Category Archives: Handbook of biofuels production

Biofuel-driven biorefineries for the co-production of transportation fuels and added-value products

R. VAN REE, J. SANDERS, R. BAKKER and R. BLAAUW, Wageningen University and Research Centre (WUR), The Netherlands and R. ZWART and B. VAN DER DRIFT, Energy Research Centre of The Netherlands (ECN), The Netherlands

Abstract: This chapter discusses the importance of biorefining for optimal valorisation of biomass in a sustainable way. The focus is on conventional and advanced biofuel-driven biorefineries co-producing transportation fuels and added-value products from biomass. A definition for biorefining is presented. A classification system, the current status and future trends of biorefineries are discussed.

Key words: biorefining, definition, biorefineries, classification, biofuels, biofuel-driven biorefineries, biomass value chains.

21.1 Introduction

21.1.1 Biofuel production processes

The production of biofuels for transport, as alternative for the conventional crude oil derived transportation fuels gasoline and diesel, has gained a lot of interest in the last ten to 15 years, as a result of both the general approach to become less dependent on politically unstable countries and the concern about the consequences of anthropogenic CO2 related global warming.

Both conventional biofuels (biofuels produced from crops that also could be used for food and feed production) and advanced biofuels (biofuels produced from non-food and non-feed crops) can be distinguished. Examples of advanced biofuels are: biochemically produced cellulosic ethanol, butanol and hydrogen, and thermochemically and catalytically produced Fischer-Tropsch diesel, methanol, dimethylether (DME), hydrogen, and synthetic natural gas (SNG). Also catalytically produced sugar derived furanics are examples of advanced biofuels for transport.

Combustion visualisation

Combustion bomb study

The study on the spray combustion characteristics of 10% CPO blended with diesel fuel was conducted in a constant volume combustion chamber. With the fixed experimental conditions such as spray ambient pressure and injection events, the effects of 10% CPO diesel at the injection line pressure of 100 MPa on spray combustion and flame structure were investigated using photo diode and ICCD camera. The two-colour method was also employed to predict combustion flame temperatures and KL factors.

The injection system used in this research was an electronically controlled accumulator type fuel injector system.13,14 With a 0.2 mm diameter single hole injector, driven by a piezo electric actuator via an extended pressure pin, we could control the needle lift and fuel injection rate shaping. The schematic diagram of the injector and details are shown in Fig. 23.1.

Experiments were conducted in a constant volume 2.2 litre vessel with 80 mm diameter quartz observation window on the side, gas mixing propeller on the bottom and injector on the top, as shown in Fig. 23.2. The ambient conditions maintained inside the vessel were high temperature and pressure by igniting hydrogen in an enriched oxygen and air mixture. The oxygen concentration after the hydrogen combustion was approximately 21% by volume.13,14

The rectangular injection rate shaping was obtained in this experiment, as shown in Fig. 23.3. Fuel injection mass was set at approximately 15 mg for all experiments. Injection pressure was 100 MPa. The fuel was injected in the vessel at the ambient conditions of 3.0 MPa, temperature around 900°C, as shown in Fig. 23.4. The calculated composition of ambient gas was O2 20.9%, N2 70.8% and H2O 8.3%.

image167

/Photo diode

Подпись:Подпись:Подпись: Photo diode S (to detect start of injection) Подпись: P.F. (310.3 nm)Подпись:image168* (to detect luminous emission) P. M.T.

(To measure OH radical photoemission in flame)

image271

23.2 Experimental apparatus.

After the hydrogen combustion, the fuel was injected into the vessel and then combusted. Fuel spray combustion flame photographs were taken by ICCD camera. Light emission of flame was measured using two photo sensors: a photo multiplier tube with a band-pass filter centres on a wavelength of 310.3 nm (FWHM: 16.3 nm) was used for measuring the intensity of OH radical emission and two photo diodes (used for measuring the luminous light intensity) at the upper and the middle of observation window. The start of spray was detected by the combination of the use of He-Ne laser with photo sensor. Using photo diode data, then the ignition delay and combustion period were evaluated.

image171

23.4 Temporal variation of gas pressure inside the vessel.15

The two-color method was applied to estimate two dimension (2D) contour of temperature and KL factor (KL factor is the factor used to indicate soot) distribution in the combustion flame. This two-colour pyrometry system was set up by placing Vari lens that has the two-different band-pass filters 488 nm in centre wavelength (FWHM: 11.3 nm) and 634 nm in centre wavelength (FWHM: 8.5 nm) for separating image to be two in front of an ICCD camera lens. The intensity data of both filters were used to calculate the true temperature and KL factor.

The data obtained from He-Ne laser and OH-radical were used to calculate ignition delay. It was found that 10% CPO diesel gave shorter ignition delay compared with diesel as shown in Fig. 23.5.

The data 10% of peak intensities obtained from the two photo diodes were selected to be the start and end of the combustion. The result shows that the observed combustion period of 10% CPO diesel at injection pressure of 100 MPa was slightly shorter than diesel as shown in Fig. 23.6.

The amount of injection fuel became slightly smaller and the injection period became slightly shorter with the 10% CPO diesel due to the higher viscosity of 10% CPO diesel.

The exposure time of ICCD camera was set at 10 psec.15 The spray combustion flame intensity data complied with two colour method.16 Some of the calculated results of true temperature are shown in Fig. 23.7.

The calculated data obtained from spray combustion flame true temperature were used for calculating the KL factor, the factor for indicating amount of combustion soot in flame. The calculated results are shown in Fig. 23.8.

о

ID

in

E

ID

Q.

О

О

————————- 1——————————

Diesel 10% CPO diesel

23.6 Fuel spray combustion period with injection pressure of 100 MPa at ambient conditions of 3 MPa.15

Подпись: 23.5 Fuel combustion ignition delay.15

Total KL factor is the summation of KL factor over the spray combustion flame area. This factor could be used to estimate the total soot of the combustion.

It was found, as shown in Fig. 23.9, that the difference in total KL factor between diesel and 10% CPO was very small.

The average KL factor, shown in Fig. 23.10, was calculated from the total of KL factor divided by spray combustion flame area at all flame area. This factor could be used to estimate the soot concentration of the spray combustion. The

image173

23.7 Spray combustion flame temperature distribution.

image174

23.8 KL factor distribution.

image175

23.9 Total KL factor of palm diesel 60% and diesel fuel at an injection pressure of 100 MPa and 60 MPa.

image176

Time after start of injection (msec)

23.10 Average KL factor of palm diesel 60% and diesel fuel at an injection pressure of 100 MPa and 60 MPa.

results have shown that the difference of average KL factor between diesel and 10% CPO was very small.

Histograms of temperature and KL factor were calculated by evaluating the value from the counted number of spray combustion flame pixel and converting them to flame area (mm2). The interval of temperature and KL factor were selected at 50 K and at 0.005 A. U., respectively. The results are shown in Fig. 23.11.

image177

23.11 Flame temperature and KL factor histogram of palm diesel 60% and diesel fuel at an injection pressure of 100 MPa and 60 MPa.

It was found from temperature histogram that spray combustion of 10% CPO started with lower temperature than diesel. Spray combustion temperature had increased close to diesel during the mid range of combustion period. Then, it became lower by the end of combustion. However, the differences were very small.

The KL factor histogram of Thai palm 10% CPO had no significant difference compared to diesel. Hence, it could be concluded that the difference in soot emission could be very small.

The effects of 10% CPO diesel at an injection pressure of 100 MPa on spray combustion and flame structure were investigated. It was found that diesel blending with 10% CPO has shorter ignition delay and shorter combustion period compared with conventional diesel fuel. High temperature combustion area (over 2400 K) of 10% CPO diesel was also smaller than diesel, especially at the end of the combustion period. The amount of soot and soot concentration affected by this blending percentage may not be significantly different from the diesel fuel.

Catalytic pyrolysis for improved pyrolysis-oil generation

The introduction of catalysts into pyrolysis processes introduces additional costs. Islam and Ani have estimated that the cost of catalytic pyrolysis process (in terms of dollar per joule of energy produced from the oil) might be around twice that of a pyrolysis only process although this would decrease with increased scale of production.79 The main reasons for the higher costs are associated with the catalyst and the additional capital and human infrastructure requirements. However, it does appear that costs of less than 1€/l may be achievable118 and this could ensure process viability as an alternative to petroleum. These and similar cost analysis do point to an important critical step in the introduction of this technology, the need to develop catalysts that demonstrate good performance and extended lifetimes or simple catalyst regenerative processes. This was also a key finding of a report by the NREL (National Renewable Energy Laboratory) in the USA.119 This report provides an excellent summary of the research up to 2003 and there are other general reviews of catalytic pyrolysis which are useful including work by Bridgwater,47 Elliott,120 Sharma and Bakshi121 and Chen et al.44

Although projected costs are higher there are a number of clear advantages in catalytic pyrolysis over conventional pyrolysis. These are:

1 The products have lower oxygen content and higher H:C ratios than simple pyrolysis-oil.122 This is an important indicator of quality; the H:C ratio should be as low as possible for fuel use and an effective ratio is often defined:123

(H/C)eff = (H-X)/C

where H and C are the number of hydrogen and carbon atoms in the product respectively. If X is the relative (to the value of C) number of hydrogen atoms in the original feedstock, then the effective ratio indicates the ‘upgrading’ of the catalytic pyrolysis product.

2 The product distribution is narrowed, i. e. the produced oil has a narrow molecular weight range consistent with use as a fuel oil with better yields in the gasoline (C5-C12 range).124,125 The catalytic action allows some control over the product distribution and by correct choice of catalyst and process conditions products of increased value can be increased in concentration.126

3 Catalytic pyrolysis increases the amount of aromatics and branching in the pyrolysis oil products compared to conventional pyrolysis methods.122,127-129 This is extremely important to the potential use of pyrolysis oil as a fuel in engines and turbines because the presence of aromatics and highly branched hydrocarbons increases the smooth running of the engine or turbine. By use of catalytic techniques the aromatic content can be increased to 30-50% with the major products being naphthalenes and toluene but there are significant amounts of benzene, indanes and substituted benzenes.

4 The pyrolysis oil product from a catalysed process (rather than non-catalysed process) is deoxygenated significantly lowering the average oxygen to carbon ratio across the product distribution.46,130-132 This is important for two main reasons.133-134 Firstly, simple, non-upgraded pyrolysis fuels have low energy content which makes them unsuitable for direct use in established liquid hydrocarbon fuel technologies. Secondly, non-upgraded pyrolysis oils are highly acidic (as above) because of the oxygen content (around 40-50%) and removing oxygen significantly reduces the acidic components and decreases the corrosive nature of the product.

5 Catalytic pyrolysis also decreases the char content of pyrolysis products135 whilst increasing the contribution of gases.136,137 Reduction of char is extremely important as the presence of char can catalyse inter-molecular reactions during storage and increase product viscosity.138

6 The use of catalysts in a pyrolysis reaction can significantly lower the reaction temperature.139,140 This is extremely important because pyrolysis reactions are endothermic and add significant costs to the overall process. Unlike most catalysis processes which are exothermic and can recover energy for any heating required, there is a true cost of running pyrolysis technology and it is, therefore, important to decrease the energy required as much as possible.

Equilibrium method

The equilibrium approach is further sub-divided into two models:

1 Stoichiometric model

2 Non-stoichiometric model.

The Stoichiometric model is based on an equilibrium constant. This method requires knowledge of the specific chemical reactions and reaction paths used for the calculation. It means selecting appropriate chemical reactions, and information concerning the value of the equilibrium constant is required. This method, therefore, is not suitable for complex problems where the chemical formulas of the feed or the reaction equations are not well known. This requires the second model, involving minimization of Gibbs free energy (non-stoichiometric model), which is an effective tool to find composition of gases when the reaction paths are unknown (Florin and Harris, 2008). It is a little more complex but advantageous, as a detailed knowledge of the chemical reaction is not needed.

The following section presents a brief discussion on stoichiometric and non­stoichiometric models of computation for the equilibrium concentration of the product gas. It also discusses energy balance, which is essential for an auto­thermal gasification reaction.

Hydrocracking of BTL-FT wax to diesel

Although different options have been proposed for the post-treatment and upgrading of the FT waxes (Dancuart et al., 2003; de Klerk, 2007; Dupain et al., 2005), it is generally accepted that hydrocracking is the most effective route to maximize the middle distillate yield and it is currently the applied option. Given the small number of commercial FT plants, little technology has been developed specifically for the refining of the FT wax products. In most commercial sites, standard crude oil refining approaches have been used without taking into account the specific characteristics of the FT wax product compared to conventional refinery streams, such as extra low aromatics content (< 1 wt.%) and virtually zero sulphur (< 5 ppm) (see Table 19.3).

Conventional hydrocracking takes places over a bifunctional catalyst with acid sites to provide isomerization/cracking function and metal sites to provide hydrogenation-dehydrogenation function. Platinum, palladium or bimetallic systems (i. e. NiMo, NiW and CoMo in the sulfided form), supported on oxidic supports (e. g. silica-aluminas and zeolites), are the most commonly used catalysts, operating at high pressures, typically over 10 MPa, and temperatures above 350°C.

In recent years, considerable research is ongoing to investigate the effect of the operating conditions, both experimentally (Calemma et al., 2005, 2010; Rossetti et al, 2009) and computationally (Fernandes and Teles, 2007; Pellegrini et al., 2004), and the catalytic material on the yield and quality of the FT wax hydrocracking products. Concerning the operating conditions, it was found that wax hydrocracking requires milder pressure and temperature, as the paraffinic nature of the wax implies higher availability of hydrogen in the unit (little hydrogen consuming aromatics) and thus suppressed coke formation (de Klerk, 2008). FT wax hydrocracking to middle distillates is favoured at pressures ranging from 3 to 5 MPa and temperatures between 250°C and 300°C (Calemma et al., 2010) and yields a product containing light paraffins up to C24, as presented in a product sample chromatograph obtained from FT wax hydrocracking experiments performed in Chemical Process Engineering Research Institute (CPERI) (Fig. 19.7). At these conditions, middle distillate yield (C10-C22) reaches up to 80-85 wt.% at intermediate conversion levels (~60 wt.%) (Calemma et al, 2010). At higher conversions, a small reduction in the middle distillate yield can be observed, indicating an increase of consecutive hydrocracking reactions leading to lighter products. Still, the consecutive reactions are limited, allowing the reaction to be carried out at high conversions without lowering significantly the middle distillate selectivity (Calemma et al., 2010).

Extensive work has also been conducted by our group as part of the EU-funded IP RENEW project that explored technology routes for the production of BTL fuels (Lappas et al, 2004). More specifically, the operating conditions (temperature, pressure, H2/oil ratio) were investigated in experiments with different commercial hydrocracking catalysts in a specially designed hydroprocessing pilot plant unit. Main conclusions were that with all catalysts, hydrocracking temperature appears to play the most important role and influences significantly the product yields, as shown in Fig. 19.8. It was shown that the yields of naphtha and kerosene in the product increase as the temperature increases and so does the conversion.

image233

19.8 Effect of temperature on product yields in the hydrocracking of BTL-FT wax.

However, the diesel yield is maximized at a certain temperature and then decreases as a result of higher conversions achieved at higher temperatures (RENEW, 2008). Moreover, it was shown that the yield of gasoline and diesel in the product decreases as the H2/oil ratio decreases and so does the conversion. The diesel selectivity is also slightly decreased as a result of the decreasing yield and

conversion. Studies by Calemma et al. (2010) showed additionally that the composition of FT diesel, specifically the ratio of iso — and n-paraffins, is also influenced by the operating parameters.

The nature of the catalyst also affects significantly the product quality and yield. Experiments performed in CPERI with three different commercial hydrocracking catalysts showed measurable differences in diesel selectivity at isoconversion as a function of the catalytic material (Fig. 19.9) (RENEW, 2008). Catalysts loaded with a noble metal (particularly Pt) were reported to show better performances in terms of selectivity for hydroisomerization and products distribution in comparison with non-noble metals-based catalyst (Archibald et al., 1960; Gibson et al, 1960). Calemma et al. (2001) reported high diesel selectivities obtained over a Pt/SiO2-Al2O3 catalyst during hydroprocessing of FT waxes and attributed the observed results to the mild Bronsted acidity, high surface area and pore size distribution of the support. Zhang et al. (2001) also showed that Pt performs better than Ni and Pd supported on tungstated zirconia for the hydroisomerization of the model compound n-hexadecane. The use of hybrid catalysts based on Pt/WO3/ZrO2 with addition of sulphated zirconia, tungstated zirconia or mordenite zeolites was studied by Zhou et al. (2003). According to the authors, hybrid catalysts based on Pt/WO3/ZrO2 provide a promising way to obtain higher activity and selectivity for transportation fuels from FT products. Given the high cost of noble metals, hydroprocessing of FT waxes has also been

* 40

3 30 «

Diesel

.9 Product selectivity at isoconversion for different catalytic materials in the hydrocracking of BTL-FT wax.

studied over nickel catalysts (de Haan et al., 2007). de Haan et al. (2007) demonstrated the benefit of using non-sulfided nickel catalysts. In conventional hydroprocessing units, catalysts are sulphated to avoid poisoning by the sulphur species in crude oil. However, in the case of the sulphur-free FT waxes, use of a sulfided catalyst implies the continuous addition of sulphur-containing compounds to avoid catalyst deactivation (de Klerk, 2008). Other advantages of developing a non-sulfided catalyst for the hydrocracking of FT waxes are a simplified, less costly and environmentally friendly process (no H2S in the tail gas) (de Haan et al., 2007). Nickel supported on a commercial silicated alumina yielded results that compare favourably with those of a commercial sulfided NiMo catalyst, with diesel selectivities of 73-77% at a conversion of approximately 52% (de Haan et al., 2007).

Biofuel-driven biorefinery

Both biomass-derived syngas (CO and H2) and SNG can be integrated into conventional existing petrochemical refinery complexes to produce both transportation fuels and chemicals from biomass. Also biochemically produced ethanol, butanol and hydrogen potentially can be used in the same existing refinery infrastructure to produce a variety of bio-based chemicals and materials.

An interesting technology for the production of the energy dense biomass — derived intermediate bio-oil is fast pyrolysis (thermal degradation in the absence of oxygen). Currently, a lot of effort is being put into the (catalytic) hydrogenation of this material to make it suitable to produce biomass-derived fuel additives. Another development is the development of catalytic fast pyrolysis processes for the production of bio-based chemicals.

The synergistic combination of aquathermolysis (hot pressurised water treatment) and fast pyrolysis is a promising thermolysis option integrating fractionation of biomass with the production of valuable chemicals (de Wild et al, 2009). Aquathermolysis causes hemicellulose to degrade and disappear from the raw materials. Lignin ether bonds are broken, but the lignin is hardly affected. Cellulose is also retained and seems to become more crystalline (see Fig. 21.6).

Catalytic cracking of highly oxygenated biomass-derived feedstocks

15.2.1 Catalytic cracking of bio-oils

Bio-oil is a chemically complex mixture of more than 300 oxygenated compounds, the main constituents being acids, aldehydes, ketones, alcohols, glycols, esters, ethers, phenols and phenol derivatives, as well as carbohydrates and a large proportion of lignin-derived oligomers. Liquefaction and pyrolysis are the two major technologies to produce bio-oils. Their properties depend on the specific feedstock and conditions of the production process such as temperature, period of heating, ambient conditions and the presence of oxygen, water and other gases. The possible utilization of bio-oil is, however, limited because of some negative attributes such as low pH, low heating value, high oxygen content and high viscosity. Bio-oil component can be converted into more stable fuels using zeolite catalysts (Bridgwater, 1994). Reaction conditions used for the above process are temperatures from 350°C to 500°C, atmospheric pressure and gas hourly space velocities of around 2 h-1. The products from this reaction include hydrocarbons (aromatic, aliphatic), water-soluble organics, water, oil-soluble organics, gases (CO2, CO, light alkanes) and coke. During this process, a high number of reactions occur, including dehydration, cracking, polymerization, deoxygenation and aromatization. However, poor hydrocarbon yields and high yields of coke generally occur under reaction conditions, limiting the usefulness of zeolite upgrading.

Bakhshi and co-workers studied zeolite upgrading of wood-derived fast — pyrolysis bio-oils and observed that between 30 and 40 wt.% of the bio-oil formed coke or char (Adjaye et al., 1996; Katikaneni et al., 1995a; Sharma and Bakhshi, 1993). The ZSM-5 catalyst produced the highest amount (34 wt.% of feed) of OLPs of any catalyst tested. The products in the organic liquid were mostly aromatic for ZSM-5 and aliphatic for SiO2-Al2O3. Gaseous products included CO2, CO, light alkanes and light olefins. However, bio-oils are thermally unstable and thermal cracking reactions occur during zeolite upgrading that leads to a high coke formation. Bakhshi and co-workers also developed a two-reactor process, where only thermal reactions occur in the first empty reactor and catalytic reactions occur in the second reactor that contains the catalyst (Srinivas et al.,
2000). The advantage of the two-reactor system is that it improves catalyst life by reducing the amount of coke deposited on the catalyst.

image127 Подпись: C3 image129

The transformation of model bio-oil compounds, including alcohols, phenols, aldehydes, ketones, acids and mixtures, has been studied over HZSM-5 catalysts (Fig. 15.2) (Gayubo et al., 2004a, 2004b, 2005). Alcohols were converted into the corresponding olefins at temperatures around 200°C; then, the olefins obtained were transformed into higher olefins (either butenes or C5+ olefins) above 250°C. At temperatures higher than 350°C, the olefins are transformed into C4+ paraffins and a small proportion of aromatics. Phenol has a low reactivity on HZSM-5 and only produces small amounts of propylene and butanes. 2-methoxyphenol also has a low reactivity to hydrocarbons and thermally decomposes generating coke (Gayubo et al, 2004a). Acetaldehyde had a low reactivity on ZSM-5 catalysts, and it also underwent thermal decomposition leading to coking problems. Acetone, which is less reactive than alcohols, converts into C5+ olefins at temperatures above 350°C. These olefins are then converted into C5+ paraffins, aromatics and light alkenes. Acetic acid is first converted to acetone, and that then reacts as above. Products from zeolite upgrading of acetic acid and acetone give considerably more coke than products from alcohol feedstocks (Gayubo et al., 2004b). Therefore, the majority of biomass-derived molecules produce large amounts of coke when passed over acidic zeolite catalysts. Gayubo et al. have recently studied the catalytic transformation of the aqueous fraction of crude bio­oil obtained via the flash pyrolysis of sawdust (from Pinus insignis) at 450°C over

image77

-> Light alkenes: C4~, C3_, C2“

HZSM-5 zeolite (Gayubo et al., 2009, 2010). Previously, the bio-oil has been subjected to stabilization treatments to minimize coke deposition on the catalyst and to attenuate deactivation. Co-feeding methanol (around 70 wt.%) minimizes coke deposition within and outside the catalyst particles, thereby increasing the viability of crude bio-oil upgrading (Gayubo et al., 2009). Furthermore, the deposition of coke might also be controlled in a specific step of thermal treatment prior to the catalytic reactor minimizing deposition on the catalyst and thereby attenuating deactivation (Gayubo et al., 2010).

The options for utilizing bio-oils in refineries are affected by its high acid number, high water content, high oxygen content and high metal content, particularly potassium and calcium. Metals can be removed with guard beds or ion exchange. Removal of metals is required before processing because these materials will typically poison catalysts. The low thermal stability, high water content and very high oxygen content make it difficult to blend the bio-oil with common refinery intermediate streams such as vacuum gasoil (VGO). The most serious problem for bio-oil processing is its high acid number that causes corrosion in standard refinery units. The industry standard for refinery vessels is that the total acid number of the blend must be less than 1.5 mg KOH/g. Bio-oil can probably be processed using 317 stainless steel cladding, which is not standard in refinery units. Therefore, bio-oils would require pre-processing in a 317 stainless steel system to reduce the acid number before processing in typical refinery units (Holmgren et al., 2007). Since the FCC is the biggest unit and the heart of most refineries, much more development work would be required to minimize refinery risk before such an approach was viable. As an alternative to blending, co-processing bio-oil with petrol feedstocks in an FCC unit might be possible if a separate feed system was used to inject the bio-oil. Hence, the direct feeding of bio-oils into standard refinery does not appear a straightforward task.

Among various upgrading processes, hydrodeoxygenation is a promising alternative to reduce the acidity and oxygen content. Bio-oil was hydrotreated at high pressures (2000-2500 psi) and low space velocities (0.1-0.2 LHSV) by Holmgren et al. (2007). At these high pressures and low space velocities, hydrodeoxygenation predominates. Large quantities of hydrogen are required to generate water during hydrodeoxygenation because of the high level of oxygen (46%) in bio-oil. The resulting hydrotreated oil was then cracked in an FCC or hydrocracker to produce gasoline. This approach is unlikely to be commercially viable because of the high hydrogen requirement and the high capital cost of the hydrotreatment step. Samolada et al. (1998) reported a two-step process of thermal hydrotreatment and catalytic cracking of biomass flash pyrolysis liquids (BFPLs). Thermal hydrotreatment of BFPLs can be effectively operated, producing liquid products that can be upgraded in a refinery. The heavy liquid product of this process (HBFPL), mixed with light cycle oil (LCO) (15/85 wt./wt.), was considered as a potential FCC feedstock. Commercially available cracking catalysts were found to have an acceptable performance. The obtained bio-gasoline quality is comparable with that of the VGO cracking but with low yields of approximately 20 wt.%. The co-processing of gasoil with a thermally hydrotreated bio-oil has also been investigated by Lappas et al. (2009). The results showed that the presence of the bio-oil favours the gasoline and diesel production but increases the coke yield. However, depending on the concentration of biomass liquids, it was shown that this option is technically viable for FCC units running with good quality feedstocks, that is the FCC unit with excess coke burning capacity.

Conceptual design of a bio waste ethanol plant

17.3.1 Introduction

This section describes the design of a small-scale mixed waste (biomass/plastics) gasifier with a synthesis gas fermentation-unit, which produces ethanol. Part of the synthesis gas will be used for heating up the gasifier and for making electricity (see Fig. 17.2).

Research has been carried out and a design has been made by Van Kasteren et al. (2005). This report shows an analysis of the fermentation of synthesis gas to ethanol process with the aid of bacteria. Before fermentation can take place first the biomass has to be gasified and the synthesis gas cleaned. The next section describes the choice of gasifier.

17.3.2 Choice of gasifier

A wide range of biomass fuels such as wood, charcoal, wood waste as well as agricultural residues — maize cobs, coconut shells, cereal straws, rice husks can be

image116

17.2 Scheme of a bio waste to ethanol plant.

used as fuel for biomass gasification. Theoretically, almost all kinds of biomass with moisture content of 5-30% can be gasified; however, not every biomass fuel leads to the successful gasification. Most of the development work is carried out with common fuels such as coal, charcoal and wood. Key to a successful design of a gasifier is to understand the properties and the thermal behaviour of fuel as fed to the gasifier. Gasification systems: descriptions are found in Chapter 16 and elsewhere (Reed and Gaur, 2000). Attention has to be given on tar (also known as creostate, is a sticky, condensable vapour whose main constituents are benzene, toluene, indene, naphthalene and phenol) minimisation during gasification. Its formation is affected by temperature, type of feedstock used and run time. Many gasifier designs produce so much tar that the gas clean up equipment cost is several times the gasifier cost. For the synthesis gas to be used in ethanol production, the level of tar has to be reduced to <50 ppm.

It is known that tar components have an influence on bacteria. A study of Ahmed et al. (2006) shows that the presence of tar inhibits the growth of the bacteria Clostridium carboxidivorans P7T during synthesis gas fermentation. However, the bacteria are not killed. After an adaptation period the bacteria start growing again. It appears that the bacteria are going to produce more ethanol and less acetic acid. So in this respect tar is even beneficial for the process. The work shows that certain amounts of tar do not cause problems for the synthesis gas fermentation process as long as enough adaptation time is taken into account. Still more research is needed to determine which tar concentrations are acceptable. Conclusion is that a gas cleaning system for removal of tars remains necessary although for synthesis gas fermentation processes this needs not to be so elaborate as for catalytic or for combustion applications.

The choice of the gasifier is mainly based on the gasifier requirements as stated before. The most important requirement is the great variety in feed the gasifier must be able to deal with. In the case of a gasification of bio waste the presence of impurities has to be taken into account. Another important issue in the feed — requirements of the gasifier is the size of the feeding material for the gasifier. For an entrained flow and/or a fixed bed gasifier, a relatively small feed size is necessary. These gasifiers need more elaborate grinding than the other type of gasifiers, although grinding is necessary for all gasifiers.

The gasifier configuration must be as simple as possible, in order to keep maintenance at a low level. Besides that it is desirable to use a technique, which is already proven and in which some experience is required. The gasifier has to be robust in order to be able to work with waste streams which differ in composition in time. The entrained flow and the fixed bed gasifiers are most sensitive for this change in feed compositions. Also quick stop and/or start up is more complicated with entrained flow and fixed bed systems.

Another demand set for the gasifiers is the synthesis gas quality. Because the produced synthesis gas will go to the fermentation unit in order to obtain ethanol, the quality and exact composition of the CO-H2 is not very important, although it is desirable to obtain a continuous composition of the synthesis gas.

Reviewing all the advantages and disadvantages of the different types of gasifiers, a Circulating Fluidised Bed Gasifier (CFB) seems the best choice:

• CFB has a simple design and requires low maintenance.

• CFB is able to tackle a large range of feed without requiring a strict feed-size.

• CFB can be scaled up to deal with larger amounts of feed.

• CFB has high efficiency because of the circulating ash.

• CFB is relatively easy to stop and start.

• CFB produces synthesis gas of reasonable quality.

• CFB is commercially available and already in use in several plants worldwide.

Feedstocks and processes

20.4.1 Biomass feedstock for reforming

Solid, liquid, and gaseous biomass can be used for reforming. Solid biomass first has to be gasified or evaporated, liquid biomass can either be first evaporated or processed in the liquid/supercritical phase. Gaseous biomass is handled as such.

Solid biomass (waste) is available from already existing industries (like agriculture, wood production, and first-generation biofuels). It is available in its bulk volume or has been densified using for instance pelletization allowing easier handling. The biomass can then be fed to a gasifier using a screw feeder or a hopper system. To convert solid biomass to a liquid has several advantages over using the solid biomass directly.

• A liquid is produced from bulky solid biomass, which is usually difficult to handle. By eliminating void volume which is inevitably present with solid biomass, the energy volumetric density is significantly increased. This makes trans-shipment and transport, especially over longer distances, much more effective.

• It can be stored in tanks. It is often more stable against biological decomposition and cannot ignite at ambient temperature.

• Liquids are easier to process especially when pressurized conversions are envisaged.

With liquefaction, a solution is being given for effectively utilizing biomass: to bridge the large gap of biomass supply and demand, and to do it in a sustainable way. Biomass is available decentralized (where it is being grown) but processing needs to be done centralized to benefit from the economy of scale. Biomass can be liquefied where the biomass is available and then be transported over long distances (road, water) to central processing units of similar scales as the current petrochemical industry. Besides technical and logistic advantages, this conversion chain will also give incentives for economical development and job creation especially in rural areas.

Fast pyrolysis (see Chapter 14) is a liquefaction technology which seems to be very attractive for handling relative dry biomass streams on a worldwide scale. Fast pyrolysis technology produces pyrolysis oil (or bio-oil) via rapid heating of biomass to approximately 500°C in absence of oxygen. In this way, the biomass is thermally decomposed and produces gases, vapors, and char. The vapors are condensed yielding the pyrolysis oil with a yield up to 70 wt%. The gases can be combusted to supply heat for the process and the char can be used as a fuel or it can directly be recycled back to the land since most minerals and metals are concentrated into it. Pyrolysis oil as such can be reformed directly. Additionally, more water-rich fractions of pyrolysis oil are co-produced within various bio-refinery concepts (see Section 20.4.2) which allows ‘milder’ reforming than the full oil.

Hydrothermal liquefaction can be used to produce oil from wet biomass streams which is the subject of Chapter 18. Very large quantities of low organic water streams are available (e. g., from municipal, fermentation, digestion waste and in the future from algae) which essentially could be used for reforming at elevated pressures. However, the organic concentration must not be too low (> 10 wt%) for the energetic efficiency of the process. Also bio-gas (CH4 rich) from anaerobic digestion (Chapter 12) can be used as reformer feed which, after a gas cleanup, is essentially a mixture of CH4/CO2.

Trace components in biomass such as sulfur and chlorine are a serious issue in both reforming and downstream catalytic conversions. Sulfur removal is manageable using commercial technologies, such as adsorption (e. g., ZnO, Ni, etc.) and hydrodesulfurization (HDS). HDS (e. g., Albemarle’s NEBULA, BASF, Haldor Topsoe) can bring S levels down to single-digit ppm, but is expensive. High chlorine levels pose a greater challenge. The best option for chlorine, ammonia, and metal contaminants is to use dedicated sorption processes for each contaminant. To summarize, various general and specialized cleanup solutions need to be developed and used, depending on the contaminants in gas and the downstream catalysts.

Production of biodiesel/biofuel via acylglycerol route

Another approach reported in the literature for the conversion of DD consists in esterification of FFA with glycerol to form acylglycerols as an intermediate step in the production of biodiesel/biofuels.

Synthesis of MAG from DD was mainly studied due to the large number of applications as additives, for enhancing plasticity of fats or as bases in the food, medicine and cosmetic industry. Among synthesized acylglycerols, the monoester has the highest surface activity, and therefore, its concentration is very important for direct usage of the reaction product as an emulsifier.

The esterification of glycerol with fatty acids leads normally to a mixture of MAG, DAG, and TAG and some amount of unreacted substrates. The proportions depend on the presence and type of catalyst, as well as the reaction conditions such as temperature and the molar ratio FFA:glycerol.

Studies showed that enzymes have an enormous potential as catalysts in the processes where high regioselectivity is required (Lo et al., 2005). However, for the large-scale synthesis, the processes are not yet competitive due to the high cost of the enzyme.

Different studies summarized hereafter describe processes for the synthesis of acylglycerols in order to decrease the acidity of the feedstocks. These processes are catalyzed either enzymatically or conducted under non-catalytic conditions. However, the step of transesterification of acylglycerols to FAME was not further described in order to evaluate the final quality of the biodiesel.