into your DC system and cause havoc with some DC instruments/equipment. Not a good idea, but a very commonly found situation. So long as there are no stray AC voltages leaking to ground all will be fine.

My IT is connected as shown in post #27, with a few exceptions.

1) Between the shorepower inlet and the RCB/ELCI I have a 1-OFF-2 transfer switch to select between the shorepower and genset.
2) My transformer (which I had custom made) is rated at 6250va and has 2 primary and 2 secondary windings. My primary windings are in series, hence 240vac primary (capable of parallel with 120v each). My secondary windings are parallel at 120v each. The 2-pole primary ELCI is rated at 30a and a single pole protection on the secondary is 50a. Also there is a 120C degree internal thermal breaker on the transformer.
3) I do have the neutral of the secondary connected to boat ground (and I'm still thinking about that).

The output feeds the distribution box with 50A of 120vac where there is a RCB main breaker. I also have a RCB on the output of the inverter.

This supports my (2) 16k btu ACs, a 750w water heater, while making coffee and toast at the same time.

[osirissail]

Dotdun - "My IT is connected as shown in post #27, with a few exceptions." Wow, small picture, had to put on my super magnifier eyeglasses.
- - Everything is quite "standard" for a 220/110 or 240/120 step down isolation transformer - except - the unprotected shore-power side winding shield. Insurance companies, surveyors, and other various regulators can get rather picky about "unprotected" high power systems. If for some unknown reason an overload or short develops in the shore-power side of the isolation transformer there is no protection to trip the shore power circuit breaker. At best you would have a "hot case" which if touched might be lethal to a human or worst a fire could develop before you can manually disconnect the shore power.
- - A suggestion would be to use a 3-pole transfer switch for the genset/shore-power selector. The third pole would be used to transfer the shore-power ground to the genset ground. The isolation transformer should have 2 internal shields: one for the shore-power side and a separate shield for the boat-side. These shields are internal and not connected to each other. They are there only to serve as a shorted winding path to trip the supply circuit breaker.

[DotDun]

Originally Posted by osirissail

Here is a color coded example of a 1:1 isolation transformer and its wiring. Notice that the shore power ground (green) wire does go to the isolation transformer. This is necessary to provide protection should the shore power side winding develop a short. However, the ship's side ground is attached to the same terminal as the ship's side neutral (white). Ne'er the twain (grounds) shall meet.

I understand. In your example, the shorepower circuit breaker would blow if the primary windings short to the shield.

In my case, I have the same level of protection except I'm getting it with the RCB vs. connecting the transformer shield to shorepower ground. If my primary windings short to the shield (which is connected to boat/seawater ground), my RCB will blow.

In my case, there is never a chance of the boat ground meeting shorepower ground. That was a goal!

[DotDun]

Originally Posted by osirissail

Dotdun - "My IT is connected as shown in post #27, with a few exceptions." Wow, small picture, had to put on my super magnifier eyeglasses.

Click on the picture! It'll get bigger!

Did you miss the "GFP" on the main breaker?

[osirissail]

I did click on it and even opened the picture as an attachment. It is still too small or too low a resolution to be able to clearly read the labels.
- - But anyway, all that is important is that the shore power supply breaker be able to automatically trip if a short in the primary winding (shore power side) occurs. An overload situation is a bit different as heat/fire can start in any of the components - cable/connectors/transformer. If that is covered then I see no problems.
- - There are many creative ways to cover the problems of short/overload protection besides the "classical" way set forth in various "codes." But if you have insurance or local regs that mandate following the "classical" way then you are sticking you financial butt out there. Nothing wrong with that, you just need to realize it. If you ever intend to sell the boat then "creative" solutions can present a major problem.

[DotDun]

Originally Posted by osirissail

- - There are many creative ways to cover the problems of short/overload protection besides the "classical" way set forth in various "codes." But if you have insurance or local regs that mandate following the "classical" way then you are sticking you financial butt out there. Nothing wrong with that, you just need to realize it. If you ever intend to sell the boat then "creative" solutions can present a major problem.

I do understand and appreciate your thoughts!

My only comment is my IT is configured per the ABYC E-8 specification that allows for no connection from shorepower ground to the transformer shield. This is the same design supported by Charles with their IT installations (see post #33).

FWIW, RCBs are accepted and used widely on European boats.

Thanks for the thoughts!

[GordMay]

Originally Posted by DotDun

... my IT is configured per the ABYC E-8 specification that allows for no connection from shorepower ground to the transformer shield...

FWIW:
ABYC "E-8" has been superceded by "E-11".

[SV Demeter]

Great thread here. Had to re-read a few bits but very informative!

[fairbank56]

http://www.paneltronics.com/Atimo_s/...I_Sept2010.pdf

[DotDun]

Originally Posted by osirissail

There are many creative ways to cover the problems of short/overload protection besides the "classical" way set forth in various "codes." But if you have insurance or local regs that mandate following the "classical" way then you are sticking you financial butt out there. Nothing wrong with that, you just need to realize it. If you ever intend to sell the boat then "creative" solutions can present a major problem.

Osirissail,

After more thinking about scenarios that could cause trouble with my installation, I came up with one that could be bad.

With my current configuration, shorepower ground disconnected at inlet, if my boat was on the hard and connected to shorepower AND there was a transformer fault (primary to shield), the boat ground is now hot. There would be the possibility of 30ma for 100ms for someone that touched the ground plate on the hull. This would not happen with the boat in the water as the water would provide enough ground to trip the ELCI.

I'll take your suggestion and put a 3 pole transfer switch in and extend either shorepower or generator ground to the transformer shield depending on switch postion. Since the transformer is mounted in a plastic (vs. metal) enclosure, I don't have to worry about the case/chassis ground.

Thanks again for the thought provoking conversation.