Valorization of Cistus ladanifer and Erica arborea shrubs for fuel: Wood and bark thermal characterization

As a form of upgraded biomass characterized by its high energy density, low production costs, and low process energy requirements, wood pellets are an environmentally friendly fuel allowing for carbon neutral heating with high energy efficiency. In this work, the suitability of a valorization of the woods from the two most representative shrub species from the Iberian Peninsula (namely Cistus ladanifer and Erica arborea) for heating has been assessed. Whereas Erica arborea met the requirements of ISO 17225-2:2014 for ENplus-B class (the calorific content for both wood and bark was high and not significantly different, and the ash content was permissible for specimens with branch diameter ≥2,8 cm), Cistus ladanifer was in the limit of the normative and only met the requirements in terms of acceptable ash percentage (1,9%) and heating value (19 kJ·g-1) for old specimens with branch diameters >3,4 cm. Consequently, while the harvest of E. arborea for its use as fuel does not need to be selective, that of C. ladanifer should be limited to the most robust specimens and foliage should be avoided.


INTRODUCTION
A significant proportion of Mediterranean forest vegetation consists of evergreen small diameter hardwood shrubs, such as Cistus ladanifer (gum rockrose) and Erica arborea (tree heath), which have been traditionally used as fuelwood for domestic heating purposes. In the geographic area under study (Castilla y León, Spain) both species are so abundant that their utilization as a biomass resource for energy purposes has aroused significant interest: In fact, since 2012, field studies aimed at this valorization, funded by the European Union through the LIFE+ and Joule programs, have been conducted in several municipalities in the province of Zamora (Spain). The ultimate goal would be to collect tree heath and gum rockrose for their combustion in district heating facilities (municipal boilers) in Fabero (Soria, Spain) and Las Navas del Marqués (Ávila, Spain), as well as for the electricity production plant located in Garray (Soria, Spain). Of the two bushes into consideration, the most potentially profitable would be E. arborea, whose heating value was recently reported to be the highest of all evergreen Mediterranean hardwood species (Barboutis and Lykidis 2014).
The use of biomass as an energy source provides substantial socio-economic and environmental benefits. However, bio-fuels have low bulk densities which limit their use to areas around their origin, being this drawback a restrictive factor for their energy use. Nevertheless, densification by pelleting minimizes this disadvantage. Global pellet production has considerably increased in the last years (from 7 to 19 million tons between 2006 and 2012 (Duca et al. 2014)), mainly in Europe and North America, and the growth in pellet consumption has resulted in more diversity. Consequently, the industry has started looking for products such as wastes obtained from forestry and scrubland wood. The doubtful quality of these materials originated the development of quality standards in some countries, so as to guarantee the right use of the different types of pellets in combustion equipment. Due to differences in chemical structure, bark and wood from C. ladanifer and E. arborea should show different properties, and -in particular-those related to their applicability as fuels. This differentiation is important because the bark of all evergreen hardwood species usually presents significantly higher ash content than wood and, in agreement with the international standard ISO 17225-2:2014 (International Organization for Standardization 2014) -which has recently superseded the European Standard EN 14961-2 for the quality characteristics of pellets (European Pellet Council 2011)-, the threshold ash content value is 2%. In this normative, the required net calorific value (NCV) or lower calorific value (LHV) is ≥16,56 kJ·g -1 and the higher heating value (HHV) is ≥18,82 kJ·g -1 .
The aims of the study presented herein have been: (i) to correlate the results of our analytical determinations and related calculations on HHV and ash content (AC) for bark and wood from C. ladanifer and E. arborea with those from other direct and indirect methods used in the literature; and (ii) to explore which bark diameters would meet the ISO 17225-2:2014 (International Organization for Standardization 2014)/ENplus (ENplus 2015) requirements for HHV and AC with a view to the valorization of these two shrub species as fuels. This is in line with the work by other authors on woods from other species (Duca et al. 2014, Miranda et al. 2017.

MATERIAL AND METHODS
The quantity known as higher heating value (also referred to as gross energy, upper heating value, gross calorific value (GCV) or higher calorific value (HCV)) is determined by bringing all the products of combustion back to the original pre-combustion temperature and, in particular, condensing any vapor produced. This is the same as the thermodynamic heat of combustion, since the enthalpy change for the reaction assumes a common temperature of the compounds before and after combustion, in which case the water produced by combustion is condensed to a liquid, hence yielding its latent heat of vaporization.
Calculations for the estimation of biomass and heating values may be obtained either by direct or by indirect methods. Direct methods involve the destruction of heavy biomass, whereas in indirect methods equations are used to estimate heating values from measurements of other variables, making the process easier (Bombelli et al. 2009).
In the first part of this study, heating values were determined by a destructive method, which comprised the selection, felling and extraction of biomass of each of the species and its subsequent combustion. C. ladanifer samples had a height of 115,3±32,4 cm and a crown width of 28,68±15,25 cm while E. arborea samples had a height of 158,2±49,0 cm and a crown width of 103,7±60,0 cm. The aerial part was separated from the roots using a saw and then, following an analogous procedure to that described by Ruiz-Peinado et al. (2012), root systems were excavated by using a tractor with a shovel and then spades to complete the job. For each plant, soil was excavated down in a circular area of twice the mean crown diameter. In addition to the main body of the roots, those remaining in the hole were also collected. Samples were transported to the laboratory (ETSIIAA facilities, Universidad de Valladolid, Spain), where they were separated into different fractions and weighed (fresh weight). In the case of C. ladanifer, they were classified into leaves, xerochastic capsules, branches (thin: 3-7 mm in diameter; thick: 7-17 mm in diameter) and roots. On the other hand, for E. arborea -given its morphology and the impracticality of leaves separation-they were divided into four fractions: leaves with flowers and fruits, fine material (<1 cm), thick material (<5 cm) and roots, in agreement with Mello et al. (2012).
In addition to aforementioned information, the stem diameter (2R), bark thickness (f) and wood and bark percentages were characterized for both species. The proportion of bark was calculated as the Valorization of Cistus ladanifer and..: Carrión-Prieto et al.
ratio of bark area in a transverse section to the total stem area of this section, according to equation (Barmpoutis et al. 2015): where Z=bark percentage (%), R=barked stem radius (cm) and f=bark thickness (cm).
For the determination of the bark percentage, the transverse surfaces were assumed to be circular. Consequently, bark and wood were separated and the materials were ground by means of a portable chipper. The resulting data is summarized in Table 1. It should be noted that for the study of C. ladanifer two sets of individuals were selected: ones with average stem diameter (1,9 cm trunk diameter) and also robust old specimens (older than 12 years, according to the equation y = 1,5496 x + 1,5342 R 2 (Valares Masa et al. 2016)), with diameters above the average, provided that this second group was more likely to meet the EN standard. Ten repetitions were carried out for each group.
Calorific values, expressed as HHV, for C. ladanifer and E. arborea fractions were calculated from elemental analysis data in agreement with the US Institute of Gas Technology (IGT) (Talwalkar et al. 1981): HHV = 0,341(%C) + 1,322(%H) -0,12(%O+%N), where %C, %H, %O, %N are the mass fractions in wt% of dry material and HHV the heating value for dry material in MJ/kg. Although originally derived from data on coal, this formula has been shown to give acceptable results for a wide range of carbonaceous materials including biomass (CHPQA 2008).
Alternatively, HHV values were also calculated from holocellulose and lignine+extractives percentages, following the guidelines of Aseeva et al. (2005) and Kienzle et al. (2001) and applying a factor of 17,5 for holocellulose and of 25,5 for the lignine+extractives mixture.
Experimental HHV values were determined in a Parr 1261 isoperibol bomb calorimeter (Thermo Fisher Scientific, Waltham, MA, USA), according to the method described in BS EN 14918:2009 standard (British Standards Institution 2010). Other experimental values, such as the total enthalpy of combustion, were obtained from differential scanning calorimetry (DSC) curves by numerical integration of the experimental signal on the whole temperature range (30-600 ºC). DSC data were obtained on a TA Instruments (New Castle, DE, USA) mod. Q100 v.9.0 DSC equipped with an intracooler cooling unit at -25 ºC (with a 1:1 volume mixture of ethylene glycol-water), at a heating rate β=20°C/min and at a N 2 :O 2 ratio of 4:1 (20 mL/min). Samples were hermetically sealed in aluminium pans, and an empty pan was used as a reference. TG/DTG analyses were conducted with a Perkin-Elmer (Waltham, MA, USA) STA6000 simultaneous thermal analyser by heating the samples in a slow stream of N 2 (20 mL/min) from room temperature up to 700 ºC, with a heating rate of 20 ºC/min. Pyris v.11 software was used for data analysis (PerkinElmer 2014). Temperature calibration was performed with high-grade standards, biphenyl (CRM LGC 2610) and indium (Perkin-Elmer, x=99,99%), which was also used for enthalpy calibration.

HHV experimental values from calorimetry
HHV, determined with an isoperibol bomb calorimeter, according to the method described in the EN 14918:2009 standard, yielded values of 19,7 kJ·g -1 for C. ladanifer and 21,0 kJ·g -1 for E. arborea.

Results from DSC and TG/DTG curves
DSC curves for C. ladanifer and E. arborea woods are shown in Figure 1 and their thermal effects (mainly due to holocellulose and lignin combustion) are summarized in Table 5. Overall enthalpy change values obtained from these curves resulted in 18,04 kJ·g -1 and 18,63 kJ·g -1 , respectively.  T peak stands for the temperature at which the maximum mass loss occurred, according to TG/DTG measurements; T offset stands for the temperature at which the maximum value of heat flux occurred, obtained from the DSC thermograms.
The ash content of the various fractions of C. ladanifer and E. arborea was estimated from the residue after heating at 700 ºC (Figures 2, 3, 4 and 5), according to the usual temperature conditions for pyrolysis in oxygen bomb calorimeters (Wang et al. 2016). Both the inner and outer parts of the stem and those of the epidermis and cortex of the roots of C. ladanifer resulted in percentage values ranging from 0,5% to 0,6% with no significant differences between fractions. Conversely, for E. arborea, the percentages for the stem outer part and the root epidermis ranged from 0,6 to 0,9%, while those for the inner parts of the root and the stem were 0,19% and 0,36%, respectively. It is worth noting that all these values were below 2%.
Whole enthalpy change values from thermal analysis were around 18,04 kJ·g -1 and 18,63 kJ·g -1 for C. ladanifer and E. arborea, respectively. These values can be assigned to low heating values (LHV), provided that they would be in good agreement with those expected from the holocellulose and lignin net calorific values (ca. 17 kJ·g -1 and ca. 21 kJ·g -1 , respectively (Energy research Centre of the Netherlands 2012)) and the percentages reported in Table 2. In fact, the value reported in the literature for the LHV of C. ladanifer is 17,9 kJ·g -1 (Martínez et al. 2000), very close to the one reported herein.
Regarding the ash content broken down for each of the fractions, the highest value was obtained for stem bark (around 6,0%), thus identifying this fraction as the one which compromises the use of these shrubs as fuelwood.
In terms of the requirements of ISO 17225-2:2014 (International Organization for Standardization 2014) for ash content of pellets (ENplus-B class) and in view of Table 7, C. ladanifer stems with a diameter of 1,9 cm would be non-compliant, while those with diameters over 3,4 cm would be acceptable. Consequently, we propose this minimum barked diameter to produce pellets of ENplus-B class. The barked diameter value proposed for E. arborea is entirely coincident with that suggested by Barboutis and Lykidis (2014) following the EN 14961-2 norm.

CONCLUSIONS
One of the requirements of current European standards concerning biofuels in the form of pellets for their use in rural district heating is the ash percentage maximum, limited to 2%. Ash content is significantly influenced by the bark and foliage percentages of the plants to be used as fuel. Both shrub species under study, C. ladanifer and E. arborea, yielded HHV values that met the requirements established in the regulations for their use as fuel. However, only the ash contents for E. arborea were compliant without ambiguity. In the case of C. ladanifer, biomass ash percentage was in the upper limit of the normative and this would be a problem for its acceptance as fuelwood. To ensure its adequacy, only old specimens (with stem diameters ranging from 2 to 4,8 cm) should be harvested, avoid foliage.