Air Quality Monitoring

I have recently had the second natural gas leak in my crawl space and concurrently discovered a mold spore concentration of 40,000+ ppm (normal approaches 0 - this was wear your hazmat suit territory). I am now properly paranoid and want to continuously monitor for natural gas and mold, store the data for trend analysis, and alert when a problem is detected.

The major project components are sensors and communications. The monitoring station will be in a fixed location with access to 110v AC so providing power or stations size is not an issue.

Methane seems to be the way to detect natural gas. Mold spores range from 1 to 100 microns so a particle sensor is the most likely way to detect them. Temperature and humidity are the major factors in mold growth so tracking them is also useful.

The available sensors can also track VOC, CO, and other pollutants. These are not essential to the current project but are useful to track if available. As a side note I would also like to track Radon but don't see anything available to do so. The topics suggest that this is a non-trivial problem and there are commercial solutions available.

The data needs to be stored. Alarms generated and transmitted when gas concentrations exceed stated limits. I'm planning an IOT approach but open to suggestion. There is strong WiFi available and a robust and protected local network.

I am looking for general guidance on hardware. All of the sensor packages offer sample code and wiring so I'm not as concerned about the software at this point.

I've looked at the threads for Air Quality and the most recent are still years old and the bulk are 8 to 10 years old. The hardware recommended is (hopefully) obsolete and much better is now available.

I'm also not opposed to buying a commercial product but have not found any that meet the project objectives.

Kidde make detectors that comply with (and exceed) residential building code.

Mold spores indicate lots of moisture, so prepare to protect all electronic stuff from moisture. That includes sensors which means they may only work properly for a short time. A much easier solution to all of the stuff you mention is to permanently mount a fan to blow the moisture out of the crawl space.

1 Like

Adafruit sells the PMSA003 particle sensor, which works pretty well to report PM2.5. At least, readings agree reasonably well with AQI PM25 values reported for my area, and others have made similar comments.

However, it has been reported by an independent laboratory that the sensor cannot detect particles > 3 um in size. It reports detection frequencies for larger particles, but presumably, these are just estimates, based on distributions of particle sizes for typical cases of air pollution.

https://www.researchgate.net/publication/320555036_Particle_Distribution_Dependent_Inaccuracy_of_the_Plantower_PMS5003_low-cost_PM-sensor

Kidde has a variety of products on the web site. All are some combination of Smoke, CO, and VOC but none have particulate or Radon. All connect to the Kidde app and will provide remote reporting and alarm. As an additional paid service they have an agent that reviews the temp/humidity profile to warn of mold likelihood. The products are in the $100 to $125 range and the additional service is $20/year.

Thanks for the suggestion but I would still have to add a particulate material (PM 1) sensor and a methane sensor for the natural gas. No Joy.

https://www.homedepot.com/p/Kidde-Firex-Plug-in-Carbon-Monoxide-Propane-Natural-and-Explosive-Gas-Detector-9-Volt-Battery-Backup-Digital-Display-21029623/301627165

Agreed, this is a grade level dirt floor crawl in NJ where the water table is high and basements are uncommon. The mold problem was caused by a poorly executed waterproofing job done 15 years ago that trapped ground water and water seeping in through the cinder block wall under a layer of plastic. This created a mold farm.

We have just completed mitigation and installed commercial grade de-humidification to prevent the problem from coming back. Longer term probably involves a more substantial water barrier than the plastic. This would either be a total foam spray or a concrete floor. Both are expensive and painful. In the short term I want to monitor conditions in the hope that the current "fixes" solve or prevent the problem.

The space is ventilated by two "arrow slit" spaces with metal grates. To a limited extent forced ventilation works in the summer although the relative humidity is often above 60% and hits 80%. To prevent mold you have to be between 30 and 50% so de-humidification is required. A correctly sized unit can keep the humidity around 40 to 45%. Providing outside air only worsens the problem. In the winter the air is cold and the humidity is low but the heating ducts and water pipes run through the space. Importing cold air means that all the heating equipment has to work harder and there needs to be a heater in the crawl to keep the pipes from freezing. Providing increased ventilation is not a workable solution during most of the year.

The reality is that I'm stuck with sealing and de humidifying the space. An accurate monitoring system with data logging seems the best way to keep track of what is going on and prevent future problems.

Thank you for the link. I hadn't found that unit on the Kidde site when I looked. The unit tests for natural gas, propane, and CO. It has no remote reporting capability so I would need to go to the crawl periodically to check on it and the alarm might not be audible elsewhere in the house.

Someone has an IOT project where they tapped into the LED display to provide remote reading/alarm capability so this has possibilities. I will keep this in mind if the build becomes too much of a problem.

Thanks for the info. I'm hoping to use a PM1 sensor but may couple that with a PM2.5 or PM10.

How did you find the research paper?

You are missing the point that any accurate monitoring system will soon become inaccurate because of the environment! And you will either be given false data or will not be given data at all. If you can work with that, then go ahead and good luck.

Paul,

I do understand that electronics degrade in hostile environments.

The crawl space itself was not a hostile environment. The humidity maintained at 45%. The problem was that a poor waterproofing design trapped ground moisture against the walls under a plastic barrier. The mold grew behind the plastic but the released spores did not sprout inside the envelope. As evidence there was no mold on wine bottles which had been stored for up to 20 years (the space was cool to cold and dark).

I can work with a considerable error range. The target is to maintain a RH of 50% or lower. Mold is not generally a problem below 60%. I can have a +/- 10% error and still be OK because the range on a 50% RH reading would be 45% to 55%.

Prior to the mitigation I had a residential de humidifier and now I have a much larger commercial unit in the same space. Independently of this project I have a temp/rh remote reading meter as part of a weather station that lives in the space. If the two start reporting widely different values I know there is a problem.

Thanks for ensuring I knew what your concern was and I hope I've taken it into account.

The very best you can do, then, is not allow temperature changes such that moisture condenses on any metal parts of your system.

Hi, @glennj0158

You don't have to put any electronics in the crawl space.

Simply have a pump that pulls the air down to your sensor.
You do not need continuous 24 hour sampling.

Tom.... :smiley: :+1: :coffee: :australia:

You are right and I had not thought of moving the air to a different location. It would also allow me to use the Kidde unit without going to the crawl.

I now have a usable solution for protecting the electronics and, if necessary, locating a commercial unit in an accessible location.

Back to the original request, recommendations for PM1, PM 10, Methane, Temp and Humidity sensors, LAN communications, data storage and alerts. Thanks to everyone and Paul in particular for participation to date and solving a potential problem I hadn't considered.

Web search.

I'm hoping to use a PM1 sensor

The Plantower sensor sold by Adafruit estimates particle sizes and reports counts in size ranges, including 1 um:


  Serial.print(F("Particles ~ 0.3um / 0.1L air:")); Serial.println(data.particles_03um);
  Serial.print(F("Particles ~ 0.5um / 0.1L air:")); Serial.println(data.particles_05um);
  Serial.print(F("Particles ~ 1.0um / 0.1L air:")); Serial.println(data.particles_10um);
  Serial.print(F("Particles ~ 2.5um / 0.1L air:")); Serial.println(data.particles_25um);

BTW it uses a fan to move air, so you could use a tube to sample the crawl space.