Extreme EMI

[PRR]

> getting the right person ... took it pretty seriously.

Yes. It IS a serious issue. Depending where the current flows, you can have dead dogs, dead water-meter men, or even dead customers. And current leaking before a customer's meter is unprofitable!

NYC had a few cases recently, dogs shocked from stepping on manhole plates. The Jersey Shore had a situation a few years back, a whole neighborhood was getting shocks and it took a long time for the electric company to find the fault.

I have heard of ground leaks so bad the snow melted all around the ground-rod.

> the neighborhood neutral looks bad.

That's a nastier problem than I expected. I don't know much about underground distribution. Do they just bury the standard overhead cable, two insulated wrapped with bare? In our area, that's begging for wire-rot. When the bare neutral rots-out, you have 240V split across whatever balance of loads you have on each 120V leg. Since your 120V loads "never" balance, a difference current flows. It can't return to transformer CT, so it flows to ground and eventual balance with the entire utility system.

> a copper pipe* from my furnace to the AC unit outside

Probably freon, not water; and that really should NOT be a current path. The furnace shell is bonded to fusebox ground. The outside compressor is bonded to fusebox ground. And you don't bond with water-pipe, much less a freon pipe. There should be ground wires (or conduit). So there should be zero voltage between them. And no current in the pipe.

> my AC pops the circuit breaker so frequently

There may be multiple problems.

It's your space, your butt. Be careful, but be curious. If you feel comfortable opening junction boxes and fusebox, look and follow the groundING conductors. These are bare or green wires. Almost every post-1961 circuit "must" have one. You should be able to follow Ground from fusebox all the way to the load. No breaks, no loose or tarnished connections. Conduit and armor-cable connectors are often untrustworthy; I try to specify a wire-ground be run in conduit.

I think grounding has been required for furnaces and 230V loads even longer. The last major "ungrounded" circuit is a legacy electric stove (and maybe dryer?): these used to be allowed "ground through Neutral" because the nominal Neutral current is small. New stove circuits are supposed to be 4-wire H-H-N-G, but an old 3-wire in good shape may stay in service.

 

[shabtek]

re: direct buried power cable. usually the secondary--240vac is a bundle of 3 conductors with a plasticy jacket on each--the primary is another story, and that is where your system neutral/ (preffered) return path to the big stepdown (substation) goes; it was a concentric cable with bare "stranded" copper as the outer layer/neutral, until in the last 20 or so years when jacketed concentric began being used. as you can imagine the bare stuff corrodes and is beginning to come to the end of its useful life, and the jacketed stuff is not impervious to moisture ingress/ thermal cycling.
that leakage current from a bad primary neutral is what is more likely to melt snow/burn wooden structures/cause injury than a bad neutral on a neighboring homes service run.

 

[kafka]

> getting the right person ... took it pretty seriously.
Yes. It IS a serious issue. Depending where the current flows, you can have dead dogs, dead water-meter men, or even dead customers. And current leaking before a customer's meter is unprofitable!

> the neighborhood neutral looks bad.
That's a nastier problem than I expected. I don't know much about underground distribution. Do they just bury the standard overhead cable, two insulated wrapped with bare? In our area, that's begging for wire-rot. When the bare neutral rots-out, you have 240V split across whatever balance of loads you have on each 120V leg. Since your 120V loads "never" balance, a difference current flows. It can't return to transformer CT, so it flows to ground and eventual balance with the entire utility system.

> a copper pipe* from my furnace to the AC unit outside
Probably freon, not water; and that really should NOT be a current path. The furnace shell is bonded to fusebox ground. The outside compressor is bonded to fusebox ground. And you don't bond with water-pipe, much less a freon pipe. There should be ground wires (or conduit). So there should be zero voltage between them. And no current in the pipe.

> my AC pops the circuit breaker so frequently
There may be multiple problems.

It's your space, your butt. Be careful, but be curious. If you feel comfortable opening junction boxes and fusebox, look and follow the groundING conductors.

Yeah, it's the freon; two copper pipes, one about 1" in diameter for expanded gas, the other 1/4" for compressed liquid.  I also checked some other AC units in the neighborhood, and the result on the houses with the neutral problem was the same - net current on the freon pipes and on the power to the AC.  With the temp neutral in place, the current is gone.  I found 0 ohms from my freon line to my cold water.

What I found was that almost no current was going through my water bond, but that there are plenty of other paths from neutral to ground; i.e. the AC, the furnace, the water heater.  I even found a couple of amps of current on my gas lines - although that was entirely inside the house, and none on the external gas line (once again, gone when the temp neutral was put up).  If I wanted to disconnect neutral from the water lines, it wouldn't be as simple as disconnecting the bond, which wasn't really conducting anything anyway.

I think one of the reasons all the other paths were preferred is that the ground bond is aluminum, which is only 65% as conductive as copper.  Given also the relative diameter of the copper pipes, the plumbing paths were just much better than the actual ground bond.  A good safety step would probably be to upgrade the aluminum bond to copper.

Before I called the power company I opened up the service panel to make sure I wasn't asking them to come out and turn a screw for me.  Everything was very tight.  I haven't seen any corrosion anywhere, except a lot of aluminum powder at the bond post the water meter.  That was probably due to the decade or so of current going through it.

The AC could also be blowing the fuse just because it's 25 years old.  It's probably one of the oldest ones in the neighborhood.  It's pretty small compared to what people are putting in these days, but with proper adjustment of the vents, it still works well.

I'll have to take a look at the cable when they do the digging.  I'm curious whether it's just plastic casing, or whether it's in a conduit.  I didn't get a close look, but the neutral bundle looked like it had a jacket as well.

The engineers at the power company were grateful for the analysis, which is why I think I got resolution so quickly.  If the information comes to them piecemeal, they'd have a much harder time putting it together into a resolution.  I think the key was mapping out the neighborhood instead of looking at it entirely from one property.  Being able to point to my neighbor's results, and having one series of houses that looked bad while the surrounding properties looked fine, made it easy to make the case.

 

[Butterylicious]

Wow!  I'm a little late to this thread.  PRR pointed me to it whereas I am an electrical contractor.  You have a very bad and potentially lethal situation.  I'm glad the power company is responding quickly.  Reading 12 amps on your water pipe is very very alarming.

I noticed you said that your main bonding jumper is aluminum.  You'd be wise to replace that.  We all know how important good grounding is in an audio circuit.  It's even more important in premises wiring.  You also mentioned your AC breaker trips frequently as well as your GFCI receptacles.  I'm 99% certain you may have more than 1 grounding problem.  Similar to a ground loop in an amplifier, if your neutrals and grounds have parallel paths, current can (and will) flow down the undesired path.  Here's a quick basic synopsys of how your system should be wired.  I'd suggest you find a "master" electrician to give it a look over.

Your main panel outside is where the earth ground is supposed to be made.  The equipment itself, all ground wires, and the neutral bar should be earthed at this single point.  Regardless if this is a main breaker only or a main panel with several branch circuit breakers, this is the ONLY point where neutrals and grounds should tie common.  Heading downstream to a sub-panel, the neutrals and grounds should be separate.  Often this is not the case.  It gets past a lot of fellas because everything appears to work, but there is a huge difference between something that works and something that works correctly.  Nuisance tripping and current back-flow are what happens.  A classic example I've seen personally was a washer fed off of a mis-grounded sub-panel and a 240V dryer fed from a main panel.  The customer complained about getting shocked between the 2 appliances.  There was 85 volts potential between the 2 units.  I fixed the ground on the sub-panel and the washer breaker tripped.  There was a short in the motor.  Thinking back, actually the sub-panel wasn't grounded at all so this isn't the best example, but my point is really the difference between something that works and something that works correctly.

If you find a good electrician who "gets it" when it comes to audio, he should be willing to oversize your main ground.  The NEC guidelines are minimum requirements and not necessarily design criteria. If you have any questions, feel free to PM me where as I don't spend nearly as much time on this forum as I used to.

-Richard

 

[PRR]

> If I wanted to disconnect neutral from the water lines

Do NOT do that ! ! ! !

Fuse-box neutral "MUST" connect to interior metallic piping.

NEC would also like it connected to exterior metallic piping, since this is frequently the most reliable dirt-ground available.

(At my house, the point is moot: the water company ran plastic from street to porch. The foot of copper stub under the dry porch may conduct enough to electrocute the meter-changer, but not to carry substantal fault current. When I found that out, I also found abandoned well-piping, and bonded to that, as well as the jumper around the water meter to ensure zero potential for meter-change.)

> the ground bond is aluminum, which is only 65% as conductive as copper.

Any metal is thousands of times more conductive than flesh or dirt.

Aluminum itself is a fine conductor. When used to replace copper, you use two gauges bigger, so the conductivity is similar. And there is a special case: an exposed groundING conductor must be #6, no matter what lame thing it bonds to, so that it has mechanical strength to resist damage. My meter-jumper can't ever leak 1 Ampere, but I am required to use a fat #6 which could easily carry 50 Amperes, so when I toss lumber around the cellar it is unlikely to break.

Aluminum DOES have issues with getting good contact. Small Al wires under cheese-head screws tarnish, resist, heat-up, tarnish faster, expand/contract, and you get huge heat in a small spot. This seems to not be an issue with the many-zero cable used for feeder, because the clamps are big and brutal.

> one series of houses that looked bad while the surrounding properties looked fine, made it easy to make the case.

Yes, a good move. Do their job FOR them. One whiny customer is hard to diagnose. A map of 3A 9A 13A 10A 6A 2A 0.1A leakages, and they can pinpoint the fault without leaving the office, order a specific repair instead of a random-search party.

> you may have more than 1 grounding problem.
> I'd suggest you find a "master" electrician to give it a look over.

I'm an experienced but unlicensed wire-bender and untrained theorist, Richard is a practical trained professional who has seen things go very wrong; yet we both come to the same idea.

You probably have more than one problem. The side-effects are often not obvious. Get a Master Electrician. This is not DIY or handyman stuff.

 

[PRR]

> the difference between something that works and something that works correctly.

I think you heard this, but other may learn something.

Bad switch in kitchen. Don't know which breaker it is on. Plug in a radio, loud. Go down cellar, turn a breaker off. Radio plays, turn it on, try the next breaker.

I flipped every breaker and the radio never stopped! (And unlike my father's house, I can see that there is no pre-fusebox tap.)

Well, there's always reverse psychology. I turn ALL the breakers off. The radio DOES quit. So I don't have any magic power source.

Turn breakers on one by one. TWO breakers pass power to the radio!

At this point I know I could sit with pencil/paper and work out what I have. But fact is, it AINT RIGHT, and I don't really care what they did, I need to do it right.

I suppose you've guessed that they connected black wires semi-randomly, not knowing sources from loads. There's a 50:50 chance two sources together will pop the fuse, they persisted and found some way everything "worked".

Multiple dangers here. I had two parallel 20A breakers; it is quite possible they could pass 40A through the circuit, double the safe value.

Actually more like triple: it was all #14 gauge 15A wire.

 

[Butterylicious]

>Do NOT do that ! ! ! !
Fuse-box neutral "MUST" connect to interior metallic piping.

ABSOLUTELY!

There are 2 schools of thought when it comes to water pipe (and now required) gas pipe bonding.  Back in the old days... well as recently as er... 15ish years ago, the cold water pipe was acceptable to use as a means of earthing your electrical system.  Early on it was often the sole means of earthing, later on it was considered a secondary means of earthing, but today it is considered a bond between the electrical system and the water piping.  The rational behind this was that if the water system was broken, plastic, or replaced, you wouldn't lose your earth reference.  Bonding is important so that zero volt at your faucet is zero volt at your coffee maker.  Same holds basically true for your gas piping system.  Of course electricity knows no codes.  All it knows is the shortest path to earth so in all actuality, your water pipe may very well be your primary ground, which I suspect is your case if you're reading current there.

There are a lot of different methods of earthing.  Probably the most common is a ground rod.  I'm somewhat hesitant to discuss this here because I got yelled at on this forum for expressing my opinion but I'm firm in my belief this is not the best method.  Here in Austin, the local authority having jurisdiction has disallowed the use of ground rods because our typical soil does not provide adequate conductivity.  NEC say this is fine as long as you can achieve 25 ohms to ground max.  If not you're supposed to add additional rods no less than 6' apart.  Here in Austin, a 12" square grounding plate buried 30" deep and encased in concrete is the replacement for a ground rod.  The preferred method of earthing here is to attach to the steel rebar within the slab with an approved clamp.  Obviously if the structure is pier & beam, this isn't an option.  On rebuild work where there is a slab, we chip into the edge of the slab and expose a section of rebar, grind it clean and clamp on.  We also supplement with a concrete encase ground plate.  Another method which is a total PITA is to dig a trench 30" deep, 25' long and concrete encase a piece of bare copper sized in accordance with the NEC.

Earthing is probably the most neglected yet IMHO the single most important part of an electrical system.  I've seen it first had where a poorly earthed system don't provide an adequate path and therefore breakers won't trip but instead buzz and arc weld.  After repairs, the same breakers trip instantaneously as they are supposed to. 

Finally as PRR said, this is not a DIY repair.  I'm cool with Joe Geek who plays with vacuum tube circuits adding a plug or a light etc., but you've got something awfully funky going on.  It sounds like you've got studio equipment and such that could easily suffer form a problematic electrical system.  It's hard to put a price tag on something like this, but I'd bet it's less than replacing all your gear..... or a casket.

-Richard

 

[dcharrison]

Could be worse...

I heard tale once of a tech who was having trouble with a home stereo installation.  Whenever he hooked things up and powered on, there was nothing but hum.  After replacing the electronics (twice) he began troubleshooting his cables.  At one point, while attempting to measure continuity in a speaker cable, he mistakenly set his multimeter on volts instead of ohms.  With one end of a wire and one test lead in one hand and the other end of the wire and the other test lead in the other hand, he discovered he could generate measurable voltage out of thin air simply by waving the wire around!

It turned out that this house was in a new neighborhood and when the power company installed the transformer for the house, they tied the hot leg of the unfinished house next door directly to ground.  14400 volts were making a round trip to ground just outside the house.  Apparently, you could actually feel the magnetic field as you approached the transformer. 

Why it didn't set off all sorts of alarms down at the power station is beyond me. 

 

[kafka]

> If I wanted to disconnect neutral from the water lines

Do NOT do that ! ! ! !

Fuse-box neutral "MUST" connect to interior metallic piping.

LOL!  Oh, of course not!  My point was just to illustrate that there are so many other, lower impedance paths between the service panel neutral and the water main that the intended bond was conducting just a small fraction of the current.  It's interesting to me that if I were to double the conductivity of that path, either by changing to copper or decreasing the gauge of the conductor, it still wouldn't be able to compete electricaly with several paths, each of which are better than the intended bond.

 

[joe-electro]

Thanks for posting this. It's one more tool in the troubleshooting toolbox.  I too couldn't believe that number of amps on your water pipes. Excellent troubleshooting on your part. You may have saved a life in the process.

---Joe

 

[chris319]

Kafka -

What make and model of gauss meter did you use?

Thanks.

 

[kafka]

You'll want to get a cheaper single-axis meter for circuit tracing.  I used the Lutron EMF-823, which goes for around $100.  It goes up to 2000µT, which is way overkill (at 8" from the source, that would be 2000 Amps).  The lowest setting, 20µT is plenty.  It was helpful to open it up to see the orientation of the coil.  For this meter, the coil is wound across the box, so if you hold it horizontally and point it to the left, it's the same as if you were holding a guitar.

The Edward Leeper book Silencing the Fields  has a lot of good info on picking and using a meter.

Once I figured out what was happening, I picked up a Greenlee clamp-on ammeter at Home Depot for < $60.  It's just a Gaussmeter with a different label.  By that point I knew I would be talking to my neighbors and the power company, so I figured it was worth a few extra $$ to be able to show the current directly, instead of having to explain the correspondence to magnetism.

 

[chris319]

I picked up a Greenlee clamp-on ammeter at Home Depot for < $60.

Now you've got me wanting to pick one up and go around checking our building which consists of 53 condominium units. With each homeowner able to hire his/her own electrician and do gawd-knows-what, who knows what's been done? The building was built in 1970 and was originally wired with aluminum.

 

[abbey road d enfer]

Here in Austin, a 12" square grounding plate buried 30" deep and encased in concrete is the replacement for a ground rod.

What resistance are you getting with that?

  The preferred method of earthing here is to attach to the steel rebar within the slab with an approved clamp.  Obviously if the structure is pier & beam, this isn't an option.  On rebuild work where there is a slab, we chip into the edge of the slab and expose a section of rebar, grind it clean and clamp on.  We also supplement with a concrete encase ground plate.  Another method which is a total PITA is to dig a trench 30" deep, 25' long and concrete encase a piece of bare copper sized in accordance with the NEC.

All these methods rely on concrete being conductive, right? Doesn't concrete's resistivity vary with time and humidity and whatever? Just curious.
Here in France, we use either rods or mesh, but never enclosed in concrete. When we use mesh, it has to be buried 3ft deep, with a red PVC mesh 1ft above so when the nextbulldozer comes in it is supposed to give some warning.

 

[Butterylicious]

Here in Austin, a 12" square grounding plate buried 30" deep and encased in concrete is the replacement for a ground rod.

What resistance are you getting with that?

  The preferred method of earthing here is to attach to the steel rebar within the slab with an approved clamp.  Obviously if the structure is pier & beam, this isn't an option.  On rebuild work where there is a slab, we chip into the edge of the slab and expose a section of rebar, grind it clean and clamp on.  We also supplement with a concrete encase ground plate.  Another method which is a total PITA is to dig a trench 30" deep, 25' long and concrete encase a piece of bare copper sized in accordance with the NEC.

All these methods rely on concrete being conductive, right? Doesn't concrete's resistivity vary with time and humidity and whatever? Just curious.
Here in France, we use either rods or mesh, but never enclosed in concrete. When we use mesh, it has to be buried 3ft deep, with a red PVC mesh 1ft above so when the nextbulldozer comes in it is supposed to give some warning.

Good questions.  I do not own a split core ground meter but I have rented them in the past.  Typically we simply short a branch circuit and use our best judgment as to if the earthing is sufficient.  You are correct about concrete's resistivity.  I guess we somewhat blindly adhere to the requirements set fourth by the "local authority having jurisdiction."  Your method sounds very good + the warning barrier is is way cool.  We're required to put caution tape 1' above buried conduits.  Our minimum burial depth is 30" with a 2" concrete cap dyed red.  They're big on concrete here.  I have always been interested in European and other regions of the world electrical requirements.  I've never been to Europe, but I've been to Mexico and saw electrical that was nothing short of frightening.

-Richard

 

[PRR]

> different methods of earthing.  Probably the most common is a ground rod.

Which is generally pretty darn poor.

> in Austin, the local authority having jurisdiction has disallowed the use of ground rods because our typical soil does not provide adequate conductivity.

"CAN'T" have a dirt-rod? Or they count it like a ground-clamp on PVC pipe (useless)?

In the early 20th century, the PRIME ground was the metallic water piping. As you said, you want the toaster and the faucet on the same reference. And the miles of usually-metal pipe deep underground (and often seeping) was a far better earth electrode than anything you would build.

That concept has got totally messed-up. Most folks have interior plastic pipe. That may actually make a brass faucet "non conductive", because tap-water conductivity is low compared to flesh. But I don't say this is safe in all cases. And a strict interpretation of NEC "could" require a bonding wire from metal faucet to service panel (never heard of that being done). Moreover most new exterior pipe is plastic. So we may have safer sinks, but we have lost the historically reliable good ground "free" with our water-supply.

Long-ago, a good water-supply was sufficient; if not available, you drove a dirt-rod to spec. Problems: in hard soil you want to cut the rod short rather than work so hard. Also, even a full-spec rod is hardly 100 ohms in damp season, and much higher in some soils and in dry areas.

There was a time when you were supposed to drive one rod, measure its resistance, and if not low enough, drive a second rod. You were not required to drive additional rods to reach the specified resistance. If two didn't do it, then clearly the dirt is so non-conductive that risk of dirt-shock is somewhat low, and it would be excessively costly to keep driving rods to meet a spec. One and maybe a second, but that's all. Sounds like in your town, even two is not good enough to count.

I think recent NEC requires several things "if available". And large concrete with metal IS an excellent "dirt electrode". Concrete is not a great conductor but neither is most dirt. Concrete is generally better than dirt, at least for the first decade, at least for the mixes we use in the US. The key thing with dirt is Large Area. Concrete is the cheapest way to cover a lot of dirt, and is usually already covering the cellar floor, or at least the foundation perimeter. Assuming it has any kind of metal to bring current together, it is an excellent dirt-electrode. So if you don't take it, a careful Inspector is going to ask "Why?"

Where rock is shallow, there is a spec for laying a rod in a trench. But shallow trenches tend to go dry and non-conducty between rains. And any metal rod has low surface area, does not contact a lot of dirt, is prone to high resistance.

Deep trench, concrete with rebar would be good where there is no alternate. Awful expensive by usual domestic standards.

Naked mesh works, but tends to corrode. Corrosion is a problem for all metal ground electrodes. (Well, not a solid gold rod....) Concrete-encased steel has a good record of low corrosion. Still a problem making a connection which will never rust-off.

When you get serious you bury money. AM radio towers have more metal underground than up to the sky, so that the bottom of the wave has a solid ground to push from, and the wave will tend to flow out over the land instead of up to the sky. Power stations have a dense grid of buried electrodes, partly to reference the power, but mostly to dump lighting-strikes. You see a 100-foot gravel lot with a couple 10-foot transformers and towers; all that "empty" land is kissing the earth as good as possible. In some areas they salt the soil to increase conductivity. And of course replace electrodes when they corrode. Stuff a homeowner won't do.

 

[shabtek]

on the part of the electrical grid I maintain, the construction specs "call for" 32' deep ground rod at every pole with a xfmr or other equipment (ie capacitor, regulator. automated switch, surge arrestor, etc) --every 3-5th pole...the planet does not always allow that much penetration. I don't know what the local code specs for penetration/resistance at a building

 

[PRR]

> I don't know what the local code specs for penetration/resistance at a building

per NEC 1996, paraphrased (do not rely on my words):

Rod 8 feet in contact with the soil. If you hit rock bottom, you may drive at 45 deg, or bury in a trench 2.5 feet deep. Acceptable size and material are given.

A plate electrode is 2 square feet.

If one rod/plate gives >25 ohms, you "shall" drive another >6 feet away.

Grounding electrode conductor may not be sliced; there is an exception for irreversable "listed" connectors and for thermite(!) welding.

Metal underground water pipe must have 10 feet in dirt or it does not count. (Perhaps related to Buttery's local authority declining to count dirt-rods in that Austin soil.) A metal underground water pipe alone must be supplemented with another electrode.

A concrete encased electrode must be at least 20 feet of 1/2" rebar or #4 wire near the bottom of a concrete foundation (no minimum depth or area).

A ground ring is >20 feet #2 wire encircling the building, >2.5 feet deep.

Lightning rod dirt-rods must be separate from power-entrance dirt-rods, but may have to be bonded to each other (this is a whole sub-set of canned worm).

I'm a mere hobbyist homeowner; I don't know how far NEC reaches. It does cover business and industrial services, apparently without limit; it is a little frightening to think that a huge factory "could" have just two dirt-rods in dry sand.

 

[emrr]

Nice brief primer on system grounding in January issue of Broadcast Engineering.  Addresses some legal aspects of grounding rods. 

 

[kafka]

OK, and here's the epilogue to this story - the cosmic "everything happens for a purpose" part, if that's the kind of thinking that appeals to you.

A week ago, I got a letter from the water company saying they had been billing us using estimated readings because the outside extension meter hasn't worked for more than a year.  Furthermore, because of the age of the house, they wanted to replace the main water meter in the basement.  So, if I hadn't taken care of this, that would have been 12 amps through the nice lady who came out to replace it this morning.