Planning multi-feedstock supply for a new territorial biorefinery

This study evaluates the possibility to create a feedstock supply area in Tuscany (NUTS2 region of Italy) that can sustain a territorial biorefinery (see Ceapraz et al., 2016, for a review). The feedstock includes four industrial crops, i.e. hemp, camelina, safflower, and flax, that have proved to be suitable for cultivation in Tuscany. The bioconversion of those crops (pre-treatment in a dedicated plant and further processing into the biorefinery) delivers a variety of value-added materials for the bio-based product market, such as biochemicals or biopolymers, among the others (e.g., Mohanty et al., 2002; Schneider, 2006; FitzPatrick et al., 2010; Iskandarov et al., 2014), thereby contributing to technological innovation (Wellisch et al., 2010). The combination of the biorefinery with a biogas-to-energy plant allows close the material cycle, by recovering energy and fertilizing potential from the by-products of the bioconversion (Kaparaju et al., 2009; Cherubini, F., 2010). The four crops under study can be included in traditional rotations with cereal and fodder crops, which make up most farming systems in Tuscany, or cultivated on marginal land. The introduction of industrial crops would offer farmers the opportunity to diversify the crop mix, raise farming profitability, and reduce uncertainty. Needing to secure continuous and optimal operations, biorefinery operators are willing to stipulate biomass procurement contracts with farmers, thereby providing them with a fixed price over a long-time horizon, generally a decade or more (e.g., Kumarappan and Joshi, 2014). Farmers with greater entrepreneurship might also decide to adopt a pre-treatment plant, thereby improving their contractual power. Agricultural residues of other crops within the rotation may also undergo bioconversion. Delivering those residues to the biorefinery would cut transaction costs at the farm level, by reducing the number of buyers. Cropping on marginal land is another way to increase farm profit (Fahd et al., 2012). Despite the popularity of the bio-based economy and the growing research interest in that field, few ex ante evaluations focus on prospected investors in bio-based industries and prospected suppliers of biomass in Italy. Many Authors deal with broad issues and feed the debate around the sustainability of bio-based industries (see Parajuli et al., 2015). The academic literature shows that various methodologies are applicable to the economic analysis of biorefinery supply, e.g. input-output analysis, multi-criteria decision making, cost-benefit analysis, cost-effectiveness analysis (Ceapraz et al., 2016). Authors have already analysed the potential of selected geographical areas in terms of biomass production. For example, Forster-Carneiro et al. (2013) uses official statistics to evaluate the suitability of the national supply of agricultural residues and livestock waste to the implementation of sustainable biorefineries in Brazil. Mapemba et al. (2007) calculate the costs associated with the upstream phases (harvest through delivery to the biorefinery) of bio-based products originated from lignocellulosic biomass in the USA. Also in the USA, Judd et al. (2012) concentrate on the transportation problem. Results from Brazil of the USA are hardly applicable to the Italian context, given the great existing differences, e.g. in pedo-climatic characteristics, average extension of farmland, and proximity of farm and biorefinery headquarters. Few economic studies are set in Italy. Bonfiglio and Esposti (2016) propose an input-output model for assessing the economic impact of the establishment of a territorial power plant in Italy. Lopolito et al. (2011) relies on a participatory approach and fuzzy cognitive mapping to help policy decisions when it comes to the establishment of a bio-based industry in agricultural districts of Italy. The problem of planning multi-feedstock supply and selecting of a suitable area for establishing a biorefinery is a principal-agent problem (Zhang et al., 2016). The prospected operator (principal) would decide upon biorefinery location based on a cost minimization strategy. Major costs subject to minimization are transaction and input procurement costs. The operator would control both costs via biomass procurement contracts with farmers. Then, an underlying condition for selecting the location for a new biorefinery is defining a plausible multi-feedstock supply area hosting those farm types (agents) that might be interested in subscribing a contract. Against that background, the aim of this paper is to design a contract to help the creation of a feedstock supply area that could support the establishment of a new territorial biorefinery in Tuscany. To that purpose, we build a principal-agent model to design an optimal mechanism for incentivising farmers to diversify their crop mixes by introducing one or more of the industrial crops needed by the biorefinery operator. The solution of the principal-agent problem involves identifying the optimal location for the biorefinery and then designing a contract mechanism that maximises principal’s profit. Suitable farm types for introducing biomass cropping and agents’ marginal costs are estimated through a mathematical programming model. Mathematical programming is also used for simulating the possibility to introduce a pre-treatment plant on each farm type. In the economic literature, combined principal-agent-mathematical-programming models are used for informing and improving decision making processes (Viaggi et al., 2009). We use real world data originating from a research project aimed at creation of a farm-to-gate supply chain for bio-based products in Tuscany, including the construction of a territorial biorefinery. Preliminary results suggest that the most suitable farm types for the integration of industrial cropping are in southern and eastern Tuscany. Those areas host the largest arable farms of the region, miss steep mountains, and offer ease of access to means of transport (existing infrastructures). The highlighted areas are hilly, with clay and often shallow soils, thus, hosting both productive and low productive land. The latter might be classified as marginal land. The small-scale biorefinerey turned to be the most cost-effective among the simulated ones, given the possibility to source all the needed feed stock within the boundaries of Tuscany and to be energy self-sufficient, thanks to the coupled biogas-to-energy plant. The selected farm types would adopt the industrial crops on different shares of farmland. At time of writing, we do not highlight significant differences among the four crops. Farmers do not seem willing to adopt a pre-treatment plant. Probably, they still show risk aversion given the novelty of the biorefinery concept. However, few forward-looking farmers with more entrepreneurship would implement a pre-treatment plant on farm.