Garage Refrigerator Trips GFCI

The GFCI is detecting current differential between what is going to and coming from the refrigerator above the threshold that would be considered safe for humans.

Refrigerators have many possible leakage paths, especially those with defrost elements. You may be able to disconnect the elements until you find the leakage.

Another issue with GFCI's is the inductive load (compressor) can have an effect on the breaker with repeated tripping.

Does it trip immediately when you plug it in, -when the compressor comes on, -when it defrosts?

We have an upright freezer in the garage with GFCI. It ran beautifully until the defrost cycle started, then it tripped the breaker. Turns out the defrost elements had developed enough leakage (water and current) to trip the breaker. I replaced the elements and it hasn't given any more trouble.

If it trips immediately, you could try disconnecting everything, i.e. fan(s), compressor, icemaker, defrost, etc., and plugging the frig back in. Then reconnect things one at a time until one of them causes the breaker to trip. Of course, you may have to unplug the refrigerator while doing this to avoid shock/electrocution! If you're lucky, it's the defrost or ice maker, which it sounds like you don't need anyway. Or replace the component.

First, a couple questions?

1. Why is there a GFCI in place. Is it building code?
2. Is this your house, do you own it? Then replace the GFCI with a standard breaker, do you think there is a GFCI on the outlet on your inside refrigerator? I think not!
3. It could be that the GFCI is so weak from tripping time and time again that it is worn out? Try replacing it with a newer model?
4. Is the refrigerator pulling too many amps?
5. Have you inspected the "tripping" GFCI? Checked the amperage draw on the refrigerator?
6. Loose wire connections can also cause GFCI issues.

Before you do perform any changes, Please consult your local building codes and maybe even hire a "QUALIFIED" electrician. Just because someone has a contractors license doesn't mean that they know what they are doing.
Check references!

I had a similar issue with an old fluorescent light fixture that would trip my GFCI outlet in the garage.

I solved it by replacing the unneeded GFCI with a standard receptacle and the problem went away.

A finicky GFCI and an old light fixture was the perfect storm. The only reason for the GFCI in the garage was the house was a rental home and at the time it was a requirement that any garage outlets were to be protected.

What a pain in the butt it was dealing with draconian building regulations.

I finally replaced the standard breaker with a GFCI breaker and the landlord never had another problem.

Sometimes I feel like building codes are put into place when people aren't smart enough to not put the frayed extension cord in their mouth!! And then plug it in!

Good luck!

Some answers:

1. All the garage and exterior outlets are on a single circuit coming from the panel mounted GFCI.

2. Yes, my house. The circuit has 5 outlets, three of which are in the garage and 2 outdoors. I cannot make the whole circuit non-GFCI unless I put individual GFCI's on all the other outlets. No, there is no GFCI on my inside (kitchen) refrigerator.

3. Yes, I could replace it with a new GFCI and preferably one that is less sensitive to high frequency "noise" if that is what is the problem.

4. No, it is not.

5. All OK

6. I have not looked for loose wires inside the refrigerator because it seems a common issue for people who try to run a fridge in the garage.

I was wondering if I could do something very simple like plug in a surge protector, or something, in series to prevent RFI or noise. If this sounds stupid then tell me - I really don't know that much about induced currents on GFCI's.

GFIs work as follows...

Under normal conditions, power is delivered to the load on the black wire and returns to the power supply on the white wire. What a GFI measures is that the current flow in the black and the white "match" each other. If the white sees less current than the black wire, current is returning to the source by either the bond conductor (the green wire) or by another means such as the concrete floor or some other conductive path. When the difference in the flows get to 5ma, the GFI trips.

You need to figure out if it is the fridge or the GFI that is faulty. How old is the fridge? If less than 25 years old... I would, slip a rubber mat under the fridge to isolate it from the floor and make sure nothing else is touching the fridge. Go get a cheap 3 wire extension cord. Strip off the outer insulation to expose the white, the black and the green wires inside. Do not cut the insulation on those ones. Borrow a clamp-on amp meter and measure the current in the three wires. If the black and the white match each other and the green shows no current, the fridge is fine and the GFI is more than likely wonky.

You need to watch it over a few hours to catch both the compressor and the defrost heaters running.

Nice description. Considering that 5 outlets are being protected by one GFI then it would only take 1mA of capacitive leakage from each outlet to trip the GFI.

I would buy a contractor pack and put individual GFI's on all outlets that needed to be protected. That way, each one has its' own 5 mA limit.

This would require a wire to be moved on the existing GFI so that the other outlets would be receiving unprotected power. Then, individual GFI's could be used properly.

Thanks...

The capacitive leakage angle gets played up a little to much if you ask me. That current flow, needed to charge the capacative nature of the connected wiring is not a permanent current flow. It happens only when the circuit is energized then falls to 0 once the the circuit is charged so to speak. Modern GFIs take that into consideration and as such, ignore the first x number of cycles.

On the contrary.

Being an alternating current circuit capacitive coupling between wires and everything else is a permanent effect that does produce a measurable although small load on the system.

To get a 5 ma leakage to ground from a common 120 VAC 60 Hz systems only about .11 micro-farads of capacitance is needed which BTW any good working surge suppressor or EMI filter on a device with a three wire power cord systems can easily add 1 - 2 ma or more load to the circuit.

Add in a bunch of wiring tightly coupled to a large grounded metal mass like how any common induction motors windings are placed in the motor core slots and 5 ma leakage to ground by pure capacitive coupling is pretty easy to create.

Agreed... but what is that would bleed the charge off once charged?

What are you figuring is getting charged and holding a charge exactly in a alternating current system that reverses its voltage polarity 60 times a second?

In theory if the power was cut from the line set at exactly the tip of the sine wave there would be around ~170 volts of charge trapped in one of the lines in reference to which ever conductor is the high side and whatever point is the low but given that the plastic insulation on normal residential wiring is not a perfect insulator the small but measurable conductivity of the plastic would quickly bleed off the residual charge.

Be interesting to try and measure this in a lab somehow.

Agreed... but... given the minute current needed to charge the wiring system from a insulation aspect, and therfore from a capacitive standpoint, the current would be ballanced between the hot and the neutral would it not? How would any of this current flow to the bond or to ground?

I agree fully with everything you have said... I just dont think the GFI would see that current as a problem as it would have to between the hot and the neutral, or do you see a current flow between the hot and the bonding system?

Just look at a typical three wire electrical cable that is used on most any American home.

The live line and the common are on opposite sides of the cable with both having their own insulation but the ground line is between them not having any insulation thus making it the line that is going to carry the greatest capacitive coupling effects to the hot line Vs the common line, which is at a near equal potential to the ground line to begin with.

This is how the capacitive coupling effect finds its easiest electrical path around the GFI that is only looking at the current differential between the live and common lines which do not share equal voltage/capacitive coupling effects in reference to the ground lead.

Some interesting reading here.

Yep unfortunately its info from fluke which..... Well lets just say..

"For example, taking a measurement on a 240 V ac circuit with all loads disconnected might result in a value of. 0.02 A (20 mA) leakage. This value represents an insulation impedance of:

240 V / (20 x 10-6) = 12 MΩ. (Ohms Law R=V/I)"

My math says 12K ohms not 12M ohms but hey they're only off by three orders of magnitude so thats about what I would expect from fluke give their quality control nowadays.

I mean its not like they had the formula for figuring out resistance based on known voltage and current values right in front of them or anything. (Oh wait they did..)

I wonder if they had any hand in the designing GFI units?

Anyone reading this... I am wrong here. The leakage current does flow all the time, in direct proportion to the level of the AC voltage available.

TechDir1997,

Thank you for the comprehensive response. I can speak for a lot of contributors in this forum that most replies from a lot of posters are lacking and sometimes actually hinder the process.

• I see why the circuit has a GFCI breaker, you have 2 outlets exposed to weather and water. Keep the circuit GFCI protected.
• You may have answered your own question!

1. Remove the GFCI breaker from the panel and replace with a standard breaker of sufficient amperage.
2. Install GFCI receptacles in the exterior outlets and install weather resistant covers.
3. From the way you have answered the questions, you appear to be fairly experienced in home repair and should be able to perform this task yourself.(I'm not trying to put you down, I don't know your experience).
4. Check all wiring connections downstream of the breaker to make sure they are tight and insulated. Sometimes a loose wire nut is the issue too?
5. Chill the beer and ENJOY!!

Good luck

Bryan

Bryan: thanks for the kind words and your input. Yes, I am fairly handy and can do what you mention. I was just wondering if there was a quick solution...

North of 60: I have another fridge in the garage which I can run on the same circuit and never have a problem, so it is a fridge problem rather than the circuit breaker. I also run power tools and never have a problem with the breaker. Incidentally when I need the tricky fridge, which is only occasionally, I run an extension cord to an indoor outlet but doing so is a bit of a pain.

Thanks to all respondents.

Then if it looks like the GFI is ok... dump the bad fridge. There may be enough leakage going on the give someone a rude awakening.

The two refrigerators while both having starting capacitors on their compressors could have different size capacitors thus creating or lending to the tripping problem.

There is also a high number of faulty, leaking capacitors currently being sold in the marketplace which is more than likely the root cause of the issue.

I would suggest you unplug the refrigerator, remove the capacitor cover and get the DATA off the capacitor then purchase an identical capacitor preferably made in Canada, Brazil, Mexico, or one of the European countries and replace the cap with the new one.

The cost is relatively low and it might well solve the problem. < $50

Reasoning: If the one refrigerator trips a GFCI breaker the odds are high that it will also trip a GFCI outlet so changing to individual GFCI's most likely will not solve your tripping issue.

I have two 22 CU.FT. Side by Side refrigerators in great condition for sale???????

However, I don't think you live anywhere near Las Vegas?

I know, the outlet replacement is a bit labor intensive but that may solve your problem. Permanently!

I have never been a fan of GFCI's except where absolutely needed, like bathrooms, kitchens and pool decks.

My home was built in 1997 and it has not one GFCI, anywhere! I have a feeling that the previous owners (Property Flippers) bought this house to fix and flip and removed the GFCI's in the kitchen and bathrooms because they didn't know how to wire them. They got hit by the Real Estate downturn in 2006 (I bought the house in 2011) so I think they went really cheap partly because they had no more money (I bought the house for HALF of what they paid) and no clue how to fix a house properly. Every aspect of the house they touched or tried to fix, they made worse.

Good luck

I hate those things as well. I understand their value when put in critical areas but other than that they are nothing but a big pain to deal with.

I have yet to ever see any of them work properly let alone reliably with any sort of medium to higher powered electric motor load or inductive/HID lighting based load either.

At work out here in the oil fields we have diesel powered light plants all over the place of which many also server as power sources for other equipment.

I would love to meet the idiot (and severely beat him) that thought that having a GFI device on a genset that by design has no real earth ground plus has a substantial part of its produced power being dump into HID lighting ballasts upstream from the GFI devices would be a good idea.

Sure they protect people very well being almost nothing of any serious power draw can be plugging to one without it tripping.

You said...

"I would love to meet the idiot (and severely beat him) that thought that having a GFI device on a genset that by design has no real earth ground plus has a substantial part of its produced power being dump into HID lighting ballasts upstream from the GFI devices would be a good idea."

Don't get hung up on the lack of an earth ground. The GFI watches for current mismatch between the hot and the return. As such, this automatically includes the bond wire or as you US guys call it the "equipment ground conductor". A GFI plug on a genset does not need a earth ground to protect the connected equipment when current is returned through the bond.

Not hung up on the circuitry concepts but rather annoyed from having way too much hands on experience with light plant gensets and other portable power sources and their GFI's causing more problems in the field than they ever solved.

The #1 problem I have found with the light plants is that most have a normal HID type ballasts for each bulb and those ballasts have a PF/current limiting capacitor in the secondary circuit of which is connected to the live primary circuit in an normal autotransformer configuration of sorts.

The problem is that those capacitors are metal can type and are solidly secured to the metal frame work of the units of which that in itself is connected to the ground /common point of the generators circuitry.

Because of that the metal case of the capacitor acts like a second smaller capacitor connected directly across the live line side and ground giving it a real measurable capacitive coupling value between the ballasts secondary and primary. The problem with the GFI's and that ballast and capacitor setup is that once the bulbs strike an arc and light up they can produce loads of high voltage higher frequency harmonics that will travel through that capacitors lamp side lead and its grounded metal case which can easily produce odd eddy currents on the GFI's through either the power lead and or the common lead imbalancing the system far enough for the GFI to falsely trip if there is any degree of load on it.

The load, the lights, and the generator windings as a whole on the genset and its GFI can easily test at 10's of megohms resistance to ground on a 1000 volt High pot test plus only show microscopic (10's of picofarads) capacitive coupling effects to ground from either and or both the live an common line yet still cause the GFI to trip due to the effects of the lighting load that is not even running through the GFI itself.

In a perfect 120/240 VAC 60 cycles power systems they may work fine given that very few light plants have perfect electrical characteristics let alone ideal perfectly clean electrical loads they make GFI's into huge nuisance devices at are not saving anyone from anything but getting work done.

I certainly cannot argue your past experience. Do you mind it I copy this post to a tech expert from a manufacturer I interact with on a regular basis? He is GFI AFI expert. I am curious to see what he has to say. He may have an easy fix.

Sure. I dont mind. If it helps get GFI's out of na application where they cause more problems than they solve I am happy to share my findings.

I snub you:

Kenmore Refrigerator Washing Machine Snubbers - Dey Parts

All outdoor and garage outlets MUST be protected by GFCI outlet per the USA NEC (national electric code). EXCEPT a refrigerator or freezer outlet in a garage. This outlet must only allow the appliance to be plugged in, so it must be a simplex (not duplex) outlet.

You may have to rewire this outlet if it is protected with a GFCI by using new strung wire to the panel, and using a simplex outlet that only the refrigerator is power by. The NEC also requires outlets used by refrigeration to be the sole circuit fed by the circuit breaker (at least for the kitchen refrigerator).

If you don't reside in the USA, or a state/county/city building authority that requires the NEC, and has not placed override exceptions on this, you may find my post non applicable.

Darn, I don't want to buy another copy.

My local building authority, required me to take a test before they would issue an electrical permit. That was 10 years ago. The test was 100 multiple guess. Most questions were on GFCI and closet lights.

The local authority has many override code additions. One being all wire in the basement ceiling must be in conduit, or require a finish ceiling. There must have been one fire or electrocution from hangers of clothes hanging from them.

Using an extension lead, plug the fridge into another circuit. Does the problem go away, stay with the fridge or stay with the garage circuit? Wise money is on the latter, with rainwater and crud inside one of the outside sockets as the true cause.

Crabtree: there is only the one GFCI circuit at the house. Outlets in other areas that need GFCI's, such as bathrooms, are protected with individual GFCI's.

As mentioned above, the "bad" fridge works fine when plugged into a non-GFCI circuit. Indeed it was the kitchen fridge for many years and was demoted to the garage when we got a new kitchen fridge. I also have a second fridge in the garage that will work fine on the GFCI circuit. Hence my suspicion that it is a fridge issue rather than a circuit issue.

Whether or not this is what you were saying, you have given me an idea: I will run an extension cord from the "bad" fridge to an individual GFCI in a bathroom and see what happens. It it works, then it means that the panel-mounted GFCI for the garage circuit is too sensitive. If it doesn't work, then it is definitely a fridge problem.

Go for it.

One thing that you said, if I read you right. The other fridge in the garage does not give a problem. Are you trying to run both fridges on the same circuit? Is this a 20 amp circuit, or possibly a 15 amp circuit? Could it be that you are just overloading the circuit with both compressors running at the same time, or worse, starting at the same time? Of course you would not see this problem if you plugged the "problem" fridge into another circuit.

Gringogreg: You make an excellent observation about two fridges on the same circuit and to be honest I didn't think about that being the problem.

Running two fridges, at odd times, is the goal but normally just the one runs all the time. Only the "bad" fridge trips the GFCI when plugged in but the "good" fridge never does.

I need to unplug the "good" fridge and try to run the "bad" fridge alone on the garage circuit. I believe the "bad" fridge always trips the GFCI circuit but I need to refresh my memory. I will report back tomorrow.

How did you get in my garage?

Remember, you got shocked and passed out...

Love it!

Here's another: Receptacles (the 5-15 for example) carry a horsepower rating. the intent is to limit the amount of arcing for appliances were the connection/disconnection is by the plug itself under load. It is a gray area for horsepower ratings applied to GFCI outlets. Some machines like a vacuum cleaner, or pipe threader, exceed the horsepower rating by a good margin. Pulling the plug under load, either by accident or intentionally, damages GFCI's

The plot thickens. Plugged both refrigerators into the same duplex socket on the garage circuit (ie has the panel mounted GFCI) last night and by this morning they were still going. I don't have a good explanation for why the GFCI did not trip except that I've never tried this in winter before when the fridges don't do much work and the humidity is very low.

Both fridges were running together for a while but did not start simultaneously. I will monitor what goes on. I thought I had this narrowed down but it is wide open again.

BTW, the GFCI garage circuit is 20 A.

I assume the duplex outlet that you are now using is not the one where the problem fridge was originally plugged into. If this is the case, check the "problem" outlet for a loose connection, cracked body, etc. If in doubt, replace the outlet. Also, check the connections to the outlet feeding the problem outlet and (worse case) the cable feeding the problem outlet. Maybe the cable got nicked by a drywall screw creating a short between the neutral and the ground wires.

I still think that two refrigerators on the same 20 amp circuit is pushing it. What if you want to put an additional load on the circuit.

The length of the wire between the GFI and the outlet can have a substantial effect on the tripping of the GFI. The longer the run after the GFI, the more suseptable it is to false ground false trips. This is particularly true when it is raining heavy outside, lightning, heavy snow, very high humidity and other similar conditions. It happens most often with lightening and heavy rain than the others. Cause is longer length of wiring after the GFI to pick up spurcious charges from outside sourses, especially from the lightning. Fix is to take panel GFI out of the circuit and replace it with a regular circuit breaker and install duplex outlet type GFI's to replace the plain duplex outlets. Length of wiring after GFI's is nil so there is no "antenna" effect to trip the GFI. It worked in my house. All bath rooms ran from one panel GFI. Replaced it with breaker and all outlets with duplex GFI's and no "down stream" outlets from them. No trips in eight years. Use this method to get rid of the refrig trippings. Good Luck, Old Salt