**4. Results**

#### *4.1. Meteorological Conditions in the Base Year*

On the basis of calculations made with the WRF/CALMET model, the analysis of meteorological conditions significantly affecting the dispersion of pollutants in 2018 was performed (Figures 7–15). Both spatial variability of selected parameters per year and monthly variability were discussed based on selected locations—Bogatynia (the largest city in the commune), Sieniawka (located in the southwest of the examined area) and Działoszyn (located in the north of the examined area).

Wind as a parameter shaping the rate and direction of the pollution spread is one of the most important parameters for the dispersion process. With complex terrain, it can be highly variable, and when dealing with such a specific object as a deep open-pit mine, it is necessary to include it in the model as a field variable in space. This is clearly marked when analyzing the prevailing wind directions, where a significant change in the air flow occurs in the area of the open pit and, while the wind is southern for the most part of the area, in the pit, it changes direction to the west or east. The significant dynamics of the dominant wind direction can also be seen in the depression in which Zittau and Hrádek nad Nisou are located, where in 2018, winds from the western sector prevailed.

The calculations show that the wind field in the analyzed area is characterized by variability of annual mean speeds in the range of 3.8 to 4.4 m/s. The highest speeds (above 4.2 m/s) characterize the northwest and southeast parts of the analyzed area (Figure 7). Reduced wind speeds are characteristic of the valley areas (around Bogatynia, as well as Zittau and Hrádek nad Nisou).

Based on the hourly wind speeds and directions, annual wind roses were prepared for selected grids within meteorological domains representing the following cities: Bogatynia, Sieniawka and Działoszyn (Figure 8). The obtained calculation results indicate that, in 2018 in Bogatynia, winds from the southern sector (19%) definitely dominated; the highest speeds were also recorded in this sector. In Sieniawka, the share of winds from the southern sector was also the largest (13%), but the share of winds from the NW and WNW sectors was also significant (around 10% per year). In Działoszyn, apart from southern winds (13%), there was a large share of SSW winds (11%). Only in Działoszyn, the share of winds from the direction in which KWBT is located (in this case, the southern sector) was significant. The highest amount of calm winds was recorded in Bogatynia (1.8%); in other cities, this value was around 1%.

In the case of the analysis of wind speed values, higher values of monthly mean wind speeds (except for the summer months, i.e., June–August) were observed in Bogatynia than in Działoszyn and Sieniawka (Figure 9). According to the distribution of monthly mean wind speeds in 2018, higher wind speeds (above 5 m/s) occur in the autumn and winter months (October, January, November, and December). The spring–summer period (May–August) was characterized by lower wind speeds (below 3.5 m/s).

On the other hand, the classification of wind speeds for selected grids from the studied area shows that the most frequently occurring were the winds with speeds in the 3–5 m/s range (30.5–38.1%), referred to as mild winds (Figure 10). Weak winds, i.e., 1.5–3 m/s, and winds with speeds above 5 m/s occurred with similar frequency in all towns (around 21%–27%). In Bogatynia, more than 1% of winds with very high speeds >10.8 m/s were observed. In 2018, wind speeds less than 1.5 m/s occurred in 9.9% of cases in Działoszyn, 11% in Sieniawka and 12.6% in Bogatynia.

According to the Climate Monitoring Bulletin of Poland, published annually by the Institute of Meteorology and Water Management, 2018 has been classified as an extremely warm year [40,41]. The analyses show that the annual mean air temperature in 2018 in the studied area varied from around 9 ◦C to over 10 ◦C, while in most of the area, it was around 9.5 ◦C. Temperatures above 10 ◦C occurred in the northern part of KWBT and in the vicinity of ELT and in the northwest of the area (Figure 11). The coldest month in 2018 and the only one with the average temperature below 0 ◦C was February (around −4 ◦C). March was also quite cool (monthly mean around 0 ◦C) (Figure 12). The month with the highest mean (>20 ◦C) was August. At the same time, from April to October, monthly mean temperatures were above 10 ◦C. The characteristics of monthly mean temperatures indicate that the Bogatynia and Działoszyn regions are thermally similar, while the Sieniawka area is slightly cooler.

**Figure 7.** Spatial distribution of annual mean speeds and dominant wind direction determined by the WRF/CALMET model in 2018 (source: own study).

**Figure 8.** The distribution of wind directions and wind speeds determined by the WRF/CALMET model in the meshes of the meteorological grid corresponding to the location of selected towns in 2018 (Source: Own study): (**a**) Bogatynia, (**b**) Sieniawka, (**c**) Działoszyn.

**Figure 9.** Monthly mean wind speeds determined by the WRF/CALMET model in selected towns in 2018 (source: own study).

**Figure 10.** The frequency of wind speeds in specific ranges in selected towns in 2018 (source: own study).

**Figure 11.** Spatial distribution of annual mean air temperature values determined by WRF/CALMET in 2018 (source: own study).

**Figure 12.** The course of the monthly mean air temperature determined by the WRF/CALMET model in selected towns in 2018 (source: own study).

The Pasquill atmospheric stability classes, which describe the vertical air turbulence associated with temperature gradient and wind speed, are very important parameters for the dispersion of pollution. The model adopts six stability classes (PGT1–PGT6). Classes 1 and 2 are unfavorable for the dispersion of pollutants due to the fact that the trail of exhaust gases rises and falls due to intense turbulence. Classes 5 and 6, in which inverse conditions occur, are very unfavorable; the pollutants remain in the given area at low altitudes because they have no conditions for dispersion. The incidence of individual classes was determined for the towns of Sieniawka, Działoszyn and Bogatynia (Figure 13). The calculations show that, in the vicinity of the towns in question, in 2018, the most common was the atmospheric stability class 4, which represents neutral conditions (around or over 50% of cases). Class 1, defined as extremely unstable conditions, was very rare (less than 1% of cases). However, unfavorable classes 5 and 6 occurred in a total of around 21–30% cases during the year, most often in Sieniawka.

The year of 2018 was a dry year, which is also confirmed by the spatial distribution of the annual total precipitation in the area (Figure 14). Such conditions adversely affect the rise of dust pollutants, which, in the case of a large-scale object such as a mine, may contribute to the occurrence of high concentrations. In 2018, the annual rainfall totals in the studied area ranged from around 600 (in the west in the area of the Nysa Łu ˙zycka Valley) to 800 mm (in the east in the area of the Izera Mountains). The analysis of the annual rainfall totals in selected locations indicates a relatively small variability: from around 590 mm in Sieniawka to around 620 mm in Bogatynia and Działoszyn. The analysis of the variability of precipitation in 2018 shows that the lowest rainfall occurred in February and November—below 10 mm—while the highest was measured in June and December (77–97 mm) (Figure 15). In Sieniawka, high rainfall was noted also in January, where the total was around 30 mm higher than in Bogatynia and Działoszyn. For the remaining period of the year, rainfall totals in Sieniawka were slightly lower than for Bogatynia and Działoszyn, especially in May and June. In the remaining months of the year, the differences in the total precipitation between individual towns did not exceed 10 mm.

**Figure 13.** Share of the Pasquill atmospheric stability classes determined by the WRF/CALMET model in selected towns in 2018 (source: own study).

**Figure 14.** Spatial distribution of annual precipitation determined by the WRF/CALMET model in 2018 (source: own study).

**Figure 15.** Monthly precipitation totals determined by the WRF/CALMET model in selected towns in 2018 (source: own study).

#### *4.2. The Results of Air Quality Modeling*

Human health protection is the main criterion for air quality assessment. This study focuses on the analysis of long-term effects; therefore, the values of annual mean concentrations of PM10 and PM2.5 were assessed, with limit levels of 40 μg/m<sup>3</sup> and 25 μg/m3, respectively. The assessment of changes resulting from the implementation of individual scenarios is presented as a relative difference.

Additionally, the shares of the main particulate matter emission sources in concentrations in two selected profiles of different emission nature were analyzed for scenarios 1 and 2. The first profile, around 40 km long, running from Działoszyn through Zatonie to Hrádek nad Nisou, clearly shows the impact of the energy complex. The second profile, around 26 km long, running from Olbersdorf through Zittau, Bogatynia and Hermanice to Frydlant, reflects the characteristics of the concentration field associated with emissions from household heating.

#### 4.2.1. Annual Mean PM10 Concentrations

The obtained results show that, in the year of diagnosis, the annual mean PM10 concentration practically in the entire analyzed area did not exceed the limit value (Figure 16). The area located near the ash storage area, where maximum concentrations reached over 100 μg/m3, was the exception. However, this area is located outside built-up areas, within a forest complex, so its range is limited. High concentrations also occur within the open pit, which is justified by the industrial characteristics of the area. Within the settlement area, the annual mean concentrations of PM10 are in the range of 18–26 μg/m<sup>3</sup> (45–65% of the permissible level). The highest concentrations in built-up areas occurred in Bogatynia and reached 34 μg/m<sup>3</sup> (85% of the permissible level). The lowest concentrations in the studied area occurred in its northwestern and southeastern regions (in elevated areas, marked in dark green).

The model calculations show that the implementation of corrective actions discussed in scenario 1 (Figure 16b) regarding the power complex facilities results in a maximum 70% reduction in concentrations nearby ash storage area. The greatest effectiveness is expected in the immediate vicinity of the ash stockyard (area marked with an arrow). Within a radius of 4 km from the main emission sources, a maximum reduction of 5% can be expected. The greatest effectiveness of actions in development areas is expected in the districts of Bogatynia–Zatonie (30–40%) and Trzciniec (10–15%), but also in Sieniawka (5–10%) and on the German side in the cities of Hirschfelde and Drausendorf and in eastern areas of Zittau (5–10%).

On the other hand, of the analyses carried out for activities indicated in scenario 2, they will be most effective in built-up areas (Figure 16c). In Bogatynia, as much as 20–40% decrease in annual mean PM10 concentrations is estimated. In larger cities, concentrations will fall by around 10–20%, and, in the remaining areas, the change will not exceed 5–10%.

The best effect was obtained for the compilation of scenarios 1 and 2, which is scenario 3 (Figure 16d), in which case at least a 10% decrease in concentrations can be expected basically in the entire analyzed area within the Polish borders. A small impact on the decrease in concentrations on the Czech and German side may, however, prove that the impact of both the energy complex and local emissions associated with the combustion of fuels in household heating devices is very limited.

Analyzing the share of emission sources in the annual mean PM10 concentrations in the profile between Działoszyn and Hrádek nad Nisou (Figure 17), it can be concluded that the inflow of pollutants from outside the computational domain has a very significant share, which is estimated at around 16 μg/m<sup>3</sup> basically along the entire length of the profile. The concentrations related to household heating, which are clearly marked only on the Polish side of the border, are also an important component of the profile. The closer the sources of the power complex, the more significant the increase in their share; however, their range of influence is very limited (a few km). In the immediate vicinity of the ash storage area, its share is similar to the share of the inflow. The impact of KWBT is practically limited to industrial areas and may be associated with local high concentrations, even exceeding the target values. At the same time, it is clear that individual operations (transshipment at a coal yard, dumping or mining) have a significant impact basically in the place of their performance.

**Figure 16.** Annual mean PM10 concentrations in 2018: modeled values and relative differences for the analyzed scenarios (source: own study): (**a**) in 2018, (**b**) relative difference for scenario 1, (**c**) relative difference for scenario 2, (**d**) relative difference for scenario 3.

**Figure 17.** Profile of annual mean concentrations of PM10 between Działoszyn and Hrádek nad Nisou, considering the shares of individual source groups in 2018 (source: own study).

The implementation of the measures from scenario 1 will result in a very large decrease in concentrations, and so air quality standards will be met throughout the entire length of the analyzed profile (Figure 18). In this case, the most important factor shaping the air quality in the studied area will be the inflow of pollution. The impact of the power complex will be comparable to the current impact of emissions from household heating.

**Figure 18.** Profile of annual mean concentrations of PM10 between Działoszyn and Hrádek nad Nisou, considering the shares of individual source groups for scenario 1 (source: own study).

In the profile from Olbersdorf to Frydlant, the inflow of pollutants from outside of the studied area has the largest share in annual mean PM10 concentrations (Figure 19), similar to the profile analyzed earlier. However, in Poland, emissions related to household heating also have a very high share in concentration. This is particularly evident in Bogatynia, where concentrations from this type of emission can reach up to 12 μg/m3. Transport is the third most important group of sources in the studied profile (maximum annual mean PM10 concentrations reach up to around 3 μg/m3). The impact of the mine is relatively small, and it is significant only within the open pit (mining). At the same time, there are no exceedances of air quality standards virtually along the entire length of the profile.

The implementation of the anti-smog resolution will practically eliminate the impact of emissions from household heating, which will result in a significant reduction in the concentrations associated with it. This is clearly marked on the analyzed profile (Figure 20).

**Figure 19.** Profile of annual mean concentrations of PM10 between Działoszyn and Hrádek nad Nisou, considering the shares of individual source groups in 2018 (source: own study).

**Figure 20.** Profile of annual mean concentrations of PM10 between Olbersdorf and Frydlant, considering the shares of individual source groups in scenario 2 (source: own study).

#### 4.2.2. Annual Mean PM2.5 Concentrations

The results of the conducted model tests indicate that the lowest values of the annual mean PM2.5 concentration occurred in the eastern part of the area, where they do not exceed 14 μg/m<sup>3</sup> (56% of the limit value) (Figure 21a). In the central part of the area, concentrations remain in the range of 14–16 μg/m3. Outside the industrial area, the highest annual mean PM2.5 concentrations occur in Bogatynia, where they reach around 23 μg/m<sup>3</sup> (92% of the limit value). In other locations in the studied area, higher annual mean concentrations of PM2.5 were also recorded in Trzciniec Dolny (84% of the limit value), Zatonie (79% of the limit value), Sieniawka (76% of the limit value) and Działoszyn (70% of the limit value).

**Figure 21.** PM2.5 annual mean concentrations in 2018: modeled values and relative differences for the analyzed scenarios (source: own study): (**a**) in 2018, (**b**) relative difference for scenario 1, (**c**) relative difference for scenario 2, (**d**) relative difference for scenario 3.

Minimizing measures implemented in scenario 1 will locally (nearby ash storage area) result in a maximum 75% reduction in annual mean PM2.5 concentrations (Figure 21b). However, the range of maximum reductions is smaller than it was in the case of PM10, which is due to the nature of the dust associated with the emitters of the power complex (in particular, KWBT and ash storage area). Emissions from this type of facility primarily concern mineral dust with larger fractions and lower volatility. The greatest efficiency is expected in the immediate vicinity of the ash storage area. Within a radius of 1.5 km from the main sources, a maximum reduction of 5% can be expected. Analyzing the development areas, the direct significant impact of the application of the measures resulting from scenario 1 can only be seen in Zatonie and Trzciniec.

The implementation of scenario 2 will result in similar reductions in PM2.5 concentrations, as in the case of PM10, but the reduction range is greater (Figure 21c).

As in the case of PM2.5, the best effect was obtained for scenario 3: a 10% reduction in concentration values was obtained practically throughout the entire analyzed area within the Polish borders (Figure 21d).

In the profile between Działoszyn and Hrádek nad Nisou (Figure 22), there are no exceedances of air quality standards set for annual mean concentrations of PM2.5. Analyzing the shares of emission sources, it can be stated that, also in the case of this pollution, the inflow of pollutants from outside the computational domain has a very significant share, which is estimated at around 13 μg/m<sup>3</sup> on the entire length of the profile. An important component of the profile is also concentrations related

to household heating. As the sources of the power complex approach, their share increases, and the impact is much smaller than it was in the case of PM10.

**Figure 22.** Profile of annual mean concentrations of PM2.5 between Działoszyn and Hrádek nad Nisou, considering the shares of individual source groups in 2018 (source: own study).

The emission reduction related to the implementation of scenario 1 will significantly reduce the annual mean concentrations of PM2.5 and the share of concentrations from local emissions will be lower than the concentrations from inflow (Figure 23). The impact of emissions related to the ash storage area will practically disappear and the impact of emissions from the mine will be limited to its area. The largest share of local emissions in concentrations will be associated with the impact of local heating sources.

**Figure 23.** Profile of annual mean concentrations of PM2.5 between Działoszyn and Hrádek nad Nisou, considering the shares of individual source groups for scenario 1 (source: own study).

In the profile between Olbersdorf and Frydlant, the annual mean concentration of PM2.5 also has the most significant share of pollution (Figure 24). Locally, however, in Bogatynia, there is a very large share of emissions associated with household heating systems. The concentrations in Bogatynia are approaching the limit value, but they do not exceed it. Other local sources (also emissions related to KWBT) are much less important.

**Figure 24.** Profile of annual mean concentrations of PM2.5 between Olbersdorf and Frydlant, considering the shares of individual source groups in 2018 (source: own study).

The implementation of the anti-smog resolution in the examined area will practically result in a very large reduction of local emissions and thus its impact on the formation of air quality in the examined area (Figure 25). The more significant impact of KWBT is practically limited to the open-pit area, but its share is similar to the share of emissions from local transport.

**Figure 25.** Profile of annual mean concentrations of PM2.5 between Olbersdorf and Frydlant, considering the shares of individual source groups in scenario 2 (source: own study).

#### *4.3. Assessment of the E*ff*ectiveness of the Implementation of Various Scenarios Based on the Results of Health Risk Analyses*

The obtained results of health risk assessment analyses indicate the 6% general improvement in air quality from the implementation of scenario 1 (all activities specified for the mine) in the entire commune leads to the reduction of health effects by nearly 5%. This confirms that the mine itself has a negligible impact on the health of residents (current estimated health impact—base scenario—is approximately 7%; see Table 7).


**Table 7.** PM2.5 concentration and related premature deaths calculated for emission scenarios.

A completely different situation is observed in the case of scenario 2 (introduction of provisions resulting from the implementation of regional measures, i.e., anti-smog resolution). A noticeable decrease of over 7% in the level of pollution in the commune when implementing scenario 2 causes over a two-fold higher (14.4%) decrease in health risk (premature deaths).

Both abovementioned scenarios can be implemented independently and simultaneously by separate units based on their competence (mine authorities and local government). The implementation of both scenarios together (scenario 3) causes a significantly higher improvement in air quality and reduction in health risk in the studied area (e.g., changes in the values of the weighted population annual average concentration). After implementing the scenario 3 measures in the analyzed area, the 18% decrease in the number of premature deaths can be expected (estimated according to the baseline scenario) when the concentration of average annual particulate matter is reduced by just over 13%.

In each case (scenarios), the spatial distribution of health effects reduction changes is not regular (Figures 26–28). The analysis of the obtained results of premature death changes distribution for the areas of the commune shows little impact on the populated area for scenario 1, where in the vast majority of areas (nearly 73% of the area), the observed changes do not exceed 5% (Figure 26). Despite this, in this scenario, there are close to 1% areas with over 30% improvement in health. Generally, in 7% of areas, the number of premature deaths is reduced by at least 15%.

The situation looks much better with the implementation of scenario 2 (Figure 27). For this scenario, a decrease in the number of premature deaths associated with long-term exposure to PM2.5 lower than 5% is already observed in nearly 58% of areas. An improvement in health—that is, at least a 15% reduction in premature deaths—has already been noticed in 12% of areas.

Implementation of scenarios 1 and 2 at the same time remains the most effective and health-promoting solution (for scenario 3, see Figure 28). Assessment results for scenario 3 show that only in 25% of the areas, the improvement in health is less than 5%, while a reduction in premature deaths of over 15% is observed almost in 33% of the commune's area.

**Figure 26.** Spatial distribution of relative changes (reductions) in the number of premature deaths for scenario 1 relative to the baseline scenario.

**Figure 27.** Spatial distribution of relative changes (reductions) in the number of premature deaths for scenario 2 relative to the baseline scenario.

**Figure 28.** Spatial distribution of relative changes (reductions) in the number of premature deaths for scenario 3 relative to the baseline scenario.

#### **5. Summary and Final Conclusions**

Mathematical modeling is one of the tools whose application within the air quality managemen<sup>t</sup> system is crucial. Commonly, mathematical models apply, among others, when developing plans and programs of corrective actions aimed at achieving and/or maintaining air quality at an appropriate level; analyses of the effectiveness of implementing specific solutions; analyses of the impact of air pollution on various elements of the environment, including human health; air quality forecasting, considering changes in the activity of emission sources and meteorological conditions, and providing the public and decision-makers with adequate information. In the case of modeling systems, it is important that the input data used in the analyses (including emission, meteorological and topographic data) are current and accurate. This has a significant impact on the results obtained by modeling and their credibility and representativeness. An important element of this type of analysis is the assessment of discrepancies in the results of model calculations, including the relative error rate of the modeling result in relation to the results of measurements made at measuring stations using devices compatible with or equivalent to the reference method specified for a given pollutant. This type of procedure was undertaken as part of the work in which the selected case study was the area of the commune in Poland, whose character was considered unique and complex due to the geographical location and diversity of the terrain and the functioning energy complex. Analyses were carried out considering the defined activities for three emission reduction scenarios: (1) scenario related to changes in emissions in the analyzed mine resulting from the minimizing measures indicated in the report on the mine's environmental impact, (2) scenario resulting from the "anti-smog" resolution in force in the Lower Silesian Voivodship and (3) scenario compiling the abovementioned scenarios. Additionally, to demonstrate the effectiveness of planned preventive and corrective actions taken in the analyzed area, a health risk analysis was performed. All analyses were made considering the changes in the distribution of pollutant concentrations within the boundaries of the commune.

The results of the conducted analyses indicated that the lowest values of the annual mean PM2.5 concentration occurred in the eastern part of the studied area and did not exceed 14 μg/m<sup>3</sup> (56% of the limit value). The implementation of activities resulting from the considered scenario 1 will result in a reduction in annual mean concentrations of PM2.5 (maximum 75% in industrial area), mainly due to the maximum reduction in activities carried out within the furnace ash storage area, as well as the construction of a new ash conveyor, thanks to which the storage area will be practically taken out of operation. The achieved range of maximum concentration reduction was smaller than in the case of PM10, which may be due to the nature of the dust associated with the emitters of the energy complex (in particular, KWBT and ash storage area). Emissions from this type of facility primarily concern mineral dust of larger fractions that are transported in the atmosphere to a much lesser extent. The greatest result is expected in the immediate vicinity of the ash storage area. Within a radius of 1.5 km from the main sources, a slight reduction in concentrations, not exceeding 5% of the current state, can be expected. For this reason, actions taken under this scenario do not have a major impact on the health of the surrounding residents.

The performed analyses have shown that the implementation of scenario 2 will result in similar levels of PM2.5 concentration reductions as PM10, but with a much vaster spatial range, which translates into a much higher impact on the population threat related to air quality, with an estimated 14% reduction.

The best e ffect was obtained for scenario 3, where, in total, the average 10% reduction in concentration values was obtained practically in the entire analyzed area within the borders of the community. Therefore, the implementation of both scenarios seems to be the most e ffective for limiting the health risk associated with the exposure of residents to particulate matter (estimated health e ffect reduction is almost 20%). The full implementation of scenario 3 shows that only in 25% of the area was the expected improvement of health lower than 5%. A significant decrease in health risk (more than 15%) was observed in as much as one third of the studied area.

Mathematical modeling as a tool should be disseminated, and the data used in the model should be available not upon special request but due to the obligation to provide access to information on the environmental impact of an installation, plant and/or the group of emitters in a given area.

**Author Contributions:** Conceptualization, I.S., M.P. and K.S.; methodology, I.S., M.P. and K.S.; software, M.P. and K.S.; validation, I.S., M.P. and K.S.; formal analysis, I.S., M.P., K.S., D.K., M.Z. and K.K.; investigation, I.S., M.P., K.S., D.K., M.Z. and K.K.; writing—original draft preparation, I.S., M.P., K.S., D.K., M.Z. and K.K.; writing—review and editing, I.S., M.P., K.S., D.K., M.Z. and K.K.; visualization, M.P. and K.S.; supervision, I.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was co-financed within the "Excellent Science" program of the Polish Ministry of Science and Higher Education.

**Acknowledgments:** The authors of the study thank for their cooperation: Anita Kuli´s from ONE WAY Anita Kuli´s (Zielonka, Poland), Rafał Skorupi ´nski, as well as Milena Gola-Kozak and Dorota Sucholas from PGE Górnictwo i Energetyka Konwencjonalna S.A., Kopalnia W˛egla Brunatnego Turów (Bogatynia, Poland).

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
