Refrigerator 38F vs 34F

[holeydonut]

I have a few days worth of data now comparing how much more energy it takes to operate my refrigerator set to 34F vs 38F. So far, it looks like the refrigerator is taking ~17% more energy to get those extra 4F's.

Some notes:

1) I have this refrigerator (Samsung RF23J9011SR/AA 23 cu ft counter depth)
2) The freezer is set to 0F
3) The "other extra cooler" on this has been set to "chill" the whole time
4) I'm only looking at data collected from midnight to 5am. This way I'm not dealing with disruptions of people opening the refrigerator (nobody snacks out of this fridge at night; we have an upstairs snack fridge for that)
5) The fridge is always about 75% full
6) Between midnight and 5am my downstairs thermostat is set to 65F
7) Sample data collected over the first couple weeks of February 2021
8) Data collected with an Emporia Vue on the GFCI circuit that the fridge is plugged into.

Here's a table of the limited sample (# in watt hours reported by Emporia each hour).

@ 38F, the thing takes about 63 watt hours of power each hour which works out to 1.51 kWh per day of just "holding" temperature

@ 34%, the thing is taking about 74 watt hours of power each hour which works out to 1.78 kWh per day.

Over a full year, this is estimated to be 98.55 kWh. And since PG&E rates fluctuate like crazy, this 98.55 kWh is like 1 share of Gamestop Stock.

I'll keep an eye on the data to see if the 34F drops into the 60's for watt hours... although I'm sure changing ambient temperatures and stuff may be more of a factor than than the actual F's of refrigerator setting.

Does any of this matter? Nah not really.

 

[holeydonut]

I guess this thread seems kind of random without context. We had a side-convo in another thread about having the refrigerator set to 34F.

PG&E wants homeowners to set their refrigerators to 38F to save energy. 38F is the upper limit for recommended safe operating temperature for your refrigerator to avoid food spoilage and still stay under the FDA guideline where food refrigeration cannot exceed 41F. A refrigerator will cycle so 38F could hit 41 at some point during operation.
Energy Saving Tips

But the flip side is PG&E may cut your power for no reason, so you'll hit the food spoilage limit faster. That means some people will play it safe and just drop the refrigerator to 34F (the lower bound of safe operating temperature for a refrigerator).

TLDR, PG&E sucks.

 

[wwu123]

Interesting, thank you for the data. Seems the power is roughly linear with delta-T within this narrow range, i.e. deltaT of 31F vs 27F is 15% difference, and you measured power delta at 17%.

The default manufacturer setting for my fridge is 38F, but a few years ago I lowered it to 36F, as I noticed the top shelf where the milk is stored was somewhat warmer than the lower shelves. 36F is the lowest the fridge can be set.

 

[holeydonut]

Interesting, thank you for the data. Seems the power is roughly linear with delta-T within this narrow range, i.e. deltaT of 31F vs 27F is 15% difference, and you measured power delta at 17%.

The default manufacturer setting for my fridge is 38F, but a few years ago I lowered it to 36F, as I noticed the top shelf where the milk is stored was somewhat warmer than the lower shelves. 36F is the lowest the fridge can be set.

I haven’t seen a Delta T chart for refrigerators, but I think it’s more of a coincidence that the difference between 31F and 27F is resulting in the power that my fridge is using to cool and hold a fridge at 38F and 34F.

Delta T tests in HVAC more of a test to see if your system is cooling properly, I don’t think the Delta T measure factors in power consumption like SEER does.

 

[getakey]

It should be proportional to the difference in room temp to refrig temp, i.e.:

(65-38)/(65-34) = 1.15 or 15% assuming efficiency is the same for 34 vs 38 degrees. Efficiency loss probably claims the additional 2%

 

[jboy210]

34F. Phttz. I borrowed a -80C freezer from my old client to store all the Pfizer vaccines I am selling. Strangly they only use 8 kWh/day. That is like drying 2 loads of clothes

 

[gpez]

34F. Phttz. I borrowed a -80C freezer from my old client to store all the Pfizer vaccines I am selling. Strangly they only use 8 kWh/day. That is like drying 2 loads of clothes

Those things will probably stay at -80C for days with how insulated they are...

@holeydonut are you also accounting for how much time your fridge is open before the data collection time? Thinking what the temp is before you start collecting (though if you don't snack for hours ahead of that then you've implicitly done so...)

 

[holeydonut]

Those things will probably stay at -80C for days with how insulated they are...

@holeydonut are you also accounting for how much time your fridge is open before the data collection time? Thinking what the temp is before you start collecting (though if you don't snack for hours ahead of that then you've implicitly done so...)

I have a 18 month old that wakes up at about 5:30am... so everyone is usually in bed by 10pm.

That's why I only sample data for the 5 hours midnight to 5am. It's basically the time window I know where the fridge has had enough time to refill the ice machine, cool to its target temp etc. And nobody is opening it from there.

 

[holeydonut]

34F. Phttz. I borrowed a -80C freezer from my old client to store all the Pfizer vaccines I am selling. Strangly they only use 8 kWh/day. That is like drying 2 loads of clothes

8 kWh per day seems low.

This article projects their annual kWh to be 8,333 kWh which is about 22.8 kWh per day. The 8,333 kWh per year would be about $1,250 energy costs using the article's $0.15 per kWh estimate.

Freezer Management

This is aligned with some spec sheets I saw on some Thermo Fisher units I sourced many years ago were 8A at 115V. This would put it at around 8,100 kWh per year since the 2 stage compressors runs non-stop.

Edit: Nevermind, these things are getting pretty efficient! The ultra high end one is down to 8 kWh per day.
-80 Lab Freezers (Ultra-Low Freezers) | Thermo Fisher Scientific - US

 

[jboy210]

8 kWh per day seems low.

This article projects their annual kWh to be 8,333 kWh which is about 22.8 kWh per day. The 8,333 kWh per year would be about $1,250 energy costs using the article's $0.15 per kWh estimate.

Freezer Management

This is aligned with some spec sheets I saw on some Thermo Fisher units I sourced many years ago were 8A at 115V. This would put it at around 8,100 kWh per year since the 2 stage compressors runs non-stop.

Edit: Nevermind, these things are getting pretty efficient! The ultra high end one is down to 8 kWh per day.
-80 Lab Freezers (Ultra-Low Freezers) | Thermo Fisher Scientific - US

Yep. They are quite efficient. And it always amazes me how much these things hold. You could put sides of beef in that thing. But more likely you will put samples in there that will not be found for decades. At one lab I worked at they were going to move one of the older versions of these freezers. They cleaned it out and unplugged it for a few days. When they came to move it they found an entire stack of samples that were hidden behind sheets of ice in the back and were exposed by the melting. Some of the samples were from 20+ years ago. The scientists whose name was on the boxes had retired over a decade before.

 

[mswlogo]

17% sounds high but maybe it’s right. Might not be linear between 34F and 38F either.

When does the fridge do auto defrost cycle? Which can be fairly expensive. Your sample is small that you might be hitting or missing defrost cycles. Just total it for a week on each setting. Usage of door open will average out.

Breaking it out and tallying it the way you have seems awkward error prone especially when you factor in defrost cycles and not knowing where they landed. Why are you averaging everything. You should be adding everything. Then determine watts per hour over the total hours measured, don’t do intermediate averages. I think that will throw things off a bit.

If you want to chase watts get a Sense House Monitor Worth every penny. I have two of them.

I added up your first 4 days of each and got 11% increase. 1474/1325 if you want to check my math.

 

[holeydonut]

17% sounds high but maybe it’s right. Might not be linear between 34F and 38F either.

When does the fridge do auto defrost cycle? Which can be fairly expensive. Your sample is small that you might be hitting or missing defrost cycles. Just total it for a week on each setting. Usage of door open will average out.

Breaking it out and tallying it the way you have seems awkward error prone especially when you factor in defrost cycles and not knowing where they landed. Why are you averaging everything. You should be adding everything. Then determine watts per hour over the total hours measured, don’t do intermediate averages. I think that will throw things off a bit.

If you want to chase watts get a Sense House Monitor Worth every penny. I have two of them.

I added up your first 4 days of each and got 11% increase. 1474/1325 if you want to check my math.

I'll keep getting some data over the next few days and see how the # comes down. But yeah, small sample size and all that.

That's one problem with the Emporia Vue I have instead of the Sense... the refrigerator is on the same GFCI as the one my mother in law uses for her electric kettle and random kitchen appliances. So if I take a long data sample, I need to manually delete the air fryer running for 40 minutes and random tea time.

I am curious to learn how the Sense "sees" a refrigerator. The compressor and various freezer/defrost cycles each have very unique signature when I look at the second-by-second data. I wonder if the Sense actually learns them all.

PS. I'm actually impressed at how "non BS" the Energy Guide sticker is on this fridge in terms of kWh energy estimated. Aside from the completely out of whack $0.12 per kWh, they've actually pegged the annual kWh pretty close to my recent experience @ 679 per year (1.86 kWh per day).

 

[getakey]

you don't believe my math? U and A are going to be the same. So the difference in heat loss is proportional to outside temp minus inside temp or (65-38)/(68-34). The heat loss has to be made up by the electricity used. Nothing else in play except if there is a defrost cycle

Heat Loss Formula:
q = (U × A) × Δt

Where,

q Total heat loss
U The overall coefficient of heat transmission
A Area
Δt the temperature difference between inside and outside temperatures

 

[holeydonut]

you don't believe my math? U and A are going to be the same. So the difference in heat loss is proportional to outside temp minus inside temp or (65-38)/(68-34). The heat loss has to be made up by the electricity used. Nothing else in play except if there is a defrost cycle

Heat Loss Formula:
q = (U × A) × Δt

Where,

q Total heat loss
U The overall coefficient of heat transmission
A Area
Δt the temperature difference between inside and outside temperatures

I tagged your previous maths as informative; I think the data is close-ish to proving your formula!

 

[getakey]

ok, thanks
I was really replying to MSWLogo. I realize that wasn't clear

 

[mswlogo]

I'll keep getting some data over the next few days and see how the # comes down. But yeah, small sample size and all that.

That's one problem with the Emporia Vue I have instead of the Sense... the refrigerator is on the same GFCI as the one my mother in law uses for her electric kettle and random kitchen appliances. So if I take a long data sample, I need to manually delete the air fryer running for 40 minutes and random tea time.

I am curious to learn how the Sense "sees" a refrigerator. The compressor and various freezer/defrost cycles each have very unique signature when I look at the second-by-second data. I wonder if the Sense actually learns them all.

PS. I'm actually impressed at how "non BS" the Energy Guide sticker is on this fridge in terms of kWh energy estimated. Aside from the completely out of whack $0.12 per kWh, they've actually pegged the annual kWh pretty close to my recent experience @ 679 per year (1.86 kWh per day).

View attachment 637398

Sense is far from perfect. Sometimes it surprises me what is does figure out and other times it disappoints me in simple stuff it can’t. It might not be great on a fridge with a variable speed compressor. But you will get a better sense of your whole house with it, guaranteed. And you’ll more likely find what is really costing you in energy use. You could add two extra zones and focus on say charging your Tesla and your fridge and move it around as needed if you wanted.

like I have a water pump. If it runs for longer than an hour something is very wrong. So I put alarm on that. Funny thing is a week after I set it up, the alarm went off and the pipe was frozen and the pump would have burned itself up. This was at a remote cabin and zip sent a neighbor over to shut it down. It’s a great tool.

 

[SMAlset]

We bought our 21.1 cu. ft capacity SubZero refrig/freezer back in 2008 when we did a kitchen remodel and wanted a very efficient unit. They are still making this model and it's an over under unit (freezer on bottom with internal ice) and the motor is on the top so as to not add heat to the refrig/freezer units. 8.3 max running amps. Haven't located our Energy Guide but believe it was rated at either 602 or 659kWh annually.

Like @wwu123 we had ours set to 37F and recently thought the top shelf with milk and egg beaters containers would be better set lower so adjusted it to 36F (preset to 38F but adjustable between 34F to 45F in the refrigerator section).

On the subject of reducing wattage, if you've had your unit for a while two things to consider. Check the seal on it and clean the condensor unit. Ours has a powerful seal (so strong in fact you can't always open quickly if you've just shut it). We're due to clean our condensor coils.

Yes, the 12 cents kWh average electricity cost is a joke. With solar and PWs now however ours has been down to $0.0. No bills since PTO.

BTW our refrig/freezer installation required it to be on a 15A dedicated circuit without a GFCI and I think I read somewhere that that had something to do with when power might go out. Saw that you indicated yours was on a shared GFCI circuit.

 

[wwu123]

It should be proportional to the difference in room temp to refrig temp, i.e.:

(65-38)/(65-34) = 1.15 or 15% assuming efficiency is the same for 34 vs 38 degrees. Efficiency loss probably claims the additional 2%

Yup, that's what I meant by delta T = difference (delta) in Temperature (T). Just a generic term, nothing to do with specifically with HVAC industry that OP was replying. Just that a lot of heat transfer is linearly proportional to the delta T, like heat loss through windows and walls.

 

[charlesj]

Would this be a getter formula for heat transfer?
Rate of heat flow - Wikipedia

Warm room temperature heats up inside refridge, not linear?
Then, how much energy to cool that.

 

[getakey]

thats the rate, but not the net amount of heat transferred. The net heat transferred has to equal the net amount of cooling