Advances in computational tools have enabled the development of genome-scale metabolic (GSM) models. Integrating GSM models with downstream processes through multi-scale models allows predicting biorefinery feasibility. This study utilizes the GSM model iCbu641 to predict the use of the wild-type strain Clostridium butyricum IBUN 158B for producing 1,3-propanediol (PDO) from Colombian biodiesel-derived glycerol. Different bioreactor and evaporation trains were evaluated, with the selected biorefinery comprising a fed-batch bioreactor and double-effect evaporation trains. Economic evaluation indicates the selected biorefinery has a payout period of 7.4 years and a production cost of around 1 US$/kg, indicating feasibility. Five additional biorefineries were assessed using in-silico knockout mutants blocking competing co-products. Simulations revealed that blocking hydrogen production could reduce unit costs and payout periods by up to 15 %, suggesting its potential in a PDO biorefinery. Concluding, the novelty of this study is the holistic integration of biological and economic assessments, reducing resources in exploratory experiments.