Sawdust collection and characterization

After a year of weathering, Pinus pseudostrobus sawdust was collected in the sawmill of ejido La Luz, located in Iturbide municipality, Nuevo León, Mexico. Fresh sawdust was obtained after the sawing of logs. Both types of sawdust were conditioned to 65 % relative moisture and 20 ºC, until they reached a constant weight. The percentage of particles retained in each one of the sieves with No. 4, 6, 8, 20 and 40 meshes was determined; furthermore, the normal distribution of the particle size was estimated using the Kolmogorov-Smirnov procedure (^{Razali and Wah, 2011}). Proximal analyses (of moisture content, volatile matter, ashes, and fixed carbon) and an analysis of the sawdust in bulk were conducted in accordance with the standards ^{ASTM D 1762-84 (2001)}, UNE-EN-14774 (^{UNE-EN, 2010} and UNE-EN-15149-2 (^{UNE-EN, 2011}).

Manufacture of briquettes

The briquettes were manufactured in a Lippel^TM laboratory briquette-producing machine, with a vertical orientation. The equipment has a solid base, with a cylinder with a 20 mm diameter and an in-built thermostat which allows to modify the temperature, and two pistons that apply the pressure and facilitate the extraction of the briquettes. Each briquette was made using 200 cm³ of sawdust; the pressures utilized were of 10 and 15 MPa, and the temperatures, of 50, 70 and 90 ºC; five briquettes with a diameter of 3.3 cm and a length of 8.0 cm were made for each treatment.

Characterization of the briquettes

Density was determined according to the standard UNE-EN-16127 (^{UNE-EN, 2012}). The measurements were carried out 48 hours after the briquettes were manufactured, in order to allow them to stabilize so that the dilation effect would not affect the results (^{Križan et al., 2015}).

Resistance to compaction was obtained by applying a load to the whole length of the briquette, as established by ^{Borowski and Hycnar (2013)} and according to the norm ^{ASTM D 143-83 (1994)} (Equation 1). The registered value was the force to the breaking point of the briquette; the tests were conducted in a universal Shimadzuc machine with a capacity for 100 kN.

Where:

RC = Resistance to compaction

L _f = Force to the point of failure

Resistance to impact was determined by the number of pieces formed when each briquette was dropped onto the ground on three occasions, from a height of two meters (^{Chaiklangmuang et al., 2008}; ^{Rajaseenivasan et al., 2016}) (Equation 2).

Where:

RI = Resistance to impact

NP = Number of manufactured pieces

The heating value of the sawdust and of the briquettes was calculated using the equation developed by ^{Cordero et al. (2001)}, which relates the percentage of fixed carbon to that of the volatile matter (Equation 3).

Where:

SHP = Superior heating power (KJ kg^-1)

FC = Fixed carbon (%)

VM = Volatile matter (%)

Statistical analyses

The normality of the data of the variables was determined using the Kolmogorov-Smirnov test (^{Razali and Wah, 2011}). The percentage variables were transformed before the statistical analyses, with the square root of the arc sine (^{Steel and Torrie, 1960}). A factorial design (2 × 2 × 3), in which the sources of variation were: a) type of sawdust, b) pressure, and c) temperature. Each treatment included five repeats.

A variance analysis was applied to the data with a normal distribution in order to determine the existence of statistical differences between treatments (p<0.05); when these occurred, Tukey tests were made. Whenever the distribution was not normal, the Kruskal Wallis test was carried out. All the statistical analyses were performed using the R 3.2.1 statistical package (^{Bolker, 2012}).

Particle size

The percentages of particles between the two types of sawdust in the sieves with mesh sizes. 4, 20 and 40 were statistically different (Figure 1). The fresh sawdust exhibited higher values for particle sizes with mesh sizes 4 and 40. According to ^{Žandeckis et al.(2014)}, particles in the interval of 0.1 to 0.85 mm make products with a better quality. ^{Grover and Mishra (1996)} point out that smaller particles are required in order to increase the adhesion with the larger particles. However, the mechanical resistance of the manufactured briquettes was estimated in terms of the balance in the proportion of the smaller and the larger particles.

Porcentaje = Percentage; Núm. de tamiz = Mesh size; Aserrín fresco = Fresh sawdust; Aserrín intemperizado = weathered sawdust.

Bars followed by the same letter in each sieve do not differ statistically for Tukey’s mean comparison test with a 95 % confidence interval.

Figure 1 Percentage of fresh and weathered sawdust of the sieves with mesh sizes Nos. 4 (4.76 mm), 12 (1.68 mm), 20 (0.84 mm), 40 (0.42 mm), and 60 (0.35 mm). 

Proximal analyses

The moisture content was lower in weathered sawdust (Table 1), which indicates that this type absorbs less moisture under similar conditions, due to hysteresis and to oxygen emission during storage (^{Alakoski et al., 2016}). The moisture content of both types of sawdust corresponds to the interval of 6.39 % to 12.4 % established by ^{Ghaffar et al.(2015)} and ^{Žandeckis et al.(2014)} for the manufacture of high-quality briquettes but is 15 % lower than that considered by ^{Grover and Mishra (1996)}.

The percentage of volatile matter in fresh sawdust coincided with the value established by ^{Brito and Barrichelo (1982)}, cited by ^{Hansted et al. (2016)}, of 75 to 85 %, while that of weathered sawdust was slightly lower.

As for to the ash content, the percentage in weathered sawdust was statistically higher (p<0.05) than that found in fresh sawdust, due to the concentration of inorganic elements during the storage and handling (^{Thek and Obernberger, 2012}; ^{Correa-Méndez et al., 2014}). The percentage of ash is an important factor to be considered during combustion at high temperatures, due to the formation of slag in the furnaces (^{Alakoski et al., 2016}). A value of 0.83 %, registered for weathered sawdust, exceeds the accepted limit (0.50 %) for the production of briquettes (^{ÖNORM, 2000}), while the result for fresh sawdust was similar to that cited by ^{Correa-Méndez et al.(2014)} for Pinus leiophylla Schiede. ex Schltdl. & Cham., P. montezumae Lamb. and P. pseudostrobus.

The percentage of fixed carbon in briquettes made with both types of sawdust exhibited no statistically significant differences; the value for briquettes made from both biomass sources was higher than the one cited by ^{Freitas et al. (2016)}.

Briquettes

Proximal analyses. The lowest moisture content (6.63 %) was found in briquettes made with fresh sawdust at a pressure of 15 MPa and at 70 ºC (Table 2). The interaction between the sawdust type and the temperature evidenced significant statistical differences (P<0.001) (Table 3). The higher the moisture content, the greater the loss of energy was during combustion, due to water evaporation, which reduced the heating power of the material (^{Onyenanu et al., 2015}). A low moisture content is also desirable in order to reduce the rot caused by fungi, as well as disintegration during transportation.

The highest percentage of volatile matter (81.01 %) occurred in briquettes made with weathered sawdust at a pressure of 10 MPa and a temperature of 70 ºC (Table 2). The variable pressure and the sawdust:temperature interaction exhibited statistically significant differences (p<0.05) (Table 3).

The reaction index was calculated based on the volatile matter/fixed carbon ratio, and it indicates the potential reaction of the material during pyrolysis (^{Purohit and Chaturvedi, 2016}); therefore, the most reactive briquette was the one made with weathered sawdust at 10 MPa and 50 ºC.

The ash content exhibited significant statistical differences (p<0.05) in the variables type of sawdust, temperature and sawdust: temperature and sawdust: pressure interactions (Table 3). The fixed carbon ranged between 10.96 and 14.07 % (Table 2), which is below the 17.87 % cited by ^{Ballarin et al. (2010)} for Cupressus lusitanica Mill. sawdust briquettes.

Heating value

The heating value corresponded to the interval of 17.35 to 18.40 MJ kg^-1, of which the highest was for briquettes made with fresh sawdust at a temperature of 70 °C and a pressure of 15 MPa (Figure 2). The results agree with the interval of 16.374 a 19.774 MJ kg^-1 by ^{Onyenanu et al. (2015)} for briquettes made with sawdust with various particle size distributions; these values were is lower than the 19.678 MJ kg^-1 indicated by ^{Cardoso et al. (2016)} for conifer sawdust briquettes, and higher than the 14.68 MJ kg^-1 documented by ^{Rahaman and Salam (2017)} for rice husk briquettes.

Poder calorífico = Heating value; Temperatura = Temperature; Aserrín fresco = Fresh sawdust; Aserrín intemperizado = weathered sawdust.

Bars followed by the same letter in each temperatura do not differ statistically for Tukey’s mean comparison test with a 95 % confidence interval.

Figure 2 Heating value of briquettes produced with fresh and weathered Pinus pseudostrobus Lindl. sawdust at various pressures and temperatures. 

Density

There were significant statistical differences (p<0.05) in the density of the briquettes for the variables type of sawdust, temperature and sawdust:temperature:pressure interaction (Table 4). The highest density was obtained with fresh sawdust at a temperature of 70 ºC and a pressure of 15 MPa (Figure 3a). In a similar assay, ^{Filbakket et al. (2011)} reported denser pellets made with fresh sawdust. On the other hand, ^{Silva et al. (2011)} point out that different percentages in particle sizes produce differences in the density of biofuels. The higher density is also ascribed to the higher percentage of extracts in fresh sawdust, which may reduce friction during pressing, generating greater mobility of the particles and a better arrangement during pressing.

Aserrín fresco = Fresh sawdust; Aserrín intemperizado = weathered sawdust; Densidad = Density; Resistencia a la compression = Resistance to compaction; Resitencia al impacto = Resistance to impact; Número de piezas = Number of pieces.

a = Density; b = Resistance to compaction; c Resistance to impact.

Bars followed by the same letter for each sieve do not differ statistically for Tukey’s mean comparison test with a 95 % confidence interval.

Figure 3 Physical and mechanical properties of briquettes made with fresh and weathered Pinus pseudostrobus Lindl. sawdust at different pressures and temperatures. 

Resistance to compaction

The resistance to compaction of the briquettes exhibited statistical differences (p<0.05) between the sawdust types, the temperature and the sawdust:temperature:pressure interaction (Table 4). Figure 3 shows that the briquettes made with fresh sawdust at 70 ºC and 15 MPa are the most resistant to compaction. The higher resistance of the fresh sawdust briquettes is the result of a greater adhesion between the particles under this condition. Another aspect that affects resistance positively is the higher percentage of particles of less than 0.42 mm (Sieve No. 40). In this regard, ^{Mitchual et al. (2013)} and ^{Chin and Siddiqui (2000)} point out that the particle size affects the mechanical durability of densified products inversely. The percentage of small particles is considered to improve the mechanical resistance up to a certain limit, after which it decreases as the resistance to impact diminishes.

Resistance to impact

The resistance to impact evidenced statistical differences between the sawdust types, temperatures, and the temperature:pressure interaction (Table 4). The fresh sawdust briquettes made at 50 °C and 10 MPa were more resistant to impact as they generated a smaller number of pieces.