I remember the UK Association for Environment Conscious Building (AECB) doing some work on this a while back, since it’s a common question when internally-insulating a building and (effectively) moving the ends of the joists into colder and damper conditions.
Looks like the study I remembered used the OmniSense sensors. This is probably worth a read if you’ve not seen it already:
It looks like the systems used might run to well over £1000 perhaps as high as £2000 Something in the region of 7m risky solid wall dwellings in England and Wales that should be insulated if it can be done safely. I think it would be very sensible to have more affordable systems available so more monitoring can be deployed. I might get in touch with a few universities to see if anyone is researching this and might be willing to contribute. Temperature and humidity sensors are really quite affordable these days. There are also very affordable soil moisture sensors but how useful they could be is a total unknown to me.
Do you think this is relevant enough to the OEM project to be taken up by the lab? I’d be happy to buy some components and install them sharing experience and results. I have never worked with embedded hardware but have developed websites and desktop software so should be able to pivot. I could use a good point in the right direction at the very least.
By Platform, do you mean emoncms (the software to record and visualise the data)?
If so yes.
You could setup a 1-wire sensor to record the humidity. Connect that to an ESP device to read the data and send it on. You can send that to an instance of emoncms to record the data.
You may be aware of Home Assistant which could provide automations to alert you of issues.
Thanks Brian,
I agree temp and relative humidity should be quite doable and would be a small pivot to what are core features of your hardware/software. How to use RH as a proxy for Wood Moisture Equivalent or how to use an alternate sensor would be really helpful too.
Thanks John, I’ve seen such curves. Would be a good starting place. They are for steady state I believe. You can register over 100 RH and still have a fairly low ME over a short period. That said I think RH might be very useful and less prone to problems arising from salts which can screw the resistance of the timber and make ME figure much less useful.
I guess I best buy a package and start trying to log RH and then once I’m happy with that plaster it into a little void at the end of a joist.
Thanks yes, that’s how I ended up here. not a member of GBF. I have emailed AECB and also bought a few components on Ada Fruit. I’ll write up my experiments. Kit was under £100 for a number of sensors but bought a load extra bits as postage a sunk cost.
I did an unusual new build that meant I was very concerned about humidity, over 10 years ago now. So I bought a temperature & humidity logging system from a company in Canada that doesn’t seem to exist any longer. It never worked properly although some bits did work well enough and for long enough to convince me that the house ‘worked’ as designed and I didn’t need to be too worried. The house hasn’t fallen down yet, anway!
As you suggest, if you’re going to build things into inaccessible places then you really want at least the power, and ideally the data, to be sent over wires rather than wirelessly; otherwise the system will have a very limited life. I don’t know what the current state of the art is; it always seems to be more expensive than reasonable.
But nowadays I use a bunch of wireless T&H meters to measure the internal humidity of our house, and use rechargable batteries. I also use some homemade pi-based units and even an EmonTH that’s plugged in to the mains. They all log to my emoncms system.
So I suppose a bunch of EmonTH units might be a reasonable option for building in to your joist pockets (maybe they’d do a good price on multiple units without the external cases and battery boxes etc?). Otherwise putting something together from Adafruit parts (or Aliexpress) might be a way forward.
One other thought,MiFlora is a sensor for measuring the dampness of soil. Might that help? It would need to be accessible, as it is a button battery, but there may be USB versions somewhere.
Good idea! I see “Soil Moisture Sensor, 5x Capacitive, Hygrometer Module, Arduino, Raspberry Pi, 98mm x 23mm, 3-pin” and various others on Amazon UK that are wired. If Craig is happy playing with hardware they might work very well. A little bit of experimentation or putting one in the same situation as various other meters for comparison/calibration could produce a very usable system.
My plan is to multiplex 8 x SHT30 Temperature And Humidity Sensor - Wired Enclosed Shell[ID:5064]
With these:
2 x Adafruit PCA9546 4-Channel I2C Multiplexer (TCA9546A Compatible)
[ID:5663]
And provide power and data over usb using this:
1 x Adafruit FT232H Breakout - General Purpose USB to GPIO, SPI, I2C (USB C
& Stemma QT) [ID:2264]
I might supplement with a board that can broadcast on BLE/ radio etc. Once I have it working I might see if I can get it logging to OEM. I don’t have a hub and my electrical system isn’t really complex enough to bother monitoring just now so will likely just run on an old debian laptop to start with.
I’m minded not to bother with the soil sensor just now as I haven’t got any of the other equipment for testing WME to affect some kind of lookup table. Salts can be a pain with the WME too and I’m working with 100 year old wood of unknown provenance. I’d be willing to give it a go longer term though. More of a desktop exercise to start with, might need to reverse engineer a proven system. Some of these meters run to £1000s but I don’t really know what the difference can be. At a surface level they just give you a clue about moisture content from a measure of resistance. I wonder if the smarter ones use a range of voltages, AC/DC, frequency or look at something else like capacitance. Just material databases? Personally believe it’s too in-exact a science with all the salt issues to bother getting too accurate a machine but someone with more experience of these things would be valuable. Perhaps they have some protocol to test and correct for salts?
I’m happy to report back next week once I have my parts delivered. I might not get it all working without help. Never used embedded systems. Looks like everything is supported with Python packages. I don’t much like Python but needs must. Looks like OEM is a C++ product, I’ve used Java and C# quite a bit so perhaps I have a bit of a headstart with C++. I’ll have a look at that once I’ve got it working with the AdaFruit tutorials that all seem to be Python.
Thanks for tip. Might come to that. I do want a cabled access though. Going to be in an inasseccible place and have yet to loose my suspision of wifi. Was CAN or USB by my initail research for the cable runs I was looking at. Could pop something like this on for everyday and simply power over usb and plug it into a computer only if/when an issue or want to edit something.
Using circuitpython/blinka doesn’t look to bad so maybe Ill just forget the c++ for a good while.
Adafruit kit behaves well IME, so I think that’s a good call. I don’t particularly like Python either - I hate the indentation rules - but I use it when I must. You can feed data into OEM very easily using any programming language via HTTP requests, so that’s not a problem when you come to it.
As you say, humidity is a bit of a black art. Once you get a system up and running, you’ll get some experience is normal in your environment and should be able to identify any problems just by virtue of them being different to the rest. So getting accurate results isn’t all that important.
