This topic popped up recently, so I thought it might be worth addressing in a short post. With the newfound prevalence of consumer-grade carbon dioxide monitors such as the Aranet4 Home and Vitalight Mini CO2 Detector, it’s natural that people question the accuracy and trustworthiness of these devices. In this short post, I aim to address this question and discuss why you can trust your carbon dioxide monitor - with a few caveats.
If you’ve researched CO2 monitors already, you will know that the gold standard for carbon dioxide sensors is non-dispersive infrared (NDIR) sensors. Other sensors, such as MOX (metal oxide) and electrochemical sensors, suffer from a range of issues that are out of the scope of this short post. The takeaway is that you want to look for a carbon dioxide sensor that uses an NDIR sensor.
There are two main types of NDIR sensors - transmissive and photoacoustic. While transmissive tends to be the better performer overall as it’s less susceptible to environmental factors, photoacoustic sensors are far smaller and are, therefore, quite popular in smaller devices. Both of these NDIR sensor types perform well, and the most common sensors have a stated accuracy of ± 3% ± 30 to ± 5% ± 50.
Of course, there are hundreds of different NDIR sensors on the market, and it’s impossible to make a blanket statement about them all. However, I have yet to come across an inaccurate NDIR sensor - even in the very affordable Vitalight Mini CO2 Detector.
This isn’t to say they don’t exist, as I’m sure they do, but NDIR sensors tend to be quite accurate. If you are still unsure, try to ensure the carbon dioxide monitor that you are considering has an NDIR sensor from a big brand such as Sensirion (SCD30, SCD40, SCD41, etc.), SenseAir (S8, S88, Sunrise, etc.), or even a brand such as Cubic. These sensors are by far the most tested in studies and have good reputations. Furthermore, they’re included in all kinds of monitors and devices, such as the Aranet4, which uses a SenseAir Sunrise.
If you want to ensure the device you purchase has an accurate carbon dioxide sensor, either check the spec sheet for a model number or contact the monitor manufacturer to ask which carbon dioxide sensor they use. Once you have this information, you can visit the sensor manufacturer’s website to check the stated accuracy - it will look something like this:
Of course, we shouldn’t take manufacturers at their word, and it’s important to look into third-party testing. Luckily, with consumer-grade CO2 monitors becoming increasingly common, there are a lot of studies which now investigate the accuracy of these monitors.
For example, this study found a range of consumer-grade air quality monitors and sensors, including the SCD40, Awair and AirVisual monitors (using the SenseAir S8) to all provide high correlations with reference equipment. There are also ample other studies investigating the accuracy of specific consumer-grade carbon dioxide sensors. I recommend looking them up if you’re unsure about the accuracy of the monitor you purchase.
While NDIR sensors from reputable vendors are accurate, it’s important to consider that there are caveats, and these sensors can lose accuracy under different conditions (air pressure, relative humidity, temperature, etc). While any quality NDIR sensor should remain accurate in normal conditions, if you live somewhere with extreme weather conditions, you will want to ensure you get an NDIR sensor that is tested and compensates for those conditions. If you’re interested in how these conditions can impact CO2 measurements, the following post is an interesting read.
On top of this, a discussion about calibration needs to be had. NDIR sensors (as with all pollutant sensors) are prone to drift, meaning they can lose accuracy over time. The good news is that this drift is easy enough to overcome by calibrating the sensor. The bad news is that some sensors have a function called ABC (automatic baseline calibration), which essentially means the sensor recalibrates itself every week, fortnight, month, or on any other schedule that the manufacturer sets.
While this ‘feature’ is used to combat sensor drift, it also means that if your CO2 sensor is never exposed to ambient air or left in a room that never reaches 420 ppm (approximately), the device will recalibrate itself, setting the lowest value it’s been exposed to over that period as the baseline. This leads to issues because if the lowest value the monitor is exposed to over a week is (for example) 600 ppm, the monitor will assume that 600 ppm is 420 ppm, skewing all future readings. This can be combated by a manual calibration done outdoors, and is fine if the monitor is in a room that is often empty, but it can cause issues in some situations. ABC leads many users to believe their monitor is inaccurate as it can lead to readings very quickly becoming very skewed.
Overall, modern NDIR CO2 sensors are very accurate if you know how to manage them and keep ABC in mind. If you have any questions, please feel free to ask. However, I hope this has helped clear up some worries about CO2 sensor accuracy!