Earth rod for electric fence

[aesmith]

Hi,

We run an electric fence configured as follows. A PEL energiser is powered from a 12V battery (with solar panel), the output is in the form of two leads, the "Live" output connects to the fence, and the "Earth" output connects to an earth stake, currently a galvanised rod driven about 15-18" into the ground.

We test the fence with a basic tester which shows up to 5 neons indicating 0-5000V in increments of 1000V. The tester has it's own earth spike that you stick in the ground at the foot of the fence, then put the business end on the fence and watch to see how many neons flick for each pulse.

The problem we have is poor readings at the fence, and after chasing round checking for shorts, or bad connections, it now appears it's the earth that's the problem. Testing between the ground and the fence shows 3000V, testing between earth stake and fence shows 5000V, and testing between ground and earth stake shows 2000V. Shown as tests (1), (2) and (3) in the diagram.

It seems to me that the issue is the earth, or the earth stake. Could anyone suggest any easy way that I could improve the earthing? What's odd is that when I'm doing tests (1) and (3) I get these sorts of readings even with the tester's earth probe placed within a couple of inches of the earth stake. Could the issue be that a galvanised rod simply makes poor contact, or do we need to get it deeper in the ground? Alternatively if it proves impossible to get the stake further down, would running a bare copper wire horizontally be any good?


Any suggestions welcome, thanks in advance, Tony S

 

[telectrix]

if anything, bang in a longer rod.

 

[Darkwood]

Sounds ironic that you require an earth rod for something designed to give you a shock if you touch it in the first place ...but its a functional earth so let it off!

 

[aesmith]

Thanks. Yes I guess it's a bit of a reversal of the normal earthing aims. Anyway, for this earth rod, does it make a lot of difference what sort of metal is used? I understand you'd use copper for a protective earth, but I was thinking that for this application we probably don't need to get down to fractions of an ohm.

Does it seem odd that there's apparently a highish impedance between the rod and the tester even though they're only a few inches apart?

Tony S

 

[Darkwood]

Not been familiar with these systems and would need to see a few results from differing set-ups but forethought to my mind would be ground surface resistance could be higher at the rod than is at fence ...

 

[Marvo]

Most earth spikes you buy at the wholesalers are actually copper coated steel. You can use galvanised steel rods or tubes for earth spikes as well, they do rust eventually but the rust doesn't affect the Ra values significantly until it compromises the structural integrity of the rod. The biggest problem with galv steel tubes or rods is that you can't usually screw them together to extend the length.

Your 15-18 inches is probably just well short of the mark required. I'd start with a 2.3 meter rod and if necessary screw another one on to make the total rod length 4.6m. You get a rod driving attachment for an SDS Max jack hammer, I suggest you should invest in one of those as well, I'm sure one of the UK guys can give you a link to a decent one.

Does the fence have alternating wires that are earth or are all the fence wires live? If there's earth wires in between you should also put an earth spike every 50 meters along the fence and link it to those wires.

 

[aesmith]

Thanks all. I'm a bit taken aback by the idea of driving in a rod down to 4.6m, I'm pretty sure I'll hit rock before then but I'll see how deep I can drive a stake. FYI the SSE guy who traced their service cable for me said that their earth is formed by laying a copper cable in a trench and back filling, that's why I thought of running a wire horizontally.

The fence wires are all live, so the deterrent relies on the intruder getting a shock between the live fence and what I'll call "real earth", ie the stuff he's standing on. I did wonder about running an earthed wire at ground level, but that's not normal practice so I assume not normally required.

Thanks again, Tony S

 

[Marvo]

You can create an earth by digging a large trench and burying a copper earth mat. The deeper and larger it is the better, I'd recommend at least a meter deep and you need a few bags of bentonite or similar around the copper plate or lattice to get a decent earth impedance.

 

[Engineer54]

That's a decent sized earth mat you have there Marvo. More often than not, just one of those latice mats would typically be used.

Is that earth mat being laid under a paved patio area??

 

[Marvo]

That particular earth mat was one I just pulled from Google images as an illustration. It's pretty similar to the general earth mats we install but I didn't have pictures to hand. If it's a new building we try to lay the earth mats in the foundation trenches before the concrete is thrown that way we don't have to do the digging . Depending on the building size with between 4 and 6 mats evenly spaced around the foundation trenches and linked with a 10mm bare copper wire or braid we often get sub-ohm earth impedances.

 

[aesmith]

Not been familiar with these systems and would need to see a few results from differing set-ups but forethought to my mind would be ground surface resistance could be higher at the rod than is at fence ...

Quick update, we checked all the continuity around the fence, and I was able to drive a rod down a good metre into the ground. That's improved matters, we're now seeing only 1000V between the rod and ground elsewhere, and 4000V between fence and ground with that reading consistent all around the perimeter.

By the way, in case you thought this was some sort of prison camp, the fence in question is an anti-fox measure for our poultry.

 

[Marvo]

Well that kinda figures, you've halved the earth impedance and you've got half the voltage to earth now which is a good result of sorts. I'd persist with at least another couple of one meter earth rods preferably several meters away from the existing rods if I were you, bearing in mind at a meter depth the rods you've installed may be negatively affected by ground freezing in winter.

 

[aesmith]

Cheers, I'll see what I have to hand as sticking in one or more additional rods is going to be easier than going deeper. I'll start to run out of resolution on the tester, it's just a series of neons which blink with each discharge, one light = 1000V up to all five = 5000V or over.

 

[Marvo]

Ahh, I know the type of tester you're using, problem is if the 1000v LED lights then it means your voltage is above that value and possibly just below 2000v. You really need an earth impedance tester but I doubt you're going to want to buy one just for this job as they're not particularly cheap. I'm sure your chickens will thank you if you keep knocking in earth rods at least until none of the LED's are lighting on your tester because without a decent earth your fence won't be much of a deterrent.

 

[aesmith]

... I'm sure your chickens will thank you if you keep knocking in earth rods at least until none of the LED's are lighting on your tester because without a decent earth your fence won't be much of a deterrent.

I'm only just realising that. I wasted quite a lot of time checking for shorts when the fence was reading low, now kicking myself for not checking the earth sooner.