An analysis of process fault diagnosis methods from safety perspectives

Highlights

• This paper analyzes the interconnection of fault detection and diagnosis, risk assessment and abnormal situation management.

• The paper reviews the existing methods and models related to fault detection and diagnosis, abnormal situation management, and risk.

• The paper also present authors’ thoughts on the research direction in the area of process safety in industrial 4.0.

• This paper aims to serve as a road map for process safety research to enable safer and sustainable development.

Abstract

Industry 4.0 provides substantial opportunities to ensure a safer environment through online monitoring, early detection of faults, and preventing the faults to failures transitions. Decision making is an important step in abnormal situation management. Assigning risk based on the consequences may provide additional information for abnormal situation management decisions to prevent the accident before it occurs. This paper analyzes the interconnections between the three essential aspects of process safety: fault detection and diagnosis (FDD), risk assessment (RA), and abnormal situation management (ASM) in the context of the current and next generation of process systems. The authors present their thoughts on research directions in process safety in Industry 4.0. This article aims to serve as a road map for the next generation of process safety research to enable safer and sustainable process operations and development.

Introduction

Modern process plants are becoming more complicated due to process units’ interconnectivity resulting from plantwide control and optimization. In such a plant, control systems connect with many sensors and actuators to control plant operations. The sensors aid in monitoring process conditions while the actuators control the process by physically adjusting the system's variables. Despite these monitoring and control measures, processes can drift beyond their safe operating range due to actuator, sensor, or system faults. These faults may lead to a system failure and ultimately cause a plant accident.

In process systems, potential accidents are prevented using layers of protection. Failures in such layers increase accident probability and lead to its consequences. Hazard identification, probability assessment, and the consequences of hazardous incidents, considering the layers of protection in a system, provide an understanding of the system safety status.

Hazard identification, FDD, RA, and mitigation action play vital roles in maintaining plant safety. Many researchers have reviewed hazard identification approaches [Dunjo et al. (2010); Cameron et al. (2017); Willey (2014)]. Also, there are several review articles on FDD methods, by [Gao et al. (2015); Zhong et al. (2018); Puncochar and Skach (2018); Md Nor et al. (2019); Hoang and Kang (2019)], and on RA in process systems, reviewed by [Khan and Abbasi (1998); Khan et al. (2010); Swuste et al. (2016); Amin et al. (2019)]. However, these articles focus on different process safety elements without defining the interrelation and the overall ASM process. For process safety management, the methods and models used must be analyzed in combination to obtain a holistic view of the safety management framework. Therefore, this article attempts to review and analyze process safety elements’ methods and models, focusing on their interrelations. Specifically, the article focuses on addressing the following questions:

• How can risk be used for fault diagnosis and abnormal situation management (process safety perspective)?

  • How can fault detection and diagnosis be used from the process safety perspective?

  • How is abnormal situation management practised from the process safety perspective?

  • How are FDD and ASM integrated with the safety system?

  • What are the key knowledge and technological gaps in the preceding areas?

• How could operational risk be a tool for process safety management for Industry 4.0?

  • What are the available approaches for operational risk assessment?

  • What are the potential uses of machine learning techniques in assessing operational risk?

  • What are the key knowledge and technological gaps in implementing novel machine learning tools in process safety management?

• What is the way forward with Industry 4.0 to make a smart process plant a safe environment?

According to Iserman and Balle (1997), based on the SAFEPROCESS committee's definition, ASM is a centralized, continuous, and comprehensive process to prevent and control the potential hazards in process systems. Moreover, ASM should identify the deviation from normal operation to faulty and failure conditions and bring the system back to normal operation.

In the process industry, determining the risk margin, using appropriate modeling such as failure models, accident models, and risk models, helps to provide information to prevent the fault from becoming a failure condition.

A failure model evaluates the accident probability by determining system failure based on a data-driven or physical model approach. Similarly, the accident model relates to the causes and effects to address the consequences. However, to develop the failure model and the accident model, hazard identification will be an initial step. When fault leads to failure, the failure model and accident model can evaluate the possible hazards and consequences.

In the process industry, FDD, RA, and ASM may apply in a closed loop. FDDs approaches to determine the fault condition and are initiated to identify the possible hazard. The failure model and accident model evaluate the probability and consequences of the system hazard when the process systems fail to identify and control the system's fault condition by ASM. Assessing a risk margin using risk models gives feedback to ASM regarding the hazardous event. With the feedback information, ASM changes the decision to control the operation.

From the Industry 4.0 perspective, interconnecting FDD, ASM, and RA help to develop an intelligent safety system by learning the risk and taking necessary action autonomously to prevent the hazard.

Several review articles on the recent methods and models for FDD and RA have been published and are available in the open literature. However, these methods have been limited in scope, mainly focusing on FDD methods. This review comprehensively studies ASM from a holistic perspective, discusses the past and present methods and techniques, and directs the future trend of process safety for Industry 4.0. The review's scope includes the topics directly related to system faults, failure analysis, RA, mitigation action, and process safety, published in journal papers. However, to discuss the FDD and ASM standards, some industrial standards and conference papers are used.

The rest of the paper is organized as follows: Sections 2 and 3 focus on past and present methods and models for FDD, and ASM's role to prevent hazards in process systems. Section 4 summarizes the past and present risk assessment models, failure models, and hazard identification to protect the plant from accident consequences. Finally, the last section highlights the next generation research needed for Industry 4.0 and its challenges.