Judge David Ruth said that while there were complexities involved in Burton’s task of installing a new rangehood and fan
Electrical inspector Simon Holmes found that the earth conductor of the socket outlet installed by
It's not my line of work and I just about install an RCD every time even when it is not required but I thought a new install with a socket outlet in a domestic situation was a requirement?
Correct that sockets added to an existing circuit must have RCD protection.
This seems to have been ignored by the prosecution; instead concentrating on lack of testing.
In particular, EFLI testing - which wouldn't be required for an RCD socket.
EC testing definitely was required either way.
Bur what gets me is that if the A was connected to body of appliance via earth contact of socket; then there can't have been any voyage acroos A & N contacts - so why didn't the owner complain that the rangehood didn't work?
I’m guessing a bit here, it’s difficult to fully understand the details from this article, but to me it sounds like he replaced a fixed wired appliance with a socket outlet. The cable at the range hood was a 2C+e. The red and green are both active (red permanently active, green switched), black is neutral. So unit is not earthed, and original range wasn’t earthed either? The original unit possibly had a permanent supply and a second switched supply, for some reason.
Likewise, I don’t understand why there’s no mention of the failure to install RCD protection, which would’ve most likely been lifesaving.
I don’t think exception 3 in clause 2.6.3.1 applies.
I’m guessing a bit here, it’s difficult to fully understand the details from this article, but to me it sounds like he replaced a fixed wired appliance with a socket outlet. The cable at the range hood was a 2C+e. The red and green are both active (red permanently active, green switched), black is neutral. So unit is not earthed, and original range wasn’t earthed either? The original unit possibly had a permanent supply and a second switched supply, for some reason.
I re-read the article and I've now come to the exact same conclusion.
When I first read the article I felt sorry for the electrician but after further thought it appears that fundamental things were missed.
Even my basic 90's tests should have picked up P-E, P-N and N-E voltage problems.
I do wonder if a significant number of electricians still don't test. re TV show Target
The original article led me to think that there was only one (3-core_ cable st the socket;
so with active on E terminal of socket there couldn't be any voltage on the A & N terminals so the appliance wouldn't run.
This second article indicates there's a separate switch for the appliance; "3-plate" wired; and using the green core of the switch cable as switched active. As used to be common - but hasn't been allowed for a long time now.
Would need to be very old for tape to have been acceptable; under 1976 Regs had to be proper sleeving [R 62].
But also could well have been non-compliant (on that point) when originally installed.
So now thinking that all this wiring was existing, with ceiling rose or similar for the original fan.
Fed by 2C cable; with "3 core" (actually 2C&E) used for the switch only.
Original appliance may well not have needed an earth.
Converting ceiling rose to socket would be the only PEW .
Arguably low risk, as simply replacing a fitting.
If the red rape had fallen off the green core of switch cable; it would look like an E
And there may not have been an E core in the "feed" cable; eg old 2-core lighting circuit.
So:
- 2 reds connected together as A ; originally in "loop" terminal but now (being red) connected to A of socket;
- 2 blacks feed connected together as N; connected to N of socket.
- 1 green, originally taped red but tape lost, originally connected to A of ceiling rise but now connected to E of socket.
Total of three connections to be made; to three terminals.
Possibly he never even realised that there was a controlling switch; as it's not unusual to find a mix of 2-core & T&E wiring at a point.
May have assumed the T&E was the feed, and the 2C going on to elsewhere
No "extra' wires at rise / socket to make him think twice.
Rangehood would operate 24/7; regardless of switch position
Switch = off; happy customer
Switch = on; body of rangehood live - perhaps for the first time since the unit was installed.
If that's (close to) what happened, then
- test of earth continuity of new socket would have failed
- polarity test of socket would have failed.
Which supports the accusation of "failed to test".
And certainly shows why testing needs to be done ALWAYS.
If RCD had been fitted as required; may have saved a life
Though with this level of incompetence, he might well have stuffed up the RCD connections, and probably wouldn't have tested it either.
I'm based in Nelson and Steve is probably one of the best electricians in Nelson. I find it hard to believe that he didn't test and pick up on this.
A $60 plug in socket and see would have picked up the fault immediately and takes 5 seconds to run it's tests.
I was talking to an inspector that spoke with Steve last week, Steve does carry out testing as per section 8 and encloses his test results with his invoices.
Personally I always get asked what I am doing when I use a trailing lead to check protective earthing continuity. I have even had other electricians ask me.
But based on Steve's reputation amongst Nelson based electricians and inspectors, I find it hard to believe that Steve doesn't carry out earth continuity testing or even polarity testing.
The inspector and myself both agreed that if Steve was at fault then why did the EWRB reinstate his licence?
Did they understand different to what the prosecution claims happened?
This may have been caused by a two way strap with the green conductor used to carry current as part of that strap to control a triple plated light fitting or switch plate, elsewhere on the same lighting circuit used to feed the range hood. If the switch or fitting was changed after Steve's work was completed.
If the strap went through the fitting (I have come across it done this way) and an installer had twisted all the PECs together unaware that one of the PECs was part of a strap at the other ends then that could liven all the other green conductors in the installation intermittently depending on the position of the two way switches. These protective earths may not go back to the earth bar so that would mean no RCD would trip, nor would any fuse blow or MCB operate.
I often find such installations where none of the 2C+E wiring is actually connected to earth and people have been adding to it without testing. I also find straps with no sleeving or tape on either the black or green sheathed conductors, often at one end if not both. I'm sure I am not the only one here that does.
I have also come across a standard switch mechanism where the active and neutral of a single 2C+E cable are sleeved as required and terminated in the common and pole one of the switch to operate a triple plated Class I light fitting.
The green PEC wire was in pole two of the switch because the loop terminal was missing and it looks like some bright spark thought, I'll put it in pole 2.
It looked like the loop terminal had lost it's screw and dropped out.
I'd say a DIYer did this changing the switch and also the light?
The green conductor was also connected at the Class I fitting.
This meant when the switch was off the fitting was live.
No RCD at the board and the circuit feeding the lights was two core.
The light wasn't two way switched but imagine if it had more 2C+E cable attached but not back to the earth bar at the board.
You can see there are many possibilities how these things can easily happen.
I don't think the reporting is accurate as to what was actually found and what actually happened.
Steve may have only done something as simple as change an old socket on a lighting circuit, behind a Class II range hood that was then later changed to a Class I range hood?
The issue was most likely always there and the earth always live when a certain two way strap elsewhere on the light circuit had it's switches in a certain position.
But it should have still shown him a fault at:
continuity testing
polarity testing
EFLI testing (that would have been required for changing an existing non RCD protected socket)
But we don't know if something else was changed by someone else after Steve finished his work.
As for the Judge claiming Steve incapable of understanding the technicalities of the incident whilst he, the judge could, I think that is pretty arrogant of the judge. I'd be surprised if the judge had any idea of the number of things that could cause such a fault.
How do we as electricians protect ourselves against others altering our work after completion or other oddball stuff elsewhere on the circuit away from what we have worked on?
Light circuits are difficult to test in older installations with all manner of differing loads and switching arrangements.
Do we have to work with GoPros permanently strapped to our heads?
Should we now start refusing to add to or maintain installations were the green earth conductor is sleeved, taped or left au-naturel?
Should we start replacing these 2C+E cables used as 3C straps?
Agree that 3rd-party interference is always a possibility.
But in absence of any supporting evidence; it certainly can't be presumed.
Also agree that the reporting of this case doesn't provide enough technical detail to be certain of exactly what was done.
We can only speculate.
But speculation should be based on the known facts, and likely scenarios; rather than remote possibilities.
It may well have been a lighting circuit without PEC. But I've seem absolutely nothing to suggest that there was any 2-way switching involved.
Yes with legacy installations there are often aspects that don't comply with today's requirements, but are allowed to continue in service under ESR 113 (as long as they still comply and have not become unsafe).
And unfortunately some that never complied - which are not covered by ESR 113; and are required to be remedied if unsafe.
The ways of EWRB can be strange, and they too often get things wrong.
But rather than wonder why they re-instated the PL; perhaps better to ask why they revoked it in the first place.
Obviously they believed he did something wrong.
Most parties found guilty by EWRB never lose their PL; but even for those that do it's seldom forever.
Re-instatement is usual; often after remedial training and / or after some time under supervision.
So it certainly isn't evidence that this HSW conviction is wrong.
I don't think any of that means we need go-pros or similar methods of trying to record exactly what we did.
But we do have to be aware that what we think we're dealing with may not be what we're actually dealing with.
And we do have to test - something far too many either don't do at all, or only partially.
While we are not required to record results ; doing so may help when challenged.
That said; a tick in a box (as on many CoC forms) proves absolutely nothing, and can't even be considered reliable evidence
Having numbers written down may be evidence, but it is not absolute proof of having tested.
All it can prove is that you know what sort of result is acceptable.
Luckily absolute proof isn't required; just beyond reasonable doubt.
We also have to recognise that Worksafe operate under a different legal regime from the rest of NZ law.
Basically HSW law holds that if an incident has happened, then someone must have been at fault.
So you're pretty much guilty until proven innocent, rather than the other way around as for most law.
Plus pleading guilty generally leads to lower penalty; while any defence generally leads to greater penalty as well as greater cost.
Plus you can't insure against penalty; so your insurance company doesn't have a financial incentive to get you off;
so may not want to fund the defence unless your legal team is very confident of victory.
For electrical issues; neither the judge, nor any of the lawyers, are likely to have any technical understanding.
Nor even a good grounding on ESRs / Standards.
Overall, when charged by Worksafe under HSW; you're on a hiding to nothing.
WRT whether an RCD would have mitigated the shock;
if properly connected it would, regardless of whether there was any PEC for the circuit / fitting.
An RCD does NOT need a PEC in order to operate.
It operates solely on imbalance of current between A & N (for single phase) - on the assumption that the 'residual' current may be following a path through a person.
The difference having a PEC makes is not about whether or not the RCD operates; it's about when the RCD operates.
Wiring Rules for (earth) fault protection require the protective device to operate on occurrence of the fault; so that exposed conductive parts do not remain live for longer than the permitted time (0.4 s for sockets). So no-one gets a shock at all, unless they happen to be touching the otem at the time of the fault.
Not having a PEC means the exposed conductive part(s) remain live until something else - eg a person - creates a path for the fault current to flow through. They then suffer a shock, but only for the short time until the RCD operates.
Given the usually-presumed impedance hand - foot of 1000 ohms, and assuming full voltage; not just 30 mA but more like 230 mA will flow.
That's more than 7 x the rated residual current; so the operating time for a Type A RCD will be < 40 ms.
My point was Alek that if there is no continuity of the earthing conductors with the installation Earth then the RCD would not trip until there was. People seem to rely on RCDs instead of actually testing earth continuity with the circuit neutral disconnected.
I pointed out to another electrician today that EFLI tests do not confirm an earth is actually connected too the earth bar.
Furthermore 30mA hurts like hell if you are fit and healthy.
If you have an underlying health condition or your knee against an earthed appliance like a free standing range for example, fault current path plays a big role,
30mA could kill you easily.
I think it was Big Clive who did a demonstration and gave himself a shock through an RCD, watching it he clearly, massively underestimated it.
It appears, from what I was told by a Nelson Inspector, that allegedly there was a fan and a light installed with a switch for each.
One 2c+E cable from the switch plate,.
Common neutral for the fan and light.
Fan on red conductor.
Light on green conductor with red tape applied.
Cable was cut and fan and light removed, possibly not disconnected first, maybe a junction box installed afterwards?
Switches obviously not looked at.
2nd visit socket was fitted to cable and obviously no earth continuity testing was done at the socket.
Circuit was relivened, range hood plugged in and was working.
Then 2nd switch was operated at some point and frame of range hood energised.
Still shocked and surprised that an electrician with such a good reputation was capable of such work.
I hadn't considered that possible configuration; but it makes sense, and fits the reported facts.
And while red tape was not compliant when installed, it wasn't unusual.
Agree testing is often skimped; with some "mandatory" tests either done using inappropriate methods or simply not done at all.
Also that ECC testing must be done with N disconnected from E
- for a new curcuit; connect and test PEC before connecting N or As;
- for existing circuit; best option is turn of main switch and open MEN link.
I'd like to believe that doing no testing at all happens less often these days.
But yes this does seem to be a case of zero testing.
An RCD properly installed, in conjunction with earthing (tested!); would have tripped as soon as the switch livened the appliance body.
without earthing, it would have tripped when someone provided a fault current path, even a very poor one, by touching the unit.
And yes, fault current path does play a big role.
With a poor path, current below 15 mA, the RCD should not trip.
The person would suffer a shock, but unlikely to be fatal at that level.
Or rather, and pedantically; unlikely to enter ventricular fibrillation .
With better contact; generally around 20 - 25 mA, the RCD would trip, and the person would duffer a slightly worse shock, for no longer that 400 ma
yes that would hurt, but again unlikely to be fatal.
With a good contact, as I said, the fault current would be in order of 230 mA (based on dry-skin body impedance approx 1000 ohms);
ie much more than 5 x rated residual current (150 mA); and the RCD would trip within 40 ms.
Not going to speculate which shock would feel worse subjectively; though there are Tables [AS/NZS 60479 series]that provide objective data.
Which is why RCDs for additional (fault) protection are required to have RRC no more than 30 mA; and those for junior education no more than 10 mA (relating to let-go threshold, rather than fibrillation); and those for medical (shock current path assumed to be lacking impedance of skin) are not just 10 mA max but faster-acting (Type I) as well.
Contrary to widespread misunderstanding; an RCD doesn't limit the shock current,
which depends solely on the voltage and the impedance of the shock current path.
An RCD only limits the duration of the shock.
So yes, a shock via a 30 mA RCD "hurts like hell"; because typically the shock current is far nearer 200 mA than 30 mA.
That's not the same as a shock at 30 mA; which is very unlikely to kill you by fibrillation.
WRT EFLI testing, a "live" EFLI test does confirm that the earth terminal is connected to either E-bar or N-bar.
It can't tell us which, and it can't tell us what colour wire is making the connection;
So by itself can't confirm polarity.
But it does prove that the item is effectively earthed.
A good example of the 'swiss cheese" theory of how things go wrong
First an error of assumption about what was there, followed by failing to test to confirm / reject that assumption;
followed by non-compliant work; followed by failure to fully test.
I to took all 4 of these lined up for the fatality to occur.
Any one of them had been different, the incident likely would have been averted- or at least non-fatal.
We are all confident that we do good work and haven't stuffed anything up
- unless / until we actually test, when we find that we're actually not perfect and do sometimes make mistakes.
I certainly do a lot more testing now than I did as a young sparky
The young know everything; the old (sometimes) know better.
These users thanked the author AlecK for the post:
A good example of the 'swiss cheese" theory of how things go wrong
First an error of assumption about what was there, followed by failing to test to confirm / reject that assumption;
followed by non-compliant work; followed by failure to fully test.
I to took all 4 of these lined up for the fatality to occur.
Any one of them had been different, the incident likely would have been averted- or at least non-fatal.
We are all confident that we do good work and haven't stuffed anything up
- unless / until we actually test, when we find that we're actually not perfect and do sometimes make mistakes.
I certainly do a lot more testing now than I did as a young sparky
The young know everything; the old (sometimes) know better.
Well said
My favourite TV program is Mayday/Air Crash investigators
Planes just don't fall out of the sky many things need to go wrong just like in this case.
Very unlucky
Part of me thinks this poor guy is being made an example for the rest of us to sharpen our processes
Part of me thinks this poor guy is being made an example for the rest of us to sharpen our processes
Disagree, this "poor guy" killed someone due to his incompetence and by not doing absolute basics. It shouldn't be a case of upping our game as the bar should not be so damn low. I can't believe a life was lost due to lack of earth continuity testing. Something that has no excuses for being skipped.
