Lightning protection in Africa

[echase]

I have a problem with lighting protection in a charity in Zambia. The earthing is TT, i.e. earth pits or rods are provided for almost every building. I have measured the resistance of these and some are half to one ohm. But others are 100 ohm which is is dreadful and we are redigging these.

The wet season has loads of lightning and when this strikes the overhead 415V 3-phase incoming grid lines huge surges enter the property. All the 10kA max Surge Protection Devices (SPDs) fitted years ago to the live incomers in distribution boards are blown. The local electrician has replaced them with 20kA max ones and now included the neutrals too.

We are adding 4 lightning spikes on 13m tall poles to protect a few single storey buildings and will dig separate new earth pits for these. What is the advice on whether to connect these pits to the main ones? Some say keep them separate and some say connect up all earths to give equipotential bonding everywhere. Problem I see is that it’s all very well bonding it all together in a nice modern UK building with modern electrics. But the idea of putting thousands of amps into a one ohm pit means the pit briefly gets to thousands of volts (V=IR) wrt any other E, N or L wire. So if you bond that pit to the house earth and you are standing on the floor and touching the fridge, or worse still the fridge and a tap which is not bonded, there will be 1000s of volts on the fridge case, yet your feet and tap hand are nearly at zero. The SPDs will limit it somewhat by sending some surge back along the grid to the substation, but it’s still a lot of volts, albeit very brief.

If it were all truly equipotentiality bonded the whole house, floor, taps, etc. reach the same 1000s of volts so it does not matter. But in Zambia where the wiring is struggling to meet IEE regs of decades ago it’s going to be hard to get it all bonded without rebuilding the place and spending more money than they can afford. So are separate pits better for less well bonded buildings?

Also what wire size should be used for connecting the spikes to pits? The local electricians say 16mm2 whereas I have measured buildings in UK and elsewhere with copper straps of 30 to 50mm2 for their spikes. If the pit is half an ohm whether the connecting wire is 0.05 or 0.01 ohm seems rather irrelevant to me, so 16mm2 is fine. But will it simply vaporise when carrying 1000s of amps from a strike? On the other hand will it actually transmit 1000s of amps if the pit is half an ohm?

 

[echase]

Just measured an old church in UK which has 2 solid 50mm2 round bars, presumably of copper, down from its lightning conductor. They go into the ground in 2 places 4 m apart either side of the Norman tower. No idea what is underground there but it passed its last test for earth resistance 5 years ago.

As far as I can see there is no connection between those 2 earths and the mains earth at the meter. It's a conventional modern main wiring system that passed its wiring inspection recently.

 

[Robson689]

What exactly do you want to know? We don't install lightning protection like this anymore we install to BSEN 62305

I will happily help you with any questions

 

[echase]

What exactly do you want to know? We don't install lightning protection like this anymore we install to BSEN 62305

I will happily help you with any questions

Thank you. I put the things I especially wanted to know in bold above to make it clearer.

Don't know much about BSEN 62305 and my electrician son does not either because lightning is not a serious problem in UK (and we don't use TT earthing much), but in Zambia it would be classed at very high risk. A man died on the site 10 years ago when stuck under a tree in the grounds and the yearly toll on electronic equipment damaged by surges and spikes is £1000s.

BSEN 62305 seems to represent an ideal that is hard to find in Africa due to old buildings and shortage of money. Another problem is that many buildings have corrugated iron or aluminium roofs and it’s not practical to bond the sheets together (if that is what BSEN 62305 recommends). So in practice it’s not possible to make the building fully comply with BSEN 62305. We just have to spend the limited money on the things that have the highest benefit, in accordance to what a local electrical contractor advises. Hence the choice to upgrade the SPDs and fit 4 lightning poles. We also will run 16mm2 wire along the roof ridges linked to the lightning down wires.

That is unless someone here says there are better ways to spend the money as I don’t fully trust the local advice, e.g. it partially conflicts with advice I’ve had from an Ugandan electrician who is familiar with TT earthing.

 

[Robson689]

I agree that lightning is not so much an issue in the UK but there is hundreds of people like myself who install, maintain and inspect lightning protection systems and that's all we do!

You could always use flat aluminium tape on the ridges which is about £35 for 50m I don't know if that price is cheaper/more expensive than your initial idea just make sure not to put it in the ground.

What is the advice on whether to connect these pits to the main ones?
I would probably try to keep it separate but I have no reasoning, it's just because that's what we do here which isn't much help...
We do however take one 70mm2 bond to the main electrical earth bar but make of that what you will


I'm struggling to picture what you'll be doing.

Will the 13m poles be next to the buildings? If so, I know it used to be generally accepted that for an air terminal to be effective it should be as tall as the building is long so is this why you mentioned linking the two poles because you're wanting one at each end?

It might be worth researching the rolling sphere method to make sure it'll work as intended
[automerge]1567862768[/automerge]
Just another thought, if it's a well known charity would it be possible to try contacting a manufacturer and see if they'd be willing to maybe sponsor at least some basic materials? Could also see if someone like Dehn would be able to provide any technical assistance I'd say it might be worth a try

 

[echase]

Following a quick read of http://www-public.tnb.com/eel/docs/furse/BS_EN_IEC_62305_standard_series.pdf I understand more. Interesting that they are talking of 10 ohm earth resistance as that means the earth is going to jump many 1000s of volts if no SPDs

One thing I don’t understand is the practice of putting a short spike (air terminal) somewhere on the roof peak as in pic below of a Zambian church roof with its 50mm2 copper down tape. Seems to be that the 90% of the rest of the roof is outside the protected zone therefore, but I guess if you use tape along the ridge this can stop this problem. Maybe I am just not observing the tape there. It would be cheaper to put our spikes on the roof ridge if ridge tape and laying the down wire on the roof is OK. But as the roof angle is shallow will it cover well down to the eaves?

The reason the contractor was suggesting 13m poles next to the building is to get around this as it puts the spike higher than the building, albeit offset. Also if the roof is metal and the spike is on the peak, laying the down wire/tape down the roof exposes the roof to the surge, or is that not a problem? The roofs are linked to house earth but only at one sheet of many and I am sure the electrical resistance between each sheet is large.

What he recommends is 1 pole each next to the 2 single storey houses, one house is in pic below (that one is not a metal roof). With a 45ish protection angle it’s not going to cover the whole building, which is quite long, but I was aiming to get it as close as possible to the solar panels to at least cover those.

Then there is a group of 3 larger single storey buildings. If the 2 poles are in the small gaps between buildings they cover a bit either side. E.g. in this pic one pole would behind the solar panels behind the flag pole and in front of the green roof, which is another building spaced 3m away. I guess the new pole would be a bit higher than the flagpole. The other pole would be to the left behind the taller section as there is another building behind there.

Clearly a lot of roof is outside the 45ish degrees. But BSEN 62305 talks of catenary wires and so would it be a good idea to run them between the tops of both poles and on to the furthermost corners of the roofs? This could obviate the need for ridge wires. But they will use 16mm2 for all this and I can see it just vaporising. But your suggested Al tape can't be that thick at that price.

My son thought that spacing the down or ridge wires/tapes more than about 0.5m away from the solar panels was needed to stop lightning jumping across to the panels, but seems to me that if the roof is metal it may jump anyway if any point of the wire is near the roof as the panel frames rest on the roof. E.g. the lightning poles may be almost touching the roof edge.

Charity giving is a good idea although we may be too small to qualify. Not an internationally known one.

[automerge]1567882901[/automerge]

But will it simply vaporise when carrying 1000s of amps from a strike?

I guess it could be said that an ionised/vapourized cloud of copper atoms makes a very good conductor. So maybe for one strike thin wire is effective. So as long as you replace the wire before the next strike maybe it is OK. One new 16mm2 wire every say 10 - 20 years is cheaper than one >50mm2 wire/tape for life.

There is no evidence the buildings on that site have been hit in the last 15 years (there would be holes in the roofs if they had as they are thin) but the grid nearby gets struck annually at least, or the strikes are near enough to induce a lot of current into the grid. (I was in Canada once when the whole of Quebec tripped out due to the aurora borealis inducing too much current into overhead wires.)

I am not sure whether lightning prefers metal things like buildings rather than tall things like trees. But on the evidence of the two known strikes there in 15 years, one hit a tree and one the long perimeter wire fence. One is taller and the other covers a bigger area target than the buildings.

 

[echase]

I have had a think about all this and come up with the attached which is my attempt to arrive at a lighting protection concept that is achievable in Africa considering the limited budgets usually available there. It includes some photos of "good" practice from Austria where I was last week. Before you scream at my incompetence please read the disclaimer at the end. Any constructive comments on it gratefully received.

I did a drawing of a rolling sphere for the house in the photo above and concluded that the proposed 13m pole beside the long roof edge would be OK for Level 2 protection (30m radius sphere) as long as we ran catenary wires from the top of the pole to the ends of the ridges, as the house would be otherwise unprotected all the way to the ends.

There is a warning on this forum that for international locations local standards and practice takes precedence over UK advice. However this is dangerous advice in much of Africa as local practice often falls way behind what is safe. For instance, if you believe IEC 62305 is correct, the local “expert” contractor is recommending a solution that does not protect the house without the catenary wires that I am suggesting. IEC 62305 may apply in Africa but I have never met anyone who uses it.

 

[Robson689]

We don't directly bond solar panels where I work as they're not able to withstand a lightning strike. The panel would be damaged so protection is offered with 3m air terminals suitably spaced.

I somewhat agree with your statement about prioritising SPD over lightning protection system as like you say the building could never be struck but being vulnerable to surges will probably hit you harder in regards to damage.

You seem to have a good understanding and I would love to help you but I don't have the technical knowledge about product specifics as you don't learn such in-depth information in college.

What is important to remember is that lightning protection is not law, it's mostly theory/research based and for thousands of years before modern regulations it was pretty well accepted that you just need a finial on the highest point with an earth rod as you said.

Whilst we may only do new install to IEC 62305 in the UK it doesn't mean that other methods aren't valid. I would generally expect labour & materials for a small level 4 installation to be in the thousands - I did repairs today to a BS 6651 system and those alone we're in excess of £3,000 just for earthing works so there are great costs involved with modern methods.

 

[echase]

Thank you for your support. I am very interested in your costing, if only because I may need to get a church's earthing in UK fixed as it is fractionally over the 10 ohm limit. In Zambia we have been quoted only £2200 for the four 15m poles, spikes/air terminals, 4 earth pits and wiring. That covers about 6 major buildings. If we add lots of catenary wires it may adequately protect all 6 buildings as long as 16mm2 wire is OK. Even that is quite expensive compared with their budgets and building costs .

The reason this is so much cheaper than UK is that labour costs may be 1/5th or 1/10th UK rates, plus they are scrimping on copper by using 16mm2 instead of UK's 50 to 75mm2.

They tend not to use earth rods there but bury coils of 16mm2 wire in a 2m deep by 1m square pit. Or use a short rod and a coil. They put a 0.5m layer of crushed salt and charcoal at the pit's bottom to increase conductivity because there can be no rain for months, so the earth get very dry.

Separately I am fascinated that the BS EN IEC standard uses quantitative risk assessment (QRA) to arrive at a probability of death based on the protection provided. QRA was my profession for decades and it astonishes me that this IEC standard is encouraging the lightning protection profession to calculate this. Normally it’s the preserve of the chemical, oil, rail, nuclear or aviation industries where multiple deaths may occur per accident and they have tons of safety professionals to work it all out. Lightning is not likely to kill more than one or two for each strike, so is not that “serious” a problem and I can’t believe the average lightning protection firm has the safety professionals to fully understand QRA. No chance they have in Africa. Some researcher has gone over the top in developing this IEC and foisting it on us.

I have seen an output from StrikeRisk software which calculated a 2.5976 E-06 probability of a death per year. The 9th commandment of the QRA profession is don’t quote more than one decimal place (no decimal place is preferred) as more places implies the answer is very accurate when in fact it is just an estimate because the input data is an estimate.

 

[Marvo]

I don't have any info specific to high risk lightning areas in Zambia but just uploaded some South African documents for you.

Here's the links to them, they're currently awaiting approval and should hopefully be available soon.

Surge Protection in Low Voltage Electrical Installations | ElectriciansForums.net | Free Electrical Advice - Electricians Forums: Electricians' Talk Forums

Surge Protection in Low Voltage Electrical Installations > Building Regulations Approved Documents > Technical Data Sheets & Downloads. Electrical Advice Forum. Electrical Advice and Electricians Downloads.

www.electriciansforums.net

SANS 61024-1 1990 | ElectriciansForums.net | Free Electrical Advice - Electricians Forums: Electricians' Talk Forums

SANS 61024-1 1990 > Building Regulations Approved Documents > Technical Data Sheets & Downloads. Electrical Advice Forum. Electrical Advice and Electricians Downloads.

www.electriciansforums.net

SANS 10313 | ElectriciansForums.net | Free Electrical Advice - Electricians Forums: Electricians' Talk Forums

SANS 10313 > Building Regulations Approved Documents > Technical Data Sheets & Downloads. Electrical Advice Forum. Electrical Advice and Electricians Downloads.

www.electriciansforums.net

SANS61024-1-2 | ElectriciansForums.net | Free Electrical Advice - Electricians Forums: Electricians' Talk Forums

SANS61024-1-2 > Building Regulations Approved Documents > Technical Data Sheets & Downloads. Electrical Advice Forum. Electrical Advice and Electricians Downloads.

www.electriciansforums.net

SANS 61024-1-1 | ElectriciansForums.net | Free Electrical Advice - Electricians Forums: Electricians' Talk Forums

SANS 61024-1-1 > Building Regulations Approved Documents > Technical Data Sheets & Downloads. Electrical Advice Forum. Electrical Advice and Electricians Downloads.

www.electriciansforums.net

Good luck.

 

[echase]

I somewhat agree with your statement about prioritising SPD over lightning protection system as like you say the building could never be struck but being vulnerable to surges will probably hit you harder in regards to damage.

The $64,000 question for us is it worth spending money on air terminals at all? Our buildings are mostly corrugated “iron” roofed and of low value. If lightning blows a hole in the thin roof it’s cheap to repair.

What we don’t want though is the occupants to be hurt or the electronics to be damaged or a fire to start. Is simply earthing the roof (plus SPDs) an adequate protection for the occupants and electronics? Pro this solution is that, as the roof is bonded in at least 2 places to a good earth, the whole roof is a “air terminal”. Against it is that if the roof is punctured or the bonding between sheets is inadequate (I am sure it is as corrugated roof sheets are only nailed together and the nails and holes get corroded, whether iron or aluminium) will the lightning jump from the roof to nearby water pipes or mains cables and the metal conduits for such (or rebar in the walls, although not sure we have any). In this case people and electronics are at risk despite the fitted SPDs. The pipes are not earth bonded.

Any thoughts?

Many thanks Marvo. Look forward to seeing those reports. Clearly South Africa will be one of most forward thinking countries in Africa in this area. But it is also has larger budgets to spend so what can be achieved there may not read across to say Zambia or worse still South Sudan, where I also visit.