Announcement

Collapse
No announcement yet.

Beginner Working on a Self Install

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #61
    So based on the above discussion, what can we advise the OP about lightning protection for his ground mount array? Since the array is supported by wood posts (a fairly good insulator) then no connection to earth is made and no additional grounding (ie grounds rods) should be installed at the array. Thus we are treating the array just as we would an external generator and would not define it as a "structure" per NEC rules. A four wire connection to the house main service entrance is all that is needed. But if his AHJ requires him (as I think they will) to sink a ground rod or two at the array then he should bond the electrodes together underground between the array and the main service entrance using a separate thick copper conductor, as max2k recommends.

    In the case of ground mounts that use metal pipes embedded in concrete a few feet in the earth, then a separate thick copper conductor would be bonded below ground from array structure to grounding electrode at the house in addition to the four wire connection. But IMO it depends on distance between the two structures. Then a risk/cost analysis is necessary. A four wire connection many hundreds of feet between the structures will have inherent resistance and thus force a side strike into the ground or reduce the step potential because of distance at either site without damaging the other. How risk adverse are you and how much money do you want to spend?

    In all cases an EGC is necessary to protect people from faults in the equipment wherever located.

    Comment


    • #62
      Originally posted by DanS26 View Post
      So based on the above discussion, what can we advise the OP about lightning protection for his ground mount array? Since the array is supported by wood posts (a fairly good insulator) then no connection to earth is made and no additional grounding (ie grounds rods) should be installed at the array. Thus we are treating the array just as we would an external generator and would not define it as a "structure" per NEC rules. A four wire connection to the house main service entrance is all that is needed. But if his AHJ requires him (as I think they will) to sink a ground rod or two at the array then he should bond the electrodes together underground between the array and the main service entrance using a separate thick copper conductor, as max2k recommends.

      In the case of ground mounts that use metal pipes embedded in concrete a few feet in the earth, then a separate thick copper conductor would be bonded below ground from array structure to grounding electrode at the house in addition to the four wire connection. But IMO it depends on distance between the two structures. Then a risk/cost analysis is necessary. A four wire connection many hundreds of feet between the structures will have inherent resistance and thus force a side strike into the ground or reduce the step potential because of distance at either site without damaging the other. How risk adverse are you and how much money do you want to spend?

      In all cases an EGC is necessary to protect people from faults in the equipment wherever located.
      EGC is required and as you noted it solves separate problem not relevant to lightning protection. Problem with EGC or any other conductor coming from remote site is in case of a strike at the remote site its remote end gets under tens of kV potential (and I don't think wooden structure will hold that potential, it will also become a conductor IMO) which it readily brings to the house due to low conductor impedance compare to the ground. To protect from this one would create alternative shorter path with even lower impedance to bypass ppl / equipment / pets. Use of correctly wired TVSS at the entrance from the site is also recommended to effectively clamp whatever arrives to the local ground potential. 'Correctly wired' here means RF like for the frequencies of up to 1 MHz due to short times these events occur at.
      Last edited by max2k; 09-23-2017, 03:24 PM.

      Comment


      • #63
        Originally posted by max2k View Post
        how is that not dangerous? In case of a strike near the solar site that site jumps in potential but the house being remote would remain where it was. Then you have wires coming from the site pretty much bringing that site 'jump' potential into the house over AC wires.
        I think it wouldn't be much difference in danger - since there could also be a lightning strike to a tree near the already existing house ground-rod. And that would have the same effect as a strike at the more remote ground rod next to the array.

        At least I'm assuming the existing ground rod is right by the main breaker panel and that's where the wires to the sub-panel for the solar sub-panel are located.


        I like the suggestion of adding a thick wire buried underground that connects all the ground rods . It's something that would reduce the current through the wires connecting the panels. And I don't see any harm in that. But I'm not a licensed electrician

        Comment


        • #64
          Originally posted by foo1bar View Post

          I think it wouldn't be much difference in danger - since there could also be a lightning strike to a tree near the already existing house ground-rod. And that would have the same effect as a strike at the more remote ground rod next to the array.

          At least I'm assuming the existing ground rod is right by the main breaker panel and that's where the wires to the sub-panel for the solar sub-panel are located.


          I like the suggestion of adding a thick wire buried underground that connects all the ground rods . It's something that would reduce the current through the wires connecting the panels. And I don't see any harm in that. But I'm not a licensed electrician
          problem is potential gradient- check out the video from the link in post #60 - it shows potential values involved vs distances. So unless the strike happens at the exact 'electrical center' between house and remote site both will experience difference in potential easily in tens of kV range relatively to each other. Equalization of this difference will happen through any conductor connecting the sites. Lightning doesn't need to strike at the array directly, it's enough to strike the ground anywhere in 300'-500' radius and grounding rods at both sites will 'pick up' potential at their locations.
          Last edited by max2k; 09-23-2017, 06:00 PM.

          Comment


          • #65
            Originally posted by max2k View Post
            Equalization of this difference will happen through any conductor connecting the sites. Lightning doesn't need to strike at the array directly, it's enough to strike the ground anywhere in 300'-500' radius and grounding rods at both sites will 'pick up' potential at their locations.
            But is there significant additional risk compared to a lightning strike near the main panel's grounding rod?

            The illustrations in that video show a ground rod that is connected to equipment - so the current goes from the ground into the ground rod, into the equipment to the EGC, back to the main panel and down into the primary ground.

            This case would be more like a sub-panel in a structure with a separate ground rod. It is fairly common to have a separate building 2-500ft away with a 4-wire feed to it and a pair of ground rods at the remote building.

            Are those a significant danger?

            Comment


            • #66
              Originally posted by foo1bar View Post

              But is there significant additional risk compared to a lightning strike near the main panel's grounding rod?

              The illustrations in that video show a ground rod that is connected to equipment - so the current goes from the ground into the ground rod, into the equipment to the EGC, back to the main panel and down into the primary ground.

              This case would be more like a sub-panel in a structure with a separate ground rod. It is fairly common to have a separate building 2-500ft away with a 4-wire feed to it and a pair of ground rods at the remote building.

              Are those a significant danger?
              IMO- it depends. If remote building has TVSS clamping incoming wires to the local ground it's OK. If they're just left there this would be safe until the first nearby strike: as the pic shows both grounding rods would 'pick up' their respective local ground potentials and here you have L1/L2 coming from another site where this potential is tens of kV apart. No insulation would sustain that so equipment will be pretty much gone without TVSS. N is an interesting case - it would 'bond' both sites together so equalizing current would flow through it and if it can sustain it the damage might be limited.

              For this reason I think equipment should always be grounded to the local ground using local EGC eventually connected to the local grounding rod. It will protect someone standing next to some equipment on the ground and touching the case for example- as EGC is supposed to. Bringing EGC from another site should be avoided as it will bring those tens of kV in case of a strike and even with EGC being connected to the local ground equalizing current might create voltage spike locally. Since EGC is also connected to equipment grounds this would be dangerous.

              In case you have multiple grounding rods around the structure all of them should be bonded together by dedicated thick buried wire creating GES which will ensure the structure will have equal potential within its perimeter but not necessarily 0 in case of a strike. Then local service panel gets connected to that GES in exactly 1 point by local GEC and then local EGCs connected to that through panel's EGC bar. This way all your equipment ends up under same potential no matter what happens.
              Last edited by max2k; 09-23-2017, 08:27 PM.

              Comment


              • #67
                Wow, sorry, I've missed a lot of discussion over the weekend. This isn't all extremely familiar to me, so I'll have to re-read through many of the posts to fully understand some of the discussion above. In general, it hasn't really "clicked" with me yet.

                Nonetheless, yes I have a 4-wire feed to what I suppose would technically be considered a sub-panel. This panel is the prewired SMA Multigate. The one thing that I do find odd about the connection is that although there are DIN terminal blocks for a N wire, there is/was nothing connected to the internal side of the neutral.

                This sub-panel (Multigate) is connected to an existing sub-panel that I installed in my garage a few years ago, so hopefully that doesn't create any other issues. That sub-panel is also connected with a 4-wire feed and has isolated ground/neutral bars.

                Now, moving off the subject of LPS for a moment. I had some issues last week with my inspector not showing up for the inspection. Then he finally did show up (a day late) and inspected things while nobody was home. All he was able to see was the exterior sub-panel (Multigate), as he had no access inside my home. He failed my inspection until he can see the remainder of the system, but also mentioned that I need to have a disconnect switch installed near my meter. I understand the logic for wanting that, but as far as I can tell, NEC 690 does not require the disconnect to be installed at or near the meter. 690.14 states "the PV disconnecting means shall be installed at a readily accessible location either on the outside of a building or structure or inside nearest the point of entrance of the system conductors." I have installed my disconnect inside my garage, at the location where the conductors enter the building. Is this acceptable, or do I really need to install a disconnect near my meter? My meter is on the complete opposite side of my house and would require hundreds of feet of trenching/wire. I have submitted a request to my power company to clarify on the matter.

                Comment


                • #68
                  Update - I received confirmation from my power company indicating agreement with interpretation of the NEC requirement and how I have things installed. I had offered to install a placard to my meter, indicating the disconnect's location. They confirmed that this is satisfactory and meets their installation standards. Hopefully the AHJ will agree.

                  Comment


                  • #69
                    Originally posted by brant2000 View Post
                    690.14 states "the PV disconnecting means shall be installed at a readily accessible location either on the outside of a building or structure or inside nearest the point of entrance of the system conductors." I have installed my disconnect inside my garage, at the location where the conductors enter the building. Is this acceptable, or do I really need to install a disconnect near my meter?
                    So you have the wires come up into the meter (mounted on outside wall of your house?) - then they go back underground, around the house to the opposite side and that's where they first enter the building?

                    Seems like a disconnect right where it enters the house would be compliant.

                    BTW - the POCO isn't a higher authority than the inspector. So just because they say "OK" doesn't mean the inspector has to. You have to make both the AHJ and the POCO happy.
                    Probably you're fine - and it's just that the inspector couldn't see the shutoff so assumed it was totally missing.

                    Comment


                    • #70
                      Originally posted by foo1bar View Post

                      So you have the wires come up into the meter (mounted on outside wall of your house?) - then they go back underground, around the house to the opposite side and that's where they first enter the building?
                      No, they enter my house on the opposite side of the meter. They never get within ~150' of the meter.

                      Originally posted by foo1bar View Post

                      BTW - the POCO isn't a higher authority than the inspector. So just because they say "OK" doesn't mean the inspector has to. You have to make both the AHJ and the POCO happy.

                      Understood. The inspector was quoting me requirements from a pamphlet that he received from the POCO. That's why I figured I needed to get them to clarify the requirements directly.
                      Attached Files

                      Comment


                      • #71
                        In the 2014 code if your AC (and DC) disconnects are not readily available next to the service entrance (see 690.13) then you fall under 690.15(A)(4) - you have to have a plaque describing exactly where the disconnect is and the plaque must be visible and nearby the service entrance. There are also requirements about the plaque - see 705.10 - pvlables.com is a good place to have them make this plaque.

                        Some of the wording is vague - lots of comments on this in the mike holt solar forums, and they made a stab at cleaning it up in the 2017 code.

                        The plaques can be as simple as this (middle sticker) or they can be line diagrams of your home with arrows showing the disconnect:

                        https://www.flickr.com/gp/149650329@N03/6dpL93
                        Last edited by tyab; 09-26-2017, 04:06 PM.

                        Comment


                        • #72
                          Thanks tyab. This is what I'm having made up by an individual that I ordered some other labels from on ebay. Hopefully it satisfies the requirements. Solar Placard.jpg

                          Comment


                          • #73
                            So I finally had and passed my inspection with the AHJ. Now I've been dealing with some issues in getting the monitoring system to work correctly, which also appears to have discovered that the system keeps running into some hiccups.

                            From the best that I can gather, I'm getting some type of signal interference within the AC cables routing between inverters and down to the multigates. There are physically 2 daisy-chained strings of 12 inverters; however, whenever I scan the system using Sunny Explorer, it finds 16 inverters on each multigate (4 overlapping on each). Unfortunately, I haven't been able to figure out what is causing this interference. As instructed by SMA, I have made sure that there is sufficient clearance between the AC cables of each multigate, but that hasn't changed anything. The multigates come in a pre-wired enclosure, which tightly bundled all of the AC cables within the enclosure. I'm am going to attempt connecting one of the multigates with relocated wiring, to see if that has any effect. Otherwise, I'm not sure what else to try.

                            This internal confusion within the multigates appears to be causing the units to cycle themselves on and off-line, causing me to lose production every few minutes.

                            Comment

                            Working...
                            X