Multiple tests failed to validate the water mass conservation within sealed small air containers. The water content grows while the container is heated and reduce when cooled. How come ? How does the humidity capacitive sensor behave if a group of water molecule coalesce together but the group remains airborne? Is the temperature a truly dry bulb temperature? Where to look to get the water mass conservation right?
I am not a programmer nor an electronic technician, so I picked up an assembled SparkFun Atmospheric Sensor BME280 with Arduino prototyping board. I used an I2C connection and Adafruit program library on Arduino program running on a Mac. The data was treated through and Excel sheet using functions from ASHRAE Psychrometric analysis to find the specific humidity W, expressed in kg of water / kg of air. Various containers were used to discard any effect of adsorbtion or generation of unwanted gas through the test setup. Here is an example of a test result.
Solved! Go to Solution.
The theory of computational mass is reasonable and verified: with the same instrumentation and program, the computed water mass in vapor followed exactly the 100% saturation line when the vessel was cooled below the dew point, reflecting the fact that vapor condensate on the wall of the vessel (the only difference from the validation test above was the use of a very big unsealed vessel). The equation used is simple and was manually tested at first: W = 0.62198 * RH * Pws / (P - RH * Pws). RH and P are the humidity and barometric pressure as measured by BME280. Pws is the partial pressure of water at saturation computed using the temperature measured by BME280 according to ASHRAE handbook equation (I assume BME280 give the dry bulb temperature, not the dew point nor the wet bulb temperature).
Please use reference device to ensure the correct environment condition inside the container and also compare with the BME280.
From the information provided from user, sensor have hysteresis for humidity sensor and stabilization time is needed.
Suggest to use lab device to verify the sensor performance.
Thank you for the reply: I do not have access to external laboratory for the moment. For information, I used other sensors, less accurate thant BME 280, that showed a similar behavior. I suspect an explanation related to the condensation/evaporation of airborne water nuclei in the atmosphere. Dew precipitation only appears when there is a high number of water molecule coalescing (100 to 1000 molecules): what about very low molecule concentration?
What would be helpful is to know how BME280 humidity sensor behave when it is in contact with a small nuclei of condensated vapor (aerosol water liquid content): does it sense it like one water molecule, thus purely an indication of vapor partial pressure, or does it adsorbs the whole nuclei (2 and more H2O molecules) and thus indicate the whole airborne water content? From the test, I suspect the first hypothesis: BME 280 senses the small airborne nuclei of condensate as a partial vapor pressure and assimilate the signal as being one water molecule.