A. A. KOUTINAS and S. PAPANIKOLAOU, Agricultural University of Athens, Greece

Abstract: Biodiesel and bioethanol constitute the main biofuels produced currently at industrial scale from renewable resources (mainly oilseeds, waste oils, starchy crops and sucrose-rich biomass). However, the limited availability of conventional raw materials and/or the direct competition with food production restricts the growth of first-generation biodiesel and bioethanol production. In the last few years, there is a growing interest in biodiesel production from microbial oil accumulated by oleaginous microorganisms cultivated on waste streams from the food industry and agricultural residues. This chapter focuses on the description of the potential of microbial oil production by yeast and fungi, the biochemistry of oil accumulation and the prospect of biodiesel production from microbial oil.

Key words: biodiesel, microbial oil, biorefinery, oleaginous microorganisms, biomass.

8.1 Introduction

Bioethanol (mainly from sucrose and starchy crops) and biodiesel production (via transesterification of triglycerides) are the main first-generation biofuels that are currently produced on industrial scale. Biodiesel is produced by transesterification of triacylglycerols with short-chain alcohols (mainly methanol or ethanol) to produce monoalkyl esters, namely fatty acid methyl esters (FAMEs) and fatty acid ethyl esters (FAEEs). The worldwide production of biodiesel is mainly dependent on the utilization of waste oils, animal fats and oilseeds such as rapeseed, sunflower and soybeans. The recent food crisis has shown that research should focus on the development of second-generation biofuels generated from lignocellulosic raw materials and industrial waste streams (e. g. food industry wastes).

In the past few years, research has focused on the development of biodiesel production from single cell oil (SCO) that can be produced via fermentation using various oleaginous microorganisms (i. e. microorganisms that are able to accumulate lipids intra-cellularly at more than 20% of the total cellular dry weight). The proposed strategy may provide a more eco-efficient and sustainable option as compared to first-generation biofuels and second-generation bioethanol production routes utilising lignocellulosic biomass. Potential advantages include:

— The raw materials that will be used for the production of SCO-derived biodiesel do not compete with food production. In this way, cultivation of land

for food production as well as industrial food processes could coincide with biodiesel production by utilizing residues and agro-industrial wastes.

— Microbial oil could be produced from various carbon sources (e. g. glucose, lactose, xylose, sucrose, glycerol) using natural microorganisms contrary to bioethanol production where natural microorganisms that are traditionally used in industrial processes utilize mainly glucose and sucrose.

— Bioethanol separation is an energy intensive technology with significant capital investment requirements, while separation of intra-cellularly accumulated SCO is likely to be achieved at significantly lower capital cost and energy requirements.

— Biodiesel production from oilseeds and waste oils will never provide adequate quantities of biodiesel to sustain the worldwide demand. In addition, the production cost of oilseeds is approximately 70-80% of the total biodiesel production cost. Biodiesel production from SCO will depend on the utilization of low-value waste streams or residues and therefore will offer a sustainable option for biofuel production.

— Transesterification of SCO results in the production of crude glycerine that could be used as a platform intermediate for the production of biofuels, chemicals and biodegradable plastics (Koutinas et al., 2007a; Aggelis, 2009).

Since 1990, an international conference on the genetics and physiology of acid — and solvent-producing clostridia is being held biannually. Clostridium 1 took place in Salisbury, UK, while the following meetings were located in Blacksburg, Virginia (1992), Evanston, Illinois (1994), Ulm, Germany (1996), Toulouse, France (1998), Champaign/Urbana, Illinois (2000), Rostock, Germany (2002), Edinburgh, UK (2004), Houston, Texas (2006), and Wageningen, The Netherlands (2008). Clostridium 11 will be held in the United States in 2010 and Clostridium 12 probably in Nottingham, UK in 2012. These symposia provide an excellent means for both young and experienced researchers to learn about most recent findings in basic and applied aspects of the solvent-forming clostridia.

An additional resource is the web page ‘www. clostridia. net’, maintained by Nigel P. Minton in Nottingham, providing information on apathogenic and pathogenic clostridia, forthcoming conferences, transnational research collaborations, and Marie Curie-workshops especially aiming at pre — and postdoctoral researchers.

Finally, six books have been published meanwhile, either dealing with specific aspects of clostridia or giving a comprehensive overview of this genus (Bahl and Durre, 2001; Bruggemann and Gottschalk, 2009; Durre, 2005b; Minton and Clarke, 1989; Rood et al, 1997; Woods, 1993).

10.6 Acknowledgments

Work reported from this laboratory was supported by grants from the BMBF GenoMik and GenoMikPlus projects (Competence Network Gottingen) and the SysMO project COSMIC (PtJ-BIO/SysMO/P-D-01-06-13), www. sysmo. net.

C. DE LUCIA, University of York, UK and Technical University of Bari, Italy

Abstract: This chapter illustrates and discusses main objectives of biofuels policies viewed under multidirectional effects on economy, energy and environment. The analysis touches multiple effects of biofuels production and use such as the need for guaranteeing energy security and supply, environmental protection and land-use change, the expansion of rural areas and food safety and the increasing institutional support for biofuels policies including the contribution of these to climate change mitigation.

Key words: biofuels, feedstock, land use, rural development, climate change mitigation.

2.1 Introduction

Since their introduction in the supply chain, biofuels contributed to the reduction of carbon emissions. It is this evidence, together with advances in technological progress for renewables use and recent development of international agreements on climate change, that suggests to governments the adoption of new practices to enhance the agricultural sector. A renovated agricultural system was launched for biofuels feedstock production. This, in turn, served as a stimulus for countries facing current unbalances of imported energy commodities to search for new energy supply and security initiatives. Additionally, current biofuels feedstock production and future bioenergy and biorefinery practices are instrumental in the enhancement of rural development and the creation of further policy tools in the biofuels industry as well as the agricultural sector. However, this scenario is not without drawbacks. The positive and negative synergies occurring across a multitude of biofuels objectives should be carefully addressed. The aim of this chapter is to illustrate and discuss main objectives of biofuels policies viewed under multidirectional effects on economy, energy and environment. The chapter is organised as follows: Section 2.2 illustrates biofuels and bioenergy seen as energy security and supply; Section 2.3 discusses environmental and land-use concerns linked to biofuels practices; Section 2.4 emphasises the risk for food safety and the need for the development of marginal areas when considering biofuels activities; Section 2.5 describes current biofuels policy support and delineates future scenarios for climate change mitigation; finally, Section 2.6 concludes.

The main feedstock for ethanol production is sugar from cane and beet. Sugar is converted into bioethanol by ethanologenic fermentation. The most employed microorganism is Saccharomyces cerevisiae due to its capability to hydrolyse cane sucrose into glucose and fructose, two easily assimilable hexoses (Sanchez and Cardona, 2008). Yeasts such as Schizosaccharomyces pombe present the additional advantage of tolerating high osmotic pressures (high amounts of salts) and high solids content (Bullock, 2002). Among bacteria, Zymomonas mobilis provides higher ethanol yield, up to 97% of theoretical maximum (Claassen et al., 1999). The disadvantage of its use during fermentation is the formation of a polysaccharide (which increases the viscosity of fermentation broth) and sorbitol, which decreases the efficiency of the conversion of sucrose into ethanol (Lee and Huang, 2000).

Utilizing any type of unused plant-derived oils, also known as straight oil (SO), as feedstock is economically infeasible, resulting in a high final cost of the biodiesel. Above that, it raises ethical questions as this feedstock competes with food stock. Shah and Gupta (2006) argued that it is more reasonable to use inedible oils such as Jatropha oil. This argument however is debatable, as a land has to be developed for plantation, and it would be more advisable then to use it to plant something that can be used as food stock. The only sensible way to overcome this dilemma is to use waste oils (WO) and waste fats (WF) as raw materials for biodiesel production. In addition to using feedstock that does not compete with food stock in this case, the use of WO and WF is considered an important waste minimization and recycling process, no less than half a million tons of which are discarded every year in Japan alone (Kaieda et al., 1999).

In comparison to SO, WO has significantly higher amounts of water, around 2000 ppm and FFA, 10-15% (Zhang et al, 2003; Lai et al, 2005), as well as higher polymerization products. As explained earlier, the high FFA content renders alkali — catalysts processes not suitable, and the use of chemical catalysts is limited in this case to the acidic ones (Zhang et al, 2003). Due to the comprehensible attractive benefits of WO, biodiesel production from this feedstock has been investigated using acidic catalysts, despite being much slower and more hazardous catalysts compared to the other chemical catalyst, namely the alkaline (Al-Widyan and Al-Shyoukh, 2002; Al-Widyan et al., 2002; Zhang et al, 2003). Methanolysis of triacylglycerols (TAGs) with a lipase is considered one of the effective reactions for production of biodiesel fuel from WO. Shimada et al. (2002) have successfully produced biodiesel from WO using immobilized lipase from C. antarctica. They have further proved that the yield of biodiesel production from WO, containing up to 2000 ppm water, was comparative to that from SO. WO containing around 500 ppm water was also successfully utilized to produce biodiesel, using lipase from bacterial, P. cepacia, and yeast, C. antarctica, sources in free and immobilized on ceramic beads forms, in the presence and absence of n-hexane (Al-Zuhair et al., 2008).

Animal fats have also been used for biodiesel production (Ali et al, 1995). However, due to the high melting point of animal fats, that is usually near the denaturation temperature of lipase, and because methanol and animal fat are immiscible, the reaction system has to take place in an organic solvent to dissolve the solid fat (Ma et al., 1999). The use of organic solvent, however, requires the addition of solvent recovery unit. To overcome this drawback, thermostable lipases, which have relatively high optimum temperature, can be used.

There is a vast amount of lignocellulosic waste material from agriculture and the forest industry which can be used for ethanol production. Lignocellulosic biomass like wood and fast growing plants like switch grass, reed canary grass or crop residues from food production such as corn stover are cellulose feedstocks which can be used in bioethanol production.

Lignocellulosic biomass is composed of polymeric structures of cellulose, hemicellulose, lignin, other organic compounds (extractives) and inorganic salts. Cellulose is the major component in most lignocellulosic biomass. In fact, it is the most abundant polymer on earth. Like starch, cellulose is a polymer of glucose molecules and the chain length varies between 100 and 14000 units. However, in cellulose the glucose units are connected to each other by b-1,4-glycosidic bonds instead of a-1,4-bonds as in starch, the structure of cellulose is shown in Fig. 9.2.

This makes a crucial difference compared to starch. In cellulose the glucose polymer is linear giving the possibility for the cellulose chains to align with each other and form multiple hydrogen bonds between the chains. In this way, cellulose can form crystalline structures. These crystalline structures are very stable and they are the reason why it is so difficult to hydrolyse cellulose: the crystals are so tight that it is very difficult for the hydrogen ions and the water that is needed for the hydrolysis to actually get to the glycosidic bonds. In fact, although cellulose consists of very polar glucose units, the tight hydrogen bonds prevent water solvating the polymer and therefore cellulose is not soluble in water. This is fortunate because otherwise cotton clothes (cotton being pure cellulose) could not be washed and would not be so useful! However, not the whole portion of cellulose is in the crystalline form, in some locations, the crystal structure is disturbed and an amorphous form of cellulose is formed. This form is not as stable as the crystalline form and is more susceptible to hydrolysis.

The cellulose chains that are held together with hydrogen bonds form what are called fibrils and a bundle of these fibrils then forms the actual cellulose fibre. In order to ‘soften’ the cellulose, the hydrogen bonds must be broken and that is why the concentrated acid method is so effective: in such a high concentration of acid or also in fact strong base, the hydrogen bonds are broken and access to the glycosidic bonds is made. The double sugar units with a b-1,4-bond between the two glucose units is called cellobiose.

Hemicellulose is a branched polymer of both 6-carbon sugars (hexoses) like glucose, mannose and also 5-carbon sugars (pentoses) like xylose. In grasses and hardwoods, the pentoses in the form of xylans dominate, while in softwoods the major hemicellulose component is the hexosic glucomannan. Since the hemicellulose polymer chain is branched, the formation of hydrogen bonds creating the crystalline

structure of cellulose is prevented. This makes hemicellulose much more susceptible to the hydrolysis of the glycosidic bond. Actually, hemicellulose in solution is as easy to hydrolyse as starch. Two different structures of hemi-cellulose are shown in Fig. 9.3.

Lignin is a polymeric structure of aromatic units (p-hydroxy-phenyl-propanoid units) and the second most prevalent polymer on earth. Lignin functions as the glue between the cellulose fibres in the lignocellulosic biomass. The amount of lignin varies depending on the type of biomass; Table 9.1 shows the typical composition of cellulose, hemicellulose and lignin in different types of biomass.

The composition of lignin also varies between different types of biomass. The phenyl ring in the monomer structure of lignin can either have no, one or two

methoxy groups. In grasses the non-methoxy monomer is predominant, in hardwoods there is a mix of all three and in softwoods the one and two methoxy rings are predominant. Since lignin does not contain as much oxygen as cellulose and hemicellulose, the energy value is much higher. Cellulose and hemicellulose have an energy value (calorific value) of approximately 17 MJ/kg, while lignin has up to 25 MJ/kg. So although lignin is only around 25% of the dry solid content in wood for example, almost 40% of the heat value comes from it. A structure of a segment of lignin in softwood is shown in Fig. 9.4.

Historically, lignin has always been utilised as an energy source, for example in the energy recovery boilers of the pulp and paper industry. In a future bio-refinery process lignin may have a more important role as a feedstock for the production of several organic compounds, e. g. phenol. One problem chemically with lignin produced in a dilute acid or enzymatic process is that it is highly condensed which reduces the number of reactive hydroxyl groups and therefore there are problems to react it further.

Switch grass (Panicum virgatum L.) is a perennial warm-season C4 species (tolerant to heat and cold), which can be used in bioethanol production. This grass

is grown in Central USA as a fodder crop or for soil conservation and is a potential long-term bioethanol feedstock to replace corn. The composition of switch grass on a dry basis is about 30-36% cellulose, 24-27% hemicellulose and 16-18% lignin. From highly adapted switch grass varieties the theoretical ethanol production potential is about 5000-6000 litre/ha. Based on the technique used in ethanol production the ethanol yield is often high (72-92% of the theoretical value in labscale). The excess of switch grass can be used to produce Kraft pulp with short fibres (Keshwani and Cheng, 2009).

Reed canary grass (Phalaris arundinacea L.) is a perennial rhizomatous grass which is mainly used as a raw material for solid biofuel production in the Nordic countries. This grass grows naturally in Europe, Asia and North America, especially in wet and humus rich soil. The grass is about two meters tall with a sturdy, upright straw, broad leaves and a long panicle. The annual production yield is eight to ten tonnes dry solid/ha in Sweden (Xiong, Landstrom, and Olsson, 2009). The harvesting starts normally some years after establishment and growth persists for at least 12 years (Xiong, Landstrom, and Olsson, 2009). The grass is usually stored and transported as bales to increase the density and reduce production costs. Reed canary grass consists mainly of cellulose, hemicellulose and lignin, but there are also proteins, lipids and a relatively high content of inorganic material. The main sugars after hydrolysis of reed canary grass are glucose, xylose and also arabinose. In reed canary grass, the amount of hexoses in the stem varies between 38% and 45% of the dry weight of the material and the amount of pentoses about 22-25%. The lignin content varies between 18% and 21% of the dry weight. Therefore, the grass has a good potential as a feedstock for ethanol production in the future (Arshadi and Sellstedt, 2008).

Reed canary grass has also been found to be a useful complement to short fibre raw materials like birch in kraft pulp production (Paavilainen, 1996; Finell and Nilsson, 2004). Alfalfa (Medicago sativa L.) is usually used for the production of fuel, feed and other industrial materials. Alfalfa stems consist mainly of cellulose, hemicellulose, lignin, pectin and proteins. Therefore, the feedstock has the potential to be used for ethanol production and also other chemicals (Diena et al, 2006).

Previous work has shown that it is possible to produce ethanol from alfalfa either by separate hydrolysis and fermentation (SHF) or simultaneous saccharification and fermentation (SSF). The yield of fermentable sugars from hydrolysis or saccharification is an important response variable in assessing the value of the feedstock. Corn seed has been used as a starchy feedstock in bioethanol production but other parts of the corn plant have not been used until recently. The stalk and the leaves, which are called corn stover, can be used as a source of lignocellulosic material in ethanol production; also the corn cob can be used. The amount of corn stover is huge since for every kilogramme of produced corn, almost the same amount of corn stover is left. The amount of corn stover available for fermentation usage is estimated to be between 60 and 80 million dry tonnes per year (Kadam and McMillan, 2003). Some of the corn stover needs to be left in the field to prevent soil erosion and also corn stover may be needed as a feedstock for bio-based materials like composite products (Kadam and McMillan, 2003), but some part can be collected and used as a raw material in bioethanol production (Ohgren, Rudolf, Galbe, and Zacchi, 2006).

Rice straw is another lignocellulosic material that can be used as a raw material in bioethanol production, the annual world production of which is about 731 million tonnes. This amount of biomass has the potential to produce 205 billion litre of bioethanol (Balat, Balat, and Oz, 2008). Actually, the use of rice straw as a feedstock for bioethanol production will increase the income of farmers in many places with a gain in rice production which is an important carbohydrate source for many people in the world.

Sawdust and wood chips from softwoods (pine, spruce) are another important feedstock for ethanol production. Until now most of the excess of sawdust in some countries (e. g. Sweden, Finland) has been used as a raw material for wood pellets, a solid biofuel, for heating. The annual amount of sawdust used for the production of wood pellets is more than three million tonnes in Sweden alone. In fact the wood pellets production in North America has been increased drastically in recent years. However, for sustainable usage of the forest resources in a future bio-refinery, the extractives from the biomass can be extracted for the production of chemicals, with then the possibility of releasing the cellulose and hemicellulose components and converting them to ethanol. The residual, which contains mostly lignin together with additional sawdust and other biomass, can still be used as a feedstock for the wood pellet industry.

Anaerobic bacteria can degrade a variety of organic compounds (carbohydrates, proteins, lipids, etc.). The methane content of the biogas mixture depends on the oxidative state of carbon in the compounds present in the feedstock; the more reduced the carbon is, the higher the content of biogas in methane is (Gujer and Zehnder, 1983). The feedstock should also be balanced with respect to the ratio of carbon and nitrogen (C:N = 20:30), since the microorganisms use carbon and nitrogen at this ratio range. Quite often, organic feedstocks contain these nutrients at lower or higher ratios. In such cases, the codigestion of selected feedstocks can adjust the required balance (‘diet’) and enhance biogas production, e. g. codigestion of sewage sludge with agricultural wastes or municipal solid wastes (Alatriste — Mondragon et al., 2006) as well as cattle manure with municipal solid wastes (Hartmann and Ahring, 2005). Apart from C and N, other elements present at trace concentrations are also crucial to the growth of anaerobic microorganisms. For example, Ni (involved in the synthesis of coenzyme F430), Fe (constituent of electron carriers), Mg (stabilising the cellular membranes), Ca (stabilising the cellular wall and contributing to the thermal stability of the endospores), Co (component of the vitamin B12), Zn (constituent of several enzymes). etc. If these trace elements are not contained in the feedstock, they should be supplied since their absence is correlated with decrease in efficiency (Zandvoort et al., 2006).

The definition of the term ‘allocation’ and the allocation methods used in LCA are given in the Appendix. The allocation issue arises in systems where biofuels are co-produced with other outputs, such as electricity and/or heat so that the impacts have to be allocated between the co-products in an appropriate way. As shown in Table 3.2, most international approaches favour either allocation based on energy

3.4 GHG savings for bioethanol from different feedstocks and country of origin (DfT, 2009; Edwards et al., 2008; Fehrenbach et al., 2007).

content (net calorific value) or system expansion. In the latter, the system is credited for producing the additional output. However, the methodological difficulty is in identifying the ‘correct’ way to credit the system. For example, if the electricity is co-produced with the biofuel in an EU country, the question is what electricity mix should be used to credit the system: best available technology, the average national or EU energy mix? The choice of the allocation method and the ‘credit’ are of the utmost importance as often very different results are obtained using different approaches. In any case, this should be examined as part of the sensitivity analysis (see the Appendix and the textbox).

The EU RED (EC, 2009) favours allocation based on the energy content of biofuels, although other allocation procedures, such as system expansion or

Rapeseed Sunflower Soya Palm oil Waste

Biodiesel feedstock

3.5 GHG savings for biodiesel from different feedstock and country of origin (DfT, 2007; Edwards et al., 2008; Fehrenbach et al., 2007). For legend, see 3.4.

economic value might be more appropriate in particular cases. For example, energy allocation cannot be applied in systems where biofuel co-products do not have an energy value but have an economic value, e. g. ash and fertilisers. In these cases, allocation based on the economic value may be more appropriate. However, this produces volatile results in line with economic values of commodities and should be used only where other allocation methods cannot be applied (ISO, 2006b).

An example of how to allocate environmental impacts using two different bases — mass and energy — can be found in the textbox on pages 46-47.

In an FT complex, the production of purified syngas typically accounts for 60-70% of the capital and running costs of the total plant (Dry, 2002). The most popular feedstock to provide syngas for the FT synthesis has been coal (German vehicles during the Second World War), but nowadays, natural gas is gaining in importance. Sources of gas are either large, remote reserves of natural gas or the so-called associated gas that cannot be flared any more due to more severe CO2 emission regulations (Dry, 2002; Prins et al., 2005).

Biomass has not yet been commercially applied as a feedstock for the Fischer — Tropsch synthesis (FTS); however, the integration of biomass gasification with FTS has been demonstrated (Boerrigter and den Uil, 2002). Prins et al. (2005) carried out an exergy analysis of biomass integrated gasification FTS, and the maximum thermodynamic efficiency achieved was 46.2%, consisting of 41.8% fuels and 4.4% electricity. The thermodynamic analysis showed that a mild thermal pretreatment of the biomass may improve gasification properties, that is heating value and moisture content. Although proof-of-concept of straw gasification technology scalable to an on-farm production has been demonstrated, little is known about differences among grasses in their suitability as gasification feedstock (Prins et al., 2005).

As mentioned earlier, carrying on the reaction in organic solvent of high alcohol solubility has been suggested as an answer to problem of enzyme inhibition by short-chain alcohols. Although this results in an increased rate of reaction by operating at higher concentrations of alcohol, it is not recommended since it requires additional solvent recovery unit. Supercritical CO2 (SC-CO2) offers the same advantages for lipase catalysis as organic solvents such as solubilization of the alcohol, simple recovery of the enzyme and favoring esterification to hydrolysis. In addition, SC-CO2 offers more, such as product separation and easy recovery of the solvent. Moreover, it is non-flammable, non-toxic and inexpensive. The production of biodiesel in supercritical fluids (methanol) has been reported in the literature; however, just recently the coupled use of lipase with SC-CO2 in the production of biodiesel has been reported (Rathore and Madras, 2007). Using supercritical fluids is usually expensive though, and more work is required in this regard to provide significant enhancement to the production of biodiesel and to offer biodiesel in competitive prices.