Evaluation of Carbon Dioxide Sensors for Indoor Air Quality Monitoring

Abstract

Carbon dioxide (CO2) is important factor of indoor air quality. In this paper we evaluate several CO2 sensors of latest generation. All sensors work on the same principle: nondispersive infrared (NDIR) detection. The sensors were evaluated in the campaign that lasted for several months. Good agreement between all sensors was observed, despite different construction, factory calibration and price. Finally, inexpensive method is proposed for continuous measurement of indoor concentrations of CO2.

Full text

30TH DAAAM INTERNATIONAL SYMPOSIUM ON INTELLIGENT MANUFACTURING AND AUTOMATION
DOI: 10.2507/30th.daaam.proceedings.061
EVALUATION OF CARBON DIOXIDE SENSORS FOR
INDOOR AIR QUALITY MONITORING
Adnan Masic, Dzevad Bibic, Boran Pikula & Sanela Klaric
This Publication has to be referred as: Masic, A[dnan]; Bibic, D[zevad]; Pikula, B[oran] & Klaric, S[anela] (2019).
Evaluation of Carbon Dioxide Sensors for Indoor Air Quality Monitoring, Proceedings of the 30th DAAAM International
Symposium, pp.0457-0459, B. Katalinic (Ed.), Published by DAAAM International, ISBN 978-3-902734-22-8, ISSN
1726-9679, Vienna, Austria
DOI: 10.2507/30th.daaam.proceedings.061
Abstract
Carbon dioxide (CO2) is important factor of indoor air quality. In this paper we evaluate several CO2 sensors of latest
generation. All sensors work on the same principle: nondispersive infrared (NDIR) detection. The sensors were evaluated
in the campaign that lasted for several months. Good agreement between all sensors was observed, despite different
construction, factory calibration and price. Finally, inexpensive method is proposed for continuous measurement of indoor
concentrations of CO2.
Keywords: Carbon dioxide; Sensors; NDIR; Indoor air quality; Calibration.
1. Introduction
Measurement of concentration of carbon dioxide (CO2) in the air is very important in many areas: from food
production to the indoor air quality assessment and control [1], [2]. In the era of massive automatization and IoT (Internet
of Things), having a good technique for CO2 determination is a necessity [3]. Historically, CO2 sensors first appeared in
the 1950’s [1]. There are several principles of work of CO2 sensors: some measure pH-value (indirect and possibly
inaccurate), some use gas chromatography and mass spectrometers (very expensive) etc. In this paper we will focus on
NDIR type (nondispersive infrared) of CO2 sensors, since we believe that they are the best choice for automatic control,
IoT and many other purposes. Figure 1 shows the principle of work of NDIR CO2 sensor [2].
Fig. 1. Principle of work of NDIR sensor [2]
- 0457 -

30TH DAAAM INTERNATIONAL SYMPOSIUM ON INTELLIGENT MANUFACTURING AND AUTOMATION
Since CO2 has resonant frequencies in infrared (IR) part of the spectrum, construction of the NDIR CO2 sensor can be
based on low-cost IR diode as source and some type of optical detector. Together with electronics and gas inlet (or
diffusion membrane), entire sensor is usually small (few cm in size) and very affordable.
Calibration of CO2 sensors is usually performed by factory, which may be good enough for some applications. But in
many cases professional (and potentially expensive) individual sensor calibration is required [4], [5]. There are dual
channel NDIR CO2 sensors where one channel contains reference gas (usually nitrogen). Such sensors should be more
accurate in the long run. The aim of this paper is to evaluate performance of several commercial CO2 sensors, with focus
on one question: are they good enough for indoor air quality assessment?
2. Sensors
We have acquired several NDIR sensors, and they are described in Table 1. Amphenol Telaire is a single channel
sensor with patented factory calibration algorithm, which requires periodical exposure of the sensor to the outdoor levels
of CO2 (about 400 ppm). Plantower offers dual channel sensor for similar price, while Alphasense offers cell of the sensor
and transmitter board separately. Price of the Alphasense sensor and the board is much higher than Amphenol and
Plantower sensors. We should also note that Alphasense board doesn’t provide digital signal (except USB), which means
that additional circuitry is necessary to translate DC current into digital signal, if the sensor is connected to
microcontroller, such as Raspberry Pi or Arduino. All sensors were successfully integrated into our in-house developed
ecosystem for data acquisition, which was primarily developed for use in drones and later adapted for other applications
as well [6], [7], [8].
Sensor
Amphenol Telaire 6703
Plantower DS-CO2-20
Alphasense IRC-A1 (with board)
Type
Single channel NDIR
Dual Channel NDIR
Dual Channel NDIR
Range
2000 ppm
5000 ppm
5000 ppm
Interface
Digital: UART and PWM
Digital: UART and PWM
Analog: 4-20 mA and USB
Table 1. Description of sensors
3. Results
The sensors were collocated in the Physics laboratory at the Faculty of Mechanical Engineering in Sarajevo for several
months, during the summer semester 2019. The room has area of about 80 m2 and height 4 m, giving total volume of 320
m3. It was periodically occupied with groups of 25 students (and 2 staff members).
Fig. 2. Results for periodic occupation of the room
0
500
1000
1500
2000
2500
CO2 (ppm)
Alphasense
Amphenol_UART
Amphenol_PWM
DS-CO2-20
- 0458 -

30TH DAAAM INTERNATIONAL SYMPOSIUM ON INTELLIGENT MANUFACTURING AND AUTOMATION
Figure 2 shows readings from two Amphenol, one Alphasense and one Plantower sensor for a period of few days where
the laboratory was occupied three times, corresponding to three peaks on the graph. We can see very good agreement
among sensors. Alphasense was giving slightly lower values, while Plantower was giving highest readings, but the
difference is acceptable. Peak values of CO2 concentration exceeded 2000 ppm.
Figure 3 shows fast build up and slow decay (without forced ventilation) some days later. Concentration of CO2 reached
3000 ppm. We can see some noise (spikes) in the Figure 3, which is a consequence of direct sunlight reaching the sensors,
which must be prevented for valid measurements. Again, agreement among sensors is good, with Alphasense giving lower
values.
Fig. 3. Fast build up and slow decay of CO2 concentration
5. Conclusion
CO2 measurements in this study were successful with surprisingly good agreement among sensors (without any
external calibration). Thus we can conclude that the latest generation of low-cost CO2 sensors is good enough to be used
for indoor air quality assessment. We recommend dual-channel CO2 sensors with digital interface (UART or PWM). It is
difficult to recommend Alphasense CO2 sensor and transmitter board for IoT applications, since it is expensive, bulky,
and analog-interfaced. The range of 2000 ppm (default value for many commercial CO2 sensors) is not sufficient for all
indoor applications, and we recommend at least 5000 ppm range. Proper enclosure for sensors is required, since the direct
sunlight affects the measurements. As a suggestion for future work, we will try to measure CO2 outdoors, which will be
much more challenging task. Network of multi point CO2 sensors connected over the internet is yet another proposal for
future work.
6. References
[1] Neethirajan, S., Jayas, D. S. & Sadistap, S. (2009). Carbon dioxide (CO2) sensors for the agri-food industry – a
review. Food and Bioprocess Technology, 2:115–121, 2009, DOI: 10.1007/s11947-008-0154-y
[2] Mendes, L. B. et al. (2015). NDIR Gas Sensor for Spatial Monitoring of Carbon Dioxide Concentrations in Naturally
Ventilated Livestock Buildings, Sensors, 15, 11239-11257, 2015, DOI: 10.3390/s150511239
[3] Yasuda, T., Yonemura, S. & Tani A. (2012). Comparison of the Characteristics of Small Commercial NDIR CO2
Sensor Models and Development of a Portable CO2 Measurement Device, Sensors, 12, 3641-3655, 2012, DOI:
10.3390/s120303641
[4] Cho, H. C. & Yi, S. H. (2010). NDIR CO2 gas sensor with improved temperature compensation. Procedia
Engineering 5:303-306, 2010, DOI: 10.1016/j.proeng.2010.09.108
[5] Kim, J.H., Park H. G. & Yi S. H. (2018). IR Absorbance as a Criterion for Temperature Compensation in
Nondispersive Infrared Gas Sensor. Proceedings. 2018, DOI: 10.3390/proceedings2130775
[6] Masic, A. (2015). Unmanned Aerial Vehicle as Data Acquisition System. Journal of Trends in the Development of
Machinery and Associated Technology. Vol. 19, No. 1, 2015, ISSN 2303-4009 (online), p.p. 181-184.
[7] Masic, A.; Musemic, R. & Dzaferovic-Masic, E. (2016). Temperature Inversion Measurements in Sarajevo Valley
Using Unmanned Aerial Vehicles, Proceedings of the 27th DAAAM International Symposium, pp.0423-0427, B.
Katalinic (Ed.), Published by DAAAM International, ISBN 978-3-902734-08-2, ISSN 1726-9679, Vienna, Austria,
DOI: 10.2507/27th.daaam.proceedings.062
[8] Masic, A.; Bibic, Dz.; Pikula, B.; Dzaferovic-Masic, E. & Musemic, R. Experimental study of temperature
inversions above urban area using unmanned aerial vehicle. Thermal Science, 2018, DOI:
10.2298/TSCI180227250M
0
500
1000
1500
2000
2500
3000
3500
CO2 (ppm)
Alphasense
Plantower
Amphenol_PWM
Amphenol_UART
- 0459 -