Page 1 of 2 12 LastLast
Results 1 to 10 of 12

Thread: earth leakage unit failure

  1. #1
    Suspended
    Join Date
    Oct 2007
    Posts
    2,346
    Thanks
    0
    Thanked 254 Times in 209 Posts

    earth leakage unit failure

    been having some challenging issues with earth leakage units lately.

    i have 2 problems at present which are still not sorted out.

    maybe i am missing something...

    1 st one...

    get a call out because an earth leakage unit has tripped...

    step 1...isolate the entire installation and remove all the sockets and switch off the isolators...
    step 2...disconnect all the neutrals
    step 3...do an insulation resitance test on all the circuits...found 2 circuits with low readings...identified the circuits and checked the installation...found an unused cable which was grounded...disconnected and isolated from the installation...2nd circuit was a cable in the garden which had been hit by a spade and the red wire was exposed but only grounding when it got wet from rain...retest...everything clear...reconnected and switch on...everything working.
    2 days later called back earth leakage tripped again...disconnected and removed the earth leakage and swopped with the outbuilding...house has not tripped since...however the out building has started tripping now.

    so i asume it is the earth leakage unit is faulty and replace the unit with a brand new one...2 days later i am called back...earth leakage tripping again ...so i do the same thing to the outbuilding...disconnect and do an insulation resistance test...find a plug in the garden full of ants and moisture...disconnect and isolate...retested...all clear except the light circuit...a low reading of just below 1 meg...normal on fluorescent fittings with electronic ballasts...so i leave...2 days later earth leakage unit tripping again...and so it has been for the past week every 2-4 days it randomly trips...reset and works for another couple days.

    the 2nd...

    do all the test like above...find no low reading...plug everything back in...4 days later it trips again...go to site and reset doesnt trip for a couple days again...

    so i do some other tests
    mA readings on the earth wires
    circuit
    1. 0.4 mA
    2. 0.0 mA
    3. 2.6 mA
    4. 7.8 mA
    6. 0.2 mA
    7. 24.3 mA
    main incoming earth 123 mA (this i dont understand)

    the earth leakage unit is warm...but the load is only 37 amps...unit is rated at 60 amps

    the installation has been working for at least 5 years without a problem the way it is wired...with everything on earth leakage.

    the only thing that has changed...is earth leakage tripped a couple of weeks ago for the first time and ever since then this random tripping occurs.

    after dicussing this issue with another person...something which i need to look into are the surge arrestors on all the plugs for the computors...that they have not been damaged...which i would asume should be causing the earth leakage to trip all the time not just randomly.

    there are a few thing i have thought about...
    replacing the earth leakage...which i am not convinced is the problem.
    seperate all the circuits and remove the lights and a/c units from the earth leakage which has been working for more than 5 years without an issue...why now suddenly is there a problem.
    take all the computers...servers...battery charges...ups etc off the earth leakage and fit red dedicated sockets...but again if it worked for 5 years the way it is why suddenly now would it need to be changed.

    i do realise because the current adds up to a point...they could eventually get higher than the rated 25-30 mA and will start tripping...

    any ideas?
    Last edited by murdock; 07-Feb-12 at 08:26 PM.

  2. #2
    Diamond Member AndyD's Avatar
    Join Date
    Jan 2010
    Location
    Cape Town
    Posts
    4,923
    Thanks
    576
    Thanked 934 Times in 755 Posts
    It's not easy to give advice about specific installations and why or where an insulation fault might be. I can give some general advice but please forgive me if I'm telling you things you already know.

    1 Global IR testing.
    When you want to check the leakage current for all the circuits together, rather than clamping every individual neutral the easiest way is to clamp the live and the neutral (on the load side of the RCD) together (at the same time) in your clamp meter. What you're looking at then is the difference in currents between them ie the earth leakage current. In an ideal world you'll get a zero current reading assuming infinite IR because the current in one direction on the live wire is cancelled out by the current flowing in the reverse direction on the neutral wire. In technical terms this results in zero hall effect so the tester would read zero unless there's leakage.

    2 Another thing is you need a clamp meter with an appropriate current range. Even a scale with 1Amp FSD is too high when you take into account the resolution of the tester and the accuracy discrepency it might not even give a reading for 5 or 10mA. You need a tester (clamp meter) with a 100mA scale or even better a 50mA scale. Many of the testers sold at general electrical wholesalers are not appropriate for this kind of fault finding. Unfortunately a good tester that is suitable will set you back around R4000.00.

    3 RCD's (earth leakage breakers) are very prone to nuisance tripping because of poor or loose terminations. Most RCD's will have up to 40 Amps through them in a domestic installation and the terminations are prone to generating heat. This can result in either nuisance tripping to start with and can end up with burned insulation on the wiring and even a fire. For technical reasons (not relevant here) the neutral is always more prone to heat damage than the live side. If you have an unexplainable tripping problem always remake or re-terminate the four connections on the RCD and the connection on the main neutral bar. There's a fair chance this might sort out the problem.

    4 Finally, testing an RCD unit is critical, it's a waste of hours and possibly days if you're chasing faults that aren't there or are so small they're inconsequential. To test an RCD you need to do a leakage current ramp test, this should determine the tripping threshold in mA. You should do this test with all load side wiring disconnected. Ignore the first ramp test because if the unit doesn't get regularly tested by the customer there will be resistance in the internal mechanical components. Take note of the second and third test results which should be between 25-30mA.

    5 The next test is the tripping time. Again you need a specific tester (also expensive) because many RCD's only monitor for leakage currents during one half cycle and not the other. So, if you introduce a leakage fault at the beginning of the 'blind' half cycle the process of tripping won't start until the beginning of the next half cycle. This means there could be a delay of up to 10 milliseconds before tripping starts depending exactly when the fault occurs. A good RCD tester will have be able to apply a fault on both half cycles so you can accurately measure the tripping times and check for consistancy. If the tripping times are over the manufacturers specs or they're not consistant then throw the RCD away. It's a sure sign it's suffered excessive wear and tear or internal damage.

    6 Something that has a large effect on the performance of the RCD is the main earth. You need to take readings of the incoming earth and the local earth rod on a TN-C-S installation. The earth impedence and stability affects the tripping times of the RCD. The neutral-earth voltage under load should also be checked. Obviously acceptable readings vary depending on incoming supply and earthing arrangements. All that said I know there's nobody here who would work on any installation an not test the earths and the bonding so you can ignore ths last point.

    Hope you find something here that might be helpful although I'm sure you know most, if not all of it already.
    _______________________________________________

    _______________________________________________

  3. Thanks given for this post:

    Dave A (08-Feb-12)

  4. #3
    Suspended
    Join Date
    Oct 2007
    Posts
    2,346
    Thanks
    0
    Thanked 254 Times in 209 Posts
    thanks andy...it just trying to remeber everything...thats the challenge...i learn new things everyday...and forget more than i learn...calling a friend is sometimes the best solution

    i do have a mA current clamp meter with a 200mA and 2000mA range.

    and i do have a fluke 1653 which has a trip time facility.

    i think i will make up a check list
    2 things i will add...loop impedance test at all the plugs
    test the surge protection plugs...i have a feeling that that is where the problem could be...there was a surge of power or a black out of sorts which causded problems in other offices a couple of weeks ago at about the same time this problem started...i believe it could be one of the plug tops which are breaking down.

  5. #4
    Diamond Member AndyD's Avatar
    Join Date
    Jan 2010
    Location
    Cape Town
    Posts
    4,923
    Thanks
    576
    Thanked 934 Times in 755 Posts
    The clamp meter even with a 200mA scale is on the high side. If you're measuring values of 3-5 mA for example you're working at only 2% of the testing range which will mean the accuracy of the tests would be very low.

    Surge protection is another story as well when it comes to IR testing. Most surge protection circuit start dumping to earth once the clamping voltage is reached. This can be the cause of misleading reading when testing, especially at 1000 volts. If there's sureg protection on the circuit then you need to test at 250v to get an accurate reflection of leakage on that circuit under normal use. It's worth noting that most computer power supplies have built in MOV surge protection and this can cause false reading and can also be damaged during IR testing.
    _______________________________________________

    _______________________________________________

  6. #5
    Silver Member
    Join Date
    Jul 2008
    Location
    Germiston, South Africa
    Posts
    397
    Thanks
    24
    Thanked 47 Times in 43 Posts
    Thanks for this information Andy. I find intermittent tripping irritating and difficult to diagnose but I will take use your test methods next time and hopefully be more successful.

  7. #6
    Gold Member Sparks's Avatar
    Join Date
    Dec 2009
    Location
    Port Elizabeth
    Posts
    890
    Thanks
    20
    Thanked 127 Times in 96 Posts
    Quote Originally Posted by murdock View Post
    thanks andy...it just trying to remeber everything...thats the challenge...i learn new things everyday...and forget more than i learn...calling a friend is sometimes the best solution
    Now you talking

  8. #7
    Diamond Member AndyD's Avatar
    Join Date
    Jan 2010
    Location
    Cape Town
    Posts
    4,923
    Thanks
    576
    Thanked 934 Times in 755 Posts
    Quote Originally Posted by Leecatt View Post
    Thanks for this information Andy. I find intermittent tripping irritating and difficult to diagnose but I will take use your test methods next time and hopefully be more successful.
    Some RCD tripping problems can be a real swine to find. Obviously it helps enormously if you have the right testing equipment and buckets of patience. There's occasions where I've had intermittant faults and used a data logger to monitor leakage current over several days just to get a handle on how large and how frequent the fault is. It can be a time consuming exercise.

    A lot of headaches can be avoided if you inform the customer thoroughly before you start. If he thinks it's going to be a ten minute job and it ends up requiring multiple visits and a full day labour time then you're going to get pressure from him, he's going to lose confidence in you and you'll probably struggle to get paid for your time as well. I always explain that the earth leakage is made up of probably a dozen different faults that are large enough to be measured and the RCD just sees the result of them all added together. The name of the game is to reduce the total leakage so it stays under 30mA 100% of the time.

    The first thing to do is to remedy the the larger faults in order to reduce the overall leakage to around 30% or less of the tripping threshold. With a 30mA RCD I would want to see less (much less) than 10mA of leakage under normal operation of the installation.

    The second thing I would do is IR test all appliances in the house. This is easier said than done because I would include things like TV mast amplifiers, air conditioners, appliances hidden away in cupboards, garden sheds, garages, irrigation systems, swimming pool systems and even PIR flood lights that have been supplied through the back of a socket. I route through cupboards, I rummage in lofts, I feel rocks, even bedroom 'toys' aren't exempt from testing. If it's got a 3-pin plug on it then it needs to be tested.

    I've found hundreds of things that have been installed by running a sneaky cable into the back of a socket and tapping into the supply. These things range from swimming pools to garden lighting to security systems to braai lights. The only way to find exactly what the RCD is looking at is to drop all sockets forward and check for spurred supplies going out of them. I also use this exercise as an excuse to draw up a complete plan of the installation which makes life easier for future testing.

    For the really elusive faults I have a 4 channel data logger that can monitor 4 circuits simultaneously at a decent resolution. Thankfully I only need to set it up maybe once a year. There's also no substitute for experience. For example there's a couple of makes of washing machine that use an arrangement of double pole relays on the PCB to supply the motor and the heating elements. What this means is the motor and elements are invisible during a normal IR test. To test the machine fully you need to remove the back cover and test the invisible items separately.
    _______________________________________________

    _______________________________________________

  9. #8
    Full Member
    Join Date
    Feb 2011
    Location
    Scotland
    Posts
    46
    Thanks
    2
    Thanked 5 Times in 4 Posts
    Some really good info from you Andy,hope you get it sorted Murdock let us know

  10. Thanks given for this post:

    AndyD (09-Feb-12)

  11. #9
    Diamond Member AndyD's Avatar
    Join Date
    Jan 2010
    Location
    Cape Town
    Posts
    4,923
    Thanks
    576
    Thanked 934 Times in 755 Posts
    I re-read your opening post now I'm back in the office;

    Quote Originally Posted by murdock View Post
    all clear except the light circuit...a low reading of just below 1 meg...normal on fluorescent fittings with electronic ballasts
    There seems to be something of an urban legend going around that electronic ballasts on fluorescent lights cause earth leakage. I think it might have been started by a manufacturing company who ran off a bad batch, they started this myth to allow them to sell the faulty units without getting warranty claims

    Seriously, if any type of lighting ballasts or switchgear tests low IR at 250 Volts then it's faulty. I would expect it to test at least 10 megs @250V end of story.

    In the UK and the US there's a similar urban legend about household fridges being expected to cause RCD tripping. I've never understood it, a refrigeration compressor is hermitic and completely moisture-free. It should have near infinite IR unless it's burned out which isn't any more common than any other appliance. I know fridges have a defrost element that operates in a moist environment but so do washing machines and steam irons and kettles. Still the myth goes on though.


    Quote Originally Posted by murdock View Post
    so i do some other tests
    mA readings on the earth wires
    circuit
    1. 0.4 mA
    2. 0.0 mA
    3. 2.6 mA
    4. 7.8 mA
    6. 0.2 mA
    7. 24.3 mA
    Your total leakage here is already more than the 30mA tripping threshold.

    It's not uncommon to get a fault on an appliance where the earth is no longer connected and the IR is low. This usually results in the customer complaining of a tingling feeling when they touch it. With this type of fault the leakage current is no longer flowing down the CPC or earth wire, it may be flowing down an extraneous earth bond somewhere or it may not even be flowing at all until the customer touches the appliance chassis. You won't see this leakage current when you clamp the individual earth wires at the DB. You'll only see this leakage current if you simultaneously clamp the L+N at the load side of the RCD. The best way to find this problem is to PAT test the appliances.


    Quote Originally Posted by murdock View Post
    main incoming earth 123 mA (this i dont understand)
    Where exactly did you get this reading?
    _______________________________________________

    _______________________________________________

  12. #10
    Suspended
    Join Date
    Oct 2007
    Posts
    2,346
    Thanks
    0
    Thanked 254 Times in 209 Posts
    the 24 mA is the light circuit...

    123 mA was the first test i did at the main DB clamping the main earth...but since i have removed the surge arrestors plug tops...has dropped

    i have been keeping records of all the tests...will post surge protected plug tops results as sooon a s i have time...very busy at the moment...got a long weekend of work...and trying to resolve these earth leakage issues...which are just tooo time consuming...like reports...people just cant understand why you need to charge for the 5 hours you spend creating reports...power quality and thermal reports...someone has to review the data collected.

    this is getting interesting...thanks again andy for sharing your vast knowledge

Page 1 of 2 12 LastLast

Similar Threads

  1. High Failure Rate in Business Start-ups
    By zivhave in forum General Business Forum
    Replies: 10
    Last Post: 22-Aug-13, 01:27 PM
  2. SME Failure Statistics
    By Gary Simmonds in forum Business Finance Forum
    Replies: 5
    Last Post: 10-Apr-13, 02:47 PM
  3. Random earth leakage problem
    By BillH in forum Electrical Contracting Industry Forum
    Replies: 39
    Last Post: 23-Jan-12, 09:37 PM
  4. CBI earth leakage units
    By murdock in forum Electrical Contracting Industry Forum
    Replies: 2
    Last Post: 29-Oct-09, 12:44 PM
  5. [Opinion] USA bailout failure
    By Morticia in forum General Business Forum
    Replies: 14
    Last Post: 10-Oct-08, 04:30 PM

Did you like this article? Share it with your favourite social network.

Did you like this article? Share it with your favourite social network.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •