Given the current policy interest in biofuels, a large number of studies are available. The term biofuels denotes plant oils, biodiesel, bioethanol and biogas. Generally, LCA studies compare biofuels with the respective petrochemical fuels, which are gasoline, diesel and natural gas. An additional interest is to compare different biofuels, which are specified as so-called feedstock-technology combinations, i. e. process chains using a specific feedstock and a specific technology, where one technology may be feasible for different feedstocks and vice versa.

100% Rape ME CH 100% Rape ME RER 100% Palm oil ME MY 100% Soy ME US 100% Soy ME BR 100% Recycled plant oil ME CH 100% Recycled plant oil ME FR Methanol fixed bed CH Methanol fluidized bed CH Ethanol grass CH Ethanol potatoes CH Ethanol sugar beets CH Ethanol whey CH Ethanol wood CH Ethanol sweet sorghum CN Ethanol rye RER Ethanol corn US Ethanol sugarcane BR Methane grass biorefinery

> Methane manure

і Methane manure+cosubstrate

> Methane manure, optimized Methane manure+cosubstrate, optimized

Methane biowaste Methane sewage sludge Methane wood

3.3 GHG emissions for several biofuels per km for biofuels and fossil fuels (Zah et al., 2007).

One comprehensive survey including the biofuels ethanol, methanol, biodiesel and biogas has been carried out in a study commissioned in 2007 by the Swiss federal administration (Zah et al., 2007). It covered several technologies as well as biomass from domestic and from main production regions worldwide. The main results are presented in Fig. 3.3.

As Fig. 3.3 shows, the environmental performance as to global warming potential (GWP) is dependent on the type of fuel, the species and the regional origin of the feedstock. The lowest GHG emissions can be found for biofuels based on waste materials. Among agricultural feedstocks, those with high yield due to the species and the climatic conditions of agriculture show better performance; this is true, for example, for sugar cane from Brazil. The other way round, low yields per hectare and high use of nitrogen fertilizer along with the emissions of N2O (nitrous oxide) for
certain agricultural techniques, e. g. for corn from the US, lead to comparably high values of GHG emissions. The assessment of biofuels is different for other impact categories, notably eutrophication. Here, fertilizer use gener­ally causes a higher impact compared to fossil fuels but also shifts environ­mental performance between different feedstock-technology combinations. In addition, specific contributions to other impact categories exist, for example by the use of chemicals in agriculture, toxicity impacts appear.

These general findings are confirmed by the majority of studies, although controversial debates on single issues are encountered in the literature. One of these debates emerged as to the net energy ratio of starchy crops, specifically on bioethanol production from corn in the US, where some authors reported negative results and stated that more fossil fuel is consumed by production than gained as bioethanol (Pimentel and Patzek, 2005). These findings, however, were not confirmed by others (Hill et al., 2006; von Blottnitz and Curran, 2007). In contrast, all studies agree that net energy ratio as well as GWP are far better for sugar cane compared to corn, and that crop yields have a major impact on the overall results, as shown in Fig. 3.4 for GHG savings per acre depending on crop yields.

The large contribution of agriculture is also confirmed for other impact categories. This is shown, for example, for sugarcane by Cavalett et al. (2011) (Fig. 3.5). Due to the use of agrochemicals, agriculture is generally seen as the main contributor to the impact categories of human toxicity and ecotoxicity (Bai et al., 2010; Cherubini and Jungmeier, 2010).

During recent years, an additional aspect of agriculture has been recog­nized as a crucial issue for assessment of biofuels, which is the concern of land use changes (LUC) due to the rising demand for biomass. Land use change implies the direct or indirect change of not-cultivated land for agricultural use. Depending on former use, actual crop and agricultural techniques, carbon contained in the soil and plants can be released to the atmosphere, which in the worst case can jeopardize all GHG savings of biofuels. There are controversial statements regarding this topic since the first publication, but all studies agree that additional contributions to GHG must be expected from the conversion of very carbon-rich areas like peatland. This is confirmed, for example, for the case of palm oil diesel by a recent meta-analytic review of LCA (Manik and Halog, 2013).

The other important issue as to environmental performance of biofuels is the conversion technology itself. Here, several aspects have to be taken into account, primarily the efficiency of the technology. So-called second generation biofuels that make use of the full plant (and not only part of it, like seeds) are expected to show better performance, but this is also controversial. Von Blottnitz considered 47 publications that compare bioethanol production systems using LCA. Some of the LCA studies show a better environmental performance for second generation biofuels than first generation biofuels (Stoglehner and Narodoslawsky, 2009; Cherubini and Jungmeier, 2010). In addition to this, conversion technologies may provide by-products, e. g. glycerin from biodiesel production. In LCA, these are accounted for as to their substitution of fossil-based products. The way this is done on the methodological level may also be decisive for the outcome of an LCA study (see Cherubini and Str0mman, 2011).

LCA results for bio-based chemicals

Bio-based basic chemicals, often called platform chemicals, are industrially produced chemicals that are used as raw material for many industrial products. Here, only few LCA studies are available (excluding ethanol, which was included above as part of the biofuels section). In 2006 a study of biotechnological production of 21 bulk chemicals from renewable resources was carried out on behalf of the EU (Patel et al., 2006). All bio­based products were compared with the respective petrochemical product using the categories of non-renewable energy use (NREU), GWP, land use in the form of land occupations and other environmental impacts. From this, significant reductions have been identified for all bio-based products. Limited availability of data and uncertainty concerning novel processes were identified as a main drawback of the assessment. Mainly based on this study, Hermann et al. presented results for the assessment of ten bio-based bulk chemicals produced by biotechnological processes (Hermann et al., 2007): 1,3-propanediol (PDO), acetic acid, acrylic acid, adipic acid, butanol, ethanol, lysine, lactic acid, polyhydroxyalkanoates (PHA), and succinic acid. In addition to that, five products produced from the aforementioned products are included: caprolactam, ethyl lactate, ethylene, polylactic acid (PLA), and polytrimethylene terephthalate (PTT). The assessment covers waste management within the system boundary and takes into account the impact categories NREU, GWP, and land use. Results show savings as to GHG and NREU for most bio-based chemicals, already for current technologies. For future technology, it is estimated that due to learning effects the savings will be 25-35% higher. This can be explained by the relatively high energy requirement for the production of petrochemical polymers.

There are some studies focusing on individual bio-based basic chemicals. Ekman and Borjesson show that propionic acid produced from by-products of agriculture leads to significant reduction of GHG emissions compared to fossil fuel alternatives. However, the contribution of propionic acid to eutrophication is higher (Ekman and Borjesson, 2011). Glutamic acid is an important component of waste from biofuel production and an interesting starting material for the synthesis of bio-based chemicals. Lammens et al. compare the environmental impacts of four bio-based chemicals from glutamic acid with their petrochemical equivalents: N-methylpyrrolidone (NMP), N-vinylpyrrolidone (NVP), acrylonitrile (ACN), and succinonitrile (SCN). The bio-based NMP and NVP show less impact on the environment, while for the ACN and SCN the petrol-based chemicals have less impact. Further optimizations indicate that the production of bio-based SCN can be improved to a level that can compete with the petrochemical process (Lammens et al., 2011).

Uihlein and Schebek (2009) compare the environmental impact of a lignocellulose biorefinery system with conventional production alternatives. The biorefinery delivers three products (lignin, ethanol, xylite), which are compared to their fossil counterparts (see Fig. 3.6). It was found that the biorefinery has the largest environmental impacts in the three categories fossil fuel use, respiratory effects and carcinogenics. The environmental impacts mainly arise from the provision of hydrochloric acid and to a lesser extent also from the provision of process heat. The optimal variant (acid and heat recovery) provides better results than the fossil alternatives, whereas the overall environmental impact is approximately 41% lower compared to the fossil alternatives (Uihlein and Schebek, 2009).

Polymers are a main product group of the chemical industry as to the amount produced. This is why bio-based polymers are also the most interesting for biorefinery systems. Vinka et al. have compared the polylactic acid (PLA) production with the production of conventional petrochemical polymers of various kinds for 1 kg product as functional unit by means of LCA. The investigated impact categories are fossil fuel use, GWP and water demand. Fossil fuel use and GWP show significant benefits for PLA. In contrast, water demand shows a much smaller difference between the compared products (Vinka et al., 2003).

Kim and Dale have tried to estimate the ‘cradle to gate’ environmental impact of bio-based polyhydroxyalkanoates (PHA) packaging film made from crop residues (a mixture of corn grain and corn stover). The PHA production from corn grain was defined as the reference system. Compared to PHA from corn grains only, the PHA made from corn husks and corn grains shows negative GHG emissions by -0.28 to -1.9 kg CO2-eq. per kg depending on the technology used. The significant reduction can be explained by the surplus energy from lignin-rich corn stover. Photochemical smog and eutrophication are related to nitrogen-induced soil pollution. PHA fermentation technology is still immature and in the development phase. The trend shows further improvements, thus reducing the environmental impact (Kim and Dale, 2005).

A cradle-to-gate LCA for PHB production was carried out by Harding et al. For the life cycle impact assessment (LCIA) GWP and ten other impact categories were selected. The LCA results were compared with the production of polypropylene (PP) and polyethylene (PE). They show that, on one hand, the energy required for the PHB production is significantly lower than for the polyolefin production, on the other hand, the acidification and eutrophication effects are lower for PE than for PHB (Harding et al., 2007).

Roes and Patel (2007) have developed an approach, which is based on classical risk assessment methods (largely based on toxicology), as developed by the life cycle assessment (LCA) community, with statistics on techno­logical disasters, accidents, and work-related illnesses. The approach has been applied to ethanol and four polymers from cradle to grave: polytrimethylene (PTT), polyhydroxyalkanoates (PHA), polyethylene terephthalate (PET) and polyethylene (PE). The results show lower risks for bio-based polymers compared to petrochemical equivalents. However, the uncertainties in the data need to be reduced (Roes and Patel, 2007).

Alvarenga et al. investigate PVC production from bioethanol as a substitute for ethylene. Two scenarios for bioethanol-based PVC for 2010 and 2018 are compared with fossil-based PVC, using several indicators for impact assessment. As to non-renewable resource use and GWP, bio-based PVC performed better; as to other impact categories, for some assumptions it performed worse for the state of 2010. As to 2018, better results turned out due to gains in efficiency and technological

oxidation processes. This causes a general increase of costs along with increasing the oxygen content in the production (see Fig. 3.7) (Lange, 2007).

The generic advantage of renewable resources is that they are often rich in oxygen functionalities. This is why renewable resources can be used best when highly functional intermediates and polymer need to be obtained. Other authors point out that the selective deoxygenation of carbohydrates is more effective and therefore cheaper than the selective oxygenation of hydrocarbonates (Alles and Jenkins, 2010).

Luo et al. dealt with the economic and environmental analysis as well as with technical design of a lignocellulosic biorefinery (LCF), which produces ethanol, succinic acid, acetic acid and electricity. The economic analysis shows that the designed biorefinery has great potential in comparison with a sole ethanol biorefinery even if the price of succinic acid drops or the investment costs double (Luo et al., 2010).

Various studies address the costs of production across the value chain. One important aspect here is transport. Several studies use mathematical models to assess the complete costs of energy use of biomass while considering the whole supply chain (biomass production, harvest, farm and road transport and conversion plant). On the basis of the assessment results, strategies have been proposed to reduce costs. In general the transport of biomass is an important aspect, so processing biomass in the vicinity of agricultural production areas has economic advantages (Yu and Tao, 2009; Akgul et al., 2012; Giarola et al., 2011).