Determining the emission impact of area sources (biofilters, wastewater tanks) on air quality and the environment by classic measurement techniques (i.e. static hood sampling), is currently lacking in pertaining uniform and representative emission data by being restricted in sampling area, time and safety. This creates an extra hindrance when emission rates from such sources need to be determined by the fluxwindow method, which implies measuring emission concentrations up- and downwind along different horizontal and vertical profiles of the area source.
In an effort to improve on this matter, the possibility of using a drone equipped with an emission detection laboratory (OLFASCAN Flying Lab) to quantify emission concentrations and rates via the fluxwindow method from a sludge buffer tank was investigated. The OLFASCAN Flying Lab is equipped with several electrochemical sensors for performing air quality measurements and was attached to a DJI Matrice600 PRO RTK drone.
N. De Baerdemaeker1, M-A. Vandenabeele2, K. Haerens1, S. Degryse3, T. Van Elst1*
Competing interests: The author has declared that no competing interests exist.
Academic editor: Carlos N. Díaz.
Content quality: This paper has been peer-reviewed by at least two reviewers. See the scientific committee here.
Citation: N. De Baerdemaeker, M-A. Vandenabeele, K. Haerens, S. Degryse, T. Van Elst, The use of chemical drone sensor technology in assessing emission rates and deposition from area sources, 9th IWA Odour& VOC/Air Emission Conference, Bilbao, Spain, www.olores.org.
Copyright: 2022 Olores.org. Open Content Creative Commons license . It is allowed to download, reuse, reprint, modify, distribute, and / or copy articles in olores.org website, as long as the original authors and source are cited. No permission is required from the authors or the publishers.
ISBN: 978-84-09-37032-0
Keyword: OLFASCAN Flying Lab, odour, volatile organic compounds, sampling techniques, meteorology, fluxwindow, modelling.
Video
Abstract
Determining the emission impact of area sources (biofilters, wastewater tanks) on air quality and the environment by classic measurement techniques (i.e. static hood sampling), is currently lacking in pertaining uniform and representative emission data by being restricted in sampling area, time and safety. This creates an extra hindrance when emission rates from such sources need to be determined by the fluxwindow method, which implies measuring emission concentrations up- and downwind along different horizontal and vertical profiles of the area source. In an effort to improve on this matter, the possibility of using a drone equipped with an emission detection laboratory (OLFASCAN Flying Lab) to quantify emission concentrations and rates via the fluxwindow method from a sludge buffer tank was investigated. The OLFASCAN Flying Lab is equipped with several electrochemical sensors for performing air quality measurements and was attached to a DJI Matrice600 PRO RTK drone. Meteorological data was locally assessed via an installed meteorological tower on a height of 10 meters. Via an air suction pipe of 8 meters in length to nullify the effect of drone downdraft on the emission readings, air was sampled in fluxwindows over the tank at both the upwind and downwind side. Results indicated that emission concentrations are mapped in detail by the drone sensor technology, allowing for a reliable calculation of emission rates by the fluxwindow method. By modelling the emission rates of the sludge buffer tank, the deposition on the environment could also be calculated. Results of the research show that chemical drone sensor technology can aid in pertaining representative emission data from area sources that are often too restricted, unsafe and time-consuming for classic measurement methodologies.
1. Introduction
To this day, sampling remains one of the main problems to obtain representative and quantitative data from emission and odour sources (Capelli et al., 2013). Especially area emission sources prove to be more difficult as the main sampling techniques are based on the principle of evaluating a fraction of the conveyed air flow (Gostelow et al., 2003; Bockreis and Steinberg, 2005).
Area sources can be divided into two categories: (i) active area sources with outward air flow like biofilters, and (ii) passive area sources without outward air flow like wastewater tanks (Gostelow et al., 2003; Capelli et al., 2013). Independent of the type of area source, hood sampling techniques are by far the most widely used and accepted technique to capture and estimate odour and emission concentrations from area sources (Capelli et al., 2013; Lucernoni et al., 2016). The hood technique isolates a section of the area source and collects odour and emissions from the hooded surface by extracting the closed-off air via the depression pump sampling technique for point sources (Capelli et al., 2013; Lucernoni et al., 2016). In case of passive area sources, this technique requires to additionally blow neutral air over the hooded surface to create a wind tunnel effect (Lucernoni et al., 2016). Though reliable emission results can be obtained, is the technique exhaustive when it comes to completely mapping emission and odour concentrations from the entire area surface.
To calculate emission rates from area sources, the emission concentrations need to be mapped in both a horizontal an vertical profile along the windward and leeward side of the area source, i.e. fluxwindow method (Baumbach et al., 1999). This technique requires methods that can determine emission concentrations of area sources on different heights (Baumbach et al., 1999), for which the classical hood methodology is not properly equipped for.
In order to determine emission concentrations and rates from area sources in an effective, accurate and fast way, a flexible and mobile measurement device that allows airborne quantification, i.e. drones, is required (Bartholmai and Neumann, 2010). Recent developments in drones (UAV) and chemical sensing technology opened this domain as chemical sensor drones are being widely used in industrial emission monitoring applications (Burgués and Marco, 2020).
In this paper, OLFASCAN and Drone Division build on this emerging state-of- the-art chemical sensing drone technology by testing the accuracy of this technique and by exploring new methodological opportunities by utilizing the fluxwindow method to determine emission rates and deposition impact from area sources.
2. Materials and methods
2.1. Experimental design
For the purpose of this research an open sludge buffer tank of a waste water treatment plant in Ghent, Belgium was made available. The experiment was conducted on March 30 2021. Next to the tank, a meteorological tower was installed. The sensor unit of the tower was installed at a height of 10 meters, and allows for ultrasonic detection of wind speed, three vectors of wind direction, temperature, relative humidity and solar radiation in 30 second intervals. Based on these parameters, atmospheric stability is also directly calculated per 30 second interval.
2.2. OLFASCAN Flying Lab
The OLFASCAN Flying Lab is equipped with 5 electrochemical sensors (Alphasense Ltd.), of which three are interchangeable with other sensor types. For the purpose of this research, following sensors were present in the Flying Lab:
-
H2S – low range (3 ppb – 3 ppm)
-
NH3 – low range (5 ppb – 10 ppm)
-
NH3 – high range (3 ppm – 100 ppm)
-
CH4 (10 ppm – 20 000 ppm)
-
PID 10.7 eV (5 ppb – 50 ppm)
The sensors are installed in a sealed-off chamber, and sample air is sucked via an 8 meter long inert PVC suction pipe into the chamber via a built-in pump with a maximum flow capacity of 4.5 litre per minute (Figure 1). The emission data is lively monitored via the cloud software (20 second sample interval), and each emission point is accompanied by a GPS- and timestamp. The Flying Lab is also equipped with a temperature and relative humidity sensor, and possesses the capability of collecting air in sample bags for further analysis. The OLFASCAN Flying lab was attached to a DJI Matrice600 PRO RTK drone piloted by Drone Division (Figure 1).
Fig. 1.: (A) Setup OLFASCAN Flying Lab on DJI M600Pro drone, (B) in-flight air sampling via 8 m long suction pipe
2.3. Data processing
The emission data collected with the Flying Lab is typically processed in ARCGIS software so that the emission measurements can be visualised on topographic maps, indicating at which GPS location the different concentrations were measured.
By incorporating meteorological data, emission concentrations of area sources measured with the Flying Lab can be converted into emission flows via the fluxwindow method. The technique consists out of constructing two imaginary fluxwindows, one upwind and one downwind of the area source. By measuring the emission concentrations on different heights and lengths along the two fluxwindows, the difference in emission concentration can be determined. Multiplying with the flow rate across the fluxwindows derived from the determined meteorological windspeed, yields the emission rate of the source. The obtained emission rate was then modelled with the meteorological data of 2012 for Flanders, Belgium with the Flemish IMPACT model (version 3.3.12), to determine the year average deposition on the environment.
3. Results and discussion
3.1. Results
The averaged meteorological parameters (± standard deviation) determined by the meteorological tower over the course of the fluxwindow experiment (26 minutes in total), illustrated that the dominant wind direction was southeast, with an average windspeed of 2.36 m.s-1 (Table 1, Appendix – Figure A1).
Table 1: Averaged meteorological parameters (± standard deviation) over course fluxwindow experiment
Based on the dominant wind direction, two flux windows were artificially made at the down- and upwind side of the sludge buffer tank (Appendix – Figure A2). The distance between both windows was 20 m. The fluxwindow grid at both sides consisted out of 3 x 7 points, with the 3 points equally taken at a distance of 1 m over a total width of 3 m, and the 7 points taken at a distance of 1 m over a total height of 5.85 m, with the exception of the first point taken at a height of 0.15 m above the water level of the sludge buffer tank. At each point of both grids, the NH3-concentration was measured (Table 2).
Table 2: Averaged NH3-concentrations ± standard deviation (ppb) fluxwindow grids sludge buffer tank
Averaging the NH3-concentrations over both flux windows attained to an average NH3-concentration of 1.56 ppm upwind, and 2.15 ppm downwind. The NH3- fluxconcentration was calculated as the difference between the down- and upwind concentration and was equal to 0.59 ppm. Based on the average windspeed and the surface area of the fluxwindow grids, a flow rate of 149 105 m³.h-1 was obtained. Incorporating the measured air temperature and relative humidity with the determined flow rate and average NH3-fluxconcentration, a NH3-emission rate of 0.062 kg.h-1 was calculated.
As stated in the Material and methods section (section 2.2), the Flying Lab was also equipped with several other sensors including a H2S-, CH4- and PID-sensor. During the fluxwindow method, no H2S-emissions were registered from the sludge buffer tank. Though CH4-concentrations were high, the variability in both PID- and CH4-emissions along the fluxwindow grids was minimal. The total average PID- concentration was equal to 0.60 ± 0.02 ppm and the total average CH4-concentration was equal to 9 627 ± 130 ppm.
For modelling the N-deposition impact of the sludge buffer tank on the surrounding environment, which includes a VEN (Flemish Ecological Network) area, the calculated NH3-emission rate of the tank was processed in the Flemish IMPACT model (Appendix – Table A1). The model calculates the N-deposition impact as year averages in kg N.ha-1.year-1 (Figure 2).
Fig. 2.: N-deposition impact (kg N.ha-1.year-1) sludge buffer tank on surrounding VEN-area
The results indicate that the NH3-concentrations of the sludge buffer tank are responsible for a N-deposition of max. 2,9 kg N.ha-1.year-1 in parts of the VEN-area that are close to the position of the sludge buffer tank.
3.2. Discussion
The fluxwindow method requires the mapping of emission concentrations of area sources in a horizontal and vertical profile (Baumbach et al., 1999; Capelli et al., 2013). Baumbach et al. (1999) investigated NO2-fluxes of a city, which required a flying device, i.e. tethered air balloons and planes, in order to map the vertical NO2- concentrations. The results showcased the precise estimation of emission rates via the fluxwindow method, but only when significant amounts of emission concentrations are measured along the horizontal and vertical profile of the fluxwindow (Baumbach et al., 1999). Classical handheld measurements are however restricted in accessibility, time and safety to map vertical and horizontal emission concentrations of area sources, and for smaller area sources such as biofilters and waste water treatment tanks, flying devices like planes and tethered air balloons are practically unusable (Toscano et al., 2011; Villa et al., 2016).
The emerging development in drone technology provides a mobile measurement device for airborne quantification that can overcome the dangerous and physical limitations of pollution sources (Bartholmai and Neumann, 2010; Villa et al., 2016), allowing for a reliable and quantitative assessment of emission concentrations in horizontal and vertical profiles when drones are equipped with chemical sensor technology (Burgués and Marco, 2020; Cao et al., 2020). Such an assessment however requires the nullification of downdraft effects on the sampled air (Villa et al., 2016), which was accomplished in this study by sampling the air above the sludge buffer tank via a 8 meter long air suction pipe centred below the drone rotors (Figure 1B; Burgués and Marco, 2020). Because the suction pipe is made from inert PVC material, and cleansed after measurements, the risk of H2S and NH3 adsorption is strongly reduced (Burgués and Marco, 2020)
Application of the fluxwindow method by drone and sensor technology in this study thus allowed for mapping emission concentrations at sufficient amounts of points alongside the constructed fluxwindow grids. The drone controls additionally allowed for steady measurements at equal distances alongside the grids, illustrating the added value of drones in the use of emission rate calculations of area sources via the fluxwindow method. Ideally, the fluxwindow experiment should be executed over a longer period of time (1 to 2 hours in total) to better account for variability in the measured NH3-concentrations along the fluxwindow grids (Table 2). The execution of the fluwxindow experiment here was restricted in time as the experiment was part of a larger emission mapping study in which the emissions of all of the waste water treatment installations of the site were measured. Further research is also required to determine the feasibility of the fluxwindow method when the emission impact of very large surface areas like pools need to be investigated.
The study results however did indicate that the effect of wind swirls inside the sludge buffer tank on emission concentrations did not significantly influence the overall NH3-emission rate. The water level of the sludge buffer tank was at the moment of measuring 1.3 m high, while the edges of the tank are 4.5 m high. To compare the effect of wind swirls, the NH3-emission rate was also calculated based on the flux window data points that were taken above the edge of the sludge buffer tank (height 4, 5 and 6 – Table 2). The average NH3-concentration up- and downwind was respectively 0.60 ppm and 2.22 ppm, attaining to a NH3-fluxconcentration of 1.62 ppm. Based on the average windspeed and the surface area of the fluwxindow grids, a flow rate of 50 976 m³.h-1 was obtained, which translated to a NH3-emission rate of 0.058 kg.h-1.
To investigate the reliability of the measured NH3-concentrations and the calculated NH3-emission rate, the NH3-concentration was additionally measured alongside the borderline of the VEN-area and the waste water treatment plant with the Flying Lab (Figure 3). The NH3-concentration was measured at a height of 8 m above ground level, which coincided to the maximal height of the fluxwindow grids above ground level.
Fig. 3.: NH3-concentration (ppb) borderline VEN-area – waste water treatment plant
Following the dominant southeast wind direction, increased NH3-concentrations were measured at the borderline in a straight line opposite from the sludge buffer tank. The average NH3-concentration in this increased concentration zone was equal to 0.43 ppm and the highest value equal to 2.06 ppm, which, taking dilution effects into account, is in line with the NH3-concentrations measured at the downwind fluxwindow of the tank. These results verify that the modelled N-deposition values in the VEN-area can indeed be related to the NH3-emission rate of the sludge buffer tank.
According to the modelled results, the sludge buffer tank is responsible for a N- deposition of max. 2,9 kg N.ha-1.year-1 in parts of the VEN-area that are closest to the position of the tank in a southwest wind direction (dominant wind direction in Belgium). The critical N-deposition value for nature in Flanders, which indicates the amount of deposition a vegetation can tolerate without being damaged, varies between 6 (very sensitive) to > 34 (not sensitive) kg N.ha-1.year-1. For the VEN-area, the critical N- deposition of the vegetation is estimated at 20 kg N. ha-1.year-1 (Departement Omgeving, 2021). The measured N-deposition at locations in the VEN area in closest proximity to the sludge buffer tank varied between 21,3 and 29,5 kg N. ha-1.year-1. (VMM VLOPS20, 2017), implying that the sludge buffer tank alone can contribute up to 10-15% of the allowed N-deposition in these parts of the VEN-area.
Results of the study thus indicate that the sludge buffer tank is a relative important N-emission source for the environment. To reduce the environmental impact of this source, it could be interesting to cover up the sludge buffer tank and treat the extracted air.
4. Conclusions
There is a need for new techniques capable of quickly and accurately determining emission and deposition impact from area sources. Classical handheld methodologies are however restricted in sampling area, time and safety, thus potentially leading to an underestimation of the real impact of these sources on the environment. Drone and chemical sensor technology presented here as the OLFASCAN Flying Lab, can prove to be a step forward in determining the emission impact of area sources, as emissions can easily and accurately be measured in both vertical and horizontal profiles. However, it is crucial for reliable assessment with this technique to account for drone downdraft effects on air sampling, which was accomplished in this study by using a 8 m long inert PVC suction pipe centred below the drone. By simultaneously assessing local meteorological data, the fluxwindow method can be executed on area sources in a reliable way to translate measured emission concentrations to emission rates. Modelling these emission rates allows to further quantify deposition impacts on the environment and scale the pollution degree of the area source. Based on the obtained results, the OLFASCAN Flying Lab highlighted that the investigated sludge buffer tank is (i) a relative important N-emission source, and (ii) drone and sensor technology can aid in identifying and quantifying important N-emission sources in a more uniform and representative way.
5. References
Bartholmai M., Neumann P. 2010. Micro-drone for gas measurement in hazardous scenarios via remote sensing. 10th WSEAS/IASME international conference on electric power systems, high voltages, electric machines and 6th WSEAS international conference on remote sensing (Japan) 4-6 October 2010. p. 149-152.
Baumbach G., Glaser K., Vogt U. 1999. Balance of the air pollutant mass fluxes of a city: first results of vertical soundings within the TFS-EVA project in Augsburg 1998. WIT Trans Ecol Environ. 28, 841-846.
Bockreis A., Steiberg I. 2005. Measurement of odour with focus on sampling techniques. Waste Manag. 25, 859-863.
Burqués J., Marco S. 2020. Environmental chemical sensing using small drones: a review. Sci Total Environ. 748.
Cao R., Li B., Wang H.-W., Tao S., Peng Z.-R., He H. 2020. Vertical and horizontal profiles of particulate matter and black carbon near elevated highways based on unmanned aerial vehicle monitoring. Sustainability. 12, 1204-1220.
Capelli L., Sironi S., Del Rosso R. 2013. Odor sampling: techniques and strategies for the estimation of odour emission rates from different source types. Sensors. 13, 938-955.
Departement Omgeving (2021) Milieueffectrapportage richtlijnenboek: landbouwdieren. https://omgeving.vlaanderen.be.
https://omgeving.vlaanderen.be/sites/default/files/atoms/files/ Richtlijnenboek_Landbouwdieren_2021.pdf. (accessed on 03/06/2021).
Gostelow P., Longhurst P., Parsons S., Stuetz R. 2003. Sampling for measurements of odours. IWA Publishing, London (UK).
Lucernoni F., Capelli L., Sironi S. 2016. Odour sampling on passive area sources: principles and methods. Chem Engineer Trans. 54, 55-60.
Toscano P., Gioli B., Dugheri S., Salvini A., Matese A., Bonacchi A., Zaldei A., Cupelli V., Miglietta F. 2011. Locating industrial VOC sources with aircraft observations. Environ Pollut. 159, 1174-1182.
Villa T. F., Salimi F., Morton K., Morowska L., Gonzalez F. 2016. Development and validation of a UAV based system for air pollution measurements. Sensors. 16, 2202- 2217.
VMM VLOPS20 (2017) VLOPS kaarten – totale vermestende depositie. www.geopunt.be.
http://www.geopunt.be/catalogus/datasetfolder/07c8bb6a-396a-40e7-98ca- a39e22b10479. (accessed on 03/06/2021).
6. Appendix
Fig. A1.: Windrose results during course of fluxwindow experiment
Fig. A2.: Setup fluxwindow grids sludge buffer tank
Table A1: Parameterization and source characteristics Flemish IMPACT model for calculating yearly average N-deposition of the sludge buffer tank
