Grounding Plans + Problems

adas

Aloha I have some transient voltage 19v from the negative of the battery (and negative of the panels) to ground. My shop is an all metal building:

1:The panel frames are directly bolted to the metal roof.
2: The combiner box is isolated
3: the FX80's are isolated
4: The battery box combiner is isolated
5: The inverter is isolated
6: 110v panel is bolted to the metal building and all sorts of conduit is bolted to the building and to the 110V panel.
7: Building is grounded in 2 places.

So in essence the panels use the same ground as the 110v.


What should I do?

Frank

BB.

Re: Grounding Plans + Problems

Ideally, the frames of the solar panels should be floated, and tied using a "star" method to your site's (or local) earth ground. In reality, floating frame ground of the panels from the metal building is probably not practical. Earthing them to the building and tying a ground from the panel frames to the building's earth ground is the best you can do... You will have a parallel path between the metal building and the safety ground wire to earth--but nothing bad will happen (check with AC camp meter--make sure you don't have more than a few amps of circulating current--do this when the arc welders and other power hogs are running).

The other issue is that with any large structure that uses lots of power (and arc welders in your case), you will see different ground and voltage potentials...

In (older?) metal hulled ships, this is specifically addressed by floating both "neutral" and "hot" for your AC distribution... Anytime there is current flow between either of the AC lines and ship ground--that is a fault that turns on/off a couple of lights and is repaired.

For your building... making sure that any energy source is floating with respect to the building's frame (AC plugs/distribution), and that all neutral/safety grounds are "star" connected back to the one "ground window" (basically the one ground point where everything is returned to). But with metal conduit/grounded wire receptacle--this is virtually impossible to do (and probably against NEC).

Otherwise, you will see some current return via the neutral, and some currents return via the building's metal frame work (typically the building has lower resistance than your copper wiring--so any parallel paths will favor the building).

Trying to make sure your arc welders don't energize the building (many people use the metal building/benches has a handy ground return for arc welding--either by accident or on purpose).

Seeing a 19 volt difference between to ground referenced potentials a hundred feet apart is not unusual... I was doing some wiring for my senior project at an old aquarium/theme park... I grabbed an earthed wire at the office building and another one that was grounded a 100' away at a salt water dolphin tank... Boy--was I SHOCKED!... Used my DVM to measure a 70 Volt AC difference between the two... In the end, I could not force anyone to fix the problem (if anyone could ever find a problem to fix)--large AC motors/pumps/lighting on a fill over an old saltwater marsh--would probably never be able to fix.

So--I had to ground reference everything next the the tank--that was were people/animals could be electrocuted, used lots of GFI (Ground Fault Interrupters) for any AC equipment, and made sure that I used isolation between the tank and the computer equipment installed a 100' away. Could achieve similar AC safety by using isolation transformers (more expensive solution).

In the end, if you do not measure large AC currents in your grounds--you are probably fine... My two cents--if I measured ground currents that where in excess of ~10% of the wire/circuit's current rating--then it is worth investigating/fixing.

-Bill

adas

Re: Grounding Plans + Problems

Aloha, is there not some issue that dc and ac shud not share the same ground?
I am grounding the building in another place now and will see if there is a difference.
Should not the Neg be grounded on the fx80 side?

and the (neutral) on the 110v panel be grounded?


BTW it was my new inverter that "found" the problem as the GFI kept tripping and the old Aims inverter did not care.
Frank

BB.

Re: Grounding Plans + Problems

Ah... If only it was so simple...

A little though problem.... Connect an AC welder's output return to once side of the building. And start arc welding on the other end of the building--without the using the return cable... You now have 100's of AC Amperes going through the sheet metal/framing of the building--and because there is some resistance and impedance--there is a voltage differential between the left and right ends of the structure.

Now, run an AC circuit from the left side of the building (distribution panel, neutral bonded at the box and grounded to the building (metal on metal) and ground wire tied to left corner of the building grounding rod. The box, the conduit, the neutral and safety ground are all at "0 volts".

Put in a GFI breaker in the distribution box and send the 120vac ciruit to the right side of the building. Say you are using metal conduit, and 3 wires--hot/neutral/gnd. Put a metal box on the right side of the building and a standard 120 VAC 3 prong receptacle. The hot/neutral are tied to the outlet, the green wire tied to the box/ground plug, and the metal box.

Now, remember you have 100's of amps going through the building so the Right side is "XX AC volts" higher than the left side. Both your conduit and your green wire are spanning a voltage gradient of the building (has to be a voltage gradient, so that current can flow). Now, without anything connected, there is current flowing through the green wire... And, if the GFI plug measures anything more than 0.01 amps or so (don't remember the exact current spec.), the GFI will see that there is current flowing through the safety ground and shut down the circuit because, it believes that a person must be getting shocked from touching the AC hot and earth/green ground (will not protect against a Hot to Neutral person contact).

In this case, the fix is:

1: remove the arc welder current through building frame
2: Install a plastic box with the receptacle to prevent current flow into the green safety ground wire (if allowed by your inspector)--note, a grounded metal drill body will still draw a spark when touched against the building and trip the GFI)--potential is still there.
3: Remove the GFI from the distribution breaker box and install a GFI outlet directly at the point of use. This makes the GFI ground reference local--ignores current to safety return to panel.
4: Install an isolation transform at point of use where ground potentials exist.
5: Move the neutral bonding to the other side of building--no local ground differential (But you may have problems back on the other end of the building if #1 is not fixed)

I can probably think of others too... Some of these address the symptoms, others address the problem. Some are cheap and easy, others are expensive but, possibly "safer"...

Really depends on your specific conditions and needs.

-Bill

n3qik

Re: Grounding Plans + Problems

adas wrote: »

7: Building is grounded in 2 places.

That may be part of the problem. Should have only one ground.