*3.2. Thermal Analysis*

The TG curves that were obtained from the temperature programmed combustion of MB, BC, and their blend MB-BC at di fferent β (0.1, 0.2, 0.4, and 0.5 K/s) are depicted in Figure S1. The resultant DTG curves are shown in Figure 2 together with the weighted calculated curves (Equation (1)) corresponding to the combustion of MB-BC. As for a typical combustion profile, mass loss occurs along with increasing temperature under oxidizing atmosphere until the volatiles and fuel content of the sample is exhausted, and then the mass of ashes remains stable. In the case of BC, the loss of volatiles and char gasification occur in a single step due to the large contribution of fixed carbon (see Table 2) to the mass loss during combustion. On the other hand, a slight weight gain that is related to oxygen chemisorption may be observed during the combustion of BC but is not present in the DTG curves corresponding to MB. Di fferently from BC, the DTG curve that corresponds to MB shows that mass loss occurs in two main stages, which has already observed by other authors for the combustion of microalgae [16,56,57]. The first stage, which is the most remarkable, is attributed to the devolatilization of MB and volatiles combustion and extends until 670 K. It is to highlight that this first stage occurs in a temperature range for which no mass loss is observed for BC. This is due to the higher volatiles/fixed-carbon ratio of MB (5.01), as compared to that of BC (0.13), which involves microalgae combustion predominantly occurring in gas-phase due to the combustion of volatiles. Above 670 K and ending at around 1000 K, occurs the second stage, which comprises three subsequent steps at β = 0.1 and 0.2 K/s that overlap at higher β. Despite evident di fferences between MB and BC, DTG experimental curves corresponding to the combustion of MB-BC mostly resemble those of BC, except for the mass gain associated to oxygen chemisorption, which is roughly appreciable. This is also true for the weighted calculated curves MB-BC (T), since they are nearly coincident with the experimental curves. This fact indicates that interaction between MB and BC during combustion is not relevant, which is favourable in terms of the practical application of co-processing. A good correlation between experimental and weight calculated DTG curves was also observed by Gao et al. [56] for the co-combustion of a lignite coal and microalgae biomass from *Chlorella vulgaris* (blending ratio 50%). These authors stated that the synergetic e ffects in the co-combustion of these materials were negligible [56]. Di fferently, and after having observed inhibitive e ffects during co-pyrolysis of a sub-bituminous coal that was blended with *Chlorella vulgaris* biomass (blending ratios 30, 50 and 70%) [73], Chen et al. [74] verified that, although no synergetic effects occurred in the initial and final stages of their co-combustion, some interaction occurred at the intermediate stage. Interactions were especially evident for the highest ratio of microalgae within the blend (70%) and at relative high temperatures, which has been also observed for the co-combustion of coal with composite biomass pellets that were made from catkins, wood waste, and rice straw [75]. Interactions between coal and biomass during co-combustion reported in the literature have been attributed to the acceleration of coal devolatilization during co-combustion, due to the reaction with the active radicals that were produced during biomass devolatilization [56]. Such acceleration may be ascribed to the higher temperature of coal particle surface during its co-combustion with biomass than during its individual combustion.

**Figure 2.** Differential thermogravimetry (DTG) curves corresponding to the temperature programmed combustion of MB, BC and their blend (MB-BC) together with the weighted calculated curve corresponding to the blend (MB-BC (T)) at the different heating rates here used, namely 0.1 K/s (**a**), 0.2 K/s (**b**), 0.4 K/s (**c**), and 0.5 K/s (**d**).

Figure 2 evidences that DTG curves that correspond to MB and BC were a ffected by increasing the heating rate (β). In this sense, it is observable an increase of the temperature at which mass loss begins; a broader range of temperature in which mass loss occurs; higher weight loss (%/s) peaks; and overlapping of sub-steps in the combustion of MB. Still, the DTG curves of the blends remain analogous to those of BC, even at increasing heating rates (β).

Table 5 depicts the characteristic parameters of the DTG combustion profiles. These parameters confirm the di fferences between MB and BC that were observed in Figure 2 and similitudes between BC and MB-BC. In this sense, the *Tv* and *Tf* in Table 5 correspond to the temperatures at which mass loss begins and ends, respectively. Meanwhile, for each DTG curve, the *Tm* corresponds to the temperature at which the maximum mass loss rate occurs, that is, the *DTGmax*. For all MB, BC and MB-BC, the temperatures *Tv*, *Tf* and *Tm* increase with β. For MB, the *Tv* and *Tm* values are around 240 K lower than for BC. Di fferently, *Tv* values for MB-BC are just slightly lower than for BC, due to the absence of chemisorption mass loss in the combustion of MB-BC, while *Tm* values are mostly coincident. On the other hand, lower *Tf* values are observed for BC than for MB-BC, which shows slightly lower values than MB. Finally, at β = 0.1 and 0.2 K/s, the *DTGmax* corresponding to MB are the lowest, while at β = 0.4 and 0.5 K/s they are the highest, which evidences the intensification of devolatilization of MB with β. Meanwhile, at each β, BC and MB-BC show very close *DTGmax*.

*B* **(K**/**s)** *Tv* **(K)** *Tm* **(K)** *Tf* **(K)** *DTGmax* **(%**/**s)** MB 0.1 400 542 1031 0.0363 0.2 410 543 1040 0.0839 0.4 433 554 1100 0.1981 0.5 440 557 1120 0.2457 BC 0.1 640 782 874 0.0890 0.2 657 820 930 0.1295 0.4 671 867 1035 0.1669 0.5 675 875 1050 0.1820 MB-BC 0.1 600 781 981 0.0873 0.2 631 823 993 0.1281 0.4 646 870 1046 0.1645 0.565488510800.1844

**Table 5.** Characteristic parameters of DTG combustion curves determined for microalgae biomass (MB), bituminous coal (BC) their blend (MB-CB).

*Tv*: onset temperature for volatile release and mass loss; *Tm*: temperature of maximum mass loss rate; *Tf*: final combustion temperature detected as mass stabilization; *DTGmax*: maximum mass loss rate.

Figure 3 represents the DSC curves that were obtained from the temperature programmed combustion of MB, BC, and their blend MB-BC at di fferent β (0.1, 0.2, 0.4, and 0.5 K/s), together with the weighted calculated curves (Equation (2)) corresponding to the combustion of the blend. As it may be seen, heat release during BC combustion occurs in two stages, namely during the chemisorption mass gain and, especially, during the fixed-carbon combustion that was observed in Figure 2.

**Figure 3.** Differential scanning calorimetric (DSC) curves corresponding to the temperature programmed combustion of MB, BC and their blend (MB-BC) together with the weighted calculated curve corresponding to the blend (MB-BC (T)) at the different heating rates here used, namely 0.1 K/s (**a**), 0.2 K/s (**b**), 0.4 K/s (**c**) and 0.5 K/s (**d**).

Heat liberation during MB combustion takes place at four subsequent steps, corresponding to the combustion of volatiles, organic material, fixed-carbon, and char, which are evident at β = 0.1 K/s and progressively overlap at increasing β. Consequently, at β = 0.5 K/s, heat release during MB combustion occurs just in two main stages. The second one, which corresponds to the combustion of the fixed-carbon content, is centred at the same temperature than the second stage of BC (Figure 3). Although DSC analysis helps to obtain a more realistic approach to the combustion process of biomass, there are few published DSC results regarding the combustion of microalgae biomass to compare with the here obtained. Yet, <sup>L</sup>ópez-González et al. [16] also observed two heat release stages during the temperature programmed combustion of biomass from three di fferent microalgae strains at β = 0.67 K/s. Regarding the combustion of MB-BC in this work, it may be observed in Figure 3 that, even when DSC curves corresponding to MB and BC are very di fferent, heat release during the combustion of MB-BC shows the same trend than BC.

The characteristic temperatures at DSC curves in Figure 3 are depicted in Table 6, together with the enthalpies of combustion, which were calculated for each MB, BC, and MB-BC by the integration of the corresponding exothermic peak. The onset temperatures for heat release ( *Ti*) are slightly lower for MB than for BC, with the latter being close to those of MB-BC. Contrarily, the final combustion temperatures ( *Te*) are higher for MB than for BC, which are again very close to those that were observed for MB-BC. Regarding temperatures of maximum energy release ( *Tmax*), the lowest value is observed for MB combustion at β = 0.1 K/s, since the volatiles combustion was the main peak of heat release. However, at increasing β, the third stage (corresponding to the fixed-carbon combustion) progressively gains more prominence, so higher *Tmax* values occur. At each β, *Tmax* values that were observed for BC and MB-BC are equivalent and correspond to the fixed-carbon combustion peak. Finally, lower enthalpy ( Δ *H*) values are obtained for MB combustion than for BC and MB-BC. In any case, the Δ *H* here obtained for MB are slightly higher than those that were determined by <sup>L</sup>ópez-González et al. [16], which were within 7.8 and 8.8 kJ/g.


**Table 6.** Characteristic parameters of DSC combustion curves determined for microalgae biomass (MB), bituminous coal (BC) their blend (MB-CB).

*Ti*: onset temperature for energy release and peak integration; *Tmax*: temperature of maximum energy release during combustion; *Te*: temperature at the end of energy release during combustion; Δ*H*: enthalpy determined by integration of the heat release peak in the corresponding DSC combustion profile.

#### *3.3. Non-Isothermal Kinetic Analysis*

The TG curves that correspond to the temperature programmed combustions of MB, BC, and MB-BC at the di fferent β here considered are represented in Figure 4. For each case, six di fferent percentages of conversion are pointed out in each curve: 10, 20, 30, 40, 50, and 60%. As it was previously observed in DTG curves (Figure 2), mass loss at relatively low temperatures was more remarkable for MB than for BC due to the high volatiles content of the first (Table 2). On the other hand, the ash yield, or residual mass after burning, was 6, 31, and 28% for MB, BC, and MB-BC. These yields are in agreemen<sup>t</sup> with ash contents in Table 2, and, in the case of MB-BC with the weighted calculated value for the blend.

**Figure 4.** Thermogravimetry (TG) curves corresponding to the combustion of MB (**a**), BC (**b**), and MB-BC (**c**) at different heating rates. Conversion percentages (10, 20, 30, 40, 50, and 60%) have been marked by straight lines crossing experimental data.

Figure 5 shows the plots of ln (β/*T*) and ln (β/*T*2) vs. 1/*T* to the several conversion degrees (α), corresponding to the combustion of MB, BC, and MB-BC, as for the respective application of the FWO model [50,51] and the KAS model [53,54]. Meanwhile, Table 7 depicts the corresponding estimations of *E*.

The *E* values estimated by the FWO model are slightly higher than those that were estimated by the KAS one, except for MB, with both models giving nearly coincident values. As it may be seen in Table 7, the average *E* values (*E\**) that were determined for the combustion of MB are more than double of that corresponding to BC. On the other hand, the *E\** values corresponding to the combustion of MB-BC are just slightly higher than what would proportionally correspond for the relative MB and BC contents (100 and 91 kJ/mol, for the FWO and KAS models, respectively).


**Table 7.** Values of apparent activation energy (*E*, kJ/mol) determined for the combustion of MB, BC and their blend (MB-BC) at the considered conversion degrees (α), together with *R*<sup>2</sup> corresponding to fittings in Figure 5.

*E\** ± standard deviation from *E* values at the considered α.

In Table 7, the change of *E* with α, which is especially evident for MB, may be related to the complexity of the fuel composition and subsequently to the complex reactions ongoing under combustion, as it has already been observed for woody biomass [76]. The *E\** values here determined for BC are close to those previously obtained for a coal of the same rank and origin [23], while those of MB are in consonance with published results on the combustion of microalgae biomasses. Using the FWA and the KAS methods, Chen et al. [70] determined *E\** values between 134 and 242 kJ/mol for the combustion of *C. vulgaris* under 20%O2/80%N2 and 80%O2/20%N2, respectively. Applying a modification of the FWO method, namely the Starink's model, Zhao et al. [77] determined *E\** values between 93 and 142 kJ/mol for the combustion of different microalgae strains, being 118 kJ/mol the *E\** corresponding to *C. sorokiniana* C74. For the oxy-fuel co-combustion of *C. vulgaris* biomass with lignite and using the FWO and the KAS methods, Gao et al. [56] determined *E\** values for the blends that were between 150 kJ/mol (lignite/microalgae = 7:3) and 197 kJ/mol (lignite/microalgae = 3:7) under air atmosphere (21%O2/79%N2). These authors [56] highlighted that the lowest and highest *E\** were obtained for lignite coal and *C. vulgaris* microalgae (respectively, 146 and 213 kJ/mol), and that a higher percentage of microalgae in the blend resulted in a higher *E\**. On the other hand, Chen et al. [74] determined *E*\* values of 68 and 107 kJ/mol for the combustion of a sub-bituminous coal and *C. vulgaris* biomass, respectively, by the FWO method and of 57 and 103 kJ/mol by the KAS one. Subsequently, these authors [74] found that with the increasing content of *C. vulgaris* in the blends, the *E\** first decreased and then increased, with the highest value (FWO: 115 kJ/mol) being that of the blend having the largest microalgae content (coal:microalgae = 3:7).

From an economic and environmental point of view, the possibility of a joint combustion of microalgae biomass and coal in power plants may be interesting, since it allows for the use of existing infrastructures, already equipped with appropriate systems for emissions control and staffed with qualified personnel. In the case of residual microalgae biomass from wastewater treatment, combustion might be the preferred and safest managemen<sup>t</sup> alternative for its valorization. Other uses, such as

fertilizer or animal feed, should only be allowed under strict toxicity controls in order to avoid the incorporation of pollutants, such as trace metals or pharmaceuticals into the soils, ground water, and food chain. Meanwhile, if properly designed and operated, the combustion of microalgae biomass can provide high stability of trace metals in the ashes and complete destruction of organic pollutants.

Overall, the results obtained in this work indicated that, under appropriate conditions, co-combustion of microalgae biomass with coal might be an option to consider. In fact, through the solution of energy and mass balances, it has been recently demonstrated that the smart integration of microalgae culturing with a large scale coal power plant is feasible in terms of net energy ratio (NER) and CO2 emissions [78]. Still, further work regarding the combustion and co-combustion of microalgae biomass in different types of boilers is actually necessary in view of their practical implementation. In a circular economy context, integrating microalgae culture-wastewater treatment and the thermal valorization of residual biomass is a challenge that would allow for closing the loop for a sustainable microalgae culture.
