The paper discusses the impact of the COVID-19 pandemic on the Italian chemical and process industries, where Directive 2012/18/EU Seveso III, for the control of Major Accident Hazard (MAH), is enforced. The Safety Management System (SMS) for the control of MAH, which has been mandatory for 20 years in Italian Seveso Establishments, has been highly stressed by the external pressure, related in some way to the COVID-19 pandemic. Fairly, most companies, in particular in oil and gas sectors, have demonstrated an adequate capability to reconcile operation continuity and health requirements. This experience is providing the establishment operators and the regulators with valuable suggestions for the improvements of the SMS-MAH. Within this framework, an innovative organisational resilience model is proposed, aiming at the development of a higher capability to face future new crisis. The current SMS-MAH already includes some basic pillars to enhance resilience, which were valuable during the pandemic crisis, but a full and rationale development is still needed. Starting from the first pandemic phase experience, this paper presents a novel tool to assess the degree of “resilience” of a SMS-MAH. It is based on a questionnaire, featuring 25 questions grouped into eight items, according to the typical SMS-MAH structure. A two level AHP model has been developed in order to define the weights to be assigned to each point. The AHP panel included industrial practitioners, regulators, authorities and researchers. The results are based on the COVID-19 experience and consequently the developed model is tailored to face health emergencies, but the approach may be easily transferred to other external crises.

The focus of the present paper is the development of a resilience framework suitable to be applied in assessing the safety of ship LNG (Liquefied Natural Gas) bunkering process. Ship propulsion considering LNG as a possible fuel (with dual fuel marine engines installed on board) has favored important discussions about the LNG supply chain and delivery on board to the ship power plant. Within this context, a resilience methodological approach is outlined, including a case study application, to demonstrate its actual effectiveness. With specific reference to the operative steps for LNG bunkering operations in the maritime field, a dynamic model based on Bayesian inference and MCMC simulations can be built, involving the probability of operational perturbations, together with their updates based on the hard (failures) and soft (process variables deviations) evidence emerging during LNG bunkering operations. The approach developed in this work, based on advanced Markov Models and variational fitting algorithms, has proven to be a useful and flexible tool to study, analyze and verify how much the perturbations of systems and subsystems can be absorbed without leading to failure.