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 34 replies
  • Subscribers 32 subscribers
  • Views 4365 views
  • Users 0 members are here
Options
You may also be interested in
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

BS7671;2018 Minor works certificates.

Sparkingchip
Are you all organised and have a supply of the new certificates to hand?


What do you actually think of them?


Andy B.
  • 0

    As for the test that produces a 220ms for a final circuit on a TT system requiring a 0.2s disconnection time that is for once clear in 643.7.1 as it says, " The effectiveness of the automatic disconnection of the supply by RCDs shall be verified using a suitable test equipment according to BS EN 61557-6 to confirm that the relevant requirements of Chapter 41 are met, taking account of the operating characteristics of the device". So a 220ms time would not meet the requirements of Chapter 41 although it would meet the product standard. The same would apply to RCDs to the older BS 4293 standard.



    220ms at 1x doesn't prove that that requirement has been met, but neither does it prove that it hasn't been either.


    You could re-do the test at say 2x or higher (say using the 1x setting on the next range up, which would normally give you around 3x) which hopefully gives you a result within 200ms. Or you could do a 5x test (using the supply N rather than PE if Zs is a problem - as that verifies the effectiveness of the RCD itself just the same) and perhaps get a result below 150ms.


    I think we have a problem though in that recording the time without recording what the test involved makes it all rather meaningless to anyone subsequently reading the schedule.


    Or perhaps we take the attitude that if the RCD is within spec. for the tests we perform then it can be deemed to be operating correctly, so it's reasonable to presume it'll also operate within specification for other residual currents. After all we don't attempt to test any other component of the installation to anything like this degree.


       - Andy.
  • 0
    Graham


    You posted a very good table above. Could this not be fettled and honed and go in to Part 6 as it would take out all the ambiguity? You might want to submit it in the next DPC?


    As for the test that produces a 220ms for a final circuit on a TT system requiring a 0.2s disconnection time that is for once clear in 643.7.1 as it says, " The effectiveness of the automatic disconnection of the supply by RCDs shall be verified using a suitable test equipment according to BS EN 61557-6 to confirm that the relevant requirements of Chapter 41 are met, taking account of the operating characteristics of the device". So a 220ms time would not meet the requirements of Chapter 41 although it would meet the product standard. The same would apply to RCDs to the older BS 4293 standard.
  • 0

    Playing Devil's advocate for a moment... on a TT system is there a requirement to test a 30mA unit at 1x at all? While I can of course see the common sense in testing at 30mA, 643.8 seems only to actually require:



    • That ADS is proved - typically that means showing a ≤0.2s disconnection time - hence a residual current exceeding 2x - so a 5x test using an ordinary instrument.

    • and for additional protection, only the 5x test for 40ms is specified (and even that's only in a NOTE).



    Well, I don't disagree with that, although the question that I think we will still be asked by JP, is whether we are verifying the device (in which case all tests in accordance with BS EN 61557-6 would be necessary) or the correct application of the electrical installation (which is at least as much about loop impedance and product selection as whether the device itself works - after all, we don't test mcb's do we ...)


    Also is there any value in using a lower multiple on a higher setting on the test instrument - e.g. using a 300mA x 1 as a better approximation to a 2x test for a 100mA device than a 5x 100mA test (and reduce the reliance on a low Zs in the process). Or is that instrument abuse?



    Perhaps, although you've got to wonder about the tests required in Chapter 64 and BS EN 61557-6. For example, I have a 300 mA RCD in a TT system protecting a final circuit containing fixed equipment < 32 A, for which Additional Protection is not required. It has a disconnection time of 220 ms at x1 test (which is OK, but doesn't meet the 0.2 s disconnection time):

    • Do I replace the RCD just because I can't verify the 0.2 s disconnection time?

      OR

    • Do I let it go, write 220 in the test cert (but then have no evidence of verification of the required disconnection time as per Chapter 64 and Table 41.1)?


    On a different tangent, should we be verifying the RCD's reaction to d.c. components in the residual current too these days?



    Yes, where necessary (noting Type AC still acceptable for general use), but you set that on the instrument in any case.



  • 0
    30 mA

    Conduct both 1x and 5x tests. Requires loop impedance less than 1364 Ω if you take into account full volt drop

    Record the 5x result.
    Verifies both additional protection to 643.7 and ADS to 643.7.1 b) for 0.2 s and 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4 and 411.3.2.1.

    Whilst the full Zs of 1667 Ω presented in Table 41.5 wouldn't provide sufficient current for a 5x test with lower voltages, earth electrode resistance should be limited to 200 Ω so no issues anticipated











    Playing Devil's advocate for a moment... on a TT system is there a requirement to test a 30mA unit at 1x at all? While I can of course see the common sense in testing at 30mA, 643.8 seems only to actually require:
    • That ADS is proved - typically that means showing a ≤0.2s disconnection time - hence a residual current exceeding 2x - so a 5x test using an ordinary instrument.

    • and for additional protection, only the 5x test for 40ms is specified (and even that's only in a NOTE).



    Also is there any value in using a lower multiple on a higher setting on the test instrument - e.g. using a 300mA x 1 as a better approximation to a 2x test for a 100mA device than a 5x 100mA test (and reduce the reliance on a low Zs in the process). Or is that instrument abuse?


    On a different tangent, should we be verifying the RCD's reaction to d.c. components in the residual current too these days?


        - Andy.
  • 0

    I think that I have been slightly mis-read here.


    I was referring to the five tests at 1/2 Ideltan, Ideltan (x 2), and 5 Ideltan (x2). Performing them and writing them in a notebook is no bother at all.





    Don't disagree with you. I think from the Table in my previous post, you will see where my thinking is coming from re 10 and 30 mA RCDs.


    And, more importantly, where JP is going with higher rated RCDs. The rub being, that a designer could select an RCD with rating 500 mA Type S to meet 411.3.2.1, and, providing the loop impedance were low enough to achieve the stated disconnection time (in this case, 200 Ohm would be good enough for the 200 ms disconnection time at 2x, or 1 A) , all would be well ...

    That is, until we come to our Initial Verification (or with existing installations Periodic Verification) in accordance with BS 7671:2018 - if we take 643.7.1 b) (2) at its verbatim word !
  • 0

    Graham Kenyon:




    Chris Pearson:




    Graham Kenyon:

    John,


    In general, I don't really disagree with you, but ...




    If an RCD does not trip at 1 x Idelta n then it is defective. This softer test is a better indicator of RCD health than the 5 x test and consumer safety.






    This does not unfortunately take account of a small number of cases where the RCD passes a 1x test, but then goes on to fail a 5x test, although I agree this is not simply a case of "sticky RCD".




    For my own peace of mind, at home, I would wish to be assured that an RCD does trip as intended. If passing at 5x means that one will also pass at 1x, then all well and good: only the one test is required. However, if not, then surely both tests are required.


    Of course, there is no reason to stop doing the 5 tests, it is just that only one time needs to be recorded in the EIC/EICR (plus the all important test button).



    I'm still not satisfied.



    I think that I have been slightly mis-read here.


    I was referring to the five tests at 1/2 Ideltan, Ideltan (x 2), and 5 Ideltan (x2). Performing them and writing them in a notebook is no bother at all.

  • 0
    JP,


    A summary of my thinking where RCDs are used to provide ADS in TT systems (or in the case of 10 mA and 30 mA RCD, Additional Protection and ADS):

     
    RCD Residual current Rating

    Tests, results and notes

    10 mA

    Conduct both 1x and 5x tests

    Record the 5x result.
    Verifies both additional protection to 643.7 and ADS to 643.7.1 b) for 0.2 s and 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4 and 411.3.2.1.

    30 mA

    Conduct both 1x and 5x tests. Requires loop impedance less than 1364 Ω if you take into account full volt drop

    Record the 5x result.
    Verifies both additional protection to 643.7 and ADS to 643.7.1 b) for 0.2 s and 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4 and 411.3.2.1.

    Whilst the full Zs of 1667 Ω presented in Table 41.5 wouldn't provide sufficient current for a 5x test with lower voltages, earth electrode resistance should be limited to 200 Ω so no issues anticipated

    100 mA

    Conduct both 1x and 5x tests. requires loop impedance less than 409 Ω if you take into account full volt drop

    Record the 5x result.
    Verifies ADS to 643.7.1 b) for 0.2 s and 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4 and 411.3.2.1.

    Whilst the full Zs of 500 Ω presented in Table 41.5 wouldn't provide sufficient current for a 5x test with lower voltages, earth electrode resistance should be limited to 200 Ω so no issues anticipated


    If you are asserting it's unsafe to test the RCD at 5x in domestic, then do and record the 1x, but now you only have verified disconnection at 1 s and compliance with 411.3.2.4, and therefore in domestic this RCD can only provide ADS for distribution circuits, and final circuits not covered by 411.3.2.2.

    300 mA

    Not practicable (or perhaps safe) to test at 5x, so we are limited here to testing at 1x.

    Record the 1x result.
    Verifies ADS to 643.7.1 b) for 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4, but not 411.3.2.1.

    This leads me to assert that the 300 mA RCD may only provide ADS for distribution circuits, and final circuits not covered by 411.3.2.2

    500 mA

    Not practicable (or perhaps safe) to test at 5x, so we are limited here to testing at 1x.

    Record the 1x result
    Verifies ADS to 643.7.1 b) for 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4, but not 411.3.2.1.

    This leads me to assert that the 500 mA RCD may only provide ADS for distribution circuits, and final circuits not covered by 411.3.2.2

    100 mA Type S

    Conduct both 1x and 5x tests. requires loop impedance less than 409 Ω if you take into account full volt drop

    Record the 5x result.
    Verifies ADS to 643.7.1 b) for 0.2 s and 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4 and 411.3.2.1.


    If you are asserting it's unsafe to test the RCD at 5x in domestic, then do and record the 1x, but now you only have verified disconnection at 1 s and compliance with 411.3.2.4, and therefore in domestic this RCD can only provide ADS for distribution circuits, and final circuits not covered by 411.3.2.2

    300 mA Type S

    Not practicable (or perhaps safe) to test at 5x, so we are limited here to testing at 1x.

    Record the 1x result.
    Verifies ADS to 643.7.1 b) for 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4, but not 411.3.2.1.

    This leads me to assert that the 300 mA Type S RCD may only provide ADS for distribution circuits, and final circuits not covered by 411.3.2.2

    500 mA Type S

    Not practicable (or perhaps safe) to test at 5x, so we are limited here to testing at 1x.

    Record the 1x result.
    Verifies ADS to 643.7.1 b) for 1 s disconnection.

    I would be able to demonstrate compliance with Regulation 411.3.2.4, but not 411.3.2.1.

    This leads me to assert that the 500 mA Type S RCD may only provide ADS for distribution circuits, and final circuits not covered by 411.3.2.2



     

  • 0
    Graham


    Yes 30mA RCDs need testing at 5 x I delta n to verify they will trip at not more than 40ms but we need to verify at a lesser current than 150mA coursing through someones body they will also trip in a safe time. Reference to the IEC shock current curves would seem to suggest this? As for testing at more than 5 x I delta n at "higher than 5 times its rated residual operating current", as suggested in the Note to 643.8, I will ask again what does that prove? How about a test current of 5A that should operate any sticky 30mA RCD.


    For fire and fault protection 643.7.1 directs us to use an RCD tester to BS EN 61557-6 which may not be able to deliver test currents at 5 times higher rated RCDs. 643.7.1 directs us to Table 3A of Appendix 3. 


    So we need to know what the rating of the RCD is and the purpose for which it is installed?


    So Graham head above the trench time. For additional protection what current(s) would you apply to the RCD and what maximum trip time would you expect? The same question for a 300mA RCD, time and not time delayed, provided for fault or fire protection?
  • 0

    Chris Pearson:




    Graham Kenyon:

    John,


    In general, I don't really disagree with you, but ...




    If an RCD does not trip at 1 x Idelta n then it is defective. This softer test is a better indicator of RCD health than the 5 x test and consumer safety.






    This does not unfortunately take account of a small number of cases where the RCD passes a 1x test, but then goes on to fail a 5x test, although I agree this is not simply a case of "sticky RCD".




    For my own peace of mind, at home, I would wish to be assured that an RCD does trip as intended. If passing at 5x means that one will also pass at 1x, then all well and good: only the one test is required. However, if not, then surely both tests are required.


    Of course, there is no reason to stop doing the 5 tests, it is just that only one time needs to be recorded in the EIC/EICR (plus the all important test button).


     




    I'm still not satisfied. There are also the cases where the 1x time passes (i.e. < 300 ms) but does not provide evidence of compliance for 643.7.1(b) - in particular the disconnection time recorded > 200 ms in cases when Chapter 41 demands 0.2 s disconnection time.

    And the even greater number of cases where 643.8 is not verified by the 1x test - in particular disconnection time recorded > 40 ms (per the Note to the regulations)?


    Surely, this is the purpose of the verification stated in the Regulations, and if the evidence is not there, from the test as specified, then the particular requirements of Chapter 41 for ADS are not met - similar to a high loop impedance reading where OCPDs are relied upon to provide ADS?


    Presented with a set of test results that record the fact that the requirements have not been met, and certified by a competent person by their signature, what else could a court decide?


    Sure, if the 1x test is below 40 ms (as might be the case) then no-one would know the difference. But this doesn't help anyone decide what is really necessary.

    For a 30 mA RCD, there is rarely a reason to not carry out the 5x test, and record the result. I agree with larger residual current ratings, the 5x test is not desirable, but the way the 18th Edition requirements are written does not really afford us a means of providing evidence compliance (unless of course you're lucky enough to be able to achieve the required disconnection time on the 1x test).


    This is why I don't subscribe to the "only record 1x" ...

     

  • 0

    Graham Kenyon:

    John,


    In general, I don't really disagree with you, but ...




    If an RCD does not trip at 1 x Idelta n then it is defective. This softer test is a better indicator of RCD health than the 5 x test and consumer safety.






    This does not unfortunately take account of a small number of cases where the RCD passes a 1x test, but then goes on to fail a 5x test, although I agree this is not simply a case of "sticky RCD".




    For my own peace of mind, at home, I would wish to be assured that an RCD does trip as intended. If passing at 5x means that one will also pass at 1x, then all well and good: only the one test is required. However, if not, then surely both tests are required.


    Of course, there is no reason to stop doing the 5 tests, it is just that only one time needs to be recorded in the EIC/EICR (plus the all important test button).

Community Rules & Guidelines