**3. Results and Discussion**

The kinetics of higher heating value combustion of untreated and torrefied barley straw have been extensively studied using ISO 1716:2018 [35]. To this end, the well-known higher heating value of combustion equation is shown below:

$$H\_{\mathcal{K}} = (t\mathcal{W} - e\_1 - e\_2 - e\_3)/m \tag{1}$$

where *Hg* represents the higher heating value of combustion, "m" stands for the mass of the sample in grams, *e*1 refers to a correction coe fficient concerning calories for heat of formation of nitric acid, *e*2 to a correction coe fficient concerning calories for heat of formation of sulfuric acid and *e*3 to a correction coe fficient for calories or heat of combustion of fuse wire. For the given case, both *e*1 and *e*2 are taken as being equal to zero since neither nitric acid nor sulfuric acid were used. Moreover, *W* is the energy equivalent of the calorimeter, which is determined under standardization and *t* is the net-corrected temperature increase, with equations following further analyzing the above variables.

$$t = t\_c - t\_a - r\_1(b - a) - r\_2(c - b) \tag{2}$$

$$c\_3 = 2.3l\_f \tag{3}$$

$$\mathcal{W} = 10,104 \text{ J/} ^\circ \text{C} \tag{4}$$

To this end, *a* stands for the time of firing, *b* for the time when the temperature reaches 60% of the total rise and *c* for the time at the beginning of period in which the rate of temperature change is constant. Next, *ta* corresponds to the temperature at firing time and *tc* the temperature at time *c*, *r*1 is the rate at which the temperature was rising until firing and *r*2 is the rate at which the temperature is rising during the 5 min period after the time *c*. Finally, *lf* is the size of the fuse wire consumed during the firing.

A severity factor was used in order to integrate the e ffects of reaction times and temperature into a single variable during torrefaction. In this context, a 'combined severity factor' for isothermal reactions was based on the 'P' factor, first introduced at 1965 by Brasch and Free [36], for the prehydrolysis-Kraft pulping of *pinus radiata*, and then at 1987 (under the name 'reaction ordinate') applied by Overend and Chornet [37] in the case of fractionation of lignocellulosics by steam-aqueous pretreatments (like wet torrefaction). The 'P' factor had units of time and was as follows:

$$(\text{'P' factor}) = \left[ \exp(T - 100) / 14.75 \right] \cdot t \tag{5}$$

where *t* is the reaction time in min and *T* is the reaction temperature in degrees Celsius.

Moreover, in the case of torrefaction for high energy density solid fuel of fast-growing tree species, the following severity factor was used [38]:

$$\mathbb{E}SF = \log\left[t \cdot e^{\frac{T\_h - T\_R}{14.75}}\right] \tag{6}$$

where *t* is the reaction time of the torrefaction in min, *Th* the reaction temperature and *TR* the reference temperature, both in degrees Celsius.

In addition to the above, a combined severity factor for non-isothermal reaction conditions was also introduced in the case of the batch autohydrolysis of wheat straw [39],

$$R\_0^\* = 10^{-pH\_\cdot} \int\_0^t e^{\frac{T\_0 - 100}{14.75}} dt\tag{7}$$

where *T*θ is the reaction temperature in degrees Celsius.

At this point, it should be noted that since, in this work, the main variables used are time and temperature, pH was removed from the equation, with the simplified severity factor used for non-isothermal reaction conditions given in the following equation:

$$R\_0 = \int\_0^t e^{\frac{T\_0 - 100}{14.75}} dt\tag{8}$$

A similar severity factor was used by Aguado et al. [40] for wet torrefaction of almond-tree pruning. On the other hand, a severity index was used by Zhang et al. [41] for spend coffee grounds and microalga residue torrefaction. Several torrefaction severity reporting methods were reported by Campbell et al. [42], while the dry mass yield was suggested as an indicator for severity presuming that was the most reliable singular severity indicator for bench and pilot scale work.

Consequently, in the present work, the severity factor values according to Equation. (8) and for each of the experiments carried out are provided in Table 1. Therefore, the gradual reduction of the test sample mass from starting time (*m*0) until the end of each experiment (*m*t) is used, with the parameter of solid residue yield showing the percentage of the mass loss over torrefaction time.

In this context, the diagram of Figure 3 depicts the temperature profile and how the latter is affected during the time stages described earlier. An example of temperature profiles at different times of the muffle furnace during torrefaction of 300 ◦C is shown in this Figure. The preheating time in the case of 300 ◦C was around 25 min, because the initial temperature in the furnace was about 30 ◦C, i.e., the furnace was cold (not preheated). Similar temperature measurements were done at each torrefaction time for the muffle furnace.

**Figure 3.** Experimental temperature profile for a typical barley straw torrefaction experiment.

Moreover, in Figure 4 the percentage of loss of mass during the torrefaction procedure following illustrates the impact of the time (Figure 4a), severity factor (Figure 4b), and logarithm of severity factor (Figure 4c) on the solid residue yield percentage decrease, with the latter showing a rapid reduction for small severity factor values which is gradually almost stabilized for higher severity factor values. Increased weight loss occurs when torrefaction temperature is also increased due to moisture removal and hemicellulose breakdown which produced H2O, CO, CO2 and other hydrocarbons. Finally, the following equations describe the exponential relation between the yield (*y*) and the time (*t*) or the severity factor (*R*0 or log*R*0), with the equation parameters given in Table 3.

$$\text{Model A1: } y = y\_e + (y\_0 - y\_c) \exp(-kt) \tag{9}$$

where *ye* is the value for *y* at infinite time, *y*0 is the value for *y* at zero time, and *k* is the pseudo-first order kinetic constant.

$$\text{Model A2: } y = y\_c + (y\_0 - y\_e) \exp(-kR\_0) \tag{10}$$

**Table 3.** The parameters and standard error of estimate (SEE) of the three models for the solid residue yield (% w/w) of barley straw torrefaction.


**Figure 4.** Torrefied barley straw solid residue yield vs. time (**a**), severity factor (**b**), and logarithm of severity factor (**c**).

It must be mentioned Model A3 is described by the same Equation (10) as Model A2, but its parameters were estimated without taking into account the experimental value for *y* at zero time. The standard error of estimate (SEE) values for these tree models are presented in Table 3, showing that the best fitting to the experimental data was for Model A3. The fitting of these three models is illustrated in Figure 4a,b and c for Model A1, A2 and A3, respectively.

Moreover, Figure 5 demonstrates the Higher Heating Value (*Hg*) of barley straw combustion vs. torrefying reaction time (Figure 5a), severity factor (Figure 5b), and logarithm of severity factor (Figure 5c). To this end, according to the experimental results obtained, the optimal time that gives the maximum output (*Hg* = 21.3 MJ.kg) was 47.5 min, where *Hg* increases by 21.7%. On the other hand, the gross heat of combustion for the untreated barley straw was measured a total of three times, with the average value found to be 17.5 MJ/kg and the standard deviation 0.17 (1.0%). Therefore there is an increase of *Hg* during conditions intensification. After all, the following equations describe the relation between the *Hg* and the time (*t*) or the severity factor (*R*0 or log*R*0) with the equation parameters given in Table 4.

$$\text{Model B1: } H\_{\mathcal{R}} = H\_{\mathcal{R}^{\text{ct}}} - \left[ (H\_{\mathcal{R}^{\text{ct}}} - H\_{\mathcal{R}^{\text{0}}})^{-1} + k\_1 t \right]^{-1} \tag{11}$$

where *Hge* is the value for *Hg* at infinite time, *Hg*0 is the value for *Hg* at zero time, and *k*1 is the pseudo-second order kinetic constant.

$$\text{Model B2: } H\_{\mathcal{R}} = H\_{\mathcal{R}^c} - \left[ (H\_{\mathcal{R}^c} - H\_{\mathcal{R}})^{-1} + k\_1 R\_0 \right]^{-1} \tag{12}$$

**Figure 5.** Torrefied barley straw Higher Heating Value of combustion vs. time (**a**), severity factor (**b**), and logarithm of the severity factor (**c**).


**Table 4.** The parameters and standard error of estimate (SEE) of the three models for the Higher Heating Value (MJ/kg) of the combustion of the torrefied barley straw.

It must be mentioned the Model B3 is described by the same Equation (12) as Model B2, but its parameters were estimated without taking into account the experimental value for *Hg* at zero time. The SEE values for these tree models are presented in Table 4, showing that the best fitting to the experimental data was for Model B2. The fitting of these three models is illustrated in Figure 5a–c for Models B1, B2 and B3, respectively.

Figure 6 illustrates the relation between the Higher Heating Value of barley straw combustion and the material's mass loss percentage due torrefaction. The theoretical curve was estimated using Models A1 and B1 in combination. Moreover, Models A2 and B2 could successfully fit the experimental data. The maximum Higher Heating Value of the barley straw combustion is expected to be at the maximum material's mass loss percentage, i.e., at the most severe torrefaction conditions. Moderate torrefaction conditions could be chosen to reduce barley straw's mass loss but with a lower Higher Heating Value of the material combustion.

**Figure 6.** Torrefied barley straw Higher Heating Value of combustion vs. the mass loss percentage.

In Figure 7, the torrefied barley straw Enhancement Factor (EF) and Energy yield (EY) vs. the mass loss percentage are presented. The Enhancement Factor (EF) is given by

$$(EF) = H\_{\mathcal{R}} \mathcal{H}\_{\mathcal{R}^{\rm II}} \tag{13}$$

where *Hgt* is the HHV for torrefied straw and *Hgu* is the HHV for untreated straw. The Energy yield (EY) is given by the following equation:

$$
\Psi(EY) = (EF) \cdot y
\tag{14}
$$

**Figure 7.** Torrefied barley straw (**a**) Enhancement Factor (EF) and (**b**) Energy yield (EY) vs. the mass loss percentage.

According to Figure 7a the Enhancement factor increases by mass loss decreasing, while according Figure 7b, Energy yield decreases almost linearly by mass loss decreasing. The theoretical curves are according to the same above-described Models A1 and A2, and Equations (10) and (12), respectively. There was no need for re-estimation of the models' parameters.

In Figure 8 are shown the Scanning Electron Microscopy (SEM) images of the untreated barley straw at (a) × 750, (c) × 7500 and (e) × 20,000 magnification, and torrefied barley straw (at optimal conditions) at (b) × 750, (d) × 7500 and (f) × 20,000 magnification. We observe that the effect of the torrefaction on the straw surface topology is the roughening of the surface. The effect might facilitate

the use of torrefied barley straw for the production of adsorbents (low-cost activated carbon substitute). This could be an alternative use to the torrefied straw as energy production material (coal substitute).

**Figure 8.** Scanning Electron Microscopy (SEM) images of untreated barley straw at (**a**) ×750, (**c**) ×7500 and (**e**) ×20,000 magnification, and torrefied barley straw at (**b**) ×750, (**d**) ×7500 and (**f**) ×20,000 magnification.

In Table 5, the Higher Heating Value of combustion, the Solid residue yield, the Enhancement factor and the Energy yield for some untreated and torrefied lignocellulosic residues according to the recent literature are presented.


**Table 5.** Higher Heating Value of combustion, Solid residue yield, Enhancement factor and Energy yield for some untreated and torrefied lignocellulosic residues.

These HHV values are comparable to the values found in the present study with regards to untreated and torrefied barley straw [34], but there are significant differences when another lignocellulosic material was used. Moreover, the EF value of torrefied barley straw [11] was similar to the findings of the present work, while the EY value [11] was lower compared to that of the present work. On the other hand, most of the other lignocellulosic materials presented in Table 5 have higher EY values (73.7–98.9%) compared to the 60.7% found herein. The high EY values were found due to high EY and/or high solid residue yield.

The higher heating values (HHV) of the barley straw samples in the present work can be calculated from their C, H and N contents (see Table 2) in a dry basis, using the following expression, as derived by Friedl et al. [45] for biomass from plant origin:

$$\text{HHV} = 3.55 \text{C}^2 - 232 \text{C} - 230 \text{H} + 51.2 \text{C} \cdot \text{H} + 131 \text{N} + 2060 \tag{15}$$

The values calculated according to Equation (15) for untreated and torrefied barley straw of this work were 18.1 and 22.4 MJ/kg, respectively. This is very close to the experimental values shown in Table 5. The EF was 1.24 very close to 1.22, i.e., the experimental one.

Lignocellulosic biomass torrefaction (dry or wet, in the absence of oxygen or not, under atmospheric pressure or not) is a pretreatment process used to overcome the disadvantages of using biomass as a fuel such as low energy density, high moisture, and oxygen contents [46,47]. The torrefaction increases energy density, hydrophobicity, and reduces grinding energy requirement of biomass. The environmental and economic aspects of the torrefaction process and torrefied product, and various applications of torrefaction products have been taken into account by various researchers. The cost competitiveness of torrefied materials is one of the major concerns of the torrefaction process. Integrating the torrefaction with other processes makes it economically more viable than as a standalone process [47].
