Validation of odor concentration from mechanical-biological treatment piles using static chamber and wind tunnel with different wind speed values

ABSTRACT

The basic principle of odor sampling from surface sources is based primarily on the amount of air obtained from a specific area of the ground, which acts as a source of malodorous compounds. Wind tunnels and flux chambers are often the only available, direct method of evaluating the odor fluxes from small area sources. There are currently no widely accepted chamber-based methods; thus, there is still a need for standardization of these methods to ensure accuracy and comparability. Previous research has established that there is a significant difference between the odor concentration values obtained using the Lindvall chamber and those obtained by a dynamic flow chamber. Thus, the present study compares sampling methods using a streaming chamber modeled on the Lindvall cover (using different wind speeds), a static chamber, and a direct sampling method without any screens. The volumes of chambers in the current work were similar, ~0.08 m³. This study was conducted at the mechanical-biological treatment plant in Poland. Samples were taken from a pile covered by the membrane. Measured odor concentration values were between 2 and 150 ou_E/m³. Results of the study demonstrated that both chambers can be used interchangeably in the following conditions: odor concentration is below 60 ou_E/m³, wind speed inside the Lindvall chamber is below 0.2 m/sec, and a flow value is below 0.011 m³/sec. Increasing the wind speed above the aforementioned value results in significant differences in the results obtained between those methods. In all experiments, the results of the concentration of odor in the samples using the static chamber were consistently higher than those from the samples measured in the Lindvall chamber. Lastly, the results of experiments were employed to determine a model function of the relationship between wind speed and odor concentration values.

Implications: Several researchers wrote that there are no widely accepted chamber-based methods. Also, there is still a need for standardization to ensure full comparability of these methods. The present study compared the existing methods to improve the standardization of area source sampling. The practical usefulness of the results was proving that both examined chambers can be used interchangeably. Statistically similar results were achieved while odor concentration was below 60 ou_E/m³ and wind speed inside the Lindvall chamber was below 0.2 m/sec. Increasing wind speed over these values results in differences between these methods. A model function of relationship between wind speed and odor concentration value was determined.

Introduction

The Polish Committee for Standardization PN-EN 13725 air quality standard (2007) does not clearly defined the method of sampling air from the surface diffusion sources for the determination of odor concentration by dynamic olfactometry. This part of that standard is limited to a general description of sampling and the transport materials, as well as methodology of filling containers. However, difficulties arise when sampling from different surface diffusion sources because of their size, shape, and required operating parameters, raising the need for specific guidelines for the design of sampling equipment, which the standard lacks. The authors only recommend that the materials used for equipment production must meet specific requirements (PCS, 2007). Due to the lack of specific guidelines for sampling, the results obtained by the dynamic olfactometry method may vary from the same diffusion source when different sampling devices are used (Navaratnasamy et al., 2004; Bokowa and Liu, 2008; Hudson and Ayoko, 2008; Hudson et al., 2009). In present work, various methods of sampling from diffusion sources were compared. In particular, examination of the waste heaps on the mechanical-biological treatment (MBT) plant in Poland, was carried out.

The surface sources of odorants can be divided into active and passive. Active sources are characterized by a minimum flow of 50 m³/m²-hr and include, above all, biofilters (Richter and Frechen, 2009). Hood devices, used for sampling from area sources, are typically made of polyethylene with a thickness of 0.10 or 0.15 mm. Inter alia, pyramid-shaped covers with a surface area of 1 m² are used, which are equipped with outlet pipes with diameters of 0.14–0.2 m and a length of at least 10 times that of the diameter. A port for air sampling is placed in the middle of the exhaust pipe, at a height of at least 5 times that of the diameter (Richter and Frechen, 2009).

The passive sources are characterized by a laminar flow. These sources are either not ventilated or have poor ventilation with a minimum flow of less than 50 m³/m²-hr (Richter and Frechen, 2009). Passive sources include, inter alia, excessive settling tanks on sewage treatment plants, couloirs, compost heaps, and aerated and nonaerated sludge tanks. To determine the emission factors of odorants for passive area sources, two different approaches can be used (Hudson and Ayoko, 2008). The first method is based on making indirect measurements using micrometeorology methods, whereas the second approach is based on the method of dynamic hood sampling devices (Capelli et al., 2013). The basic principle is based primarily on knowing the amount of air drawn from the dome covering a specific area of the malodorous compounds source.

Devices used for sampling from passive sources include flow chambers modeled on the Lindvall cover, static flow chambers, and dynamic flow chambers. Representative samples are collected at the outlet of the cover. The static chambers are sealed domes with an open base, which are placed over the source of the odorants. To obtain representative data across the surface source, it is important to sample from different points that are equally spaced on the surface of the emitter. It is assumed that area from which a single sample is taken should represent about 1% of the total area of the emitter (Belgiorno et al., 2013). A drawback of these devices is that the pressure prevailing under the chamber is often different from the pressure outside, which may lead to errors, due to the fact that there is a relationship between the pressure source and the surface emission surface (Frechen et al., 2004). In the flow chambers, clean air (directly from the bottle, or purified through a carbon filter) is supplied with a fixed flow. Under the cover of the chamber, the clean air is mixed with the fragrant air. In addition, some of the dynamic flow chambers comprise components for mixing air inside the dome (Gostelow et al., 2003).

Streaming chambers modeled on the Lindvall cover—tunnels—are primarily used to collect samples from passive sources. The ratio of the longer chamber edges to shorter edges should be at least 2:1 (Frechen et al., 2004). The wind tunnel provides a controlled environment that can be replaced, so that small differences in emission can be measured with statistical significance (Sommer and Misselbrook, 2016). This system is based on the assumption that the innate character of the flow, which is encountered at the air/soil/water boundary, can be simulated under the cover. Smith and Watts (1994) observed that the odor concentration obtained by longer and narrower chambers was about 20% lower than that obtained using the smaller chambers. This is likely due to the fact that increasing the volume of the chamber results in a reduced number of stagnant zones as well as turbulent zones. However, the increased volume also limits the sampling surface (Jiang et al., 1995).

Frechen et al. (2004) and Capelli et al. (2009) suggested that it is reasonable to make measurements at low wind speeds inside the tunnel (1–10 cm/sec). Hobson and Yang (2001) found that to get the most reliable results, the speed of the air inside the Lindvall chamber should be kept at the level of 0.2–0.3 m/sec, since the flow rate and odor concentration measured at the chamber outlet are important parameters to determine the level of odor. Liu et al. (2015, 2016), after examination of the emission rates of volatile organic compounds with sweeping velocity inside wind tunnel from 0.1 to 1.07 m/sec, suggested that 0.28 m/sec is the recommended practical operation wind speed value. The linear correlation was obtained between the sweeping velocity and emission rates when wind speed values were less than 0.5 m/sec. Di Petra et al. (2016) compared wind tunnels with perforated plates inside chamber, guide vanes, wire gauses, and guide channels. Those researchers demonstrated that tunnels with guide channels represent the best solution among them, without using a main chamber longer than 1 m. Wind speed inside tunnel could be even up to 5 m/sec (di Petra et al., 2006).

The discrepancy between the results obtained from samples taken using the flow cells may be the effect of factors such as the small area of the chamber surface, change of flow turbulence as compared with natural conditions, and insufficient mixing of the clean and polluted air (Smith and Watts, 1994). It is advantageous that chambers without forced airflow shall not be applied to aerated tanks (Bockreis and Steinberg, 2005), due to the fact that in liquid sources with forced air as well as mixed sources, the turbulences of the liquid are large. Furthermore, the air under the dome can quickly saturate the odorants. These devices are primarily used to collect highly volatile compounds, such as hydrogen sulfide (H₂S) and other sulfur compounds (Leyris et al., 2005). When the odor concentration is greater than 100–200 ou_E/m³, tunnels are best suited for surface sources (Zarra et al., 2013).

Parker et al. (2013) wrote that according to long-standing debate (Fowler et al., 2001; Blanes-Vidal et al., 2007; Cole et al., 2007; Acevedo at al., 2009; Hudson et al., 2009), and despite their shortcomings, wind tunnels and flux chambers often are the only available, direct methods of evaluating the fluxes from small area sources. Furthermore, there are still no widely accepted chamber-based methods, and there is still a need for standardization to ensure full comparability of these methods.

Materials and methods

Object of research

The object of the current research was the installation of the mechanical-biological treatment plant—the regional municipal waste processing facility in Poland. Its total capacity is 920 Mg/day (230,000 Mg of processed municipal solid waste per year). Regarding the biological component, on the open square, aerobic stabilization of organic fraction of the waste is performed on piles 44 m long, 6 m wide, and 2.5 m high. They are actively aerated and are covered with a semipermeable (vapor-permeable) membrane. Each of the piles is equipped with three aeration channels and a channel for the leachate collection. A computer unit controls the fan that aerates the piles to minimize the possibility of anaerobic zone formation. The process continues until reaching the 4-days respiration activity (AT₄) ratio of less than 10 mg O₂/g of dry mass, which lasts for 5–9 weeks.

Characteristics of the materials

Scentroid SM 100 (Ontario, Canada) is a field olfactometer, using compressed air from the cylinder under high pressure (31 MPa) to dilute the test sample. The apparatus consists of a dilution valve control. Its high accuracy is used to provide a constant flow of diluted air through the device, which allows the user to select one of the 15 positions, which correspond to ratios of clean air to the dilution of the test air sample. The range of the device is between 2 and 30,000 ou_E/m³; the detection limit of the olfactometer is 3 ou_E/m³.

The measurement of wind speed and direction was made using a hand anemometer Kestrel 4500 NV with a wing rotor, whereas the measurement of temperature and relative humidity was carried out using the psychrometer Rotronic HygroPalm sensor HygroClip2 HC2-S3.

The Scentroid wind tunnel is a device modeled after the Lindvall chamber, which is composed of a set consisting of a Lindvall chamber, air blower, and carbon filter to ensure an odorless air supply to the system. The air blower allows the force of wind speed under the cover in the range of 0.13–0.66 m/sec, for a continuous operation time of 1.5 hr. The Scentroid wind tunnel device allows for the measurement of emission from liquid or solid surface sources and meets the standards specified in the standards The Association of German Engineers VDI 3880, European Committee for Standarization EN 13725, and Standards Australia/Standards New Zealand Commitee AS/NZS 4323. Its dimensions are as follows: length: 167 cm, width: 33 cm, height: 23 cm without fixed floats, weight: 10 kg. The static chamber used for the static test was made of an odorless material and has the shape of a truncated cone with bases of 18 and 28 cm diameters. The height is 45 cm, topped with a 5-cm nozzle with a 5-cm diameter. The chambers have similar active volumes (volume of the air column over the tested surface): the volume of Lindvall chamber is 0.078 m³, whereas the volume of static chamber is 0.075 m³. Background odor of sampling is one of the main limiting factor for measuring odor concentration. The equipment used in this study is made of stainless steel, which produces zero background odor and is easily decontaminated. Only virgin polytetrafluoroethylene (PTFE) tubing was used for both sampling and clean air delivery. Also, pure, technical air was cleaned by a carbon filter to achieve “odorless” medium and avoid background odor. The instrument used in this study has a lower dilution of 2 and employs only PTFE and stainless steel surfaces, making it easy to decontaminate to nearly zero background odor. Also, background odor is intended as the odor of the sampling site itself. The examined methods are based on the principle of isolation of a portion of the polluted air and measurement of the odor concentration at the outlet. It should be noted that the sampling isolates the background odors, and, in fact, the detection threshold of this type of sampling could be less than the background odors already present in the environment. The wind tunnel was equipped with the carbon filter to ensure an odorless air supply to the system and elimination of the background (environmental) odor.

Methodology

Tests were performed on piles located on the MBT installation. The interval between the series was a maximum 7 days, and the complete cycle lasted 7 weeks, starting from the dump of the pile and covering the geomembrane until completing the biostabilization (removal of waste to a landfill). Samples were taken from five places on the plateau of the piles (corners and geometric center). Sampling of air emitted from the surface sources of odorants was carried out as follows:

• Using a Lindvall chamber with the wind speed at the outlet of the Lindvall cover—from 0 to 5.00 m/sec (wind tunnel air speed between 0 and 0.66 m/sec) as well as such external wind speed

• Using a static chamber

• directly, without using chambers

For each test, from the time of setting the static chamber as well as initiating the flow of air inside the Lindvall chamber and the measurements, 10 min were needed to equalize, respectively, pressure and airflow. Samples were taken from a distance of approximately 10 cm above the membrane. In the case of the Lindvall cover, the sampling point was in the center of the outlet section of the chamber when the air speed measured in the outflow was between 1 and 5 m/sec, and in the center of the top basis of chamber when the value of wind speed was 0 m/sec. The total number of tested samples was 1175.

Calculation

Comparing the results, the analysis of variance was performed to assess the impact of an independent, qualitative factor—the sampling method—on the distribution of the dependent variable—the odor concentration.

To determine whether the data came from a normal distribution, the Shapiro-Wilk statistical test was selected. This test can be used for small samples. Furthermore, it was chosen because the χ² test and the Kolmogorov test do not use the full range of information resulting from the sample. In addition, a modified Kolmogorov-Smirnov test (Anderson-Darling test) was considered, in which the critical values depend on the specific distribution (Dodge, 2010). Therefore, this test is more sensitive to the shape of the “tails” of the distribution. It was recognized that the Shapiro-Wilk test, because of its power, is the most appropriate for this study. One of the basic assumptions in the analysis of variance is the homogeneity of variance in groups. To verify the hypothesis of equal variance in all populations, Bartlett’s test was employed, the most commonly used test, which concerns the variance of k normal populations, N(µ_i, σ_i²). The Bartlett test is very sensitive to deviations from the normal distribution, so two additional tests were performed: the Flinger-Killeen test (more resistant to outliers) and the Levene test. Although the analysis of variance answers the questions of whether or not the group means differ, it does not indicate which of them are significantly different. Therefore, after rejecting the null hypothesis, a detailed post hoc test was performed. Among post hoc tests, the honestly significant difference (HSD) Tukey test was chosen.

Additionally, the method of the one-dimensional spline curves had been used, which is considered to be an additive model (Hastie et al., 2001):

(1)

where X is explanatory variable, Y is dependent variable, α is a direction of projection, and f_k is the degree of polynomials. This method consists of dividing the domain of X into K disjointed intervals using a structured set of points, called nodes:

To observe the values belonging to each interval [ξ_k−1, ξ_k] (for k = 1, …, K), where ξ₀ is the smallest of the variable X and ξ_K the highest, the function f_k is adjusted, using the classical least squares method. In the splines method, it is assumed that the sought functions f_k are constants:

(2)

where

(3)

Due to the potential discontinuity of f, the degree of polynomials f_k was increased and conditions of continuity at the nodes and related functions were imposed. Hastie et al. (2001) has shown that such a modified-order spline function of order M can be represented as a linear combination of the basis functions h:

(4)

and

(5)

where l = 0, …, M; k = 1, …, K − 1; and

(6)

In the issue of spline functions, a commonly used parameter is the number of degrees of freedom, df, which equals the number of base functions. Most often, the spline functions of the third order (cubic splines) are used, so spline third-order function with K − 1 nodes, ξ_k, can be represented by the basic functions as follows:

(7)

The strength of the correlation relationship between variables was determined using the Spearman correlation coefficient. The choice of the applied factor was based on the type of features, the number of observations, and the shape of the relationship. The linear Pearson’s correlation coefficient was rejected a priori because it is sensitive to the sample variability, especially for small samples. Therefore, for the examined population, it was decided that the better test would be the Spearman correlation coefficient, which is used to describe the strength of correlation features when they are measurable, qualitative, and the tested community is small.

The correlation coefficient is in the range of −1 to 1. These results were visualized by the so-called “heatmaps.” This is a visualization of the matrix of numbers, in which the numerical value of the cell matrix is shown in a color (or shade of the color).

Results

Table 1 contains the means and standard deviations of odor concentration values achieved using different sampling methods. The results show that there is a difference between odor concentration values obtained in separate sampling methods, as well as during increases in the wind speed in the Lindvall chamber. In particular, the odor values decrease with increasing wind speed in the Lindvall chamber outlet. The difference between the mean of the odor concentration with wind speeds in the outlet of 1 and 5 m/sec is 18.2 ou_E/m³. Results are visualized in Figure 1, which contains box plots of the data, whereas Figure 2 contains a graph of spline functions for different values of degrees of freedom, df.

Table 1. Means and standard deviations of odor concentration values achieved using different sampling methods.

	Methods		Lindvall chamber, wind speed in the outlet (m/sec)
Parameter	Static	Direct	0	1	1,5	2	2,5	3	3,5	4	4,5	5	vout
Mean (ouE/m^3)	19.9	18.1	19.5	18.6	15.1	12.8	10.8	6.7	2.9	1.5	0.9	0.4	11.9
SD (ouE/m^3)	20.3	18.7	20.2	19.8	14.9	12.5	9.9	7.3	4.5	3.5	2.7	1.6	11.2

Figure 1. Box plots of odor concentration values achieved using different sampling methods.

Figure 2. Graphs of spline functions of wind speed and odor concentration relations for different values of df.

Table 2 contains the parameters of regression curves representing the relationship between odor concentration values measured using a static chamber and of the other sampling methods.

Table 2. Parameters of regression curves showing the relationship between odor concentration values obtained using the static chamber and the other sampling methods.

Method	Slope	χ^2	R^2	R^2adj.	RMSE	RSS
Direct	0.897	4351	0.947	0.946	4.321	4351
v = 0	0.996	460	0.995	0.995	1.405	460
v = 1,0	0.970	1157	0.987	0.987	2.229	1157
v = 1,5	0.723	1835	0.965	0.965	2.807	1835
v = 2,0	0.590	3307	0.910	0.909	3.767	3307
v = 2,5	0.451	3256	0.857	0.856	3.738	3256
v = 3,0	0.323	2350	0.810	0.808	3.176	2350
v = 3,5	0.202	886	0.816	0.814	1.950	886
v = 4,0	0.156	515	0.820	0.819	1.486	515
v = 4,5	0.109	534	0.680	0.677	1.514	534
v = 5,0	0.059	256	0.570	0.567	1.047	256
v = vout	0.494	6148	0.793	0.791	5.137	6148

Note. Slope = slope of the curve; χ² = value of χ²-test statistic; R² = coefficient of determination; R²_adj = adjusted coefficient of determination; RMSE = root mean square error; RSS = residual sum of squares.

Table 3 contains results of variance analysis of odor concentration values achieved using different sampling methods.

Table 3. Results of the variance analysis of odor concentration values achieved using different sampling methods.

Measure	df	Sum Sq	Mean Sq	F value	Pr(>F)
Method	12	157.319	13.110	75.454	>2.2 × 10^−16
Residuals	3042	528.542	173.7	—	—

Note. df = degrees of freedom; Sum Sq = sum of squares; Mean Sq = mean of squares; F value = value of F-test statistic; Pr (>F) - probability value for F-test statistic.

Further, the value of F-test statistic is 75.454, so according to the probability test, the zero hypothesis is rejected in favor of the alternative hypothesis; thus, the odor concentration values are dependent on the sampling method.

Importantly, to determine which of the above results differ significantly from each other, a post hoc Tukey test was performed. Additionally, Pearson’s correlation matrix heat map was performed (Figure 3).

Figure 3. Pearson’s correlation matrix heat map.

Discussion

Many researchers have previously compared the odor concentration values obtained with the Lindvall chamber and those with a dynamic flow chamber (Hudson and Ayoko, 2008). In particular, they have studied the concentration of odor coming from the effluents from livestock farms and feed lot (wet and dry). Amongst these studies, Bokowa and Liu (2008) compared the results of the odor concentration of samples taken from two locations, using the previously mentioned devices, whereas Navaratnasamy et al. (2004) compared both devices while examining various concentrations of butanol.

With respect to chambers, Hudson et al. (2009) compared a wind tunnel and U.S. Environmental Protection Agency (EPA) dome-type flux chamber (Kienbusch, 1986) to assess odor emissions and proved that wind tunnel emission rates were 60–240 times greater than those found in the flux chamber.

The results obtained by Hudson and Ayoko (2008), who examined leachates from animal production and farm fattening units, were 4–8 times smaller when the sampling was performed using a Lindvall chamber (from 51 to 469 ou_E/m³) than when the samples were obtained using the intake flow chamber (from 406 to 3360 ou_E/m³). Those obtained by Bokowa and Liu (2008) were 11 times smaller (Lindvall chamber: 709–845 ou_E/m³, intake flow chamber: 14760–23840 ou_E/m²), whereas those obtained by Navaratnasamy et al. (2004) were 3–6 times lower (Lindvall chamber: 30–502 ou_E/m³, intake flow chamber: 30–1644 ou_E/m³). In addition, Leyris et al. (2005) compared methyl sulfate concentrations in the samples obtained via the previously mentioned types of chambers with different airflows. In those cases, the results of concentrations of collected samples using a dynamic flow chamber were larger than those obtained using the Linvall chamber (differences between odor concentration values were from 2.7% to 106.7%), except for samples containing methyl sulfate concentrations of 100 mg/dm³ (flow rate 5 m/sec) and 550 mg/dm³ (flow rate 3 m/sec). The difference between odor concentrations in the samples obtained through both Lindvall and static chambers from the present studies are summarized in Table 4.

Table 4. Comparison of odor concentration values in the samples obtained through two types of chambers (Lindvall; static) with different airflow in the outflow of Lindvall chamber.

Wind speed (m/sec)	Difference		Wind speed (m/sec)	Difference
	%	Static/Linvall		%	Static/Linvall
0	2.01	1.02	3	66.33	2.94
1	6.53	1.06	3.5	85.43	6.86
1.5	24.12	1.32	4	92.46	13.27
2	35.68	1.55	4.5	95.48	22.11
2.5	45.73	1.84	5	97.99	49.75

Lucernoni et al. (2017) compared the odor emission rates achieved by wind tunnel and flux chamber and reported that measurements carried out with wind tunnel gave emission value of 248,050 ou_E/sec, whereas with flux chamber it was 6150 ou_E/sec. Parker et al. (2013) compared the sampling methods of collecting the following five compounds from the surface source: phenol, 4-methylphenol, 4-ethylphenol, indole, and skatole. The results were corrected with the coefficients for either the Lindvall chamber or the dynamic flow chamber, respectively, 3.84 and 2.42. The evaporative flux ratio correction was calculated as I_out/I_in, where I represents the flux (µg m⁻² min⁻¹). The researchers conducting the study observed considerable variation in uncorrected fluxes between the dynamic flow chamber and the wind tunnel. In particular, there were differences of 61%, 57%, 56%, 53%, and 49% between the measurements for, respectively, phenol, 4-methylphenol, 4-ethylphenol, indole, and skatole. After the evaporative ratio correction factors were applied, the differences dropped to, respectively, 17%, 13%, 12%, 9%, and 4%. Thus, Parker et al. (2013) proved that corrected flux values are more precise and more accurate than uncorrected flux values. In the present work, comparing the results of static and Lindvall chambers, the smallest difference is between odor concentration values in the static chamber and the wind speed in the outflow of Lindvall chambers, equivalent to 0, 1, and 1.5 m/sec. The R² values are, respectively, 0.995, 0.987, and 0.965, whereas when the wind speed is 5.0 m/sec, the R² value is 0.570. Although there is no significant difference between the odor concentration values obtained by direct sampling (without any chamber) and the condition when wind speed in the outlet of the Lindvall chamber equals 0, there is a difference between direct sampling and when the wind speed in the outlet of Lindvall chamber is the same as the wind speed of the outdoor air; here, the percentage difference in the results varies by 34.25%. Means and standard deviations of odor concentration values achieved using different sampling methods between static, direct, and Lindvall chamber method are comparable when the maximum value of wind speed in the wind tunnel outlet is 1 m/sec. There are statistically significant differences when that value is higher than 1 m/sec. Also, the values of the third quartile (75%) of the odor concentration are the same in case of direct sampling, wind tunnel speed 0 and 1 m/sec. There is also a perfect positive linear relationship (correlation coefficient value = 0.99) between odor concentration values achieved by a static method and those two values of wind speed. While wind speed value is 0.2 m/sec, that correlation coefficient value is 0.95, whereas in case of wind speed equals 5 m/sec, the correlation coefficient equals 0.75. According to the graphs of spline functions for different values of degrees of freedom, df, the highest coefficient of determinations were achieved for df = 5. In that case, the coefficient of determination R² = 0.996, adjusted coefficient of determination R²_adj. = 0.989, and single smoothing parameter = 0.19, which gives a cubic spline interpolation of the data. So, the fitted function formula was Y = 23.0 − 9.15 X + 14.09 X² − 8.723 X³ + 1.91 X⁴ − 0.14 X⁵.

Conclusion

Several researchers have demonstrated that there is a significant difference between odor concentration values obtained by the Lindvall chamber and by a dynamic flow chamber when concentration values are between 30 and 23 8640 ou_E/m³. The main objective of the present study was to compare sampling methods using a chamber modeled after the Lindvall chamber (using different wind speeds), a static chamber, and a direct sampling method in the MBT plant. Odor concentration values from the plant were between 2 and 150 ou_E/m³. Also, the second objective was to examine the differences of the achieved odor concentration values while increasing of the wind speed inside the wind tunnel. Volumes of the chambers employed were similar at ~0.08 m³. The author showed that both chambers can be used interchangeably in these conditions: odor concentration is below 60 ou_E/m³ and wind speed in the Lindvall chamber outlet is between 0 and 1.5 m/sec, and the speed inside the chamber is below 0.2 m/sec with a flow below 0.011 m³/sec. Also, a model function of the relationship between wind speed and odor concentration values was determined. By increasing the wind speed above those values, there were significantly different results between those methods. In each of the cases, the odor concentration values from the samples taken using the static chamber were higher than those obtained in samples taken by Lindvall chamber.

Additional information

Notes on contributors

Mirosław Szyłak-Szydłowski

Mirosław Szyłak-Szydłowski, DSc. is a research scientist and academic teacher in the the Faculty of Building Services, Hydro and Environmental Engineering at Warsaw University of Technology, Warsaw, Poland.