Isolating Transformers, The Earth wire connection conundrum

[goboatingnow]

This discussion has surfaced from time to time , from 2008 onwards , especially as S/V Jedi has described his system.

The figure below gives my best understanding of what most would regard as best practice, and outlines the differences between EU boats to ISO 13297 and US boats to ABYC ( including the use of whole boat RCBs(ELCIs))

Firstly there is teh issue of nomenclature

'GND', 'Ground' , etc is a common return path for current , it may or may not be at earth potential

'Earth' , a point at earth or near earth potential

'Grounded'. usually synonymous with 'earthed '

'RCD" Residual Current Device , ( GFCI, ELCI etc)

'RCBO' an RCD and a overload breaker combined ( RCCB) This is UK terminology

'AC Protective Earth Wire' the common building wire at earth potential





This drawing represents an amalgam of US/EU practice. IN the EU, a shoreside RCBO is virtually always present certainly on marina pillars in recent memory.

The whole boat RCD ( usually RCBO in EU) is now effectively mandatory in both US and EU boats.

I have postulated best AC panel practice, which shows double pole circuit breakers on the AC feeds to the boat, optionally one could also have as is common in EU, a RCD, to provide protection for 'local' AC generators. equally a protective wire is run connected to the earth pins of outlets and teh cases of appliances in order to trigger breakers and RCDs if either side of the secondary touches these devices.

without an isolating transformer, we have two major differences

• DC negative is not commonly bonded to shore power earth wire
• Underwater metals are not bonded together and connected to the boat GND.

In teh US a deliberate AC seawater earth is sometime installed, but typically the bonding system and engine/prop shaft produce an effective GND to EARTH connection.


( one can see of course why US boats have huge issues with impressed corrosion, but thats an aside)

What I want to address here is the floating /non floating system as a result of installation of an isolating transformer

In ABYC solutions, where a local seawater earth is effectively established, the major safety factor of an isolating transformer, ie that earth is not in the fault path is removed. BUT, there is a robust debate, that isolating power transformers should not have floating secondaries. for example the common 110V tool transformer on UK building sites, is a centre tapped 110V secondary , with the centre tap connected to primary earth wire. Hence earth is in the fault return path but the max voltage that can be experienced in such a fault is 55V.

So we have a debate unpon installing an isolating transformer

1. A true floating AC output , no neutral, both are HOT, but earth is not in the return path, no danger to boat occupants if they touch one or the other HOT wires.

2. The inadvertent establishment of seawater(earth) referenced secondary output, in effect producing boat L & N . This could occur due to dampness, teh protective earth coming in contact with underwater or engine/prop shaft

3. The deliberate attempt to establish not only a local boat 'Ground", but to attempt to earth that ground, via underwater metals or even a hull plate.


Its clear that (3) is the preferred ( ABYC) US approach, yet such an approach adds earth as a fault path. ( note not a fault path TO earth). with this system if the secondary HOT ( L) touches anything on the boat and the user touches that, then a fault path to the neutral side of the transformer exists via seawater which causes a shock. ( this assumes the 'thing' is not connected to the protective wire).

What I dont understand is , by deliberately defeating the isolation it reintroduces the swimmer in freshwater risk, as fault currents can now flow through seawater , ie out of one underwater fiting and back through another, especially if the protective wire is poor made or connected.

ABYC, does not like (1) above and to be fair , several other code bodies do not like floating 1:1 POWER transformers. Why?

(a) a concern for excessive common mode voltages, but in practice the secondary does not float excessively above earth potential due to leakage resistance and capacitance .

(b) A rather strange concern ( as Calder as described) of the floating HOT touching an appliance, and raising the potential of that appliance unbeknownst to teh occupant, then a subsequent fault on the other HOT causes the occupant to be inserted into the circuit.

Now this is rather a bizarre claim, and seems to completely ignore two things
(a) A RCD on the secondary side covers this
(b) double pole breakers and the presence of the protective fault wire, will mean a trip will occur in most cases.

I have seen some fault sceanarios debated where tingling can be experienced on flating secondaries , because of stray capacitive connections causing low fault currents to return to the secondary 'neutral'

So an RCD and double pole CB protected secondary does deal with the secondary HOT to appliance ( or hull) issue, in that if the rather bizarre double fault occurs, then the RCD will sense the fault current returning via the protective wire. Thats assuming the CB hasn't tripped anyway.

SO while I can understand the theoretical ABYC concern, it does seem to effectively predate RCDs and proper protective (earth) wiring.

ABYC in effects , defeats a claimed safety feature of isolating transformers, to simply deal with the galvanic corrosion issue.

S/V Jedi on the other hand , went to a lot of bother to ensure that the safety aspect, ie no earth return path was maintained in his AC system.

some misc things

ISO 13297 offers no help by the way as it ignore isolating transformers completely

UK codes for liveaboards, requires equipotential earth throughout the boat, ie a polarising traffo is all that can be installed. earth must be continous from primary to secondary and out to shore.

Interesting in many European countries Neutral is floating, ie it is not at or near earth potential, Hence maintain L,N symmetry is not a priority and in fact the mains sockets are not polarised. ( resulting in lots of reverse polarity lights when a US boat plugs in , whereas reverse polarity isnt often detected on European boats). This obviously requires double pole circuit breakers to provide overcurrent protection in a floating neutral


Which in reality is right.

(ps Ive left the case earth connection issue for another day)

dave

[transmitterdan]

Just did a cursory look at your drawing but it looks right to me. Will study more details tonight.

[Zil]

Why not have the "protective ground" on the drawing right connect through to the E (earth) on the drawing left.?

[transmitterdan]

Originally Posted by Zil

Why not have the "protective ground" on the drawing right connect through to the E (earth) on the drawing left.?

Because that can cause to galvanic corrosion of the metal bits of the boat. And because there is no need for an earth reference inside the boat.

[transmitterdan]

Originally Posted by goboatingnow

In ABYC solutions, where a local seawater earth is effectively established, the major safety factor of an isolating transformer, ie that earth is not in the fault path is removed.

dave

Dave,

I did not understand this statement. What safety feature of an iso transformer has been defeated by your sketched hookup?

[RaymondR]

Why is it thought necessary to connect both the AC and DC systems together at a common earth. I have a steel hulled vessel and having suffered an electrolysis problem am in the process of insulating all DC systems from the hull?

[goboatingnow]

Originally Posted by transmitterdan

Dave,

I did not understand this statement. What safety feature of an iso transformer has been defeated by your sketched hookup?

The fact that earth is now in the return path of the output of the secondary. whereas with a floating output , earth cannot be used as a return path to the secondary

dave

[goboatingnow]

Originally Posted by RaymondR

Why is it thought necessary to connect both the AC and DC systems together at a common earth. I have a steel hulled vessel and having suffered an electrolysis problem am in the process of insulating all DC systems from the hull?

Its ABYC practice, in Europe , under ISO 13297, as long as a whole boat RCD is fitted there is no need to connected AC earth to DC negative,

The stated aim of that connection is to ensure that any metal item onbard that isnt in the AC circuit , but is in the DC circuit , will casue the AC fuses to blow if a hot AC wire, connects to a DC device.

IN Europe the RCD does that protection.

And yes in a steel boat , a fully floating AC and DC system bot interconnected is teh best best, even if I believe it doesnt comply with ABYC.

With whole boat RCD, and ensure really good wiring standards.


dave

[transmitterdan]

Originally Posted by goboatingnow

The fact that earth is now in the return path of the output of the secondary. whereas with a floating output , earth cannot be used as a return path to the secondary

dave

Maybe you could make another version of the drawing with a stick man showing how he could become injured. That would help me understand this point.

[goboatingnow]

Originally Posted by transmitterdan

Maybe you could make another version of the drawing with a stick man showing how he could become injured. That would help me understand this point.

Note, in the drawing below, The earth wire ( protective wire) is always connected to Hot2, or L .

[RaymondR]

Since the function of the earth wire in the AC circuit is to provide a high reliability current path to ground in the event that the case/housing/enclosure of a relatively high voltage AC appliance becomes connected to a high voltage, current carrying conductor, that is, either of the live conductors in the power supply to the appliance, it appears that the earth on the secondary side of the transformer in your sketch should be either connected to that on the primary or have it's own a return conductor to the earthing point along with that on the primary side. The AC earth(s) should not be connected to any possible source of DC current as this may cause electrolysis, and a consequent reduction in the efficiency of, the current path into the earth stake, or earth wires connection to it.

Many years ago I purchased a TV set in Asia and on returning to Australia was instructed by customs to take it to the local electricity authority for testing. The technician tested it and explained that the power supply put DC current back into the AC circuits earth line and that a deterioration of the earthing lines connection to ground by way of electrolysis had been observed where these TVs had been used thereby creating a dangerous circumstance.

[goboatingnow]

Originally Posted by RaymondR

Since the function of the earth wire in the AC circuit is to provide a high reliability current path to ground in the event that the case/housing/enclosure of a relatively high voltage AC appliance becomes connected to a high voltage, current carrying conductor, that is, either of the live conductors in the power supply to the appliance, it appears that the earth on the secondary side of the transformer in your sketch should be either connected to that on the primary or have it's own a return conductor to the earthing point along with that on the primary side. The AC earth(s) should not be connected to any possible source of DC current as this may cause electrolysis, and a consequent reduction in the efficiency of, the current path into the earth stake, or earth wires connection to it.

.

Secondary Isolating transformer earth to my knowledge is never connected to shore earth. Its connected to the secondary neutral and optionally earth.

As per ABYC. AC earth is connected to DC neg , whereas under ISo 13297 for euro boats , once a whole boat RCD is in place there is no need for n AC DC bond. Yes you are right the presence of such link, is a major cause of impressed current corrosion and is the major reason the issue is far more prevalent on the US.


Dave

[CarinaPDX]

Dave - Thanks for this thread - Just what I needed.

At last year's boat show I had a long discussion with the ABYC rep about my electrical installation. I was not thoroughly convinced of his arguments, and your thread covers my concerns.

I have purchased an isolation/stepdown(up) transormer, and plan to connect the primary safety ground to the transformer's shield (only) so that is where the shore side ends. I have a 2-pole 30A circuit breaker as well as a 30A ELCI (RCD plus 2-pole breaker). The ABYC recommends putting the ELCI between the shore power input and the transformer, and have the simple breaker on the secondary side. This seems odd to me: the chance of leakage in the few feet of cable (ELCI>XFMR) or inside the transformer is very small. OTOH with the recommended connection of the secondary neutral to the secondary safety ground and the ship's seawater "earth" connection does create the opportunity for electrocution through a seawater return path, which would seem to benefit from the ELCI instead of the recommended breaker on the secondary side.

What do you think?

Greg

[RaymondR]

Dave and Greg, this thread has certainly started me thinking about AC on boats and is something I should have thought about a good bit more than I have.

The AC in my boat was installed on the assumption that I would only ever connect to shore power for AC. As time has gone by this has become the least used source as I now permanently live aboard and have added an inverter and an engine driven AC generator.

My understanding of multipole AC systems is that the sum of the voltages at the neutral is zero and that at the power station and the various transformers neutral is generally earthed. I know in my house wiring there are a number of earths and I suspect that both the neutral and earth circuits are both earthed as a precaution.

However I am not sure that in the instance of the secondary circuit of an isolation transformer this situation should prevail as it appears that the assignment of earth to one or the other of the outputs of the secondary is arbitrary and whichever you choose could result in making the earth system live.

I am not sure if it was on this forum but I recall a thread regarding a number of unexplained drownings where it was thought that swimmers coming into contact with a vessel may have experienced a stunning electrical jolt and subsequently drowned.

However I can see the logic of connecting the secondary side earth to the transformer secondary if there was an earth leakage breaker in the circuit.

[transmitterdan]

Edit: Dave, thanks for taking time to post the pictures. It helps understand the topic very well.

I believe residual current devices or whatever the nom de jour is today are an excellent safety feature. I would never trust my life to one. Likewise I would never trust my life to a circuit breaker.

In the EU configuration the entire A/C power distribution system is not referenced to any other metal part of the boat or DC system. Therefore, the hot and/or neutral (and thus safety wire) can take on literally any potential limited only by whatever random leakage current happens to be present. Thus the safety ground could be 220V or more relative to the DC power system and other metal bits on the boat. A person on board touching AC safety wire and DC ground could get a tingle or a static discharge but probably not a lethal shock.

The ABYC think this is a bad idea and I agree although there is room here for disagreement. ABYC also think that if AC hot is accidentally connected to DC + or - that this should trip a breaker. In many older boats this possibility exists because AC hot wires are colored black and DC - wires are also black causing some owners to make the mistake of connecting them together. In the ABYC configuration this will trip the AC circuit breaker. Modern color coding tries to remove this confusion but there are many older boats in service with the old color system.

In the EU system as I understand it connecting AC hot to DC ground will cause no observable effect. But unexpectedly the safety wire will now be 220V relative to DC and other metal boat bits. This requires the residual current breaker to trip if a person should come across a DC wire or bus and AC safety. In the ABYC style it is virtually impossible for DC - and AC safety to ever have lethal potentials and it does not require an RCD to make it so.

I think this is the crux of the arguments but there may be additional ones that are more subtle.