An integrated approach for real-time hazard mitigation in complex industrial processes

•Optimization of safety-threshold for complex industrial processes.•Consider joint probabilities of multiple process variables leading to an accident.•Enables dynamic risk assessment based on multiple real-time process variables. Modern engineering systems give paramount importance to safety in orde...

Full description

Saved in:
Bibliographic Details
Published in:Reliability engineering & system safety Vol. 188; pp. 297 - 309
Main Authors: Rebello, Sinda, Yu, Hongyang, Ma, Lin
Format: Journal Article
Language:English
Published: Barking Elsevier Ltd 01.08.2019
Elsevier BV
Subjects:
Accidents
Bayesian analysis
Bayesian network
Contamination
Economic impact
Genetic algorithm
Genetic algorithms
Hazard mitigation
Industrial hazard
Markov chains
Optimization
Organic chemistry
Process monitoring
Process variables
Real time
Real variables
Reliability engineering
Risk assessment
Safety
Safety-threshold optimization
System safety
Process monitoring
System safety
Safety-threshold optimization
Genetic algorithm
Bayesian network
Industrial hazard
ISSN:0951-8320, 1879-0836
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Optimization of safety-threshold for complex industrial processes.•Consider joint probabilities of multiple process variables leading to an accident.•Enables dynamic risk assessment based on multiple real-time process variables. Modern engineering systems give paramount importance to safety in order to avoid or mitigate hazardous accidents which can lead to huge economic losses, environmental contamination, and human injuries. This paper proposes an integrated approach that uses both Hidden Markov Model and Bayesian Network to estimate an optimum safety-threshold for complex industrial processes. In order to estimate the safety threshold, the proposed approach considers different cost factors and the joint probabilities of multiple process variables leading to an accident. In addition to the system level threshold, it also estimates the safety-threshold for components. This helps in identifying the component that needs maintenance to enhance system performance and safety. Furthermore, it proposes a dynamic risk assessment methodology based on multiple real-time process variables. The optimum safety-thresholds are estimated using Genetic Algorithm which aims at minimizing the system running cost over a finite time horizon. A case study on Tennessee Eastman Chemical Process is presented to demonstrate the proposed methodology for optimizing process safety-threshold.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0951-8320
1879-0836
DOI:10.1016/j.ress.2019.03.037