Programmable electronic systems applied for risk control in petrochemical plants

Abstract

The chemical industry today is in transition, with increasing emphasis on total quality control along with needs to meet even more stringent levels of health, safety and environmental management. Fortunately, these needs are being paralleled by major developments in process monitoring and control instrumentation. Microprocessor-based process sensors, programmable electronic systems (PESs), and precision throttling valves make possible the implementation of complex process control strategies in DCSs where operators interact with the process through video display human/machine interfaces. The automation of past, manual operating tasks is justified under the umbrella of reducing the likelihood of human errors. Although automation does reduce the sources of human error in plant operations, these programmable control systems and software introduce new and different potential sources of error, leading to new implementation considerations. Many of the hazard identification and risk assessment methodologies used today treat the process control system as a ‘black box’. Furthermore these methodologies are based on techniques that assume independence of failures. However, when PESs are used for both regulatory and safety interlock controls, the possibilities for common mode failures and covert faults are greatly increased over that demonstrated by older technologies. Data highways, common software, central supervisory control computers and subtle deficiencies in design all introduce new potential sources of shared failures. Today, achievement of plant safety is a systems issue [1]. Process hazard assessment and control requires an integrated analysis with inputs from each of the disciplines involved in process design and plant operation, as well as the guidance of safety and risk specialists. This paper presents an approach to total process safety assessment in which instrumentation is applied as one means for risk mitigation. A methodology for reduction of hazardous event likelihood by using modern programmable electronic monitoring and control systems is described, and applications of the methodology to industrial examples are presented. The methodology applied is that documented in [2].