The Importance of the Microclimatic Conditions Inside and Outside of Plant Buildings in Odorants Emission at Municipal Waste Biogas Installations

Municipal waste biogas plants are an important element of waste treatment and energy policy. In this study, odorant concentrations and emissions were measured together with the air temperature (T) and relative humidity (RH) to confirm the hypothesis that the microclimatic conditions have an important impact on the level of odorant emission at municipal waste biogas plants. A simple correlation analysis was made to evaluate the strength and the direction of the relationship between the odorant concentration and emission and air temperature and relative humidity. The mean volatile organic compound (VOC) and NH3 concentrations vary depending on the stage of the technological line of the analysed municipal waste biogas plants and are in the following ranges, respectively: 0–38.64 ppm and 0–100 ppm. The odorant concentrations and emissions correlated statistically significantly with T primarily influences VOC concentrations and emissions while RH mainly affects NH3 concentrations and emissions. The strongest correlations were noted for the fermentation preparation section and for emissions from roof ventilators depending on the analysed plant. The smallest influence of microclimatic factors was observed at the beginning of the technological line—in the waste storage section and mechanical treatment hall. This is due to the greater impact of the type and quality of waste delivered the plants. The analysis of correlation between individual odorants showed significant relationships between VOCs and NH3 for most stages of the technological line of both biogas plants. In the case of technological sewage pumping stations, a significant relationship was also observed between VOCs and H2S. The obtained results may be helpful in preparing strategies to reduce the odours from waste treatment plants.


Introduction
One of the main goals of waste management is to optimize the collection of municipal solid waste (MSW). The introduction of an appropriate MSW economy system in the service area is of strategic importance for improving services, minimizing emissions of pollutants into the air, and reducing the economic costs of general waste management [1][2][3].
One of the methods of municipal waste management is its processing in a mechanical-biological treatment (MBT) installation. This method is used to minimize the amount of waste sent for landfilling [4]. In the case of municipal waste biogas plants (MWBP) which are a part of an MBT installation, an additional benefit of the plants' operation is the production of energy from biogas captured during the methane fermentation process (a biological process of biodegradable waste: parameters, they obtained a statistically significant correlation (p ≤ 0.05). Similar studies were also conducted by Le et al. [35], Miller et al. [36], Romain et al. [37], Schauberger et al. [38] and Sun et al. [39]. The results obtained by scientists show different levels of correlation between air temperature and relative humidity and odour emission depending on e.g., on the method of breeding. The example of research conducted on pig farms shows the complexity of the influence of the analysed parameters on the concentration of odorants. This dependence is not obvious and only comprehensive and meticulous research can give an answer to the real influence of both air temperature and relative humidity on odour emission in each situation, including those related to waste management.
Despite numerous scientific studies accompanied by analyses of the microclimate parameters impact on the emission of various chemical compounds, none of them refers to the municipal waste treatment and are carried out in a much shorter period of time-usually three months.
The literature review and our previous research show that the main factors causing odour emissions are following: organic substance decomposition processes, the dynamics which depend on the type of organic substance (biodegradable waste, animal faeces, etc.), microclimate parameters and technological regime [15,34].The study attempts to verify the above dependencies for individual odour emission sources in two municipal waste mechanical-biological treatment plants equipped with a biological part with fermentation chambers, which are a source of energy production from biogas. The obtained research results may be useful in the search for new technological solutions minimizing the emission of odorants, as well as in improving the functioning solutions. As Júnior et al. [40] concluded, there is not much information about the impact of MWBP on the environment is therefore important to provide an overall picture of the technologies used at the plants. The results presented in this paper are innovative, because so far there have been few scientific studies on odorant emissions from installations such as biogas plants processing municipal waste. The research presented in this paper is a continuation of earlier studies [15,27]. These papers contain the results of pilot research at six plants, which were the basis for the selection of more intensive research facilities, conducted on an annual basis.

Characteristics of the Analysed Plants
Both biogas plants under analysis are located in the eastern part of Poland and thus have the same climatic conditions, which is beneficial from the point of view of the research conducted. Both plants are equipped with installations designed for mechanical and biological treatment of municipal waste. The flowcharts of processes at the mechanical-biological waste treatment installations are shown in Figure 1. The particular flow-chart elements are also identified as odour sources, what is described in previous paper [15]. The mechanical part of both installations consists of similar processes, including the storage of waste delivered to the plants (mixed and raw materials collected selectively at the source) and their mechanical treatment (mainly sieving and mechanical sorting using special separators). The differences occur in the biological part, consisting of the processes of preparing the material for fermentation, methane fermentation and aerobic stabilization of the digestate. At biogas plant A, the input material for fermentation chambers is biodegradable waste collected, selectively and prepared for the process on the technological site. In the case of biogas plant B, the input material for fermentation chambers is the fraction of 20-80 mm sorted mechanically from the mixed waste stream (using sieve systems and special separators). This fraction is mainly biodegradable what is typical for municipal waste and is the basis of MBT process. According to [41] in the MSW fraction of <80 mm intended for the bioprocessing step of the MBT system, "organic waste" was the component with the largest share. Its share was, on average, at the level of 27.9 ± 10.5%. At biogas plant B, this fraction is prepared inside the technological plant building. In both plants the fermentation process lasts 21 days and then the aerobic stabilisation of the digestate lasts approximately four weeks.

Study Methodology
The research presented in this paper covers chemical determinations of compounds causing a negative olfactory sensation (odorants): volatile organic compounds (VOCs), ammonia (NH3), hydrogen sulphide (H2S) and methyl mercaptan (CH3SH) as well as the prevailing microclimatic conditions, temperature and relative air humidity using the streak method. In accordance with European standard which was developed by the Association of German Engineers (Verein Deutscher Ingenieure)-VDI 3940 [42], measurements at a height of 1.5 m were carried out using a portable Kestrel 4500 NV weather meter. Odorants were monitored using a MultiRae Pro portable multigas gas detector (RAE Systems, Inc.; San Jose, CA, USA) equipped with four individual compound sensors. The device has a built-in pumping system that allows the user to take gas samples directly from the source. The detector characteristics are shown in Figure 2. Five parallel measurements were made at each of the indicated measurement points ( Table 1). The tests were carried out in thirteen measurement series (from July 2019 to March 2020, Table 2) at two municipal waste mechanical and biological treatment plants with an installation for biogas collection and its energy use. The length of the research period is connected with the variability of seasons in Poland, as well as the variability of the temperature (T) and relative air humidity (RH) of inside and outside air.

Study Methodology
The research presented in this paper covers chemical determinations of compounds causing a negative olfactory sensation (odorants): volatile organic compounds (VOCs), ammonia (NH 3 ), hydrogen sulphide (H 2 S) and methyl mercaptan (CH 3 SH) as well as the prevailing microclimatic conditions, temperature and relative air humidity using the streak method. In accordance with European standard which was developed by the Association of German Engineers (Verein Deutscher Ingenieure)-VDI 3940 [42], measurements at a height of 1.5 m were carried out using a portable Kestrel 4500 NV weather meter. Odorants were monitored using a MultiRae Pro portable multigas gas detector (RAE Systems, Inc.; San Jose, CA, USA) equipped with four individual compound sensors. The device has a built-in pumping system that allows the user to take gas samples directly from the source. The detector characteristics are shown in Figure 2. Five parallel measurements were made at each of the indicated measurement points ( Table 1). The tests were carried out in thirteen measurement series (from July 2019 to March 2020, Table 2) at two municipal waste mechanical and biological treatment plants with an installation for biogas collection and its energy use. The length of the research period is connected with the variability of seasons in Poland, as well as the variability of the temperature (T) and relative air humidity (RH) of inside and outside air.     * Surface sources for which the gas sample was taken from under cover; ** in the mechanical treatment hall the measuring points were located at different heights in order to capture the influence of the exhaust system (ventilators); *** pre-RDF-sorted combustible fractions from municipal waste (presorted fuel from waste)-it is one of the waste fractions sorted in the analysed biogas plants. In the case of VOCs, the methodology for converting concentrations was used according to [44]. For a common VOC mixture, there is a close relationship between the VOCs measurement by a PID sensor used in the MultiRae Pro detector and that by charcoal sampling followed by gas chromatography with using flame ionization detector (GC-FID) analysis (R2 = 0.95): where C PID (in ppb) is the isobutylene-based concentration measured by PID sensor and C GC (in ppb) is the concentration determined by GC. Assuming C GC is a toluene-equivalent concentration, C GC in µg/m 3 can be converted to C GC in ppb using the ideal gas law: C GC µg/m 3 = C GC (in ppb) · 92.14 24.45 (7) C od kg/m 3 = C GC µg/m 3 ·10 −9 , where 92.14 is the molecular weight of toluene.

Statistical Analysis
The statistical analysis was made using the Statistica 13 software. The analysis of measured data was carried out by means of correlation analysis and linear regression. The dependent variables were odorant concentrations. The total number of independent variables considered in the regression analysis was two: air temperature and relative air humidity.

Microclimatic Parameters
To determine the effect of microclimatic conditions on the concentration and emission of odorants at biogas plants, VOC and NH 3 concentrations were recorded together with simultaneous monitoring of air temperature and humidity. Figures 3 and 4 show the results of the measurements of temperature (T) and relative air humidity (RH) (inside the halls-T i , RH i and outside the halls, T o , RH o values) in individual measurement series.

Odorant Concentration
The results obtained (average values with standard deviations) are presented in Tables 3-10 with a division into individual elements of the process line and including emission from ventilators. Table 3 presents the test results (mean values with standard deviations) for the waste storage section for points a-inside the hall-at 1.5 m, b-mixed waste and c-selectively collected waste.

Odorant Concentration
The results obtained (average values with standard deviations) are presented in Tables 3-10 with a division into individual elements of the process line and including emission from ventilators. Table 3 presents the test results (mean values with standard deviations) for the waste storage section for points a-inside the hall-at 1.5 m, b-mixed waste and c-selectively collected waste. At biogas plant A, VOC concentrations from 0.21-12.24 ppm and NH 3 concentrations from 0-6 ppm were recorded. The highest concentrations of VOCs and ammonia (for biogas plant A: VOCs = 12.24 ± 0.03 ppm, NH 3 = 15 ± 1.85 ppm; for biogas plant B: VOCs = 18.66 ± 1.28 ppm, NH 3 = 48 ± 4.91 ppm) were observed for both analysed biogas plants for point b-mixed waste. The higher level of VOC concentrations for point b at biogas plant B may be justified by the waste collection system in the service area-during the research period, the fraction of biodegradable waste was not collected selectively and was probably directed to mixed waste. In the case of the biogas plant service area A, selective collection of biodegradable waste is carried out, which is dictated by the need to prepare the input material for the fermentation process.

Waste Storage
When analysing point c, selectively collected waste, higher VOC values were observed at biogas plant B, which may be due to irregularities related to the method of waste collection and, consequently, the purity of raw material waste. The highest standard deviation was observed for point b for the ammonia concentration in the third series: 48 ± 4.91 ppm. The obtained results at the presented level probably come from the substance contained in the waste (noticeable smell of the solvent). The lowest values of the tested compounds were observed at point a. Table 4 presents the test results (mean values with standard deviations) for the mechanical treatment section for points d (inside the hall at 1.5 m) and e (inside the hall at 4.0 m).   Table 5 shows the test results (mean values with standard deviations) for the fermentation preparation section for points f (fermentation fraction) and g (over the wastewater tank of the pumping station). During the tests, the presence of hydrogen sulphide and methyl mercaptan was observed only at point g.  At biogas plant A, VOCs were found at the level from 0.02 ppm to 14.79 ppm, NH 3 in the range of 0-100 ppm, H 2 S in the range of 0 ÷ 1 ppm and CH 3 SH between 0 ÷ 1.5 ppm. When analysing the presented research results, the highest concentrations of VOCs were observed for the point (fermentation fraction). In the case of biogas plant A, measurements for this point were made only in four series, which resulted from the availability of this fraction during the tests (biodegradable waste collected selectively screened through a mobile sieve with a mesh size of 0-20 mm). The prepared fraction in this biogas plant is directed to the buffer on an ongoing basis, excluding storage operations in the yard. A different situation occurs in biogas plant B (concentration ranges: VOCs = 0.01-25.41 ppm, NH 3 = 0-100 ppm, H 2 S = 0-100 ppm and CH 3 SH = 0-10 ppm). In this case, the fraction for fermentation (biodegradable fraction released mechanically from the mixed waste stream) is stored in the hall or, in the case of insufficient space, on the technological site with exposure to microclimatic conditions. The high level of ammonia at this point may be the result of anaerobic changes taking place under uncontrolled conditions. Sulphur compounds, H 2 S and CH 3 SH, are present only for point g (over the wastewater tank of the pumping station), mainly in biogas plant B, which is conditioned by the technology used in the plants and the origin of the wastewater. At biogas plant A, the tank receives sewage from the technological site, where there are piles of waste subjected to aerobic processes (leachate from heaps and sewage generated as a result of precipitation) and from the biodrying hall. The concentration range of sulphur compounds for point g in biogas plant A is: H 2 S = 0-1.0 ppm and CH 3 SH = 0-1.5. In the case of biogas plant B, leachate from the fermentation process is directed to the tank, and the concentration range of the tested compounds is: H 2 S = 0-100 ppm and CH 3 SH = 0-10 ppm. Table 6 shows the test results (average values including standard deviations) for the oxygen stabilization section for points h (waste subjected to an oxygen stabilization process) and I (biodegradable waste selectively collected). At biogas plant A, VOCs were recorded at levels from 0.06 ppm to 16.22 ppm, and NH 3 at the level 0-100 ppm. At biogas plant B, VOC levels varied from 0.46-11 ppm and NH 3 from 0-100 ppm.

Other Elements of Technological Line
Pre-RDF Storage Table 7 shows the test results (average values with standard deviations) for the pre-RDF storage section for measurements points: j (shredded preRDF fraction at the storage shelter) and j (RDF fraction subjected for biodrying).
At biogas plant A, VOCs ranging from 0.02 ppm to 18.77 ppm were recorded, and NH 3 in the range 0-8 ppm. At biogas plant B, VOC concentrations of 0.13-17.81 ppm and NH 3 of 0-30 ppm were recorded. The technological process of biodrying of the pre-RDF fraction is carried out exclusively in the biogas  Measurements in Front of the Technological Plant Buildings Table 8 shows the measurement results (average values including standard deviations) in front of the plant buildings for measurement points o (in front of the waste storage hall entrance) and p (in front of the mechanical treatment hall entrance).
0.74 ± 0 0.55 ± 0 1 ± 0 1 ± 0 At the entrances to the buildings, the concentrations of the tested chemical compounds in each of the analysed plants are at a similar level. In the case of biogas plant A, the concentration of VOC can be observed at the level of 0-2.73 ppm and NH 3 at the level of 0-3 ppm. In the case of biogas plant B, the VOC concentration was recorded in the range of 0-1.75 ppm and NH 3 0-2 ppm. Table 9 shows the measurement results for section emissions from roof ventilators for measurement points at biogas plant A: l (ventilator 1 process gases captured from over mixed waste), m (ventilator 2 process gases captured from the hall centre) and n (ventilator 3 process gases captured from the mechanical treatment building); and at biogas plant B: l (ventilator 1 process gases captured from over mixed waste and selectively collected waste), m (ventilator 2 dust-extracting-process gases captured from the hall centre), n (ventilator 3 process gases captured from the mechanical treatment building) and n (ventilator 4 process gases also captured from the same building).

Emission from Roof Ventilators
The analysis of the results obtained from the roof ventilators shows that greater emission occurs from the ventilators in the B biogas plant: VOCs = 0.012-0.106 kg/h, NH 3 = 2.3 × 10 −5 -6.5 × 10 −5 kg/h. In the case of biogas plant A, the emissions from the ventilators are at a lower level: VOCs = 0.011-0.275 kg/h, NH 3 = 0-0.271 kg/h.

Air Temperature and Humidity Influence on VOC and Ammonia Concentrations and Emission
The correlation analysis between the odorant concentrations and measured microclimatic parameters was made and presented with division into individual stages (sections) of the technological sequence.

Waste Storage
In the waste storage section, some significant correlations (p ≤ 0.05) were found between NH 3 concentration and both the air temperature and RH. However, no influence of microclimatic factors on VOC concentration was observed. This is most likely due to the influence of other factors-primarily the type of waste stored (to be processed in the biogas plant) and the duration of storage. Ammonia concentration plotted against T and RH are shown in Figures 4 and 5. Energies 2020, 13, x FOR PEER REVIEW 16 of 33

Air Temperature and Humidity Influence on VOC and Ammonia Concentrations and Emission
The correlation analysis between the odorant concentrations and measured microclimatic parameters was made and presented with division into individual stages (sections) of the technological sequence.

Waste Storage
In the waste storage section, some significant correlations (p ≤ 0.05) were found between NH3 concentration and both the air temperature and RH. However, no influence of microclimatic factors on VOC concentration was observed. This is most likely due to the influence of other factorsprimarily the type of waste stored (to be processed in the biogas plant) and the duration of storage. Ammonia concentration plotted against T and RH are shown in Figures 4 and 5.

Mechanical Treatment
In the case of mechanical treatment, no significant correlations were also found, both between T and RH, and the odorant concentrations. Again, the most likely cause is the high variability in the quality of the waste. Only in one odour source (d, biogas plant A) the study showed that the concentration of VOCs was statistically significantly dependent on RH (p ≤ 0.05). The concentration of VOCs in relation to RH i is shown in Figure 6.
In the case of mechanical treatment, no significant correlations were also found, both between T and RH, and the odorant concentrations. Again, the most likely cause is the high variability in the quality of the waste. Only in one odour source (d, biogas plant A) the study showed that the concentration of VOCs was statistically significantly dependent on RH (p ≤ 0.05). The concentration of VOCs in relation to RHi is shown in Figure 6.

Fermentation Preparation
The impact of air temperature and humidity is clearly visible in the fermentation preparation section. Significant correlations were found between odorant concentrations (both VOCs and NH3) and both T and RH-as shown in Figures 7 and 8.

Fermentation Preparation
The impact of air temperature and humidity is clearly visible in the fermentation preparation section. Significant correlations were found between odorant concentrations (both VOCs and NH 3 ) and both T and RH-as shown in Figures 7 and 8.
In the case of the fermentation preparation section, a significant effect of both T and RH on the concentration of odorants from the fraction prepared for fermentation and over wastewater tank was observed. In most cases, the results show a positive correlation of the VOC and NH 3 concentrations with T (from R = 0.57 to R = 0.66) and a negative correlation with RH (from R = −0.58 to R = −0.77). A strong negative correlation with T was only observed in the case of the VOC concentration captured from the fraction prepared for fermentation in biogas plant A (R = −0.99).

Mechanical Treatment
In the case of mechanical treatment, no significant correlations were also found, both between T and RH, and the odorant concentrations. Again, the most likely cause is the high variability in the quality of the waste. Only in one odour source (d, biogas plant A) the study showed that the concentration of VOCs was statistically significantly dependent on RH (p ≤ 0.05). The concentration of VOCs in relation to RHi is shown in Figure 6.

Fermentation Preparation
The impact of air temperature and humidity is clearly visible in the fermentation preparation section. Significant correlations were found between odorant concentrations (both VOCs and NH3) and both T and RH-as shown in Figures 7 and 8.  In the case of the fermentation preparation section, a significant effect of both T and RH on the concentration of odorants from the fraction prepared for fermentation and over wastewater tank was observed. In most cases, the results show a positive correlation of the VOC and NH3 concentrations with T (from R = 0.57 to R = 0.66) and a negative correlation with RH (from R = −0.58 to R = −0.77). A strong negative correlation with T was only observed in the case of the VOC concentration captured from the fraction prepared for fermentation in biogas plant A (R = −0.99).

Oxygen Stabilization
In the oxygen stabilization section, almost no statistically significant correlations were found

Oxygen Stabilization
In the oxygen stabilization section, almost no statistically significant correlations were found both between T and RH and odorant concentrations. Most likely, the reason for this situation is the influence of other factors-mainly the type of waste going to the plant and, consequently, subjected to oxygen stabilization at the end of the technological line. Only at biogas plant B did the test show a statistically significant correlation between odorant concentrations and measured microclimatic parameters (p ≤ 0.05). VOCs and NH 3 concentrations plotted against T and RH are shown in Figure 9. The results show a positive correlation of VOC and NH3 concentrations with T (respectively R = 0.64 and R = 0.75) and the negative correlation with RH, but not high (R = −0.59). The correlations observed are related to the concentrations of odorants that have been captured from green waste directed to aerobic stabilization together with the digestate in biogas plant B. This waste stream is characterized by lower variability compared to other biofractions-hence the more visible influence of atmospheric factors on odorant concentrations. No influence of air temperature and humidity on concentrations of odorants from the digestate was observed. In this case, the type of waste delivered to the plant is most likely to have a major impact.

Other Elements of the Process Line
The figures show the relationships between the results of measurements coming from the pre-RDF storage and in front of the technological plant buildings.
In the pre-RDF section, some significant correlations (p ≤ 0.05) were found especially between NH3 concentrations and both air temperature and RH. Odorant concentrations plotted against T and RH are shown in Figures 10 and 11. The results show a positive correlation of VOC and NH 3 concentrations with T (respectively R = 0.64 and R = 0.75) and the negative correlation with RH, but not high (R = −0.59). The correlations observed are related to the concentrations of odorants that have been captured from green waste directed to aerobic stabilization together with the digestate in biogas plant B. This waste stream is characterized by lower variability compared to other biofractions-hence the more visible influence of atmospheric factors on odorant concentrations. No influence of air temperature and humidity on concentrations of odorants from the digestate was observed. In this case, the type of waste delivered to the plant is most likely to have a major impact.

Other Elements of the Process Line
The figures show the relationships between the results of measurements coming from the pre-RDF storage and in front of the technological plant buildings.
In the pre-RDF section, some significant correlations (p ≤ 0.05) were found especially between NH 3 concentrations and both air temperature and RH. Odorant concentrations plotted against T and RH are shown in Figures 10 and 11. In the case of measurements made at the technological plant building entrances (storage hall and mechanical treatment hall), significant correlations were only observed in the case of VOC concentration measurement results at the mechanical treatment hall-both in the case of biogas plants A and B.
The tests show statistically significant correlations between VOC concentrations and measured microclimatic parameters (p ≤ 0.05). VOC concentrations plotted against T and RH are shown in Figures 12 and 13. In the case of measurements made at the technological plant building entrances (storage hall and mechanical treatment hall), significant correlations were only observed in the case of VOC concentration measurement results at the mechanical treatment hall-both in the case of biogas plants A and B.
The tests show statistically significant correlations between VOC concentrations and measured microclimatic parameters (p ≤ 0.05). VOC concentrations plotted against T and RH are shown in Figures 12 and 13.

Emission
The Figures 14-16 show the relationship between T and RH and emissions from roof ventilators and from biofilters. In the case of emissions from roof ventilators, a lot of significant correlations (p ≤ 0.05) were found (in 18 measuring sources in total) between odorant emissions (both VOC and NH3) and both measured microclimatic parameters (T and RH), especially at biogas plant B. In Figures 14  and 15, the relationships with the greatest significance and the best linear fits for both biogas plants are shown.

Emission
The Figures 14-16 show the relationship between T and RH and emissions from roof ventilators and from biofilters. In the case of emissions from roof ventilators, a lot of significant correlations (p ≤ 0.05) were found (in 18 measuring sources in total) between odorant emissions (both VOC and NH 3 ) and both measured microclimatic parameters (T and RH), especially at biogas plant B. In Figures 14  and 15, the relationships with the greatest significance and the best linear fits for both biogas plants are shown. In Figure 16, the relationships with the greatest significance and the best linear fits for both biogas plants are shown.
The results show a negative correlation of the VOC emissions with RH and a positive correlation with T. The observed correlations were not high: R = −0.62 and R = 0.57 (p ≤ 0.05), respectively. In Figure 16, the relationships with the greatest significance and the best linear fits for both biogas plants are shown.
The results show a negative correlation of the VOC emissions with RH and a positive correlation with T. The observed correlations were not high: R = −0.62 and R = 0.57 (p ≤0.05), respectively.

Comparison of Air Temperature and Humidity Influence on Odorant Concentrations and Emission
The best obtained values of correlation coefficient of odorant concentrations and emission with measured microclimatic conditions (equal and higher than R = 0.7, p ≤ 0.05) are shown in Table 11.

Comparison of Air Temperature and Humidity Influence on Odorant Concentrations and Emission
The best obtained values of correlation coefficient of odorant concentrations and emission with measured microclimatic conditions (equal and higher than R = 0.7, p ≤ 0.05) are shown in Table 11. The above comparison clearly shows that, at biogas plants treating municipal waste, there is a statistically significant influence of air temperature mainly on VOC concentrations and emission. As for the influence of RH, there is a statistically significantly influence mainly on NH 3 concentrations and emissions. All correlations between odorant concentration and emission and RH are negative correlations. Almost all correlations between odorant concentration and emission and air temperature are positive correlations. Only in one odour source in the fermentation preparation section (biogas plant A) was a negative correlation between VOC concentration and air temperature observed. Well-fitted dependences have been observed in all elements of the biogas plant technological line, with the exception of the mechanical treatment building. In the remaining technological sections, statistically significant correlations were found between air temperature and VOC concentrations and emission as well as between air humidity and NH 3 concentrations and emissions with significance (p-values) at the level of 0.0003-0.0217. Some authors [34] report that well-fitted dependences should correspond to R 2 ≥ 0.60. Assuming this, the best correlation coefficients (marked in grey in the table) were recorded mainly between VOC concentrations and emission and air temperature and are related to the fermentation preparation section, emission from roof ventilators and measurements in front of technological plant building entrances. The influence of air humidity at the indicated level was also observed for VOC concentration.
The obtained results indicate that two measured microclimatic parameters-air temperature and RH-did not explain all of the variability of the odorant emissions at biogas plants. The impact of microclimatic parameters on the odour and odorant concentration has been studied in other research e.g., in pig production systems and the results of it are also varied. The effect of inside temperature and relative air humidity was studied by Choi et al. [31]. The authors obtained correlation coefficients between odorant concentration and air temperature at the level of R = 0.62 and with relative humidity at the level of R = −0.43. The signs of the correlation coefficients were the same as in this study. Also, Wang et al. [44] and Hugle and Andree [45] noted a strong positive correlation between the concentration of odours/odorants and air temperature. Similar dependencies were found in research by Le et al. [33] and Schauberger et al. [38]. They found a positive correlation between the air temperature and odour emissions. In turn, in [46], similar observations were made in relation to the relationship between air temperature and odour and ammonia concentration (increased temperature was found to significantly increase odour concentration in the exhaust air p = 0.004, odour emission p = 0.002, ammonia concentration p = 0.043 and ammonia emission p = 0.052), but different in relation to the relationship with air humidity. In turn, in research obtained by Mielcarek-Bocheńska [34] and Guo et al. [30] the changes of odour concentration were opposite to temperature changes. They showed a negative correlation of the odour concentration with the air temperature. In addition, Yao et al. [31] stated that the concentrations of VOCs were correlated negatively with microclimate parameters (among others inside temperature). Miller et al. [36] found that odour emissions decreased as temperature and relative humidity increased. According to Yao et al. [33], the relationships between microclimate parameters and odour compounds concentrations (ammonia, dimethyl sulphide, dimethyl disulphide) change during the year and were not clear. Furthermore, Sun et al. [39] noted that the simple correlations between odour concentration and microclimate parameters (among others inside temperature) were not statistically significant. In reference [46], increase of RH in the exhaust air was found to significantly increase ammonia emission and ammonia concentration in the exhaust air, thus obtaining different results from this work. Similar conclusions were drawn in reference [47]-change of temperature and humidity in the environment may lead to change of odorant emission e.g., by affecting the volatilization or by affecting the activity of microorganisms producing odorous compounds. The authors suggest that reducing the moisture content reduces the odour production in manure, and this effect may be explained by less anaerobic conditions at low moisture levels [47]. The literature review shows that various settings of the ventilation system or the lack of it may also be the cause of different results of the relationship between odour and odorant concentration and meteorological parameters [22].

Correlation between Pollution Factors
The analysis of the correlation between the tested chemical compounds with the division into individual sections showed a significant relationship for biogas plant A for the mechanical treatment section: NH 3 -VOCs R = 0.6137, p = 0.0009, for the oxygen stabilization section; NH 3 -VOCs R = 0.5589, p = 0.0030 for emission from roof ventilators; and NH 3 -VOCs R = 0.7930, p = 0.0000. In the case of biogas plant B, the correlation analysis showed significant relationships for the waste storage section: NH 3 -VOCs R = 0.4379, p = 0.0053, for mechanical treatment; NH 3 -VOCs R = 0.4001, p = 0.0429, fermentation preparation for point g; VOCs-H 2 S R = 0.9727, p = 0.0000 and oxygen stabilization; NH 3 -VOCs R = 0.3891, p = 0.0494, for emissions from roof ventilators; and NH 3 -VOCs R = 0.3016, p = 0.0298. Most of the above-mentioned correlations observed for biogas plant B are lower compared to the dependences for biogas plant A (except for the dependence in the fermentation preparation section).

Conclusions
In this study, odorant concentrations and emissions were measured together with the air T and RH to determine the impact and importance of the microclimatic conditions on the level of odour emissions at municipal waste biogas plants. Simple correlation analyses were made to evaluate the natures of the relationships between the odorant concentration and emission and air temperature and relative humidity. The conclusions are summarised below.
The mean VOC and NH 3 concentrations vary depending on the stage of the technological line of the analysed municipal waste biogas plants and are in the following ranges: 0-38.64 ppm (0-0.169 mg/m 3 ) and 0-100 ppm (0-69.653 mg/m 3 ), respectively. According to best available techniques (BAT) conclusions for waste treatment channelled VOC and NH 3 emissions to air from biological waste treatment should not exceed values 40 mg/m 3 and 20 mg/m 3 , respectively [48]. The odorant concentrations and emissions correlated with statistical significance with air temperature and relative humidity. The air temperature primarily influences VOC concentrations and emissions. Most correlations between these two variables are positive. Air humidity primarily influences NH 3 concentrations and emissions. Most correlations between these two variables are negative. The impact and importance of microclimatic factors varies depending on the individual stages of the technological line of a biogas installation. The highest correlation coefficients were observed in the fermentation preparation section; in front of the technological plant building entrances; and in the air discharged by roof ventilators. Correlations between T and VOC reached values of up to R = −0.98, R = 0.79, R = 0.82. Similarly, correlations between RH and NH 3 peaked at R = −0.77 and R = −0.84, −0.76, respectively. The smallest influence of microclimatic factors was observed at the beginning of the technological line of the biogas plant in the waste storage section and mechanical treatment hall. This is due to the greater impact of the type and quality of waste accepted at the plants and the use of an extraction system in these sections of the process line. Not all variability of the odorant concentrations and emission are explained by the air temperature and humidity. The differences in odorant concentration may be due to a lot of other factors: type of waste processed; waste storage time; type of technological processes used in a biogas plant; use of negative pressure installations; and other microclimatic factors.
The analysis of correlation between individual odorants showed significant relationships between volatile organic compounds and ammonia for most stages of the technological line of both biogas plants. The strongest correlation was noted for the fermentation preparation section at biogas plant B and for emissions from roof ventilators at biogas plant A. In the case of technological sewage pumping stations in biogas plant B, a significant relationship was also observed between VOCs and H 2 S. All these correlations were positive correlations.
A literature review of the subject and the results obtained in this study indicate that the natures of the relationships between microclimatic conditions and odorant concentration and emission are not clear and require further research. The obtained results may be helpful in preparing strategies to reduce odours from waste treatment plants. Considering the practical term of the research, it may be Energies 2020, 13, 6463 25 of 27 of great importance, especially for the managers of MWBPs. The analysis of individual technological processes, following the example of two biogas plants using different technologies, may be helpful in controlling these processes depending on the prevailing microclimatic conditions. The research could also be applied in formulating the principles of plant policy aimed at minimizing the emission of emitted compounds.
Author Contributions: M.W.: data collection, investigation, writing-original draft preparation; A.K.: conceptualization, methodology, supervision; K.L.-S.: writing-review and editing, visualisation, validation. All authors have read and agreed to the published version of the manuscript.
Funding: Work was co-financed by Warsaw University of Technology, Scientific Discipline of Environmental Engineering, Mining and Energy.

Conflicts of Interest:
The authors declare no conflict of interest.