The experiments using TGA, as mentioned before, were performed with different heating and air flow rates to study their influence on the thermal decomposition as a function of different particle sizes. The next sections will present the results.
3.1.1. Influence of the Heating Rate
One of the significant factors in the thermal decomposition of pine wood particles is the heating rate. In this way, to analyze its effect
Figure 3a–c presents the variation in mass loss during combustion in a thermogravimetric (TG) curve (dashed lines) and shows the heat flow (continuous lines) during the experiment for the three pine wood particle sizes examined. The first derivative of the TG curve—called a derivative thermogravimetric (DTG) curve—showing the rate of change in mass is also presented in
Figure 3d–f. Based on the TG, DTG, and heat flow curves, the characteristic temperatures and characteristic parameters of the combustion process for each particle size under all heating rates were determined. All these parameters are presented in
Table 2 and
Table 3. These data include the most relevant variables that identify the various transitions occurring during the couplet combustion of biomass.
The dashed lines in
Figure 3a–c allow the identification of three different stages of biomass combustion: moisture evaporation (drying), devolatilization, and char combustion. It is possible to observe that the heating rate affects the behavior of the weight loss curves. The figures also include a horizontal dashed line at the moisture content of the samples (≈7%). As can be observed in
Figure 3d–f, the thermal decomposition for all the particle sizes examined starts at approximately 260 °C for lower heating rates and 211 °C for higher heating rates. Therefore, as the heating rate increases, the initial decomposition temperature decreases. This is because, at low heating rates (<10 °C/min), the heat release took place at a slower rate and allowed enough time for water to be removed completely from the sample.
However, at higher heating rates (>20 °C/min), the initial decomposition starts at a lower temperature, as increasing the heating there is not enough time for the moisture to be fully removed from the sample and, therefore, the drying phase is controlled by the water diffusion inside the biomass structure. This was indicated by the mass loss of the sample that is below 7% after the initial decomposition.
A major mass loss follows, where the main devolatilization occurs with a maximum combustion rate between 319 to 365 °C for particles smaller than 0.063 mm—321 to 401 °C for particles between 0.125 and 0.25 mm and from 321 to 369 °C for particles larger than 1 mm. In this way, as the particle size and heating rate increase, the maximum combustion rate increases, although the effect of the heating rate is stronger. This means that the residence time of the sample in the furnace decreases with an increasing heating rate.
Furthermore, during the thermal decomposition, as can be observed in the heat flow curves, there are two different exothermic reaction regions. The first region is associated with the combustion of the light volatile matter which provides the reactivity of biomass fuels. The second region is associated with the char combustion [
33]. Consequently, heat, in general, is released between approximately 300 °C and 500 °C, reaching the highest value close to 450 °C for all pine wood particles analyzed at lower heating rates (5, 10, and 20 °C/min). However, for higher heating rates, as in the case of 243 °C/min, the temperature range is wider and the highest value is close to 680 °C. Therefore, according to the DTG curves, as presented in
Figure 3d–f, when the heating rate was increased, the thermal decomposition peaks were shifted to a higher magnitude. According to Kok and Özgür [
24], the reason for these shifts is due to the different heat transfer and kinetic rates delaying the sample decomposition. Additionally, as the thickness of the gas cushion around the particle and the intensity of degassing increase, a temporary slow-down in the intensity of convective heat transfer to the interior of the particle occurs [
53]. Another effect is that heat transfer is not as effective and efficient at slower heating rates and, therefore, the minimum heat required for particles’ cracking is reached later at higher temperatures [
21]. Conversely, at lower heating rates, the heating of pine wood samples occurs more slowly, leading to a better heat transfer to the particles and, consequently, a more effective fuel cracking. The heat flow was higher at higher heating rates and the temperatures at which the maximum heat flow occurred corresponded to the char combustion stage where the volatiles were almost completely oxidized.
The peak temperature in the DTG curves is a measure of the reactivity of the char. As the peak temperature decreases, the easier the ignition of the pine wood particles will be. The combustion reactivity regions are proportional to the height of the DTG peak and, therefore, at higher heating rates a more reactive combustion takes place. The reactivity is due to the combustion of the volatiles and the energy released is mainly due to the combustion of the fixed carbon, as can be concluded from the third peak of the heat flow curves for all the pine wood particles examined.
However, when it is applied higher heating rates on large particles, as can be observed in
Figure 3b,c, only one peak is visible. In such conditions, the heat released was very high, the material reached high temperatures in less time, and the thermal decomposition started earlier than at lower heating rates (see
Table 2). For instance, at 243 °C/min, the time to complete the combustion was only about 4 minutes. It is also observed that, as the heating rate increases, there is a shift to higher ignition temperatures. This increase in temperature is independent of the particle size, which suggests that the temperature gradients inside the sample push the initiation of the decomposition to higher temperatures [
54]. This is a result of the decrease in the heat transfer efficiency in the sample [
20,
54]. Furthermore, the combustion rate and temperatures of the second and third peak in the DTG curves increase significantly with the heating rate. Consequently, the maximum combustion rate increases as the heating rate increases and, therefore, the heat released during the experiments also increases. It is important to note that similar results were obtained for all particle sizes.
Also, increasing the heating rate increases the residue at the end of the experiments, as revealed by Mani et al. [
21]. However, as presented in
Figure 4, this feature was not observed in the present work, which shows that there is no consistent trend as far as the heating rate is concerned.
This result may be influenced by the uncertainty either from the instruments or the random selection of the sample from the batch of pine wood samples. To verify the randomness hypothesis, three tests with the same conditions at 10 °C/min, 100 mL/min, and the middle particle size range (0.125 and 0.25 mm) were conducted.
Figure 5 presents the results of the three different experiments and the different values of the remaining mass that were obtained. This observation approves the previous hypothesis as the effect of the randomness in the sample.
Following the procedure previously described, the ignition (
D) and combustion (
S) indexes are presented in
Figure 6 for different heating rates. The results show that the
D index is correlated with the
Tig. Higher ignition indexes were observed when a better ignition performance was obtained. The
S value also follows a similar trend, where the value increases with the increasing heating rate. Furthermore, as reported by Vamvuka et al. [
20], there was no significant effect of particle size on the ignition and combustion indexes.
3.1.2. Influence of the Air Flow Rate
Figure 7 presents the TG, DTG, and heat flow curves for the three pine wood particle sizes at four different air flow rates: 10, 50, 100, and 150 mL/min. Based on the TG and DTG curves, the characteristic temperatures and characteristic parameters of the combustion process for each particle size under all air flow rates were determined. All these parameters are presented in
Table 4 and
Table 5.
The dashed lines in
Figure 7a present the mass loss for particles smaller than 0.063 mm. It is possible to see that the mass loss at different air flow rates has similar behavior. However, after the devolatilization stage, there are differences between the mass loss behavior. In this way, at an air flow rate of 50 and 100 mL/min, the mass loss occurred earlier when compared with other air flow rates. These differences can be seen in the DTG profile in
Figure 7d–f in the second peak for the 100 mL/min air flow rate. Therefore, at 100 mL/min during the char combustion, the mass loss increased and, consequently, in the heat flow curve (continuous lines of
Figure 7) it is possible to see that more heat was released at 100 mL/min during the devolatilization and char combustion stages.
For particles between 0.125 and 0.25 mm, similar results were obtained (
Figure 7b,e). However, during char combustion, the mass loss is independent of the air flow rate and the heat flow decreases significantly when compared with the results presented previously for 100 mL/min. Furthermore, as can be seen in
Figure 7b, at 50 mL/min the heat flow was higher, which means that increasing the particle size caused the air flow rate necessary to release more energy to be lower.
Regarding the experimental results with particles larger than 1 mm, in
Figure 7c,e at 10 and 100 mL/min, the mass loss during the drying process occurred earlier and during devolatilization and char combustion the mass loss was slower, at 150 mL/min.
Figure 7f, at the second peak, presents a similar trend, as the results were obtained with particles smaller than 0.063 mm. However, the second peak occurred at 100 mL/min and also at 10 mL/min and 150 mL/min. Consequently, the heat flow curve (
Figure 7c) displays a higher value at the second peak for 10, 100, and 150 mL/min.
It should be noted that although the maximum heat flow peak is higher in the case of particles smaller than 0.063 mm at 100 mL/min, more heat was released in the experiments with particles greater than 1 mm for various airflows rate in this study, except at 50 mL/min. From these results, it can be said that there is a negligible influence of the airflow rate on the thermal decomposition of the biomass particles. The main devolatilization occurs at a maximum combustion rate between 328 to 332 °C for particles smaller than 0.063 mm, 332 to 334 °C for particles between 0.125 and 0.25 mm, and 330 to 335 °C for particles larger than 1 mm. In this value, as the particle size and airflow rates increase, the maximum combustion rate increases slightly. This may result from the surface area of the particle, where an increase in the surface area of the particles minimizes the mass and heat transfer limitations and improves the conversion efficiency [
55].
However, as presents in
Table 5, the heat released was in general higher when the air flow rate of 100 mL/min was applied to the different particle sizes. Consequently, with this air flow rate the oxidation took place close to the stoichiometric conditions and, with the other values, the combustion happened in an excess of comburent in the case of 150 mL/min and with a lack of stoichiometric comburent in the remaining cases.
Regarding the ignition and combustion indexes,
Figure 8 presents their variation with different air flow rates used in the experiments with distinct particle sizes. The results show that there is no consistent trend regarding the effect of increasing and/or decreasing the air flow rate on the indexes. However, an interesting trend was observed when decreasing the particle size, as higher values of the combustion indexes were obtained for high air flow rates. For instance, for particles smaller than 0.063 mm the ignition and combustion indexes were higher for 100 mL/min, while for particles larger than 1 mm the highest indexes were observed for 10 mL/min.