The moisture content of the coniferous wood, cereal straw, and miscanthus pellets was on a similar level of 4.02, 3.91, and 3.45 %, respectively. The coal was characterized by a moisture level of 5.32. The pellets were characterized by a high content of volatile matter in comparison to that of coal. The volatile matter (VM) of the pellets ranged between 72 and 82% wt. Coal had a higher ash content than the pellets. The ash content of the pellets was 6.64, 2.06, and 0.37% for cereal straw, miscanthus, and coniferous wood, respectively. The lower heating value (LHV) for coal was higher than that for biofuels and was 27 MJ/kg. For the pellets from biomass, the LHV ranged from 17.4 (coniferous wood) to 15.3–15.7 MJ/kg for cereal straw and miscanthus, respectively. Elementary analysis showed that coal was characterized by a higher content of carbon, nitrogen, and chlorine and a lower content of hydrogen in comparison to biofuels. The chemical composition of biofuels and coal is similar to that reported in other studies [
6,
11].
3.1. Emission Factors
The emission factors were calculated by dividing the pollutant mass flow by a fuel consumption per unit of time [
25]. The fuel consumption per unit of time was calculated by dividing the amount of fuel burned (the fuel in the fuel feeder was weighed before and after combustion) by the boiler operation time. The emission factors (EFs) for basic gaseous pollutants, i.e., CO
2, CO, and SO
2, are presented in
Table 5.
As for the biofuel combustion, the lowest CO emission was observed for the cereal straw combustion and amounted to 16.8 g/kg, while for the miscanthus and coniferous wood combustion, it amounted to 56 g/kg and 62 g/kg, respectively. On the other hand, the CO emissions during coal combustion was at a level of 6.7 g/kg. The CO/CO
2 ratio for the coniferous wood and cereal straw pellets was on a level of 0.03. However, that for the pelletized miscanthus was 2 times higher. In the case of coal combustion, the CO/CO
2 ratio was 0.01 and was lower than the values obtained for biofuels. This result can be explained by the different fuel oxidation efficiency in the boiler. A similar relationship has been demonstrated by Verma et al. [
26]. As for the SO
2 emissions, the three biofuels were similar and ranged between 0.3 and 0.6 mg/kg, whereas coal achieved a value of 5.1 mg/kg. It is well known that the emission factor for SO
2 depends on the sulphur content in fuel. In the presented studies, the sulphur content in biofuels was lower than that in coal (0.08–0.4% and 0.81% for biofuels and coal, respectively).
The PM, TC, OC, and EC emission factors, obtained during the combustion of biofuels and coal, are presented in
Table 6. For the biofuel combustion, the average PM emission factors ranged from 1.7 g/kg (coniferous wood pellets) to 2.7 g/kg (cereal straw pellets) and were lower than the emissions from the coal combustion, which amounted to 10.2 g/kg. It should be noted that the average PM emissions during the combustion of biofuels was about 5 times lower than that during the combustion of coal in the same boiler.
It is known that three types of particles are mainly emitted during combustion, i.e., soot, organic particles, and inorganic particles. Soot and organic particles are the result of the incomplete combustion of fuels. As in the case of CO, the emission of organic particles depends on the combustion efficiency [
26]. The PM emitted from combustion processes contains both elemental (EC) and organic (OC) carbon. The total carbon content (TC) in the exhaust gas is defined as the sum of OC and EC. The obtained results indicate that the emission factors of TC (EF
TC) from the combustion of biofuels ranged from 203 to 454 mg/kg, and they were much lower than those from coal combustion. In the case of coal combustion, the EF
TC was 4,674 mg/kg (
Table 6).
The research results indicate that the OC emission factor (EFOC) ranges from 127 to 232 mg/kg in the case of pellet combustion, while for coal, it is 1,265 g/kg. In the case of coal combustion, the EC emission factor (EFEC) was higher than the EFOC. A different correlation was observed for pelletized coniferous wood. On the other hand, the EFOC and EFEC emissions were comparable to those for pelletized cereal straw and miscanthus.
The measured EF
OC and EF
EC for the pelletized coniferous wood are at a similar level to those reported by Shen et al. [
15] for residential wood combustion in a typical cooking stove.
The contents of OC and EC in PM differently depended on the fuel type (
Figure 1). The EC mass fractions in the particles emitted during coal combustion (33.3 wt%) were higher than those emitted during biofuel combustion (2.4–10.7 wt %), while the opposite was observed for OC (12.4 wt% and 4.4–11.2 wt% for the coal and biofuels, respectively). The mass concentrations of OC in PM were very similar for coal (12 %), pelletized miscanthus (11%) and pelletized coniferous wood (10%), whereas for pelletized cereal straw, the OC content in PM was at a level of 4.4 wt %. This phenomenon can be explained by the properties of biofuels, since herbaceous plants, such as straw, have a higher burning rate than ligneous plants (pine and pellet fuels).
The OC/TC and TC/PM ratios are useful indicators in the identification of the source emission for carbonaceous PM [
27,
28]. The determined OC/TC ratios were as follows: 0.8, 0.4, and 0.5 for coniferous wood pellets, cereal straw pellets, miscanthus pellets, respectively. In the case of miscanthus pellets, the OC/TC ratio was similar to that for miscanthus pellets, as was reported in the literature (0.52 ± 0.26) [
29]. In the case of coal combustion, the OC/TC ratio amounted to 0.3.
The OC/PM ratios for coal, pelletized coniferous wood, and pelletized miscanthus were at a level of 0.1–0.12 and were higher than those for pelletized cereal straw (0.04). The results indicate that the OC/EC ratios for biomass burning are higher than those for coal (
Figure 2). A similar correlation was observed by Novakov et al. [
30].
In the present study, the OC/EC ratio ranged from 0.9 to 4.0 for biofuels. A lower OC/EC ratio has been reported for pelletized cereal straw (0.9) and pelletized miscanthus (1.1) in comparison to pelletized coniferous wood combustion (4.0) (
Figure 2). A similar correlation was observed by Fernandes et al. [
31], who presented an average OC/EC ratio of 0.85 for softwood combustion and between 3.14 and 4.39 for hardwood combustion in a woodstove. On the other hand, Zhang et al. [
32] provided an OC/EC ratio of 2.8 ± 1.3 for burning wood in a similar furnace.
Some studies have shown EC/TC ratios of between 0.01 to 0.11 for fuel combustion [
33,
34,
35]. Bølling et al. [
35] reported that the EC/TC ratio for incomplete combustion ranges from 0.5 to 0.8.
3.2. Molecular Diagnostic Ratios
The molecular diagnostic ratios of the defined pairs of individual compounds have often been used as tracers of different source categories of polycyclic aromatic hydrocarbons [
36,
37,
38,
39]. During combustion, PAHs may be formed from organic matter under oxygen-deficient conditions. Pyrosynthesis and pyrolysis are the two main mechanisms that can explain the formation of PAHs. It is well known that the mechanism formation of PAH during combustion includes: (1) Radical reactions, (2) Diels–Alder condensations, and (3) an ionic reaction mechanism. The thermodynamics of combustion favor the mechanism of radical formation. The emission factors of PAHs (EF
PAH) are strongly dependent on the properties of the fuel, the type of the furnace, and the combustion conditions. Venkataraman et al. [
18] reported that the emission factors of the total PAH from wood, briquette, and dung-cake combustion range from 2.0 to 3.2 mg/kg, 2.8 to 3.0 mg/ kg, and 3.1 to 5.5 mg/ kg, respectively. PAHs and CO are both products of inefficient combustion and are therefore dependent on the temperature of combustion [
40]. Levendis et al. [
41] found a relationship between the CO and PAHs from the combustion of coal and waste tires. On the other hand, Rantanen [
42] and Saez et al. [
43] confirmed higher PAH emissions under inefficient combustion conditions, but did not find a correlation between the PAH and CO levels. Several authors [
44,
45,
46] used molecular diagnostic ratios (MDRs) as an indicator of the distribution of PAH sources in the environment, especially in the air. For example, the ratio of ANT/(ANT+PHE) has been suggested as an indicator of petrogenics for pyrogenic sources. This ratio has been criticized in the past [
46,
47], because ANT is more reactive than PHE, and their environmental behavior is therefore very different. The BaA/(BaA+CHR) ratio allows one to discriminate between the same sources as those in the ANT/(ANT+PHE) ratio, but it is more representative. Kim et al. [
48] pointed out that a value below 0.20 suggests petrogenic emissions, while a value >0.35 indicates combustion (pyrogenic emissions). The FLT/(FLT+PYR) ratio distinguishes between petro- and pyrogenic sources (<0.40 and >0.40, respectively), but it can also indicate whether the pyrogenic emissions result from fuel combustion (0.4–0.5) or from the combustion of other materials (>0.50). The IP/(IP+BgP) ratio is to be interpreted in a similar way to FLT/(FLT+PYR). Several authors [
49,
50,
51] showed that an IP/(IP+BgP) ratio >0.5 indicates that the PAHs probably originated in the emissions from coal combustion. In
Table 7, the diagnostic ratios of the PAHs, used as the source indicator of fuel combustion, as found in previous studies, are presented.
On the basis of the PAH concentrations in flue gases, MDR indices for the tested fuels were calculated. The relationship between ANT/(ANT+PHE), FLT/(FLT+PYR), BaA/(BaA+CHR), IP/(IP+BgP), BbF/(BbF +BkF), and BaP/(BaP+BgP) are presented in
Table 8.
The obtained results indicate that the ANT/(ANT+PHE) ratio is very different, depending on the fuel type. For pelletized coniferous wood and pelletized miscanthusm, the ratio is about 0.22 and is higher than that for pelletized cereal straw and coal, which are 0.07 and 0.1, respectively. The FLT/(FLT+PYR) ratio ranges from 0.49 (pelletized coniferous wood) to 0.6 (coal) and is similar to the results presented by Shen et al. [
53] and Cheruiyot et al. [
54]. A higher BaA/(BaA+CHR) ratio was observed for coal in comparison to that for biofuels. This ratio was 0.5 for coal and 0.41, 0.36, and 0.47 in the case pelletized coniferous wood, pelletized cereal straw, and pelletized miscanthus, respectively. The BbF/(BbF+BkF) ratio for pelletized coniferous wood and pelletized miscanthus was at the same level (0.47), while for hard coal, it was 0.58. In the case of BaP/(BaP+CHR) ratios, a similar level of 0.4–0.45 was achieved for coal, pelletized coniferous wood, and cereal straw, while the ratio was lower than for pelletized miscanthus. For the IP/(IP+BgP) ratio, a similar correlation was observed. This ratio was at a level of 0.5–0.53 and 0.57 for coal, pelletized coniferous wood or cereal straw and pelletized miscanthus, respectively. The determined MDR are similar to the results presented in other publications despite different combustion conditions and physicochemical properties of biofuels. This similarity confirms their usefulness for identification of combustion sources.