Will Zs measurements be redundant under the 17th Edition?

[TonyM58]

now theres a thought.....what would be the point in taking Zs measurements in domestic premises when the 17th edition kicks in. Everything will be on RCDs (pretty much most of the time)

So as long as the Ze is less than the magic 1,666 ohms (which it almost certainly will be) what would be the point (or relevance) of the Zs measurement?


Discuss please!

 

[Cirrus]

he he good point Tony - hadn't thought of that

 

[Minky]

Supposing Disconnection times still have to be met then Zs measurements are the very least you can do before even considering testing the RCD.Lets face it without a path to earth,what good is an RCD!!.Having had no chance of reviewing 17th edition,i would assume that 0.4secs would still apply to socket outlets supplying portable appliances,and that no other factors require to be taken into consideration.Earth loop is also a fundemental test to establish the condition of all installation wiring.When looking at a commercial or industrial installation Zs is of greater significance as loading factors/fluctuating impedance/noisey paths to earth and ambient temperature all greatly alter a good reading on one day and a bad one the next.After thats all said and done,although it isnt the hardest test to do,it certainly can be one of the more awkward.

 

[TonyM58]

Minky, thanks for the reply

Of course disconnection times still have to met, but because all circuits are going to be on 30ma rcd anyway, the only disconnection times that matter will be 200ms at IdeltaN and 40Ms at 5IdeltaN. The RCD is always gonna go before the MCB.

The Ze would still be measured/established, and R1/R2 would still be measured. The RCD tests would confirm a path back to earth of less than 1,666 ohms - so even if the Zs reading was taken, it would be completely irrelevant, because the MCB would NEVER trip under a fault to earth condtion. Even a Type B requires 3 to 5 times In before it blows, and the RCD would be gone long before that happened.

of course it will be relevant for commercial/industrial but that's why i mentioned domestic in my question! Even commercial/industrial most 13a sockets (and 16A BS4343) will be on an RCD anyway, so the same will apply

good to dicuss it though!

Oh, and another thought, because the circuits will on RCD's then the Zs measurements would have to be done on the 'low current no trip loop' setting on modern meters, so its not as if we would even be able to 'stress' the circuit by pulling 23A as with the normal Zs tests. In fact the 5IdeltaN RCD test will be the highest fault current applied to the circuits, and then only 150mA!!

 

[Girlyspark]

Another spark I know recently had his annual assessment with the NIC. They found a socket in a house where the Zs was too high and failed him because of this - even though there was an RCD in place as well protecting the circuit.

 

[TonyM58]

Girlyspark,

really intrigued and interested by that one!

been scratching the old grey matter to think WHY the NIC inspector would have failed him/her.

All I can think of is:-

The NIC inspector took a Zs reading that was SIGNIFICANTLY different to that which the sparks entered onto the certs, ( suggesting the reading was falsified or not taken correctly)

OR


The socket was on a VERY long spur (such as taken to an outside building) and the Ib was such that the circuit would fail on V drop - i.e. should have used 4mm instead of 2.5 (Table 7.1 in the OSGis always a good 'quick' design guide!)

OR

It was a TT system, the socket was upstairs (ie not reasonably expected to feed equipment outdoors) and the NIC inspector was invoking the 'should be on min 100mA trip) reg.

OR the sparks had lots of 'iffy' areas on the inspection and any combination of the above 'finalised' the failure!

Just interested, that's all! Be good to find out more, such as did the sparks install one socket as a spur? Was a complete ring/radial installed and only one socket failed?

good to talk!

regards

tony

 

[Girlyspark]

Hi Tony!

The socket was 0.5 over the max in the Regs - not sure how much higher it was than the one he put down on his certificate. The work he had done was change of board at the property only, and in reality he probably didn't measure at every single socket and missed a high reading on this one - probably just a loose terminal!

It wouldn't have been a long spur and it wasn't a TT system.

The inspection at that property was fine otherwise, but the spark failed at the inspection at another property also (no main bond in place - he had written a warning letter saying it was dangerous but shouldn't have changed the board unless customer agreed to having bond put in too!). However, the inspector did say that the revisit would need to go to not only the place with the bond, but also the place where the Zs reading was high at one socket.

Watcha reckon then? Do you think insisting on revisiting the socket with the high Zs was reasonable (it was RCD protected).

And a query for you in light of this discussion - if you find a property where Zs is too high, is it sufficient to ensure RCD protection on the circuit, thus ensuring disconnection time, or do you need to make further attempts to lower the Zs (improvement on Ze by adding earth stake even if not a TT system, upgrading size of conductors etc?).

 

[Minky]

Girlyspark:-
Sounds to me the spark was asking to be failed!.
Regarding your query of high Zs for individual circuits,an attempt can be made to :-

1. Connect a 4mm earth or 2.5 mm earth if mechanically protected to other paths of earth e.g. water pipes,or structural steelwork,if applicable.Before doing so continuity to earth must be checked.
2. Assuming circuit conductors can be increased,i.e pulling new through conduit.
3. Further enquiry of protective device,different values for B curve than C curve,D curve.

I would not attempt to add an earth rod as previously mentioned for obvious reasons,best discussed in another topic forum.

 

[TonyM58]

Girlyspark,

I agree with Minky, an earth spike isnt the answer.

Bottom line is it must be established WHY the Zs is too high

We all know Zs = Ze + (R1 + R2) so, either Ze is too high, and if its a TNS or TNCS and outside 0.8 or 0.35 ohms then i would address this with the DNO. If the Ze is ok, then logically R1+R2 is too high (assuming all connections are good etc). This is quite simply down to cable length being to long, or cable CSA being too small.

But back to my original discussion. The answer to a circuit which has a Zs which fails to meet the disconnection time is supplementary protect with an RCD. Under the 17th virtually all domestic circuits will be on RCCB / RCBO, so what difference will the Zs reading make!!

 

[Girlyspark]

Hee hee - nicely brought round back to the start!

To your original question - I read the other day that RCDs have a 7% failure rate. So I suppose even if RCD protected, if the Zs indicates that disconnection time will still be met if the RCD fails, then it is reassuring.

Belt and braces!

 

[alan]

Problem with RCD's very few operate the test button , so they can be stuck latched on.

 

[Des 56]

Zs still needs to be within specs for 17th
There seems to be a very heavy reliance on RCDs in the 17th
They are not the most flawless of items and need that earth path
Waste of time having one if Frank Spencer changes the light bulb and dead shorts with his fingers
The primary protection is still the main and best protection.
Low enough Zs for the overcurrent device to operate before the killing time is up

 

[rumrunner]

Best get Shake to answer this ,i dont think he would agree with Tonys answer,

im agreeing with the girly spark ,only because i found out a long time ago that its the best policy,

 

[Shakey]

Zs still needs to be within specs for 17th
There seems to be a very heavy reliance on RCDs in the 17th
They are not the most flawless of items and need that earth path
Waste of time having one if Frank Spencer changes the light bulb and dead shorts with his fingers
The primary protection is still the main and best protection.
Low enough Zs for the overcurrent device to operate before the killing time is up

Hi Des, but if Frank shorts with his fingers, he has nothing to do with the Zs path

'Overcurrent device' is a misleading term

An overcurrent is simply a current exceeding the rated value, and can occur anywhere in the circuit (not confined to the CPD)

If we are talking about short circuit current, again Zs is irrelevant

If we are talking about earth fault current, again thats what the RCD is there to protect

I know there is a failure rate of RCD's (not laeast because of the emulsifying of the lubrication due to lack of testing by the user), but lets be honest, how many of you have actually tested a 60898 to see if it ACTUALLY trips within the required time for a given fault current?

And back to my (sorry, 'the') original point, if the circuit doesnt comply, the action to be taken is supplementary protection (additional protection under the 17th) with an RCD

But if the circuit is on an RCD anyway, the Zs measurement becomes irrelevant

And I think 'assuming' devices are going to fail to operate when required to do so is a dangerous road to go down.

Godd to talk though, nice to see some deep and meaningfull discussions on here, now where's Mr Bane when you need him.....

 

[Des 56]

But if the circuit is on an RCD anyway, the Zs measurement becomes irrelevant

The way I see it is
R1 + R2 + Ze = Zs
The R1 may have a very high resistance perhaps through a loose connection or faulty accessory that may cause a fire if the overheating is allowed to continue un noticed
The presence of an rcd will not help if the zs of the circuit is not checked and just assumed to be under the maximum for the rcd
If the standard test is carried out this would be noticed
So sorry to take the opposing view but I still think the test is relevant

 

[PC Electrics]

Here's a clue:

RCD = supplementary protection......

 

[Shakey]

But if the circuit is on an RCD anyway, the Zs measurement becomes irrelevant

The way I see it is
R1 + R2 + Ze = Zs
The R1 may have a very high resistance perhaps through a loose connection or faulty accessory that may cause a fire if the overheating is allowed to continue un noticed
The presence of an rcd will not help if the zs of the circuit is not checked and just assumed to be under the maximum for the rcd
If the standard test is carried out this would be noticed
So sorry to take the opposing view but I still think the test is relevant

Des, I didnt say dont check the circuit

You would still do R1+R2 (and find the loose connection), and still do Zs calculated or measured (EAWR's permitting)

What I am saying is lets say the Zs is 1.6 ohms for a 32A 60898, what are you going to do? You are going to put an RCD on

But the RCD is (probably) gonna be there anyway, so I am saying, what differnece does it make if you measure 1.2 or 1.6 or 2.4 or 3.6 or 4.8 .........all the way up to 1667 ohms

I am not saying dont measure (or calculate)

I am saying that the actual reading will be largely irreleveant

Obviously if you are working on a PME system and come out with 47ohms Zs for your ring, you would investigate the problem (although you should already have done that with you ring continuity checks)

But barring any wiring problems (ie any high reading is down to Ze) the Zs will have little meaning

And lets be honest -you dont even have to measure it anyway!!!!

Its good to talk chaps.....

 

[mijo]

My understanding is that an RCD will not cope with large earth faults currents. For example if the earth loop impedance (ELZ) was say 0.5 ohms you could have an earth fault current of 460 amps.
It is difficult getting manufacturers data on RCDs especially on this point ( max breaking current ) but I have seen a Hager spec for an RCD which quoted max of 200 Amps If you take an inline or plugtop RCD to bits it is fairly obvious that they will not cope with large earth fault currents and many installation type RCBO's have their RCD operation mechanically linked to the MCB the RCD will trip first but arc under heavy earth fault currents but it's mechanical trip actuates the MCB whose contacts can handle the heavy earth fault current without arcing over and break the fault current obviously the mechanical tripping of the RCD and it's arcing under large earth fault currents and the RCD's mechanical tripping of the MCB takes the breaking time outside that permitted for an RCD thus the ELZ for the MCB is required by the regs.
It should also be noted that the MCB deals with short circuit current I know this is Neutral/Live loop impedance which often in domestic installations comes under PFC (prospective fault current) but PFC must be checked both for earth fault current (earth loop impedance) and short circuit fault current (neutral-live loop impedance.

 

[jimmus]

Think the confusion rises when we talk of RCD or RCBO

Having this out with WYLEX at the moment.

17th ed- 411.4.9 Where RCD used to satisfy reg 411.3.2.2 (max disconnection times) max Zs value from table 41.5 can be used. However table 41.3 states "overcurrent characteristics of RCBO's to 61009-1"as its the overcurrent part of the RCBO that is being looked at not the earth fault part.

I am trying to get type B RCBO's to comply with 41.3 but their rep states they dont need to manufacture type B as type C comply with Table 41.5 (1667 ohms Zs) so thats what they make.

So if you have a 32A ring final circuit with Zs of 0.9ohms it is ok on type B (1.44 max) but not on type C (0.72).

I may be wrong and only the "magic 1667" figure is the only one that matters, but even if this is the case table 41.5 only applies to circuits not exceeding 32 amps so if you have for example a 9.8KW shower wired in 10mm on a 40 amp RCBO you still have to comply with 41.3 as 1667 is not applicable.

So Zs becomes relevant again.

 

[Shakey]

Ok chaps

just watched a joint NICEIC/Megger DVD on intitial verification

They state on there that where circuits are RCD protected, the Zs reading need only be taken with a low resolution meter capable of measuring 1667 ohms

I also have in front of me a copy of "Guide to the Wiring Regulations BS7671:2008"
published by no less than the ECA

Page 63 has the following quote under the heading "Irrelevant ELI specification"

"A common source of misunderstanding is that of either specifying or measuring values of ELI where the circuit also has an RCD fitted. ELI measurement under these circumastances is a futile exercise. The circuit will have been checked for continuity, and this is what is needed, together with, of course, functional checks of the RCD...........and for clarification, circuits with RCBO's do not need to meet the specified ELI values"

So back to the original post - will Zs values be irrelevant under the 17th edition? (for RCD protected circuits)


Well I think they will
NICEIC think they will
Megger think they will
The ECA think they will

and for information, the IET often adopt the ECA's reccomendations

*shakey takes a bow*

Hate to say i am right chaps, but, ........erm........point proved i think