Evaluation of a heated inlet to reduce humidity induced error in low-cost particulate matter sensors

The number of low-cost particulate matter (PM) sensors available on the market has increased in recent years. They are affordable, compact and require no maintenance which means they easily can be deployed and utilized for real-time air monitoring. However, the accuracy of low-cost PM sensors has been reported to vary depending on several factors. This thesis aims to investigate and reduce errors induced by relative humidity. Several hardware methods to reduce the relative humidity of air were considered. These include heating, diffusion, dilution and membrane drying. Heating was chosen as the most suitable method to use with low-cost PM sensors because of low maintenance, small size and low price. The chosen concept was developed into two identical prototypes consisting of a heated pipe that was attached to two Alphasense OPC-N2 sensors. To evaluate how the sensors behaved with varying humidity with and without the prototype, tests were performed in an urban environment (Malmö, Sweden). Tests in a controlled aerosol laboratory were performed as well (LTH, Lund, Sweden). The results of the urban field test show that, when compared to the reference instrument, the native OPC-N2 sensors are susceptible to relative humidity errors especially when the RH goes above 75%. Moreover, constant offset between the two identical sensors was observed. Attachment of the heated inlet prototype improved the correlation between the OPC-N2 sensors and the reference instrument. The results of the laboratory tests were inconclusive, possibly because of the majority of the generated particles being outside the specified detection range for the low-cost PM sensors.Particulate matter (PM) sensors are very helpful in determining the quality of the air we breathe in. They measure the amount of small and often harmful particles suspended in the air. Cheaper particulate sensors have recently been introduced to the market. These sensors, however, have shown to be susceptible to measurement errors when the relative humidity of the air is high, which is often the case in the Swedish climate. By attaching a heated pipe to the inlet of a low-cost particulate sensor, the relative humidity of the air measured by the sensor was lowered. Our tests showed that this led to a reduced humidity-induced error in the sensor resulting in more accurate PM readings. This study strengthens the knowledge towards the possibility of deploying multiple low-cost PM sensors in order to collect more accurate and high resolution air quality data in a cost-efficient way