A novel kpi for continuously monitored safety barriers based on probabilistic safety margins

In oil and gas upstream plants, several barriers (technical, procedural and organizational) are in place to prevent and mitigate accidents. Proper safety barriers functionality is, then, important to control the risk during the life of the plants. Safety barriers modelling is, then, required for risk assessment. In this work, we model the barriers functionality by discrete Health States (HSs) and their stochastic process of transition by a multistate Bayesian Network (BN). For each barrier, the HS is defined with reference to properly defined Key Performance Indicators (KPIs). Here, for technical barriers that can be continuously monitored, we propose a specific KPI based on Probabilistic Safety Margins (PSMs). Its application is illustrated with respect to the Process Control System (PCS) of the slug catcher of the upstream plant, which continuously controls the process pressure of the system within a specific operational range.