It is reported in various publications and press articles that biofuels (incl. bioethanol and biodiesel) do actually contribute to global warming (Fargione et al., 2008; Reijnders and Huijbregts, 2008; Searchinger et al., 2008) under certain conditions. Other WtW studies, such as that of Beer and Grant (2007), found only marginal advantages for E10 blends (of the order of 4% reduction on GHG emissions). Looking at the results presented in this chapter so far, however, it does not seem to be the case for bioethanol from wheat produced in the Swiss context. This is not always true and is really the result of the default choices made in this chapter (which correspond to the most likely situation in the present European context). Let us assume A-2 as the allocation method (i. e., economy) and E-3 as the vehicle/fuel per­formance option (i. e., energy basis) and evaluate the net GHG emissions of bioethanol from wheat under various land-use change scenarios. This framework is among the most unfavor­able set of options and actually corresponds to that of EMPA (2007a). Economic allocation is the default method in the ecoinvent database and the method chosen by the Swiss authorities to evaluate the sustainability of fuels in the frame of the Ordinance on the ecological balance of fuels. Vehicle/fuel performance based on the energy content of fuels is undoubtedly the most frequent hypothesis in LCA studies of biofuels (e. g., CONCAWE-EUCAR-JRC, 2008; EMPA, 2007a; GM-LBST, 2002; IFEU, 2004; VIEWLS, 2005). The results are presented in Table 5 and illustrated in Figure 5.

Under these assumptions, the variation of life-cycle GHG emissions with respect to gaso­line ranges from —24% to +383%. Unless the land-use change leads to an improved annual carbon stock (i. e., LUC-5 LUC-8 and LUC-9 as shown in Table 4), the net GHG emissions of

wheat to ethanol are indeed larger than those of gasoline. In the worst scenario (LUC-6, i. e., forested land to cultivated land), the net GHG emissions of bioethanol can be as large as

3.8 times that of gasoline.

2 CONCLUSIONS

The default case in this chapter for fuel-bioethanol production from wheat in the Swiss context considers allocation based on energy content (A-1), the switch from set-aside land to cultivated land (LUC-1) and vehicle/fuel performance based on actual vehicle test with fuel-ethanol used as E5 (E5-1). With net GHG emissions of 0.066 kg CO2 eq./km (i. e., —72% with respect to gasoline) and a net energy use of 0.567 MJp/km (i. e., -84% with respect to gasoline), this default case may seem particularly advantageous compared to other simi­lar studies. This default case, however, is realistic and corresponds to the most likely situa­tion in the European context. Energy allocation is the methodology adopted by the European Union in its Directive on the promotion of the use of energy from renewable sources. Set-aside to long-term cultivated is a reasonable option when considering the pro­duction of biofuels from agricultural crops. Finally, fuel ethanol in the EU is mainly used as E5 at present. If the set of methodological choices as in EMPA (2007a) is applied to the same system, meaning economic allocation (A-2) and vehicle/fuel performance based on the energy content of fuels (E-3), the resulting net GHG emissions and net energy use are 0.273 kg CO2eq./km (i. e., +15% with respect to gasoline) and 1.944 MJp/km (i. e., —44% with respect to gasoline), respectively. These results are much more unfavorable and signifi­cantly different from those of the default case.

Various authors have demonstrated the significant effect of methodological choices on the GHG and energy balance of biofuels through review papers and other similar studies (Borjesson, 2009; Farrell and Sperling, 2007; Reijnders and Huijbregts, 2003). The present chapter quantifies these effects, based on a case study concerned with the production of fuel-ethanol from wheat in the Swiss context. In addition, it demonstrates and quantifies the effects of the fuel blend and the choices regarding vehicle/fuel performance.

The results presented in this chapter show a large variation of the net GHG emissions of wheat-based ethanol for transportation with a high sensitivity to the following factors: the method used to allocate the impacts between coproducts, the type of reference systems, the type of land-use change, and the type of fuel blend. Depending on the allocation method (energy content, economy, dry mass or carbon content), the net GHG emissions of ethanol may vary by a factor of up to 2.6 (with carbon content being the most favorable and economy the least favorable). When substitution is applied, the net GHG emissions of ethanol may even be negative when both straw and DDGS are used as fuels, thereby making the difference even more significant. Depending on the land-use change situation, the net GHG emissions of ethanol may vary by a factor of up to 6.4. Similarly, the hypotheses regarding actual fuel blends and vehicle/fuel performance may result in a variation of net GHG emissions by a factor of 2.2. Depending on the combinations of methodological choices and land-use change situations, the variation of life-cycle GHG emissions with respect to gasoline may range from —112% to +120% for the same ethanol production pathway.

In face of missing data and time stress, many studies use pragmatic approaches to evaluate the energy and GHG balance of biofuels. Thus, several studies are not transparent enough and methodological choices can turn a positive GHG balance into a negative one and vice versa. As policymakers will take decisions by using these results, it is important to establish the rationale of the evaluation methods. Some items need further research works, for exam­ple, rationale of allocation methods, indirect land-use change (Gnansounou et al., 2008b). Others are till now subject to low transparency and consistency requirements.

Especially concerning the boundaries of the system, the authors recommend to use a WtW approach. One should not mind if the implementation of the WtW should be simplified; uti­lization stage must be taken into account as long as comparison of different qualities of fuels is concerned, that is, fuels associate with different mechanical efficiencies. The functional unit must be appropriate, reflecting the fact that these fuels must be compared for the same service (e. g., the distance traveled). Finally, for transparency purpose, the reference system must be explicitly defined.