Abstract
Trends in the production and applications of torrefied wood/biomass are reviewed in this article. The thermochemical conversion of biomass is a promising technology because biomass is an environmentally friendly fuel that produces substantially lower CO2 emissions compared to fossil fuel. Torrefaction is the thermal treatment of biomass at temperatures from 200 to 300 ˚C in the absence of air or oxygen to liberate water and release volatile organic compounds, primarily through the decomposition of the hemicelluloses. Torrefied biomass has a higher heating value, is more hydrophobic, resists rotting, and has a prolonged storage time. The different torrefaction technologies and reactors are described. An overview of the applications of torrefied biomass, the economic status, and future prospects of torrefaction technology are presented and discussed. Currently, torrefaction demonstration plants have technical problems that have delayed their commercial operation. Torrefaction reactors still require optimization to economically meet end-use requirements and attain product standardization for the market. Several characteristics of torrefaction need to be demonstrated or scaled up for successful commercialization.
Download PDF
Full Article
Current Trends in the Production and Applications of Torrefied Wood/Biomass – A Review
Anthonia E. Eseyin,a,* Philip H. Steele,a and Charles U. Pittman Jr. b
Trends in the production and applications of torrefied wood/biomass are reviewed in this article. The thermochemical conversion of biomass is a promising technology because biomass is an environmentally friendly fuel that produces substantially lower CO2 emissions compared to fossil fuel. Torrefaction is the thermal treatment of biomass at temperatures from 200 to 300 ˚C in the absence of air or oxygen to liberate water and release volatile organic compounds, primarily through the decomposition of the hemicelluloses. Torrefied biomass has a higher heating value, is more hydrophobic, resists rotting, and has a prolonged storage time. The different torrefaction technologies and reactors are described. An overview of the applications of torrefied biomass, the economic status, and future prospects of torrefaction technology are presented and discussed. Currently, torrefaction demonstration plants have technical problems that have delayed their commercial operation. Torrefaction reactors still require optimization to economically meet end-use requirements and attain product standardization for the market. Several characteristics of torrefaction need to be demonstrated or scaled up for successful commercialization.
Keywords: Torrefaction; Torrefied wood/biomass; Torrefaction technologies; Torrefaction reactors
Contact information: a: Department of Sustainable Bioproducts, Mississippi State University, MS 39762 USA; b: Department of Chemistry, Mississippi State University, MS, 39762 USA;
* Corresponding author: eseyinae@gmail.com
INTRODUCTION
Biomass, which can be defined as lignocellulosic material from plants, has been recognized as the fourth largest energy source in the world; it is an important source for both renewable fuels and valuable chemicals (Saxena et al. 2008). A primary use of biomass is for power generation. It can be used directly as solid fuel or processed into gaseous or liquid fuels (Briens et al. 2008). There are many biochemical and thermochemical processes available to convert biomass to different fuels and chemicals.
Biomass is renewable, since new biomass can be grown to replace that used for energy. This growth removes CO2 from the atmosphere, neutralizing the CO2 emission generated when converting biomass to energy. However, several problems and challenges are unavoidably encountered during biomass utilization because of the diversity of its physical and chemical compositions, which depend on the origin of the raw material.
When biomass is used as feedstock for power generation, it often exhibits undesirable properties. Some types of biomass have high ash content, which leads to the agglomeration of the bed material inside the boiler as well as fouling the surface of heat transfer tubes in combustion chambers (Oehman et al. 2005; Pronobis 2006; Romeo and Gareta 2009). Raw biomass generally has low calorific value because of its high moisture and oxygen contents (Pimchuai et al. 2010; Chen and Kuo 2011).
Due to rigidity, mechanical strength, poor flow and fluidization properties, biomass requires high grinding energy. It is also difficult to feed into boilers (van der Stelt et al. 2011; Li et al. 2012; Ohliger et al. 2013). Other challenges to biomass use include the large land surface required to grow it (Higman and van der Burgt 2008) and high costs for collection and transportation (Biagini et al. 2005). After drying, biomass can regain moisture and may rot during storage (Bergman 2013). Biomass is hygroscopic and forms a considerable quantity of soot during combustion.
To enhance biomass utilization efficiency and limit the challenges mentioned above, a torrefaction pretreatment is beneficial (Mosier et al. 2005; Zwart et al. 2006; Acharjee et al. 2011; van der Stelt et al. 2011). Torrefaction technology and its applications have advanced significantly over the last decade (Stamm 1956; Kamdem et al. 2002; Tjeerdsma and Militz 2005; Stanzl-Tschegg et al. 2009; Chen et al. 2011; Phanphanich and Mani 2011; Agar and Wihersaari 2012a; Dhungana et al. 2012; Huang et al. 2012; Syu and Chiueh 2012; Makarov et al. 2013; Becer et al. 2013; Johnston 2013; Wilen et al. 2013; Doassans-Carrere et al. 2014; Halina et al. 2014). However, there are unresolved issues and an incomplete understanding of the scientific process of torrefaction.
This review discusses current trends in the production and applications of torrefied wood/biomass. First, the advantages of wet and dry torrefaction are discussed. Laboratory scale studies on torrefaction including mass and energy balances of wet lignocellulosic biomass torrefaction, torrefied wood/biomass gasification and characterization of torrefaction products as well as torrefaction kinetics are presented. An overview of current torrefaction technologies and reactors, advantages and disadvantages of these technologies then follow. In the fourth part of this review, applications of torrefied wood/biomass are described. These include: gasification, co-firing of torrefied biomass with coal, combined heat and power generation, standalone combustion, production of bio-based fuels and chemicals, heating blast furnaces and industrial applications. The final sections enumerate the economic status and future prospects of torrefaction technology.
The following excellent reviews on biomass torrefaction have been published: (Chew and Doshi 2011; van der Stelt et al. 2011; Koppejan et al. 2012; Batidzirai et al. 2013; Chen et al. 2015).
WHAT IS TORREFACTION?
Torrefaction is the thermal treatment of wood/biomass in the low-temperature range to achieve biomass energy balance optimization, to promote grindability, and to reduce the hygroscopic nature of biomass. This in turn reduces its susceptibility to biological decay (Kamdem et al. 2000; Almeida et al. 2010; Acharjee et al. 2011). The effect that torrefaction has on reducing the grinding energy needed is a primary consideration in many energy-producing applications. These include the co-firing of lignocellulosic materials in pulverized coal-fired power plants and other industrial kilns (e.g. cement, coke and steel industry kilns) (Phanphanich and Mani 2011; Koppejan et al. 2012). In addition to these impacts, irreversible material property changes such as reduction of strength, toughness, and abrasion can occur (Stamm 1956; Kamdem et al. 2002; Tjeerdsma and Militz 2005; Stanzl-Tschegg et al. 2009).
Torrefaction research was performed in France in the 1930s. Then, in the 1980s, the results of torrefaction experiments using two temperatures and two tropical wood samples at 270 to 275 ºC were published (Bourgeois and Doat 1984). This research led to the building of a continuous wood torrefaction plant in 1987. Torrefaction technologies can be divided into either the wet or dry process.
DRY TORREFACTION
Dry torrefaction is the thermal treatment of wood/biomass at temperatures of 200 to 300 ºC in the absence of air or oxygen. The process results in the liberation of water and volatile organic compounds. This primarily occurs through the de-volatilization of the hemicelluloses. Dehydration and decarboxylation reactions occur during torrefaction. Cellulose and lignin in woody biomass are decomposed at temperatures above 300 ºC (Chouchene et al. 2010). In spite of the fact that 30 wt.% of biomass can be lost during torrefaction, the torrefied product may retain up to 90% of the initial energy content (van der Stelt et al. 2011).
Dry torrefaction technology has been rapidly developed to the stage of market introduction and commercial operation. Several torrefaction installations have recently been built in Europe and North America. Market analyses have predicted that in 2020, torrefaction technology market will be about 130 million tons per year (Walton and Van Bommel 2010).
Once biomass has been torrefied, it has become a distinctly different material that has several advantages and disadvantages when compared to the original biomass. Figure 1 is an overview of an integrated torrefaction plant.
Fig. 1. Overview of an integrated torrefaction plant (Kiel et al. 2012)
Table 1. Logistics, Advantages, and Disadvantages of Handling Torrefied Wood/Biomass (Stelte 2013)
Torrefaction can be used as a pretreatment method prior to fast pyrolysis of biomass to bio-oil. This pretreatment improves the quality of pyrolysis oil by lowering its water content and the proportion of low molecular weight compounds (Meng et al. 2012; Zheng et al. 2012). Torrefaction in general, is an effective method for reducing the water, acid, and oxygen contents of bio-oil when derived from fast pyrolysis of torrefied biomass. Removing oxygen raises the heating value and pH of bio-oil. Torrefaction-aided fast pyrolysis is a stepwise biomass pyrolysis through which the major biomass constituents are more selectively decomposed into a variety of chemicals at each stage. Products from each stage can be less complex and more stable than the products from direct fast pyrolysis, where all biomass constituents are decomposed synchronously at the same temperature (Czernik and Bridgwater 2004; Mohan et al. 2006). Torrefaction is influenced by the biomass chemical properties, treatment temperature, reaction time, and the apparatus used (Prins et al. 2006a; Chen et al. 2011). Meanwhile, the effects of torrefaction severity on the structure of torrefied biomass, its corresponding fast pyrolysis behavior and pyrolysis mechanism are currently not well understood (Zheng et al. 2013).
WET TORREFACTION
Wet torrefaction (also referred to as hydrothermal pretreatment), is the treatment of biomass in hydrothermal media or hot water at temperatures between 180 and 260 ºC and pressures up to 4.6 MPa (Yan et al. 2009, 2010; Bach et al. 2013; Runge et al. 2013; Chen et al. 2012). Biomass is immersed in water at these conditions from 5 to 240 min (Lynam et al. 2011), resulting in the formation of solid fuel, aqueous compounds, and gases (Yan et al. 2009; Sasaki et al. 2003; Ando et al. 2000). The resulting solid product contains about 55 to 90% of the original mass and 80 to 95% of the fuel value of the original feedstock. Water soluble compounds, consisting primarily of monosaccharides, furfural derivatives, and organic acids, make up approximately 10% by mass of the by-products. Gaseous products make up the balance (Bobleter 1994; Petersen et al. 2009).
In wet torrefaction, the hemicellulose can be hydrolyzed and completely solubilized into the aqueous phase, while the lignin binding is disrupted. However, cellulose is almost entirely preserved in this solid product. Nonetheless, the enzymatic digestibility of cellulose is enhanced because the cellulose in wet torrefied biomass is now more readily accessible to enzymes. Wet torrefaction is therefore an effective pretreatment technology for enhancing subsequent enzymatic hydrolysis of cellulose (Yu et al. 2011; Cybulska et al. 2012; Rohowsky et al. 2013). The term autohydrolysis is typically used in this context.
Wet-torrefied biomass has more fixed carbon (proximate analysis) and elemental carbon per unit of dry matter (ultimate analysis) than raw biomass. Thus, a higher weight fraction of biomass is transformed into a fuel with properties that resemble low-rank coal. With reduced equilibrium moisture content, the pretreated solid is more hydrophobic than the original biomass. Wet torrefied biomass can be easily stored to accommodate seasonal availability because it has far less propensity to absorb water, swell, or decompose. Wet-torrefied biomass is also very friable. Since it contains lignin, it can be pelletized for feeding to a thermochemical conversion process (Yan et al. 2009).
Wet torrefaction is similar to hydrothermal carbonization (Goto et al. 2004; Funke and Ziegler 2010; Parshetti et al. 2013; Liu et al. 2013; Hoekman et al. 2014), and it is sometimes discussed under the general term ‘‘hydrothermal conversion’’ (Knezevic et al. 2009; Wang et al. 2011; Kruse et al. 2013) or “hydrothermal treatment’’ (Karagoez et al. 2004; Nonaka et al. 2011; Murakami et al. 2012). In spite of the fact that wet torrefaction and hydrothermal carbonization have sometimes been used interchangeably, there is a significant difference between these terms. Wet torrefaction is primarily used for the production of upgraded solid fuels for energy applications only. In contrast, hydrothermal carbonization is applied to the production of charcoal that has much higher carbon content. This can be used not only as fuel but also as activated carbon, soil enhancers, or fertilizers, etc. Compared to dry torrefaction, (200 to 300 ºC) the reaction temperature used for wet torrefaction is lower (180 to 260 ºC). The pressure used is the saturated water vapor pressure, generated at the temperature applied.
ASSESSMENT OF DRY AND WET TORREFACTION
There are differences in the chemical structures of dry and wet torrefied biomass. These differences often give rise to subsequent divergent pyrolysis behavior observed in these two types of pretreated biomass. After wet torrefaction, the wet hydrophobic solid product can be effectively dried mechanically and/or by natural dewatering. These options are attractive and significantly reduce the energy requirements for the post-drying step. Valuable organic compounds such as acetic acid, formic acid, lactic acid, glycolic acid, levulinic acid, phenol, furfural, HMF, and sugars are found in the aqueous phase products of wet torrefaction, accounting for up to approximately 10 wt% of the feedstock (Yan et al. 2010; Hoekman et al. 2011). These water-soluble organic fractions might potentially be separated as valuable by-products to further improve wet torrefaction economics.
The significant reduction in ash content of fuel made by wet biomass torrefaction suggests that the procedure can be employed in the production of “cleaner” biomass solid fuels as well. Regression analyses and numerical prediction showed that wet torrefaction can produce solid fuel with greater heating value, higher energy yield, and better hydrophobicity at much lower processing temperatures and holding times than dry torrefaction (Bach et al. 2013). Wet torrefaction leads to easier pelletization than dry torrefaction because wet torrefied biomass does not require water addition to improve the pelletability and binding capacity (Reza et al. 2012).
The yields and solid fuel quality obtained from wet torrefaction have been reported to be better than those from dry torrefaction. At 200 °C, for example, wet torrefaction of loblolly pine can give mass and energy yields that are as high as 88.7% and 95% respectively, compared to 83.8% and 89.7% respectively, for dry torrefaction at 250 °C with the same holding time (Yan et al. 2009). In addition to the solid fuel, some water, CO2, small amounts of CO, H2, hydrocarbons, and dissolved organic and inorganic compounds are released from biomass during wet torrefaction (Erlach et al. 2012). Another advantage of wet torrefaction is its ability to dissolve and extract inorganic components from solid biomass fuels. In spite of the numerous advantages of wet over dry torrefaction, relatively few studies on wet torrefaction have been reported in the literature, compared to an increasing number of recent studies on dry torrefaction.
OXIDATIVE TORREFACTION
Oxidative torrefaction involves the torrefaction of biomass in an oxidative environment. In this process, there is a reduction in operating costs as well as N2 consumption by employing air as the carrier gas. Wang and co-workers (2013) investigated the oxidative torrefaction of biomass residues and densification of torrefied sawdust to pellets. The properties of torrefied sawdust and its pellets, including density, energy consumption for pelletization, higher heating value, and energy yield in oxidative environments were similar to those of the biomass torrefied in inert atmospheres. The use of oxygen-laden combustion flue gases as carrier gases in torrefaction was beneficial, avoiding the need for inert gasses application and additional thermal energy input.
Chen and co-workers (2013) determined the reaction characteristics of biomass torrefaction in inert and oxidative atmospheres at various superficial velocities. The reaction was controlled by heat and mass transfer in biomass torrefied in nitrogen. However, for that torrefied in air, surface oxidation was the dominant mechanism in the torrefaction process. The surface oxidation intensified the internal heat and mass transfer rates when temperature and superficial velocity were raised, resulting in a significant drop in solid and energy yields.
OVERVIEW OF LABORATORY SCALE STUDIES ON TORREFACTION
Mass and Energy Balances of Lignocellulosic Biomass Wet Torrefaction
Many studies have examined wood/biomass torrefaction. Mass and energy balances are of significant importance for the economic design and optimization of torrefaction technology. A particularly interesting study was carried out on the wet torrefaction of loblolly pine in the temperature range of 200 to 260 ºC and at saturated vapor pressures of (225 to 680 psi) in a Parr reactor (Yan et al. 2010). Researchers reported that: a) gases accounted for 9 to 20% of the product and the quantity produced rose with increasing temperature, b) temperature also significantly affected mass yields and characteristics of the pretreated solid according to ultimate analyses and fuel-value measurements, c) organic acids were produced and accounted for 2 to 9% of the raw biomass, d) the quantity of precipitates dropped with increased temperature from 14% at 200 ºC to 9% at 260 ºC. The mass balances for wet torrefaction are shown in Table 2 for three temperatures.
Table 2. Mass Distributions in the Wet Torrefaction of Loblolly aPine (Yan et al. 2010)
a Reactants and products are given per the mass of dry wood. Uncertainty is shown in parentheses.
b Precipitate compositions in the aqueous product stream are summarized in Fig. 2. As the wet torrefaction temperature was raised, the overall quantity of monosaccharides found in the aqueous output stream decreased due to their conversion to 5- hydroxymethylfurfural (5-HMF).
Fig. 2. Composition of precipitates in the aqueous output stream from loblolly pine wet torrefaction at various temperatures. Each mass fraction is reported as a fraction of the dry biomass feed (Yan et al. 2010).
The enthalpy and heat of reaction of loblolly pine wet torrefaction were also determined. The heat of formation was accurately measured with a calorimetric bomb, while the heat of reaction was determined by the difference of the heats of formation of the products and reactants at each temperature (Yan et al. 2010). The magnitude of the heat of reaction was less than 2% of the heat of combustion for the untreated biomass. The reaction seemed to become less endothermic with an increase in temperature (Table 3).
Table 3. Enthalpy and Heat of Reaction in the Wet Torrefaction of Loblolly aPine (Yan et al. 2010).
a All data were obtained on the basis of 1 g of biomass feedstock. Uncertainty (not further specified in the source work) is shown in parentheses. The errors associated with each variable may play a significant role in determining whether the reaction is endothermic or exothermic because the estimated heat of reaction is relatively small (0.25 kJ) compared to the energy in the reactants to the reaction temperatures. It is therefore necessary to conduct uncertainty analysis of the heat of reaction calculations.
An uncertainty analysis was also performed for the heat of reaction estimation. It showed that the effect of temperature on the heat of reaction was not significant (Fig. 3).
 |
||
|---|---|---|
Fig. 3. Frequency distributions of the heats of reaction, during wet torrefaction of loblolly pine at three temperatures (Yan et al. 2010)
The effects of torrefaction on the chemical structures within torrefied wood derived from loblolly pine at different temperatures and times were examined (Ben and Ragauskas 2012). Solid-state cross-polarization/magic angle spinning (CP/MAS)13C, nuclear magnetic resonance (NMR) spectroscopy, and carbohydrate analysis were employed. The NMR results showed that the aryl-ether bonds in lignin were cleaved during torrefaction. The methyl carbons in hemicellulose acetyl groups were absent after torrefaction at 250 ºC for 4 h. The torrefied wood had a higher heating value (HHV) that was greater than the original wood feed. The HHV (20.16 MJ kg-1) of wood feed (dried at 75 ºC, 48 h) was far lower than that of the torrefied wood, which was increased by 60% (32.34 MJ kg-1) after torrefaction at 300 ºC for 4 h ( Table 4). This value is higher than for anthracite coal (31.84 MJ kg-1) and Pittsburgh seam coal (31.75 MJ kg-1), and much higher than Converse School-Sub C coal (21.67 MJ kg-1), German Braunkohle lignite (25.10 MJ kg-1), and Northumerland No. 81/2 Sem. Anth. Coal (24.73 MJ kg-1) (Channiwala and Parikh 2002).
With an increased wood torrefaction time from 0.25 to 8 h, at 250 ºC, the mass and energy yields decreased linearly from 94.97% to 64.36% and 99.79% to 79.12%, respectively (Ben and Ragauskas 2012). By contrast, the HHV increased from 21.22 to 24.78 MJ kg-1 (Table 4). The mass yields of torrefied wood samples decreased significantly upon raising the torrefaction temperature from 250 to 300 ºC. Less than 50 wt% of biomass remained after torrefaction at 300 ºC (Ben and Ragauskas 2012). This magnitude is similar to other literature reports (Deng et al. 2009; Pimchuai et al. 2010; Chen and Kuo 2011).
Table 4. Influence of Temperatures and Residence Times on the Mass Yield, HHV, Energy Densification Ratio and Energy Yield of Torrefied Loblolly Pine Wood (Ben and Ragauskas 2012)
a Mass yield = mass of dried torrefied wood/mass of dried wood × 100%. b Energy densification ratio = HHV of dried torrefied wood/HHV of dried wood. c Energy yield = mass yield × energy densification ratio. d The original loblolly pine wood sample was dried at 75 oC for 48 h before analysis of higher heating value.
Torrefied Wood/Biomass Gasification
Torrefied wood/biomass gasification has been extensively studied, especially the resulting gas composition and heating values. Syngas composition can be influenced by several process parameters including feedstock composition, particle size, and gasification conditions: mainly temperature, steam to biomass ratio, pressure, and gasification reactor design (Gil et al. 1997; Kandiyoti et al. 2006; Higman and van der Burgt 2008; Pereira et al. 2012; El-Emam et al. 2012).
The influence of pressure and biomass feed composition on gasification product yields and composition during fluidized bed O2/steam gasification was investigated (Berrueco et al. 2014). Two different biomass feedstocks: GROT (forest residues) and VW (virgin wood) were gasified. Three different torrefaction levels were applied: raw biomass, lightly torrefied (LT), and significantly torrefied (ST). A laboratory scale pressurized fluidized bed reactor was used. The main observed trend for both biomass feedstocks was that gas yield increased with increased pressure and torrefaction levels. Tar yield increased with the experimental pressure, and this occurred with a decrease in char yield. Also, raising the pressure shifted the gas composition towards higher CH4 and CO2 contents, while H2 and CO levels decreased. VW-derived materials (VW-LT, VW-ST) yielded higher levels of H2and CO and lower levels of CH4 than the corresponding forest residue (grot) feeds. As pressure and torrefaction level increased (more severe conditions), the differences between VW and forest residues became less relevant.
The gasification of wood pellets in a bubbling fluidized bed reactor at various temperatures, pressures and steam to biomass ratios (S/B) was reported (Mayerhofer et al. 2012). High temperatures (750 to 840 ºC) promoted H2 formation, while CH4 and CO2content decreased. Additionally, higher S/B ratios shifted the gas composition to higher H2 and CO2 concentrations and lower CO and CH4 contents in the gas produced. An increase in gasification pressure led to higher CH4 content, due to the enhancement of methanation at high pressures. Raising pressure also resulted in a slight increase in H2 content and lower CO/CO2 ratios.
Torrefied wood and conventional wood gasification were compared (Prins et al. 2006a). Untorrefied willow was compared to torrefied willow at both 250 °C for 30 min and 300 °C for 10 min, respectively. These reaction times excluded the heating times required to go from 200 °C to the reaction temperature at 8.5 min and 17 min, respectively. Torrefaction raised the lower heating value (LHV) of willow from 17.6 to 19.4 and 21.0 MJ/kg, respectively (Table 5).
Table 5. Composition of Untorrefied and Torrefied Willow (Prins et al. 2006a)
Air-blown gasification of these untorrefied and torrefied wood feeds were conducted both at 950 ºC in a circulating fluidized bed as well as at 1200 ºC during oxygen-blown gasification of torrefied wood in an entrained flow gasifier. Both gasification processes were run at atmospheric pressure. The overall exergetic efficiency of air-blown gasification of torrefied wood was lower than that of untorrefied wood because the volatiles produced in the torrefaction step were not utilized (Prins et al. 2006a).
The optimum gasification temperature of untorrefied wood is rather low (below 700 oC), and this feed is not an ideal fuel for gasifiers. Untorrefied wood becomes over-oxidized in gasifiers because of its high O/C ratio and moisture content. It also produces low optimum gasification yields, leading to thermodynamic losses. Considering gasification of wood at 950 ºC, there is a considerable amount of over-oxidation, which negatively influences the gasification efficiency. If wood is modified by torrefaction, its composition becomes more favorable so that it is over-oxidized less in the gasifier.
Figure 4 is a CHO illustration of untorrefied and torrefied wood produced at 250 and 300 ºC which was then gasified at 950 ºC (Prins et al. 2006a). The triangular CHO-diagram presents another reason for the increased gasification efficiency of torrefied wood. In this figure, the so-called carbon boundary line is shown at a temperature of 950 ºC. In order to avoid the formation of solid carbon, this line has to be crossed. Excess oxygen is added to achieve complete gasification on the carbon boundary line where a thermodynamic optimum exists. C, H, and O are present in all the gasified products where I, IIa, and IIb represent their mole percentages in gasified untorrefied wood as well as wood, torrefied at 250 and 300 ºC respectively.
Fig. 4. CHO diagram illustrating gasification of wood and torrefied wood (TW), where the torrefied wood was produced at 250 and 300 oC, then gasified at 950 oC (Prins et al. 2006a)
Gas yields and reaction kinetics during torrefied beech wood gasification were studied (Couhert et al. 2009). Beech wood was subjected to mild torrefaction (240 °C) and severe torrefaction (260 °C), using a specially designed crossed fixed bed reactor. A 2 s gasification at 1400 °C of the torrefied wood produced approximately the same quantities of CO2, 7% more H2, and 20% more CO than gasifying the untreated parent wood. Under these conditions, true equilibrium was reached. When gasification experiments were performed at a lower temperature, (1200 °C), the kinetics of torrefied wood gasification were comparable to that of the parent wood. However, the chars from torrefied wood were less reactive towards steam than the char from untreated wood.
Characterization of Torrefaction Products
Torrefaction products have been examined extensively (Felfli et al. 1998; Gaur and Reed 1998; Pach et al. 2002; Nimlos et al. 2003; Tumuluru et al. 2012; Hilten et al. 2013; Lin et al. 2013; Saleh et al. 2013; Keipi et al. 2014). The volatile species released during torrefaction of deciduous and coniferous wood at 270 ºC have been analyzed (Prins et al. 2006b). Deciduous xylan-containing wood (beech and willow) and straw are more reactive during torrefaction than coniferous wood (larch). The solid mass conserved in the torrefied deciduous wood ranged from 73 to 83% (depending on residence time) versus 90% for the coniferous wood. The difference in the volatile species released during torrefaction of deciduous versus coniferous wood originated from the difference in their hemicellulose structures. Deciduous wood contains mostly the acetoxy- and methoxy-substituted xylose units of their hemicelluloses. These groups generated more acetic acid and methanol volatile products than coniferous wood.
The effect of varying torrefaction temperature on spruce wood and bagasse from 260 to 300 °C in an auger reactor was investigated at a 10 min residence time (Chang et al. 2012). The treatment temperature and original biomass chemical composition significantly influenced the solid/liquid/gas product distributions (Fig. 5).
Fig. 5. Effect of torrefaction temperature on the distributions of biomass torrefaction product phases (Chang et al. 2012)
The concentration of carboxyl or carboxylic acid groups in torrefied spruce wood was determined, using both methylene blue sorption and potentiometric titration, after torrefaction for 30 min at temperatures of 180, 200, 220, 240, 260, and 300 °C (Shoulaifar et al. 2012). They also determined the equilibrium moisture content of the torrefied samples along with dehydration reactions. The degradation of carboxylic acids is a key reaction that reduces hydrophilicity of torrefied biomass. This occurs primarily by decarboxylation, which increases as torrefaction severity increases. The equilibrium moisture content also decreased with increase in torrefaction temperature and a drop in carboxyl content.
The influence of torrefaction on different biomass sources was investigated by (Arnsfeld et al. 2014). Torrefaction at about 300 °C lowered the amount of oxygen in biomass significantly. Furthermore, the values of the ultimate content and proximate analysis after torrefaction depended strongly on the biomass origin. Integrated versus external torrefaction via thermodynamic modeling were analyzed and compared (Clausen 2014). The biomass to syngas efficiency increased from 63% to 86% (LHV – dry) by switching from external to integrated torrefaction at 300 °C. Integrated torrefaction at 250 °C and gasification without torrefaction yielded biomass to syngas efficiencies of 81% and 76% respectively.
Fast Pyrolysis of Torrefied Biomass for Bio-oil Production
Torrefaction of corncobs was carried out as a pretreatment before fast pyrolysis to generate bio-oil. Corncob torrefaction was conducted in an auger reactor at 250, 275, and 300 °C, and at 10, 20, and 60 min residence times for each temperature (Zheng et al. 2013). These torrefied corncobs were then fast-pyrolyzed in a bubbling fluidized bed reactor at 470 °C to generate bio-oil. Using solid state 13C NMR and FTIR, the structural changes of the torrefied corncobs were probed before fast pyrolysis. Employing torrefaction prior to fast pyrolysis improved the quality of the resulting bio-oil. When torrefaction severity was elevated, the heating value of bio-oil was increased and its acidity was lowered due, in part, to a drop in water content. However, the bio-oil yield decreased significantly. The decrease in bio-oil yield likely resulted from the crosslinking and charring of corncobs during torrefaction. Such pretreatment changes would require more fragmentation to occur when generating bio-oil in the fast pyrolysis step. This slows vaporization allows more time for solid phase condensation reactions to advance, producing more char.
Figure 6 illustrates, in a simple scheme, the effect of torrefaction on the subsequent fast pyrolysis mechanism of cellulose. The left side of Fig. 6 illustrates the thermal fragmentation of raw cellulose during fast pyrolysis. Fragmentation to lower weight polymers and oligomers of glucose occurs, while dehydration-cyclization simultaneously occurs to give anhydrosaccharides. Where fragmentation occurs all the way to the monosaccharide level, levoglucosan is produced. Further fragmentation to 5-hydroxy-methylfurfural, hydroxyacetone, hydroxyacetaldehyde and other products occurs. Competing with decomposition to vaporizable molecules, larger cellulose fragments both partially dehydrate and crosslink. These larger solid phase species are unable to vaporize and instead, dehydrate and partially carbonize to char. On the right side of Fig. 6, the dehydration process is underway during the lower temperatures of torrefaction. Sufficient thermal energy is not available to cause extensive fragmentation to lower molecular weight oligomers, glucose, levoglucosan, and further decomposition products. Thus, bio-oil is not produced but some of the cellulose begins to crosslink and condense. After torrefaction, the 425 to 500 ºC temperatures used in fast pyrolysis generate smaller amounts of volatiles as further crosslinking and char formation occur instead.
Fig. 6. Effects of torrefaction on the fast pyrolytic decomposition of cellulose (Chaiwat et al. 2008). Figures used by permission of copyright holder.
KINETICS OF TORREFACTION
Several attempts at correlating torrefaction properties with process conditions have been made in order to gain insight into this process. Most of the current torrefaction studies focus on the biomass property changes in batch-scale reactors. The degree of torrefaction is calculated based on the measured weight loss (Arias et al. 2008; Shang et al. 2012). In large-scale production facilities, torrefaction is usually performed continuously within a closed collector. This creates an inert atmosphere that makes process control more challenging. In order to improve the process control in continuous torrefaction reactors, mathematical models that accurately describe the torrefaction reaction under different heating rates need to be developed.
The weight loss kinetics of deciduous and coniferous wood types was studied (Prins et al. 2006c). The kinetics of torrefaction reactions in the temperature range of 230 to 300 °C were described accurately by a two-step mechanism, with the first step being much faster than the second step. The first step was hemicellulose decomposition, while the second step represented cellulose decomposition. The solid yields for the first step were higher by 70 to 88% than for the second step.
Linear regression mass loss was used to predict changes in fixed carbon (and thereby volatile matter) as well as the gross calorific value (and thereby energy yield) for three species of eucalyptus wood and bark (Almeida et al. 2010). Temperature and feedstock moisture were correlated as independent variables to predict mass loss and energy yield with a quadratic surface methodology for corn stover (Medic et al. 2012). A response surface methodology was used to correlate torrefaction severity (time and temperature) with weight loss, energy value, and energy yield for mixed softwoods (Lee et al. 2012). All three empirical data correlations performed well for their respective feedstocks. However, these empirical data did not offer a prediction for the behavior expected for additional feedstocks. These studies were purely empirical and not based on any fundamental understanding. Nonetheless, it is necessary for new empirical data to be modeled to adjust correlations for additional feedstock use.
Combustion kinetic studies of dry torrefied woody biomass materials using multiple pseudo-component models have been reported (Brostroem et al. 2012; Tapasvi et al. 2013). Brostroem et al. used a global kinetic model, while Tapasvi et al. employed a distributed activation energy model. Both studies showed that the degree of feed torrefaction had little effect on the combustion kinetic parameters of the torrefied biomass regardless of the torrefaction conditions. However, Brostroem et al. reported that hemicellulose, cellulose, and lignin activation energy values were constant at 100.6, 213.1 and 121.3 kJ/mol, respectively. This was true for both raw and dry-torrefied spruce. Tapasvi et al. reported that the activation energy values for the cellulose, non-cellulosic fractions, and char remained constant at 135, 160, and 153 kJ/mol respectively, for various types of feedstock and their degree of torrefaction.
A torrefaction model for wood chips in a pilot-scale continuous reactor with a two-step series, first-order reaction model was developed to study the two-step kinetics of wood chip torrefaction in a TGA setup (Shang et al. 2014). The first step was much faster than the second step. This study took into account the mass loss during the heating period in calculating the kinetic parameters, unlike other studies, that were based on kinetic parameters obtained from the isothermal part of torrefaction. These other approaches neglected sample degradation during the heating period. Shang et al.’s model was useful in predicting the HHV of wood chips torrefied in a continuous pilot scale reactor.
A simple first-order kinetic model was applied to estimate the activation energy and pre-exponential factor of both raw and dry-torrefied eucalyptus samples in a two-stage combustion process (devolatilization followed by combustion) (Arias et al. 2008). Both the activation energy and pre-exponential factor increased in stage 1 and decreased in stage 2 after dry torrefaction. Nonetheless, the model was based on an empirical method, which was not validated because the model itself could neither reproduce simulated curves nor give any information about the fit quality between the predicted and experimental data.
The fuel properties of typical Norwegian birch (hardwood) and spruce (softwood) were assessed after both dry and wet torrefaction (Bach et al. 2014). TGA experiments were employed. The thermal degradation kinetics under dry torrefaction conditions were investigated and the torrefaction kinetic parameters were determined. A two-step kinetic model was employed to simulate the recorded mass loss curves. In the first step, decomposition of the initial biomass to form an intermediate solid and volatiles exhibited a higher rate than the second step.
The determination of kinetic constants is often difficult. Kinetics derived from TGA experiments involve conditions that are dramatically different from those in torrefaction reactors. The TGA temperature ramp rates are very slow, and the particle size of the samples is much smaller (Narayan and Antal 1996). Most fast pyrolysis reactors operate with a fixed heat source temperature, whereas the actual temperatures reached by the reacting samples are higher than those calculated for TGA studies. If several elementary reaction processes with different activation energies are involved, then the controlling chemical processes in TGA studies may differ markedly from torrefaction processes (Lede and Villermaux 1993; Lede 1996, 2010; Lede and Authier 2011).
Recent kinetic and mechanistic literature on thermal reactions, particularly the use of thermal analysis (TA), was evaluated (Galwey 2004). Major problems with kinetic studies based on thermal analysis experiments were described. Ambiguities exist in the definition of the essential terms (“mechanism”, “rate constant,” and “activation energy”). Also, a lack of order in the results was identified. The lack of physical meaning for the calculated Arrhenius parameters was noted, and the impossibility of finding a real kinetic mechanism from thermal analysis data alone was emphasized. Galwey concluded that supplementary tests (X-ray analysis, microscopy, etc.) are required.
The main bottleneck in using kinetic models expressed with calculated Arrhenius parameters to determine thermal decomposition pathways remains unsolved. “If we are to interpret them in terms of the transition-state theory, they are not applicable to solid state reactions” (Vyazovkin and Wight 1997). Because of the stable and tightly packed array-structure of samples in solid state reactions, the Maxwell-Boltzmann’s energy distribution functions are not suitable. However, other energy distribution functions such as the Fermi-Dirac function for electrons and the Bose-Einstein function for photons can be applied (Vyazovkin and Wight 1997). In this case, Ea refers to the activation enthalpy and Arefers to the frequency of lattice vibrations. Another major issue with these approaches is the empirical nature of the kinetic models tested (White et al. 2011).
In order to identify realistic solids’ degradation mechanisms, further investigations are required because thermal changes (including chemical changes) are often more complex than is recognized. The origins of modern thermal analysis kinetics are located in a specialized branch of chemistry that is concerned with the thermal decomposition of solids, known as “crystolysis reactions” (Galwey 2004). The theory applied in cases that were evaluated was based on geometric models that are applicable to heterogeneous reactions in crystals, where the stoichiometries were regarded as already well established. For solid state thermal degradation kinetic studies employing thermal analysis methods, the calculated values of the Arrhenius parameters describe a given step of the process in a general manner. Arrhenius parameter values have a different physical meaning for reactions in crystals than for gas or liquid phase transformations.
The torrefaction technologies available today are basically designed and tested for biomass. Further research in the area of kinetic modelling for large scale reactor design and also for the optimization of product characteristics is absolutely necessary. The choice of torrefaction technology is exceptionally difficult because of the absence of practical comparative assessment of different types of reactors.
TORREFACTION TECHNOLOGIES
Most torrefaction technologies now being developed are based on already existing reactor concepts designed for other purposes such as drying or pyrolysis. These technologies are being modified for torrefaction. The reactors being developed in most cases are established technologies that developers are familiar with. They are simply being optimized for torrefaction applications. Some torrefaction technologies are capable of processing feedstock with small particles such as sawdust, while others are capable of processing large particles. Only a few can handle a large spectrum of particle sizes.
Many torrefaction technology developers are companies with extensive backgrounds in biomass processing and conversion technologies including carbonization and drying. This is an indication that technology selection needs to be based on feedstock characteristics. Alternatively, the feedstock needs to be pre-processed before torrefaction, using size reduction equipment such as scalpers for handling over-sized material or sieves for extraction of particles of smaller materials. These considerations all influence the capital and operating costs of a torrefaction plant. All of these technologies have their advantages and disadvantages. No single technique is fundamentally superior to the other. Since each reactor has unique characteristics and is well suited to handle specific types of biomass, proper reactor selection is important for specific biomass properties and application.
Table 6. Torrefaction Reactor Technologies and the Companies that have Developed them
Key: AT-Austria, BE-Belgium, CA-Canada, CH-Switzerland, FR-France, IT-Italy, NL-Netherlands, GER-Germany, S-Spain, SWE-Sweden, UK-United Kingdom, US-United States of America.
Sources: Kleinschmidt 2010; Beekes and Cremers 2012b; Kleinschmidt 2011; Nordin 2012; Melin 2011; Torrefaction of biomass 2012; and the respective company websites.
Current torrefaction technology can be categorized into two groups based on heating: 1) methods in which the biomass is heated directly, and 2) those in which it is heated indirectly. Those employing direct heating utilize a non-oxygen gas loop with a heat exchanger in a moving bed, drum, vibrating belt, or multiple hearth furnaces. Direct heating can also employ a low-oxygen gas loop linked to a burner in a tunnel, moving bed, or a torbed. Processes that are heated indirectly use an auger or a drum. Of these eight technology types, only a few are able to produce more than five tons/h of torrefied wood/biomass. Table 6 provides an overview of the eight most important torrefaction reaction technologies and the companies that have developed them. The following eight reactors are briefly described.
Fixed Bed Reactor
In a fixed bed down-flow reactor, the raw solid biomass is fed to the top of the reactor. The biomass undergoes drying and torrefaction, exiting at the bottom of the reactor. Neutral (oxygen free) hot gases enter the bottom of the column and travel upwards. The loaded gases exit at the top of the reactor. A condenser extracts water vapor and other condensable substances from the gas.
Fig. 7a. Fixed bed reactor (http://www.autoclaveengineers.com)
Fig. 7b. Schematics of a fixed bed reactor
The dry gas is combusted in a burner to generate hot gases for recirculation through the reactor. Excess gas is filtered and released to the environment during operation. In a fixed bed reactor, the biomass particles remain stationary (Dhungana et al. 2012).
Rotary Drum Reactor
The rotary drum is a proven technology for various applications (Koppejan et al. 2012). The rotary drum is heated directly or indirectly, using a superheated stream of flue gas resulting from the combustion of volatiles. The rotary drum of an indirectly heated reactor tumbles the biomass in an inert gaseous environment where heat is transferred from the hot drum wall to the biomass particles. In torrefaction applications, the process can be controlled by varying the torrefaction temperature, the rotational velocity, and the length and angle of the drum. If the rotational velocity is low, biomass will be carbonized instead of being torrefied. However, if the rotational velocity is too high, biomass will have low product quality because of partial torrefaction (Beekes and Cremers 2012b). As the drum rotates, biomass in the bed mixes homogeneously and heat is distributed between biomass particles. Because the scalability of the rotary drum is limited, a modular setup is required for higher capacities.
Fig. 8a. Rotary drum reactor (Thamer 2013)
Fig. 8b. Schematics of a rotary drum reactor
Screw Reactor
The screw torrefaction reactor is stationary. It could be vertical, horizontal, or inclined with a circular or rectangular cross-section. In this reactor, a rotating screw churns and moves the biomass through the reactor to enhance heat transfer between the wall and the bulk of the biomass. It also moves the biomass along its length at the same time (Dhungana et al. 2012). A screw reactor is often heated indirectly, using a heat transfer medium inside the hollow wall. Since the hot outer wall heats the biomass indirectly, there is no direct contact with an oxygen-laden heating medium. Some designs may have holes in the shaft for the volatiles to escape. Biomass may also be heated by applying heat to the screw itself viacirculating hot fluid within its hollow structure. For torrefied biomass to achieve a good product quality, biomass with very low bulk density and high moisture content requires pre-treatment before feeding into the screw reactor (Beekes and Cremers 2012b).
Fig. 9a. Screw reactor (www.syncoal.com)
Fig. 9b. Schematics of a screw reactor
Multiple Hearth Furnace (MHF)
The multiple hearth furnace consists of a series of circular hearths placed one above the other and enclosed in a refractory-lined steel shell. Hot gases are input to the hearth to heat the biomass during torrefaction. Biomass is stirred and moved in a spiral path across each hearth by a vertical rotating shaft that goes through the center of the furnace, carrying arms with rabble blades (Lonardi et al. 2008). Biomass is fed to the top hearth, passing across it and then, through drop holes to the hearth below. Biomass passes continuously over and across each hearth to the bottom of the furnace, where the product is discharged through one or more ports.
Fig. 10a. Multiple hearth furnace (fgcgroupllc.com)
Fig. 10b. Schematics of a multiple hearth furnace
In some operations, combustion of charge-elements supplies the heat. In others, heat is provided by the combustion of auxiliary fuel in burners on certain hearths (direct firing), or in a separate combustion chamber (indirect firing) where the hot gases are fed by counter current flow from bottom to top of the furnace (fgcgroupllc.com).
A multiple hearth furnace offers operational flexibility, permitting a wide range of processing options and allowing for many structural variations to accommodate special operations. The residence time and temperature of the desired treatment can be closely controlled and also varied within wide limits. Fresh biomass can be loaded from the top or at any other hearth. Heat supply can be directed to the most effective area. Gases can escape from the top, bottom, and intermediate hearths, or any combination of these. The atmosphere can be oxidizing, reducing or neutral and can be varied in different parts of the furnace.
Torbed (Rotating Fluidized Bed) Reactor
In a torbed reactor, heat is carried via a gaseous medium that is blown in from the bottom of the bed with a high velocity (50 to 80 m/s) past stationary angled blades. The biomass particles inside the reactor acquire vertical and horizontal movements, resulting in toroidal swirls that rapidly heat the biomass particles in the reactor. This improves the heat and mass transfer among solids and gases, resulting in lower retention times and the production of a more homogeneous product (Beekes and Cremers 2012b). The relatively intense heat transfer enables torrefaction to be completed within a short residence time (around 80 s).
Intense heat transfer can also be used to operate the reactor in a controlled manner at elevated temperatures (up to 380 ºC), resulting in a higher loss of volatiles from the biomass feed.
Fig. 11a. Torbed reactor (www.torftech.com)
Fig. 11b. Schematics of a torbed reactor
Microwave Reactor
The microwave reactor employs microwave irradiation with electromagnetic waves, which can operate at a specific frequency in the range from 300 MHz to 300 GHz (Dhungana et al. 2012). The electric component of the electromagnetic microwave radiation causes heating by two main mechanisms: dipolar polarization and ionic conduction. Typical industrial microwave reactors usually operate at 2.45 GHz. The microwave irradiation produces efficient internal heating of biomass particles. The heating depends on the ability of the materials being heated to absorb the impinged microwaves and convert them into heat. The microwave reactor is different from the other reactors that utilize direct heating. In direct heating systems, by contrast, biomass is heated externally, such that heat from the heat carrier (gas, solid, liquid, or reactor wall) first contacts the surface of the biomass particles and moves by conduction into their center. This mode of heating is less efficient because biomass is a poor thermal conductor. Microwave heating occurs simultaneously throughout the sample.
Fig. 12. Microwave reactor (Thamer 2013)
Moving Bed Reactor
The moving bed reactor consists of an enclosed reactor vessel with the reactor length usually determined by the retention time that is required to produce the desired product. In torrefaction, the time required is 25 to 30 min (Beekes and Cremers 2012b). Biomass enters from the top of the vessel and moves downward gradually while torrefaction takes place. The heat for torrefaction is carried by a gaseous medium that enters from the bottom and flows counter to biomass to the top of the reactor. The torrefied product leaves the reactor from the bottom and is then cooled. Gaseous reaction products (volatiles) are removed from the top of the reactor.