In the debate on the future of transport biofuels, environmental matters have also become important. Slowing down climate change is often mentioned as an impor­tant reason for expanding the use of biofuels, as CO2 that is released on burning this biomass is supposed to be rapidly sequestered again by re-growth of biomass. How­ever, this is not the whole story. Because fossil fuels are used for powering biofuel production, biofuel production may be associated with the emission of greenhouse gases other than CO2, such as N2O and CH4, and biofuel production can be associ­ated with changes in the carbon content of ecosystems. Thus, a lively discussion has originated on whether promoting transport biofuels does indeed slow down climate change. And there is a longstanding discussion as to whether the overall environ­mental impacts of biofuels are positive (e. g. Healy 1994), which has focused on the impacts of agricultural chemicals used in biomass production and water use. Also, the impact of transport biofuel production on nature has emerged as an issue. In particular, the importation by industrialized countries of palm oil from Southeast Asia, of biofuels from South America and the cutting of tropical forests for the es­tablishment of biofuel plantations in Africa have sparked a debate on the impact of transport biofuels on living nature. In turn, environmental concerns contribute to the emergence of regulations and certification schemes that aim to address such concerns (Mathews 2008b; van Dam et al. 2008).

On the basis of available data, it is also possible to draw general conclusions as to the question of whether categories of transport biofuels are better or worse regarding life cycle greenhouse gas emissions than conventional fossil transport fuels, such as diesel and petrol. The latter have a life cycle greenhouse gas emission of about 3.6kg CO2 equivalent kg-1 fuel (EUCAR et al. 2007; Reijnders and Huijbregts 2008b). In the following, comparison between fuels will be made on the basis of equal energy generation.

Transport Biofuels from Crops on Peat

Transport biofuels from crops grown on peat land do badly as to greenhouse gas emissions. This is linked to the large carbon losses from peat land when used for cultivation. This has been shown for bioethanol in Europe (Reijnders and Huijbregts 2007) and for palm oil from Southeast Asia (Danielsen et al. 2008; Fargione et al. 2008; Reijnders and Huijbregts 2008a). This will also hold for produce of the sago palm, which often grows on peat land (Melling et al. 2005a, b; Singhal et al. 2008). Sago has been called the ‘ starch crop of the twenty — first century ’ and suitable for the production of the transport biofuel ethanol (Singhal et al. 2008). The yield of sago plantations is 2-3 Mg starch year-1ha-1 (approximately 0.6-0.9 Mg C) (Singhal et al. 2008), but the annual loss of C from peaty soil is approximately 11 Mg C year-1 ha-1, partly as methane (Melling et al. 2005a, b). When mineral soils have a high carbon content, as in the case of peaty clay, net greenhouse gas emissions will also be relatively high (Reijneveld et al. 2009), making transport biofuel production on such soils relatively unattractive for mitigating climate change.

As pointed out in Chap. 1, a variety of options for producing biofuels from marine biomass have been suggested, such as biofuels from Macrocystis pyrifera or giant kelp (Wilcox 1982; Bungay 2004), Laminaria (Hornetal. 2000; Chopin etal. 2001) and Dunaliella (Ben-Amotz et al. 1982). Dunaliella has been found more suitable to cultivation in open ponds (Joint et al. 2002; Ugwu et al. 2008). As to Macrocystis pyrifera, it seems doubtful whether the energy balance for biofuel can be positive (Bungay 2004).

Near-shore cultivation of macroalgae is substantial (Neushul and Wang 2000; Wikfors and Ohno 2001; Chopin et al. 2001; Critchley et al. 2006; Troell et al.

2006) . For Gracilaria in Taiwanese coastal waters, average yields of 4Mgha-1 year-1 (dry weight) have been reported (van der Meer 1983). Yields of commer­cial Eucheuma cultivation in the Philippines, Indonesia and Kiribati are about 6 Mg (dry weight) ha-1 year-1 (Ask and Azanza 2002). Such yields suggest relatively low solar energy conversion efficiencies if compared with cultivated terrestrial plants (see Table 2.1). Cultivation is vulnerable to invasions of competing algae and her­bivores, and major interventions may be necessary to limit losses in such cases (Buschmann et al. 2001; Ask and Azanza 2002; Neill et al. 2006). As pointed out in Chap. 1, prices for cultivated macroalgae are high, and thus it is hard to see the emergence of a practical large-scale biomass-from-the-sea-for-transport-fuel scheme based on macroalgae cultivation (Neushul and Badash 1998; Buschmann etal. 2001).

The possibility exists that crops that are to serve as lignocellulosic feedstocks for transport biofuel production may turn out to be invasive species. The selection for ‘weedy characters ’ in such species, which allow for cultivation on marginal lands with relatively low inputs of nutrients, is conducive to such risk (Barney and DiTomaso 2008). The impacts on ecosystem services of invasions by species involved in biofuel production are strongly dependent on the nature of the invader and the extent of the invasion. However, effects may be considerable. One of the species considered for lignocellulosic biomass production is reed canary grass (Phalaris arundinacea L.). This grass species is able to invade wetlands and impact their hydrology. For the most part, the outcomes of such invasions are considered detrimental (Zedler and Kercher 2004). Stream banks may also be invaded by reed canary grass (Lavergne and Molofsky 2004). There is, furthermore, some evidence that sweet sorghum and giant reed (Arundo donax) are invasive in specific ecosystems in the USA (Barney and DiTomaso 2008; Royal Society 2008). Jatropha curcas, a source of biodiesel, is considered as invasive in South Africa and as weedy in Australia (Achten et al. 2009).

Also, the production of seaweeds for biofuel production may give rise to invasive species. A case in point is the macroalga Kappaphycus alvarezii. This native to the Philippines has given rise to invasions of coral reefs in Hawaiian and Indian waters as an unintended effect of cultivation (Bagla 2008).

A wide variety of plants produce lipids that can be used as a basis for transport fuels. Main suppliers of lipids are currently plants producing edible oils such as rapeseed or canola, sunflower, oil palm, coconut and soybean. There is also limited use of non-edible lipids such as Jatropha oil and of animal fats (‘yellow grease’) and very limited use of algal lipids. A number of additional potential vegetable sources of lipids have been suggested (Sims et al. 2006; Shao and Chu 2008), and lipids have also been produced microbially from sugars (Zhou et al. 2008). The use of edible oils often has the advantage that co-products may be used in ani­mal feed production. This may be different for non-edible oils. For instance, the oil cake of many of the current Jatropha varieties is not suitable for feeding live­stock because of the presence of toxic compounds such as phorbolester and curcin (Carvalho et al. 2008; Sujatha et al. 2008), but such toxic Jatropha oil cake can be anaerobically converted to methane (Achten et al. 2009) or burned to supply en­ergy.

The lipids used for biofuel production mainly consist of triacylglycerol, in which the acyl groups are fatty acids (Agarwal and Agarwal 2007). In principle, a variety of lipids can be burnt as such in diesel motors, more of them in warm climates. Vegetable oils are used to a limited extent as transport fuel. For instance, there is a significant use of coconut oil in motorcars on the Pacific islands (Cloin 2007), and in Europe there is limited use of rapeseed oil in heavy-duty vehicles. However, for most applications, viscosity is too high. This problem may be solved by dilution, microemulsification and transesterification (Canakci and Sanli 2008). In practice, the solution is mostly transesterification to produce biodiesel, which is compatible with fossil diesel fuel and leads to a more limited increase in maintenance costs. In transesterification, the glycerol OH groups are replaced by the OH groups of either ethanol or, more commonly, methanol.

Transesterification to produce biodiesel from lipids can proceed with the help of an inorganic base catalyst (e. g. NaOH, KOH or NaOCH3). This approach is widely applied in commercial biodiesel production (Canakci and Sanli 2008). Potential al­ternatives are the use of insoluble inorganic catalysts (Shu et al. 2007; Li et al. 2007; Vasudevan and Briggs 2008) and the use of an enzyme: lipase (Harding et al. 2008). These alternatives are under development (Abdullah et al. 2007; Ranganathan et al. 2008). Transesterification by superheated or supercritical alcohols (that are not sen­sitive to free fatty acids and water) has also been studied (Marchetti et al. 2007; Joelianingsih et al. 2008). In the case of lipids that are characterized by the presence of greater than 0.5-1% free fatty acids (that react with base catalysts to soap) — of­ten waste lipids — the use of both homogeneous and heterogeneous acid catalyzed transesterification has been advocated (Abdullah et al. 2007; Vasudevan and Briggs 2008; You et al. 2008; Canakci and Sanli 2008; Park et al. 2008). Alternatively, free fatty acid levels can be reduced to less than 0.5% by the use of ion-exchange resins (Ozbay et al. 2008) or the admixture of virgin lipids.

An alternative option to transesterification is to catalytically remove oxygen from the triacylglycerol, while adding hydrogen. This gives rise to the synthesis of propane and mixtures of hydrocarbons (paraffins) that have diesel-like properties and can also be used in kerosene blends (Holmgren et al. 2007). Such a deoxy­genation process is currently commercially exploited (Rantanen et al. 2005). Still another way of converting virgin or used vegetable oil in transport fuels uses cat­alytic cracking or pyrolysis (heating in the absence of oxygen). In the latter case, this has to be followed by upgrading of the bio-oil that is a product of pyrolysis. In this way, one may produce fuels that are suitable for application in diesel or Otto motors or as a substitute for kerosene that is applicable in air transport (Milne et al. 1990; Knothe 2001; Demirba§ and Kara 2006; Dupain et al. 2007; Ooi and Bhatia 2007; Tamunaidu and Bhatia 2007). It has been noted that in the case of cracking unsatu­rated lipids, the product may contain relatively large amounts of aromatics (Dupain et al. 2007). Also, alkane synthesis from lipids by the bacterium Vibrio furnissii has been reported (Fortman et al. 2008). Finally, there are efforts to produce fatty acid ethylesters and hydrocarbons by re-engineering metabolism in heterotrophic micro­organisms (Wackett 2008). Ultimately, this way of producing biofuels is critically dependent on cheap sources of carbohydrates (Rotman 2008).

In Germany, a major producer of biodiesel, glycerol that is generated by trans­esterification of oils and fats is anaerobically converted into methane (see below). Glycerol can also be gasified to synthesis gas (mainly CO and H2). Synthesis gas (also called ‘syngas’) may be converted into methanol. Methanol, in turn, can be mixed into gasoline up to 15% by volume and applied in Otto motors without major adaptations. Methanol can also be used for the production of biodiesel, as an admix­ture in diesel (Cheng et al. 2008) and to produce methyl tert-butylether (MTBE) to be applied in petrol for use in Otto motors or to produce dimethylether (Arcoumanis et al. 2008). Methanol can, moreover, be reformed on board means of transport in a way that fits the use of H2 in fuel cells (Ferreira-Aparicio et al. 2005).

Converting glycerol into methanol via syngas is now commercially applied. Fur­thermore, synthesis gas derived from glycerol may be turned — via the Fischer — Tropsch reaction — into hydrocarbons that may serve as diesel, petrol or kerosene (Scott et al. 2007; Simonetti et al. 2007; Valliyappan et al. 2008). Also, syngas may be used as a source of H2 (Yazdani and Gonzalez 2007; Valliyappan et al. 2008). There are other options for converting glycerol into transport biofuels, too. Anaer­obic fermentation may convert glycerol into ethanol and/or butanol (Coombs 2007; Yazdani and Gonzalez 2007). And glycerol may be converted into propanol, which can be mixed with conventional gasoline (Coombs 2007; Fernando et al. 2007).

Studies regarding the prospects for future modern biomass production tend to rely fully or overwhelmingly on land as the place where biomass is grown for this pur­pose. Studies with high estimates regarding the technical potential of biomass sup­ply often have most of that potential met by energy crops that have high yearly yields per hectare (Hall and Rosillo-Calle 1998; Berndes et al. 2003; Hoogwijk et al. 2003; de Vries et al. 2007). Hoogwijk et al. (2003), for example, assumed a dry weight productivity of biomass-for-energy plantations on surplus agricultural land of up to 20Mgyear-1ha-1. Such yields may be achieved. Actual experience shows that breeding efforts may increase biomass yields (dry weight) in the range of 6.7-11.3Mgha-1year-1 to yields greater than 16Mg(Volk etal. 2003). And there is much research aiming at further yield increases, for example, by lengthening the growing season without risking frost damage, limiting remobilisation of nutrients following senescence and improving drought resistance (Karp and Shield 2008). However, in general, highly productive species and varieties tend to be relatively inefficient in their resource use (Wood 1998), which is not in line with sustainable resource use (Pimentel et al. 2002; Reijnders 2006). Indeed, sustainable productiv­ity is limited due to restrictions on water and nutrient use and the need to maintain adequate soil carbon levels.

In a first approximation to the levels of biomass production that may be produced in a sustainable way on land, it would seem useful to focus on natural net primary production (NNPP), which varies geographically (Havstad et al. 2007; Campbell et al. 2008). Kheshgi et al. (2000) estimated average natural NNPP on land at 4 Mg (= 106 g) of dry biomass per year. Campbell et al. (2008), who studied abandoned agricultural soils, estimate that potential production rates on such soils average 4.3 Mg dry biomass ha-1 year-1. As pointed out earlier in this chapter, it may well be that recycling nutrients in the case of biofuel use is less efficient than in natural systems and that a part of carbon fixed in NNPP may be necessary to maintain soil organic carbon in a steady state. Thus, it is likely than on average, a lower amount of biomass can be harvested sustainably than 4-4.3 Mgha-1year-1. Pimentel et al. (2002) have suggested that in tropical and temperate areas, on average, approxi­mately 3 Mg ha-1 of woody biomass can be harvested in a sustainable way per year.

Again, there are geographical differences in sustainably harvestable biomass due to climate and water and nutrient availability (e. g. Nabuurs and Lioubimov 2000; Gough et al. 2008).

To get an idea of what a sustainable yield of feedstock may mean for energy sup­ply, it would seem interesting to focus on agricultural land that has been abandoned (including currently fallow land). Field et al. (2008) and Campbell et al. (2008) es­timate that the total area of such land is about 385-472 x 106 ha. We further assume that, after restoration of nutrients and soil organic matter, on these lands, a yearly sustainable yield of about 3 Mg (Pimentel et al. 2002) biomass with a lower heating value of 20MJkg-1 (Field et al. 2008) may be achieved. This would correspond with about 23-28 EJ (= 1018 J) year-1. As pointed out in Chap. 1, use of primary energy for the transport sector is currently about 100 EJ.

Another option that may be considered in the context of sustainable supply re­gards biofuels produced from what are currently ‘wastes’, such as organic urban wastes, biomass from forest remediation and residues from forestry and agricul­ture which are not used as animal feed. The worldwide amount of such wastes is currently estimated at between 50 and 100 EJ (Swedish Environmental Advisory Council 2007; Lysen and van Egmond 2008). Unfortunately, it is not clear how much thereof is necessary for maintaining the future productivity of arable lands and forests in line with the sustainability requirements for soil organic matter dis­cussed in Sect. 3.2. However, even when only 10-20% thereof could be diverted to transport biofuel production, this would represent a substantial contribution to the transport fuel supply.

Social concerns have been raised, too, in the context of expanding transport bio­fuel production. These relate to land tenure, especially by native and small farmers confronted with expanding large-scale cropping of biofuel feedstocks, to the fate of such farmers at the hands of oppressive governments favouring large-scale bio­fuel projects, to labour relations, to working conditions and to the exploitation of child and migrant labour (Cooke 2002; Nicholls and Campos 2007; Smeets et al.

2008) . Such social concerns have, for example, been raised about Burma, Malaysia, Indonesia, Brazil, Colombia and parts of Africa, in the context of ethanol produc­tion from sugar cane and palm oil and Jatropha oil production (Oxfam 2007; Eth­nic Community Development Forum 2008; Gross 2008; Mayer 2008; Smeets et al. 2008). Social concerns contribute to the emergence of regulations and certification schemes that aim to address such concerns (Mathews 2008b; van Dam et al. 2008).

This Book

The next chapters of this book will deal with matters that have a bearing on the debate about the future of biofuels with a main focus on environmental issues. Chapter 2 deals with cumulative fossil energy demand and solar energy conver­sion efficiencies of transport biofuels and of other ways to convert solar radiation into usable energy. These are important for the potential to displace fossil fuels and the area required for biofuel supply. Chapter 3 takes a look at the use of non-energy resources for transport biofuel production, such as water, plant nutrients and fertile soils. Chapter 4 considers the emissions linked to the life cycle of biofuels. Chap­ter 5 discusses the impact of transport biofuel production on living nature. Chapter 6 looks at the future of transport biofuels in view of the previous chapters and tries to answer questions that are frequently asked about biofuels.

Biodiesel made from vegetable oil which may also serve as food or feed is usu­ally inferior to conventional diesel, when the arable soil is subject to tillage without addition of large amount of fresh C (e. g. residues, manure) and land use changes are factored in. Ethanol from starch and sugar crops is usually inferior to conven­tional gasoline when arable land is subject to tillage, ethanol production is powered by fossil fuels, land use change is factored in and Intergovernmental Panel on Cli­mate Change (IPCC) guidelines are applied. It has been argued that no foreseeable changes in agricultural or energy technology will be able to achieve meaningful benefits as to the emission of greenhouse gases if annual crop-based biofuels are produced at the expense of tropical forests (Gibbs et al. 2008).

Among the transport biofuels from sugar and starch crops, ethanol from sugar cane does relatively well. For Brazilian sugar-cane-derived ethanol produced in 2005/2006, a greenhouse gas emission has been published of somewhat more than 0.4 kg CO equivalent per litre ethanol (Macedo et al. 2008). The latter energetically equals about 0.751 gasoline, which has a life cycle CO2 emission of about 2.5 kg CO2 equivalent. The estimate of Macedo et al. (2008) includes an N2O emission from soils, though linked to a lower input of N fertilizer than used by Machado et al. (2008), and a greenhouse gas emission linked to burning ‘trash’. Changes in aboveground and belowground C were not accounted for by Macedo et al. (2008), and the value for the emission of N2O used by them was well below the 3-5% of N input suggested by Crutzen et al. (2007).

There is no clarity regarding the impact of sugar cane cultivation on soil C stocks. It has been suggested that a conversion of rainforest to pasture and then to sugar cane plantation reduced the soil C stock by about 40% (Groom et al. 2008). But there are also reports that current practices are not associated with reductions in soil C of current arable soils with limited tillage (La Scala et al. 2006; de Resende et al.

2006) . However, a major loss of carbon is associated with converting the wooded Cerrado to arable land for growing sugar cane. Fargione et al. (2008) estimate that it will take 17 years to pay back the carbon debt for this by producing sugar cane ethanol. If the IPCC guidelines are followed, ethanol from sugar cane for which the Cerrado savannah has been cleared would probably do somewhat better than fossil gasoline. If arable land remains in use for many decades, sugar cane ethanol will do substantially better than conventional gasoline.

Most proposals for microalgal biofuels from open ponds or bioreactors focus on biodiesel made from algal oil (Scragg et al. 2002; Chisti 2007; Huntley and Redalje 2007; Wijffels 2008; www. oilgae. com; Liu et al. 2008). However, there have, for instance, also been proposals to convert algal biomass into methanol via synthesis gas or into bio-oil via pyrolysis (Hirano et al. 1998; Sawayama et al. 1999). The alga Botryococcus braunii has been looked into, in view of its ability to produce substantial amounts of hydrocarbons, which may be turned into transport biofuels by catalytic cracking (Bachofen 1982; Banerjee et al. 2002). As pointed out in Chap. 1, current strains of this microalga are slow growing, which has not been conducive to its application (Banerjee et al. 2002).

Of the microalgae commercially grown in open ponds, Spirulina apparently has the best yields per hectare per year in commercial cultivation (Belasco 1997). Max­imum productivities in open ponds are achieved under tropical or subtropical con­ditions (Jimenez et al. 2003). Yields currently obtained in industrial facilities for the cultivation of Spirulina located in these regions range from 10 to 30 Mg dry biomass per hectare per year (Vonshak and Richmond 1988; Jimenez et al. 2003). Low yields of, for example, Spirulina may however occur due to, for example, phage infections and rainfall conducive to the growth of unfavourable organisms (Shima — matsu 2004). For instance, Li and Qi (1997) reported that the 80 Chinese Spirulina production plants had production on average of 3.5 Mg ha-1 year-1.

It may be that in the future, microalgal yields from raceway ponds may be in­creased over current levels, for instance through improving photosynthetic activ­ity by minimizing light harvesting chlorophyll antenna size (Neidhardt et al. 1998; Mussgnug et al. 2007). On the other hand, a focus on algal lipids for transport bio­fuel production may well lead to biomass yield limitations, because nutrient lim­itations are conducive to high lipid contents but not to maximizing biomass yield (Wijffels 2008; Liu et al. 2008).

Hirano et al. (1998) studied Spirulina production and processing to supply methanol (via synthesis gas) and assumed a yield of approximately 110Mgha-1 year-1. When both fossil fuel inputs in infrastructure and operation are considered, this would correspond with an overall solar energy to biofuel conversion efficiency of about 0.12%.

Actual yearly yields much exceeding 30Mgha-1year-1 have been claimed for microalgae growing in water that has been saturated in CO2 (Kheshgi et al. 2000; Wang et al. 2008). Algal ponds that are to be saturated in CO2 have been proposed to capture the CO2 of power plants (Kheshgi et al. 2000). Also, closed bioreactors have been proposed for algal capture of CO2 from power plants (Skjanes et al. 2007). The efficiency of algal CO2 capture in open ponds has been estimated to be in the order of 30% (Benemann 1993; Kadam 2002), whereas an efficiency of 40% has been suggested for algae in photobioreactors (Ono and Cuello 2006). Whether such per­centages can be achieved is not certain. Yields from open ponds saturated with CO2 have proved disappointing, and maintaining desired algal cultures in such ponds has turned out to be difficult (Benemann et al. 2003). There is also the matter of the efficiency of CO2 sequestration by algae. The suggested efficiency for photobiore­actors of 40% is, for instance, higher than efficiencies so far reported by Hsueh et al. (2007) and Jacob-Lopes et al. (2008) for flue gases with high concentrations of CO2 handled by photobioreactors. Moreover, the latter efficiencies were achieved under good irradiation, whereas the CO2 emission of power plants may also occur at night and when solar irradiation is poor. CO2 capture and sequestration (CCS) in aquifers or abandoned natural gas or oil fields would be able to reduce the emission of power plants with an efficiency of about 90% (Odeh and Cockerill 2008). Thus, whether the application of CO2 capture by algae will be important in the future depends to a large extent on the emission requirements for such plants.

Microalgal yields from closed bioreactors subject to solar irradiation may be much higher than from current commercial open ponds (Eriksen 2008). For the production of algal oil, a value of about 16Mgha-1year-1, has been suggested as ‘possible with state of the art technology’ in closed systems (Wijffels 2008). However, growing algae aiming at high outputs in bioreactors requires large inputs of energy for building the reactors and for nutrients and intensive mixing. It has been estimated that this could lead to a negative energy balance for flat panel bioreactors and an even more negative energy balance for tubular bioreactors (Wijffels 2008).

6.1 Introduction

As pointed out in Chap. 1, there is a lively debate about the future of transport biofuels. In this debate, many matters have been raised. Some of these refer to the question of whether biofuels do what they have been promised to do, e. g. increase energy security and tackle climate change. Others refer to the side effects of trans­port biofuel production, for instance on biodiversity, natural resources such as water and food prices. And there are questions that relate to what governments should do about transport biofuels and how to proceed with specific feedstocks: should they be processed in biorefineries, or rather be converted into one biofuel?

In this chapter, we try to answer several of the questions that have frequently been raised in the transport biofuel debate. In doing so, we will draw on the previous chapters.

The frequently asked questions that we try to answer in this chapter are:

• Should the focus be on one-fuel output or on multi-output biorefineries?

• Can transport biofuels significantly contribute to energy security?

• What is the effect of transport biofuel production on food security and food prices?

• Is expanding biofuel production a good way to tackle climate change?

• What is the effect of biofuel production on nature conservation?

• How to use natural resources in biofuel production in a sustainable way?

• What government policy should one aim at for biofuels?