Findings from performance tests

The engine power (P) and BSFC at full load of the same engine running with market diesel (Do) and biodiesel B5 (B5) are given in Fig. 23.20. It is observed in Fig. 23.20 that engine power is higher and BSFC is lower with B5 fuel at all measuring points, although the improvement is not much due to low percentage of biodiesel in the blend. The average engine power was increased 1.34% while averaged BSFC was reduced 1.29%. The detailed explanation of this effect will be shown together with impacts of B5 on exhaust emissions below.

Impacts of biodiesel B5 fuel on exhaust emissions in comparison with the market diesel can be observed in Fig. 23.21.

It is depicted in Fig. 23.21 that using B5 fuel the HC, CO and PM were reduced 12.29%, 8.60%, 2.25% respectively, while NOx was increased 1.93%. The reduction of HC, CO and PM, and the increasing of NOx emissions with biodiesel fuels have already been mentioned by many researches. Those shown in Fig. 23.22 by U. S. Environmental Protection Agency (EPA)40 are an example of these effects.

Average emission changes found by the EPA for B20 (a blend of 20% biodiesel with conventional diesel) also showed significantly lower levels of emissions of specific toxic compounds for biodiesel and biodiesel blends, including aldehydes, PAH (polycyclic aromatic hydrocarbons), and nitrated-PAH.40 However, a number of factors such as different fuel system designs, engine calibrations, fuel

23.21 Comparison of exhaust emissions as the same engine was run with market diesel and biodiesel B5 fuel.

23.22 U. S. Environmental Protection Agency evaluation of biodiesel effects on pollutant emissions for heavy-duty engines.22

quality and blending rate can cause biodiesel emissions to differ significantly from the average values.

The increasing of NOx was shown to be related to a small shift in fuel injection timing caused by the different mechanical properties of biodiesel relative to conventional diesel. Because of the higher bulk modulus of compressibility (or speed of sound) of biodiesel, there is a more rapid transfer of the fuel pump pressure wave to the injector needle, resulting in earlier needle lift and producing a small advance in injection timing.41

Of the testing case, 5% of biodiesel in the blend did not affect much in the reduction of the energy content comparing with that of the market diesel, while the structural oxygen content of a biodiesel fuel improved its combustion efficiency due to an increase in the homogeneity of oxygen with the fuel during combustion. Because of this the combustion efficiency of biodiesel is higher than that of petrodiesel. The results were that, with the biodiesel B5 fuel, the engine power was increased, CO, HC and PM emissions were reduced. Because of low energy content of the biodiesel, higher biodiesel blends may lead to lower engine power and higher fuel consumption.

Whilst the zeolite catalysts described above are highly effective and form the basis of the majority of large scale process plants (see below for FCC catalysts), scientists are still searching for catalysts to improve the efficiency of catalytic pyrolysis. Probably, the major driving force has been the need to reduce gas yields and, hence, improve liquid content. In order to generate more efficient processes, a further type of molecular sieve materials has begun to be more extensively studied. These are mesoporous materials (normally silicates) that were first detailed by researchers at the Mobil Research and Development Corporation in New Jersey.191 These are now heavily researched for applications in many fields and readers are referred to a recent excellent collection of papers that provide a comprehensive review of the field.192 A typical example is shown in Fig. 14.3. The first mesoporous materials were given the abbreviation MCM (Mobil Composition of Matter) and, like zeolites, consisted of tetrahedral silica-oxygen linkages through which ran regularly arranged (periodic), uniform sized pores. Like zeolites, mesoporous materials are formed from organic templating or

structure-directing molecules but in mesoporous solid synthesis chemistry it is generally thought that the templating species are aggregates of the organics into micellar forms and this results in pore sizes that are around ten times (i. e. in the range of a few nanometre) that of the general zeolitic microporous solids (although much recent work has been carried out to extend this). Where the mesoporous materials differ is that the inorganic framework is amorphous and has no crystalline order and periodicity only arises from the pore structures. The most common pore arrangement is an hexagonal honeycomb arrangement which is more robust than a lamellar and cubic form. Materials that follow the Mobil synthesis route and have the hexagonal structure are normally described as MCM-41 and it is this material (and analogues) that dominates the mesoporous catalytic pyrolysis literature. However, the synthesis chemistry is well-developed and routes to complex combinations of macroporous-mesoporous structures,193 thin films194 and complex particle shapes195 as well as precise control of pore size are well documented.196

One of the important steps in providing active mesoporous catalysts is the incorporation of aluminium to provide acid sites and the usual material investigated for catalytic pyrolysis is Al-MCM-41.197 However, more acidic samples gave greater amounts of coke and gas products suggesting that an optimum aluminium content exists.197 These authors also report a lack of hydrothermal robustness of these materials and that would have a major effect on the industrial use of these materials.197 The relative hydrothermal stability of mesoporous silicates compared to zeolites and amorphous aluminosilicates for hydrocarbon cracking is well known.198 However, they do show good activity for the production of fuel oils from palm oil.199 Triantafyllidis and co-workers have shown that the nature of the acid site in Al-MCM-41 (since both Brpnsted and Lewis acid sites are formed) can have effects on the molecular distribution of the resultant fuel oils.200 One of the major reasons for the use of large pore systems is the possibility of improving the yield of higher molecular weight products and reducing the gas products. The rationale for the use of mesoporous systems (compared to microporous systems) is that the larger pore systems could allow diffusion and reaction of larger moieties and there is evidence in the MCM system for this pore size effect.201

Larger pore mesoporous systems include SBA-15, which is known to be more hydrothermal robust (because of thicker pore walls than MCM-41) and can be readily synthesised with alumina content.202 Qiang et al. have used various SBA-15 catalysts to study pyrolysis of sawdust.203 As might be expected, it was found that Al-SBA-15 significantly outperformed SBA containing no aluminium analogues and that catalytic activity improved with the amount of aluminium added.203 Aguado and co-workers have shown that Al-SBA-15 can also be used for the catalytic pyrolysis of polyolefins and shows very promising characteristics.204 In particular, hydrothermally stabilised SBAs outperformed ZSM-5 and this was attributed directly to the larger pore size which reduced diffusional limitations (as described above). To support the conclusion that SBA-15 may be a better pyrolysis catalyst than MCM-14, Cao et al. have shown that SBA-15 measurably improves the product fuel quality compared to MCM type materials.205

Figure 16.13 is a schematic diagram of the entrained bed coal gasifier when the solid fuel and gaseous stream are flowing in same direction. Entrained bed gasifiers are generally modeled as plug flow reactors. Now, as the solid fuel flows along with gaseous stream, it undergoes several reaction intensities depending on different operating parameters (like temperature, pressure, concentration, and time). During the whole process, there is continuous exchange of mass and energy between the gaseous and solid fuels.

Development of the model for the entrained bed gasifier is divided into three parts: (a) mass balance, (b) determination of the equation for calculating reaction rates, and (c) heat balance. Thus, the iteration of the equations developed from these three parts will give the desired output.

Figure 16.13 shows the model chart for the entrained flow reactor. Here it is assumed that the solid particles and gas are in one-dimensional plug flow in the axial direction and radially well mixed. For this case, the mass balance equation can be written as:

Mass balance for solid component

^ = — NArATs, L [16.38]

^ tv » /■

W = W (1 — X),

5 soK s

where Ws = flow rate of the solid (g/s), Wso = initial solid (g), NV = number of coal particles per unit volume, A = cross sectional area of the gasifier (cm2), Ts = temperature of the solid (K), L = gasifier length (cm), r = rate of solid gas reaction (gs-1), Xs = solid conversion.

Mass balance for the gas component

where Fgl = flow rate of gas (mol s-1), vlk = stoichiometric coefficient for the th gaseous components in kth solid gas reaction, | = extent of reaction (mol s-1).

Now the diffusion of the gaseous component into the solid component is governed by the reaction rates. Thus, the overall reaction rates can be defined as:

5

,</./. sJ’Kl. lul,. I 16.421

A=l

where, rk(Ts, L) = xsk(T)(Pyi)n4nr2ps, k = 1,. . .,5, N = number of coal particles per sec (s-1), x = surface reaction rate coefficient (gs-1 cm-2 atm-1), y = mole fraction of the gaseous component.

Using the above equation the coal conversion can be predicted and the size of the particles can be known from the relation:

" 5

Х0«г*(7’.’

_ *=i

where a = amount of gaseous reactant required to react with the unit mass of coal (mol g-1), D = diffusion coefficient of the gas (cm2 s-1), R = universal gas constant (KJ mol-1 K-1).

The expression for the rate of energy transfer (considering conduction and radiation) between the solid and the gas phases is as follows.

-еиа(г;-Т„4)^о,-л.(т.-Г„>о,

where cps = specific heat capacity of the solid (J g-1 K-1), DH = heat of the reaction (cal g-1), Я = thermal conductivity (cal s-1 cm-1 K-1), є = emissivity, о = Stefan — boltzman constant (cal s-1 cm-2 K-4), Di = internal diameter of the gasifier (cm).

Thus, the system of non-linear equations is developed by conducting mass and energy balance. By solving this system the composition of different product gases can be obtained.

The relation of the dimension (length, diameter, and thickness of the entrained reactor and the primary and secondary nozzle diameter) with coal capacity has been developed by Kim and Kim (1996). The results are shown in Fig. 16.14.

16.14 Design graphs for entrained bed gasifiers (Kim and Kim, 1996).

16.5 Conclusions

To secure a quality of life for current and future generations, sufficient water, land, and energy must be available. It is generally recognized that human development cannot continue to depend on fossil fuels in the present manner forever. Therefore, the issue is not whether renewable biofuels will play a role in providing energy but to what extent, and what the implications of their use will be for the economy, for the environment, and for the global security. What is seldom mentioned is that even in a ‘sustainable world’ not only energy but also carbon for organic chemicals, including plastics, is required.

Over the years, we have seen that the principal roles of syngas have shifted from domestic heating fuel, to feedstock for Fischer-Tropsch (F-T), to petrochemical feedstocks, to starting materials for alternative fuels, to IGCC, and to hydrogen sources. The only way to produce these useful resources from waste and biomass is to first gasify them in order to make syngas. In electric power generation, IGCC has contributed tremendously to improvement of power generation efficiency, thus keeping the cost of electric power competitive against all other forms of energy. Interest in methanol and dimethylether is revived due to the ever-rising cost of conventional clean liquid fuel. With the advent of hydrogen economy, there is no doubt that the use of hydrogen in combination with fuel cells as a transport fuel will improve the climate by eliminating CO2, NOx, CO, hydrocarbon, and soot emissions — and this is a prospect that could become reality within two decades.

The issue here is how to produce this hydrogen and make it available in a useable form. Gasification, coupled with water-gas shift, is the most widely practiced process route for biomass to hydrogen; however, it needs to be refined further. It is our opinion that gasification can and will have an important role to play in the coming decades. Therefore, more advances are expected in the areas of product gas cleaning, separation and purification, feedstock flexibility and feeding, disposition of ash/slag, plant availability, economics of scale, and integrated or combined process concepts.

Within few years, we have witnessed large steps towards the commercialization of the BTL-FT process. There are several companies active in technology development and commercialization of individual steps in the BTL-FT process sequence. A number of companies have large-scale biomass gasification technologies including Conoco Phillips, Siemens, VTT, TPS, CHOREN, Lurgi, Shell, GE, Kellogg Brown and Root, Prenflo, Advantica BGL, Noell, Winkler and KRW (E4tech, 2008). Additionally, there are companies focusing on the production of fuels from syngas, such as Sasol, Shell, JFE Holdings in Japan (slurry bed FT reactor producing dimethyl ether (DME)), Fuel Frontiers Inc. (ethanol from syngas) and Syntroleum (focus so far on CTL and GTL) (E4tech, 2008).

Very few companies, however, are active in the whole BTL process chain. The most important player in the BTL market is CHOREN, a German-based technology company. With its Carbo-V patented biomass gasification process for converting biomass to syngas, the company partnered with Shell and Volkswagen to construct the first commercial BTL plant in the world based on the Carbo-V gasification process and the Shell SMDS FT process. The CHOREN beta demonstration plant in Freiberg, Germany, has been operating since 2005, with a capacity of 45 MW thermal and 15 000 tons of BTL fuel per year. CHOREN is currently constructing the first commercial BTL plant in Schwedt, Germany, with a capacity of 640 MW thermal and 200 000 tons of BTL fuel per year using these technologies, with fuel production scheduled to start in 2012 (Rudloff, 2005).

The efforts of CHOREN for commercialization of the BTL process are complimented by substantial research activities. CHOREN is participating in the OPTIFUEL demonstration project, a 42-month project funded by the EU with 7.8 million Euros within the 7th Framework Program for Research and Technological Development. The project kicked off in February 2010 and is expected to establish the technical basis for the large-scale production of BTL-FT fuels with a consortium comprising besides CHOREN, the automotive companies Ford, Renault and Volkswagen, CONCAWE, representing the European mineral oil industry, Invensys Process Systems as simulation technology provider, research institutes IFP (France), CERTH (Greece), IIT Delhi (India) and the German project consultant SYNCOM. Performance data from the CHOREN Freiberg demonstration plant will be modelled to identify improvement opportunities compared to the current production processes and to create the technical basis for large-scale BTL production facilities. Using BTL products manufactured in the Freiberg plant, the automotive manufacturers and oil industry will work together to blend the BTL liquids, evaluate their exhaust emissions and explore their potential in current and future engine technologies. In addition, the economic aspects and the potential to reduce energy and greenhouse emissions from all parts of the BTL production process will be evaluated. (OPTIFUEL, 2009).

Currently, the main focus in both conventional and advanced biofuel production processes is on the production of the specific biofuels rather than on the development of processes that maximise the valorisation of the raw materials to both bioenergy and bio-based products in a sustainable way.

Valorisation of both primary and secondary biofuel chain and process residues to added-value bio-based products (chemicals, materials) is the short term option for upgrading of these processes to integrated biorefinery processes, maximising the overall valorisation of the raw materials concerned, minimising the biofuel production costs, and thereby increasing their market competitiveness.

Integration of both biochemical and thermo-chemical biomass conversion processes into already existing petrochemical infrastructures is another short term option to valorise biomass to both bio-based products and biofuels, greening their fossil counterparts.

Development of relatively small-scale concepts seems to be a favoured option to introduce more advanced (green and whole crop) biorefinery processes into the market at mid-term. These concepts require less initial investment which is an advantage for the industrial stakeholder support for the introduction of new risky initiatives and because of the economy-of-duplication these concepts are expected to become market competitive soon. These concepts will create the perceptional, socio-economic and environmental framework for the introduction of even more advanced (lignocellulosic feedstock and marine) biorefinery concepts at larger scale on the longer-term.

While direct and indirect photolysis systems produce pure hydrogen, dark — fermentation and photo-fermentation processes, produce a mixed-biogas — containing primarily hydrogen and carbon dioxide (CO2), but which may also contain lesser amounts of methane (CH4), carbon monoxide (CO) and/or hydrogen sulphide (HjS) or ammonia (NH3). Moreover, the hydrogen content in the gas phase is in general lower than 50%. PEMFCs require hydrogen at a high purity (>99%) and cannot tolerate CO at concentrations higher than 10 ppm. In order to remove diluting (CO2, CH4) and/or contaminating (CO) gases, purification of the biogas is essential. Up to now, membrane technologies based on palladium have been proposed as hydrogen purifier in industrial scale applications (Shu et al, 1991).

Although the cracking of vegetable oils into liquid fuels has been studied in detail, the cracking of triglycerides molecules under realistic FCC conditions is less described in the literature. However, certain number of authors have performed studies about the processing of vegetable oils (Bhatia et al., 2007, 2009; Chew and Bhatia, 2009; Dupain et al., 2007; Li et al., 2009; Melero et al., 2010b; Tamunaidu and Bhatia, 2007; Tian et al., 2008) and animal fats (Lummus, 1988; Melero et al., 2010b; Tamunaidu and Bhatia, 2007; Tian et al., 2008) under conditions that try to simulate operating conditions of the FCC unit. In these studies, the reaction system employed is usually based in a riser reactor and an FCC catalyst. After the catalytic cracking reactions, conversion is usually over 75% (Bhatia et al., 1998; Chew and Bhatia, 2009; Melero et al., 2010b; Tian et al., 2008). Furthermore, there are no remarkable amounts of oxygenated hydrocarbons in the final cracking products, as almost all the oxygen initially present in the triglyceride molecule ends forming water or carboxylic gases (CO and CO2) (Dupain et al., 2007; Melero et al., 2010b; Tian et al., 2008).

Figure 15.5 shows the yields towards different products for the catalytic cracking of crude PO in a fixed bed reactor of short contact time at 565°C and a catalyst-to-PO mass ratio of 4 (Melero et al., 2010b). Besides the oxygenated compounds detected (water and carboxylic gases), main hydrocarbon products are gaseous hydrocarbon products, such as dry gas (H2, methane, ethane, ethylene) and liquid petroleum gases (LPG, propane, propylene, butenes, butanes), and liquid hydrocarbon products such as gasoline (GLN; C5, 221°C), which is divided into light naphtha (LN; C5, 90°C), medium naphtha (MN; 90-140°C) and heavy naphtha (HN; 140-221°C), LCO (221-360°C) and decanted oil (DO; > 360°C). As observed in Fig. 15.5, water is the main oxygenated compound in the cracking of vegetable oils because it involves approximately 70% of the initial oxygen in the triglyceride molecule, which means a yield of water in the final product cracking of ca. 10% when a 100% crude PO feedstock is processed. Similar results have been described in the catalytic cracking experiments performed by different authors (Dupain et al., 2007; Marker, 2007; Ramakrishan, 2004). Water is produced by means of decarboxylation reactions (Idem et al., 1996) as well as catalytic dehydration reactions (Chang and Silvestri, 1977) or condensation processes (Adjaye and Bakhshi, 1995). Carboxylic gases are also important oxygenated compounds with a yield of ca. 5%. Carboxylic gases are formed by CO in 60% mass percentage and CO2 in 40% (Melero et al., 2010b). CO is formed through decarbonylation reactions from different molecules such as ketenes, aldehydes, fatty acids and esters. By-products of this reaction depend on the original oxygenated compound. On the one hand, in the case of ketenes and aldehydes, decarbonylation reactions lead to reactive species such as free radicals and, on the other hand, in the case of fatty acids and esters, they produce alcohols

(Idem et al., 1996). CO2 is formed through fatty acid and ester decarboxylation reactions, producing water and ketenes as by-products (ketene usually loses its oxygen molecule because of molecular decarbonylation reactions to form ethylene). These data mean that around 17% of the initial oxygen ends as CO and 11% as CO2. Hence, 15% of the renewable raw materials that are being fed to the FCC reactor end up as non-valuable products (water and carboxylic gases) under the tested reaction conditions used in this work (Melero et al., 2010b).

Dry gas is mainly a thermal cracking product, although it can be obtained by means of catalytic reactions, especially in the case of ethylene. Dry gas is not an

important cracking product because it is obtained in a small percentage (never higher than 5%) and it has a low commercial value. Ethylene is the main compound, leading to more than 40% of the final dry gas yield, and ethane and methane production is always close to 30% for both compounds. On the other hand, LPG production in the case of PO cracking is a very important fraction with a yield of ca. 25% under the tested reaction conditions (Melero et al., 2010b). High yields of gaseous products have also been achieved by other authors working under the FCC conditions (see Table 15.2). Tamunaidu and Bhatia (2007) achieved yields of gaseous hydrocarbons ranging between 19.9% and 38.1% in the cracking experiments of PO using a riser reactor (temperature = 400-500°C and catalyst-to-oil mass ratio ranging from 5 to 10). Similar experiments were performed by the research group of Chew and Bhatia (2009). These authors obtained yields of gaseous products of 16.2% and 15.9% for crude and used PO, respectively, using a riser reactor at 450°C and a catalyst-to-oil mass ratio of 5. Finally, Li et al. (2009) confirmed these results, reaching yields to gas of 28.8% in their cracking experiments of cottonseed oil in a fluidized bed reactor (temperature = 400-500°C and catalyst-to-oil mass ratio of 6-10).

LPG gases are mainly a catalytic cracking product obtained through dealkylation reactions, in which the hydrocarbon chain bonded to an aromatic ring can be broken to end up as gases (Dupain et al., 2007), or through the initial cracking of higher molecular weight products. LPG hydrocarbons are usually produced by means of b-scission reactions in which a primary carbenium ion and an olefin are formed. Afterwards, it is quite probable that hydride transfer reactions will be produced, transferring the charge from a small carbenium ion onto a large hydrocarbon and, as a consequence, forming new olefins, which can be protonated again by a Bronsted acid site and cracked further or isomerized. Obviously, after hydrogen transfer reactions, paraffins are produced. However, LPG composition is mainly olefinic and much based on propylene (more than 35% of the total LPG), although there are also important amounts of isobutane and, in a less relevant amount, C4 olefins, which are produced in the same quantity between them (Melero et al., 2010b).

The liquid product of a catalytic cracking process is usually composed of cyclic and linear aliphatic hydrocarbons as well as aromatic compounds. The main hydrocarbon liquids considered are GLN, LCO and DO (Melero et al., 2010b; Tian et al., 2008). DO is the heaviest reaction product and, in the case of the renewable raw materials, it is obtained by means of condensation or polymerization reactions (Horne and Williams, 1996; Idem et al., 1996). This fact explains the low yield towards DO of around 2-4.5%, as shown in Fig. 15.5 and Table 15.2, in the results obtained by the research groups of Melero et al. (2010b) and Tian et al. (2008). On the other hand, GLN is the main liquid compound, with a yield that can be close to 40% of the total product distribution (that means more than 75% of the OLP) (Melero et al., 2010b). The LCO presence is less important, and it implies a yield of ca. 10-15% (Melero et al., 2010b; Tian et al., 2008). Both

Research group Experimental conditions Feedstock Product (wt.%) OLP Gases GLN LCO Tamunaidu and Riser reactor Palm oil 19.9-38.1 49.5-59.1 0.1-10.4 Bhatia (2007) T = 400-500°C Catalyst-to-oil ratio = 5-10 Chew et at. (2009) Riser reactor Crude palm oil 16.2 43.5 4.2 T = 450°C Catalyst-to-oil ratio = 5 Used palm oil 15.9 33.0 4.5 Li etal. (2009) Fluidized bed T = 400-500°C Catalyst-to-oil ratio = 6-10 Cottonseed oil 7.5-28.8 25.1-33.7 49.5-64.0 Gases OLP Dry gas LPG GLN LCO DO Coke Tian etal. (2008) Riser reactor Chicken fat 4.48 34.34 32.75 11.40 2.95 2.31 T = 400-500°C Palm oil 6.35 41.56 28.14 8.90 1.97 2.20 Catalyst-to-oil ratio = 6-10 Soybean oil 4.59 29.24 32.27 15.28 4.50 3.98

gasoline and LCO are involved in b-scission, isomerization and hydrogen transfer reactions of the hydrocarbons, which come from decomposition of heavy hydrocarbons. Furthermore, cracking under FCC conditions involves high contents in aromatic hydrocarbons in the organic liquid phase. The high number of dehydrogenation reactions to remove oxygen in the form of water leads to an increase in the olefins formation, which leads to the aromatic compounds formation under the FCC reaction conditions. Concretely, an aromatic content of 30-40% has been reported in the gasoline fraction (Melero et al, 2010b; Tian et al., 2008).

Last reaction product is coke, which is mainly produced by a thermal pathway. Most catalyst deactivation associated with coke formation is produced in the initial reaction period because some of the free radicals formed by thermal processes are not able to go within the catalyst pores and are deposited in the most external part of it (Dupain et al., 2006). Coke can also be obtained from thermal direct polycondensation of either triglyceride molecules or primary heavy oxygenated hydrocarbons (Katikaneni et al., 1997). Furthermore, coke might also be obtained by a catalytic route that involves the formation of polyaromatic compounds coming from a successive hydrogen elimination of aromatic molecules. Nevertheless, coke coming from a catalytic route is always lower than that obtained thermally.

Under HTL conditions, deoxygenation can, to a certain extent, be realized without hydrogen. This is often stated as one of the major advantages of this technique, but apparently, the products would be different than those obtained by catalytic hydrotreating. Depending on the operating conditions and the feedstock, oxygen contents of the hydrophobic phases as low as 4-7 wt.% are reported in the literature.32,33 However, such low oxygen content is rare and the average reported oxygen concentration is 15-20 wt.%. Oxygen removal under the HTC conditions occurs via the following reactions: dehydration, decarboxylation and decarbonylation. CO added to the reactor or formed in the reactions appears largely converted to CO2 in the water gas shift or reduction reactions. Consequently, the net effect of deoxygenation during the HTC is the CO2 and HjO formation, the first one being favourable from an energetic point of view. It has been shown that the reactions of mono-sugars in hot compressed water are dominated by dehydration reactions,1 resulting in significant water production.

Separation of the primary conversion, namely the evaporation of pyrolysis oil, from the catalytic steam reforming seems to have a few advantages:

• Fixed bed commercial catalysts can be directly used which have been proven to be active and do not need additional mechanical strength for fluidization.

• Pyrolysis oil re-evaporation can be done at a lower temperature than the catalytic conversion to syngas, which is beneficial to the overall exergy efficiency of the process.

• Primary pyrolysis oil conversion seems to be mainly thermally driven, followed by catalytic gas upgrading. Actual splitting of these two processes makes separate optimization possible.

• Formed carbonaceous deposits and particles and other impurities like residual ash can be separated before the catalytic fixed bed, making energy utilization possible by burning the carbon or allowing them to gasify using steam and or CO2.

The first staged conversion of pyrolysis oil was reported and tested by Van Rossum et al4,35 where a methane free and low tar syngas was produced at ~810°C and a S/C of 1.5 (see Figure 20.6). Here an ‘inert’ fluidized sand bed was used followed by a fixed bed with a commercial catalyst. Other proposed and/or tested staged systems include the usage of a pre-catalyst (dolomite,50 char gasification enhancing,51), ‘inert’ gasification coupled with a current-enhanced catalytic reforming system52 and a pre-oxidation step to facilitate reforming.53

Sampathkumar (1986) described a process for recovering tocopherols from deodorizer sludge for use in pharmaceutical and food applications without the loss of vitamin E activity with an overall yield of 97%. It was found that urea forms an inclusion complex with all the fatty acids and glycerides of fatty acids without entrapping the sterols, tocopherols and other bulky molecules present in the deodorizer sludge or distillate.

The process comprises the heating of deodorizer sludge and a solution of urea mixture to form a urea complex of the fatty acids and glycerides of fatty acids. The mixture was cooled to precipitate the urea complex from the mother liquor containing the tocopherols, and separating the mother liquor from the precipitate. The mother liquor was further concentrated and the residual solids were separated. Extracting the mother liquor, an oil rich in tocopherol could be obtained. It was suggested that the instability of the urea complex of fatty acids in the presence of water and an acid can be used for further purification of FFA and their separation.

A similar process was described by Maza (1992) where urea was used for the separation of mixed fatty acids, sterols and tocopherols from DD with an increased yield and reduced use of organic solvents. The process comprised the sequential steps of (1) melting the DD, (2) adding the melted DD to a refluxing solution of urea and alcohol to form a mixture, (3) mixing the reaction mixture while cooling to allow formation of crystals, (4) separating the crystals, (5) drying the crystals, (6) dissolving the crystals in water to form an organic layer which is rich in mixed fatty acids and an aqueous layer containing urea and (7) separating the fatty acid layer. By applying this process, a light fraction enriched in FFA and a heavy fraction enriched in tocopherols and sterols were obtained. Urea is recovered for reuse by combining the separated aqueous solutions containing urea and evaporating the water. The key element of the invention is the first step of the process, i. e. melting of the DD providing an easily dispersed reactive liquid which is not diluted in the solvent.