Comparative study on the thermal behavior of untreated and

Preparing to load PDF file. please wait...

0 of 0
Comparative study on the thermal behavior of untreated and

Transcript Of Comparative study on the thermal behavior of untreated and

Applied Energy 204 (2017) 1043-1054
Comparative study on the thermal behavior of untreated and various torrefied bark, stem wood, and stump of Norway spruce
E. Barta-Rajnaia, L. Wangb, Z. Sebestyéna, Z. Bartac,
R. Khalilb, Ø. Skreibergb, M. Grønlid, E. Jakaba, Z. Czégénya*
aInstitute of Materials and Environmental Chemistry, Research Centre for Natural Sciences, Hungarian Academy of Sciences, Magyar tudósok körútja 2, Budapest, H-1117, Hungary
bSINTEF Energy Research, Sem Sælands vei 11, Trondheim, NO-7034, Norway cDepartment of Applied Biotechnology and Food Science, Budapest University of
Technology and Economics, Műegyetem rakpart 3, Budapest,H-1111, Hungary dDepartment of Energy and Process Engineering, Norwegian University of Science and
Technology (NTNU), Kolbjørn Hejes v 1B, Trondheim, NO-7491, Norway

E. Barta-Rajnai L. Wang Z. Sebestyén Z. Barta R. Khalil Ø. Skreiberg M. Grønli E. Jakab Z. Czégény

[email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

* Corresponding author. Tel.:+36-13826510 E-mail address: [email protected] Postal address: Magyar tudósok körútja 2, Budapest, H-1117, Hungary

Abstract In this work, the torrefaction of different parts of Norway spruce (stem wood, bark, and stump) was studied. Three different torrefaction temperatures were applied: 225, 275, and 300 °C with 30 and 60 minutes isothermal periods. The thermal stability as well as the evolutions of the decomposition products of the untreated and torrefied samples were measured by thermogravimetry/mass spectrometry (TG/MS). The TG/MS results are interpreted in terms of the chemical composition, namely the cellulose, hemicellulose and Klason lignin content. The inorganic components of the samples were measured by inductively coupled plasma-optical emission spectroscopy (ICP-OES) technique. It was found that the effect of torrefaction temperature was greater than the effect of residence time up to 275 °C, while at 300 °C the residence time had a significant influence on the

composition of the torrefied samples due to the intensive decomposition of cellulose. Principal component analysis has been applied to find statistical correlations between the torrefaction temperature, the residence time, the chemical composition and the thermal parameters of the samples. The results of the principal component analysis confirmed that the chemical composition and hence the thermal properties of the studied samples changed to a greater extent at higher torrefaction temperature than at lower torrefaction temperature.
Keywords: torrefaction, thermogravimetry/mass spectrometry, spruce, bark, stump, stem wood
Highlights:  Comparative study on the thermal behavior of torrefied bark, stem wood and stump  Thermal stability of the samples is interpreted in terms of the chemical composition changes  The residence time has larger effect at higher torrefaction temperature  Hemicellulose side groups are split at milder torrefaction conditions compared to the galactomannan chain  Principal component analysis has been used to identify statistical correlations
1. Introduction The Paris Agreement aims to involve all nations to combat climate change and to
keep the global temperature rise this century well below 2 °C above the pre-industrial levels [1]. The Norwegian national energy strategy has a goal of reducing the domestic greenhouse gas emissions by 30% by 2020, and the long-term goal was to become climate neutral by 2050 [2], which was changed to 2030 later. This strategy indicates that the utilization of lignocellulosic biomass and biofuels (bioethanol, biodiesel, and biogas), as sources of energy, will have to increase substantially in the next few years. Biomass is a carbonaceous renewable energy source, and therefore, it has attracted considerable attention as a replacement for fossil fuels.
Various thermal conversion technologies exist to produce bioenergy from lignocellulosic biomass, such as combustion [3], gasification [4], pyrolysis [5], as well as cofiring of biomass and coal [6]. However, in energetic applications the properties of raw lignocellulosic materials create challenges for their efficient utilization. One of the main difficulties is the high moisture content of the untreated biomass, which reduces the efficiency of the conversion process and increases the fuel transportation costs. Some of the

other problems with raw biomass materials are the following: low calorific value, low energy density, hydrophilic nature, and high oxygen content. Furthermore, the transportation, storage, and grinding are costly due to the low density and the fibrous nature of lignocellulose. Torrefaction is a mild thermal treatment method performed between 200 and 300 °C in an inert atmosphere for reducing the mentioned disadvantages [7]. A major goal of torrefaction is to upgrade the quality of the solid product by decreasing the moisture content and increasing the hydrophobicity, grindability, and energy density of biomass. The volumetric energy density of torrefied biomass can be increased by a combined grinding and pelletizing step after torrefaction [8-9]. In this way, the torrefied material can be handled and stored like coal.
While pelletization of lignocellulose is an established technology, torrefaction is still a developable process for the production of solid energy carriers. Recent research papers focus on the viability of torrefaction as a part of integrated approaches [10-13]. The major technical challenges are the predictability and consistency of the product quality, the flexibility related with using different input materials, and the densification of torrefied biomass [14]. The applied torrefaction condition (temperature and residence time) and the moisture content have significant influences on the pellet production (e.g., compression and friction energy) and pellet quality (e.g., strength) [15]. In order to estimate the feasibility of a commercial torrefaction system in a particular region, local and abundant biomass resources should be investigated.
During tree harvesting, stem wood is the main product, while the other parts of the tree (including bark and stump) are considered as by-products. According to the literature, stump constitutes 23-25% of the stem volume of a coniferous tree [16] and bark can reach 620% of the total volume of the stems [17]. These forest residues represent an abundant and underutilized source of renewable energy. Many studies have been carried out on the thermal characteristics of stem wood [7, 18-19]. These papers focus on the effect of torrefaction on the properties of the solid product, such as mass yield, energy content, hydrophobicity, grindability, and particle-size distribution. Only a few papers are available on the thermal decomposition of forest residues, such as bark and stump. The thermal behavior of bark and wood of Eucalyptus tree has been studied during torrefaction [20-21]. Almeida et al. [20] concluded that the mass loss is an excellent indicator of the treatment severity. It was suggested [21] that the most feasible torrefaction temperature was between 298 and 310 °C for Eucalyptus wood and bark. The torrefaction of stump has been studied focusing on the kinetic evaluation [22] and the thermogravimetric results [16]. In the literature, there is a lack

of papers, which compare the thermal behavior of different parts of the coniferous tree during torrefaction. A profound understanding of the thermal behavior of stump and bark is essential for the efficient utilization of these abundant energy sources in the future.
Thermoanalytical methods are suitable to determine similarities and differences between the compositions of the lignocellulosic materials without separating the main fractions [23]. Several factors may influence the thermal decomposition of lignocellulosic materials. The alkali ions are known to exert a great influence on the thermal decomposition of cellulose [23-25] and lignin [23, 26-27]. As a consequence of the difference in the relative amounts of cellulose, hemicellulose, lignin, extractives, and inorganic materials, the different biomass materials behave differently during thermal decomposition. Therefore, monitoring the changes in the chemical composition is essential during torrefaction. Nevertheless, comparison of chemical analysis and thermal analysis results is rarely carried out in the biomass literature.
The aim of this work has been to gain information about the thermal behavior of untreated and various torrefied bark, stem wood and stump of Norway spruce, which is the most abundant wood species in Norway and in the Northern hemisphere. The thermal stability and the formation of the volatile products of untreated and torrefied samples have been studied by thermogravimetry/mass spectrometry (TG/MS). The main differences between the thermal decomposition of the studied samples are interpreted in terms of the chemical composition (cellulose, hemicellulose and Klason lignin) with the goal of understanding the mechanisms of the decomposition of biomass components during torrefaction. The obtained data were evaluated by principal component analysis (PCA) to identify correlations between the temperature of torrefaction, the residence time, the chemical composition and the thermal behavior of the studied samples.

2. Materials and Methods 2.1. Materials
Different parts of a representative single Norway spruce (Picea abies) tree were selected for the torrefaction study: bark, stem wood and stump. The samples originated from a Norway spruce forest in South Norway. The trees in the forest site have high ages, more than one hundred years old on average. After harvested, the trees were divided into three parts: trunk, stump, and forest residues. The trunk was further debarked to obtain stem wood and bark. The stem wood was first cut to strips, and then further chopped into cubes with sides of 1 cm. The bark was chipped into pieces and those with length of around 5-7 cm were used for the torrefaction experiments. The stump was shredded into pieces and the pieces with length of 3-5 cm were torrefied.
2.2. Methods
Torrefaction experiments
The torrefaction experiments were carried out in a batch tube reactor placed in an electrical furnace in nitrogen atmosphere using flow rates of 1 L/min. Approximately 80 g samples were heated up at a heating rate of 15 °C/min to temperatures of 225, 275 and 300 °C in the tube reactor followed by 30 and 60 minutes isothermal periods, whereafter the reactor was cooled down to room temperature. For further experiments the untreated and torrefied samples were ground by a cutting mill to <1 mm particle size.
Higher heating value determination
The higher heating value (HHV) was determined using an automatic IKA C 5000 bomb calorimeter. The combustion of approximately 0.5 g dried sample in pure oxygen atmosphere was performed under 30 bar pressure. The heat capacity of the calorimeter system was determined by benzoic acid calibration. All heating values were calculated using the average of three replicates.
Inductively coupled plasma-optical emission spectroscopy (ICP-OES)
Approximately 2 g biomass samples were ashed at 550 °C in a furnace according to CEN/TS 14775:2004 standard method. The ashes were fused at 920 °C with a fusion blend (Li2B4O7:LiBO2, 2:1) and digested by 25 mL 33% nitric acid. The calcium, potassium, sodium, and magnesium contents of the samples were determined by a Spectro Genesis ICP-

OES (Spectro Analytical Instruments) with axial plasma observation. The amounts of the ashes have been determined according to the CEN/TS 14775 EU standard method.
Carbohydrate and Klason lignin content determination
The contents of carbohydrates were determined according to the method of Sluiter et al. [28] applying slight modifications. The milled samples (<1 mm) were dried at 40 °C for 1 day. The raw and torrefied biomass samples were treated in a two-step acid hydrolysis with 72% H2SO4 for 2 hours at room temperature, and then with 4% H2SO4 for 1 hour at 121 °C. The gained suspensions were filtered and washed with distilled water through G4 glass filter crucibles. The sugar concentrations (glucose, mannose and galactose) of the filtered supernatants were analyzed by high performance liquid chromatography (HPLC) using an Agilent 1260 system with a Hi-Plex H column (Agilent, CA, USA) at 65 °C. An eluent of 5 mM H2SO4 was used at a flow rate of 0.5 mL min-1. The solid residues obtained after washing were dried at 105 °C until constant weight. The dried residues consisted of acidinsoluble organics and acid-insoluble ash. The total ash and acid-insoluble ash contents were measured by ashing the sample at 550 °C for 5 hours until the sample weight was constant [29]. The Klason lignin content was calculated by subtracting the acid insoluble ash content from the acid insoluble residue content. All experimental data were determined using three replicates.
Thermogravimetry/mass spectrometry (TG/MS)
The TG/MS system consists of a modified Perkin-Elmer TGS-2 thermobalance and a Hiden HAL quadrupole mass spectrometer. About 5 mg samples were analyzed in argon atmosphere. The samples were heated at a rate of 20 °C min−1 from 25 to 900 °C in a platinum sample pan. The evolved products were flushed through a glass lined metal capillary heated at 300 °C by argon gas using a flow rate of 140 mL min-1. The ion source of the mass spectrometer was operated at 70eV electron energy. The mass range of 2-150 Da was scanned. The ion intensities were normalized to the sample mass and to the intensity of the 38Ar isotope of the carrier gas (used as an internal standard). Since the MS intensities of various products have different magnitudes, they have been scaled to gain comparable peak heights in the plots. The curves of the individual species developed from bark, stem wood and stump are plotted using the same scale in each of the TG/MS figures.
Principal component analysis (PCA)

Due to the large number of samples and experimental data, principal component analysis (PCA) using the Statistica 12 software (StatSoft, Inc. Tulsa, Oklahoma, USA), was employed. PCA has been used to reveal correlations between the TG data and the chemical composition of the studied samples. PCA is a technique for reduction of data dimensionality, which allows detecting patterns and visualization of patterns retaining as much important information present in the original data as possible [30-31]. The values that represent the samples in the space defined by the principal components (Factors) are the component scores. Factor loadings show the correlation between the original variables and the Factors, and it may help understand the underlying nature of a particular Factor.

3. Results and Discussion 3.1. Comparison of the three untreated samples
Table 1 summarizes the higher heating value, the ash content and selected data of the ICPOES characterization of the untreated samples. As the results illustrate, the heating values are rather similar, while the bark has significantly higher ash content than stem wood and stump. The bark has an order of magnitude higher K+ and Si contents than stem wood and stump. Furthermore, the Na+ and Ca2+ contents of the bark are also higher compared to stem wood and stump.

Table 1. Characterization of the untreated samples.

Higher heating value (MJ/kg, dba) Ash content (% m/m, arb) Inorganic components (mg/kg, dba) Ca2+ K+ Na+

Bark Stem wood 20.14 19.78 2.43 0.31

7803 1030

2011 272



Stump 19.51 0.43
1235 245 36

Si adry basis, bas received

3602 82


Fig. 1a shows the chemical composition of the three untreated samples. The sum of the mannan and galactan contents represents the hemicellulose fraction, whereas the glucan content of the samples mainly characterizes the cellulose fraction of biomass. The Klason lignin content is defined as the acid insoluble residue of the samples without the acid

insoluble ash. Besides the acid insoluble lignin, the Klason lignin contains all acid insoluble components of the sample, excluding ash. The fraction denoted by “Other” represents the sum of unquantified components and includes extractives, acid soluble lignin and acid soluble minerals. As Fig. 1a shows, the bark sample has the highest Klason lignin content, stem wood has the highest cellulose content and stump has the highest hemicellulose content. The thermogravimetric (TG) and derivative thermogravimetric (DTG) curves of the untreated samples are shown in Fig. 1b. The main DTG peak is dominated by the decomposition of cellulose, while the shoulder at lower temperature (around 320 °C) can be attributed mainly to hemicellulose decomposition. The lignin decomposes at a lower rate in a wide temperature range (200–600 °C). The evaporation and decomposition reactions of extractives start at lower temperatures and it is visible as a shoulder on the main DTG peak from approximately 160 °C. The comparison of the three untreated samples shows that bark releases the most extractives, in the low temperature range. The untreated stump has the most characteristic hemicellulose shoulder, which is in agreement with the chemical composition results showing that stump has the highest hemicellulose content. The decomposition of bark starts at the lowest temperature, the DTG peak maximum occurs at the lowest temperature, and the maximum rate of decomposition is considerably lower than in the case of the stem wood and stump. The high lignin content (41%) of bark results in the formation of a high yield of char during thermal decomposition. The different thermal behavior of the different untreated samples can be explained by their different composition as well as by the fact that alkali ions have catalytic effects on the decomposition mechanism of cellulose [23-25] and the charring reactions of lignin [23, 26-27].

Fig. 1. (a) Composition and (b) TG and DTG curves of untreated bark, stem wood, and stump. Regarding the thermal behavior of the studied samples, further information is given by the mass spectrometry curves. The thermal decomposition of extractives, cellulose, hemicellulose and lignin results in a high yield of low molecular mass compounds at low heating rate, hence the evolution profiles of these products are characteristic to the decomposition of the different parts of the tree. Fig. 2a, c and e presents the DTG curves as well as the evolution of water and the main permanent gases, while Fig. 2b, d and f presents the evolution of some typical organic products measured by the mass spectrometer during the thermal decomposition of the three untreated samples.

Fig. 2. The DTG curves and the evolution profiles of the most characteristic decomposition products and fragment ions released from untreated bark (BA U), stem wood (SW U) and stump (ST U) (m/z 2, hydrogen; m/z 16, methane; m/z 18, water; m/z 28, carbon monoxide; m/z 44, carbon dioxide; m/z 27, C2H3+; m/z 30, formaldehyde; m/z 31, CH3O+; m/z 45, COOH+).