CO₂ logging

[Ed Davies]

In another thread I wrote that the only known building regs change which would affect my house if I got a warrant extension which required updating to the more recent regs would be the one for a logging CO₂ monitor which I'm doing anyway. @PeterW expressed some surprise at this so I referenced the relevant (Scottish) building regs section part 3.14.2.

 

In case anybody's curious, I have a few vAir monitors made by a chap in Sofia. Unfortunately, he's stopped doing them for the time being but they're fairly simple things with off-the-shelf sensors and Wi-Fi equipped processors (ESP8266). I have two with just temperature and humidity sensors (one in the small bedroom I use as my study and the other in the kitchen) and a more fully-loaded one in the bedroom with temperature, humidity, CO₂, pressure and light-level sensors. These and a couple of other sensors send reports to an MQTT broker running on a Raspberry Pi where some of my own software stashes them away in a Sqlite database then servers the data over HTML where some other programs of mine plot graphs from it.

 

Here's the bedroom CO₂ level for last week or so. I was away Sunday and Monday and Tuesday so that gives a nice baseline. Note the non-zero base to the Y-axis. A little “research” project I've been meaning to get round to is comparing the overnight CO₂ levels (at, say, 04:00) with the winds at Wick airport (which I'm also logging) to get some idea of how much difference that makes to the ventilation. Note, this is all in my rented house, not the one I'm building which is still even more ventilated.

Edited October 25, 2018 by Ed Davies
Correct days I was away (Sunday night starts just before “Mon 00:00” - doh!)

[PeterW]

I’m fascinated by this sort of stuff @Ed Davies but I don’t get the logging bit as that would normally indicate someone was doing something with the data ..? In your instance you are,  but what does Mr Joe Average make of this bit of random kit in his new developer home..??

[ProDave]

I take it this house does not have MVHR?

[Ed Davies]

Dunno, you can read that as well as I can. Still, I assume what they're expecting is a little LCD in the hall with a graph of the last 24 hours so people can see what the levels were overnight, not just when they happen to glance at it when they get home at 18:00.

Edited October 24, 2018 by Ed Davies

[Ed Davies]

3 minutes ago, ProDave said:

I take it this house does not have MVHR? 

 

The one I'm renting? No, upstairs isn't too bad but downstairs has a wooden floor which provides sufficient ventilation that MHRV would be a tad redundant.

[Jeremy Harris]

A few years ago I built a small CO₂, temperature and RH logger (some here have borrowed it) and was a bit surprised at how high the bedroom CO₂ level got in the middle of the night at our old house, even with a small window that's always left open.  Interesting to correlate events like the bedroom door opening with a sudden decrease in CO₂ concentration; demonstrates well how important cross flow ventilation is in a room (our bedroom door in the old house was almost diagonally opposite the small open window).

 

The concentration levels were similar to those above, and enough to make me feel a bit muggy headed in the morning.  Leaving the bedroom door ajar made a significant difference, and lowered the peak concentration down to around 800ppm.

 

The new house has CO₂ monitoring and logging built-in, with a display in the hall where things like outside temperature, inside CO₂ concentration, RH, plus GPS derived date and time, are displayed.  The highest concentration I've seen is under 700ppm, and it only gets that high when there are several people in the house.  The bedroom CO₂ concentration rarely gets over 500ppm, so is a massive improvement on the old house. 

 

I'm convinced that the significant lowering of CO₂ level in a house with MVHR contributes a great deal to the "fresh air feeling" that visitors often notice.  Damon HD borrowed my portable monitor and did some experiments with the environment in a primary school classroom a couple of years ago, and IIRC he mentioned that there was a pretty good correlation between attentiveness (or the lack of it) with CO₂ concentrations of over around 800ppm.  I can't remember whether the thread on that is here or lost when Ebuild went down, but it was an interesting read.

[Ed Davies]

I don't leave the bedroom window open but usually leave the door ajar. Maybe the higher levels of CO₂ Sun/Mon night were due to not doing that, I don't remember. I'm really hoping that MHRV in the new house will give better CO₂ levels without needing the heat loss of an open window or door.

[Jeremy Harris]

Just found one of the 24 hour plots from when I had the logger located in the bedroom of the old house.  I managed to correlate some of the dips with known events, that illustrate how cross ventilation from the bedroom door being opened affected the CO₂ concentration.  The data seems pretty close to yours:

 

 

[Ed Davies]

Yeah, this domestic logging stuff can soon wander into TMI territory ?

 

My original logging was to text files. Closed one and opened the next when they reached 500 kB, an arbitrarily chosen size which worked out quite well at about 10 hour's worth of data. About 500 MB for the roughly 13 months before I switched to Sqlite. I did wonder about encrypting, and therefore also compressing, the files more than a day or two old. Doing that's one minor reason why I'm considering switching back to flat files.

[Jeremy Harris]

I just log data to a USB stick, one file = 1 calendar month.  The various sensors are all sampled every 6 minutes and the file format is just .csv, so that I can play around with it easily in a spreadsheet or whatever.  I may get around to poking the data down a wifi link at some point, as there's a RPi file server that's on all the time, providing us with a local cloud storage solution.  It would be handy to be able to take a look at the data whenever I wish to, rather than have to swap out a USB stick. 

 

One problem I found early on, and the reason I now save files monthly, is that it's pretty easy to end up with more lines than a spreadsheet application will handle.

[Ed Davies]

I currently have 22 million lines in a 1.8 GB sqlite database with entries since 2016 November. Most data is recorded every minute or two, power consumption every 6 seconds because that's what the Current Cost meter gives. Would be a bit awkward in most spreadsheet programs, I think.

 

In the background on my second monitor I usually have a graph of various temperatures, humidities and overall electricity power consumption updated every 30 seconds from midday the previous day to the current time. I originally just put it up for fun but actually I find I use it quite a bit to see what the central heating's up to and for things like when the washing machine's finished.

 

I've also learned what I think will be useful lessons for running an off-grid house. E.g., I plan on having 6 kW worth of inverters and had previously though of complicated schemes to make sure my consumption didn't go over that. What I've found in practice is that it doesn't anyway so they wouldn't be needed (except for use of the 7 kW electric shower and a few odd occasions when the central heating wasn't working and I was using the cooker and boiling a kettle with a 2 kW convector heater going at the same time).

 

The last few minute's data, just as the central heating starts up:

 

2018-10-25T16:26:02.66Z ['cc', 'power'] 73
2018-10-25T16:26:03.77Z ['study', 'radiator', 'temperature'] 26.75
2018-10-25T16:26:08.75Z ['cc', 'power'] 71
2018-10-25T16:26:14.84Z ['cc', 'power'] 71
2018-10-25T16:26:21Z ['cc', 'power'] 71
2018-10-25T16:26:28.1Z ['cc', 'power'] 74
2018-10-25T16:26:33.1Z ['cc', 'power'] 73
2018-10-25T16:26:39.2Z ['cc', 'power'] 122
2018-10-25T16:26:40.61Z ['kitchen', 'temperature'] 16.04
2018-10-25T16:26:40.77Z ['kitchen', 'humidity'] 66.64
2018-10-25T16:26:41.69Z ['bedroom', 'temperature'] 18.69
2018-10-25T16:26:41.85Z ['bedroom', 'humidity'] 59.02
2018-10-25T16:26:41.95Z ['bedroom', 'co2'] 925
2018-10-25T16:26:42.06Z ['bedroom', 'pressure'] 1002.23
2018-10-25T16:26:42.17Z ['bedroom', 'altitude'] 92.15
2018-10-25T16:26:42.28Z ['bedroom', 'light'] 8
2018-10-25T16:26:42.39Z ['bedroom', 'eng', 'rssi'] 31
2018-10-25T16:26:45.36Z ['cc', 'power'] 127
2018-10-25T16:26:51.53Z ['cc', 'power'] 129
2018-10-25T16:26:58.12Z ['cc', 'power'] 129
2018-10-25T16:26:58.58Z ['study', 'temperature'] 22.47
2018-10-25T16:26:58.74Z ['study', 'humidity'] 57.95
2018-10-25T16:27:00.06Z ['study', 'radiator', 'temperature'] 26.6875
2018-10-25T16:27:03.71Z ['cc', 'power'] 132
2018-10-25T16:27:15.88Z ['cc', 'power'] 126
2018-10-25T16:27:22Z ['cc', 'power'] 125
2018-10-25T16:27:28.17Z ['cc', 'power'] 127
2018-10-25T16:27:33.96Z ['cc', 'power'] 129
2018-10-25T16:27:40.07Z ['cc', 'power'] 127
2018-10-25T16:27:45.94Z ['cc', 'power'] 130
2018-10-25T16:27:48.91Z ['kitchen', 'temperature'] 16.04
2018-10-25T16:27:49.07Z ['kitchen', 'humidity'] 66.64
2018-10-25T16:27:52.1Z ['cc', 'power'] 127
2018-10-25T16:27:56.45Z ['study', 'radiator', 'temperature'] 26.625
2018-10-25T16:28:04.39Z ['cc', 'power'] 126
2018-10-25T16:28:06.78Z ['study', 'temperature'] 22.44
2018-10-25T16:28:06.93Z ['study', 'humidity'] 57.89
2018-10-25T16:28:10.45Z ['cc', 'power'] 127
2018-10-25T16:28:16.29Z ['cc', 'power'] 129
2018-10-25T16:28:22.36Z ['cc', 'power'] 130
2018-10-25T16:28:28.44Z ['cc', 'power'] 127
2018-10-25T16:28:34.49Z ['cc', 'power'] 127
2018-10-25T16:28:40.6Z ['cc', 'power'] 126
2018-10-25T16:28:48.1Z ['cc', 'power'] 125
2018-10-25T16:28:52.74Z ['study', 'radiator', 'temperature'] 28.1875
2018-10-25T16:28:52.89Z ['cc', 'power'] 133
2018-10-25T16:28:55.55Z ['bedroom', 'temperature'] 18.68
2018-10-25T16:28:55.71Z ['bedroom', 'humidity'] 59.05
2018-10-25T16:28:55.81Z ['bedroom', 'co2'] 925
2018-10-25T16:28:55.92Z ['bedroom', 'pressure'] 1002.25
2018-10-25T16:28:56.03Z ['bedroom', 'altitude'] 92.02
2018-10-25T16:28:56.14Z ['bedroom', 'light'] 7
2018-10-25T16:28:56.25Z ['bedroom', 'eng', 'rssi'] 31
2018-10-25T16:28:57.24Z ['kitchen', 'temperature'] 16.03
2018-10-25T16:28:57.4Z ['kitchen', 'humidity'] 66.67
2018-10-25T16:28:58.96Z ['cc', 'power'] 127

 

Edited October 25, 2018 by Ed Davies