$2.70/yr saved! Redundant GFCI removal

[TradEddie]

On another thread I mentioned that our home inspector identified that the bathroom outlets had redundant GFCIs. All three outlets had individual GFCIs, but all went through the same GFCI breaker too. He recommended that we remove the redundancy, so we got a small discount on the house price to pay for an electrician to fix it.

Ten years later I finally got around to removing those local GFCIs so I wired one up to the Kill-a-watt to see if the phantom draw was as bad as some websites suggested. The GFCI draws 0.8W with an open circuit, by removing all three I saved a whopping 2.4W, 21kwh or $2.70 a year. Now all I have to do is figure where the remaining 9,979kwh are going...

TE

 

[lilbertie]

On another thread I mentioned that our home inspector identified that the bathroom outlets had redundant GFCIs. All three outlets had individual GFCIs, but all went through the same GFCI breaker too. He recommended that we remove the redundancy, so we got a small discount on the house price to pay for an electrician to fix it.

Ten years later I finally got around to removing those local GFCIs so I wired one up to the Kill-a-watt to see if the phantom draw was as bad as some websites suggested. The GFCI draws 0.8W with an open circuit, by removing all three I saved a whopping 2.4W, 21kwh or $2.70 a year. Now all I have to do is figure where the remaining 9,979kwh are going...

TE

They are in my garage shop too. 4 of them because they are on individual circuits.

 

[Highbeam]

My meter measured 6 watts on each gfci when I was chasing phantoms. I went around and pressed the test button to shut off the ones I didn't need. Waaaay different than 0.8 watts.

 

[Highbeam]

Wait a sec, how did you use a kill-a-watt to measure phantom load of an outlet? I used a clamp on Klein ammeter on the circuit wire right next to the breaker.

 

[TradEddie]

I wired it up to an old lamp cable with plug. I've no idea of the low end accuracy of the kill-a-watt, so take that 0.8W with a grain of salt. The one I measured didn't have an LED, maybe those use a little more. I did notice that the GFCI was slightly warm to the touch when I removed it, so they certainly consume some electricity. On an unrelated job, I was opening the breaker box yesterday and I felt the GFCI breaker, which was not noticeably warm.

TE

 

[Highbeam]

I wired it up to an old lamp cable with plug. I've no idea of the low end accuracy of the kill-a-watt, so take that 0.8W with a grain of salt. The one I measured didn't have an LED, maybe those use a little more. I did notice that the GFCI was slightly warm to the touch when I removed it, so they certainly consume some electricity. On an unrelated job, I was opening the breaker box yesterday and I felt the GFCI breaker, which was not noticeably warm.

TE

I got an IR temp gun for Xmas and shooting a GFCI outlet shows that heat is being generated, shooting a regular outlet shows no heat generated. They are phantom loads but you need some.

 

[TradEddie]

I got an IR temp gun for Xmas and shooting a GFCI outlet shows that heat is being generated, shooting a regular outlet shows no heat generated. They are phantom loads but you need some.

I'm okay with the phantom loads in a GFCI, but it would be nice if manufacturers would declare that load so that we could consider it when purchasing. In my case I had three GFCIs that were not needed and may even have been counterproductive. I would never replace a GFCI where one was required just to save a few dollars a year.

TE

 

[Highbeam]

I'm okay with the phantom loads in a GFCI, but it would be nice if manufacturers would declare that load so that we could consider it when purchasing. In my case I had three GFCIs that were not needed and may even have been counterproductive. I would never replace a GFCI where one was required just to save a few dollars a year.

TE

When buying new gfci outlets i would be interested in choosing the one with a one watt phantom load over one with a six watt phantom load.

 

[Where2]

I've got a few different GFCI units kicking around. I'll have to see whether they have measurable differences in current draw...
Any idea what brand and model the one tested was?

 

[woodgeek]

I just spent 10 idle minutes googling this issue to find the solution....and came up dry. The answer does not appear to be known to the internet. And we can figure all the GFI outlets in the US are burning through a gigawatt of standby load....

Bottom line: US Energy Star program has not targeted GFCI outlets, so it appears no manufacturer is going to stick their neck out and list their draw. 99% of the earnest greenie kids that have a kill-a-watt and a blog are not going to pull the outlet out and measure the draw as Highbeam has.

So, anyone here want to get some pitchforks and torches, um, or an internet petition to get Energy Star to start pulling down that GW phantom load monster?

 

[woodgeek]

Ok, lets assume that the average home has 5 of these and they draw at 2W standby, that's 10W per home * 100M homes in the US = 1 GW. Times 8600 hrs/ year, that is 8.6 TWh of electricity a year. Since in the US, we generate ~1.2 lbs CO2/kWh, this is 8.6 Billion kWh/yr * 1.2 lbs/kWh = 10 Billions lbs CO2, or 5 million tons CO2/year.

For another point of comparison, since the average capacity factor of solar PV in the US is ~20%, the GFI outlets are using as much power as 5 GW of PV panels. This is equal to all the solar power installed in the US before roughly the end of 2012.

So the cost/electricity/climate impact of GFI outlets is nearly equal to the benefits of all solar power in the US .

 

[TradEddie]

So the cost/electricity/climate impact of GFI outlets is nearly equal to the benefits of all solar power in the US .

Depressing.
We can only hope that the number of ER visits prevented by the presence of GFCIs somewhat offsets those numbers, but I doubt it.
It's hardly a massive issue compared to the pointless phantom loads of everyone's microwave or stove for which there is no benefit or value from the phantom load.

Anyway, I'll measure each of the three I removed with the Kill-a-watt again, and find some low draw nightlight or something to benchmark the kill-a-watt at such a low reading.

TE

 

[TradEddie]

OK, FWIW:

My kill-a-watt reads 0.0 to 0.3w with no load, and does not detect the presence of a 0.3w LED night light, but reads 4.5w for a 5w incandescent nightlight.

Leviton Class A GFCI Issue No. B343488 reads 0.5w.
Leviton Class A GFCI Issue No. B281027, (with LED) reads 0.8w.

TE

 

[Where2]

I just rigged up a Leviton (No. 802-6599-W) GFCI outlet I had kicking around to a grounded cord. According to the Kill-A-Watt, it draws 0.01A (sometimes fluctuating to 0.00A). On the Watt scale, the Kill-A-Watt doesn't read anything but 00 Watts. I thought maybe I had wired it up incorrectly. So, I pulled out my Leviton 6185 GFI Polarity Tester. The outlet checks fine. If I dial the tester switch up to "GFI Trip Test", it does trip the GFCI outlet, as it should.

 

[woodgeek]

TE, thanks for the data. I found an unreferenced comment that the Leviton brands had the lowest draw.

HB, what were your brand/model?

I am ok with GFI technology. It seems that cartoons and movies pre-1960 have an inordinate number of cases of folks getting zotted by forks in toasters and hairdryers in the bath. And I know that my 1960 house had no outlets in any of the bathrooms, think it was a code requirement....had to snake wire and install GFIs myself, in 2010.

 

[Highbeam]

TE, thanks for the data. I found an unreferenced comment that the Leviton brands had the lowest draw.

HB, what were your brand/model?

I am ok with GFI technology. It seems that cartoons and movies pre-1960 have an inordinate number of cases of folks getting zotted by forks in toasters and hairdryers in the bath. And I know that my 1960 house had no outlets in any of the bathrooms, think it was a code requirement....had to snake wire and install GFIs myself, in 2010.

I'll take a look. Note that I did not test with a kill-a-watt but used a Klein clamp on that displays amperage to enough decimals to allow me to calculate wattage that low. It's not a fluke or anything really expensive but much more expensive than a kill-a-watt.

 

[woodgeek]

Got it. If you think your 6W is accurate, that's good enough for me.

 

[Where2]

I just rigged up a Leviton (No. 802-6599-W) GFCI outlet.

After 137 hours on the Kill-a-Watt, I've measured 0.08kWh used.

 

[woodgeek]

ONe way to reconcile the high reading from HB and lower (<1W) readings, other than saying its a brand difference is if the Kill-a-Watts are reading true watts, and the clamp-on ammeter is reading VA. IOW, 6VA (reactive) versus 1W (dissipation)??

 

[Highbeam]

I looked at the gfcis and there is no brand name on the front, only a lock symbol with a lightning bolt through it. I only shop at home depot for electrical stuff so the gfcis must be their brand.

 

[Bill1234]

I stumbled across this thread. I know it is an old thread, but it got me thinking about my GFIs. I checked the one by the main breaker panel and was reading aprox. 150mA with an the meter in-line. That works out to 18 watts @ 120V. I thought that was quite a bit since there are seven GFIs in my house that I count. That would be like having a 100 watt bulb on all of the time. Then I got to thinking about phantom reading due to the impedance in the GFI.

Today I turned off all the breakers except the ones feeding the GFI circuits. I then went to the service entrance meter (the one the power company reads) and watched the spinning disk (you know the one that looks like it is going to fly out of the meter when your AC is running). It was not moving at all. This means no power consumption by the GFIs. If there actually was 150mA consumption by just that one GFI then that would mean 18 watt-hours consumption over a 1 hour period just for one power stealing GFI. That means the disk on the entrance meter would have revolved 4.6 times in one hour (1 revolution equals 7.2 watt hours for my meter). I left only the GFI circuits energized for over two hours and absolutely no movement of the disk in the service entrance power meter even though I was still reading about 150mA (or 16 watts) at just that one GFI. This phantom current reading happens when trying to read the current in a high inductance or high capacitance circuit with a meter that is not designed for such readings. The amperage read with most DVM's (I was using a Fluke 83).

If you are concerned about you GFI outlets then do the same experiment. If you turn off all of the circuits except the one powering the GFI and have everything else unplugged from the circuit and are see the spinning disk moving at your power companies service meter try tripping the GFI using the test button. If the power companies meter then stops this means you probably have a bad GFI outlet or still have something plugged into the circuit drawing power.

I powered on one at a time my circuits which supplied my "always on" devices (Cordless phone-answering machine, alarm clock, door bell, smoke detectors, range (the clock), microwave, radon vent fan, furnace control circuit, landscape light power supply (daylight so lights were off) and my two garage door openers. The total consumption was aprox. 62 watts per hour. The spinning disk was revolving once every 7 minutes which equals about 8.5 revolutions per hour or about 62 watts/hour. The biggest surprise when powering each circuit on was the consumption of my two garage door openers in stand-by mode (overhead lights were off). They are consuming a combined 25 watts per hour which is a large portion of my "always on" appliances. That is significant as it equals about $30 a year just for the door openers to be standing by. That is a circuit I am now considering adding switches to to cut the power when they are not in use, especially to the one which is never used during the winter.

 

[EatenByLimestone]

I think the Govt involved in GFI power consumption would translate into $60 outlets in order to save .24w a day.

 

[georgepds]

....The total consumption was aprox. 62 watts per hour. The spinning disk was revolving once every 7 minutes which equals about 8.5 revolutions per hour or about 62 watts/hour. ...

At 62 watts/hr , and 168 hrs in a week, you will have accelerated to a 10.4 kW draw.... wow

By next year this time you'll be drawing 541 kW... now that's a phantom load

(Hint.. there is power measured in watts, and energy measured in watt hours, but no watts per hour)

 

[Highbeam]

I stumbled across this thread. I know it is an old thread, but it got me thinking about my GFIs. I checked the one by the main breaker panel and was reading aprox. 150mA with an the meter in-line. That works out to 18 watts @ 120V. I thought that was quite a bit since there are seven GFIs in my house that I count. That would be like having a 100 watt bulb on all of the time. Then I got to thinking about phantom reading due to the impedance in the GFI.

Today I turned off all the breakers except the ones feeding the GFI circuits. I then went to the service entrance meter (the one the power company reads) and watched the spinning disk (you know the one that looks like it is going to fly out of the meter when your AC is running). It was not moving at all. This means no power consumption by the GFIs. If there actually was 150mA consumption by just that one GFI then that would mean 18 watt-hours consumption over a 1 hour period just for one power stealing GFI. That means the disk on the entrance meter would have revolved 4.6 times in one hour (1 revolution equals 7.2 watt hours for my meter). I left only the GFI circuits energized for over two hours and absolutely no movement of the disk in the service entrance power meter even though I was still reading about 150mA (or 16 watts) at just that one GFI. This phantom current reading happens when trying to read the current in a high inductance or high capacitance circuit with a meter that is not designed for such readings. The amperage read with most DVM's (I was using a Fluke 83).

If you are concerned about you GFI outlets then do the same experiment. If you turn off all of the circuits except the one powering the GFI and have everything else unplugged from the circuit and are see the spinning disk moving at your power companies service meter try tripping the GFI using the test button. If the power companies meter then stops this means you probably have a bad GFI outlet or still have something plugged into the circuit drawing power.

I powered on one at a time my circuits which supplied my "always on" devices (Cordless phone-answering machine, alarm clock, door bell, smoke detectors, range (the clock), microwave, radon vent fan, furnace control circuit, landscape light power supply (daylight so lights were off) and my two garage door openers. The total consumption was aprox. 62 watts per hour. The spinning disk was revolving once every 7 minutes which equals about 8.5 revolutions per hour or about 62 watts/hour. The biggest surprise when powering each circuit on was the consumption of my two garage door openers in stand-by mode (overhead lights were off). They are consuming a combined 25 watts per hour which is a large portion of my "always on" appliances. That is significant as it equals about $30 a year just for the door openers to be standing by. That is a circuit I am now considering adding switches to to cut the power when they are not in use, especially to the one which is never used during the winter.

I don't know how many old school disc meters are still in use. I know I'm on to a smart meter with no disc to watch.

My home's stand by load is about 200 watts these days. Nothing running. As measured by a clamp on whole house monitor made by efergy. The clamps are around my main inlet between the meter and the main panel. If your tool can't measure an 18 watt draw correctly then what device would measure no draw on that device? Does the power company's meter not work on the same flow measuring principle?

 

[CliffStoll]

This page has been quiet, but I've made a few measurements, so I'll post 'em here.

I purchased a new Leviton GFCI (July 2018) - it has tamper-resistant flaps over the slots, and a little green LED. Leviton Model GFNL1.
I spent some time carefully measuring its power consumption. These figures are actual measurements, using calibrated Hewlett-Packard voltmeter and AC ammeter.

All measurements are at 121 volts, 60 Hz. Temp was 65 degrees F (18 C)

GFCI not tripped, no load (ie, idling with the green led on): 0.0096 amperes (9.6 ma). = 1.16 watts (rms)
GFCI tripped (ie, no output, green led is off): 0.0063 amperes (6.3 ma) = 0.76 watts (rms)

The Fairchild RV4145A integrated circuit that's the brains of most modern GFCI's needs about 1.5ma (full time). A dim LED typically eats 3 to 5 ma. This accounting leaves some 4.8ma of current unaccounted for.

So every GFCI outlet is a phantom load of about 1.16 watts, all day and night.

When my kids were little, I went around our house and installed GFCIs everywhere. Result? My 2 bedroom home has 23 GFCI outlets! That's 26.7 watts, full-time phantom load.

There's 8,760 hours in a year so that 26.7 watt phantom load turns into 234 KiloWatt-Hours of energy used each year.

In the Bay Area, Pacific Gas & Electric charges around twenty cents per kilowatt hour .... so those 23 GFCI outlets cost us about $47 each year.

(A single GFCI has a power cost of about $2 per year. Amazon sells a GFCI outlet for $10 to $12. So in 5 or 6 years, the cost of the outlet equals as the cost of the electricity it's used. Perhaps its time to add energy star ratings for GFCI outlets? )

Making things worse: At our home, there are many full-time loads that just idle most of the time: several internet routers/wifi modems, wired smoke alarms, computers, printers, a full time doorbell transformer, cellphone chargers. There's also a set of security cameras, a burglar alarm (with a battery charger). Two microwave ovens that chew up 9 watts of phantom power each. A solid-state sump-pump switch that needs 8 watts of power all the time. The refrigerator uses 15 watts even when its motor isn't going. Our furnace uses 12 watts all the time.

All together, including those 23 GFCI, my house shows a full time, baseline load of about 400 watts -- hey, that's $700 per year!

Moral: I'm pruning my phantom loads.

A simple switched power strip knocked the home-entertainment center phantom load to zero. Same with the computer printers. The furnace? Unplug it during the summer months. Replace old power supplies with higher efficiency Class V wallwarts.

And do I really want all my house outlets GFCI protected?