How Ground Fault Protection in a PV system works

Discussion in 'General Electronics Chat' started by roma, Jun 10, 2011.

  1. roma

    Thread Starter New Member

    Apr 20, 2011
    12
    0
    I'm going through the Photovoltaic Systems book in preparation for the NABCEP Solar PV entry level exam.

    I'm stumped by the following. I understand that a grounding system is integrated into the PV system so that if a component blows up, current dissipates and does not shock a person.

    I also understand the role of a ground-fault protection device in the PV system. If the insulation on hot wires is damaged, the hot may energize the ground wire. This presents a fire hazard.

    But, how is the ground-fault protection device not activated when current flows through ground because of an actual fault in the system? How does the ground-fault protection actually work?

    Thank you.
     
  2. #12

    Expert

    Nov 30, 2010
    16,247
    6,743
    Without knowing your schematic, which would be very helpful, GFCI generally works by comparing the current in the "hot" wire and the "current carrying ground" wire. If they are not equal, some current must have escaped and the protector shuts off the circuit.
     
  3. roma

    Thread Starter New Member

    Apr 20, 2011
    12
    0
    #12, thank you for the input.

    Please examine this link.

    A ground fault protection device in a PV system is a fuse in the inverter that shuts off when it detects a ground fault through the GEC. This is different from a GFCI, which you were referring to.
     
  4. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,905
    2,157
    That might be code but fusing ground connections seems a pretty dangerous thing to do when surge currents from lighting or ground potential rise could cause a large voltage differential from the frame ground , DC ground and power terminals causing a shock and equipment hazard.

    Others seem to share my feelings. http://www.midnitesolar.com/pdfs/DC-GFP-Draft3-5.pdf
     
    Last edited: Jun 11, 2011
  5. roma

    Thread Starter New Member

    Apr 20, 2011
    12
    0
    NSAspook, thanks for the response.

    As I understand (I often am wrong), according to what you write, the code does not take into account whether the electron flow legitimately needs to be dissipated to ground, or whether it is a ground fault that must be stopped.

    This was my initial question. According to what I've read, the ground fault protection device is usually activated when the ground fault exceeds 1 A. As one solar module produces around 7 A, this doesn't make sense to me.

    Thank you!
     
  6. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,905
    2,157
    Because current doesn't flow through the grounding wiring under normal conditions they just picked a value that would trip quickly even if the fault wasn't a zero ohm dead short but something like a small bare spot in the wiring touching the frame with a resistance in the tens of ohms.

    Fire protection is the reason for DC ground fault protection on PV systems. The writers of this misguided section of code seems to have forgotten about all the possible bad side effects of unbonding the DC ground.

    http://www.nmsu.edu/~tdi/pdf-resources/pdf version divided PV:NEC/APPENDIX H.pdf
     
  7. Kingsparks

    Member

    May 17, 2011
    118
    5
    This may or may not help, I don't know from the information you gave but I'll post it anyhow.

    I Used to run into this all the time doing in-service for new equipment. One problem people have is confusing system ground and system neutral.

    In a proper working circuit, system ground and the "hot" wire will have the same current. Going to the supply, AC plug or whatever, you have the normal three wires. Hot, Neutral and Ground. Ground here is not circuit ground. The ground wire for an AC system under normal use with no faults has "NO" current flow. All current is between hot and neutral. Then if a "GROUND" fault occurs, the current in the "Ground" wire is sensed and the circuit is shut down.

    Since the system you describe is set up to protect people from being shocked, or electrocuted maybe, it could be using a similar system, maybe not.

    Another thought, could the fuse you mentioned in the inverter be part of a "crowbar" system where current to ground would blow the supply fuse?
     
    Last edited: Jun 15, 2011
  8. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,905
    2,157
    The fuse in the GEC on a DC side of a solar inverter circuit is only for fire protection on the DC powered wiring and operates a bit differently from the AC utility power. Placing a fuse in the supply circuit to protect from a short doesn’t work because a solar panel is a current source that will only supply a set current even when shorted, so if a panel will supply 16 vdc at 10A under normal conditions when shorted it will still only supply about 10A.

    As stated before, when there is a positive to frame ground fault the small GEC fuse will interrupt the fault current but this now creates a new voltage reference of the negative power wire to ground that might be several hundreds of volts on wire that before was at the system ground potential when bonded by the fused circuit.
     
  9. Kingsparks

    Member

    May 17, 2011
    118
    5
     
    Last edited: Jun 15, 2011
  10. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,905
    2,157
    The single wire ground fault (in a negative ground DC system) is the positive (ungrounded conductor) wire shorting to ground (with the negative wire bonded to ground via the fuse). The disconnect (per code on the ungrounded conductor) is usually near or in the inverter. If the ground fault is anywhere before the disconnect (very likely because the bulk of the interconnect PV wiring is from the panels to the inverter) opening the power wire disconnect does nothing to stop the fault because the fault shorts the ungrounded conductor to ground before the disconnect circuit.

    I think a better way to design the fault detector would be to remove the power conductor grounding requirement and use a ungrounded DC fault detection system instead.
    http://www.nmsu.edu/~tdi/pdf-resources/IAEI%20September-October%202010.pdf
    http://www.littelfuse.com/products/Protection+Relays/Ground+Fault+Relays/PGR-2601.html
     
    Kingsparks likes this.
  11. Kingsparks

    Member

    May 17, 2011
    118
    5
    Naspook.

    Thanks for the patients. Got to admit I was not considering the amount of interconnecting conductor need for a number of panels. I can readily see that would present the most probably source for a short.

    I have not had time to review all three links you listed although I have the first. A well written presentation and as soon as I get a bit of time I will peruse the other two. Good to know and now I have learned something new, my entire reason for joining this forum.

    Thanks again.
     
Loading...