Resistance Heaters and GFCI

[gradstdnt_99]

Looking for some experience with people having issues with resistance heating elements run off of a GFCI breaker.

In my lab I have a number of resistance heating elements hooked up to a 50 amp GFCI breaker. Some are radiant heating elements like the ones found in an electric oven. It appears some heating elements for some reason have developed enough current leakage thir outer housing to trip the GFCI circuit. Some elements have been running fine, other over time appeared to have gotten worse. Some just didn't want to work from the start. Resistance between conductor and outer housing is in the mega ohm range. Apparently still enough current leakage trip the GFCI.

Any insight into why this happens is appreciated.

[Conrad Hoffman]

I don't remember the numbers, but GFIs are very sensitive. There's obviously some film of something on your insulators that's conducting enough current to trip the thing. If you can even read it with a DVM, it's probably too much. Is there any way to clean whatever's between element and ground? Wire brush? Torch? Sandblast? Also, check the line cords and plugs for wear or contamination.

[rjibosh]

Could you describe what a "number of heating elements" hooked up to a 50 amp breaker means. If you have a few 100 watt elements on a 50 amp breaker, you are not very safe. Good thing the GFI is working. If it wasn't those little guys really do not have any interruption protection. (fire hazard)

Rick

[PeteM]

Could be the problem is the heaters cycling on and off through a thermostat; generating pulses??

[SAG 180]

In Queensland Australia, the stove is the one appliance that doesn't have to be hooked up via a GFCI which I assume is an earth leakage device like an earth leakage circuit breaker. The reason is the ceramic potting inside the heater element is porous and moisture can get in and leakage will occur until it gets hot enough to dry out.

[atty]

My guess would be that you are getting leakage to ground when the heating elements distort when heated up. Let 'em cool down, apply a meter, and everything looks fine. Since it's a little difficult to take megger readings when power is applied, I think I would start disconnecting the elements one at a time, and see if I could find the troublemaker.

[JST]

Some materials have differing conductivity when hot vs cold. With ovens that should not be an issue unless contaminates are on the ceramic,.

heaters also tend to be physically large, and may have significant capacitive leakage. That can trip a GFCI also.

Megohm resistances do NOT trip a good GFCI. it usually takes over 3 mA to trip, since the typical UL standards in the US allow up to 3.5 mA leakage.

The GFCI is a useful device, but it is also one initially designed in defiance of various physical laws. It is a wonder that it works at all, let alone that it works reasonably well, given the types of innocuous leakage that occur with certain equipment types. They would work much better with balanced power wiring like US 240V.

I assume that is what yours run from, so it is possible that there really is a fault, or that your heaters are not balanced as well as they might be, or that there is a moisture problem.

Anything with a large EMI filter will tend to trip GFCIs, anything with significant capacitive leakage currents will tend to trip GFCIs.... anything.......

[gradstdnt_99]

Originally Posted by rjibosh

Could you describe what a "number of heating elements" hooked up to a 50 amp breaker means. If you have a few 100 watt elements on a 50 amp breaker, you are not very safe. Good thing the GFI is working. If it wasn't those little guys really do not have any interruption protection. (fire hazard)

Rick

50 amps, 240 volts, that's 12,000 watts right there. Don't know why you think a couple of 100 watt elements would be much of an issue??? Isn't the bulb inside the oven that much??? The large heating element on my electric stove top is rated at 2600 watts.

I am heating up a lab instrument with three resistance heating elements. They come in countless configurations. Currently I'm using tubular heating elements like this.

http://www.instrumart.com/Product.aspx?ProductID=21697

This isn't far from what you find on a stove top and in an electric oven. Have also used a smaller more flexible version called a cable heater that is 12 feet long.

The cable heater worked very well for a time and then for some reason it started tripping the GFCI. Don't really have a good explanation for this. Runs fine if it's not grounded but as soon as the housing touches the ground, it pops the GFCI.

I prefer the added protection the GFCI provides for the environment this is currently used in. There is during part of it's operating cycle water present. Keep in mind water is not the issue here. Everything is bone dry and it's the same elements tripping the GFCI. Would be nice if I was able to "tune" the amount of leakage allowed before the GFCI tripped.

[WHHJR]

Most electric panel/breaker vendors have a 50 MA (leakage current) GFCI for Heat Trace applications. Please note they are for Equipment protection Not People. They are expensive ~$200+ if I recall correctly. I have installed many on HT applications with no problems.

[mark thomas]

Originally Posted by gradstdnt_99

50 amps, 240 volts, that's 12,000 watts right there. Don't know why you think a couple of 100 watt elements would be much of an issue??? Isn't the bulb inside the oven that much??? The large heating element on my electric stove top is rated at 2600 watts.

I think his point was that 50 amps is much higher than the normal expected load (if the load were a couple 100W elements.) In theory, that 100 watt element could suck 49 amps all day (in some failure mode) and not trip breaker, and such a failure mode is certain to be a hazard of one sort or another, probably including fire. As a general rule, breakers and fuses are sized to just above normally expected loads, although this is admittedly a very vague statement and actual code requirements are much more complex.

[The real Leigh]

Originally Posted by Conrad Hoffman

...GFIs are very sensitive.

It's GFCI gentlemen, as in Ground Fault Circuit Interrupter.

NOT GFI.

- Leigh

[The real Leigh]

Originally Posted by JST

The GFCI is a useful device, but it is also one initially designed in defiance of various physical laws. It is a wonder that it works at all, let alone that it works reasonably well, given the types of innocuous leakage that occur with certain equipment types. They would work much better with balanced power wiring like US 240V.

The GFCI simply compares the magnitude of the current in the two supply lines.

If they're unequal, it __assumes__ the missing current is flowing through some undefined leakage path. Not an unreasonable assumption.

Not sure what "physical law" is being violated here, since the GFCI is just measuring current at two points in a series circuit.

- Leigh

[Steve Riley]

I am currently working as a catering engineer and replace a large number of heating elements for various reasons. It is reasonably uncommon to have one which trips a breaker when it heats up, but it does happen, maybe one in a thousand will have that problem.

I assume it is a manufacturing fault.

regards, Steve

[Conrad Hoffman]

Too much coffee Leigh? Both Wikipedia (ok, not the final word as a reference) and other references refer to the things interchangeably as both GFCI and GFI.

FWIW, I also remember having a drying oven that someone had dropped a tiny wire into. It came to rest between the element and a grounded plate. Tripped the GFCI every time! I have an ultrasonic cleaner with a huge line filter to keep noise out of the power line, and that's unusable with a GFCI as well. Still, my guess is there's just a film of dirt on one or more insulators somewhere.

CH

[JST]

Originally Posted by The real Leigh

Not sure what "physical law" is being violated here, since the GFCI is just measuring current at two points in a series circuit.

- Leigh

The one about capacitors passing AC.......... So many things have significant capacitive leakage, which is to some extent "normal"...... so the GFCIs have to be desensitized to where they actually only "sort of" act as people protection for 120V circuits, or 230V overseas.

The GFCI will pass a lot more unbalanced current than it really takes to kill you. It's basically playing the odds, protecting some, and, yes, failing to protect others. Whoever it saves is happy, and the others were dead without it anyway.

The principle is fine, theory says it works, and it DOES work. But it can't be made to work as well as it ought to because of the physical laws that present problems.

With balanced current such as the heaters are probably getting, the capacitive leakage from one 120V half should be balanced by similar leakage from the other. Leakage would act like load current, and if reasonably equal, will be rejected.

The fact that there is STILL tripping of the GFCI suggests that there is either am unbalanced heaters issue, or a genuine problem.

if, for instance, the heater setup varies the number of heaters, and their connection, for different settings, that might cause it. If that adjustment is not balanced, one setting may pop the GFCI due to unequal leakages.

if, for another instance, one heater element in a series/parallel setup is not working, the effective capacitance may be changed, and the formerly balanced leakage is now unbalanced, again potentially popping the GFCI.

[jim rozen]

I have not yet heard the answer to the issue of a thermostat cycling the elements on
and off. If this is happening that WILL trip a ground fault unit.

The next question not asked, is how was the megohm range leakage resistance
actually measured? With a low voltage DVM? Or with a megger?
The megger might be the better instrument.

At what temperatue is the leakage resistance measured? It MUST be measured hot.
Any other measurement won't tell the story.

Run the element to heat it. Disconnect, and put the meter on it while still hot.

You may be suprised.

BTW GFI units don't really measure the current at two points, and compare them,
that's close, but what they do is take one or two turns of both hot and neutral
lines through a ferrite torride. There's a many-turn sense winding on the same
core.

As long as the current in both wires is EQUAL then there is no signal in the sense
widing. The unit does not care what the value of the currents are, they're not
meausured. What IS measured, is the DIFFERENCE between the two. If that gets
bit, the unit trips out.

Jim

[Steve Riley]

" It appears some heating elements for some reason have developed enough current leakage thir outer housing to trip the GFCI circuit. Some elements have been running fine, other over time appeared to have gotten worse. "

"Some" I still think its a manufacturing fault, there should not be a conductive path from the resistive wire to the casing of the element.

Could also be a faulty thermostat, or the insulation on the connecting wires breaking down due to temperature or a faulty gfi.

Millions of heating elements are connected to gfi 's without problems. I agree with replacing the troublesome elements.

regards,
Steve