Evaluating Probabilistic Chances of Overpressure-Induced Mechanical Explosion in Fixed Bed Reactor: A Case Study of Hydrogenation of Heavy Base Oils Using Fault Tree and Bayesian Network

The heavy base oils are traditionally treated using concentrated hot acid to remove various polycyclic aromatic hydrocarbons. The hot acid treatment poses operational safety issues at the supplier end whereas a cancer hazard at the consumer end. An alternative to the hot acid treatment is the hydrogenation of heavy base oils. The feasibility for hydrogenation of various base oils was studied in a semi-pilot-scale Fixed Bed Reactor (FBR) by us. For some of the highly viscous heavy base oils, the serious issue of choking inside the semi-pilot-scale setup was observed. This problem motivates us to further investigate the issues associated with FBR at the pilot-scale where less men power and minimum automation are involved. The FBR is vulnerable to overpressure and overtemperature scenarios when a multiphase exothermic reaction is carried out under severe conditions. The present study is the preventive measure taken by us at the early stage of process development in which we evaluate probabilistic chances of overpressure-induced mechanical explosion in FBR. The factors leading to an overpressure scenario were identified using the qualitative approach of HAZOP and Directed Acyclic Graph (DAG). The prior failure probabilities were estimated using Fault Tree Analysis (FTA) and Bayesian Network Analysis (BNA). Later the posterior probabilities were estimated using BNA and different scenarios were compared. For an overpressure scenario, high gas inflow, low liquid outflow and operator actions are critical factors. The reliability of the safety relief valve, safety relief valve tension and pressure measurement is the most critical and shows maximum improvement indices.