Floating service panel ground

Question

The main lug panel in my parents garage appears to have a floating ground. This is in Florida.

Earlier today my father observed some sounds from the office. By the time I got there you could smell a fried IC and there was s buzzing sound. I unplugged everything and plugged them back in one by one. I found that the buzzing was coming from a power strip with surge suppressor. It’s LEDs indicated a circuit fault. I suspect one of the MOVs has been fried.

This was three hours before my departing flight so I scrambled to buy some basic diagnostic tools and poke around.

A plug-in AC outlet tester reports an open ground in multiple branch circuits.

The panel in the garage has service wires going to separate neutral and ground bars but I observe no strap bonding these together.

Using a DVM I measure 240V H1-H2, 120V H1-N and H2-N, and what I assume are ghost voltages of 120V and 240V between these conductors and G. I didn’t have a two-wire AC contact tester to definitively rule out these ghost voltages but based on the plug-in tester at the outlets I have an open ground not a hot ground.

Outside the house three insulated wires in buried RMC feed the meter. This is connected to another panel by PVC with a main disconnect switch and single (20A?) breaker that feeds the sprinkler system. This shutoff panel then feeds the main lug panel in the garage. The shutoff panel resisted my limited attempts to open so I don’t know if neutral and ground are strapped together inside.

Earlier today I was making repairs near the service entrance and as a result may have disturbed the grounds. The conduit running into the meter had rusted thru at the soil level. The three insulated service wires were exposed but appeared intact. I dug around this conduit, rigged up a rigid plastic cover to provide some safety from weed trimmers and rodents, surrounded it with gravel and packed dirt. My parents will need to hire an electrician to replace this conduit properly.

There is a solid Cu grounding wire coming out of the shutoff panel and it is bonded to a buried grounding rod. The rod is heavily corroded. Earlier in the day I “cleaned out” what was either dirt or corrosion in the area of the rod and bonding connector and as a result it’s possible I may have lost continuity to the rod.

There was a second thinner Cu wire also bonded to this rod that runs off somewhere, perhaps to the swimming pool pump. What I think is the other end of this smaller grounding wire was recently disconnected when a new pool pump was installed by a pool supply contractor. They told my parents it was not necessary since the pump is connected internally to a grounding wire going back to the branch circuit origin in the garage service panel. I did in fact observe a green grounding wire in the junction box but did not have tools at the time to check continuity.

Earlier today when I was fiddling with the grounding rod this second wire popped loose. It’s connector has essentially rusted off. I have since attempted to bond this wire to the grounding rod using the other connector but I’m not sure I’m getting a good connection.

The service panel is fed from the top by rigid metal conduit. There are four wires coming out: H1 and H2 are bonded at the top of the breakers, N goes to a bus on the left and G goes to a bus on the right. The panel grounding lug is not used but presumably the conduit feeding the panel grounds it. There is a grounding lug connected to the panel body at lower right. This was installed a few years back to feed a new AC circuit (exits upper left with H1/H2 from the 60A 240V breaker top right). The original AC wires are capped off at upper right and are no longer energized.

N and G need to be bonded either in the garage panel or the outside panel. I suspect that they were effectively bonded to the GEC somewhere outside and when I disturbed the grounding rod I somehow broke this connection.

Another clue is that the pool pump breaker used to trip all the time. My parents had a pool technician look at it and they ended up replacing the pump. This was a few weeks back. Still the problem persisted so they called in an electrician. They say the electrician tested the breaker and confirmed it works. I’m unclear how but at some point the pump breaker stopped tripping. Maybe because one of these techs tightened a bad connection. Maybe because there is an poor/open ground to the offending circuit somewhere and so no current to trip the breaker. I did not have time to test voltages in the pool pump panel and my parents don’t know electricity so their reports are fuzzy at best.

I’ve asked my parents not to plug in any of the office surge protectors and to get an electrician to debug this ASAP. Any ideas what’s going on or what to ask my parents to ask the electrician?

Sub-panel in garage is max 200A, 40 breaker poles. Shutoff panel outside has a 200A main breaker and a single 20A branch circuit. Both panels are FPE Stab-Lok. I've added an image of the service entrance to illustrate where there was corrosion at the ground rod and service entrance RMC (both of which were recently disturbed).

Two electricians have opened the shutoff panel so I'm answering Harper's questions based on their observations:

There exists a neutral-ground bond inside the actual main panel, which again is the main breaker box outside. The N-G bonding is corroded and has failed.

Ground extended to the meter, just to ground it. Yes.

The bare wire to your Grounding Electrode System (ground rods) comes in to the main "panel" and is firmly electrically attached to the ground (i.e. steel frame). Proper cable clamp on entry. Probably not. This bonding point inside the shutoff box is corroded.

Bottom line: The N-G bond at the service entrance has failed due to corrosion. The G-Earth bond may have failed previously or once disturbed. Prior to a few weeks ago there may have been a path from garage sub-panel ground to swimming pool pump, to buried grounding rod to earth. This was disconnected when the pool pump was replaced.

Answer 1

Your grounding problem is due to the open N-G bond in the service disconnect

An outlet tester's "open ground" indication is an indication of an open somewhere between the ground contact on the receptacle and the N-G bond in the service equipment, as the light in the tester that corresponds to this is connected from Hot to Ground, and if the bond is intact, that circuit will be completed, even if the grounding electrode system is missing altogether. However, if that Neutral-Ground bond's gone, that'd explain the open ground reading you see.

As a result of that, I suspect the grounding electrode system to the disconnect is not the source of your issue; note that the pool tech was correct in disconnecting the redundant GEC to the pool pump as the pool pump and pool bonding grid are bonded to the mains equipment grounding system through the pool pump circuit's equipment grounding conductor (if nothing else). The grounding electrode should still be examined while you have qualified eyeballs in there, though.

REPLACE THAT PANEL!!!!!!! (and the service disconnect too!!)

The busted RMC for the service is actually not the largest safety concern with your parents' electrical installation, though! That subpanel was made by Federal Pacific Electric, and is an outright fire hazard! Many of those breakers cannot be relied upon to trip if a circuit is overloaded (either due to common-trip jams or simple miscalibration), and the Stab-Lok design is prone to contact faults in the breaker-to-busbar connections as well, which can lead to fires within the panel itself. As a result, the only fix is to replace the entire panel with a new one.

So, the verdict is:

1. Replace the FPE panel with something decent, see below for replacement suggestions
2. Replace the disconnect with something decent since it's also a FPE atop having its guts rusting out on you -- note that it can be a different flavor than the subpanel
3. Have the damaged RMC service conduit permanently repaired or replaced

Panel replacement options

There are two criteria I use when choosing replacement panel types:

1. It needs to be available in >42 slots (42 is a bare minimum slot count for a main panel)
2. It needs to be available in both single and three phase version (this is a sign of a commercial grade panel, not something that's purely builder/residential grade)

This leaves you with a few options, depending on what you can get (i.e. some supply houses are partisans to one manufacturer or another) and whether you can rip out the existing enclosure:

• Square D QO and Eaton CH are both considered "industrial grade" lines, but they are slightly costlier due to using a 3/4" form factor. They do allow you to get a better slot-to-space ratio, which is a factor if you're using an Eaton retrofit kit in the existing enclosure.
• Eaton BR is a midrange line (some would say it's builder-grade with industrial aspirations). It has readily available exotics and a good variety of options, while being inexpensive and widely distributed.
• Siemens (QP) is one of the few "industrial grade" 1" breaker lines out there. It is less costly than a 3/4" form factor panel, while providing basically the same level of performance, and access to >42 slots.
• Last but not least, if you can't rip out the existing panel enclosure at all, Eaton has retrofit interiors available in both the BR and CH product lines that can be used in this application. You're limited on slot count by the existing enclosure, though, so this is kind of a last resort compared to putting in a new panel that will let you take advantage of the code changes over the past 15 years that liberalized panel sizing.

As to the service disconnect, it can be replaced with a type of device called an "enclosed circuit breaker" such as an Eaton ECC series if you reroute the swimming pool circuit to the new subpanel, or with a standard 200A loadcenter of some sort -- in fact, this'd be a perfect application for one of the small feed-through-lug equipped loadcenters Eaton makes, such as a BR816B200RF. That way, you don't need to burn 4 slots on a 200A subfeed lug in the outdoor panel, while the swimming pool circuit can stay routed the way it is. In addition, you may want to have the conduit rerouted so its not trying to drain water that gets in straight into the panel, or at least ensure that the new install has the top conduit fitted correctly so it isn't leaking water into the panel.

Answer 2

I suspect you have a double problem

First, you have a faulty grounding electrode system, i.e. the copper wire coming out of the main breaker/shutoff, and all it connects to. There might also be a problem with the one, proper neutral-ground bond inside that same main shutoff cabinet.

However that wouldn't matter unless you also had a second problem, probably a ground fault of some kind. Any (appreciable) current flow on ground is a ground fault and shout not be there. A clamp ammeter might help find it. You could also fit a GFCI breake-- oh wait. FPE.

---

Your panel, pictured, is a main lug panel, meaning it has lugs where a main breaker is expected to be. This makes it by definition a subpanel. Never bond neutral and ground in a subpanel.

That main 200A shutoff switch is the main service and the correct point for the neutral-ground bond.

You are dealing with a system that is grounded by the metal, which is really ooky-spooky to deal with! "You don't see any grounds" even though the entire thing is enveloped in ground, actually. Good on you, it is the safest system and very correct here.

Here's what I expect to see from a healthy, modern grounding system:

• ground* and neutral separated at the subpanel. Check.

• Two hots, neutral AND GROUND from service to subpanel. I see Rigid conduit which is valid as a grounding path, even outdoors. You report it in good shape. Check.

• The main breaker/main "panel" grounds are in good order - and in Rigid it's super simple - just all the steel bits are secure and not damaged by rust etc. Check.

• There exists a neutral-ground bond inside the actual main panel, which again is the main breaker box outside. Check? ?? ????

• Ground extended to the meter, just to ground it. Check?

• Only 3 wires (no ground) from meter to pole. Check. Your rusted out Rigid conduit is not an electrical defect, but the jagged edges are at risk to damage the insulated wires therein. The far side of the meter is the power company's responsibility, just report it. That would be a good time to replace a meter pan/main breaker, not that you need to.

• The bare wire to your Grounding Electrode System (ground rods) comes in to the main "panel" and is firmly electrically attached to the ground (i.e. steel frame). Proper cable clamp on entry. Check??

• The Grounding Electrode System itself is in good order, with clean cable going to ground rods or other piping to secure a solid reference to earth. It sounds like you have a problem here.

I would shut off the main breaker (for good reason), tear the Grounding Electrode System apart outside the panel, and "go to town" with a wirebrush and clean it all up. You can move the contact point.

If in doubt, replace the split-bolts or pipe clamps which clamp it all together. Avoid big-box for odd bits like this, they'll spend 30 minutes searching and hand you a 3-pack of the wrong thing. Hit a real electrical supply house, they'll have you out the door in 5 minutes at sane price.

If the ground rod is irreparable, inventory what sort of water pipes etc. are reachable, photograph them and come back here and ask what you can use for a grounding electrode.

---

Now, about that panel.

First, thoroughly evaluate the meter, meter pan and main breaker box. Unless they are FPE, Zinsco, rusted out, or damaged, then there's nothing wrong with them and I would not be in any hurry to replace or deprecate them.

I would not spend new money on an outside main breaker as they're expensive. But since you already have one, an outside main breaker is a feature not a bug. It means you can turn that off and fully de-energize the main-lug "main" subpanel inside your garage. That means you can work on it yourself safely. Most people can't do that.

The FPE panel! Kill it with fire before it kills you with fire.

When I saw your top picture, I was like "uh oh". They're not available because the company was run out of business after they cheated like Volkswagen on their UL testing. They were never safe even by any old standard, the busbars are faulty by design so it's not even possible to engineer a safe breaker for that panel.

Some manufacturers make a retrofit kit where you drop in one of their panels (minus the outer steel box) into the outer steel box of an FPE Stablok. This retrofit kit needs to be listed for your exact panel, so get the numbers off it. Huge time saver -- fold back the wires, swap the guts, voila new panel, 1/4 the work.

Otherwise the panel must go, so find a new panel and get to work.

I am terrified of disturbing drywall but nonetheless I would cut out some drywall to make it easier to not damage wires. Drywallers are cheaper than electricians.

I agree with my colleague to get as many spaces as possible. Your panel space is already quite tall, so shouldn't be too bad. I would much rather you DIY a 42-space panel than pay an electrician to put in a 30-space or gasp a 30-circuit 20-space, which they will want to do because they want to score the low bid.

As I'll discuss, avoid double-stuff breakers, but systems with 3/4" tall breakers are good for fitting many breakers in less vertical space. Those are Square D QO and Eaton CH, both industrial-grade panels, which you want, and industrial-grade barely costs any more. Avoid builder-grade lines, which Stablok was.

GE Qline is alright, but avoid their whole concept of 1/2" "double-stuff" breakers. I also like Siemens, which makes clip-on and bolt-down variants, the latter is well-suited for generator subpanels (since back-fed breakers must be bolted down).

Panel size

See how your panel is full? That's Super Bad. That means when your family talks about an on-demand water heater, they get their hopes up and then nope. They talk about a plug-in hybrid car, man comes out to install the level 2 charger and nope. That is a failure of craft, and your goal as a craftsman is never let that happen.

Some people say "just use double-stuff breakers". Nope. Today the vast majority of breakers must be GFCI, AFCI, dual GFCI+AFCI, or whatever they manage to come up with next. Those won't fit in double-stuffs. Design your panel so you never have to use a double-stuff, and still have plenty of extra spaces when you're done. Ideally 50% extra.

That said, there's a "42-space wall". Panels over 42 spaces are insanely priced. Here's a trick others can't do: you can feed 2 panels from the main breaker, or daisy-chain off your subpanel's main lugs to a second subpanel. Just change the lugs to 2-wire lugs, and use the full size 200A wire for both.

You could even install a second subpanel alongside your StabLok and move one circuit at a time over at your leisure. But I would hurry on this one.

Wanna generator?

I personally loathe generators with a scathing passion, for reasons you will realize at 2am. (you can shut yours off, you can't shut off your neighbor's.) However if your family is considering a generator, panel-change time is the time to choose a panel with a built-in generator interlock. In this case I would recommend a subpanel for the gen-operated circuits, either a paralleled sub as discussed, or a "sub-sub-panel" fed off a breaker in the main subpanel.

---

* By "Ground" I refer to the Equipment Safety Ground (which NEC calls Equipment Grounding Conductor). Not to be confused with the Grounding Electrode System (or NEC's bizarre term "grounded conductor" which refers to nothing to do with grounds).

Answer 3

While FPE breakers are infamous, your problem appears to be a grounding/grounded bond or connection. The first step- Call your power company. Many times, even if the problem isn't theirs, they can be helpful and a good source for advice. If your equipment works, why fix it? Namaste