Making a small Off-grid Cabin-style install that would satisfy NEC?

KE5YD · February 2015

Subject describes what I'm wanting to do. I have 510 watts of PV from 2 grid-tie panels mounted on a metal ground-mounted rack 85 feet from the mppt controller

and batteries presently in the cabin. This install will probably always be 12V DC Only. I would like to put my PV current wires underground and do "steps" that

would suppress radio-noise. If anyone has or knows of an OUTLINE of "necessary steps" ; I would appreciate the info. I have read/looked at some older versions

of the NEC; but there's a number of things I'm not sure of. I have primarily been a Lurker on the Forum for a couple years or more.

HandyBob · February 2015

Howdy,
About all I will say is that I was in electrical engineering way back when, so I figured the way to do things would be to by all of the parts and tack it together myself. After spending many years in a solar powered RV and starting to build a home I realized that trying to find all the things needed and connecting it together, even for me, was just not practical. I bought a small Magnum panel and would do it again. Take a look at their AE catalog. They have things figured out. I would consider Midnite, but theirs is bigger than you need and I don't have many good things to say about their charge controllers. However, their surge protector is the only one to buy. I wish I had gone to 48V, not 24V, but I'm running a real house & shop building with 2000W on the roof. The only thing I don't like about this is having all of my eggs in one basket. I'd love to buy a separate charger some day so that I'm not relying on the Magnum. It seems to me that having to power the whole house from the generator while also powering the charger is ridiculous. So far this is just theory since the only generator I own is a used one out of an RV and it is sitting on a pallet with the cylinder full of oil. The house is probably about a year in the future.

BB. · February 2015

As a start, from our host NAWS:

http://www.solar-electric.com/reducing-electromagnetic-interference-pv-systems.html/

What brand/model of solar charge controller do you have?

Will you have an AC Inverter?

I assume this is for HF HAM mostly?

-Bill

KE5YD · February 2015

Thanks! for your comments Bob; and Bill the controller is a Morningstar 45mppt. Bill, if I were to use an inverter; it would be small, outboard(not house wired), and temporary. I'm interested in more as what to use and how to do things with the install that would satisfy NEC if I were to have it inspected. I suspect that most
personal off-grid installs do not meet code??

I would like to start with the metal ground-mounted panel rack. I know each panel must be grounded using proper hardware, the rack grounded with the panels.
Now, 1st question: Do THESE tie to a ground rod/grounding system at this point?
2nd question: Do I need or should I have a disconnect(one or both) conductors at this point OR can I just use the breakers in the cabin?
3rd question: For the buried conductors, I know there is direct burial wire with a ground conductor...I figure #10 is marginal for this distance and current.
#8 UF would be good; but more difficult to obtain. Does USE-2 wire have to be in a conduit? If I use 2 individual wires for the positive and
negative; should I also carry a 3rd wire for ground?
4th question: Does NEC require a Ground Fault Protection Device for This install? I believe that IT might meet the Exception?
5th question: Bob has expressed his opinion for a Surge Protection Device, and I believe his choice is good. Does anyone have any feedback on Others?

That's enough questions for this post. Your comments?

Photowhit · March 2015

KE5YD wrote: »

Thanks! for your comments Bob; and Bill the controller is a Morningstar 45mppt. Bill, if I were to use an inverter; it would be small, outboard(not house wired), and temporary. I'm interested in more as what to use and how to do things with the install that would satisfy NEC if I were to have it inspected. I suspect that most personal off-grid installs do not meet code??

Likely even those that pass inspection don't meet code!

KE5YD wrote: »

I would like to start with the metal ground-mounted panel rack.

Ground mounting 'Grid tie' type panels likely need to have the wiring run in conduit or be fenced, Wiring above 30 volts DC must have limited access.

KE5YD wrote: »

I know each panel must be grounded using proper hardware, the rack grounded with the panels.
Now, 1st question: Do THESE tie to a ground rod/grounding system at this point?

At what point? panels and metal enclosures and boxes must be grounded at the same point.

KE5YD wrote: »

2nd question: Do I need or should I have a disconnect(one or both) conductors at this point OR can I just use the breakers in the cabin?

1 usually positive, 2 panels of the same size don't need fusing when combined, but a disconnect breaker ahead of the charge controller makes life easier, but I don't think is 'required'

KE5YD wrote: »

3rd question: For the buried conductors, I know there is direct burial wire with a ground conductor...I figure #10 is marginal for this distance and current. #8 UF would be good; but more difficult to obtain. Does USE-2 wire have to be in a conduit? If I use 2 individual wires for the positive and
negative; should I also carry a 3rd wire for ground?

The loss is in voltage drop and might consider if the expense of heavier wire is worth the added 'wattage' or if you would rather save the money toward more panels.

KE5YD wrote: »

4th question: Does NEC require a Ground Fault Protection Device for This install? I believe that IT might meet the Exception?

KE5YD wrote: »

5th question: Bob has expressed his opinion for a Surge Protection Device, and I believe his choice is good. Does anyone have any feedback on Others?

Midnite's surge protectors are topnotch, as are their charge controllers in my opinion! (Not sure what Handy Bob uses or will use, but they are likely stand alone Like Morning star or buggy Like Xantrex or designed by boB and Robin... Though perhaps the newer Outback has little input from them...

BB. · March 2015

KE5YD wrote: »

Thanks! for your comments Bob; and Bill the controller is a Morningstar 45mppt. Bill, if I were to use an inverter; it would be small, outboard(not house wired), and temporary. I'm interested in more as what to use and how to do things with the install that would satisfy NEC if I were to have it inspected. I suspect that most
personal off-grid installs do not meet code??

I am not a code guy--But if your wire/conduit/protect you can get the basic requirements of the NEC. Your local building department may have others--There have been one or two that even required a meter socket to meet local code (yea--for off grid).

I would like to start with the metal ground-mounted panel rack. I know each panel must be grounded using proper hardware, the rack grounded with the panels.
Now, 1st question: Do THESE tie to a ground rod/grounding system at this point?

In general, there are two or three reasons ground the array frame.

First is to ground exposed metal from becoming "energized" -- By a short from your solar array, from an AC power line, a distribution line falling across the area, etc. Driving an 8-10 foot ground rod at the base of the off grid array (and bonding to cold water pipe, etc.). is a good start.

However, the connection of the ground rod to the earth can be as high as 25 Ohms or so--So it does not trip a circuit breaker back at the AC panel or DC panel... To do that, connecting a 6 awg ground wire from the remote array ground rod to the local/house/main ground rod (which is used to ground the rest of your system). Run the rest of your green wire grounds (from your AC panel, DC battery bus, cold water pipe, gas lines, etc.) to the local ground rod. The tie from the remote ground rod to the local ground rod will trip any breakers/fuses in the AC mains/DC mains if there is a short to the panel frames/racking.

Next grounding reason is for lightning. A direct run (short, no right angle bends, etc.) direct to the local ground rod. this also can help reduce static build up on the insulated metal array too.

There is no reason to ground reference the solar panel +/- leads on the array at the remote array location. You can use these surge suppressors if you have lightning in the area:

Midnite Solar Surge Protection Devices and Accessories

2nd question: Do I need or should I have a disconnect(one or both) conductors at this point OR can I just use the breakers in the cabin?

If I remember correctly, you have to use metal conduit to the disconnect. Once past the disconnect, you can use PVC inside the building (check NEC/Code--I am not a code guy).

If you have three or more parallel solar panel strings, then you need a combiner box +fuse/breaker per string. You really do not need a disconnect/breaker on the panel input to your controller--However, if you have breakers in the combiner box and/or a disconnect/breaker, people like this so they can turn off the array when working on the controller. The charge controller to battery bus should have a wire+fuse/breaker rated for 1.25x continous current (i.e., a 60 amp charge controller should have a 75 amp (round up to 80 amp) branch circuit. Again, many people like to use a circuit breaker -- Easy on/off -- Plus large fuses+holders are not cheap either.

Always check the manual for the charge controller--They should give you wire gauge/fuse/breaker ratings/etc.

3rd question: For the buried conductors, I know there is direct burial wire with a ground conductor...I figure #10 is marginal for this distance and current.

If you have parallel solar panel connections, you can run several parallel runs of 10 awg and have the combiner box back at the cabin.

#8 UF would be good; but more difficult to obtain. Does USE-2 wire have to be in a conduit? If I use 2 individual wires for the positive and negative; should I also carry a 3rd wire for ground?

I like to bond the remote and master/cabin ground rods together. You can use a direct burial 6 AWG (bare) cable for that next to the insulated run. I don't think there is any reason for a green wire in the DC circuit (solar panel circuit).

4th question: Does NEC require a Ground Fault Protection Device for This install? I believe that IT might meet the Exception?

Yes, NEC does require some form of Arc Fault prevention. the DC GFI was one method to do this--But, personally, I do not believe that DC GFI is safe at all... I would not install a DC GFI system (or I would short out the "sense fuse" between DC and Earth grounds.

If you "need" DC GFI / Arc Fault prevention system... Midnite has that built into their Midnite Classic charge controllers (Arc Fault detection monitors for electrical "noise" associated with an arc/spark in the wiring--And shuts down the controller if it occurs).

5th question: Bob has expressed his opinion for a Surge Protection Device, and I believe his choice is good. Does anyone have any feedback on Others?

Yes, those appear to be some of the best choices out there at this time.

Sorry I could not answer your other questions--Hopefully some others here can.

-Bill

Photowhit · March 2015

By the way - If you intend to have DC outlets in the cabin, the cigarette lighter outlet do not meet NEC code. I believe, and I couldn't find the reference (Backwoods home use to have an article about this) NEMA 5-20 (T slot?) outlets can be used when no 240volt AC is installed in the same building. They still sell the outlets and plugs but they should be available at or through large hardware stores.

oil pan 4 · March 2015

The NEC code for 2014 is only a few pages long, I am sure you could google search it, then you would need several more pages of tables and lists.
If I remember correctly I saw nothing in the code book about DC receptacles.
Easiest way to get around that would be to put the receptacle in some kind of housing connected to the fixed wiring by a disconnect. Then it would be considered "equipment" and then from that point almost anything goes.

Photowhit · March 2015

oil pan 4 wrote: »

The NEC code for 2014 is only a few pages long, .....

I believe it's over 900 pages now! More than what I consider a 'few' but I read slowly...lol

Ethan Brush · March 2015

KE5YD wrote: »

I would like to start with the metal ground-mounted panel rack. I know each panel must be grounded using proper hardware, the rack grounded with the panels.
Now, 1st question: Do THESE tie to a ground rod/grounding system at this point?

A ground or pole mounted array needs a grounding electrode. The structure itself may be able to serve as the electrode if it meets the requirements of a concrete encased electrode. Generally you would not bond one of the SYSTEM conductors here, see answer to #4

2nd question: Do I need or should I have a disconnect(one or both) conductors at this point OR can I just use the breakers in the cabin?

You will need a disconnect for the array conductors either outside or inside near where they enter the structure. You are exempt from this if the conductors are in metal raceways or MC cable.

3rd question: For the buried conductors, I know there is direct burial wire with a ground conductor...I figure #10 is marginal for this distance and current.
#8 UF would be good; but more difficult to obtain. Does USE-2 wire have to be in a conduit? If I use 2 individual wires for the positive and
negative; should I also carry a 3rd wire for ground?

The most cost effective wiring method is aluminum XHHW or USE in PVC conduit. USE can be direct buried but I dont recommend it. I generally stay away from UF just because it is very expensive for the ampacity you get out of it since it is sized from the 60 degree temperature column not the 75 degree like most other conductors. THHN/THWN (copper) can be used in conduit also. size your conductors for 156% of the array short circuit current, and check you voltage drop too. Yes run two "line" conductors and an equipment grounding conductor.

4th question: Does NEC require a Ground Fault Protection Device for This install? I believe that IT might meet the Exception?

Generally you need GFP. The exception is for if you have no more than two paralleled strings and all DC source and ouput conductors are located outside the building. Doesnt seem very likely you would meet this unless you had a remote equipment room. Note that whether the array is a grounded or ungrounded system and the connection of the grounded conductor to the grounding electrode system is typically made in/by the charge controller as part of its GFP scheme.

HandyBob · March 2015

Magnum's installation manual does a great job of showing the grounding & neutral bonding details. Yes, most small off grid solar does not meet code. The code is out in left field with many of it's requirements. For instance, fusing of solar panels that have no fault capability. They cannot provide the current needed to blow a fuse even if shorted. It is virtually impossible to meet wire protection requirements when using small inverters that don't have enclosed compartments. Even 12V is supposed to be enclosed.

OK, the following is just my humble opinion. Just like may of the things I say it may not be accepted by many, but it is the brutal truth.

I don't have the time to get into all of the details or have any extended conversations about this, but that one comment about MidNite controllers being top notch....... I've gone through a year and a half of frustration with the Classic 200. I'm not the only one. I've had conversations away from the forum with folks who own them and I put another on a system done for a friend. I've also seen some comments made here that were shouted down. I got shouted down about this myself before. It is time for some truth. Basically, they know how to make MPPT work, but writing computer code seems to be beyond them and foisting products on the market before they have a manual written or sufficient testing behind them is inexcusable. One ridiculous thing required by the code is ground fault and arc fault. It is included in the Classic (great marketing feature) and it nuisance trips. I had to disable both. The fan is noisy. The "Turbo" fan buzzes like a hair clipper... I finally disconnected that. It won't do automatic equalization and once in a while will start an EQ when I push buttons, even though that is turned OFF. The first time that happened I called them and yelled. People world wide set their Classic to do auto EQ and think it is happening, but they are ruining batteries and don't know it. The screen screws up occasionally, doing a split down the middle, requiring a reboot. The time at absorption is supposed to automatically change as needed.... It does not. I've had three that all did the same things and so did the one on my friend's system. My second one puked. Now I have a forth one furnished under warranty. It is supposed to have a quieter variable speed turbo fan but why should I go to the work of installing it? The worst thing about the Classic is the #4 lugs, in an 80 amp product. Morningstar has been using #2 lugs for years. The classic also is missing remote voltage sensing. Morningstar puts that on their cheap Prostar, the Sunsaver MPPT and the Tristar. Morningstar's automatic algorithm works seamlessly. The comparisons could go on & on.

I spent weeks emailing back & forth with the founder of MidNite before buying the thing and I regret believing him. Marketing using flappers and misspelling absorption should have told me.

Does anybody want to buy a new Classic 200 cheap? Send me a PM.

KE5YD · March 2015

Attachment not found. Gosh.......Thanks! for everyone's comments. I have put together a VERY crude sketch of which WE can talk about. I have a question about placement of the GFPD

if needed? About placement of lightning and static protection device(s)? About placement of ferrites or Other RFI protection devices?

Ethan Brush · March 2015

KE5YD wrote: »

Attachment not found. Gosh.......Thanks! for everyone's comments. I have put together a VERY crude sketch of which WE can talk about. I have a question about placement of the GFPD

if needed? About placement of lightning and static protection device(s)? About placement of ferrites or Other RFI protection devices?

A few comments: Note that you will need an EGC with the PV output conductors back to the charge controller in addition to the grounding electrode at the array that you have shown. The GFPD will be built into the CC and will make the system bond for you most likely. That is also where you grounding electrode system will connect in most cases. It does depend on your equipment a bit and a fair amount of off grid equipment is technically not NEC compliant.

Photowhit · March 2015

HandyBob wrote: »

Does anybody want to buy a new Classic 200 cheap? Send me a PM.

What's cheap? sure!

FWIW-I'm a fan of auto equalizing, and though I have Classic lites, I don't chart them, but I can be sure that at least once in the 2+ years it has done an equalizing cycle with out me starting it! I caught it! Surprisingly just a few of us here believe in maintenance/monthly equalizing.

I knew going in that the Classic was a work in progress, and I hope it continues to evolve, We have needed a charge controller that measures charging at the battery/across a shunt for a long time. I promise you I asked over 10 years ago, maybe closer to 15 now....

So what are you using instead? Have heard issues with Xantrex, and much of Outback is from the same designers. Every thing else are stand alone, so no opportunity loads without watching and starting your self or having double relays for peak and drop.

Drop me a PM on the price, though I don't need the 200, I have 3 lites and a screen might be nice, and a backup, assuming I ever install the last array, would be nice.

Ethan Brush · March 2015

HandyBob wrote: »

Magnum's installation manual does a great job of showing the grounding & neutral bonding details. Yes, most small off grid solar does not meet code. The code is out in left field with many of it's requirements. For instance, fusing of solar panels that have no fault capability. They cannot provide the current needed to blow a fuse even if shorted. It is virtually impossible to meet wire protection requirements when using small inverters that don't have enclosed compartments. Even 12V is supposed to be enclosed.

I certainly agree that both the NEC and manufacturers are behind the times when it comes to off gird code compliance. The NEC is getting a little better at addressing off grid installations. There really aren't overcurrent protective device requirements for PV source and output circuits. There is a requirement for series fuses in some cases but that is a valid requirement as they prevent current from several paralleled strings "backfeeding" another string and ruining the modules' bus bar strips. Because PV is current limited, we cant detect and clear faults like we normally do (as you correctly state) so we need some other way to detect and deal with faults. This is why we have GFP. There are many possible ways we could set these fault detection schemes up and every scheme seems to have some disadvantages.

OK, the following is just my humble opinion. Just like may of the things I say it may not be accepted by many, but it is the brutal truth.

I don't have the time to get into all of the details or have any extended conversations about this, but that one comment about MidNite controllers being top notch....... I've gone through a year and a half of frustration with the Classic 200. I'm not the only one. I've had conversations away from the forum with folks who own them and I put another on a system done for a friend. I've also seen some comments made here that were shouted down. I got shouted down about this myself before. It is time for some truth. Basically, they know how to make MPPT work, but writing computer code seems to be beyond them and foisting products on the market before they have a manual written or sufficient testing behind them is inexcusable. One ridiculous thing required by the code is ground fault and arc fault. It is included in the Classic (great marketing feature) and it nuisance trips. I had to disable both. The fan is noisy. The "Turbo" fan buzzes like a hair clipper... I finally disconnected that. It won't do automatic equalization and once in a while will start an EQ when I push buttons, even though that is turned OFF. The first time that happened I called them and yelled. People world wide set their Classic to do auto EQ and think it is happening, but they are ruining batteries and don't know it. The screen screws up occasionally, doing a split down the middle, requiring a reboot. The time at absorption is supposed to automatically change as needed.... It does not. I've had three that all did the same things and so did the one on my friend's system. My second one puked. Now I have a forth one furnished under warranty. It is supposed to have a quieter variable speed turbo fan but why should I go to the work of installing it? The worst thing about the Classic is the #4 lugs, in an 80 amp product. Morningstar has been using #2 lugs for years. The classic also is missing remote voltage sensing. Morningstar puts that on their cheap Prostar, the Sunsaver MPPT and the Tristar. Morningstar's automatic algorithm works seamlessly. The comparisons could go on & on.

I spent weeks emailing back & forth with the founder of MidNite before buying the thing and I regret believing him. Marketing using flappers and misspelling absorption should have told me.

Does anybody want to buy a new Classic 200 cheap? Send me a PM.

I cant comment on midnite products as I do not have any (yet), but I generally hear very good things about them and they seem to be a great company. Im sorry to hear you are not happy with the classic. I do agree with you about cooling fans. I dont know what some (most?) of these manufactureres are thinking with the noisy fans. I remember a discussion about this a while back and someone said, "what are you doing, sleeping with your charge controller?" Actually yes, many of us off griders live in tiny houses and dont have dedicated isolated equipment rooms.

xsnrg · March 2015

Ethan, On the topic of NEC, 2014 in particular, there is the question of the full system shutdown ability to 30v or less within x amount of time... Sorry, don't recall X right this minute. I read through the 2014 NEC around this area, and since it may apply to the OP, can you go into some detail on when it is required and when it is not? Is it any PV wiring entering a dwelling, or only some types, is it building mounted PV arrays only vs Stand alone, etc. I ask because you seem to know your way around the code book.

Thanks

Photowhit · March 2015

In my 5 setups over 20 years, I've only had one where the batteries and charge controller were in the same room as I was and that was on a boat!

It's pretty simple and often aids in the setup to have these located outside. I believe your planning a ground mounted system, I'd suggest parking your charge controller, batteries and inverter near the panels and run the AC to the cabin. This puts the charge controller near the batteries, and gives your longest run high voltage. I live in Missouri, perhaps the only weak link is lower battery capacity in the cold, I haven't found this a problem, though some might.

This is a simple battery box built in the shade of my array, at that time about 1000 watts maybe 10 years ago, the left side is the 'power center' with inverter charge controller and breakers, the right held 4 golf cart batteries originally and held my forklift battery for a year;
Attachment not found.

This is my current setup, power center, the lift battery is to the right;
Attachment not found.

BTW- This inverter carries UL 1743(?) for residential use, but has no facility to run battery cables in conduit, The only way to make it code compliant is to have it totally enclosed in a UL box. I understand it was once used in such a box for an emergency backup for a stair lift chair. I have no codes to worry about, I may well fence the area under the array once I switch to an inverter that could be made code compliant. Thinking I might like to use as a demonstration setup.

BB. · March 2015

KE5YD wrote: »

Attachment not found. Gosh.......Thanks! for everyone's comments. I have put together a VERY crude sketch of which WE can talk about. I have a question about placement of the GFPD

if needed? About placement of lightning and static protection device(s)? About placement of ferrites or Other RFI protection devices?

I would suggest that the two (or more) ground rods be directly connected together. Then only run one green wire from the closest ground rod to your ground bus in the DC panel. If ou have two grounds from two rods a large distance apart--You are risking bringing the high differential voltages into your building--Don't want to do that

I would suggest that most systems do not really recommend DC Ground bonding the solar panel negative lead. They should run directly to the charge controller Vpanel input. The the only time I have seen this sort of grounding is when using a DC GFI breaker assembly (which I do not like).

Normally, if you wish to earth ground the DC power system, you do it at the negative battery bus to the same "master" ground rod.

Double check the DC cable wiring/grounding with your Charge Controller manual.

Placing a surge suppressor in the combiner box or where the solar panel cables enter your power run back to the main building. And/or you can put a suppressor where the DC cables enter the building (or in the main DC panel if it is very close to the building wire entry (again, you don't want to bring lightning into the building).

-Bill

Ethan Brush · March 2015

xsnrg wrote: »

Ethan, On the topic of NEC, 2014 in particular, there is the question of the full system shutdown ability to 30v or less within x amount of time... Sorry, don't recall X right this minute. I read through the 2014 NEC around this area, and since it may apply to the OP, can you go into some detail on when it is required and when it is not? Is it any PV wiring entering a dwelling, or only some types, is it building mounted PV arrays only vs Stand alone, etc. I ask because you seem to know your way around the code book.

Thanks

690.12 certainly has far reaching consequences and is contentious. This has been a problem in states that adopt the new code edition immediately or quickly (many states lag 6 years, 2 cycles, behind) because equipment to accomplish this is only just starting to become available. It is my understanding that some jurisdictions are just plain waiving the 690.12 requirements. Note that micoroinverter systems already meet this requirement.

690.12 applies to any building - not just dwelling units. Basically, the requirement is that there be a means to de-energize all PV system conductors no further than 5 feet from where they enter the building, or 10 feet from the array. So basically if you are on the 2014 code and your AHJ has no ammendments saying otherwise, you need to comply with 690.12. Note that this applies to PV DC conductors - PV source, PV output, PV combiner outputs, etc. Inverter AC output doesnt need anything special as these will be deactivated by the utility service disconnect due to the inverters anti islanding requirements.

As 690.12 evolves, I predict we will see systems meeting this requirement being coordinated with the ground fault protection system. Current GFP schemes are not entirely effective and have blind spots and perhaps more importanly, they dont de-energize a faulted system. If a GFP device could activate the PV rapid shutdown system, it could be much more effective at de-energizing a system with faults.

Photowhit · March 2015

Ethan Brush wrote: »

This has been a problem in states that adopt the new code edition immediately or quickly (many states lag 6 years, 2 cycles, behind) because equipment to accomplish this is only just starting to become available.

You can check here to see what code your state uses, some local cities may adopt codes before states do. I live in Missouri and though this states we don't have a state wide, I believe we are (or at least were) under 2005. Though may counties have no inspections...

Once up and running, you will typically be grandfathered in for future updates.

jcheil · March 2015

I have a Midnight E-Panel, pre-wired with a bunch of additional breakers, 3 shunts and AC/DC surge protectors that I was about to put on eBay.
That would take care of everything and it's to code, etc.
If you are interested send me a message.

Ethan Brush · March 2015

BB. wrote: »

I would suggest that most systems do not really recommend DC Ground bonding the solar panel negative lead. They should run directly to the charge controller Vpanel input. The the only time I have seen this sort of grounding is when using a DC GFI breaker assembly (which I do not like).

Normally, if you wish to earth ground the DC power system, you do it at the negative battery bus to the same "master" ground rod.

Double check the DC cable wiring/grounding with your Charge Controller manual.

-Bill

This is an interesting question/dilemma Bill and gets into a can or worms (I know you know most of this, this is more of a general response). The NEC states that PV systems can be either grounded or ungrounded, with stipulations and requirements for each. If the PV system is to be grounded the system bond location shall be "at any single point on the pv output circuit." Then there is an exception stating that the bond can be made in and by the GFP equipment. For both grid tie and off grid systems, I have never seen the bond made on the PV output circuit and I dont even think there is GFP equipment that would accept that setup. Ground mount arrays with no more that two paralleled strings and all DC isolated from buildings are exempt from the GFP requirements so that is the only set up I can think of where you could actually system bond to the PV output circuit. But regardless of all that, It appears the NEC does not allow bonding the PV system at the battery or anywhere "After" the charge controller. This may be at odds with what some manufacturers say. Then there is the issue of isolation. I think we really need UL standards that state whether a given piece of equipment has a isolated topology or not. Equipment manuals hardly ever tell you, some infer it. This is a big problem IMO. Solar systems of course typically have three "sections": PV, DC (battery), and AC with equipment between them. Whether the equipment is isolated determines the relationship between them and if system grounded is "Transferred" through the equipment. In NEC terms this is called a separately derived system. IF one assumes there is isolation when there isnt, or vice versa, you could have multiple neutral-ground bonds, no system grounding (which isnt necessarily a bad thing but is important to know), faults not clearing, or even a short circuit. An example, and what brought this whole rant up, is the practice of bonding at the battery, where the charge controller is an isolated topology and thinking the PV output is grounded through the charge controller. Again, the PV system not being grounded is necessarily bad, but not knowing is - thinking there is a conductor at earth potential. Also there are some differences in fuse specifics, color coding, and wiring methods between ungrounded and grounded systems.

I think most charge controllers are NOT isolating which means your system grounded transfers from the battery side to the array side (or vice versa but note bonding the battery side is not a code compliant place to make the PV system bond as mentioned above). The only isolated charge controller I know of for sure is the xantrex MPPT 80-600. I assume the morningstar 600V CC is too (I think when you need to change voltage a lot, it is more efficient to do with PWM and transformers, transformers isolate). As far as off gird inverters go, I think they are all isolating.

P.S. I already mentioned that PV system may be grounded or ungrounded but for completeness, 120 volt and 120/240 volt AC systems are required to be grounded, battery system are generally allowed to be grounded or ungrounded.

Ethan Brush · March 2015

Bill, I would like to hear if you have specific knowledge or general thoughts on isolated and non isolated off grid equipment. Also do you know which CC"s (if any) make a system bond and thus operate the array as grounded (I know the connection is typically made through a fuse, but that still makes it a grounded system, until it blows).

BB. · March 2015

Ethan,

Here is a long thread/discussion across two forums about the DC GFI version of grounding:

http://forum.solar-electric.com/forum/solar-electric-power-wind-power-balance-of-system/solar-beginners-corner/9345-system-grounding?t=10650
http://www.midnitesolar.com/Forum/index.php?topic=142.0

You will find a link to a PDF paper I wrote on why DC GFI is dangerous and should never have been suggested as a solution.

Most Solar Charge controllers, the negative Vpanel connection is the same as the negative Vbatt connection. And grounding done on either side of the charge controller will not affect the operation of the charge controller.

HOWEVER, there is at least one charge controller (older version of Schneider/Xantrex XW 60 amp MPPT charge controller???? Don't know about the current version--if different), which uses a current shunt to measure current in the negative PV circuit. So, grounding "incorrectly" can confuse (at least) this controller.

I am not in the solar business--But I don't recall ever seeing an isolated (solar array to battery bank) solar charge controller.

Grounding--A fundamental North American electrical safety theme uses a Grounded "return conductor" for safety.

The return wire never goes above ground (white neutral) and any "Hot" to Neutral or Hot to Ground will pop an over current protection device (fuse/breaker).

With floating power systems--They require a breaker/fuse on both leads. Yes, you "short" one lead to ground, and nothing happens. A second short to ground on the other lead now pops the breaker.

For example, floating AC power in large marine systems do float both Hot Leads, but they measure the float voltage to ensure that one lead is not shorted (i.e., put a 120 VAC filament bulb from H1 to ground and H2 to ground--Both bulbs will glow at low output--If there is a short, on bulb will light 100% and the other will go out--I am sure they have breakers/electronics these days to better detect faults).

Or, if you have an isolated (transformer for example) circuit--You have to have a test plan/jig to test once a day/week/etc. to ensure all is well.

If you choose a ground bonded DC power system (off grid system with battery bank), then, in my opinion, you should make that "hard system ground" at the source of high current--The Battery Bank Negative Connection.

I, personally, would not connect a safety ground to the Vpanel negative lead... The battery can supply 1,000+ amps of current--The solar array can supply 10's to 100's of amps... The wiring is not that heavy and I would not want to have a short on the DC Inverter Leads go through my solar panel wiring....

-Bill

KE5YD · March 2015

Thanks...Bill, Ethan, and others. I have made notes of the changes up to this point that need to be done. Seems Grounding and Ground Fault Protection are BIG
issues to consider and can be controversial. I find it ironic about the theory of "tying ALL ground rods together" versus " don't tie THEM if they are far apart"?
I NEED a better understanding of THIS?

I also find it interesting, of the examples I've seen for small off-grid with battery DC only systems; that the placement of a Ground Fault Device opens ONLY the
panels to controller input conductor? What happens if one has a fault on the battery + lead from the battery to the OCP, which is located between the battery and the controller output?

BB. · March 2015

Tying or breaking connections between ground rods... That is a lightning "only" related question.

The concern is if you put two rods (for example) 100' apart, and there is a lightning strike, the voltage dissipates across the earth... And there will be a votlage difference between the two rods (rod closest to the strike has the highest voltage relative to the other rod).

Originally, I was in the keeping ground rods that are separated by a "large" distance should be separate.

However, I have changed my mind... If I was to copy what the US power companies do, they earth ground the neutral at the power pole, and at the main panel in each house connected to the transformer. And each ground rod basically connected through the Utility Neutral.

Lightning itself will not go more than a few 10's of feet (as I understand) down a 6 AWG cable. So, longer distances of tying ground rods together may not have "that much" of an effect as the lightning/higher frequency effects are pretty much dissipated/blocked by inductance as the wire runs get longer.

So the 6 AWG tie between ground rods really is for the 60 Hz safety grounding (popping a breaker if there is hot to "ground" short somewhere in the system).

-Bill "my 2 cents" B.

Photowhit · March 2015

BB. wrote: »

-Bill "my 2 cents" B.

Worth at least a buck and a quarter!

Making a small Off-grid Cabin-style install that would satisfy NEC?

Comments

Categories