Do GFCI outlets waste power?

[Highbeam]

So being a good boy I was adding GFCIs to protect each outlet in my shop. I'm a bit of an energy conservation nerd and while trying to reduce my standby load at my main panel I noticed a pretty hefty vampire load on the shop circuit that feeds the subpanel. WTF? Everything is off in the shop. So I pulled off the shop subpanel cover and clamped my ammeter on each circuit and with the Leviton GFCI "on" I was burning 0.06 amps or 7 watts. Pop the test button and amperage read zero. Repeat for each outlet circuit.

Are my expensive GFCIs broken? These do have the ultra tiny LED light on them to indicate protection. I have 9 watt LED flood lights in my house that put out 650 lumens. So wasting 7 watts for each GFCI chaps my ****.

What gives?

 

[larry_g]

I'd be chapped to, that is near 2 cents a day.

lg
no neat sig line

 

[Highbeam]

I'd be chapped to, that is near 2 cents a day.

lg
no neat sig line

Each! Adds up to about 30 watts all the time. Then I bark at kids to turn off the lights when they leave a room. The fridge almost uses less than that.

 

[LXCam]

Sooooooooo turn off the breaker Mr. Tightwad.

 

[benjamintmiller]

According to this article, a GFCI contains a small transformer and some transistors. No electronic component is perfect, so it certainly seems possible for there to be some leakage.

Using our power rates, those 4 GFCIs would cost about $25 per year to have. It may be worth replacing the outlets with a GFCI breaker or another brand of outlet to see if the consumption goes down.

 

[larry_g]

Think of it as paying insurance. One hopes they don't need it and all that money is wasted. But we pay it and sometimes we collect. For me after 40+ years of paying I am nowhere near ahead of the game. I've been tickled by electricity in the past and set off a couple of GFCI's so I'll pay the 2 cents without complaint. If you don't want to pay the 2 cents then replace them with a standard outlet.

lg
no neat sig line

 

[Speedy Petey]

If this bothers you that much I'd hate to see what other lengths you go to in your house. I bet it's comical.

 

[Stuart in MN]

If the wiring in the wall is daisy-chained from one receptacle to the next, you don't need to put one in very spot. You only have to put one GFCI in the first spot and the rest will be protected. This will cut down on your initial cost as well as the parasitic electricity costs.

 

[ddawg16]

What kind of amp meter were you using that can measure that low? At that low of a current, I would be using a 1 ohm resistor with one leg and measure the current across it.

I would be very surprised if that was in fact the real current....7 w is a lot of energy.....

Actually....in re-reading your post....I think you have a load down line....you said you hit the test button? All your doing is disconnecting the load. I bet the internal electronics are still alive. Case in point...you have to have power on the line side in order to reset a GFCI. Sure you don't have a wall wort plugged in somewhere? Charger for a battery?

 

[Highbeam]

I'll check to be sure about downstream stuff. We're talking about a line of 8 outlets down a 60 foot long wall with a single GFCI protecting them all. I pigtail for the outlets.

7 watts is a lot of juice I think. I use a Klein clamp on ampmeter.

My main concern is whether this is normal or a hazard. I measure a 34 watt drop at the main panel when I shut off the subpanel breaker and my efergy whole house monitor. Then when I go to the subpanel and use the clamp on meter I can locate the vampire loads pretty easily. Verified by shutting off the GFCI using the reset button. When I press the button, the indicator LED goes out and amperage goes to zero.

I hate waste. Look at how much money is being spent to save 2-3 watts switching from CFL to LED. 34 watts is enough to put a lot of light in my shop. It's a real waste. It's not just money.

 

[wyliesdiesels]

According to this article, a GFCI contains a small transformer and some transistors. No electronic component is perfect, so it certainly seems possible for there to be some leakage.

Using our power rates, those 4 GFCIs would cost about $25 per year to have. It may be worth replacing the outlets with a GFCI breaker or another brand of outlet to see if the consumption goes down.

Incorrect. If a GFCI has leakage, then youve got a problem! Since GFCIs have transformers, the correct term would be 'transformer losses!' Transformers inherently have operating losses because of winding/coil resistance and the resultant heat. But I'd be surprised if the transformer in a GFCI had a loss of 7w! Who knows though..

If i were the OP, i would pull the GFCI out of the j box and measure the 'load side' of the GFCI with a clamp. Then do the same on the 'line side' to figure out if the GFCI is indeed consuming 7 watts! (OCD if u ask me! I think people concerned about energy losses would save more money by cleaning the lint out of the inside of their dryer and vent lines, cleaning gunk out of the central A/C condensor and evaporator coils and sealing leaks such as at junction boxes, windows, doors and vents!)

Another thing to keep in mind is if youre using a TRMS(true RMS) clamp meter, you will see eratic readings sometimes(eg. a small amount of current) because of the HIGH impedance of the inputs on the meter! Non TRMS meters dont exhibit this behavior but also dont give as accurate of readings as a TRMS meter!

 

[where2]

When I stop and add up how many different GFCI's I have around the house, now I see a 50-70W vampire load. Thanks for the heads up.

As others have noted, I might have to be selective about using a GFCI breaker, rather than an outlet in some instances.

Those who think the OP is obsessive about consumption haven't paid enough to the pockets of the POCO yet. I just got sign offs on my 4400W PV array, and sent in the paperwork for my net meter 5 hours later. I will have a net-zero electric bill in an all electric house before I am done... The inspectors asked what I was going to do with the other PV panels I had: I responded "That's phase two".

The OP recently received a "smart meter", and had the POCO "guestimate" his usage for last month... (yeah, I read that other forum too)

 

[LXCam]

I hate waste. Look at how much money is being spent to save 2-3 watts switching from CFL to LED. 34 watts is enough to put a lot of light in my shop. It's a real waste. It's not just money.

Hey I know you, your that dude that does 48.3 mph in the fast lane ain't ya??.

Ya know, being frugal is one thing, but considering 7watts as a huge lose for a device that is there to protect life and limb is padded room territory. Just saying.

K, serious question. What do you pay per kilowatt hour?.

 

[madosta]

I think someone needs to measure this without a current clamp. I am interested.

 

[justsam]

I think someone needs to measure this without a current clamp. I am interested.

I agree! Clamps are great but I do not trust them for such low current draws. Use a decent Fluke true RMS inserted in series to be sure!

 

[Dick in Wisconsin]

If the 7 watts is continuous at the GFCI outlet, wouldn't the outlet get hot to the touch eventually? Those 7 watts are going somewhere. The little LED isn't drawing 7 watts. What about Leviton, do they admit to this device having that much vampire draw?

 

[Keyrick]

I would not rely on the method of using a clamp meter trying to read current flow that is that small. I think it would be prudent to place a dmm amp meter in series with the circuit and see what you have. That would be a more accurate test.

 

[Gary S]

http://forums.mikeholt.com/showthread.php?t=103687&highlight=GFCI+current

Somebody took some measurements here and got similar results.

 

[Spudland_Dave]

Pretty simple fix...remove the flippin GFCI. ESPECIALLY in a shop....in the Kitchen or Bathroom...fine, I'll pay the piper on the small usage as there, but for a garage/shop? Its lose lose...you get to pay 10.00+ per recep, AND get a monthly fee for each one.

 

[nehog]

...
7 watts is a lot of juice I think. I use a Klein clamp on ampmeter.
...

I would not trust a clamp-on meter for very low current draws, unless you are 100% confident of its calibration at that current level.

But I also wonder how many GFCIs you have in your shop? Can't be that many, right--one per circuit is typical.

 

[rlitman]

I would not rely on the method of using a clamp meter trying to read current flow that is that small. I think it would be prudent to place a dmm amp meter in series with the circuit and see what you have. That would be a more accurate test.

I agree that that amount of current is on the low end of the scale for a clamp meter. But I disagree that a DMM in series is a better choice.

First, with a clamp meter, you can always multiply the reading. Wrap the wire 10 times around the clamp, and it will read ten times the current. 0.6A is still low on the scale for a clamp meter, but much less in the range where you truly cannot trust the number (0.06A is almost surely a phantom number), so a clamp meter can still get good readings. But on top of the likelihood that the reading is false because it is in the least accurate part of the scale, if your meter is not "true RMS", you can expect that a device which has a highly harmonic power draw (which this likely does), will display a number that is higher than the actual current.

The biggest issue though is converting that current draw to wattage. I can safely expect that the power factor of this device will be very low. Perhaps as low as 0.3. So if your current reading were in fact correct (which I doubt) you might still only be 2W, not 7W.

What you need to do, is pull one of these GFCI's out of the wall, wire up the line side to a plug, and plug it into a "Kill-a-watt" meter.

http://www.amazon.com/dp/B00009MDBU/?tag=atomicindus08-20

This is the least expensive device you can get your hands on, that actually reports power usage in watts. It has the electronics to measure the current AND power factor, and give you a correct wattage reading that will directly correspond with what you are billed by the power company. A DMM under $600 cannot do this (and no device that simply measures current can).

Run it through one of these meters, and report back to us. I am eager to hear what the real vampire power draw of a GFCI is.

 

[BFBOB]

SEVEN WATTS!!??
Now, think, you old farts. How HOT did those old-time 7W Christmas tree lights get? If that one GFI (remember, the OP said there was just one daisy-chained to 8 downstream outlets) is really consuming 7 Watts, it would be too hot to touch. I'm betting the ammeter is off at low values (maybe auto-ranging and he's misreading it by a factor of 10 or 100?) or there's something else drawing power downstream of the GFI. Doorbell transformer hidden away somewhere? Garage door opener with its WL receiver always powered up? As one suggested, small charger, as for a battery tire inflator?

For a relatively easy sanity check (inserting a series ammeter is a pain), disconnect the downstream outlets and recheck.

 

[ddawg16]

I read the article.....his readings were a factor of 10 smaller....more like 0.7w....big difference....and a lot more believable....

On the clamp on vs series method.....a lot depends on the meter being used. Most clamp on meters are not the most true RMS meter. And, I would suspect that the linearity at the low in would be in question. Yes, you can wrap the wire around it 10 times....but who is really going to do that.

Series is a more accurate way to go using a true RMS meter.

 

[Highbeam]

I am a little nutty, I drive the 60 mpg motorcycle to work when the weather is good but I drive the 15 mpg diesel F350 otherwise. I'm not a tree hugger but an efficiency enthusiast.

I do my own electrical work so I am sure to double check for errors and one check is that there are no leaks. 30+ watts is a leak and my measurement method is the best available for regular guys. I verified the 7 watts for each GFCI by measuring the sum of those losses at the main line. At that point, the wattage adds up to 30+ watts and is becoming much less of a phantom load. As I shut off each GFCI, the load goes down. I disclosed my equipment and methods. Further, my whole house energy monitor picks up the loss as well.

So the load is real. Going beyond theory, is it expected to have a loss from these devices or are my GFCIs broken? Yes, they feel slightly warm to the touch. Maybe the LED is making heat?

I have several outlet circuits. As of last night, I'm up to 5 GFCIs in the shop. Nothing is plugged in. No loads. I own a kill-a-watt of course but the whole house monitor is more user friendly and also picks up 240 loads.

Our power is cheap. It's not about money. It's about waste, safety, and equipment working properly. Are they supposed to leak?

 

[Teken]

I would suggest you find another method to measure this power draw. Either these devices are defective, or the measurements are skewed. I have 12 different types and levels of protection and none of them draw 7 watts.

This is a snap shot of the Leviton Island Self Test GFCI.

It draws down 1 watt with nothing connected. When the self test kicks in which you see the two spikes it jumps to 2 watts for a few moments. I would use a Fluke True DMM to measure the voltage drop.

Teken . . .

 

[Highbeam]

Thanks teken, your posts are what got me interested in whole house monitoring. My standby power loss with no active consumption is 143 watts so 30+ is a large portion.

Do you have any records of a regular bathroom type GFCI outlet?

 

[Teken]

Thanks teken, your posts are what got me interested in whole house monitoring. My standby power loss with no active consumption is 143 watts so 30+ is a large portion.

Do you have any records of a regular bathroom type GFCI outlet?

I am happy that the Energy Monitoring thread got you motivated. 30 watts is not a small amount of standby power to piss away as some would suggest.

I do have a reading for my bathroom and will post it up here:

This unit consumes aprox 0.43 watts. This Leviton Self Test GFCI was purchased more recently than the other one you see above. So, its apparent the power consumption was reduced to about half.

If the one's you have are indeed consuming 7 watts. I would suggest you return them to Leviton and they will replace them for free. Your standby power consuming (idle power) is very close to mine.

So, I can absolutely relate to why this would concern you as to how much is being sucked away when not in use. As the other GJ member has mentioned I would kill the breaker while not in use, until you can get this all sorted out.

Please do keep us all in the loop as this interests me a great deal and to see others doing the same thing is quite inspiring. If we all do our little part there may very well be something left for the kids.

Teken . . .

 

[sands35]

I am a little nutty, I drive the 60 mpg motorcycle to work when the weather is good but I drive the 15 mpg diesel F350 otherwise. I'm not a tree hugger but an efficiency enthusiast.

I do my own electrical work so I am sure to double check for errors and one check is that there are no leaks. 30+ watts is a leak and my measurement method is the best available for regular guys. I verified the 7 watts for each GFCI by measuring the sum of those losses at the main line. At that point, the wattage adds up to 30+ watts and is becoming much less of a phantom load. As I shut off each GFCI, the load goes down. I disclosed my equipment and methods. Further, my whole house energy monitor picks up the loss as well.

So the load is real. Going beyond theory, is it expected to have a loss from these devices or are my GFCIs broken? Yes, they feel slightly warm to the touch. Maybe the LED is making heat?

I have several outlet circuits. As of last night, I'm up to 5 GFCIs in the shop. Nothing is plugged in. No loads. I own a kill-a-watt of course but the whole house monitor is more user friendly and also picks up 240 loads.

Our power is cheap. It's not about money. It's about waste, safety, and equipment working properly. Are they supposed to leak?

Either you have something plugged in someplace you haven't found yet (wall warts?), or you have a pinched line inside a NM cable leaking to ground. I'd isolate the run between outlets that has the issue and try and find a way to replace the wire.

 

[Keyrick]

As an experiment, I tested a single Leviton, 20 amp, GFCI receptacle, a newer one with a LED, using a Fluke 79 DMM in series with the line, set to measures amps. With 126 volts applied, I measured .01 amps. So, yes, current is being drawn, but in my test, it was 1.26 Watts. Power Factor is not an issue as the GF circuitry has no transformer, rather 2, 15k resistors and a bridge rectifier to supply the DC power for the sensing and latch circuit. FWIW.

 

[BFBOB]

I read the article.....his readings were a factor of 10 smaller....more like 0.7w....big difference....and a lot more believable....

.

From Post 1:

"I pulled off the shop subpanel cover and clamped my ammeter on each circuit and with the Leviton GFCI "on" I was burning 0.06 amps or 7 watts. Pop the test button and amperage read zero. Repeat for each outlet circuit. "

 

[Old Engineer]

Are you sure about that reading??

I too own a Klein clamp meter, and it can't read anything under 100 milliamps. The display will just stay at 00.00, until I exceed 100mA, then it will read 00.10.

 

[rlitman]

Power Factor is not an issue as the GF circuitry has no transformer, rather 2, 15k resistors and a bridge rectifier to supply the DC power for the sensing and latch circuit. FWIW.

And what powers the LED? The bridge rectifier will also create harmonics that reduce the PF due to cutoff voltage. I still cannot see these having a PF over 0.8. I would love it if somebody actually measured that.

 

[Keyrick]

It's not a SMPS. It is a simple bridge rectifier with a couple of resistors that power the logic circuit. There are no reactive components in the DC power source. No reactive components, no cosine of the phase angle (PF). The logic circuit simply senses whether or not there is an imbalance of current between the Hot and the Neutral when the receptacle is in use and opens the circuit if there is an imbalance. And just for arguments sake, P=E x I x PF, 125 x .01 x .8 = 1 watt. The DC supply can easily power the LED. This is not a big deal.

 

[nehog]

It's not a SMPS. It is a simple bridge rectifier with a couple of resistors that power the logic circuit. ...

Also, for the record, seven watts in a confined space will generate noticeable heat. Power the suspected GFCI for an hour or so, then SHUT OFF THE POWER and grab the unit in your hand... Does it feel warm or hot? If it feels room temperature then it is not drawing seven watts of power.

 

[ddawg16]

Are you sure about that reading??

I too own a Klein clamp meter, and it can't read anything under 100 milliamps. The display will just stay at 00.00, until I exceed 100mA, then it will read 00.10.

And I'm willing to bet that if you put it on a calibrator.....that reading would not be accurate.

Unless your spending some serious $$ on an instrument, I doubt it's very accurate.

I used to run a calibration lab...so I 'kinda' know what I'm talking about.

On the 'less expensive' measuring instruments, accuracy usually drops off at the upper and lower ends. In fact, they will frequently spec the accuracy of an instrument for 90% of range...so the bottom 5% and upper 5% are way out of spec.

While I agree that looping 10 loops of wire around the clamp on will yield 10x resolution....there are other factors that come into play, hence, I would not put a lot of faith in the measurement.

The only real way to make an accurate measurement is with a precision wire wound 1 ohm resistor...say 1% or better. I personally would measure the resistance with a 4 wire ohm meter and then use that measured value in the calculation. Then, using the same meter, measure the voltage drop across the resistor.

Nice thing about 1 ohm....a 1 volt drop = 1A.

Now, if we know the current is going to be way down there....improve the accuracy by using a 10 ohm resistor.

To be sure there are no issues with power factor, besides using a true RMS meter, I would hook an oscope across the resistor and line and compare current to voltage. waveforms.

Just to be absolutely sure that this is all accurate....I would do the measurements with no load wires hooked up.

 

[Banjorear]

If the wiring in the wall is daisy-chained from one receptacle to the next, you don't need to put one in very spot. You only have to put one GFCI in the first spot and the rest will be protected. This will cut down on your initial cost as well as the parasitic electricity costs.

This.

 

[Highbeam]

If the wiring in the wall is daisy-chained from one receptacle to the next, you don't need to put one in very spot. You only have to put one GFCI in the first spot and the rest will be protected. This will cut down on your initial cost as well as the parasitic electricity costs.

And of course, as I wrote, that is what I did. Each GFCI has at least 8 downstream outlets.

 

[Highbeam]

I have the panel cover off now since my inspection is set for today. I can check again about the reading on my meter. Maybe I'll even take a photo.

I can accept that 0.06 amps is very low and is in the lower accuracy part of the meter's ability but it's not just one. I was reading 30+ watts of consumption when a few GFCIs were "on". Now we're getting up near .3 amps.

Honestly, a lot of this is way over my head. I don't want to waste power and I don't want a fire. I appreciate folks like Keyrick that took time to verify that they had low or no leakage through their GFCI.

 

[porschedude996TT]

Similar subject. What about those Wall chargers for a cellphone or an iPad? My wife leaves her's plugged in all the time whether the charger is plugged into the device or not. The primary transformer winding is powered all the time. How much does that eat?

 

[wyliesdiesels]

Similar subject. What about those Wall chargers for a cellphone or an iPad? My wife leaves her's plugged in all the time whether the charger is plugged into the device or not. The primary transformer winding is powered all the time. How much does that eat?

Depends on the efficientcy rating of the transformer and its impedance rating. Im sure they're not the most efficient, so i bet they waste a good amount!