Authors
Larrea Gallegos, Gustavo; Benetto, Enrico; Marvuglia, Antonino; Navarrete Gutiérrez, Tomás

Abstract
A supply network (SN) can be defined as a system where parties interact to satisfy costumer’s orders and their own objectives. Firms and their interactions are represented as nodes and edges, respectively, while the network topology is constantly changing. The recent disturbances in global logistics and production generated by the COVID-19 pandemic have shown how companies need to re-adapt to still generate value. While this kind of affectations can be directly measured in terms of economic losses, there is still a gap in the understanding of their consequences from a sustainability point of view. To address this topic, we developed a sustainability assessment method that embraces the complex nature of a SN. The computational core of the method is an agent-based model programmed in python that simulates the interactions and disruptions among firms. We verified the capacity of the tool to recreate feasible SNs and account for flows. The Peruvian fishmeal sector was selected as a case of study, especially because Peru is a leading country in the worldwide supply of fishmeal. Using primary data from a fishing company and secondary import and export data, we modelled the network of agents and their interactions. The model allows to calculate the evolution of metrics associated with different dimensions of sustainability. Moreover, the tool was used as a simulation environment to test different variations in the adopted behaviours for selecting suppliers after experiencing disruptions. Preliminary results show that network’s emission profiles hardly adopt linear behaviours, and that under given initial parameters the SN can depict resilient behaviours while still having a performance considered as sustainable.