Quantitative Inherent Safety Assessment of Crude Palm Oil Production Using the NuDIST Methodology

Crude palm oil (CPO) is the most widely consumed vegetable oil globally due to its high agricultural productivity and its increasing use in the food and energy sectors. However, its production process involves intensive operations, severe temperature and pressure conditions, and work environments that increase the risk of industrial accidents, highlighting the need for rigorous process safety assessments from the early design stages. Traditionally, these analyses have relied on inherent safety indices, which, while allowing for quick comparisons, have limitations related to subjectivity and the simplification of complex phenomena. In this context, the present study aims to evaluate the inherent safety of the CPO production process using the Numerical Descriptive Inherent Safety Technique (NuDIST), a quantitative approach based on logistic functions. The methodology considers chemical safety parameters—flammability, explosiveness, toxicity, and reactivity—and process safety parameters—temperature, pressure, heat of reaction, and process inventory—applied to an industrial process with a capacity of 30 t/h of fresh fruit bunches (FFB). The results indicate that linoleic acid is the substance that contributes most to chemical safety, with a Chemical Safety Total Score (CSTS) of 26.14, while process inventory dominates process safety, reaching a Process Safety Total Score (PSTS) of 101.95. The total NuDIST score obtained was 128.09, demonstrating that the process risks are mainly associated with operating conditions rather than the hazards of the substances. Comparisons with other industrial systems reveal that CPO extraction is inherently safer than processes such as chitosan production from shrimp exoskeletons and six (6) possible routes for methyl methacrylate (MMA) production. Taken together, these findings position NuDIST as a robust and objective tool for strengthening the safe and sustainable design of agro-industrial processes.