Greenhouse Electrics

[JohnW2]

OK, following my earlier posting, I’ve been thinking about the greenhouse electrics business, and also looking around the forums and seeing an awful lot of differences of opinion!

As far as I can see, the only really ‘safe’ option is to isolate the greenhouse’s electrical installation from the house’s CPC/earth, have a local TT electrode as the earth for the greenhouse’s electrical installation and bond the metal frame etc. of the greenhouse to that. I presume that if one does that, it’s still OK to rely upon upstream RCD protection (i.e. in the house), since the L-N imbalance that would cause the RCD to operate does not care what path to earth the ‘lost current’ has taken.

If, as I suspect is the approach most likely to be taken by a DIYer, one simply exports the house’s earth to the greenhouse (and bonds the greenhouse metal to that earth), then I see theoretical problems. Whether by virtue of the CPC or even if one ran an adequate bonding cable from house’s MET to the greenhouse frame etc.(which I suspect very few people would do), there would, under some fault conditions, be a dangerous potential difference between the greenhouse metal and the soil below. With an L-E fault (whether TT or TN at the house), a significant pd should only exist for a fraction of a second before the fault was cleared by a protective device, so ‘not too unsafe’, but with the much-discussed ‘lost neutral’ fault with a TN-C-S/PME supply, the pd between greenhouse metal and soil below would presumably be persistent, and hence very dangerous.

So, it seems to me that if someone chose not to have a local, isolated, TT system in the greenhouse, it could be argued to be ‘not unreasonable’ to bond the greenhouse metal to the exported earth if it were TT, but not acceptable if it were TN-C-S. In fact (and many of those here may not wish to comment on this!), if one did decide not to have a local TT system for the greenhouse, the ‘least of the evils’ may actually be to leave the greenhouse metalwork floating.

As for ‘outside sockets’, I really can’t think of anything one can do apart from ensure there is adequate RCD protection and encourage/hope that only Class II equipment will be plugged into the sockets – since I don’t see what else can realistically be done; one obviously can’t ‘bond’ a whole garden!

Is what I say reasonable? Any comments?

Kind Regards, John

 

[Taylortwocities]

Earlier posting, what earlier posting.??????????

Its best to add to your existing topic or at least include a link to it. Then folk have half a chance of understanding the context.

 

[JohnW2]

Earlier posting, what earlier posting.??????????
Its best to add to your existing topic or at least include a link to it. Then folk have half a chance of understanding the context.

Apologies. Bernard unintentionally sort-of hijacked my thread right at the start with talk of his cottage, whereafter everyone seemed to overlook (or deliberately ignore ) my initial post, which can be found here

All comments will be welcome!

Kind Regards, John.

 

[ericmark]

There has been a lot of talk about a TT supply from a TN supply and although in some cases I can see it is required it is not an easy call to make.

What we look at is the likely voltage gradient. With an outbuilding close to the house one would normally use same earth type as house. Where the outbuilding is surrounded by houses again normally same supply as house.

Where the outbuilding is a distance away from the house and there are open fields then one has to think carefully about supply type. It is a risk assessment. The questions asked are if non double insulated tools are likely to be used from the supply and could the user of the tools end up using them near to items earthed to the house earth.

With caravans one is not really considering the user having an old lawn mower plugged into caravan supply, and fire regulations means it can't be parked too near to buildings using a TN supply. With boats the electrolysis is a consideration so only a TT or a TN-S supply is used. And even with the TN-S supply the use of an isolation transformer often means this does not come from the building. Often no earth is connected to boat and even if it is likely through diodes because of electrolysis.

The other method to reduce the chance of a voltage gradient causing a problem is to install an extraneous-conductive-part which is protected from direct contact as would any earth rod. Although it is an earth rod with a TN supply it's not called an earth electrode. Although it is really.

As I said before one makes a risk assessment and decides which is the most appropriate for that outbuilding.

The problem is of course it is not just you who has to decide but also who ever issues the completion certificate. So my advice is to ask that person what they want. And unless there is something special as with the case of boats where the supply is IT to a transformer and the TN-S after the isolation transformer then I would accept what they say.

I have worked as an electrician for 40 odd years for 100's of firms and only three has the equipment to test the earth electrodes. With that one can test the prospective gradient but I never did. It was just a case of bang in rod and test it. But I needed 30 feet to test rod to put in test probes. In most domestic buildings there is not the area around the rod to test with the stand alone equipment. So the only way to test is the use the suppliers earth as a reference.

This is old story of chicken or egg. They should not provide a supply until earth is tested but you can't test earth until you have a supply. Which is why with new supply the supply normally wants your electrician on site.

Now this is a DIY site so if we assume you are going to DIY then if using a TT system how will you test the earth electrode? Even when using the suppliers earth as reference you still need a meter costing around £200. The stand alone meter more like a £1000 to buy. So if you can't test the earth in theory you should not do the job. Yes and that is really going to stop you? So if you can't test the earth then only safe way is using the TN earth.

So for most DIY people question is really should I stick an earth rod in as well? My answer is it will not do any harm. Especially if under the floor of the outbuilding where no one can touch it. They are not expensive so just stick one in as well as the TN earth from the house.

That's not saying in all cases that's the best option. But unless you have all the test equipment you should have, then likely the option with least risk.

 

[bernardgreen]

As I said before one makes a risk assessment and decides which is the most appropriate for that outbuilding.

That is the best advice. But few are trained in risk assessment.

Unfortunately all too often the process of risk assessment and subsequent decision seems to take second place to the easier process of force fitting an interpretation of the regulations to get a best fit solution. It might be a good compromise but it may also not be a safe method.

I mentioned the shop ( TT ) and cottage ( PME or TT ? ) with linked plumbing as it is another paradox that requires a good risk assessment.

The risks are those involved with earthing a PME system to a ground rod that is inside the premises.

In this case the rod is next door (in the shop) but is linked by solid high fault current capability pipe work. The bond of the PME MET to the water pipes inside the premises introduces the shops rod onto the PME earth/neutral which ( in theory ) is just one more ground in the multiple earthing of the system. I can see that if the pipe work was not capable of carrying high earth currents from PME to rod then a fire risk might exist if a network fault drives the PME neutral and "earth" derived from that neutral bove ground leading to high currents in the pipe work to reach the ground rod.

Connecting the PME earth to as many earth rods as possible is acceptable, that is the basis of the PME system.

A risk assessment on connnecting to a ground rod ( or to a metallic supply pipe acting as a ground rod ) indicates that possible high fault currents in the cable to the rod could create a possible fire risk.

Unfortunately the risk is higher than imagined as safety disconnect devices will not be triggered as the fault current is from network neutral via the CPC ( taken from neutral before the RCD ) to the ground rod. It will continue until the neutral is restored or the wire melts in the worse case.

The next complication to the risk analysis is when the metallic water pipe is acting as a ground rod ( OK it is NOT described as a ground rod, it cannot be used as a ground rod but it is a ground rod in practise ) is metallic to the next house where is is bonded to that house's MET. Not a serious problem with 10 mm bonding if the houses have the same network neutral for their source of PME earth. But if only one house loses its neutral then the 10mm bond and metallic water pipes take over the function of the missing neutral.

Risk assessment is complicated isn't.

Back to original. Metal green house. Feed with Live and Neutral via RCD and use TT with effective ground rod and bond frame to the ground rod.

 

[bernardgreen]

Back to original. Metal green house. Feed with Live and Neutral via RCD and use TT with effective ground rod and bond frame to the ground rod.

Then I remembered.......what if there is a water supply pipe to the green house ? A pipe bonded to the neutral via the PME MET in the house

Not a problem if it non conductive plastic pipe. But if metal and in electrical contact with the metal frame ?

 

[Taylortwocities]

In that case you are going to have to bond the frame, including a strap to bridge all of the joints in the frame

 

[holmslaw]

..

 

[JohnW2]

There has been a lot of talk about a TT supply from a TN supply and although in some cases I can see it is required it is not an easy call to make. .... As I said before one makes a risk assessment and decides which is the most appropriate for that outbuilding.

Indeed so. It is clearly anything but a 'black and white' situation. Perhaps the most worrying thing is that if one looks around at the many discussions about this topic, what one tends to see is not a discussion about risk assessment but, rather, different electricians asserting a variety of different opinions as to 'the way it must be done (always)'!

Now this is a DIY site so if we assume you are going to DIY then if using a TT system how will you test the earth electrode? Even when using the suppliers earth as reference you still need a meter costing around £200. The stand alone meter more like a £1000 to buy. So if you can't test the earth in theory you should not do the job.

I can't disagree with that. I personally wouldn't install an earth electrode without testing it - well, testing the loop impedance it provides (I have in my time tested earth electrodes using the 'transformer method', but I don't really see any great point if the Zs is acceptable), but I agree that most DIY folk wouldn't.

So if you can't test the earth then only safe way is using the TN earth.

That brings me back to the main question I was asking. If one simply uses the earth from the house's supply, does one then bond the greenhouse's metalwork to it? As I said before, if the supply is TN-C-S, that would seem to me to present a significant hazard. I have to say that this arrangement (with bonding) is what I have in my greenhouse (not the friend's one I was talking about) - but my house has a TT supply.

So for most DIY people question is really should I stick an earth rod in as well? My answer is it will not do any harm. Especially if under the floor of the outbuilding where no one can touch it. They are not expensive so just stick one in as well as the TN earth from the house.

You seem here to be talking about an option which I was not even considering - use the house's earth (which may be TN, but could be TT) and also have a local electrode which one connects together to the house earth. Is that what you're suggesting? ... and what then about bonding the greenhouse metalwork? Your mention of having the rod somewhere "where no-one can touch it" tends to suggest that you are thinking of not bonding the metalwork.

Although I agree that such an approach would, at the worst, 'not do any harm' as far a the greenhouse environment was concerned, would there not then a problem with the installation as a whole in terms of the regs? Whilst connecting a metal rod in the ground to a TN earth is essentially just 'PME', the regs would presumably regard it as an 'extraneous-conductive-part' and therefore require 'Main Protective Bonding', the required sizing of which would probably not be satisfied by the cabling likely to be used for connection to a greenhouse

That's not saying in all cases that's the best option. But unless you have all the test equipment you should have, then likely the option with least risk.

I am very happy to have my mind changed but, as I've said, in the case of a TN-C-S supply I would have definite concerns about the greenhouse metalwork being bonded to its earth. In terms of safety, my present inclination would be to put the options in the following order ('safest' first), in all cases assuming that there is adequate RCD protection in place:

• 1...Isolate house earth from greenhouse. Greenhouse has local (tested) TT electrode, with bonding to greenhouse frame.
  
  2...If the house's supply is TT or TN-S, use just that supply's earth and bond the greenhouse metal to it. An L-E or L-N fault anywhere in the installation should be cleared by a protective device, hence minimising the time during which there could be a dangerous pd between greenhouse and soil. As (I think) suggested by you, one could potentially make it even safer by adding a local earth rod to that, but I'm then not sure about the reg's MPB requirements.
  
  3...As (2), but with the greenhouse metal not bonded (i.e.floating).

Of those, as I've, said, I would be uncomfortable with (2) if the supply were PME.

Any thoughts/comments on that?

Kind Regards, John.

 

[JohnW2]

Unfortunately all too often the process of risk assessment and subsequent decision seems to take second place to the easier process of force fitting an interpretation of the regulations to get a best fit solution. It might be a good compromise but it may also not be a safe method.

Indeed so. As I've just written, when one looks around at the many discussions on this topic, one tends to see varying asserions as to 'what must be done' based on (varying) interpretation of 'the regulations', often without much apparent application of thought or common sense, or understanding of the underlying principles.

I mentioned the shop ( TT ) and cottage ( PME or TT ? ) with linked plumbing as it is another paradox that requires a good risk assessment.

Having thought a bit more about this, I think the nature of the risk is actually very simple, as per what I wrote before. As I see it, the only problem with the cottage having PME and the shop TT (with pipework joining the two) is that in the event of a lost neutral in the PME supply to the cottage, the problem experienced in the cottage (high potential metalwork) would also be experienced in the shop. I was also wondering - did you not say that the shop's water supply is sub-metered? If so, is the meter plastic?

Kind Regards, John.

 

[ericmark]

I have attended one house where there was a problem due to the TN-C-S earth being connected to an earth mat on the premises.

The earth mat was very good consisting of four earth rods with earth tape 1/8 x 1.5 inch copper strip and was primary for the radio transmitter, the earth wire was under size at 2.5mm from the house to the shack.

Workmen hit the supply cable and cut the neutral only and the 2.5mm cable did melt.

However back to risk assessment the likely hood of:-
1) Such a good earth in shed
2) Under size cable being used for TN-C-S earth to shed
3) All services being in plastic
4) Workmen cutting just the neutral
Is very slim.
With an earth rod with a resistance of likely 60 ohms for enough current to flow to melt even an under size 2.5mm cable is just never going to happen.

For 20A to flow the rod needs to be down to 20 ohms and that is for a 400 volt supply. It would be rare to get such a low reading. If we consider 6mm as min size then your looking at 10 ohms before one is really getting into a problem area.

Using a all pole RCD will of course still auto disconnect and fail safe. You will note for caravans all pole is required where we often have to use a TT supply from a TN-C-S supply it therefore does disconnect the neutral if there is a fault.

But it is not an easy call. And it does need consideration. Metal work is always bonded so that goes without saying but as to services if all in the street have metal pipes then if a neutral is broken the current is shared between many earths. It is only where as with the radio ham his was only earth for whole group of houses is there any real danger.

 

[bernardgreen]

that in the event of a lost neutral in the PME supply to the cottage, the problem experienced in the cottage (high potential metalwork) would also be experienced in the shop.

Which is the reason the DNO new supplies department raised the problem of linking a PME CPC to a TT CPC

I was also wondering - did you not say that the shop's water supply is sub-metered? If so, is the meter plastic?

Sub metering is a possibility. That has to be sorted out as well. But even if the meter was plastic I would still be concerned about sneak circuits between the two CPCs Sneak circuit are likely to be thin low current capacity routes ( compared to a 15 mm copper pipe ) which would present a higher fire risk in the situation of a floating neutral sending current into the shop's ground rod.

 

[JohnW2]

I With an earth rod with a resistance of likely 60 ohms for enough current to flow to melt even an under size 2.5mm cable is just never going to happen.
For 20A to flow the rod needs to be down to 20 ohms and that is for a 400 volt supply. It would be rare to get such a low reading. If we consider 6mm as min size then your looking at 10 ohms before one is really getting into a problem area.

I agree with all that, but it's not the melting of conductors that I was considering as the greatest worry. My concern is the p.d. between greenhouse metalwork and the soil below it if it is bonded to the earth of a TN-C-S supply which loses its neutral - and that risk would not be much reduced by an additional earth rod at the greenhouse.

Using a all pole RCD will of course still auto disconnect and fail safe.

I don't really see why an RCD would operate in response to a supply-side loss of neutral on a TN-C-S supply, nor if a human body provided an addition return path to earth ('true' earth - the greenhouse soil) in the presence of such a fault. Given the installation's neutral's effective connection to the installations's MET (main earthing conductor), in those situations, an RCD would surely see equal L and N currents (even though N currents were returning via earth, rather than a DNO N conductor), and therefore not operate

Metal work is always bonded so that goes without saying ......

You say that, and it would certainly be 'standard wisdom/teaching' - but, as I've said, when it comes to risk asessment, it looks to me that (perhaps ironically) it is the bonding of greenhouse metalwork which, under certain fault conditions, may bring about one of the greatest of the potential risks.

Kind Regards, John.

 

[JohnW2]

that in the event of a lost neutral in the PME supply to the cottage, the problem experienced in the cottage (high potential metalwork) would also be experienced in the shop.

Which is the reason the DNO new supplies department raised the problem of linking a PME CPC to a TT CPC

Indeed - but my point was that this is probably, realistically, the 'only' significant risk involved.

Sub metering is a possibility. That has to be sorted out as well. But even if the meter was plastic I would still be concerned about sneak circuits between the two CPCs Sneak circuit are likely to be thin low current capacity routes ( compared to a 15 mm copper pipe ) which would present a higher fire risk in the situation of a floating neutral sending current into the shop's ground rod.

I'm not sure that all this talk of 'fire risks' is all that justified. As far as I can see, with an average earth rod, it's unlikley that we're talking about maximum currents of more than 5-6A, and even that only if some very 'high powered' loads were connected to the TN-C-S installation (and switched on) at the time - or am I missing something?

Kind Regards, John.

 

[bernardgreen]

it's unlikley that we're talking about maximum currents of more than 5-6A,

If the neutral derived CPC of a PME is connected to a ground rod then in the event of a network fault leaving the neutral open circuit the network will try to return the net neutral current to the supply transformer star point via any and all available paths to ground and the star point. The net current being the result of un-balanced loading on the phases, this could be tens of amps, maybe hundreds if one or more phases break at the same point as the neutral has broken.

The impedance ( rod to ground ) of most ground rods won't allow that much current to flow so the whole CPC goes high voltage. Something such as the signal lead between a PC earthed to the CPC and other equipment earthed directly to ground would take a proportion of the net current, that cable might burn out.

Yes I do accept these are not common occurances.