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 9 replies
  • Subscribers 27 subscribers
  • Views 444 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

Loss of mains for type tested generation units

Jam
Hi all,


I'm reviewing the Loss of Mains protection for a number of old (pre-2018) G59 type tested inverters. While for site commissioned units it is required to note the LOM detection method, the "G59 certificate" only records operation time. I haven't managed to find (yet) a statement that it must be a certain method; G59 only seems to say that the parameters must be XX for each method (depending on age) and leaves it open to the manufacturer to decide. Does anyone know where it states which method is used? I'm also trying manufacturers but not all of them still exist!


It's always been a niggle at the back of my mind, but it's not been an issue... Now with the ENA's requirement to update old settings it's relevant.


Ta,

Jam
  • 0
    I've a feeling I've got the wrong end of the stick here, but I'll try anyway.... (it'll bring the thread to the top of the list at least)


    As I see it an inverter will disconnect itself from the grid whenever it decides that the mains has been lost - and there are a number of criteria for making that decision - voltage, frequency, rate-of-change-of-frequency, vector shift and so on... The inverter will continually monitor all such variables and disconnect when any one (or deemed combination) go out of bounds.


    When comissioning on site however I would have thought it would be difficult simulate many of those conditions (you obviously can't mess with the real grid so you'd almost need your own little power plant with controllable frequency to simulate nasty grid conditions) - so would have imagined that the test would be just to open a handy disconnector between the inverter and the grid and time how long it takes the inverter to notice and disconnect - in effect simulating an extreme case for all the monitored variables all at the same time. So in a way the on-site test produces a time, but it can't be related to any particular detection method.


        - Andy.
  • 0

    AJJewsbury:

    . . . When comissioning on site however I would have thought it would be difficult simulate many of those conditions (you obviously can't mess with the real grid so you'd almost need your own little power plant with controllable frequency to simulate nasty grid conditions) . . . 




    Indeed, and this is the reason behind the certificated “type tested” inverters, whether they are certified to G.59, G.83, or the newer G.98 or G.99 standards. If however you modify one or more of the inverter parameters, the type-test certificate becomes invalid and my understanding is that the inverter can no longer be legally connected to the public supply without re-certification. 


    Suitable test kit can be purchased, and auto test routines aren’t too difficult to write if you are familiar with testing of power system protection devices. The test kit will probably set you back over £25,000 (I haven’t bought any for a while), plus you will need annual calibration for it, which doesn’t come cheap. You would probably also need to submit your results to the appropriate DNO on a site by site basis for approval, before the inverter is reconnected to the public supply. 


    Regards,


    Alan. 

  • 0

    Jam:

    . . . I'm reviewing the Loss of Mains protection for a number of old (pre-2018) G59 type tested inverters. . . 


    . . . Does anyone know where it states which method is used?. . . 




    It will be stated on the type-test certificate, either directly, or due to the values quoted. 


    Regards,


    Alan. 

  • 0
    Hi all,


    Firstly thanks for the replies and apologies for not coming back for far too long.

     

    When comissioning on site however I would have thought it would be difficult simulate many of those conditions (you obviously can't mess with the real grid so you'd almost need your own little power plant with controllable frequency to simulate nasty grid conditions) - so would have imagined that the test would be just to open a handy disconnector between the inverter and the grid and time how long it takes the inverter to notice and disconnect - in effect simulating an extreme case for all the monitored variables all at the same time. So in a way the on-site test produces a time, but it can't be related to any particular detection method.



    As an aside, an example of what I've normally seen used for commissioning external LOM protection relays. I've never seen it done for type-tested kit though, and only with VS or ROCOF for the LOM.

     

    Indeed, and this is the reason behind the certificated “type tested” inverters, whether they are certified to G.59, G.83, or the newer G.98 or G.99 standards. If however you modify one or more of the inverter parameters, the type-test certificate becomes invalid and my understanding is that the inverter can no longer be legally connected to the public supply without re-certification. 



    There has been a change to the Grid Code that requires all mid-sized generators to change their settings, retroactively. Yes, the DNO requires notification, and yes there are specific considerations re type tested units. Finding out what the LOM method is just the first step!


    It will be stated on the type-test certificate, either directly, or due to the values quoted. 



    That's what I had thought! But for example, see the attached test certificate... I can see that it remains stable for certain VS and ROCOF events, but the actual detection method isn't stated, insofar as I can tell. If I'm looking in the wrong place, please say!


    Needless to say, I have heard nothing back from this manufacturer.


    Thanks,

    Jam
  • 0
    do uksales@ginlong.com  have anything to say ? Perhaps they offer a reprogramming service?

    Being in Liverpool there will be less of a problem with time shift, though the language barrier may still apply... But equally they may not know either.


    This has not been on my RADAR, but reading quickly, it seems there is a desire to make inverter generation more tenacious and not drop off-grid so quickly, which I presume relates to grid stability concerns.

    Finding
    Where the requisite RoCoF setting cannot be achieved without additional investment, remove LoM protection from all generation except for synchronous and doubly-fed induction (DFIG) units

    Is a bit of a surprise, I guess it acknowledges that some kit cannot just be re-programmed, but to say OK, in that case disable Loss of Mains detection altogether seems a bit extreme. I presume it does not apply to many.
  • 0
    Nothing coming back so far! But the question applies to other manufacturers too; this is just an example.


    Indeed you're right re the intent (last August's load shedding event being an example of when it would have helped). I too feel that disabling LOM is brave... perhaps cleverer minds have determined that it'll mainly apply to smaller plant that is unlikely to be stable and over-/under-/volts/freq (which do stay) will kick in quickly.


    Thing is... lots of smaller inverter G59 type test certs look like this (it's based on the model form after all). So perhaps most of them do need changing or perhaps I'm getting overexcited (ha) and they're all mandated to use "frequency shift with resonant circuit" (picking up a different inverter certificate which does actually state the method) instead of VS/ROCOF in some standard somewhere.


  • 0
    That test certificate specifies three types of frequency trip. Absolute frequency at the top of page 6 and ROCOF and Vector Shift protection in section b on page 7. I would therefore say that the inverter uses all three possible methods. 


    Regards,


    Alan.
  • 0
    Strictly 7b is just a test for a no trip, so it may actually never sense VF or rate of change at all, and pass that test. We do not know for what it is setup for if anything

    I agree section 6 is clear and the nominal settings of 47 and 52 Hz for fast trip match the test results well, and are already near the proposed new settings.
  • 0
    If the protection is on the inverter, it will be mentioned on the type-test certificate. If it isn’t fitted, then it shouldn’t be mentioned. ROCOF and Vector Change can be difficult to simulate correctly. They have obviously mentioned the non-trip settings, so I would try these as they are proving stability. I would  then use something well out, perhaps by a fact of 10 or more to see a trip. 


    Regards,


    Alan.

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