transformer inverter vs. transformerless inverter - worth the savings?

[Sunking]

And if I have misstated anything above, I am sure Sunking will correct me.

Nope you got it. Mike summed it up pretty good and if you use these 3 thoughts gets you going in the right direction.

1. Bond everything together below dirt. This is called your Ground Electrode System. (GES)

2. Bond everything above dirt together. These are called equipment bonding conductors (EBC), They use to be called equipment grounding conductors (EGC) or the third green wire in your 3-conductor power cords.

3. Bond 1 and 2 together at only 1 place. This wire is called Ground Electrode Conductor (GEC)

Use that mental picture and you are off to a good start.

[Jest Waitn]

Think of all ground electrode conductors tied directly together as being "earth". The problem with the auxiliary electrode is that it doesn't tie directly into that "earth", but instead a separate electrode in soil that could provide a current path to that "earth" via the panel frames, rail, inverter ECG connection, and AC panel ground bar. So don't ever float an electrode/rod that doesn't connect directly into the overall GEC interconnected system, with solar or medical equipment, residential or commercial.

with this note, then what about a system with solar array employing a sub-panel from the main box in which the ground wire is not present, only 240ac (3-wire)?

[adoublee]

Think of all ground electrode conductors tied directly together as being "earth". The problem with the auxiliary electrode is that it doesn't tie directly into that "earth", but instead a separate electrode in soil that could provide a current path to that "earth" via the panel frames, rail, inverter ECG connection, and AC panel ground bar. So don't ever float an electrode/rod that doesn't connect directly into the overall GEC interconnected system, with solar or medical equipment, residential or commercial.

[Jest Waitn]

And you are misrepresenting the principle of "single point grounding" which says that you can have multiple electrodes forming a single Ground Electrode System but only bond it to the EGC system (also distributed) at one point. That is what prevents earth currents from flowing in the EGC net. Not having only one electrode.
Any system which connects an isolated ground rod directly to a local EGC does create the risk you describe, but the NEC does not currently prohibit such designs.

The NEC does not require single point grounding (it is a design choice) but the only place (2017) explicitly specifies multiple ground electrode connections to the EGC net is for separate buildings.
Some have expressed reservations about that design too, but it is not as egregious as the "auxiliary electrode" required by the ill fated 690.47(D).

inetdog was paraphrasing what was said in the video. after reading it a 2nd time it seems he is referring to an office park scenario. mike was using a residential example. don't really see how the distributed comment pertains in the residential.

here is the quote from mike in the video: 'bond everything together and then connect to the earth at one point'. he is talking about auxiliary electrodes (an auxiliary electrode may cause equipment failures by providing a path for lightning to travel through equipment) in a residential setting.

so perhaps we are unintentionally 'talking around' each other.

[inetdog]

NO. he said ONE earth ground. period. not 2 or 3 or 'n' grounds. you can have spurious voltage flow from ground rod to ground rod via the path between them in the event of a lightning strike which can then destroy electronics and/or start fires or worse.

And you are misrepresenting the principle of "single point grounding" which says that you can have multiple electrodes forming a single Ground Electrode System but only bond it to the EGC system (also distributed) at one point. That is what prevents earth currents from flowing in the EGC net. Not having only one electrode.
Any system which connects an isolated ground rod directly to a local EGC does create the risk you describe, but the NEC does not currently prohibit such designs.

The NEC does not require single point grounding (it is a design choice) but the only place (2017) explicitly specifies multiple ground electrode connections to the EGC net is for separate buildings.
Some have expressed reservations about that design too, but it is not as egregious as the "auxiliary electrode" required by the ill fated 690.47(D).

[Jest Waitn]

What MH is saying is that if there is a local ground rod driven which is physically (and therefore electrically in terms of induced voltage in local earth from enormous lightning currents) distant from the main ground electrode grid it must be connected via a GEC (wire or structural steel or other excellent conductor) back to the primary GES of the building. Just connecting the panel frames, racking, and array end electrical wiring enclosures to an isolated local electrode, with only an EGC to bond them back to the main building bond grid, will cause high currents to flow in that EGC with equally high offset voltages that can destroy electronics and/or start fires.
It is easier to simply avoid the local earth electrode in the first place.

And if I have misstated anything above, I am sure Sunking will correct me.

NO. he said ONE earth ground. period. not 2 or 3 or 'n' grounds. you can have spurious voltage flow from ground rod to ground rod via the path between them in the event of a lightning strike which can then destroy electronics and/or start fires or worse.

[inetdog]

What MH is saying is that if there is a local ground rod driven which is physically (and therefore electrically in terms of induced voltage in local earth from enormous lightning currents) distant from the main ground electrode grid it must be connected via a GEC (wire or structural steel or other excellent conductor) back to the primary GES of the building. Just connecting the panel frames, racking, and array end electrical wiring enclosures to an isolated local electrode, with only an EGC to bond them back to the main building bond grid, will cause high currents to flow in that EGC with equally high offset voltages that can destroy electronics and/or start fires.
It is easier to simply avoid the local earth electrode in the first place.

And if I have misstated anything above, I am sure Sunking will correct me.

[Jest Waitn]

I know John Wiles quite well. We both sat on NEC Code Making Panels.

Anyway that article has nothing to do with your application. and as John discovered, dangerous. That article is written for Off Grid Battery Systems which means no commercial AC service on the property. John bless his heart is a Phd and Instructor, he has no real world experience. A lot of what John writes has been blasted out of NEC like 690.47.D in this nice 26 minute video by Mike Holt. You will see my smiling face in the video as I am one of his moderators and he will discuss just how dangerous John can be. Mike has never made such a strong statement ever, and is supported by all AHJ's and engineers as most have amended out John's requirements he snuck through the 2014 cycle. This is what happens when you have advocates get involved in real world applications.

Anyway that article has nothing to do with your Grid Tied application.

Here is the video. Listen to Mikes opening statements and OPINION.

ok. wow. let me make certain i understand what he is saying about grounding wrt to a solar array.

the main panel has a bonded ground. the ground rod used for a chassis or frame ground at the pv array, should NOT be connected back to the bonded ground. it should be terminated at the disconnect switch or connection box.

[bcroe]

It is sort of interesting that NEC has finally gotten around to accurately defining your DC system as "functionally grounded", distinct from "solidly grounded" systems that have a true bond to ground. In this new definition, your ground fuse connected DC conductor is no longer allowed to be white... in any previous code version, white (or gray) was required.

Hmm, between combiner and inverters my installer put in 2 pairs of 6 ga for DC, black for + and white for -. Ground
is 8 ga white, wonder how far off the colors are? From the inverters on, grounds are green. All the panel wiring is
10 ga black PV wire, with pairs numbered and polarity indicated at both ends. Bruce Roe

[Sunking]

here is a good article on grounding i just found, and yet he is grounding the -dc?

http://africanenergy.com/new/wp-cont...outh-forty.pdf

I know John Wiles quite well. We both sat on NEC Code Making Panels.

Anyway that article has nothing to do with your application. and as John discovered, dangerous. That article is written for Off Grid Battery Systems which means no commercial AC service on the property. John bless his heart is a Phd and Instructor, he has no real world experience. A lot of what John writes has been blasted out of NEC like 690.47.D in this nice 26 minute video by Mike Holt. You will see my smiling face in the video as I am one of his moderators and he will discuss just how dangerous John can be. Mike has never made such a strong statement ever, and is supported by all AHJ's and engineers as most have amended out John's requirements he snuck through the 2014 cycle. This is what happens when you have advocates get involved in real world applications.

Anyway that article has nothing to do with your Grid Tied application.

Here is the video. Listen to Mikes opening statements and OPINION.

[sensij]

here is a good article on grounding i just found, and yet he is grounding the -dc?

http://africanenergy.com/new/wp-cont...outh-forty.pdf

That article is more than 17 years old. I'm not sure when 690.42 was added to the code, ground requirements have changed a lot over the years, but it contains none of the exceptions suggested in your link.

Here is the 2014 NEC language, emphasis mine. I'm not sure how it could be any clearer.

690.42 Point of' System Grounding Connection.
The dc circuit grounding connection shall be made at any single point on the PV output circuit.

Informational Note: Locating the grounding connection point as close as practicable to the PV source better protects the system from voltage surges due to lightning.

Exception: Systems with a 690.5 ground-fault protection device shall be permitted to have the required grounded conductor-to-ground bond made by the ground fault protection device. This bond, where internal to the ground fault equipment shall not be duplicated with an external connection.

The inverter counts as a 690.5 device. Per the exception shown below, if you found an inverter without integral ground fault protection, you could do the DC grounding at the array. As you've already said, though, the inverter you are considering includes ground fault protection, as do all currently available grid-tie inverters that I'm aware of.

690.5 Ground-Fault Protection. Grounded dc PV arrays
shall be provided with dc ground-fault protection meeting
the requirements of 690.5(A) through (C) to reduce fire
hazards. Ungrounded dc PV arrays shall comply with
690.35.

Exception: Ground-mounted or po!e-mounted PV arrays
with not more than two paralleled source circuits and with
all the source and the output circuits isolated from buildings
shall be permitted vvithout ground fault protection.

[Jest Waitn]

here is a good article on grounding i just found, and yet he is grounding the -dc?

http://africanenergy.com/new/wp-cont...outh-forty.pdf

[Sunking]

i have never used or wired a transformer inverter. it seemed from the language i read that the d/c ground was a 'direct ground' from either the +/- side of the pv array to the ground rod.

That is where you went wrong and got thrown off track. Sinsij explained it fairly well. Kudos and Warm Fuzzies.

[Jest Waitn]

This is what you said:



This is what the article says:



In grid tie systems, you don't ground the DC conductor directly, separately from the AC ground. The DC ground is provided through the inverter, shared with the AC ground, and in 2017 NEC language, is considered "functionally grounded", not "solidly grounded" because there is a fuse in the circuit.

There is more flexibility in how you handle the equipment ground, but that is true regardless of whether you use a transformer based or transformerless inverter.

i have never used or wired a transformer inverter. it seemed from the language i read that the d/c ground was a 'direct ground' from either the +/- side of the pv array to the ground rod.

[sensij]

Lets not confuse the grounding of the PV framework (solid), with the PV DC conductors, which in my system
ground negative through a small fuse in the inverter, in turn monitored by a GFI circuit. Bruce Roe

Yeah, I hesitated to bring it up, but I don't understand why the OP seems to be struggling. Equipment ground (framework ground) is a completely separate topic tfrom the system ground differences between different types of inverters, but could be why he is thinking another ground is required.

It is sort of interesting that NEC has finally gotten around to accurately defining your DC system as "functionally grounded", distinct from "solidly grounded" systems that have a true bond to ground. In this new definition, your ground fuse connected DC conductor is no longer allowed to be white... in any previous code version, white (or gray) was required. By commonizing the wire requirements between transformer based and transformerless inverters, that is one less hurdle to switching between them.