Grounding Inverters NEC

Grounding a receptacle to the Earth is useless at best and dangerous at worst. This is what I talked about in the OP. There's different kinds of grounding, and everyone is confused because of it. GFCIs don't need a ground to function. They work perfectly fine on an ungrounded system. The test button doesn't look for a ground. It sends a little current around the sensor instead of through it, which looks like a fault to the circuitry. Dirty power can trip a GFCI, and the high frequency noise on the output of a HF inverter certainly could do it.

Here's the bottom line: If you have a transformer sending power into your building, the neutral of that transformer should be bonded to the grounding system of the building. Where that bond happens at is variable, depending on where the transformer connects and how the utility service, if any, is connected.

Late comment, but still worth it.

Don't assume you can take the designated neutral or arbitrarily decide to call one side of a floating supply neutral and tie it to the ground pin/lug/wire with high frequency inverters.  There are designs out there where the supposed earth wire is actually at half AC supply voltage WRT to the two current carrying wires (single phase supply).  If you join one of those wires to the earth pin you run the risk of damaging / writing off the inverter as you'll effectively be shorting one 'side' out.  Some inverters will detect the problem and not destroy themselves though.

If you are wrestling with proper grounding of a high frequency inverter and the manual doesn't explicitly state what to do you can test reasonably safely.  Place a low wattage light bulb (rated for your mains supply voltage) across a current carrying wire and earth.  If it lights, even dimly, you can not join those two wires.  Repeat the test for the other current carrying wire.  If in either case the light does not glow at all, measure the voltage across the light.  It should be well under 5% of your supply voltage if it is just leakage via noise suppression capacitors (or any other number of leakage points, high frequency inverters leak like a sieve).

You can not measure for voltage between the current carrying wires and earth with a digital meter, or a FET based analogue meter either, as the meter will not load any leakage current sufficiently and indicate voltage even though there is only microamps of current available.

If you are attempting to wire a RCD/GFCI to an inverter where the earth is not safe to bond you may find your RCD/GFCI will still function normally if you stake the nominated earth point on the inverter to ground as sufficient current can still flow between the two current carrying wires and that 'live earth'.  Of course a proper test of the installation needs to be done to confirm that current shunted from the two current carrying wires to earth will actually cause the RCD to trip.

This issue affects both cheapie Chinese inverters, assuming they even have earth pin connected which is no guarantee, and name brands alike including some respected brands in the USA in case you were thinking you were safe.  The big difference there is the respected brands spell it out clearly in the manual that 'thou shalt not connect these together'.

In my humble opinion, and of course there are exceptions, high frequency inverters should be left in the automotive field plugged into a cigarette lighter. In that case, grounding isn't really important. When you start connecting them to buildings, the game changes and the risk goes up. A low frequency inverter with a transformer output is the "right" way to power a non-mobile structure.

Each to his own but all the big boys of the industry have high frequency inverters, even Victron, intended to be wall mounted for premises wiring.  Portable HF inverters in vehicles still need to be wired to the vehicle's chassis for safety.  The AC output in most is isolated so you can't get a shock by coming between one of the current carrying conductors and the vehicle chassis if all is well.  There is still a risk of an AC line cord becoming pinched, say in a door, and one of the current carrying conductors being shorted to the vehicle chassis and then you have a dangerous situation..  There is also the risk of a HVDC fault in the inverter placing HVDC on the inverter's chassis WRT the vehicle's chassis and if you get across that you are just as dead.

A portable HF inverter can be used in a premises perfectly safely provided it is used as intended.  Take a look at a typical portable HF inverter (having holes for screwing it down doesn't make it not portable), it has sockets.  That's a big clue as to how it should be used.  In order to one-up the competition cheap inverter manufacturers have often just bodged on a terminal strip suggesting the unit can be hard wired to a normal premises installation and that just isn't the case for some of them as the neutral/earth bond in the premises will cause problems for the inverter.

Portable inverters often fall into the IT class of wiring,  I = All live parts isolated from earth or one point connected to earth through an impedance, T = Direct electrical connection of the exposed-conductive-parts to earthing independently of the earthing of any point of the power system.  If the output remains isolated there is no more risk of AC shock than in any premises wiring system where an RCD (GFCI) is in use.  The trick is maintaining that isolation.  If you start running extension leads around the place for convenience it becomes much harder to guarantee those leads won't be nicked and cause the AC output of the inverter to be referenced to the earth so people want extra protection of an RCD.

There used to be a design of portable inverter where there was no isolation between the DC and AC side.  Instead of transformer(s) being used in the HVDC stage they used buck converters.  Insane I know, but they did exist.  If the AC output ever became tied to ground there would be lethal voltage between the battery terminals and the earth.  Likewise any fault that caused the AC supply to appear on an appliance chassis would put lethal voltage between the appliance and battery terminals.  From what I've seen it's not possible to buy them any more, most countries have banned them but I'm sure someone in China still makes them and the responsibility for safety with direct import falls on the purchaser.

Grid tied solar inverters often use buck/boost transformerless designs which is why earthing and fusing properly with grid tied inverters is essential.

My own portable HF inverter is a fully isolated design, sockets only no terminal strip, with the inverter chassis earth being used for RFI suppression (safety caps tie active/neutral to chassis) and HVDC protection.  It's permissible to join neutral and the inverter's chassis ground and stake that to the ground to allow an RCD to function.  Like your intent though, I'd not be using this unit to power a premises fixed wiring system.  That's where a proper wall mounted fixed wiring only (no sockets) HF inverters should be used.

HF is better

LF is better

My mopar can trash your chevy

The point to remember here is that those of us using powerjack or powerjack style LF inverters to power our houses need to understand that when we connect the inverter output to the breaker box, we are connecting the center tap of the transformer output directly to ground.  (assuming USA wiring)  We have accepted responsibility.  We can no longer blame problems on the power company.

Running your residence with an inverter is not something to be undertaken lightly.  Potentially lethal forces are being manipulated.  Obviously from the discussions above there are differences of opinion on how best to connect things.  Each of us needs to be very aware of exactly what we are connecting, and how to remain safe while doing so.  The rest is preference and detail.  (and the devil is in the details)

I come away with all this back & forth not sure if there's any benefit grounding the inverter chassis. I haven't grounded my existing system & unless I'm told otherwise by Genetry staff, I won't ground my new inverter they're sending shortly. I will say this tho, I have my inverters 120 neutral tied directly to my grid neutral. When I switch between grid & utility power, I'm just swapping the hot sides & haven't had any trouble with that set up so far after a year in use.

There is no back and forth really. Personally, I have only been speaking about what the electrical code says. As an installer and troubleshooter, I've seen enough diverse installations to know that what the code says makes sense, at least in the area of system grounding. The benefit to grounding your inverter case is that it will save you from being electrocuted should a wiring fault occur that energizes the case. It may never happen. Or it might happen to someone else in the house. Ultimately, it's your inverter and wiring to do with as you please, but know that over 100 years of electrification has proven the benefits of grounding and bonding.

I personally am not a fan of grounding, due to a large amount of equipment damage as a result of "properly grounded" systems.  (Thanks a lot, local lightning strike.)  Up to you; the chassis really doesn't connect to anything internally anyways.  (A few decoupling caps to both the battery input and AC output, that's it.)

You know, there's a function in the GS inverters for that: ATS 😉.  I probably will need to add some features to the inverter to fully take advantage of this function, but that's what customer feedback is for...

While I don't particularly like the idea of connecting an off-grid inverter to the grid (potential damage due to lightning strikes), that's up to you.

It's another case of misunderstanding between earth ground and equipment ground. The ground that would bring a bolt of lightning into your internal wiring is a different ground than the ground that clears a fault in the supply. They are tied together but are not the same thing. The code for a "proper" grounding system is a minimum. It allows for a single electrode connection to the Earth. But the reality is, a bolt of lightning can induce thousands of volts of potential difference over a very small distance. There isn't much that you can do with that on a utility-supplied system because it's impossible to isolate it completely and the service neutral is tied to earth on both ends, so there's a parallel resistance with another uneven voltage drop. And your neutral is tied to your neighbor's neutral. And your copper water pipe is tied to your service neutral, which is tied to every other neutral on the block.... And on it goes. But an off grid system is different. 

For example, I worked on a project where a remote geophysical monitoring station was blowing through electronics almost monthly. The powers that be decided they needed a bigger supply, so they put 1 kW of PV on it. It got worse. It was always arcing damage in the electronics but no visible sign of lightning striking the outside. The problem got worse due to the minimal grounding. The system was grounded with a 4 foot rod in the dirt to the negative battery terminal. The new PV array was also grounded on the negative to its steel pole so distance from the shack. Even nearby strikes can cause enough current to flow in that loop to do damage.

After getting the funding, the solution that worked was to run "spokes" radially from the equipment shack of 2/0 bare copper. They were Cad Welded to a ring of copper around the shack. Another ring encircled the other end of the spokes. It would look like a bicycle wheel if viewed from above. The radius was something like 15 yards. Ground rods were driven at the middle and each end of each spoke. The rods were the worst part. We had to haul everything in on 4 wheelers, but that still wasn't as bad as driving the rods. Anyway, it's been at least two years since any unexplained blow ups have occured. With earth grounding it's all about creating an equipotential zone where the local impedance is so low that no significant potential can exist between distant points.

And I stress that grounding metal cases of appliances is a different animal. Completely different. Just named the same🤬

RFI is something to be considered too.  Completely floating systems can radiate crap like crazy unless a heap of time and effort has been put into containing it.  Ask any amateur (ham) radio operator and I'm sure they'll talk your ears off about that switching power supply / inverter / thing that caused them no end of grief.

When it comes to inverters it's a case of horse for course.  Sometimes a LF (big transformer) type is appropriate, sometimes a HF (wee transformer) type is.  I use which ever suits the application.