How does Touch voltage work without RCD protection

[King84]

Hi everyone
I asked my assessor about touch voltage and he was not keen on explaining properly and gave me a rough answer which made me think probably he was not sure himself or was hiding it.

basically my understanding is afcourse that how 30mA RCD works to limit touch voltage to 50V is by 50/0.03 as this gives enough 1667ohms resistance margin so as long as resistance is within this limit, the voltage will remain under 50V.

My question was with assessor as in TT system as we do need to have 100mA RCD afcourse for fault protection as due to low fault current and higher Ze but what if we dont have 30mA RCD for additional protection then what gives additional protection in that situation and upon hearing this, the assessor just gave me Touch volatage will be limited to 50V as use the formula so that got me confused how Ra=50/Ia so lets say we have higher resistance due to which we will be having lower fault current then how additonal protection is acheived in that scenario ? am I missing out something

please if anyone could briefly explain how touch voltage occurs with RCD and without RCD on normal MCBs how touch voltage is acheived? thanks tons for your great help

 

[timhoward]

I think you might be chucking too many different things into the same blender.
You need fault protection. That is achieved either through low impedance to earth so there is sufficient fault current to cause ADS or detecting leakage to earth using an RCD.
On TT systems the touch voltage shall be limited exactly as you said.
On TN systems ADS should occur instantly when a fault occurs so the touch voltage concept isn’t particularly relevant in most cases.

The choice of RCD technology is a separate matter again, and the general rule is that for discrimination you need X3 upstream. So if individual circuits are protected by 30ma RCDs which the regs require for many things then you need >90ma so in reality 100ma RCD to cover the entire installation.
If you only have the 100ma RCD for fault protection then that isn’t meeting the requirements of the regs for additional protection of E.g. socket circuits or domestic lighting circuits.
(Some older TT installs have a main switch RCD rated at 30ma.)
Does that help at all?

 

[davesparks]

The choice of RCD technology is a separate matter again, and the general rule is that for discrimination you need X3 upstream. So if individual circuits are protected by 30ma RCDs which the regs require for many things then you need >90ma so in reality 100ma RCD to cover the entire installation.

Just having a tripping current of x3 doesn't give you discrimination with RCDs, you need a time delay long enough to allow the downstream RCD to operate before the upstream one does.

 

[timhoward]

Just having a tripping current of x3 doesn't give you discrimination with RCDs, you need a time delay long enough to allow the downstream RCD to operate before the upstream one does.

Sure, I’m on holiday typing on my phone….
I should indeed have said S type or R type
or manually configurable for the upstream device. ( X3 I delta N or above also needed)

 

[pc1966]

You also have to remember that RCD do not limit the current, or by implication limit the voltage. What they limit is the exposure time at currents above the trip threshold.

The maximum values of Ra typically given from 50V/In for the RCD are the upper limit of resistance on a running system with a controlled leakage of current that in not guaranteed to trip the RCD and which will keep the continious 'touch voltage' below 50V. In earlier versions of the regs I think it was reduced 25V for agricultural areas, etc (i.e. half the resistance).

That is why you can't simply use RCD in very hazardous areas, such as in or very close/in a bath or swimming pool, since your RCD disconnection time might still allow a seriously high shock current through a wet conductive body that can kill or cause injury even if it only lasts for tens of milliseconds. Hence you use ELV there.

 

[King84]

I think you might be chucking too many different things into the same blender.
You need fault protection. That is achieved either through low impedance to earth so there is sufficient fault current to cause ADS or detecting leakage to earth using an RCD.
On TT systems the touch voltage shall be limited exactly as you said.
On TN systems ADS should occur instantly when a fault occurs so the touch voltage concept isn’t particularly relevant in most cases.

The choice of RCD technology is a separate matter again, and the general rule is that for discrimination you need X3 upstream. So if individual circuits are protected by 30ma RCDs which the regs require for many things then you need >90ma so in reality 100ma RCD to cover the entire installation.
If you only have the 100ma RCD for fault protection then that isn’t meeting the requirements of the regs for additional protection of E.g. socket circuits or domestic lighting circuits.
(Some older TT installs have a main switch RCD rated at 30ma.)
Does that help at all?

Thankyou very much for detailed reply and it all makes sense but I have one more question thats bothering me and couldnt find the answer.

Lets say RCD operates in 40ms at 5 times to give aditional protection which means accidently someone touches live cable , RCD operates in 40 ms at leakage of 30mA current which will obviously protect from electric shock.

My question is there are many old installations which obviously are still installed and it is said because they are old doesnt mean they are unsafe.
If anyone accidently touches live cable and considering Zs are all below 80% then I do understand MCB would operate within 0.4s but would that prevent electric shock?
I thought maybe touch voltage play any part here .sorry but I am still trying to grasp this part and i appreciate your help

 

[pc1966]

My question is there are many old installations which obviously are still installed and it is said because they are old doesnt mean they are unsafe.
If anyone accidently touches live cable and considering Zs are all below 80% then I do understand MCB would operate within 0.4s but would that prevent electric shock?

Without an RCD if someone touches the cable it will not disconnect.

The Automatic Disconnection of Supply (ADS) that has been in the regulations for decades is about a fault live to earthed metalwork (i.e. to the CPC) with someone in contact with the CPC (i.e. that metal object) and also the true Earth (i.e. defined zero voltage).

That is what the Zs "fault impedance" is all about, that if you get a fault L-E then so much current flows that the MCB or fuse disconnects in under 0.4s (for a TN supply, nominally 230V) or 5s for sub-main.

If you have a TT system, so the CPC goes to an earth rod, then the disconnection times drop to 0.2s and 1s respectively.

Why? Well in the TN case your R1 and R2 are usually of the same order, so a L-E fault causes roughly half the supply voltage, say 115V to appear during a fault as the voltage drop on R1 and the voltage rise on R2 are about the same. Whereas for a TT system R2 is usually very much bigger than R1 so under fault conditions almost all of the 230V appears on your metalwork, and so the disconnection times have to be less so the shock risk is less since the unfortunate handler of the metalwork gets about* twice the current for a given sweaty palm's resistance, etc.

Also the TT level of 'R2' is usually so high that even a 6A MCB won't trip. Ever. You are usually reliant on a RCD for fault protection, and that traditionally for domestic supplies was a 100mA delay RCD as an incomer along with possibly a 30mA 'instant' RCD for high risk circuits like socket outlets. High risk in that prior to RCDs being the norm and cheap-ish, every summer a dozen or so folks in the UK would die from mowing an extension lead, etc, in the garden.

As already mentioned above when you have RCD in series and want selectivity you need each level up-stream to be around x3 higher threshold current, as well as about an extra 0.2s delay so a big fault can be isolated downstream if possible, and if not then shortly after the next level up goes. Few systems have more than a 300mA/100mA S-type then 30mA instant, but if you had a larger system then the next level up would typically be an adjustable RCD set to 1A/300mA and about 0.4s delay.

[*] Human body resistance is not linear, above a certain voltage it drops making life harder and potentially very much shorter.

Electrical injury - Wikipedia

en.wikipedia.org

 

[Aaron b]

Thankyou very much for detailed reply and it all makes sense but I have one more question thats bothering me and couldnt find the answer.

Lets say RCD operates in 40ms at 5 times to give aditional protection which means accidently someone touches live cable , RCD operates in 40 ms at leakage of 30mA current which will obviously protect from electric shock.

My question is there are many old installations which obviously are still installed and it is said because they are old doesnt mean they are unsafe.
If anyone accidently touches live cable and considering Zs are all below 80% then I do understand MCB would operate within 0.4s but would that prevent electric shock?
I thought maybe touch voltage play any part here .sorry but I am still trying to grasp this part and i appreciate your help

That's not the point of ads. Basic protection is to stop contact with conductors which used to be referred to as direct contact. Ads for fault protection or indirect contact, it's also all based on a single fault. Having all exposed-conductive-parts all connected to the MET should help with the temporary rise in potential.

 

[davesparks]

My question is there are many old installations which obviously are still installed and it is said because they are old doesnt mean they are unsafe.

Not quite, there is a line in the regulations that says installations carried out to previous versions of the regulations may not necessarily be unsafe.
This does not mean that all old installations are automatically safe, it means that they may or may not be.

If anyone accidently touches live cable and considering Zs are all below 80% then I do understand MCB would operate within 0.4s but would that prevent electric shock?

If someone touches a live cable the MCB won't operate because of that alone, or if it does they will be very much dead.

Also, RCDs don't prevent an electric shock, they reduce the severity of the shock.

 

[LastManOnline]

Hi everyone
I asked my assessor about touch voltage and he was not keen on explaining properly and gave me a rough answer which made me think probably he was not sure himself or was hiding it.

Touch voltage is a very simple concept when you look at it from the perspective of ohms law.. current flowing in the circuit multiplied by the resistance (earth rod in TT system). This voltage will be present on all earthed /bonded metalwork connected to the rod.

basically my understanding is afcourse that how 30mA RCD works to limit touch voltage to 50V

The rcd has no affect on the "touch voltage" whatsoever. It simply limits the time that voltage is present on the affected metalwork.

please if anyone could briefly explain how touch voltage occurs with RCD and without RCD on normal MCBs how touch voltage is acheived?

Again, the safety devices have no ability to limit the touch voltage. Their purpose is to limit the time factor that said voltage will be present on affected metalwork or persons.

thanks tons for your great help

You are welcome

 

[King84]

Thanks alot everyone, you guys have been very helpful I really appreciate. I wish my books could explain the way you guys explained.

This makes all sense now RCD does not have impact on the current but effects time exposure and ADS is caused by fault current so it has nothing to do with touch voltage.

I am only wondering like you guys mentioned just because regs says as older installations have no RCD does not make them necessarily unsafe so I understand as ADS is there which is caused by L-E . afcourse along side with main protective bonding etc.

If there is a metal frame light switch and property is old with no RCD and there is some contact between Line conductor and earth or lets say metal work, in the situation how would the disconnection is acheived to prevent electric shock?

MCBS only provide overloading and short circuit protection but would ADS provide protection in the situation mentioned ?

sorry guys I am still in learning stage , maybe I am overthinking about it

 

[Aaron b]

I am only wondering like you guys mentioned just because regs says as older installations have no RCD does not make them necessarily unsafe so I understand as ADS is there which is caused by L-E . afcourse along side with main protective bonding etc.

If there is a metal frame light switch and property is old with no RCD and there is some contact between Line conductor and earth or lets say metal work, in the situation how would the disconnection is achieved to prevent electric shock?

It gets thrown around a lot that 'the regs aren't retrospective' as if it's the 11th commandment but you need to consider the change of use due to technology too. You will see a lot of properties with only one socket in a room there just wasn't the same amount appliances back then and no electronics. If it's a old lady that still lives like it's the 70s it might still be safe...

In the situation you mention the metal frame would be earthed and the Zs would be low enough to disconnect within 0.4 secs. We only consider one fault so the external conductive part (which I assume isn't in contact with earth or bonded) couldn't become live without two simultaneous faults. A fault outside of the enclosure would be avoided by double insulated wiring which should be adequately protected against mechanical damage.

 

[timhoward]

Keep the questions coming.

If there is a metal frame light switch and property is old with no RCD and there is some contact between Line conductor and earth or lets say metal work, in the situation how would the disconnection is acheived to prevent electric shock?

This would cause very high current flow.
Nominal voltage (230) divided by resistance to earth (under 5.82 ohms if installed to regs and it’s a B6 breaker) gives fault current of 39.5 amps.
The graph for a B6 breaker shows this would trip instantaneously.

On a TT system it’s because the resistance to earth will be much higher and there won’t be enough fault current that a different method is needed (an up front RCD)

(Feel free to keep asking questions - but for basic theory John Ward made some very good YouTube videos.)

 

[King84]

This makes sense now as in TT system we have higher earth resistance so we cannot rely on ADS to occur hence we need time delay 100mA RCD for fault protection and then for higher risk areas 30mA RCD for additional protection.

In TN as you guys explained ADS provides fault protection by disconnecting in 0.4s for final circuits and submain in 0.5s hence RCD is not required as long as we have additional protection in baths such as supplementary bonding or if the sockets are not in a place suspected of feeding outdoor then even without RCD older installations are safe as long as ADS conditions are met.

Sorry the above definitions were just for myself that I understood the concept from you guys properly lol.

Thanks alot everyone, you guys made this so much easier and I really appreciate you guys taking the time and effort to explain.
I wish the videos and books were interactive like you guys made the whole concept so much easy even someone like me understood this.
I tried asking elsewhere on another platform and guys were not helpful.
cheers to you All!

 

[Aaron b]

It's good practice to install a 100mA type S upfront but not a requirement of bs7671. A single 30mA can serve both fault and additional protection.

 

[King84]

It's good practice to install a 100mA type S upfront but not a requirement of bs7671. A single 30mA can serve both fault and additional protection.

Thanks Aaron but if 30mA can provide Fault and additional protection then why 100mA RCD is required upstream?

 

[Aaron b]

Thanks Aaron but if 30mA can provide Fault and additional protection then why 100mA RCD is required upstream?

It's not but most would feel safer having two devices.

 

[davesparks]

It's good practice to install a 100mA type S upfront but not a requirement of bs7671. A single 30mA can serve both fault and additional protection.

Since when has that been good practice?

I can't see any logic to it other than inviting unnecessary nuisance tripping.

 

[timhoward]

It's good practice to install a 100mA type S upfront but not a requirement of bs7671. A single 30mA can serve both fault and additional protection.

You often find older 5 way consumer units with an 30ma RCD main switch providing fault protection and additional protection. (These are the least likely RCD's to ever be tested and I've known quite a few not work)
It's worth adding that for a few reasons it's good if each circuit (or at least group of circuits) have it's own RCD.
a) with the general increase of leaky equipment, if you gather all the leakage of every circuit together this can increase nuisance tripping.
b) with the increase in the amount of DC leakage it's good not to gather it all together in case it blinds the RCD and it fails to operate.
c) It can help if there's a fault to know which instantly know which circuit or group of circuits is the likely culprit. It isn't always this simple but it's a good start.

Most of the time these days every circuit has it's own combined MCB and RCD (RCBO) 30ma protection and this is acceptable for TT installations.
If the consumer unit is a older split load board, with some circuits NOT RCD protected, then the main switch would be an S-Type 100ma RCD on a TT system, or there would be a similar RCD in it's own enclosure before the board.

 

[Aaron b]

Since when has that been good practice?

I can't see any logic to it other than inviting unnecessary nuisance tripping.

I suppose it each to there own when going over and above spec. Your right with considering nuisance tripping with the increased earth leakage from electronics.

 

[davesparks]

I suppose it each to there own when going over and above spec. Your right with considering nuisance tripping with the increased earth leakage from electronics.

I have no issue with going over and above to improve a situation if it is genuinely well thought out and actually improves a situation. However you have not presented this as a way of going over and above, you presented it as good practice, this implies that the industry as a whole approves of it rather than being your personal opinion.

I still fail to see the point of installing an extra 100mA RCD on a normal TN installation.

A 100mA RCD does not provide additional protection, it will allow a fatal electric shock before it operates so does not provide any back-up to a 30mA RCD. It may however provide a false sense of security for someone who doesn't understand this.

For earth fault protection you already have multiple means of detecting those faults either by MCB and 30mA RCD or 30mA RCBO.

Nuisance tripping will occur with neutral to earth faults because RCBO's (assuming they are used) generally do not disconnect the neutral, so the 100mA RCD will trip as well as the 30mA RCBO. This will defeat the whole object of dividing the installation up across multiple RCDs. Again, in my opinion, it gives a false sense of security whilst having the potential to create a larger problem.

 

[nicebutdim]

I still fail to see the point of installing an extra 100mA RCD on a normal TN installation.

I may have mis-read the comment, but my take was that the suggestion was in relation TT installations.

While not required, it does add an additional layer of protection where ADS may rely entirely upon on electronic components.

 

[Aaron b]

I still fail to see the point of installing an extra 100mA RCD on a normal TN installation.

The OP was referring to a TT installation. I'm assuming he has come across the 100mA type S and was just trying to provide some insight to why it might have been done...

 

[King84]

Dave so if on TT assuming we are not using RCBOS but splitboard RCD 30mA and for mainswitch 100mA then neutral is no longer an issue so for selectivity time delayed RCD 100mA is a good choice alongside with 30mA RCD ?
Sorry I am just trying to understand why on existing so many TT people are using 100mA if it will cause nuisance tripping and as you mentioned RCBO scenario so I am assuming this is the main reason ?

 

[davesparks]

Dave so if on TT assuming we are not using RCBOS but splitboard RCD 30mA and for mainswitch 100mA then neutral is no longer an issue so for selectivity time delayed RCD 100mA is a good choice alongside with 30mA RCD ?
Sorry I am just trying to understand why on existing so many TT people are using 100mA if it will cause nuisance tripping and as you mentioned RCBO scenario so I am assuming this is the main reason ?

Yes if double pole 30mA RCDs are installed, or DP/SPSN RCBO's, then selectivity will occur on either a L-E or N-E fault with an upstream 100mA S type RCD.

If on a TT system you have circuits which do not require additional protection at 30mA then a 100mA time delayed RCD would be used to provide fault protection.
A 100mA time delayed RCD may also be installed to provide protection for the wiring within the DB which is not protected by the RCBO's etc.


With a TT system in some respects you have to choose the lesser of 2 evils as far as the neutral to earth fault problem goes, the tripping problem still exists but the danger of not having a 100mA RCD can outweigh the danger of tripping causing an entire installation to go off.

Also consider how much diverted neutral current there will actually be in a N-E fault on a TT system.

Why you will find a lot of TT systems with a 100mA RCD mainswitch and all RCBO's after it may well be due to them being installed by people who blindly follow a book rather than thinking, or people who just don't bother to keep up to date with their knowledge, or even people who woukd rather listen to gossip and rumours than actually read and understand the rules.

 

[nicebutdim]

Why you will find a lot of TT systems with a 100mA RCD mainswitch and all RCBO's after it may well be due to them being installed by people who blindly follow a book rather than thinking, or people who just don't bother to keep up to date with their knowledge, or even people who woukd rather listen to gossip and rumours than actually read and understand the rules.

Given the fact that RCD failure isn't particularly uncommon, would you consider additional 100mA Type S protection to be unwarranted for TT systems?


Edit: This assumes the installation of DP RCBOs

 

[Aaron b]

I think it's fair to assume double pole breakers would be used on a new TT installation as it would be needed for isolation, right?

 

[davesparks]

I think it's fair to assume double pole breakers would be used on a new TT installation as it would be needed for isolation, right?

Would it be?

 

[davesparks]

Given the fact that RCD failure isn't particularly uncommon, would you consider additional 100mA Type S protection to be unwarranted for TT systems?


Edit: This assumes the installation of DP RCBOs

I don't consider it to be unwarranted on a TT installation.

Personally if I was installing a domestic TT installation I would prefer to install a 100mA S type main switch and SPSN RCBO's for all circuits.
I'd also install the 100mA RCD in a separate insulated enclosure as I feel it is a justifiable departure to do so. I'd probably use the enclosure from a REC2 or similar as they clamp the tails quite nicely and are pretty sturdy.

This is of course personal opinion

 

[timhoward]

Why you will find a lot of TT systems with a 100mA RCD mainswitch and all RCBO's after it may well be due to them being installed by people who blindly follow a book rather than thinking, or people who just don't bother to keep up to date with their knowledge, or even people who woukd rather listen to gossip and rumours than actually read and understand the rules.

All agreed. I'm not quite ready to say that are no circumstances with an RCBO board where it makes sense though.
I know you've mentioned single vs double pole RCBO's, but if double pole are used it's a legitimate 2nd chance at fault protection should for whatever reason the downstream device not operate.

It covers some rather extreme scenario's that I accept should not happen in the first place such as a bus bar to case fault or someone slipping with a screwdriver.

Finally several brands of CU have an SPD kit that is fed from an MCB. It's highly unlikely, but a faulty SPD would otherwise have no fault protection. I've not had to consider this one yet, and whether I'd change the MCB for an RCBO etc.

I also prefer an up-front enclosure as whatever the regs say I've never liked metal board + TT.
It used to be such an absolute no-no and very quickly became fine as long as a tails gland and clamp is used.

 

[pc1966]

This makes sense now as in TT system we have higher earth resistance so we cannot rely on ADS to occur hence we need time delay 100mA RCD for fault protection and then for higher risk areas 30mA RCD for additional protection.

Strictly speaking it is always ADS as the supply is automatically disconnected on a fault.

The difference is on a TN system then usually you can achieve it on the over-current protection device (fuse, MCB, etc) due to low Zs if a reasonable design permits it, but on TT it almost always has to be an RCD due to the relatively high Ra (and so high Zs and low PFC).

But you can get situations on TN with high current final circuits or sub-mains where it is not feasible to use OCPD to meet the disconnections times and so then you are back to using an RCD in combination with OCPD. For many final circuits that can just be an RCBO, but for sub-mains you might be looking at fuse-switch or MCB + delay RCD combinations, or fancy (and expensive) MCCB that feature adjustable earth leak trip settings, etc.

 

[pc1966]

Finally several brands of CU have an SPD kit that is fed from an MCB. It's highly unlikely, but a faulty SPD would otherwise have no fault protection. I've not had to consider this one yet, and whether I'd change the MCB for an RCBO etc.

My concern is the RCBO tripping on any surge and nobody notices the SPD has not been isolated, then next surge and some hardware gets broken.

An up-front delay RCD should ignore a short spike anyway, and if it did trip on a failed-to-short SPD then at least you find out. I suspect most SPD would fail open, as they usually have solder joints that are designed as ultimate thermal disconnection, but I would never rule out a short-fault as impossible.

I also prefer an up-front enclosure as whatever the regs say I've never liked metal board + TT.
It used to be such an absolute no-no and very quickly became fine as long as a tails gland and clamp is used.

The use of a proper tail gland and the 19-strand flexible tails should avoid any risk of a short before the RCD/RCBO but I can see the attraction of an up-front RCD being in its own insulated enclosure.

Many of those rules are to try and reduce the inevitable impact of rubbish workmanship:

• A well-installed metal CU should be perfectly safe on TT
• A a well-installed plastic enclosure for RCD the same against fire

But given the possibility of SPD getting very hot under major faults (lightning hit to pylon/building, or worst-case open-PEN allowing ~400V L-N) I personally would only ever fit SPD in non-combustible locations.

 

[Aaron b]

Would it be?

Double pole isolation is required under 462.2, and as you mentioned earlier disruption should be avoided by separating circuits. Since double pole breakers are practically attainable what would be the reason not to install them?

It might be worth noting it was an earlier version of the regulations that said that an upfront RCD was required with a class 1 CU (before the change to metal cu's in dwellings). Now it mentions double insulation of the conductors supplying the RCCB. That might make split load boards harder to use on TT systems and a cheep board upgrade might not comply.