Optimisation du transport et de l’allocation d’un co-produit de la biométhanisation au sein d’une chaine d’approvisionnement collaborative en boucle fermée

Abstract

This research explores the logistics anaerobic digestion production and transportation in the context of a collaborative and closed-loop supply chain (SC) in a region of Quebec (Canada). The raw materials suppliers provide an anaerobic digestion plant with organic residues, which can be in two forms: liquid or solid. The residues are transported to the plant by two types of truck (tanker and solid bulk) of different capacities, where they are transformed by anaerobic digestion. During the anaerobic digestion process, these residues are decomposed in the absence of oxygen into biogas and digestate, which is a high-value organic fertilizer used on farms. Most of the suppliers are also customers requiring to recover the digestate for their own farms. Moreover, they need to obtain the digestate in the same form as the organic residues they have provided (liquid or solid). However, when it exits the anaerobic digestion plant, the digestate is in liquid form and therefore requires passing through a separator to be dewatered. This additional operation is costly. The first aim of this project is to size a fleet of trucks adapted to the needs and capacity of the plant under study and minimize transportation costs. The second objective is to determine the amount of digestate to be allocated to supplier-customers to minimize the total costs of SC, i.e., transportation and separation costs, in order to achieve maximum collective savings. Finally, the third objective is to determine compromise strategies to ensure fairness in digestate allocations for all supplier-customers. Two sets of data were used in this project: the first is from a real case study and the second is a fictitious case study based on the real case. After defining and modeling the problem using mathematical programming, four scenarios reflecting different logistics practices (backhauling and heterogeneous truck fleet configuration) were tested. This step identified one scenario as being more efficient than the others in several respects. Compared to the baseline scenario, which did not include these logistics practices, applying both return loads and a heterogeneous fleet to our real-world case study resulted in potential savings of 17%, a 42% reduction in total kilometers traveled, and a 34% reduction in greenhouse gas (GHG) emissions. For the fictitious case study, the potential savings are 29%, the reduction in the total number of kilometers traveled is 43%, and the reduction in GHG emissions is 35%. To determine the optimal amount of digestate to allocate to supplier-customers, we started with the most efficient logistics scenario obtained previously and established eight different digestate allocation strategies. The initial distribution is considered fair and is calculated based on the pro rata wet tonnage of organic residues supplied by each supplier-customer. We then wanted to compare it with other distributions. The strategies were determined and calculated either manually or using an algorithm designed to minimize total costs. We observed encouraging results in terms of potential savings, reduction in total kilometers traveled, and GHG emissions, but no single strategy stood out above the others, either for our real-world case study or our fictional case study. We then explored two avenues for developing compromise strategies that would ensure good performance on the indicators while maintaining fairness between all suppliers and customers. First, we implemented a financial compensation system designed to use the savings achieved by the entire SC through allocation strategies to compensate supplier-customers who would receive less digestate than expected under the initial distribution. Applying this system to the real-world case study revealed that of the nine allocation strategies tested, only one allowed for savings on total SC costs. However, when applying this system to the fictitious case study, three strategies allowed for savings on total SC costs. In a second step, we implemented a dualobjective strategy that mathematically determines the allocations that strike a balance between a fair distribution based on the initial distribution and a distribution aimed at minimizing total SC costs. In the real-world case, this strategy did not produce effective results. However, in the fictitious case, this strategy enabled us to determine allocations that combined the two objectives in a more balanced way.

Date	17 Nov 2025
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Tasseda Boukherroub (Supervisor) & Jean François Audy (Co-supervisor)