Bonding in greenhouse

Building or not?

If red houses are made of red bricks, and blue houses are made of blue bricks, what are green houses made of? Try it next time that you see a 5 year old!

So if a greenhouse is a type of house, and a house is a building, is a greenhouse not a building?

Whilst "house" is usually associated with a dwelling, it might refer to a church (house of God), part of a zoo (e.g. reptile house), part of a farm (e.g. cow house), parliament, etc.

So I think that a greenhouse is a building.

Graham. It is extremely rare for me to do a TT install (almost never) , however I have decided to put in two earth rods and verify each on its own then connect together (around my neck of the woods I would usually expect around 40 to 65 ohms on a single 4` rod).  The two rods to be about 8` apart therefore outside each others sphere of influence.

My reasons for two rods :-

1/ Redundancy can`t hurt.

2/ Lowers total Ra - not to half the value of one rod but quite a good difference though.

Is there some merit to my reasoning or am I barking mad?

I would welcome your comments on my reasoning (as an approx first thought then if say each rod were to be 100 ohms to the rest of the planet as measured at the antipodes then 80 ohms might be just local and the other 20 ohms to the rest of the earth. Therefore two 100 rods have 80 ohms/2 = 40 ohms total and the other 20 ohms takes it to around 60 ohms with two rods. Very rufly speaking of course                           

Zoomup The greenhouse in question is only something like 1.8m by 2.5m, this is a good general discussion as I probably look at 4 or 5 small domestic greenhouses with power per year.

Most have plastic, normally weather proof sockets, don't think I have ever found one with the frame bonded.

The ones I work on are normally in rural locations with supply coming from a DB in a garage or shed, from experience I don't trust RCD's to be reliable once they get past about 20 years old in these damp environments and it's unusual for the occupiers to test the RCD's. 

In these installations at least having a source of ground potential rise anywhere near the earth rod or greenhouse is fairly unlikely.

I think having operation RCD's and a good installation with a plastic socket is the key to safety. If the greenhouse is in contact with earth there are different risks for bonding or not but overall the risk is probably similar. I think if I was installing I would bond as it's following the basic interpretation of the regulations. For existing installs I would note on the certificate that it's not bonded, not sure i would code it as a C3.

Out of interest I am also going to call NAPIT technical tomorrow to see what answer they give me.

Hi :ebee 

A general consideration (noted in BS 7430) is that any portion of an earth electrode which is not 600 mm deep is at risk of drying out or freezing.

Assuming top of rods are below ground level in a suitable pit, this may turn 4' rods into 3' or 2' rods in dry summers, or freezing conditions in winter.

Given that, your redundancy seems like a reasonably good idea. It may not be necessary with other types of electrode (e.g. earth mat or tape buried at least 600 mm deep) or other configurations.

The logical method is right, the numbers will not be as bad as that unless both rods are in some isolated geological structure that in effect creates a bowl of conductive soil inside a insulating rocky depression, perhaps a caldera. Given your usual stomping ground, I think that is unlikely.

It is easy to underestimate how much of the voltage drop/ effective electrode resistance is due to material very near indeed to the electrode structure.

To that end, it is worth trying to visualize how much the current spreads so that the cross-section carrying say, 90% of the current flow  is hundreds of square meters even at a distance of perhaps 10m from the electrode. (consider the area of a sphere - 4pi R^2  but then ignore the air, so half a sphere, and maybe assume a sense of direction so cut away half again, you are still looking at 300m2 at 10m range. The addition of the extra resistance to get to say 20m range is the series resistance of the shells of radius 11m, 12m, 13m etc.) Those who can recall how integrate from their applied maths lessons can go straight there, but those who cannot, bear with me.

Each of those thin shells of 1m thick can be imagined ironed out flat and the resistance of each one calculated in terms of the resistance of 1 cubic metre of earth or whatever (being the resistance of 1 cube divided by the area as that is how many metre cubes are in parallel.). The electrode resistance to the infinite plate at the end of the universe,  as you get further out from the electrode under test is the sum of these rapidly decreasing  resistances.

So let us hang some numbers on this and take 100 ohm-metres (so a cube of that soil, with metal plates between 2 opposing sides and the other sides insulated would read 100 ohms between the plates to a voltage large enough not to be confused by electrolytic or battery effects.)
Let us assume the current is all going to one octant so a 90 slice of the circle on the surface only.

the 10m shell has an area of 1/8 of 4pi *157m2  is 0.63 ohms, the 11m shell an area of ~190m2 adds 0.52 ohms ... out to 20 metres 630m^2 and  0.159 ohms. ..

So it is quite sensible to consider that nearly all the rod resistance of 80 ohms or whatever is due to resistance in the first few rod length radius - as much beyond that the current is expanding spherically.

Of course the conductivity of real soil is not uniform over depth or distance, the last ice age saw to that, but the visualisation is not incorrect, just the current tends to bunch a bit in the regions that conduct better.

It also means that standing bare foot near an earth rod while a fault is running current to earth is only really a problem while you are actually more or less on top of it again by a rod length away the voltage gradient is well down.

Mike

ebee said:

or am I barking mad?

That really is a, "Do you beat your wife?" question!

In fact, there is nothing wrong with your double electrode, but it may be unnecessary.

Zoomup said:

So we will bond it to the earth then Chris, blindly following an illogical mis-interpreted regulation.

Eh? All I said was that, IMHO, a greenhouse is a building.

I certainly wouldn't be able to justify asking a client to pay for a second rod if the first ne gave a results as low as 80 ohms. I typically get between 100 and 150 ohms in my area from a single rod, if it gets to 200 or above I put a second rod in.

I just posted at the base of the thread of this dis-jointed forum  Chris.....nothing personal.

Z.

Loop test the existing earth rod, install a new 30mA R.C.D., replace the metal clad 13 Amp socket in the greenhouse with a plastic one,  then sleep well at night. Keep it simple.

And please do not ignore Table 41.5. Ref. the loop impedance  of the earthing arrangement (Zs). The earth rod will not have to sink much current, just enough to trip off the 30mA R.C.D. Just what is the significance of the 1667 Ohms earth fault loop impedance relating to 30mA R.C.D.s?

Before R.C.D.s a large fault current was needed to melt a hot wire or cartridge fuse link, but now with R.C.D.s a much smaller current is required for an earth fault.

Z.

"how are those parts not already so? 

Not just faults in the installation, but ground potential rise from conditions outside the installation."

................................................................

That will not be the case here. The small greenhouse will be at the end of a rural garden.

Ref. ground potential rise, is yet another reason not to bond the metal frame. Keep it separated. You don't want the equivalent of exported P.M.E.

I live in TT land and come across TT installs on a regular basis, I have not found any problems with the conventional 4 ft. long earth rods being used for most installations, e.g. farms, marinas, caravan sites etc.

Z.

"how are those parts not already so? 

Not just faults in the installation, but ground potential rise from conditions outside the installation."

................................................................

That will not be the case here. The small greenhouse will be at the end of a rural garden.

Ref. ground potential rise, is yet another reason not to bond the metal frame. Keep it separated. You don't want the equivalent of exported P.M.E.

I live in TT land and come across TT installs on a regular basis, I have not found any problems with the conventional 4 ft. long earth rods being used for most installations, e.g. farms, marinas, caravan sites etc.

Z.