Rules and Guidelines
Want to participate in the discussions? Please read our Community Rules and Guidelines. Need some help? Visit Support and Answers.

  • State Not Answered
  • Locked Locked
  • Replies 66 replies
  • Subscribers 32 subscribers
  • Views 27850 views
  • Users 0 members are here
Options
You may also be interested in
This discussion is locked.
You cannot post a reply to this discussion. If you have a question start a new discussion

Minimum values of insulation resistance

Wombat
What is the science behind the 1 MOhm minimum insulation resistance? What is the basis for this particular value?
  • 0

    gkenyon:




    AJJewsbury:




    Therefore each circuit can pass an insulation test, but the whole board can fail, so what do you do next?



    But there is no BS 7671 value for a single circuit to pass/fail - just one for the complete DB's worth - 1MΩ*


    If the whole DB fails, then it's a fail.


       - Andy.


    * for initial verification at least.

     




    Regulation 643.2 says nothing about DBs


    The requirement is:


    643.3.2 The insulation resistance measured with the test voltages indicated in Table 64 shall be considered satisfactory if the main switchboard and each distribution circuit tested separately, with all its final circuits connected but with current-using equipment disconnected, has an insulation resistance not less than the appropriate value given in Table 64.



    It's not exactly the plainest of plain English! However if the main switchboard is the one and only CU, then I agree that the whole board IR test must comply with Table 64. Whether you stop there and grade unsatisfactory/FI or go on to test each circuit is a matter of judgement (and contract). If you do test each circuit, you will (presumably) find one whose IR value is low, in which case I would certainly pause having graded unsatisfactory/FI.


    Now 134.2.1:

    During erection and on completion of an installation ... appropriate inspection and testing shall be carried out ...


    Bearing in mind the above discussion, would anybody be happy with only 1 MΩ for a new circuit? I wouldn't.


    (IR testing of my brand new isolator at home gave only 80 MΩ. I spoke to Eaton about it and they said: (1) it is well above the minimum requirement; and (2) things like releasing agent or a trace of damp may lower the value.)

     

  • 0
    I cannot answer that, but I can pour water on troubled oil by adding another observation.

    The 1 meg limit is for a DC test and as such ignores any capacitance that in many cases is the larger cause of earth leakage currents anyway.

    Normal twin and earth manages 50-100pF or so per metre from L and N to E, so an installation of perhaps a few tens of metres, will have a significant leakage at 50Hz, and nothing at DC - as we sometimes see with 'ghost voltages' on disconnected cores running beside live ones that are enough to cause LEDS to glow dimply or CFLs to flicker.


    It is all  very well keeping the DC test result to  a level that means you might expect leakage from 230V of below 250uA, but when 50Hz is applied it may well rise to  a few mA, being a megohm in parallel with so many nanofarads of incidental capacitance.
  • 0

    I don't agree. What about installations with multiple levels of distribution?


    For example, even the simple case of Main three-phase switchboard feeds a number of three-phase boards, which in turn feed some equipment plus some single-phase boards. To comply with 643.2, the minimum requirement is for each of the following to be 1 MΩ (unless of course the entire switchboard with everything connected complies):

     - The main switchboard including any final circuits it supplied

     - Each distribution circuit of the main switchboard (individually) through the DB it supplies, having all final circuits connected, along with the sub-mains this DB supplies to the single-phase DBs, along with the supplied single-phase DBs with all final circuits connected.


    So, whilst all DBs will have to meet the requirement, in the case of the three-phase DBs in the tier above, their tests must include not only their submains and final circuits, but also the submains (and associated DBs and all of their final circuits) too.



    Interesting - I'd read 'the main switchboard and each distribution circuit tested separately' to mean that each distribution circuit should be tested individually - whether it originates from the main switchboard or from a sub-board (the wording doesn't specify distribution circuits hailing from the main switchboard). So in each case, the test is effectively the DB together with the circuit that supplies it (whether a distribution circuit or connection to the origin) and its final circuits - but not distribution circuits it feeds as they'll be tested along with the downstream DB. That way the approach is consistent and you're testing similar-sized units regardless of how many tiers of distribution the installation has.

     

    It's not exactly the plainest of plain English!



    you're not wrong there!


       - Andy.

  • 0
    This certainly is a very interesting discussion going on here.

  • 0
    Just in case it adds anything, I've been looking back over the old regs. As far as I can tell the requirement for insulation testing started with the 3rd Ed in 1897 - then the limit was 10MΩ divided by the maximum demand of the installation in amps.


    In the 4th Ed of 1903 it changed to 30MΩ divided by the number of 'points'.


    By the 8th Ed of 1924 it became 25MΩ divided by the number of points, but with the added proviso that 1MΩ for the entire installation was acceptable even if the formula produced a higher number.


    In the 10th Ed of 1934 the demand was for 100MΩ divided by the the number of points for the fixed wiring alone (before light fittings etc were connected) and 50MΩ divided by the number of points for the completed installation (without appliances), but again that entire installations need not exceed 1MΩ.


    The 12th just added the proviso that readings below 0.5MΩ weren't acceptable because instruments of the time wouldn't be able to provide 500V into resistances that low.


    By the 13th Ed (1955) it was 50MΩ divided by the number of points generally, but 12.5MΩ divided by the number of points if PVC-insulated cables used. Again 1MΩ for entire installation was acceptable.


    The 14th (1966) made the requirement a flat 1MΩ but large installations could be sub-divided into groups of not less than 50 outlets.


    The 16th introduced wording similar to what we have today (main switchboard and each distribution circuit tested separately) but the limit was reduced to 0.5MΩ


    And the 17th introduced what we have today - main switchboard and each distribution circuit tested separately - and the limit restored to 1MΩ


      - Andy.
  • 0

    AJJewsbury:

    Just in case it adds anything, I've been looking back over the old regs. As far as I can tell the requirement for insulation testing started with the 3rd Ed in 1897 - then the limit was 10MΩ divided by the maximum demand of the installation in amps.


    In the 4th Ed of 1903 it changed to 30MΩ divided by the number of 'points'.


    By the 8th Ed of 1924 it became 25MΩ divided by the number of points, but with the added proviso that 1MΩ for the entire installation was acceptable even if the formula produced a higher number.


    In the 10th Ed of 1934 the demand was for 100MΩ divided by the the number of points for the fixed wiring alone (before light fittings etc were connected) and 50MΩ divided by the number of points for the completed installation (without appliances), but again that entire installations need not exceed 1MΩ.


    The 12th just added the proviso that readings below 0.5MΩ weren't acceptable because instruments of the time wouldn't be able to provide 500V into resistances that low.


    By the 13th Ed (1955) it was 50MΩ divided by the number of points generally, but 12.5MΩ divided by the number of points if PVC-insulated cables used. Again 1MΩ for entire installation was acceptable.


    The 14th (1966) made the requirement a flat 1MΩ but large installations could be sub-divided into groups of not less than 50 outlets.


    The 16th introduced wording similar to what we have today (main switchboard and each distribution circuit tested separately) but the limit was reduced to 0.5MΩ


    And the 17th introduced what we have today - main switchboard and each distribution circuit tested separately - and the limit restored to 1MΩ


      - Andy.




    Thanks Andy for that interesting history.


    Whoever was writing the 16th edition must have had some very good reason that he could justify to reduce the minimum to 0.5MOhm .


    Similarly, whoever was writing the 17th edition must have had some very good reason that he could justify to increase the minimum requirement to 1MOhm.


    This 1MOhm does not seem to fit well with the threshold of reaction of IEC/TR 60479-1 or threshold of perception of BS EN 61140 values of 0.5mA ac and 2mA dc for the limitation of steady-state touch current.


    So why this very elusive value of 1MOhm for minimum insulation resistance?


     

  • 0
    A bit of an aside but I was given an ancient mains filter I tested it and got 2.5 nF between live an neutral and half that value from each side to earth also oddly just over 2Megohms neutral to earth and to high to read on live to earth anyhow befor plugging it into the mains I calculated worst case leakage and at 240 volts it just over a milliamp now I know it won't be that much because the neutral has no voltage on it but it seems that the 1 Megohm limit for installation resistance is sensible because not even the most sensitive RCD will be troubled by that and no watt hour meter will register it so really a win win I think
  • 0

    AJJewsbury:

    Just in case it adds anything, I've been looking back over the old regs. As far as I can tell the requirement for insulation testing started with the 3rd Ed in 1897 - then the limit was 10MΩ divided by the maximum demand of the installation in amps.




    Andy, thank you for the interesting info.


    They must have been doing well to measure IR in 1897 - the first Megger came a few years later.


    I suspect that 1 MΩ was picked because it wasn't practically possible to test much higher and that value gave an adequate margin for error. After 100-odd years, why change?

  • 0
    I remember reading somewhere that the fire brigade were responsible for demanding that a suitable value of 1.0MCopyright  insulation resistance exsisted between live-earth and live-live after an installation had been completed and then retested 2 weeks later to verify that it hadn't dropped off.

    Legh
  • 0
    Round numbers like 1 million have the feel of some one trying to make a good stab at an order of magnitude.


    If the limit was 21.679 there may be some sums behind it


    The values for bleed resistors are based on a sum, and it is that we get must get down from peak voltage to below 50V (a safe voltage on the pins of a plug perhaps) within 5 seconds of disconnection - so if you pull the plug out from kit with a filter inside, you do not get a bite off the pins half an hour later when you want to plug it in again.

    Longer bleed times are permitted on things that are enclosed and hard wired.

    Even then we calculate and unless it is going to run really hot, fit half that value for a bit of margin.

Join us to get the best from IET EngX.

Joining EngX lets you personalise your experience so you stay up to date on the topics that interest you, plus you’ll be able to make connections who are looking to collaborate, exchange ideas and more.

Join us Not now

Community Rules & Guidelines