cancel
Showing results for 
Search instead for 
Did you mean: 
Replies are disabled for this topic. Start a new one or visit our Help Center.

What are the variable definitions for Nest Thermostat output?

mendicant98
Community Member

My Nest Thermostat controls an old furnace which uses natural gas. The furnace is either ON or OFF; whether the furnace is ON or OFF, depends upon the measured temperature at the Nest Thermostat, relative to the temperature setpoint.

I am trying to convert my furnace to an electric heat pump.

In order to size the heat pump properly, I need to find out the ON and OFF durations for the furnace, during times when the outdoor temperature is near zero degrees Celsius/32 degrees Fahrenheit.

I requested, and received from Google, the output file for the Nest Thermostat on a series of days in January 2022, which match the above criterion.

However, I have no idea what the column labels/variable names in the output file mean. I cannot tell whether the furnace is commanded ON or OFF.

Can anyone point me to a document which describes the output variables in the Nest Thermostat datafile which Google maintains?

Thanks....

13 REPLIES 13

Houptee
Bronze
Bronze

That is not how you size a system to your house.

You can do what they call a Manual J calculation to determine the "heat loss" of the building.

You enter the dimensions and the type of construction (wood frame, brick, etc) and if there is any insulation in the walls and ceilings etc.

How many windows and sizes and what kind are they double pane or old single pane with storm windows etc.

Then after you enter all the data about the house it calculates the "Heat Loss" of the building.

That is how many BTUs are lost thru the walls and ceilings and windows etc etc so if it says you lose 45K BTUs then you need to put in a system that can provide 45K BTUs to keep the house at the design temp.

The "design temperature" which is already determined based on your zip code which is the avg winter outdoor temp that you want to use as the basis for the calculated BTUs input needed to keep the house heated at the design temp.

You can use this free version from Lennox to do a calculation of what size system you will need.

https://www.lennoxpros.com/load-calc

 


Houptee -- NJ Master HVAC Licensed Contractor

mendicant98
Community Member
 
With respect, you haven't answered my question. Still looking for someone who knows how to interpret the output file variables from a Google Nest Thermostat.
 
To your comment:
 
Up to a point, what you say is true. I understand how to calculate an aggregate heat conductance, accounting for windows, walls, floors, and most importantly, ceilings (most heat loss is through the ceilings, making the insulation R-value there of primary interest). Thank you for the link to the Lennox tool. It's not bad; my hand tools/spreadsheet are more quantitative.
 
However, to size the heat pump system properly, one needs to look at the relationship between the LOWEST temperature outside, relative to an inside setpoint. If one uses the avg winter outdoor temperature (as you have suggested), then the system will be undersized.
 
In truth, it's even more complicated. At night, our habit it to turn our thermostat down to 50F. And, where we live, the coldest temperatures occur ONLY at night, usually between 3am and 6am, on a relatively few days in December, January and February. The mean temperatures are up to 20 Fahrenheit degrees higher than the lowest temperature; the lowest temperature is about 32F. So, the delta-T is (50 - 32 = 18) Fahrenheit degrees. During the day, we will set the thermostat to about 70F (68F, really), while the outside temperature during winter months is, at worst, about 40F (and not a constant at that temperature). So, in that instance the delta-T is 30 Fahrenheit degrees. 
 
But, the next person to own the home may want the night time thermostat to be 70F. Making the delta-T (70 - 32 = 38) Fahrenheit degrees.
 
In sum: the sizing of the heat pump should NOT be set by an average temperature, but by the lowest consistent outside temperature experienced, relative to the highest desired inside setpoint. In my professional opinion, for what it may be worth.

The Manual J calculation is the industry standard way to calculate the system size.

Most building depts around the USA require a printout of the Manual J report before they issue permits for new construction homes.

Depending on where you are in the country you will most likely have Aux heat source in your heat pump so it uses that in an extreme cold event.

As far as the Nest data file I have no idea. I don't work for Nest. I volunteer to help people on this forum get Nest thermostats working.


Houptee -- NJ Master HVAC Licensed Contractor

Thanks.

Our home dates to 1947. Second floor added 1984. Another room added 2000. While I'm a little familiar with Manual J, it is not required in order to pull a permit for replacing a gas furnace with an electric heat pump in our city.

The home is located in the San Francisco Bay Area, specifically along the Peninsula. Our normative design temperature for winter is about 30 F. (For others looking to find your design heating and cooling temperatures, see: https://codes.iccsafe.org/content/ICC3102020P1/appendix-a-normative-design-temperature-limits-by-sta....) (For example, the winter heating design temperature in Middlesex County, NJ is 10 F.) 

As a consequence, we do not need any auxiliary heating strips. 

I understand you do not work for Nest, and that you are a volunteer. I asked a very specific question, related to achieving the correct size heat pump for our climate and location and heating/cooling requirements. Extracting historical data from our Nest is possible; in fact, Google did send me a datafile which they had stored from several years ago. However, the file they sent contains variable name headers which are unintelligible. Basically, I need to know when my furnace when ON, and when it went OFF. Simple question. Apparently, however, either the answer is complicated, or the 'correct' person has not yet read my post. Sigh.

Best wishes...

If you are going with a heat pump I suggest you look into the Inverter drive systems.

They have heat pumps now that vary they speed of the compressor and vary the flow of the refrigerant to produce SEER 18 and higher.

They can still produce heat in very low temperatures without kicking on electric coils Aux heat.

Look into the Daikin Fit, Fujitsu HyperHeat, Gree Flexx (this is sold under several private label names ACiq, Mr Cool Universal, Cooper and Hunter, and others but all the same Gree unit).


Houptee -- NJ Master HVAC Licensed Contractor

This is helpful. Thank you.

Hey folks,

 

Thanks for the help, Houptee.

mendicant98, we'd like to take a deeper look into this — could you fill out this form and let us know once you're done?

 

Thanks,
Mel

I have filled out and submitted the form.

BTW one of the Nest Thermostats output files, which I received from Google after making a request, has the following column headers:

 

Date

Time

avg(temp)

avg(humidity)

max(pir)

max(nearPir)

min(ch1)

max(ch1)

min(ch2)

max(ch2)

min(als)

max(als)

min(tp0)

max(tp0)

min(tp1)

max(tp1)

min(tp2)

max(tp2)

min(tp3)

max(tp3)

2022-10-01

00:00

23.74

59.63

 

 

 

 

 

 

32.00

40.00

-48.00

51.00

-47.00

45.00

-45.00

46.00

-45.00

47.00

2022-10-01

00:15

23.80

60.57

19.00

 

 

 

 

 

8.00

32.00

-52.00

48.00

-48.00

49.00

-48.00

45.00

-53.00

50.00

This is a really good article by this man who specializes in energy and building science issues.

https://www.energyvanguard.com/blog/my-undersized-heat-pump-in-an-arctic-blast/

 


Houptee -- NJ Master HVAC Licensed Contractor

Agree, really good article.

I was trying to use the Nest Thermostat output data, to perform a similar analysis. Absent that (so far), I've been able to do the following:

1) I've extracted the annualized Heating Degree Days for each of 2020, 2021, and 2022, for our location. I use the web site degreedays.net, to perform this extraction. (Units: degrees Celsius, times days.) I use a base temperature (thermostat setpoint) of 68F.

2) I've used the monthly Nest Thermostat reports, to calculate the total furnace ON time, annualized, for each of 2020, 2021, and 2022. The furnace plate rating is 125,000 BTU/hr when ON; the bonnet heat output (that is, the amount of heat transferred to the air ducts) is 100,000/hr BTU. (That means, the furnace efficiency is 80%.) 

3) I have used the utility company's report, of the number of therms of gas consumed each month, to determine how much gas our furnace consumes each month. (I've been able to extract/subtract out, the gas used for the hot water heater, and the gas used for cooking.) The monthly consumption, plus the Nest Thermostat report of the number of hours the furnace is ON, gives an energy consumption for the furnace which is within a few percent of the plate rating for the furnace.

4) So, I take the annual energy consumed for heating (converting the units first to kWh, then to Watt-days), and I divide this number by the heating degree days for the year, to obtain an effective heat transfer coefficient for our house for purposes of heating (not cooling). The units of this calculation are W/K, or Watts per degree Celsius (or, degree Kelvin; same thing).

5) The results for our house: 96, 85, 134 W/K, for (respectively) 2020, 2021, 2022. (Why the wide variation? a) our thermostat set point isn't always 68F; b) the heat transfer coefficient concatenates heat transfer via conduction, convection, and radiation, and I can imagine in some years, heat transfer due to wind/convection is higher than in others; c) other factors (open windows, open thermal blinds) not accounted for by this kind of calculation.

6) At the end of the day, I'm trying to size the electric heat pump I want to specify. Using the calculations above, I can do so. If I use the highest (most conservative) heat transfer coefficient (134 W/K), and multiply it by the worst-case delta-T (the inside thermostat set point, which is 20 C [68 F], minus the 99% heating temperature for our region, which is -1 C), then I get the size of the heat pump I need, which is 2.8 kW, or 9800 BTU/h. (This calculation assumes the heat pump is 100% efficient. To refine the sizing calculation, divide the 9800 BTU/h by the efficiency of the proposed electric heat pump.)

7) Compare the 9800 BTU/h, to our present furnace's useful heat output, which is 100,000 BTU/h, or about 10X higher. What does that mean? a) 9800 BTU/h is 10% below the output of a 1-ton heat pump; most heat pumps in the US are specified in terms of 'tons', not kW, and not BTU/h. One cannot easily buy a 1-ton heat pump with an air handler, to couple into our present duct system. So, I would start with a 2-ton heat pump, in terms of my sizing. b) This spec means, it would take five times longer for the electric heat pump, to bring the house to a given set point, than our present furnace. c) On the other hand, it is unlikely the temperature swings in our house would be as large as they are today. The Nest Thermostat, in Eco/Leaf mode, seeks to keep the furnace OFF as long as possible. The house temperature will fall to 63 or 60 F, and only if I adjust the set point upward by hand, will it kick the furnace ON. With the electric heat pump, the temperature of the house will be kept much closer to 68 F, the compressor in the heat pump will not be switching ON and OFF anywhere near as frequently as the present furnace switches ON and OFF (which will extend the lifetime of the heat pump, according to my understanding of heat pump reliability and longevity).

That's where my thinking is at, today. As Houptee has suggested, a variable speed heat pump, with high SEER, with an inverter drive; combined with an air handler unit which is as efficient as possible (and also quiet); with a 2-ton rating; seems to be what I want.

Yet: we've gotten six quotes, and ALL specify either a 4-ton, or 5-ton, system. Given my measurements on the thermal performance of our house (and, we're about to increase the R-value of our attic insulation), I just don't understand why doubling, or more, the power of the heating system, is the 'correct' thing to do.

Maybe, Houptee, you have thoughts on that dilemma.

Thanks to all.

Oversizing on the AC is actually bad because it wont run long enough to remove humidity it will short cycle.

If you are confident in your calcs for heat the Mr Cool Universal / Gree Flexx / ACiq / and various other brands that all seem to be the same unit made by Gree.

The 2 ton has dip switches to make it a 3 ton so you can adjust it up to 3 tons if 2 ton is struggling.

This guy did in depth test on the Mr Cool

https://youtu.be/n_fVbtqav2o

 

 


Houptee -- NJ Master HVAC Licensed Contractor

Markjosephp
Community Specialist
Community Specialist

Hello folks,

 

I'm dropping by to ensure that everything is covered here.

 

Thanks for the continued support, Houptee, and for assisting, Mel.

 

I appreciate your thoughts and thank you for your patience with us, @mendicant98. We've created a case for you. Your issue has been escalated to our upper support team. So, please be on the lookout for an email from us in your inbox soon and continue the conversation there, as this thread will be locked within 24 hours.

 

Regards,

Mark