Getting your OATs

[SteamyTea]

When working out thermal losses through walls, it is usual to use outside air temperature (OAT).

Would it be more accurate to use actual outside wall temperature (OWT). This would more accurately describe the thermal flux as it would account for solar gain, wind, evaporation etc.

Should be easy enough to bury a sensor a few mill deep into a wall.

May also be worth adding sensors to opposing, or all external walls even. This would give a much better picture of what is happening.

Thoughts anyone?

[Radian]

The only thing I can think of is that the temperature there is directly coupled to the fabric loss - put another way, to take an extreme example; if the wall was made from a really good thermal conductor like copper then your delta would be very small indeed. Another case of regression at work.

[SteamyTea]

4 minutes ago, Radian said:

The only thing I can think of is that the temperature there is directly coupled to the fabric loss

Was my thinking, maybe a trial is the answer.  Shaded area and compare temperatures.

It has come about as my OAT sensor (DHT22) has eventually got too wet and packed up.

So probably time to change it to a DS18B20, should be fairly easy to do.

[Radian]

Much better sensor for temperature alone IMO, 5 for £11.75  seems reasonable. Got one in the ground here.

[SteamyTea]

8 minutes ago, Radian said:

Much better sensor for temperature alone IMO, 5 for £11.75  seems reasonable. Got one in the ground here.

What I have.

Had some buried last summer, but they soon packed up, or maybe the cable to them broke.

Pulled then up a while back so shall see what they are up to.

[Radian]

Probably had dry heatshrink sleeving. Capillary action will pull moisture through in time. Needs to have some decent Glue lined heatshrink over the original although I can't vouch for the stuff in that link. Farnell or RS might be more trustworthy. We used to keep the Navy's hydrophones watertight with the stuff we got from RS.

[SteamyTea]

Well I have cobbled together some kits, so BME/P280s and some DS18B20.

Left them on the window cills logging away, while I am away.

Went I am back, later in the week, I shall fit them in the proper places and see what comes out in the wash.

[JohnMo]

Wouldn't you just end up chasing your tail for ever, as things will change on a minute by minute basis.  And still leak heat to atmosphere at its temperature?

 

The better insulated the building the closer the OWT, will be to OAT and vise versa for poorly insulated.  All confused if the sun is out.

 

[SteamyTea]

8 hours ago, JohnMo said:

Wouldn't you just end up chasing your tail for ever, as things will change on a minute by minute basis

This is what I hope to find out, can log at the minute interval, or finer.

8 hours ago, JohnMo said:

And still leak heat to atmosphere at its temperature

Yes it will, it will do nothing to change the thermal properties of what is already there.

8 hours ago, JohnMo said:

The better insulated the building the closer the OWT, will be to OAT and vise versa for poorly insulated

Yes, see above, but all data is useful when it comes to modelling.  It may show that it is worth having extra insulation on north facing wall, rather than assume that the difference will be too small to bother with.

8 hours ago, JohnMo said:

All confused if the sun is out

Same as above, but the other way around, it may show that less is needed, and what the wall, rather than the windows, contributes to overheating.

[TerryE]

On 22/02/2022 at 12:31, SteamyTea said:

Should be easy enough to bury a sensor a few mill deep into a wall.

If you have a flush meter box in your outer skin then another alternative is to put in there. 

 

I also have an Aqara ZigBee temperature and pressure sensor in my garden shed, which feeds into my RPI4-based Home Assistant setup.  Below is a quick screen grab from HA, but the data goes back a couple of years in InfluxDB.  The shed is unlagged so tracks outside temperature well, but the sensor if in a dry environment.