Isolating Transformers, The Earth wire connection conundrum

[goboatingnow]

Originally Posted by chala

Any appliance that requires to be earthed may leak or leak continuously. If that appliance cannot leak then the potential of that appliance will raise and so will anything else that is conductive and in contact with the appliance.

Given chala that you insist RCD s are three wire when in fact only ever live and neutral are fed through it. An RCD does not require ground to operate . Merely that current flows to the power supply , via a or any path that does not contain the RCD. That path may or may not include earth

Dave

[goboatingnow]

Originally Posted by chala

For an Electrical Engineer you really surprise me. You are supposed to know and understand the above. Not have to ask a forum of cruisers.

Gobo have you made up your mind, do little or do virtually nothing?

Whatever the answer, you seem to have found a reason for that “ bizarre claim”.
Maybe the ABYC knew that a RCD would do little or virtually nothing.

ABYC quite frankly has made poor show of AC on craft , is responsible for the considerable impressed current problems , discovered RCDs 20 years after the rest of us. Best ignored.

An RCD does little or virtually nothing as a whole boat RCD on the secondary of an it.

Dave

[goboatingnow]

Originally Posted by Saltyhog

The thing about these threads that I'm sure is very frustrating to many readers is who do you believe? Anybody can say what they like on this forum. However, as I like to remind the junior engineers in my department, though you may be entitled to your own opinions, you're not entitled to your own facts! To Chala, you are disagreeing with a number of electrical engineers with quite a bit of experience in the field. You also make statements which clearly show your lack of understanding. Perhaps it's just terminology differences, but if so, you need to fix your terminology. Some of us have designed power systems in safety critical medical applications (myself for one). Others just seem to have opinions based on what they've seen or based on their own misunderstanding of the subject. Dave (GBN), Jedi, and myself all see the relative merits of a totally isolated AC power system. ABYC is concerned with a much broader audience than electrical engineers. Their standards must reflect that. They must prescribe techniques that provide the safest approach for the broadest audience. This implies simple. In absolute terms, the totally isolated approach is the safest. In broader terms, some analysis may be required to determine that a potential problem exists or what to do about it. ABYC cannot prescribe analysis. They can only prescribe process. For example, if an isolated system yields "tingling" when touched, it's not really isolated (completely). There may be a problem with the transformer, but it's likely a wiring error. There is a leakage path somewhere that is going to require some analysis to nail down. That analysis is difficult to do via internet forum, and even more so via prescribed process. I believe that is the reasoning behind ABYC's standard. As the equipment and methods gain acceptance and standardization, I believe ABYC will come around.

As a follow on discussion, some of us are simply unwilling to add 60 lb. chunk of iron to our boats that is only useful in the planned somewhat rare occasion we are on dock power. For us, I think the best, safest method is whole boat RCD, GFCI's on every branch circuit, no ground to seawater. Opinions?

In the absence of a safety IT. I believe earthed appliances are a must , but there should be no AC to DC connection , whole boat RCD and no need for outlet GFCIs. This is the European legislated standard for a much higher voltage.

Yes an IT is hard to justify , but does remove earth faults , which are by far the most common form of shock , removes any danger of swimmer electrocution and completely removes impressed current corrosion

Dave

[goboatingnow]

Originally Posted by chala

Any appliance that requires to be earthed may leak or leak continuously. If that appliance cannot leak then the potential of that appliance will raise and so will anything else that is conductive and in contact with the appliance.

Electrical hogwash. An appliance can only leak to earth if earth is a return path to the source. In an IT earth is not such path. Common mode voltage is determined by many factors , but in practice is not an issue. If you don't believe me , buy a simple transformer and do some simple tests with a multimeter.

Dave

[goboatingnow]

Originally Posted by FlyingCloud1937

remember I've always said Potential!

With a transformer the output voltage is not referenced to ground - see diagram (a) below. There is no "return path" so you could (stupidly) safely touch the "live " conductor and ground and not received a shock.

I say "stupidly" as, while this arrangement is safer it is not safe unconditionally. This is because, if there is leakage or hard connection from the other side of the transformer to ground then there may still be a return path - as shown in (b) above. In the diagram the return path is shown as either capacitive or direct. If the coupling is capacitive then you may feel a "tickle" or somewhat mild "bite" from the live conductor. If the other conductor is grounded then you are back to the original transformer situation. (Capacitive coupling may occur when an appliance body is connected to a conductor but there is no direct connection from body to ground. The body to ground proximity forms a capacitor.)
SO a transformer makes things safer by providing isolation relative to ground. Murphy / circumstance will work to defeat this isolation.
This is why, ideally, an isolating transformer should be used to protect only one item of equipment at a time. With one item a fault in the equipment will probably not produce a dangerous situation. The transformer has done its job. BUT with N items of equipment - if one has a fault from neutral to case or is wired wrongly this may defeat the transformer such that a second faulty device may then present a hazard to the user. In figure (b) above, the first faulty device provides the link at bottom and the second provides the link at top.

Which we have said is protectable by outlet of appliance RCD. In practice it's an extremely unlikely event. I think Nick used the term moonrock. It could also be solved by a centre tap connected as a local appliance pseudo earth or fault wire ( not earth referenced) and breakers

Dave

[goboatingnow]

Originally Posted by FlyingCloud1937

Yep, that's my point we are on a boat floating in the sea. One day everything is fine and no faults

The next day a fault potential is there.

An ISO on a boat isn't about protecting against faults, it's a magnitude in trying to stop corrosion.

Lloyd

No that's ABYC s interpretation. It's not other code bodies interpretation. Your scenario can be simply protected by a combination of RCD on the secondary and outlet RCDs , providing the safety of both systems

Dave

[goboatingnow]

Originally Posted by chala

They do not have to believe any one they just have to comply with the regulations in force in their own countries. To peddle murderous idea in this forum is irresponsible.

Experience in what field? To me they do not look that bright they promote devices that they realize later will offer no protection and do not know how to connect them.

I hope they are satisfactory too many people have died in hospital out of electrical experiment.

Someone will say “not really ready for prime time”

There is nothing new about this method. I note that you are not a proponent of the isolation transformer. I have RCD on my boat and no ground to seawater and that for over ten years.

What specific law do you quote. ABYC is merely a recommendation , it's not code. European RCD and its associated ISO standards are law , would you like to tell us what they say about ITs. Would you care to comment on the fact that many European mains AC systems have unpolarised on almost floating neutrals.

Given your complete electrical mis understanding ( and your amusing 3 wire insistence for RCDs. ) I fins your second line deeply insulting

Dave

[goboatingnow]

Originally Posted by pbiJim

I would be interested to hear the response to this question from this gentleman's most vocal critics. I think that this is a valid & productive question. The correct answer to this question should be useful to anyone with an IT & a GFCI on their boat.

An RCD has TWO wires TWO , everyone understand. That's all that ever connected to it.

Dave

[goboatingnow]

Originally Posted by chala

I do not have concern against Isolation Transformer, I have concern about how people wire them up and how they may endanger other people. I really did not had the time to analyse and test your suggestion it is maybe a great idea but you will need to convince the Regulatory Authority that it is. You are quite right with your earth lamps but they where already superseded in the 1960 by Ground Fault Indication. I personally feel that in a ship where home appliances are used, RCD’s should be used. It is to late to go and investigate an alarm and find someone dead in the laundry.

What regulatory authority is that , since ABYC isnt one. And I have a copy of the ISO spec in front of me ( which is a regulation)

Dave

[goboatingnow]

I brought up this topic , the earth wire conundrum, because there are major differences of opinion , even amongst regulatory authorities , pseudo authorities , and people like Calder over the correct wiring of ITs in boats. Regularly they are installed incorrectly. ( like in front of a generator).

I recognise the advantages of an earth referenced secondary , but I also recognise the advantages of a floating output. Funnily one of the strange things given ABYC predilections is why they didn't specify a earthed 110v centre tap secondary , like the UK building site ITs. This at least reduces the voltage fault to earth to 55vac .

The only real issue with floating IT secondaries , is has been pointed out the live fault to case, which can go undetected , though in itself it's not a shock hazard. I would ask yourself just how many case faults you've experienced. I can't remember one.
In practice actually in such case the capacitive couplings using result in a little hum, when touched so in fact you do get a warning.

The moon rock scenario of alternative secondaries faulting to cases of different appliances which are then bridged by a person is protectable by an outlet RCD , but I really don't see the point

Alternatively an earth referenced IT secondary , with RCDs and breakers or RCBOs , in the secondary , really does nothing for on board safety. But only deals with impressed currents. An expensive solution which could easier be solved by breaking the AC DC interconnect and not bonding ( on grp hulls) and relying on RCDs to handle the safety issue. This is current European legal practice as per the ISO standards and mandated in the Recreational Craft Directive.

Both systems have fault modes that can cause issues. On balance if I went to the considerable expensive of installing a marine IT , I would want more safety then just preventing impressed current corrosion. In that regard I personally feel S/V Jedis solution is a good compromise.

Dave

[FlyingCloud1937]

Originally Posted by goboatingnow

No that's ABYC s interpretation. It's not other code bodies interpretation. Your scenario can be simply protected by a combination of RCD on the secondary and outlet RCDs , providing the safety of both systems

Dave

From your side of the pond

Conflicting Philosophies for electrical safety
There is some controversy as to why ISO 13297 allows the use of RCDs as the sole means of protection against electric shock without any other safeguards such as equipotential bonding (i.e. connecting AC protective earth to craft earth).
All other electrical safety standards expressly forbid the use of RCDs as a sole means of protection against electric shock, and insist on equipotential bonding, especially in high risk areas.
Here is the relevant paragraph from EN ISO 13297 (which applies only to small pleasure craft in the EU area)
ISO 13297 section 4.2
The protective conductor shall be connected to the craft's d.c. negative ground (earth) as close as practicable to the battery (d.c.) negative terminal.

NOTE: If an RCD (whole-craft residual current device) or an isolating transformer is installed in the main supply circuit of the a.c. system (see Section 8.2), the negative ground terminal of the d.c. Systems need not be connected to the a.c. shore ground (protective conductor)
This particular part of ISO 13297 puts electrical safety totally in the hands of the RCD, but RCDs are not 100% reliable. Using an RCD alone without protective earthing is not acceptable on dry land or in a houseboat.
For example BS7671 section 415.1.2 states that;
The use of RCDs is not recognised as a sole means of protection and does not obviate the need to apply one of the protective measures specified in Sections 411 to 414
The only measure that can realistically be applied here is Section 411 which requires that fault protection is provided by protective earthing, protective equipotential bonding and automatic disconnection in case of a fault... (BS 7671 411.1)
Note that BS 7671 is the UK version of the international standard EN 60364 electrical installation for buildings which covers the EU area. It has sections applicable to marinas and houseboats but is not applicable to small pleasure boats. However, the general philosophy of how to use mains electricity safely is consistent throughout the EU area;
a) use good quality earthing and Equipotential bonding.
b) an RCD provides additional protection, particularly in high risk areas.
c) RCDs are not allowed as a sole means of protection.

.................................................. ...........................................
Seems inconsistent ? Why ?
The reason behind this earthing inconsistancy where EN ISO 13295 is different from all other advice is to reduce galvanic corrosion by preventing galvanic current flow from boat to shore through the (nonexistent) earth connection. However in the case where an RCD is the sole protection it does this at the expense of safety.
There is a perfectly simple way to have a safe protective earth connection without increasing corrosion and that is to connect the AC protective earth to the boat earth (battery negative, engine block, any underwater metalwork etc.) and fit a galvanic isolator. or an isolating transformer which will be safe and will not increase galvanic corrosion.
.................................................. ...........................................
What are the consequences of a faulty RCD when there is no equipotential bonding ?
Let's say you are in the engine room, perhaps doing some maintenance work using an electric drill or a lamp powered from the AC mains.
Your boat is correctly wired in accordance with the electrical regulations for small pleasure craft (ISO 13297 etc). In order to save money on a galvanic isolator or isolating transformer your AC mains earth and the boat's earth (engine block etc) are not connected together... this is OK according to the applicable regulations, but...
A fault develops in the drill or lamp and a live wire touches the engine block.
This type of fault should trip the RCD on the shore power inlet, but the RCD has a fault of it's own and doesn't trip (see below for info about RCD reliability).
The engine block and anything connected to it is now live. That means the DC system is now live, the prop shaft and propeller and sacrificial anodes are now live, and so the water near the boat is also live.
The water around the boat will tend to conduct the live to earth but the connection is not reliable (especially on a plastic or wooden hulled boat) and is unlikely to be good enough to blow the main fuse or circuit breaker quickly enough (or at all) to protect anyone who happens to be in the water near the boat. Or a person on board who is touching the AC earth and the craft earth at the same time, maybe somebody is in the galley touching the sink (boat earth) and the fridge (AC earth)...
This type of fault might kill immediately or it may go unnoticed for a long time, minutes, hours, days. Fire or electric shock are real possibilities and the longer the fault goes unnoticed the greater the chances of a serious outcome...
.................................................. ...........................................
So what's the difference when equipotential bonding (earthing) is fitted ?
Go back to the previous scene in the engine room... if the AC earth was bonded to the boat earth (engine block etc) as soon as the live wire touches the engine block a fuse will blow (or the main circuit breaker in the shore power outlet). This disconnects the power immediately. ... No danger...
This has at least two valuable consequences, the first is that it immediately removes any dangerous voltage from the faulty item and therefore the engine block etc don't become live. The second is that it lets you know something is wrong... mmm drill doesn't work... why ?
Even if the RCD and all the fuses or circuit breakers were faulty and the main earth connection at the shore power outlet is faulty as well, there is a much reduced danger when equipotential bonding is fitted because all the exposed metalwork and the water around the boat are at the same potential.
.................................................. ...........................................
How reliable are RCDs ?
There have been at least three recent studies on RCD reliability. The following is quoted from a 2007 report on RCD reliability from the Electrical Safety Council in the UK. These studies were all done in domestic houses and all produced a failure rate of about 3 per hundred units tested.
The Electrical Safety Council's own survey during 2006 of approx 600 RCDs in domestic properties showed a failure rate of 2.8 per hundred tested.
(ref The Reliability of RCDs in Domestic Properties, ERA Report Number: 2007-0274, produced for the Electrical Safety Council 2007)
From research published in Italy in 1996, electromechanical RCDs in Italian residential properties had an average failure rate of 7.1%. When the RCDs were subject to regular testing, the figure fell to 2.8%
From the available evidence, the primary mode of failure of the electromechanical RCDs tested in Italy was ingress of fine particles of dust and moisture causing the moving components within the RCDs to stick or to operate more slowly than intended.
(Ref. Cantarella G., Caressin V., Tammasini R., Quality of Residual Current Operated
Circuit Breakers, ETEP. Vol. 6, No. 3, pp 149-156, 1996)
Electronic RCD manufacturers claim that their products are more reliable than electromechanical RCDs,. However, research carried out in the US in 2001 suggests that the reliability of electronic RCDs may be similar to that of electromechanical RCDs, The RCD failures in the US were attributed to the failure of electronic components. The failure rates were higher in cities with conditions of high humidity, which may not be relevant to conditions found in the UK.
(Ref. GFCI Field Test Survey Report, NEMA, Rosslyn, USA, Jan 2001)
Note that the above failure rates where in nice dry houses, increased failure rates are noted due to humidity and moisture ingress. There is no data for marinas available however it seems reasonably to assume failure rates will be the same as or higher than in domestic properties on dry land.
The UK Electrical Safety Council report on RCD reliability is available here... RCD reliability

[FlyingCloud1937]

Originally Posted by goboatingnow

I brought up this topic , the earth wire conundrum, because there are major differences of opinion , even amongst regulatory authorities , pseudo authorities , and people like Calder over the correct wiring of ITs in boats. Regularly they are installed incorrectly. ( like in front of a generator).

I recognise the advantages of an earth referenced secondary , but I also recognise the advantages of a floating output. Funnily one of the strange things given ABYC predilections is why they didn't specify a earthed 110v centre tap secondary , like the UK building site ITs. This at least reduces the voltage fault to earth to 55vac .

The only real issue with floating IT secondaries , is has been pointed out the live fault to case, which can go undetected , though in itself it's not a shock hazard. I would ask yourself just how many case faults you've experienced. I can't remember one.
In practice actually in such case the capacitive couplings using result in a little hum, when touched so in fact you do get a warning.

The moon rock scenario of alternative secondaries faulting to cases of different appliances which are then bridged by a person is protectable by an outlet RCD , but I really don't see the point

Alternatively an earth referenced IT secondary , with RCDs and breakers or RCBOs , in the secondary , really does nothing for on board safety. But only deals with impressed currents. An expensive solution which could easier be solved by breaking the AC DC interconnect and not bonding ( on grp hulls) and relying on RCDs to handle the safety issue. This is current European legal practice as per the ISO standards and mandated in the Recreational Craft Directive.

Both systems have fault modes that can cause issues. On balance if I went to the considerable expensive of installing a marine IT , I would want more safety then just preventing impressed current corrosion. In that regard I personally feel S/V Jedis solution is a good compromise.

Dave

Dave, We crossed posted.

But we hit the same nail.


Nothing is perfect, but on a boa, anything that can go wrong usually does.

All of the systems, are at a potential for failure, especially the ground fault devises. That's why testings regular is required for their correct operation.

BUT THAT JUST DOESN'T"T HAPPEN.

So Each boat can be designed to it's purpose. "A DockSide Queen", always plugged in, and 2. an almost always away form shore side connection.

but things change, so should the remedies, but they don't.

The boat I spoke of below, is a prime example. I will have my megger on the boat and find the cause, and my solution will only be temporary.

Because things will change.

Lloyd

[pbiJim]

Quote:
Originally Posted by chala
Here are two GFCI as you may notice each one have 3 wires. How do you connect them to yours two wires floating wiring?

Originally Posted by pbiJim

I would be interested to hear the response to this question from this gentleman's most vocal critics. I think that this is a valid & productive question. The correct answer to this question should be useful to anyone with an IT & a GFCI on their boat.

Originally Posted by goboatingnow

An RCD has TWO wires TWO , everyone understand. That's all that ever connected to it.

Dave

Chala is talking about GFCI. Dave is talking about RCD. The last GFCI that I wired had 3 connections, not two. Is everyone talking about the same thing here? Is a GFCI different from a RCD? Is the US version different from the EU version?

[FlyingCloud1937]

Originally Posted by pbiJim

Quote:
Originally Posted by chala
Here are two GFCI as you may notice each one have 3 wires. How do you connect them to yours two wires floating wiring?

Originally Posted by pbiJim

I would be interested to hear the response to this question from this gentleman's most vocal critics. I think that this is a valid & productive question. The correct answer to this question should be useful to anyone with an IT & a GFCI on their boat.




Chala is talking about GFCI. Dave is talking about RCD. The last GFCI that I wired had 3 connections, not two. Is everyone talking about the same thing here? Is a GFCI different from a RCD? Is the US version different from the EU version?

Lacking the ground/earth safety/reference; there is not fault, so no trip. But intercede as humans do on a boat and they can create the path. If it's a 30 mil/amp and a week heart, there may be a problem in rare instances.

Now if the devise providing the fault trip is not working correctly then all bets are off.

lloyd

[s/v Jedi]

A GFCI outlet is a combination of the protection device and an outlet. All the US versions I have seen (and use in my boat) have a double grounded outlet. The GFCI outlet has a ground connection for the outlet part, but the GFCI part is not connected to that at all.

Also, check out this old thread that describes how my AC system was done incl. diagrams etc. It adds my most favorite part, the auto-transformer (which is only usefull for US boats with 115V outlets)

Except for the auto-transformer, all the rest is usefull for every boat. The inverter/charger is actually an European version which I reprogrammed to 60Hz.

http://www.cruisersforum.com/forums/...tem-29667.html