Common neutral between two different inverters

SolarMusher
SolarMusher Solar Expert Posts: 176 ✭✭✭
Hi all,
I would like to add to my Magnum 48V 4.4kw inverter system an old/retired Outback vfx3648 to supply a few more loads and to charge batteries at a higher rate. Each inverter will be mounted on its own Epanel with common neutral (in and out) and gen input connected. Is that correct to share this neutral between both Magnum and Outback inverter or should I give up this common gen input configuration? I have a doubt.
Thanks for any help,
Erik

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    As long as the AC loads are "separate" between the L1/L2 Magnum vs the L1/L2 Outback outputs--And only the common connection is the neutral+earth bond, it should work.

    You will need to make sure that the genset does not bond neutral to earth in its panel. And double check that the other AC inverters do not bond their Neutral to earth inside themselves either. You need to pick one point where neutral and earth connection is made (probably one of the Epanels).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • SolarMusher
    SolarMusher Solar Expert Posts: 176 ✭✭✭
    BB. wrote: »
    As long as the AC loads are "separate" between the L1/L2 Magnum vs the L1/L2 Outback outputs--And only the common connection is the neutral+earth bond, it should work.

    You will need to make sure that the genset does not bond neutral to earth in its panel. And double check that the other AC inverters do not bond their Neutral to earth inside themselves either. You need to pick one point where neutral and earth connection is made (probably one of the Epanels).

    -Bill

    Hi Bill,
    The only neutral/ground bond I have actually in the system is in my main electrical panel, I will check it to be sure. Do you think it would be better to bond it only in the "main" Epanel (Magnum)?
    So, I just have to connect the Magnum single in/out neutral bus in the Epanel to the Outback single in/out neutral bus in the second Epanel, as long as there's only one neutral ground bond in the Magnum epanel and all loads separated. Did I get it right Bill?
    Thanks,
    Erik
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Yes, just one main neutral+earth bond somewhere.

    With to "independent" AC inverters--You sort of have two main panels (at least two main inverter supported load panels).

    For example, if you have two 120 VAC TSW inverters. You put one inverter on the Black+White bus. and the second inverter on the Red+white bus in the same electrical panel.

    I would be very nervous. Because the two inverters are not "in phase" (or 180 degrees out of phase like 120/240 VAC split phase power is)--The common neutral bus in the panel can end up carrying more current than you expect.

    For example, if you have two 50 amp AC B+R circuits with a split phase 120/240 VAC AC supply, the Neutral will be from 0-50 amps worst case current (Black loads "subtract" from Red loads for neutral current).

    However, with two independent AC Inverters, the 50/50 amp output currents can (worst case) add up to 100 amp in the neutral bus...

    Be very careful on what you do here. Also--You can end up with some confusion--AC inverter #1 fails and AC inverter #2 works--You will have a mix of "dead" and "live" circuits in your panel. Could confuse somebody working on the system later.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • SolarMusher
    SolarMusher Solar Expert Posts: 176 ✭✭✭
    BB. wrote: »
    Yes, just one main neutral+earth bond somewhere.

    With to "independent" AC inverters--You sort of have two main panels (at least two main inverter supported load panels).

    For example, if you have two 120 VAC TSW inverters. You put one inverter on the Black+White bus. and the second inverter on the Red+white bus in the same electrical panel.

    I would be very nervous. Because the two inverters are not "in phase" (or 180 degrees out of phase like 120/240 VAC split phase power is)--The common neutral bus in the panel can end up carrying more current than you expect.

    For example, if you have two 50 amp AC B+R circuits with a split phase 120/240 VAC AC supply, the Neutral will be from 0-50 amps worst case current (Black loads "subtract" from Red loads for neutral current).

    However, with two independent AC Inverters, the 50/50 amp output currents can (worst case) add up to 100 amp in the neutral bus...

    Be very careful on what you do here. Also--You can end up with some confusion--AC inverter #1 fails and AC inverter #2 works--You will have a mix of "dead" and "live" circuits in your panel. Could confuse somebody working on the system later.

    -Bill

    Bill, don't worry. I don't want to build a big "OutMag" inverter from the Magnum and the Outback. I want to add a AC fridge to my system but the magnum is already full loaded, so I'd want to connect all the resistive loads (opportunity load, water line heater, toaster) in a small electrical panel powered from the Outback and keep the Magnum on the real main panel to supply pumps, fridge, laundry and the rest of the lodge. As you said I'll need another "main" panel specificaly for the Outback loads. My main concern was that Magnum AE/PAE and Outback VFX are common neutrals and that somewhere in the system, they will be all connected to the same bus. Obviously you're thinking that's this won't be a problem even if they are electrically "independent" from each other and that's the answer I was looking for. In fact, I'm wondering if it wouldn't be better/clean to separate neutral In and neutral out on two neutral busses in the magnum epanel and to connect the outback epanel neutral in/out bus bar directly to the magnum neutral In only. Or it doesn't matter anyway as neutral are common on each inverter?
    Erik
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    I don't think it matters where exactly you bond the two neutrals (or bond Neutral A to ground and Neutral B to ground). As long as it is done in one place.

    And tying both neutrals together (in one place) or tying each neutral to a common ground point should not matter either. Electrically, it is pretty much the same thing.

    However, when you run your AC1 and AC2 loads--Make sure that you run the Black+Neutral (blk) to Loads #1 and Red+Neutral (red) to your loads #2.

    With some wiring following typical North American Split Phase wiring rules, you can get Romex with Black+Red+Neutral+Ground... You do not want to run this type of wiring (common neutral) to your various (mixed) Black+Red loads. Each 120 VAC load should have its own dedicated Hot+Neutral+Ground Romex (or equivalent).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Ethan Brush
    Ethan Brush Solar Expert Posts: 235 ✭✭
    Erik, so will you have two separate electric panels, one for each inverter? These inverters are isolated from the dc Inut side so you would make a NG bond in each panel, and typically run a tap from your grounding electrode conductor into each panel.
  • SolarMusher
    SolarMusher Solar Expert Posts: 176 ✭✭✭
    Ethan, I will have two separate 48V inverters sharing the same 830Ah battery bank.
    The first one is a magnum 120/240Vac which is actually supplying L1/L2 to its own dedicated electrical panel (main panel).
    The second one will be a Outback vfx3648 120Vac only that will supply L1 to another dedicated electrical panel.
    Both inverters are mounted on Midnite Epanels and will have a common 48V negative.
    Neutral/ground bond (x1) is actually in the main electrical panel but I will make it inside the magnum Epanel if these inverters have to share a common neutral.
    There are two options:
    #1: Actually, the 240Vac gen input is connected to the magnum Epanel and charges at something around 55Adc. I would like to take advantage of the Outback charger to increase the charge rate to 80Adc.
    To do that, I'd need to wire the two Epanel gen inputs to each other with a 120V wire. My main concern is to connect neutrals from Magnum and Outback inverters on the same single bus bar. As said before, each output would be separated and would supply its own dedicated electrical panel.
    #2: Forget the Outback charger, use the vfx3648 to supply its own dedicated electrical panel and keep neutrals from both inverters separated.
    Question is: is that ok to wire all neutrals from both different inverters on the same bus bar?
    Erik
    PS: are you talking about 2x Neutral/Ground bond (NG), one bond in each Epanel?
  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    As said above the ac side of any decent inverter is electrically isolated, thus neutral bonding multiple invertes is fine. The complication is really around how to achieve it code wise. From my perspective the NEC is already a little vague about off grid PV system bonding. Articles 250/690 offers 3 variants, which can be interpreted in various ways. Bills point is that two NG bonds electrically amounts to the same as one NG bong and a NN bond. And that you want only one NG bond per inverter. So that still leaves you with options. I would take the one that involves the minimum number of joins in the overall earhting system. The NEC likes the GEC to be unjoined or irreversibly joined. Once the GEC hits the earth bus system, maintaining good connections is the key to honoring the spirit of the NEC when it comes to grounding. For instance i have a hop from the GEC to the DC earth bus to the AC earth bus. Its a crimped lug and stainless post and nut solution, that is the next best thing to permanenty joined. However if you read articles 250 and 690 it really says that the GEC should go to the AC earth bus directly. It is worth a read, you can access it for free now at https://archive.org/details/nfpa.nec.2014
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    And to be really clear...

    Most TSW/PSW (true/pure sine wave) inverters have electrically isolated outputs.

    Most MSW (modified square/sine wave) inverter are not electrically isolated from DC input to AC output and you cannot ground bond a neutral or connect one leg of two different inverters together safely.

    As always, check the manual/manufacture to be sure. Many inverter manuals are very opaque about grounding (mostly MSW inverters).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Ethan Brush
    Ethan Brush Solar Expert Posts: 235 ✭✭
    Ethan, I will have two separate 48V inverters sharing the same 830Ah battery bank.
    The first one is a magnum 120/240Vac which is actually supplying L1/L2 to its own dedicated electrical panel (main panel).
    The second one will be a Outback vfx3648 120Vac only that will supply L1 to another dedicated electrical panel.
    Both inverters are mounted on Midnite Epanels and will have a common 48V negative.
    Neutral/ground bond (x1) is actually in the main electrical panel but I will make it inside the magnum Epanel if these inverters have to share a common neutral.
    There are two options:
    #1: Actually, the 240Vac gen input is connected to the magnum Epanel and charges at something around 55Adc. I would like to take advantage of the Outback charger to increase the charge rate to 80Adc.
    To do that, I'd need to wire the two Epanel gen inputs to each other with a 120V wire. My main concern is to connect neutrals from Magnum and Outback inverters on the same single bus bar. As said before, each output would be separated and would supply its own dedicated electrical panel.
    #2: Forget the Outback charger, use the vfx3648 to supply its own dedicated electrical panel and keep neutrals from both inverters separated.
    Question is: is that ok to wire all neutrals from both different inverters on the same bus bar?
    Erik
    PS: are you talking about 2x Neutral/Ground bond (NG), one bond in each Epanel?

    Note that we are talking about system grounding here, and that you have multiple systems. If you want to make the DC side a grounded system you may, but it is not required by code. System grounding makes one conductor of the system common and at the same potential as all non current carrying metal parts. This in conjunction with a fuse or circuit breaker will, in the event of a fault on the other system conductor, provide a low resistance path back the source, tripping the fuse and de-energizing the circuit. Note that The Earth connection is just for equipotential between the earth and the metal parts of the electrical system and has no role in the fault clearing process. Back to the DC system, I personally prefer not to ground it because 1)the voltage is low enough that is extremely unlikely to be a shock hazard 2) I dont like having the increased chance of a DC positive to case short which working/servicing the batteries or other controls particularly since I am always tweaking/adjusting/fiddling with a live system 3) the system may already be grounded by the charge controller 4) it is actually against code to bond the PV system (which is common to the battery system since all but the 600 volt charge controllers are un-isolated) at the battery bus. 5) even if there is a first fault on an ungrounded system, it now becomes a grounded system so running a grounded system is kind of like having a "get out of fault free card." Note that even if you dont ground the system, you still have equipment bonding and earthing which is tying all the metal parts/ground lugs of ALL SYSTEMS together and to the earth electrode(s). This is to create equipotential between all the parts and the earth from potential differences that might result from capacitance, inductance, and lighting induced currents.

    So that is my rant on the DC system grounding. Now lets consider the AC side. You have two isolated AC systems, and 120/240 volt AC systems are required to be grounded by the NEC. So you need to bond both. So you will have a neutral to ground bond in each AC panel. Note that this is different than making multiple N-G bonds on THE SAME system, which you should avoid. So as far as the grounding electrode conductor connection, when you have multiple service or separately derived system enclosures (the latter of which is what the NEC would call yours - a service is utility supplied), you would generally run a tap from you grounding electrode conductor to each service or separately derived system enclosure - in your case the two AC panels and the DC system negative if you choose to ground that. Even per the NEC the taps do not need to be irreversably spliced to the main GEC, a split bolt is fine. Another method is to mount a bus bar near your panels and run your panel ground taps, and all you grounding electrode conductors and bonds (rods, underground water pipe, gas pipe bond, building steel, metal piping system, etc) to that.

    I agree with the others on the DC side inverter connections - it doesnt matter specifically how the neutrals connect or where the GEC connection is made.

    I hope this helps and I didnt just confuse you more.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    BB. wrote: »
    And to be really clear...

    Most TSW/PSW (true/pure sine wave) inverters have electrically isolated outputs.

    Most MSW (modified square/sine wave) inverter are not electrically isolated from DC input to AC output and you cannot ground bond a neutral or connect one leg of two different inverters together safely.

    As always, check the manual/manufacture to be sure. Many inverter manuals are very opaque about grounding (mostly MSW inverters).

    -Bill

    And if all goes well and you have isolated outputs on both inverters, you still need to size the common neutral to handle at least the sum of the two maximum AC outputs, or even higher than that if the outputs have a high harmonic distortion content (which a good TSW will not have a problem with.)
    SMA SB 3000, old BP panels.
  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    If you want to make the DC side a grounded system you may, but it is not required by code.

    Id need to reread NEC, but i know for sure that, here, for PV Vocs over 120V, you are required to earth bond one of the current carrying dc conductors. Obviously this is focused on the higher array side voltages, but such systems inevtiably share PV and battery grounds.
    to the DC system, I personally prefer not to ground it

    Thats contrary to the prevailing belief in the PV industry, with the likes of Outback strongly recomending all DC systems using their inverters be earth bonded.
    because 1)the voltage is low enough that is extremely unlikely to be a shock hazard

    See above, many PV systems have Vocs as high or higher than US AC line voltages. The same logic for preventing electric shock hazards apply on the dc side as the ac side, For 12v systems on boats etc, sure no worries.
    2) I dont like having the increased chance of a DC positive to case short which working/servicing the batteries or other controls particularly since I am always tweaking/adjusting/fiddling with a live system

    Its like working on a car, a spanner on the battery positive is much more likely to come to grief on the car body than vice versa. However im not sure this is reason enough to decide either way, as care should always be exercised when working around batterys and DC disconnects in particular. You should always remove one of the battery terminals before working on disconnects.
    3) the system may already be grounded by the charge controller

    This is a seperate issue. Charge controllers that employ ground fault protection devices, or disconnects containing same should not be otherwise grounded, as the grounding is effected by the GFPI device itself, and hard grounding defeats those devices. Such devices are becoming increasinly required in many jurisdicitons.
    4) it is actually against code to bond the PV system (which is common to the battery system since all but the 600 volt charge controllers are un-isolated) at the battery bus.

    True, but what they mean is that the grounding shouldnt be connected on the PV side, but on the battery side, primarily so that only a single ground point is used,
    5) even if there is a first fault on an ungrounded system, it now becomes a grounded system so running a grounded system is kind of like having a "get out of fault free card."

    This is exactly why equipotential bonding is used. Without it a first fault can go undetected, then a second fault can then expose dangerous voltages to the user.
    Note that even if you dont ground the system, you still have equipment bonding and earthing which is tying all the metal parts/ground lugs of ALL SYSTEMS together and to the earth electrode(s). This is to create equipotential between all the parts and the earth from potential differences that might result from capacitance, inductance, and lighting induced currents.

    I know Bill has mentioned that one of the advantages of floating systems is that lightning induced surges in current carrying conductors may be netrualised where neither conductor is grounded. However the precursor for that is identical length, twisted pair type conditions. The flip side is that floating systems can have stray voltages induced in them from static and the like, causing the occasional tingle here and there, or risk to esd sensitive electronics.

    Of course where floating systems are employed, circuit protection muct be used on both conductors. This factor alone can tip the blance in favour of DC grounding., Even where PV voltages are 50 voltages or lower, and the primary benefit of DC grounding is not a factor.
    So that is my rant on the DC system grounding.

    Bit OT yes!
    Now lets consider the AC side. You have two isolated AC systems, and 120/240 volt AC systems are required to be grounded by the NEC. So you need to bond both. So you will have a neutral to ground bond in each AC panel. Note that this is different than making multiple N-G bonds on THE SAME system, which you should avoid. So as far as the grounding electrode conductor connection, when you have multiple service or separately derived system enclosures (the latter of which is what the NEC would call yours - a service is utility supplied), you would generally run a tap from you grounding electrode conductor to each service or separately derived system enclosure - in your case the two AC panels and the DC system negative if you choose to ground that. Even per the NEC the taps do not need to be irreversably spliced to the main GEC, a split bolt is fine. Another method is to mount a bus bar near your panels and run your panel ground taps, and all you grounding electrode conductors and bonds (rods, underground water pipe, gas pipe bond, building steel, metal piping system, etc) to that.

    I agree with the others on the DC side inverter connections - it doesnt matter specifically how the neutrals connect or where the GEC connection is made.

    I hope this helps and I didnt just confuse you more.

    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • Ethan Brush
    Ethan Brush Solar Expert Posts: 235 ✭✭
    If you want to make the DC side a grounded system you may, but it is not required by code.


    Id need to reread NEC, but i know for sure that, here, for PV Vocs over 120V, you are required to earth bond one of the current carrying dc conductors. Obviously this is focused on the higher array side voltages, but such systems inevtiably share PV and battery grounds.

    The NEC allows both grounded and ungrounded systems. There are some different requirements for each. Note that I am referring to the PV output circuit. The NEC seems to have no requirement for grounding battery systems, but unless you are using one of the two 600 volt charge controllers, the PV and battery are not isolated.


    to the DC system, I personally prefer not to ground it


    Thats contrary to the prevailing belief in the PV industry, with the likes of Outback strongly recomending all DC systems using their inverters be earth bonded.

    I dont agree with that statement. Transformerless inverters are rapidy taking over the gird tie inverter market, and those systems must be ungrounded. Also I try not to listen to what manufacturers say about code and earthing theory. I have found they frequently get these two topics wrong.


    because 1)the voltage is low enough that is extremely unlikely to be a shock hazard

    I would like to delete my #1 statement from the previous post. The reason is that making that statement implies that there IS an increased shock risk from an ungrounded system but the increased risk is ok because the voltage is low. That is not what I meant to say. I DO NOT believe there is an increased risk of shock from an ungrounded system. So that statement doesnt make sense.

    System grounding isnt about safety or shock prevention. It is just two different ways to set up electrical systems. Many parts of the world use ungrounded systems. Both have advantages and disadvantages.




    3) the system may already be grounded by the charge controller




    This is a seperate issue. Charge controllers that employ ground fault protection devices, or disconnects containing same should not be otherwise grounded, as the grounding is effected by the GFPI device itself, and hard grounding defeats those devices. Such devices are becoming increasinly required in many jurisdicitons.

    It is a separate issue only if someone really knows what they are doing, but with off gird that often isnt the case. If someone decides to ground the DC system and doesnt know what they are doing, and there already is a N-G bond made by the CC, then they can defeat the GFP and/or have current flowing on normally non-current carrying parts. I think this is a valid case for not grounding the DC battery bus.
    4) it is actually against code to bond the PV system (which is common to the battery system since all but the 600 volt charge controllers are un-isolated) at the battery bus.



    True, but what they mean is that the grounding shouldnt be connected on the PV side, but on the battery side, primarily so that only a single ground point is used,

    ??? Im lost. You seem to agree with me that it is against code to bond at the battery bus but then you say it should be made at the battery side. Maybe there is a typo there.
    5) even if there is a first fault on an ungrounded system, it now becomes a grounded system so running a grounded system is kind of like having a "get out of fault free card."



    This is exactly why equipotential bonding is used. Without it a first fault can go undetected, then a second fault can then expose dangerous voltages to the user.

    I agree. A second fault on an ungrounded system is like a first fault on a grounded system - a breaker or fuse will trip. There are three reasons we still bond all metal together and to earth in an ungrounded system: 1) so ground detectors can detect a fault anywhere on the system (these are used in facilities where a loss of power can have very negative consequences so it is important to find the first fault before there is second fault and a breaker kills the power), 2) to provide a path back to the source so a second fault will trip a breaker, 3) to provide equipotential between all the metal and the earth to lower the risk of touch potential between parts. Such potentials may result from capacitance,induction, less than perfect insulation and high impedance faults.


    I know Bill has mentioned that one of the advantages of floating systems is that lightning induced surges in current carrying conductors may be netrualised where neither conductor is grounded. However the precursor for that is identical length, twisted pair type conditions. The flip side is that floating systems can have stray voltages induced in them from static and the like, causing the occasional tingle here and there, or risk to esd sensitive electronics.

    I dont see how earthing/not earthing a system conductor will result in these effects.




  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    The NEC seems to have no requirement for grounding battery systems, but unless you are using one of the two 600 volt charge controllers, the PV and battery are not isolated.

    I can speak more authoritaively for NZ regulations and as i said all DC systems above ULV require grounding. Maybe you can lay your hands on the NEC articles that allow ungrounded LV DC systems.
    I dont agree with that statement. Transformerless inverters are rapidy taking over the gird tie inverter market, and those systems must be ungrounded.

    What i should have said was in the off grid world, and the OP is an off grid application. I agree with you that "ungounded" grid tie systems are the norm. However in general this is a product of other factors. Grid tied systems are pretty much all professionalyl installed, and the sector leans on the use of GFPI systems, in built to the inverters, which for non TL grounds the system anyway. Thats the solution that the grid tie world chose, rightly or wrongly. This topic has been discussed heavily on this forum and many of us feel that GFPI use in solar introduces as many problems as it solves. Im not going to repeat that all here, except to say that when a ground fault occurs it unbonds the system, leaving the system open to more serious faults if the first fault is not addressed. In the grid tie world this is less of an issue because the inverter shuts down and the system should then get the service it requires. In the off grid world this is not the case as the GFPI systems are not built into the inverters.

    As for TL inverters the reason they are not input side bonded is that the input is not isolated from the output, and grounding those would create fireworks. But again i reiterate that this forum, when we talk bonding, we are talking largely about DIY installations that are typically off grid, and i stand by the view that grounded DC side is the industry standard.
    Also I try not to listen to what manufacturers say about code and earthing theory. I have found they frequently get these two topics wrong.

    A good way to go. But just ask Ryan at midnite, hes a pretty fair indication of the off grids worlds stance on this.
    I DO NOT believe there is an increased risk of shock from an ungrounded system. So that statement doesnt make sense.
    System grounding isnt about safety or shock prevention.

    What makes you say that. In general terms electrical safety is about fire and shock risk reduction. By grounding one of the current carrying conductors, if the live conductor if shorted to metal surfaces, blows circuit pro, and delivens the hot surface, protecting the user that may touch that surface. I shoudl note that the origins of MEN type protocols do date from pre RCD/GFPI technology and to much of an extent the use of GFPI/RCD on the AC side does reduce shock risk to users. The same logic applies to DC as AC, with the proviso that high PV side voltages are limited to ingress points, disconnects, and what have you, not distributed through living environments as with AC. Also note that with solar being current limited instead of the cicruit pro popping, the voltages effectively get clamped to low levels acheiving the same effect.

    And i repeat that for ungrounded systems you must fuse both current carryng conductors. Otherwise two seperate faults can bypass all circuit pro, resulting in fire, or electric shock.


    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    The grounding one conductor vs floating a system...

    A floating system, a nearby lighting strike will probably create a "common mode" surge current/voltage. Basically both +/- (or L1 and L2) will both go up together (or down). Look at a 12 VDC system. + to - voltage is limited to about 15 volts differential voltage before something gets damaged.

    Common mode voltage (between the two power leads and ground), you can easily get hundred of volts or more between the power leads and ground.

    A ground referenced system lead--Now the + and - (or L1 and L2 for AC) are "different". It is easy to create a large differential voltage between the two power leads and fry the loads (and energy sources).

    This is called converting a common mode "signal' to a differential mode signal.

    For AC lines (residential), usually we use 600 VAC for design common mode voltages, and the equipment is actually tested at 3x (600 VAC) or 1,800 VAC at the factory to ensure that it will safely accept 600 VAC without failing/shocking somebody.

    However--While common mode signaling (not ground referenced) is used a lot in Engineering (using transformers, floating circuits, optical isolation, etc.)--For power systems, grounding one of the powerOn conductors (aka the Neutral for AC or Return Leg for DC). does have lots of advantages.

    First, as many people have said here, the grounded neutral (and tying plumbing, natural gas lines, sinks, etc. to common earth ground) means that nothing can be brought over ground reference. If something (like a shorted "AC or DC hot lead") attempts to raise the voltage of some metal/plumbing/electrical box, excessive current will flow and the breaker will trip. All is (relatively) happy.

    And a ground bonded neutral plus green wire safety ground--We have redundant conductors for power flow (neutral for normal and over current & shorts to neutral, and green wire for shorts to ground).

    What also happens with ground referenced power, it (more or less) ensures that the power "hot" leads will never exceed nominal voltage (line to line and line to ground). Floating power systems can "float" or be "driven" (by a short circuit) to other than ground referenced voltage. Lightning, cross from utility power lines (12kV line falls on 120/240 VAC transformer), static charge buildup (ungrounded metal structures 10's of feet or more above ground). Can "surprise" somebody working on the system.

    Also, many communication/signaling systems assume signal ground reference is very close to chassis ground (RS 232, RS 422, etc.). If you have a floating computer system power (floating 12 VDC), the "signal" ground is now floating with respect to earth. And if you connect this signal to a terminal/printer/other computer/etc., very often that device will act like a bonding point (tying signal ground to earth ground). With large computer systems--We very often will have terminals and printers with fried RS 232 cabling and/or fried interface boards because of DC/AC grounds that where not tied together. And old chain printers and such that developed a lot of static electric charge because of paper moving through the printer.

    So, many DC systems (12/24/48 VDC) use DC signalling as a ground reference (battery monitors, system component communications, etc.).-- So you really do not want the DC return (typically negative) to "float" relatively to the other components/people working/using the system.

    Also, my own theory is that the DC Battery bank is one massive capacitor. And the "Common Mode vs Differential Mode" problems are not very important for DC power systems. The batteries themselves do a very good job of keeping the plus and minus leads at rated voltage. And with DC return grounded, it keeps the lightning (and other) surges from propagating through the system and damaging other components/devices/DC loads.

    This is sort of confirmed by Windsun (now retired from NAWS)--He had said that the majority of lightning damaged for AC inverters was the AC output stage, not the DC input section. One reason I suggest paying attention to using Surge Suppressors and grounding AC neutral on even cabin power systems.

    Frankly, if there is a direct lightning strike--Nothing is going to prevent damage/destruction of your electrical equipment. However a good grounding scheme will help reduce bringing the lightning into the home and exposing the occupants to dangerous electrical conditions.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    I had a quick look, and it seems that prior to 2005 the NEC was even stricter than 'wiring rules' is here, actually requiring at that time all DC circuits over 50V to be grounded (article 690.41). However with the subsequent expansion of the pv industry since that time, that article has added provisions for ungrounded systems where all current carrying conductors are fused, the use of ground fault protection and some other clauses.

    About as much as you ever want to know about this can be had here and here:

    http://www.homepower.com/articles/solar-electricity/design-installation/ungrounded-pv-systems?v=print
    http://solarprofessional.com/articles/design-installation/ungrounded-pv-power-systems-in-the-nec?v=disable_pagination

    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • Ethan Brush
    Ethan Brush Solar Expert Posts: 235 ✭✭
    zoneblue wrote: »

    I can speak more authoritaively for NZ regulations and as i said all DC systems above ULV require grounding. Maybe you can lay your hands on the NEC articles that allow ungrounded LV DC systems.

    I had a quick look, and it seems that prior to 2005 the NEC was even stricter than 'wiring rules' is here, actually requiring at that time all DC circuits over 50V to be grounded (article 690.41). However with the subsequent expansion of the pv industry since that time, that article has added provisions for ungrounded systems where all current carrying conductors are fused, the use of ground fault protection and some other clauses.

    690.41 in the 2014 code was changed a bit more and they removed the voltage restriction. I still dont really like the way it is set up. I basically says that systems must be grounded than has an exception for systenms complying with 690.35.



    What makes you say that. In general terms electrical safety is about fire and shock risk reduction. By grounding one of the current carrying conductors, if the live conductor if shorted to metal surfaces, blows circuit pro, and delivens the hot surface, protecting the user that may touch that surface.
    First, as many people have said here, the grounded neutral (and tying plumbing, natural gas lines, sinks, etc. to common earth ground) means that nothing can be brought over ground reference. If something (like a shorted "AC or DC hot lead") attempts to raise the voltage of some metal/plumbing/electrical box, excessive current will flow and the breaker will trip. All is (relatively) happy.

    Right but with an ungrounded system, the first fault doesnt result in any shock hazard. After a first fault on an ungrounded system, it is now a grounded system. Its where you "would have started" with a grounded system. Yes are you touching one "live" conductor of the system whenever you touch a faulted metal object on an ungrounded system, but that is exactly what you do all day with a grounded system. The grounded/neutral conductor on a grounded system isnt any "less hot" than any other conductor. With an ungrounded system all metal piping, etc, and the ground are still bonded together to provide equipotential from capacitance and static. As I said, in broad terms, system grounding is not about safety and shock prevention. Many parts of the world use and have used since the beginning of electricity ungrounded systems with as good a safety record as grounded systems. In the US we chose grounded systems for most buildings (although there is a lot of ungrounded utility distribution). I dont think the grounded/ungrounded choice is about safety, its just two different electrical systems that each have pros and cons. One of the big disadvantages of grounded systems is it makes it much easier (possible) to get a line to ground shock ( arc flash/blast). A big disadvantage of an ungrounded system is first faults can be hard to detect and may not automatically clear, but then the argument can be made that what does a first fault matter? I read some figures once that some surprisingly high percentage of systems in parts of the world that use ungrounded systems, have a first fault - who cares it a grounded system now and the next fault will trip a breaker. Anyway so that debate has gone on for decades and I dont see any clear compelling argument or statitics that one system is superior to the other. ITs like ford vs chevy, liberal vs conservative policies, etc....
    Also, many communication/signaling systems assume signal ground reference is very close to chassis ground (RS 232, RS 422, etc.). If you have a floating computer system power (floating 12 VDC), the "signal" ground is now floating with respect to earth. And if you connect this signal to a terminal/printer/other computer/etc., very often that device will act like a bonding point (tying signal ground to earth ground). With large computer systems--We very often will have terminals and printers with fried RS 232 cabling and/or fried interface boards because of DC/AC grounds that where not tied together. And old chain printers and such that developed a lot of static electric charge because of paper moving through the printer.

    So, many DC systems (12/24/48 VDC) use DC signalling as a ground reference (battery monitors, system component communications, etc.).-- So you really do not want the DC return (typically negative) to "float" relatively to the other components/people working/using the system.

    Bill what you say is certainly valid, but I would argue that what it boils down to is not a problem with ungrounded systems, but just that the equipment was designed around or intended for use with a grounded system. Sure if you have something that is expecting a voltage reference to ground or is "cheating" and using the ground as a return conductor like say a dimmer of similar control, sure using those on an ungrounded system can cause problems. Similarly, US utility practices are just not compatible with ungrounded systems - its not that a deficiency with ungrounded systems, its just that our grid wasnt set up around that system.

    IT sounds like we basically agree that there are pros and cons to each system

    I just really like to fight the overemphasis on the importance of connecting things to dirt. IT is a widespread problem in our industry and I can definitlty say that probably more than half the electricians, utility works, and engineers I meet think grounding (both system and equipment) does so much more than it actually does. Many books say equipment earthing protects persons from shocks during faults which it cannot do. The NEC states that system grounding helps limit line surges and voltage imposed by lightning, but I dont see how that is possible and many very knowledgable people agree with me It is bonding everything together that is important. Connecting equipment to dirt is actually of very minor importance and highly unlikely to prevent shock or fire. Connecting a system dirt, well that is more a matter of your utility, local codes, and perhaps what the equipment is designed for.





  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    Right but with an ungrounded system, the first fault doesnt result in any shock hazard.

    Yeah, an awful lot of stars have to line up to actually come to grief, such is the nature of regulation these days.
    As I said, in broad terms, system grounding is not about safety and shock prevention. Many parts of the world use and have used since the beginning of electricity ungrounded systems with as good a safety record as grounded systems.

    Sure, partly just tradition.
    Many books say equipment earthing protects persons from shocks during faults which it cannot do.

    What anout this scenario. Appliance A shorts phase to chasis. Appliance B shorts neutral to chasis. Neither pops circuit pro. User grabs hold of both appiances and recieves fatal shock.

    Ok, ok, granted a lot has changed over the years. Double insulation, for one. Widespread and now mandated use of RCDs in all domestic installations, here now since 2005 as mentioned above.

    But ill say it a third time, before you go talking people out of grounding, the NEC clearly requires all ungrounded current carrying conductors to be fused. Thats exactly twice as much circuit pro. People being people always trying to cut corners. Be careful what you wish for.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • Ethan Brush
    Ethan Brush Solar Expert Posts: 235 ✭✭
    zoneblue wrote: »

    What anout this scenario. Appliance A shorts phase to chasis. Appliance B shorts neutral to chasis. Neither pops circuit pro. User grabs hold of both appiances and recieves fatal shock.

    Hmmmm, trying to figure this out. On a grounded or ungrounded system? The use of the term "neutral" implies a grounded system, but if a grounded system a breaker would have tripped due to the appliance A phase to chassis short. IF you were talking an ungrounded system, the neutral/second fault would trip the OCPD.

    But ill say it a third time, before you go talking people out of grounding, the NEC clearly requires all ungrounded current carrying conductors to be fused. Thats exactly twice as much circuit pro. People being people always trying to cut corners. Be careful what you wish for.

    I dont disagree about the OCPD required in all ungrounded conductors. In general though I am very opposed to codes and strategies that try to "brother in law proof" things. There are countless dangerous things that can happen from people cutting corners and/or not knowing what they are doing.
  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    OK, but what exactly was it again that you are saying is the advantage of ungrounded systems...i forget.

    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Grounding to "dirt" is really only helpful for lightning/static control. However ground/neutral bonding to local plumbing/gas line/RV chassis is what is important for safety reasons.

    With a floating system, yes, there is less issue with the possibility of shock--But unless the power system is tested/instrumented to validate that it is "still" isolated--Then you pretty much have to assume that there is still the possibility of shock.

    Ground Fault protection is really neat for AC systems... Very easy to design a device that will trip with ~5 mAmps of leakage current. However, for DC systems, there is (as yet) no real reliable/cheap GF detection circuitry.

    With Isolated power systems, you have to ensure that you use isolated loads (and charging system) too. If you have an isolated power system and a ground referenced load (such as digital signalling, AM/FM car type radios, etc.)--Then the system is no longer floating. It just has an "incidental" ground reference that probably is not rated to manage fault current needed to trip the breaker.

    There can be issues with grounding (positive vs negative grounding, cathodic protection, etc.).

    I will also add a comment about "non-isolated" GT type solar power systems. Yes, the array does not have a DC leg grounded... But remember that the 120/240 VAC split phase system is still ground referenced via the Neutral/Ground Bond.

    And (from what I remember when doing some earlier readings), the non-isolated GT inverter still needs to "test" the array (something like every startup?) to ensure that it is still "floating".

    In general, for typical DC power systems--I would still, personally, recommend DC bonding the return to earth ground rod (the same ground rod/earthing as the AC power systems).

    With floating/high current DC power systems that are not ground referenced (or use the NEC ground fail detection with ~1 amp fuse between DC return and earth ground), it makes for a much more complex set of failures--And as said before, without breakers/fuses on both + and - leads, these systems are not really safe (per the white paper I wrote a while ago).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Ethan Brush
    Ethan Brush Solar Expert Posts: 235 ✭✭
    zoneblue wrote: »
    OK, but what exactly was it again that you are saying is the advantage of ungrounded systems...i forget.

    That is a good question, LOL! Looking back and to recap, it started with the discussion on system grounding the DC side of an off grid system and I expressed some reasons for the case of NOT grounding the DC side: some codes, issues with isolation/nonisolation of the PV and Batt systems, and the GFP system implications. Then we got talking about grounded vs ungrounded systems in general, and I was just trying to point out some of the advantages and that ungrounded systems arent necessarily unsafe (lots of folks hear "ungrounded" and think danger).
    However ground/neutral bonding to local plumbing/gas line/RV chassis is what is important for safety reasons.

    I agree that regardless of the system being grounded or ungrounded, all the metal parts and piping systems should still be bonded together. Whether or not all this stuff is connected/referenced to one of the system conductors gets back to the grounded vs ungrounded system debate.
    With a floating system, yes, there is less issue with the possibility of shock--But unless the power system is tested/instrumented to validate that it is "still" isolated--Then you pretty much have to assume that there is still the possibility of shock.

    Good point
    I will also add a comment about "non-isolated" GT type solar power systems. Yes, the array does not have a DC leg grounded... But remember that the 120/240 VAC split phase system is still ground referenced via the Neutral/Ground Bond.

    Yes I almost mentioned that. It is somewhat silly to refer to such things as "ungrounded system" as they are the complete opposite, its just that the DC side conductors are the ungrounded conductors of the system. The NEC does allow actually ungrounded AND isolated PV systems and some transformer inverters do operate that way (I havnt worked with one but I have heard that is the case).