This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Diverted Neutral Current

So for the past 5 years if so I have been seeing and experienced a phenomena known as diverted neutral current or diverted networks current.

It has a few names. Much like the incredibly slow realisation that type AC RCDs have had their day. Even though type A has been about over twenty years.

I get at least two / three messages a week from electricians who are now using clamp meters to check the earth conductors of LV installations. Low and behold they find anything from 2A up to 169A in a few cases.

I have research this at work for the last few years and some of the work I have ensured is published via the following link.

We have also managed to produce a visual guide for measuring this issue. This work also helped with the Broken PEN IET article. 

It appears the aged life expired state of the LV network is a bigger problem and will only continue to grow. The use of CNE cable has been a tad o IMO. Years ago the RECs admitted cost was the key factor, but made the decision based upon a REC. a government owned network. Since privatisation the DNOs have fixed on fail, leaving the LV side unmonitored to a large extent. This is now slowly changing. But we assume a broken PEN conductor or failure of this can cause hazardous over voltages but also large amounts of network currents bypassing the broken PEN and importing and exporting on domestic earthing in some cases.

whilst the number are small in the scheme or things. It’s a growing issue. The network know about it. But the industry appears to not be discussing it. Why do we not get a safety alert for all electricians nationally. If we find this why are the DNOs and yes some are taking horrific decisions. And in some cases charging. I’ve issued a safety alert for the Railway where I work. 

be good to gauge folks thoughts as a broken pen gives you CNE consideration in a home, not allowed by BS 7671 and also brings up the whole what is safe isolation with this in mind. We never consider earth as a energised conductor and only ever consider rise of voltage. Maybe with changing and ageing networks, non linear loads etc. We need to change the way we see things.

I live in hope the networks will be more open on this.


  • I'll try and be quicker next time. Wow lots to take in there. I am very aware of SA as its a problem there for sure. not so much the use of RCDs and rods.... intresting thought.

  • Sorry if it reads as a great dump of random thoughts - it is sometimes the way I work. Happy to PM if you like. And here is that lost neutral detector redrawn for the 2020s. Sadly I no longer have a tame contact in ZA so I have no easy way of knowing if the idea has died a death or gone mainstream - but there is no obvious reference to it any more in anything I can find on the web so I am assuming the problem has probably been solved another way. Subtleties like company cut-outs, metering and so on are omitted..


  • Hi Paul,

    This matter is something which concerns me, in respect of connecting coaxial cables to equipment installed within a domestic environment, to Class 1 (earthed chassis) devices such as amateur radio transmitters.

    The grounded coax is then taken outside, and often connected to a grounded antenna, where a PEN fault could give rise to shock to the installer, during installation, to the user whilst connecting the cable to his equipment, or could result in high fault current in the coax, with possible fire risk.

    I have prototyped a system which I am currently using, which introduces two high impedance points within the coaxial cable, ensuring the cable itself is not earthed to 50Hz, but does have a low impedance to ground for RF.

    Not being qualified in Part P, I would appreciate comments on the validity of this approach and whether it would be acceptable to the present Code. If not, what should I further consider?

    A rough outline of my thinking is available at

    Many thanks in anticipation,


  • Welcome G4JBD, de G7VZY.

    The ham radio case is close to my heart too.  I'd not get too hung up on the fire risk to the mains wiring - the electrode resistance of your typical mast planted in cement and an RF earth  will be several tens of ohms at 50Hz unless you live in a salt marsh.  (or Essex it seems - link to amusing past test results)  But overheating a coax braid is possible if it becomes the site neutral for a while as the fault current could be 10-20A

    The safety of life case is the one to consider.

    Personally I run the radio gear as a TT island and would unplug the lot breaking all 3 poles of mains and all antenna feeds when it is not in use. No one who is not licensed should be touching it anyway, and my shack is a 'controlled environment' for which I am the design authority ;-) I'm lucky like that, or just pushy.   And I'll take it all away and fill the rawl-plug holes if I move house so no need to trouble the part P scorers I think.

    However your idea will certainly work, and is similar to the sort of thing the old live chassis TVs used to do on the coax. For HF and certainly LF you may prefer the alternative of double wound RF transformers if you are a QRO person as the caps that can take RF amps can get a bit big.


  • Regarding situations caused by broken, deteriorated, or in the case of South Africa, stolen, PEN conductors ...

    South Africa and Australia have looked at options using smart meter data from single-phase properties. Basically, voltage is monitored by the meters, and the distributor alerted if certain thresholds are met.

    If the distribution main is single-phase only, this method is almost fool-proof, but also on unbalanced mains, three-phase and single-phase meter data can also be used ... less foolproof and gives some false positives.

    I understand Australia have mandated their largest distributors to monitor for MEN faults in this way going forwards.

    A trick we appear to have missed in the UK going forwards, particularly as this fault is due to failure of component parts of the distribution network itself.

  • I guess that is very like the electrode-less O-PEN sensing based on line voltage, but of course better because aggregating the results of more than one meter is likely to avoid false trips from the amateur welder causing a lot of voltage drop, as well as helping to zoom in on the fault by district and automatically reporting it, even when there is no-one in to phone in a problem.

    As you say we seem to have missed a trick, or more likely missed a chance at joined up thinking and spending a little money to save a lot.


  • the subject has been kicking about since 2014

    If you look back in history, it's been going on a lot longer than that.

    The DTI guidance on ESQCR 2002 acknowledges the issue, but it's been discussed much earlier. CP 1013:1965 has discussion on PME, There are some interesting snippets, such as recommendations for consumer's earth electrodes, but most telling. Well worth a read. But issues must have been known at the time the Electricity Act 1947 was drafted, otherwise the provisions in the legislation would not have been put in place.

    On and off over the years, since at least as long as I have held C&G "wiring regs" qualifications, i.e. > 30 years, there have been various debates regarding "exporting PME" outside what we used to call the "equipotential zone". The discussions on EV and PME are a continuation of this particular issue, to be honest.

    "Neutral current diversion" is not new either, although perhaps we are seeing more effects due to the prevalence in plastic pipes, and the age of the network?

    Finally, if we think that TN-S would resolve the issue ... there are such things as protective conductor currents. May only be mA from a single dwelling, but in other premises, and parts of distribution networks, we are looking at many amperes, perhaps 10s of A. And of course, these currents have the same capability for wreaking havoc if the protective conductor breaks (just like the CNE conductor in PME systems - at least with PME the lights go dim rather than relying on a person alone to find out by touching an exposed metal part connected to the installation earthing system).

    I personally would like to see more transparency on such a important safety matter, even if it is a work in progress.

    Agreed, although coming to agreement on the seriousness of the issue is quite difficult - particularly, as I pointed out in the immediate para above, protective conductors are not necessarily "safe to touch", and never were - especially if you disconnect them in a live system.

  • Now Graham, you know I have read CP 1013 and BS 842, plus every edition of G12 and associated docs. 

    The term throwing ones rubbish over the fence comes to mind (DSO).

    Neutral current diversion" is not new either, although perhaps we are seeing more effects due to the prevalence in plastic pipes, and the age of the network?

    Now that I totally agree has possibly exaserbated the issue. I am bemused the ENA have been able to influence the new electrode recommendation as per told by IET a colleague who we both know well. cautions were always supposed to be taken when installing electrodes as to the effects on the network currents, but this seems to have gone out the window. 

    In the old old days, stuff was debate at the institute with peer feedback, this is the only place we now have to attempt to recreate that transparency.

  • HI,

    Not sure if it has already been mentioned but what would be your thoughts on the installation of a 1:1 isolating transformer at the origin, effectively creating a TNS system. I know this would be an expensive option. 


  • it is certainly is an option, and was used many years ago in some industries, strangly they get removed due to initial outlay and maintaince costs.