Should a giant steel framed human aquarium be earth bonded?

[Deuce]

I have a large, portable steel framed tank. approx 6m x 2m x 2m. It is used for actors/models to perform in for a variety of productions.

For many years now, I have had conflicting advice about whether or not it should be earth bonded. The initial response is "it's a giant steel structure, bond it!!". But then invariably someone else comes along and points out that, should the tank itself become energised, the people in the water within the tank would become part of the return to earth - the current would flow through the tank and them. If it was not bonded, as it is by design insulated, they would be ok. At least until they tried to leave the tank.

To date our solution has been to be super cautious of anything that could even present the possibility of bringing current to the tank itself. But how do we make a once and for all decision on whether it should be bonded without any professional consensus of opinion?

 

[static zap]

Have pondered this ,as occupants of a hot tub , reach ..Double insulated status ... but who knows with water around ..
__Some form of alarm of Risky potential - Should be available
( but this "Specialist situation is sure to be Expensive" )
If studio /stage lighting is not on RCD , I understand your reluctance to add an earth .
Should a light / cable break free and go fishing !

 

[Deuce]

Have pondered this ,as occupants of a hot tub , reach ..Double insulated status ... but who knows with water around ..
__Some form of alarm of Risky potential - Should be available
( but this "Specialist situation is sure to be Expensive" )
If studio /stage lighting is not on RCD , I understand your reluctance .
Should a light / cable break free and go fishing !

The hot tub analogy is a good one, and having modified a few I certainly never saw any method of earthing the water body directly, although the pump/heater would be earthed which would pass through to the water in the system. The same is true of the pump/heaters attached to the water body in my tank, so it too is already technically earthed - albeit a pretty poor connection (relatively clean water isn't really a sound way of making a connection).

I think the crux of the debate comes down to:

What is the risk of a none RCD protected piece of equipment falling in to an unearthed tank vs the risk of RCD protected equipment falling into an earthed tank.

Having seen literally millions of lights etc rigged at high level I can say I have never seen one fall. If one did fall, it would almost certainly break the power connection on the way down as the power cables are invariably threaded through long lengths of truss, or mechanically clamped in some way at a high level. And then as the "all on RCD" rule is so widely enforced, the chances of a light falling, the power cable reaching the tank in tact and not being RCD protected is incredibly small. It would truly be a freak incident.

The only other scenario is a mechanical failure of the tank itself, leading to instantaneous flooding of the area. That could contact power ahead of RCD in theory, but the grounding effect of water of such a large area would dissipate current very efficiently, not to mention it would be in contact with several other grounded steel structures - so long as no one tries to rescue the situation by picking a power cable...

So I think I have come to a conclusion, the tank should be earthed, that seems to be the majority view on here and on site. Even if a freak occurrence did result in a shock that wouldn't have happened had the tank remained isolated, it is still true to say that the greatest likely risk was considered and mitigated by earthing it.

Anyone who feels this thinking is still worth challenging please do speak up.

 

[Nigel]

This is quite an interesting thread.

As I see it there are differnet risks:

1. The risk of someone entering the pool and a potential difference between the pool and the ground.

This could be mitigated by ensuring the rubber matting you have provided at least a 1.5m overlap so that nobody could be touching true earth and your bonded pool.

2. Users inside the pool completely and equipment possibly bringing the water to mains potential.

This can be mitigated by using Class 2 equipment only or SELV only. Preferable the latter. An RCD plug on all your equipment also which would need to be tested periodically to ensure it works. I would not like to rely on the customers RCD protection not knowing the state of it.

Another option would be to provide an earth free equipotential bonding zone within the pool. This would mean all exposed conductive parts and extrenuous conductive parts would need to be bonded together so that they are at the same potential. The risk here is to people that are not in the pool and could touch the metal frame and true earth.

3. Users in the pool and able to touch electrical item outside of the pool that form a part of the building electrics.

This is an issue as you cannot really ask the customer to move their lights etc. So this would have to be surveyed and if any electrical equipment fell inside of the zones then you would have to decline that position of siting your pool.

As you can see there are so many things to consider each site will need to be thoroughly surveyed each time.

To be honest I would be inclined to suggest the following:

Pool only to be sited on insulated matting with a 1.5m overlap and so that no electrical iteams are within zones as per BS7671.

All electrical equipment to be Class 2 or SELV and be RCD protected by your own RCD plugs.

When people are in the pool all electrical items to be disconnected from the electrical supply and reinstated when all people are clear. Completely remove the risk when people are in there.

 

[Deuce]

This is quite an interesting thread.

As I see it there are differnet risks:

1. The risk of someone entering the pool and a potential difference between the pool and the ground.

This could be mitigated by ensuring the rubber matting you have provided at least a 1.5m overlap so that nobody could be touching true earth and your bonded pool.

2. Users inside the pool completely and equipment possibly bringing the water to mains potential.

This can be mitigated by using Class 2 equipment only or SELV only. Preferable the latter. An RCD plug on all your equipment also which would need to be tested periodically to ensure it works. I would not like to rely on the customers RCD protection not knowing the state of it.

Another option would be to provide an earth free equipotential bonding zone within the pool. This would mean all exposed conductive parts and extrenuous conductive parts would need to be bonded together so that they are at the same potential. The risk here is to people that are not in the pool and could touch the metal frame and true earth.

3. Users in the pool and able to touch electrical item outside of the pool that form a part of the building electrics.

This is an issue as you cannot really ask the customer to move their lights etc. So this would have to be surveyed and if any electrical equipment fell inside of the zones then you would have to decline that position of siting your pool.

As you can see there are so many things to consider each site will need to be thoroughly surveyed each time.

To be honest I would be inclined to suggest the following:

Pool only to be sited on insulated matting with a 1.5m overlap and so that no electrical iteams are within zones as per BS7671.

All electrical equipment to be Class 2 or SELV and be RCD protected by your own RCD plugs.

When people are in the pool all electrical items to be disconnected from the electrical supply and reinstated when all people are clear. Completely remove the risk when people are in there.

We always have a minimum 2m extension of the rubber matting all round the tank, so that's already in place. we also move all electrical outside of this area.

Everything on the supply side and our own equipment, and any other equipment attached anywhere in the space is on RCD and is tested as it is bought together and connections made.

I can and do keep to elv/selv within the 2m zone as the only stuff in there is generally my own design of SFX, which is all 12/24v. Beyond that 2m area though, none of the kit that you would find in a studio exists as class 2 or SELV.

In your scenario you're suggesting not earthing anything on or immediately around the tank then? That feels a little like, as I think you inferred yourself, taking one danger away from those in the tank and gifting it to those around the tank where the zone ends.

 

[DPG]

I know you said you can't drop to 110V, but are you able to feed everything in the area via a 10mA RCD?

 

[marconi]

Deuce: I turned up this reference from the IET wiring matters series (but can see no date on it):

Swimming pools: to bond or not to bond? - https://electrical.------.org/wiring-matters/issues/51/swimming-pools-to-bond-or-not-to-bond/

As this is a safety matter (and you are not paying me) I have only given you my thoughts. The article above is written by someone who could give you some professional advice on your special situation.

 

[Deuce]

I know you said you can't drop to 110V, but are you able to feed everything in the area via a 10mA RCD?

No chance. My kit, yes - sort of... Couple of 3hp motors on the pumps though so I'd have to upgrade them to soft start to avoid knocking out the RCD (capacitor driven single phase pumps do that below 30ma)

Everyone else kit though... Could be 30 teams of people I have never met before arriving on site, all of whom will have 30ma. Not to mention most distro's are hired in and will have 30ma as standard.

Not to mention, if the tank is at Excel in London, or Palexpo in Geneva, there could be upwards of five thousand RCD's in the same room as the tank. That's a lot of foreign speaking folk to persuade!

 

[Deuce]

Deuce: I turned up this reference from the IET wiring matters series (but can see no date on it):

Swimming pools: to bond or not to bond? - https://electrical.------.org/wiring-matters/issues/51/swimming-pools-to-bond-or-not-to-bond/

As this is a safety matter (and you are not paying me) I have only given you my thoughts. The article above is written by someone who could give you some professional advice on your special situation.

That's an excellent idea! At least that way if ever anything did happen there can be no question I've gone to an undeniable expert for advice.

Reading the article it looks very much like he will conclude to bond it, but I will make an approach and go from there.

 

[plugsandsparks]

Apart from dropping the RCD down to 10mA, which appears to be a non starter, my only other suggestion is to make sure you have a de-fib available and someone trained to use it, that way you have covered the after shock issues. Everything else like SELV has been covered and yes earthing to the building met on a TNCS system has its own risks associated with making all earth metal live at 240V if you loose the neutral. You definetly need to make sure you know for certain what earthing system is in place for every building you go to. Tricky one for sure.

 

[Deuce]

Apart from dropping the RCD down to 10mA, which appears to be a non starter, my only other suggestion is to make sure you have a de-fib available and someone trained to use it, that way you have covered the after shock issues. Everything else like SELV has been covered and yes earthing to the building met on a TNCS system has its own risks associated with making all earth metal live at 240V if you loose the neutral. You definetly need to make sure you know for certain what earthing system is in place for every building you go to. Tricky one for sure.

To be fair, the sparks working on each production have the final say over what is and isn't safe, so it's really their call and they will know what earthing system is in place, and can debate the pro's con's of their final choices. But I thought I would ask on here and get as many eyes as possible on the problem as, honestly, even the sparks on site feel a bit baffled sometimes I think

My plan now is to get a report written up by a specialist on the things to consider when making such decisions, and the implications etc.. And I will circulate that amongst the sparks in advance. I have installed a tapped earth fixing point on the tank already, that they may use to connect to whatever they judge is best.

That's about all I can do, and arguably more than I need to do as the reality is I don't have control over the site. But I of course want to do all I can.

 

[Nigel]

We always have a minimum 2m extension of the rubber matting all round the tank, so that's already in place. we also move all electrical outside of this area.

Everything on the supply side and our own equipment, and any other equipment attached anywhere in the space is on RCD and is tested as it is bought together and connections made.

I can and do keep to elv/selv within the 2m zone as the only stuff in there is generally my own design of SFX, which is all 12/24v. Beyond that 2m area though, none of the kit that you would find in a studio exists as class 2 or SELV.

In your scenario you're suggesting not earthing anything on or immediately around the tank then? That feels a little like, as I think you inferred yourself, taking one danger away from those in the tank and gifting it to those around the tank where the zone ends.

I am struggling to see a case for your pool to require earthing or bonding to be honest. If no conductive parts can be touched while touching the pool structure I fail to see where the risk is.

Or am I missing something?

 

[Lucien Nunes]

I would earth it.

Hazards for people in the tank basically stem, as you say, from live equipment falling in. The tank itself is quite a good Faraday cage or equipotential zone of its own, due to the metal structure. Unless you have something live underwater, unprotected by an RCD and not enclosed within a substantially earthed metal casing (which most ents lighting is), there won't be a potential gradient through the water. You could probably drop a live PAR can into an earthed tank without lethal harm, as almost all the leakage current would be collected by the can's earthed casing.

Hazards for people outside the tank are less easily defined. for example (and I know this is very unlikely given your operation method) a live cable were gashed against the tank, the whole frame could become live without the fault being cleared unless it is earthed. The people inside will be safe either way as they are in the tank's own equipotential. But if the floor membrane were wet, as I am sure it sometimes is, a shock path could be formed for someone approaching and touching the tank, that only the RCD would save them from. From a distance, I think this is the higher risk due to the greater probability.

when we're in the middle of nowhere relying on generators, TT.

Genny supplies are usually TN-S. The rod at the genny provides an earth reference just as the earth mat or rod does at a substation, but the distribution cable CPCs take the earth from the genny all the way to the loads. TT means that the load CPCs are connected to the star-point only via the rod.

 

[Nigel]

That's an excellent idea! At least that way if ever anything did happen there can be no question I've gone to an undeniable expert for advice.

Reading the article it looks very much like he will conclude to bond it, but I will make an approach and go from there.

No it doesn’t. It says only bond if extraneous. Which if on a rubber mate is certainly is not.

 

[Wilko]

Hi - if the supply is DNO PME, then one concern (for me) is their N being lost upstream. That would be an off-site incident totally out of your control but it could be dangerous for everyone on set. With local TT it's just a power outage.

 

[Nigel]

Hi - if the supply is DNO PME, then one concern (for me) is their N being lost upstream. That would be an off-site incident totally out of your control but it could be dangerous for everyone on set. With local TT it's just a power outage.

The op has said that it will be on a rubber mat so not really an issue if PME.