To a great extent, safety is ensured in the design and operation of hazardous production facilities (HPF) through identifying, analyzing and predicting the risk of accidents (failures), involving, where possible, a more complete quantitative risk estimation in determination of the HPF condition [1], which is the responsibility of the Federal Service for Environmental, Technological and Nuclear Supervision (Rostechnadzor). Among the HPFs, where multifactor risks exist at the stage of design, a special place is occupied by nuclear power installations, shelf development facilities, oil and gas platforms, as well as critical infrastructure facilities as the assets essential for the healthy state of society and the national economy in conditions of impacts from the catastrophic risk factors [2–4].The issues involved in the estimation and prediction of hazards from unfavorable situations, emergencies, accidents and failures are considered in [2,3,5–7] where the safety of HPFs is defined by two major factors: probability of an unfavorable event (situation) and the damage from such event, using different risk identification methods, including recent advances in the asymptotic theory of the probability of extreme values.To solve the risk estimation problems, issues involved in the estimation of risk parameters have been considered with different options of the HPF state graphical space interpretation. Peculiarities of estimating the risk sensitivity and the risk degree have been described and the evolution of approaches to the estimation of risk in the HPF design and operation has been shown. Big data analysis methods for risk management have been proposed.
Read full abstract