Why is the CO2 level lower in my house than outside?

[mess]

TL;DR

I purchased a high accuracy IR co2 sensor. Outside it reads ~400ppm co2 inside I am getting about ~300ppm co2.

How is this possible? I can't think of what could be absorbing Co2 inside my house, when I am producing it at a pretty high rate.

I have my HVAC running on a frequent cycle mode also. I have no plants.

The sensor: https://www.sensirion.com/en/enviro...n-dioxide-sensors/carbon-dioxide-sensors-co2/

Here is the data:

Inside:��������������

2020/12/05 12:12:22PM	342
2020/12/05 12:12:19PM	346
2020/12/05 12:12:17PM	349
2020/12/05 12:12:14PM	350
2020/12/05 12:12:12PM	353
2020/12/05 12:12:09PM	364
2020/12/05 12:12:07PM	370
2020/12/05 12:12:04PM	372
2020/12/05 12:12:02PM	375
2020/12/05 12:11:59PM	377
2020/12/05 12:11:56PM	378
2020/12/05 12:11:54PM	381
2020/12/05 12:11:51PM	382
2020/12/05 12:11:49PM	384
2020/12/05 12:11:46PM	388

Outside:������������������

2020/12/05 11:28:28AM	432
2020/12/05 11:28:26AM	433
2020/12/05 11:28:23AM	434
2020/12/05 11:28:21AM	437
2020/12/05 11:28:18AM	438
2020/12/05 11:28:16AM	440
2020/12/05 11:28:13AM	442
2020/12/05 11:28:11AM	443
2020/12/05 11:28:08AM	444
2020/12/05 11:28:06AM	445
2020/12/05 11:28:03AM	444
2020/12/05 11:28:01AM	446
2020/12/05 11:27:58AM	449
2020/12/05 11:27:56AM	450
2020/12/05 11:27:53AM	450
2020/12/05 11:27:50AM	453
2020/12/05 11:27:48AM	463
2020/12/05 11:27:45AM	464
2020/12/05 11:27:43AM	463

 

[Ivan Seeking]

My first guess is that the levels are cycling with the time of day and your levels inside are lagging those outside.

 

[mess]

My first guess is that the levels are cycling with the time of day and your levels inside are lagging those outside.

im not sure what you mean by that :(

 

[Ivan Seeking]

im not sure what you mean by that :(

Take data over a 24 hour period and see how much the levels vary. It may be rising and falling outside, which causes it to rise and fall inside but after a delay.

 

[mess]

Take data over a 24 hour period and see how much the levels vary. It may be rising and falling outside, which causes it to rise and fall inside but after a delay.

The outside concentration can only reduce the inside concentration if the inside concentration was higher.
because it is higher outside, that can only mean there is something constantly absorbing CO2 inside my house or the sensor is defective. I understand there is a correlation between the outside CO2 and inside, but I am trying to understand why it is on average lower.

 

[Ivan Seeking]

The outside concentration can only reduce the inside concentration if the inside concentration was higher.
because it is higher outside, that can only mean there is something constantly absorbing CO2 inside my house or the sensor is defective. I understand there is a correlation between the outside CO2 and inside, but I am trying to understand why it is on average lower.

Take data inside and outside over a 24 hour period and I'll bet it all makes sense. ;)

 

[Stephen Tashi]

Is the ppm rating independent of the temperature of the gases being sampled?

 

[mess]

Is the ppm rating independent of the temperature of the gases being sampled?

It comes with a temp and humidity sensor built into compensate

 

[mess]

Take data inside and outside over a 24 hour period and I'll bet it all makes sense. ;)

Lets say over 24 hours the concentration inside is higher than outside. I still don't understand why the smaller sample size would not match

 

[anorlunda]

The outside concentration can only reduce the inside concentration if the inside concentration was higher.

Suppose the outside concentration varies between 200 and 600 though the day, but the inside concentration responds so slowly that it varies only 300 to 500.

CO2 generated or absorbed inside the house could bias the inside reading. However, it's easy to generate C02, and difficult to absorb it.

 

[Stephen Tashi]

It comes with a temp and humidity sensor built into compensate

Perhaps you can think of a simple test to see if that function actually works.

What's the standard definition of "ppm"? Parts-of-what per million what?

Is it a volume measurement ? - volume of CO2 per volume of space?

Is it a fraction-of-volume measurement? - fraction of volume of CO2 per total volume of sample?

Is it a weight measurement? - weight of CO2 in sample per weight of whole sample?

Is it a count of molecules? - number of CO2 molecules in sample per total number of molecules in the sample

 

[Bystander]

However, it's easy to generate C02, and difficult to absorb it.

You're certain? Construction materials? Sheetrock/drywall effects? You're familiar with the (non-)results from "Biosphere 2"?

 

[BillTre]

I was thinking exposed amounts of water inside the house could absorb the CO₂.
CO₂ is highly soluble in water (like open aquariums or inside pools). A humidity sensor would only sense water in the air for its corrections.

But I like the idea of construction materials better.
Lots of those in houses. Not so much, bodies of water.

Some plastics can absorb water to some extent also, but they have limited capacity and would become saturated.

 

[fresh_42]

My first thought was a broken device. There is definitely more $CO_2$ inside than outside: breathing or smoking are two sources. As I read about the correction automatism of the device, I thought: what does it correct where to? If this automatism is meant to make different sources comparable, then it might well tune down actual values inside.

An easy test would be to compare two measurements directly after an air exchange, and then after the room is back to its normal temperature.

 

[BillTre]

Some things I noticed about the numbers:

1. the time points were fairly tightly clustered in time (not much time for variations to arise, unless due to wind)
2. the numbers seemed to vary quite a bit, but not enough to overlap the two distributions
3. Outside and inside readings both generally trended down over a period of less than a minute

This indicates possible problems/tests WRT the readings:

• Accuracy not down to the tens of PPMs? Find a CO2 standard to test your instrument on (or make one), or get a test based on a different process for comparison.
• There maybe some kind of warm up or equilibration period for the probe in a new condition (some pH meters can be like that); take readings over a longer time period, wait until meter settles down to take more accurate readings. Consult those who made it for details.
  On the other hand, such a trend in one direction could be an electronic problem in the sensor system (in pH and I think redox measuring systems, a leak of a current into the body of water being measured can off-set the readings an electronic sensor like a pH electrode would detect).
• Wind blowing variable air conditions to the sensor (making for a series of different air samples being measured); take a sample of air in a large garbage bag. Take a series of readings of that specific air sample to see whether or not the readings are stable.

 

[cmb]

Summary:: I purchased a high accuracy IR CO2 sensor. Outside it reads ~400 inside I am getting about ~300.

How is this possible? I can't think of what could be absorbing Co2 inside my house, when I am producing it at a pretty high rate.

I have my HVAC running on a frequent cycle mode also. I have no plants.

The sensor: https://www.sensirion.com/en/enviro...n-dioxide-sensors/carbon-dioxide-sensors-co2/

Here is the data:

Inside:��������������

2020/12/05 12:12:22PM	342
2020/12/05 12:12:19PM	346
2020/12/05 12:12:17PM	349
2020/12/05 12:12:14PM	350
2020/12/05 12:12:12PM	353
2020/12/05 12:12:09PM	364
2020/12/05 12:12:07PM	370
2020/12/05 12:12:04PM	372
2020/12/05 12:12:02PM	375
2020/12/05 12:11:59PM	377
2020/12/05 12:11:56PM	378
2020/12/05 12:11:54PM	381
2020/12/05 12:11:51PM	382
2020/12/05 12:11:49PM	384
2020/12/05 12:11:46PM	388

Outside:������������������

2020/12/05 11:28:28AM	432
2020/12/05 11:28:26AM	433
2020/12/05 11:28:23AM	434
2020/12/05 11:28:21AM	437
2020/12/05 11:28:18AM	438
2020/12/05 11:28:16AM	440
2020/12/05 11:28:13AM	442
2020/12/05 11:28:11AM	443
2020/12/05 11:28:08AM	444
2020/12/05 11:28:06AM	445
2020/12/05 11:28:03AM	444
2020/12/05 11:28:01AM	446
2020/12/05 11:27:58AM	449
2020/12/05 11:27:56AM	450
2020/12/05 11:27:53AM	450
2020/12/05 11:27:50AM	453
2020/12/05 11:27:48AM	463
2020/12/05 11:27:45AM	464
2020/12/05 11:27:43AM	463

Is it a new house and/or with new concrete?

Do you have many house plants?

 

[mess]

Is it a new house and/or with new concrete?

Do you have many house plants?

house from 1950, no plants

 

[fresh_42]

house from 1950, no plants

weed counts as plants

 

[cmb]

house from 1950, no plants

A mystery then.

These are very low levels of CO2 for accurate measurements. What is the calibration and precision of the instrument, and what is the temperature offset of the gases it is measuring?

 

[mess]

Suppose the outside concentration varies between 200 and 600 though the day, but the inside concentration responds so slowly that it varies only 300 to 500.

CO2 generated or absorbed inside the house could bias the inside reading. However, it's easy to generate C02, and difficult to absorb it.

From my understanding its not possible under standard environmental conditions for the outside air to be under 400ppm CO2 since that's the global level for a few decades now. Therefore the outside air could never influence the inside air to be under 400 ppm.

 

[BillTre]

Have looked more closely at your numbers.
There are two time separated sets of closely clustered points, at two different locations.

The first (based on the time stamp) set (outside) has a smaller downward (wrt time) slope (-0.688 ppm/s)
than the second set (inside; -1.277 ppm/s).
They both slope down.

The period between the readings (2658 s) is too large account for the overall differences inside/outside just based on either of the slopes, but it is in the same temporal direction.

I suspect something instrumental.
Is this a leave it on all the time type meter or a turn it on and use it for spot checks kind of meter?

 

[OmCheeto]

I've never owned nor operated a CO2 device, but the rapidity with which the CO2 levels dropped, and the sequence of the readings, indicates that your batteries might need changing.

Just a guess, of course.

 

[jrmichler]

Do one continuous ABAB test: Inside for one hour, outside for one hour, inside, outside. ABAB tests are good for separating sensor drift from actual reading changes.

The datasheet states a 20 second time constant response to a change in $\rm{CO_2}$ concentration, so the test should be long enough for the sensor to stabilize, plus longer to look for concentration fluctuations.

 

[Frodo]

The data sheet specifies: Accuracy: ± (30 ppm + 3%)

It's a bit ambiguous what that means. If I assume the 3% applies to the measured value and not to the error itself I have, at 400ppm, an accuracy of ± (30ppm plus 12ppm), or ± 42ppm. So, an actual value of 400ppm could display as 358 to 442 ppm.

Also, the measurement range starts at 400ppm so values below that are out of spec. The present global CO2 level is about 412ppm so you are unlikely to get external measurements differing much from that.

Never confuse accuracy with resolution!

Second, I would have thought the outside values to remain fairly stable assuming no close by variable CO2 emitters. What readings do you get far from potential sources?

 

[Rive]

Apart from addressing the seemingly systematic change @OmCheeto found I think you should also check the functionality of the instrument with some obvious CO2 sources.

 

[Barakn]

The o.p. is frustratingly light on details. Assuming the times are the local time zone, the measurements were taken close to solar noon. One important consideration, especially considering the instrument utilizes IR, is whether it was exposed to direct sunlight outside (presumably was shaded inside). I also wonder about contamination from the user's breath or heating effects if it was hand-held.

 

[Frodo]

Evolution and variation of atmospheric carbon dioxide concentration over terrestrial ecosystems as derived from eddy covariance measurements says

Using 355 site-years of CO2 concentration observations at 104 eddy-covariance flux tower sites in Northern Hemisphere, we presented a comprehensive analysis of evolution and variation of atmospheric CO2 concentration over terrestrial ecosystem (ACTE) for the period of 1997–2006. Our results showed that ACTE exhibited a strong seasonal variations, with an average seasonal amplitude (peak-trough difference) of 14.8 ppm, which was approximately threefold that global mean CO2 observed in Mauna Loa in the United States (MLO).

So, they see a peak-to-trough variance of 15ppm across seasons, which equates to ± 7.5ppm. Without access to the paper this suggests the daily or hourly or minute by minute variability is low.

I strongly suspect that your instrument and/or your method of use are faulty.

Edit: I have accessed the paper and the daily variability is much higher than I supposed. The text states that they took "half-hour CO2concentrations" and " spike screening was applied for all dataset. Any data outside ±3 standard deviations within 6 hours (12 points) were regarded as outliers and excluded from the records".

Figure 2 shows significant daily variations.

It seems something interesting happened at the end of 2005 ...

It also says

The difference between intensive human activity areas and natural ecosystems was similar to the increase in CO2 concentration found in Phoenix, USA, where an increase of 111-185 ppm was reported from a pristine rural site to the city center (Idso et al., 1998, 2001).

 

[.Scott]

I don't think there is any problem with these measurements at all. And they do not reflect less indoor CO2 than the outdoor level. In fact, it suggests the opposite.

Before I tell what I believe is happening, let me first say that you need a lot more data. @jrmichler suggested an ABAB measurement pattern. I suggest you collect measurements at a rate of one sample every 30 minutes for 24 hours, then switch between indoors and outdoors. Do this for 4 days so that you have the two examples of the daily pattern each for indoors and outdoors.

Plants consume CO2 in the day and exhale it at night.
Plant "dark respire" all the time - releasing CO2.
They photosynthesize only with light.
https://www.bbc.co.uk/bitesize/guides/zpwmxnb/revision/3

Your samples were taken midday, when the consumption is at its greatest - and thus the CO2 levels are dropping the fastest. You saw them drop while taking the outdoor readings, they continued to drop as you relocated the instrument, and then you saw them continue to drop as you measured them indoors.

Here is a plot of both sets of readings, outdoor (green) and indoor (blue):

Extrapolating, it would appear that had you left the instrument outside, you would have picked up even lower reading that what you got indoors.

I would come to 2 conclusions:
During the time you took these measurement:
1) It was quite sunny; and
2) Your house was ventilated - but not completely open to the outdoors.

 

[FMPeck]

The size of the NDIR sampling chamber is fixed and is open to the atmosphere so that air can move in and out. The number of air molecules in a given volume is affected by temperature and air pressure but not the concentration of CO2. At low pressures or high temperatures, there will be fewer air molecules in the sample chamber, so there will also be fewer CO2 molecules, even though the ppm of CO2 hasn’t changed. Fewer CO2 molecules “fools” the sensor into thinking that the CO2 concentration is lower than it really is. At high pressures or low temperatures, there are more air molecules in the sample chamber and more CO2 molecules, even though the CO2 concentration hasn’t changed. More CO2 molecules “fools” the sensor into thinking that the CO2 concentration is higher than it really is. Therefore a CO2 sensor calibration will only be accurate at one temperature and one air pressure. Odds are the ratio of PV=nRT (ideal gas equation) for inside and outside conditions may serve as a useful correction for your results.

 

[Ray Rebeiro]

I have owned 4 CO2 meters. Three different manufacturers. 2 units from the same manufacturer provide readings that were different by almost 40%. They insisted that I should recalibrate both units. I did and the results did not change. They accepted return of the so called defective unit and refunded the purchase price
although I did not ask for this. I was expecting a functioning unit to be returned. All of the 3 units I now have provide different readings and the differences in the readings are not trivial. The readings respond differently to changes in placement location. The units all cost more than $75 each.
I hope that this helps.