Edgard Gnansounou *, Arnaud Dauriat2

xBioenergy and Energy Planning Research Group (BPE), Ecole Polytechnique Federale de
Lausanne (EPFL), CH-1015 Lausanne, Switzerland
2ENERS Energy Concept, P. O. Box 56, CH-1015 Lausanne, Switzerland
*Corresponding author: Prof. Gnansounou; E-mail: edgard. gnansounou@epfl. ch

1 INTRODUCTION

During its earliest stage, the development of biofuel production in the industrialized countries was mostly driven by agricultural policies. The overproduction and low prices of crops called for diversification. Fuels derived from agricultural feedstocks were considered an ecologically valuable option for price stabilization in addition to fallowing. It was even seen as an alternative to the set-aside strategy. Two more motivations were highlighted. The perspective of oil depletion and concentration of petroleum resources in a limited num­ber of regions which are politically instable increased the concerns about energy insecurity risks. Furthermore, due to global climate change, several industrialized countries committed to reduce their greenhouse gas emissions. The transportation sector was one of the priorities for public incentives.

Indeed, that sector is vulnerable to petroleum products which represent 98% of its final energy consumption worldwide. The high volatility of oil prices and the low competitiveness of sustainable biofuels when oil prices decrease under a certain threshold, make a claim for stable incentives in the early development stage of biofuel markets. However, the fast growth of biofuel production and the rise of the prices of agricultural commodities in 2008 fed some controversies about the sustainability of biofuels. In addition to the risks of competition with food and animal feed, the energy and greenhouse gas (GHG) balances of biofuels were debated. In response to the spread of reluctance to continue supporting publicly the utiliza­tion of biofuels, public authorities in several countries have imposed minimum sustainability targets for biofuels to be eligible for incentives (Escobar et al., 2008; Van Dam et al., 2008).

As an example, on 23 April 2009, the European Union issued a Directive on the promotion of the use of renewable energy with the requirement that each Member State shall ensure by 2020 at least 10% sustainable renewable energy in the final energy consumption of its transport sector. Article 17 of that directive notified the minimum sustainability criterion for biofuels (EC, 2009). For instance, concerning GHG emission reduction with respect to fossil fuels, increasing minimum targets were imposed: 35% in the year of entry into force of the Directive, 50% in 2017, and 60% for biofuels produced from plants that will start from 2017 onward.

The concerns about energy balances are related to both the life-cycle energy efficiency of biofuels and the saving of nonrenewable energy between biofuels and fossil fuels. The latter aspect is relevant with respect to the substitution efficiency of biofuels.

Monitoring the application of minimum targets on GHG emission reduction to biofuels, as well as estimating their substitution efficiency with respect to fossil fuels, is subject to sig­nificant uncertainty and inaccuracy associated with the methodology applied. Assessments of the environmental impact of biofuels (ADEME-DIREM-PWC, 2002; ADEME, 2010; Beer and Grant, 2007; CONCAWE-EUCAR-JRC, 2008; Elsayed et al., 2003; EMPA, 2007a; GM-LBST, 2002; Gnansounou and Dauriat, 2004, 2005; Kim and Dale, 2008; Macedo, 2004; Malca and Freire, 2006; Shapouri et al., 2002; VIEWLS, 2005; Wang, 2005) often signifi­cantly differ in methodological choices and consequently in their results. Table 1 shows an overview of the methodological choices in these studies.

For example, while some studies (Elsayed et al., 2003) are limited to a Well-to-Tank (WtT) approach (thereby excluding the utilization phase), other studies employ a Well-to-Wheel (WtW) approach. When included in the system, the utilization phase is taken into account either in a simplified way (usually by merely considering the difference in the LHV of fuels) or with more details (by considering the actual performance of fuels according to a specific engine technology and/or fuel blend). As far as the functional unit is concerned, the distance traveled (in km) is the unit of choice in most studies (in agreement with the principles of the WtW approach). Allocation by system expansion is the most widely used method, although in some studies a combination of methods is used. For instance, system expansion is combined with allocation by mass in ADEME-DIREM-PWC (2002). Economic allocation is the second most common approach. Fuel blends considered vary from one study to the other (usually between 5% vol. and 15% vol.), depending on the most frequent use of fuel-ethanol in the region of study. All the reviewed studies however also consider ethanol as a fuel component on its own, even though the way this is done does not depend on the actual fuel blend but rather on the difference in the LHV of fuels. Finally, land-use change is included with details in only a few studies (EMPA, 2007b; Elsayed et al., 2003; IFEU, 2004), based on IPCC (2003a) guidelines.

In the particular case of GHG balance, the magnitude of the discrepancy among the results is tremendously high. Farrel et al. (2006), based on a review of corn-based ethanol studies in the USA, attributed the main differences to the way coproducts are accounted for, the value of some input parameters, and some omission/inclusion of ambiguous inputs. Reijnders and Huijbregts

(2003) focused on forest-based biofuels, analyzing the effect of the considered time frame on the emission factors, the choice of previous land use, the allocation of carbon sequestration and emissions during forest growth, and the fate of sequestered carbon after fuel wood harvesting. Bhrjesson (2009) focused on methodological choices and the influence of local conditions in wheat-based ethanol production in Sweden. He addressed the problem of coproducts allocation choices, the choice of the fuel used, and biogenic GHG emissions from cultivated soils.

In fact, quantitative investigations of the differences in the results from one study to the other are not straightforward due to the lack of information concerning the inventory data,

TABLE 1 Comparison of Methodological Choices in Reviewed Studies

System Approach Well-to-Tank

the assumptions made to complement unavailable data and modeling choices about system definition and boundaries, functional units, reference systems, and allocation methods. In the research presented in this chapter, an assessment platform was developed based on an exten­sive review of literature. Combinations of assumptions and modeling choices were defined to investigate the sensitivity of the results to several factors. The focus is mainly put on choices regarding the allocation method, the previous land use, the fuel blends, and the vehicle/fuel performance. This chapter aims to contribute to current discussions on how methodological choices and local conditions influence LCA results, addressing some important points often limitedly treated in the literature. The chapter considers wheat-based bioethanol production in Switzerland as a case study, with the aim of quantifying the variation in GHG emissions and nonrenewable energy use depending on methodological choices. The chapter is an updated version of a previous paper by Gnansounou et al. (2009).