Best practices for monitoring sensors within building assembly
Hello, now that I am getting better at utilizing control layers in my construction, what are best practices for where to place monitoring sensors in building assemblies? I have no aspirations of doing research to write scientific papers. I will leave the experts to that, but I do want to learn from my own buildings. I do not get much of a chance to unbuild them for information. How and where can I place sensors in the building so the assemblies may “Talk to me” to see if they are doing what I want them to do. This will allow me to get information specific to the locations I tend to build in. Can I teach the homeowner what to notify me of when a monitor reaches a critical threshold and before expensive damage occurs to their structure? Think moisture trapped in a roof assembly as an example. Is it cost effective to place sensors in a building assembly for the purpose of preventing damage or identification of design flaws that might be remedied before they become design catastrophic failures? Will you be considerate enough to make specific recommendations for homeowner grade monitoring equipment that may not be adequate for scientific research but will help me learn or may tip me off to possible problems? It you don’t feel comfortable promoting specific brands or pointing out the best bang for my buck for this application, can you at least recommend what degree of accuracy you feel is recommended so I don’t waste my money and time collecting unreliable data and forming false conclusions?
Thank you in advance.
Andy
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
This question comes up periodically. There are few major points to keep in mind:
1- Most homeowners will either not check embedded sensors at all, ever, or they'll quickly tire of it after the novelty wears off.
2- Unless you install sensors EVERYWHERE, you can't rely on them to catch everything. An example is a leak in one or several stud bays in a wall. Murphy's law dictates that your sensor(s) will always be in the NEXT BAY OVER from wherever the leak is. Mr Murphy is a very devious guy after all...
If you do want to put some sensors in, make sure they go in spots most likely to see trouble. This is the same concept as putting smoke detectors at high points where smoke will collect first, and heat sensors near things that might overheat prior to starting a fire. I would put moisture sensors in a roof out near the eaves, and above difficult to vent areas (rafter bays over a dormer, for example). Those are places most likely to see issues early. You'll have to think about your walls, but I'd suggest areas around trim details that are most likely to leak, probably things like window and door frames.
Bill
Thank you for your reply. Do you have any idea of actual hardware to buy?
Andy
I am doing the same because I become extremely ill from small amounts of mold. I am thinking more for fixture leaks though. For example, we are putting plumbing access panels behind all the fixtures. And also some moisture sensors, although haven't chosen a brand yet. I've never heard of something that could monitor condensation. Even in the Build America test beds they have to deconstruct the assembly to see where there is water and rot. The humidity in the air will be high around such an area though so maybe a humidity sensor could help if there is a cavity?
Yes, if I can not learn from what my buildings are trying to tell me then how can I get better? I would rather rectify problems before there is to damage my work. I think monitors only make sense even if I do not have occupant buy in. I would like to offer my own monitoring service at my expense. It would answer questions I have whether I am doing a good job and doing things right. There is no worse feeling than doing what one thinks is an excellent job only to find out this was not the case. I think the initial cost and effort would pay dividends but this question sadly did not get many bumps and I don’t think I have the technological knowledge to set it all up in a cost effective manner. I have very little experience in this. I know people are out there that do have experience but how I can find them evades me.
There are humidity sensors out there. Honeywell and Humirel make sensors I'm familiar with. The voltage output type are easier to interface, since you just need an analog input that can read a varying voltage. I have used both types, but if you want an easy one to use I'd use this one:
https://sensing.honeywell.com/honeywell-sensing-hih4010-4020-4021-series-product-sheets-009020-1-en.pdf
It is a bit shy of $17 in small quantities. Most humidity sensors of reasonable reliability are expensive, unfortunately. Note that some sensors won't necassarily read correctly at the extremes (usually within 5% or so of the low or high limits of their range), so you don't want to set an alarm trip point at something like 97%, for example.
I found ONE "moisture" sensor, but no data sheet. If I needed to monitor for moisture, I would probably just build something. The easiest would be to etch a small PC board with two (or more) closely spaced traces, and have the board made with the ENIG finish, which is a gold plated finish over the copper. Gold won't corrode like other materials, so it will improve reliability of the sensor. You'd then use a regular water sensor that would be detecting "condensation" when water droplets formed and bridged the closely spaced traces on the PC board together. I have not tried this myself, but I'd expect it to work. You'd want to characterize the sensor though so that you knew how much "condensation" would actually have to be occuring before the sensor would trigger.
Another option would be to make a sensor like I described, then place it facing the sheathing with an insulating membrane that was able to dry out. This would detect moist sheathing, but would "reset" when it dried out (which could take a while). an "insulating membrane" might be something as simple as Tyvek, or rip-stop nylon like is used for tents and kites. You want something that moisture could wick into, trip the sensor, but still dry out in a reasonable period of time without degrading.
If you want to read the mositure content of wood, you can use electrodes (nails :-) driven into the wood, then monitor resistance changes between them. Lower resistance = higher mositure content in the wood. Ideally you'd monitor the resistance using an AC voltage, not DC, to avoid galvanic issues that would corrode the electrodes. Using AC complicates the monitoring circuitry though, and if you try building something like this, I would advise capacitively coupling to BOTH electrodes, driving them with a low frequency signal in the audio range that is not harmonically related to either 50 or 60 Hz line frequencies (I think I used 445 Hz, but it was a long time ago when I tried doing this, I was sensing water levels in a lake), then monitor for chaning CURRENT, not voltage, in the sensor circuit. The way I did it was to digitally create a very, very clean sine wave at the correct voltage and something like 1V RMS, then monitor the current. The current was sampled many times over several cycles of drive waveform, then the RMS value was calculated in a simple DSP algorithm. From this info, you can RELIABLY monitor mositure content of something, but it takes some effort to implement.
Bill
I think that what others have done is to embed electrodes in wood, and then have leads come out to connect to a commercial moisture meter, rather than a homebrew circuit.
We'll be discussing this topic on The BS + Beer Show on March 11, with guest experts Kohta Ueno, Ben Bogie and Dan Kolbert.
Awesome info above and also looking forward to the BS and Beer show. I always enjoy listening when Kohta speaks because he cuts to the chase and give practical advice I can use. Everything I have picked up from him either 1st or 2nd hand was solid reproducible science. Nothing anecdotal.
Thanks,
Andy
I've used a product from Govee. Measures temp and RH. They are super easy to use and will automatically pull historical data for viewing on your phone or downloading to a spreadsheet. About $15/ea. However the batteries are not replaceable. I've had mine running for more than a year, but they won't run for ever.
https://www.govee.com/products/107/smart-thermohygrometer