(40aj) Understanding the Domino Effect in Process Industries from Rigid Body Mechanics Models

With the developing trend of agglomeration in the process industry, the hazard sources are becoming increasingly diverse and centralized. The "domino effect" containing multiple accidents is now becoming a new feature of chemical accidents in comparison to the individual accident that occurs independently. Domino effect refers to the chain effect of one or more secondary accidents caused by the primary chemical accident. Due to the characteristic of synergistic and chain effects among multi-hazard accidents in domino effect, the accident consequences are generally more severe than those of independent chemical accidents. For instance, the explosion accident that happened in 2019 in Xiangshui, China, was a typical domino accident in which the initial fire triggered multiple explosion and toxicant release accidents, resulting in 78 deaths and 716 injuries.

In recent decades, the study of domino effect has become a concerned topic for scholars around the world. Based on existing studies of the hazard analysis and risk identification, scholars have successively proposed methodologies such as "threshold theory", "probit model", and "vulnerability analysis" for the study of domino effect, and further derivate various methods and strategies for risk analysis, evaluation, and reduction. Although the research on domino effect has become gradually mature, there is still a lack of a complete model that can encompass all of the elements in the research of the principles of occurrence, development mechanism, chain-breaking strategy and risk reduction measures of the domino effect. In this paper, we analyze the domino effect from the perspective of rigid body mechanics, aiming to establish a unified physics model for the analysis of domino effect. Based on the model, this paper analyzes the influencing factors in the accident escalation process, which is the core element of the domino effect, and proposes risk management and loss prevention strategies for the domino effect.

The concept of domino effect, named after the ancient game of dominoes, graphically describes the process of a series of serious chain reactions triggered by a tiny initial energy in an interconnected system. The domino effect concept is not only applied in the field of risk analysis in process industries, but also widely introduced in various fields such as politics, meteorology, economics, and architecture. However, most subjects commonly applied only with its literal meaning, ignoring the underlying mechanic principles, and thus failing to reflect the potential positive effects of the physical model analysis of domino effect in solving specific real-world problems.

As a commonly used concept, the domino effect model contains complex mechanical principles, for example, the rotation and tilting of the rigid body, the conversion of kinetic energy to mechanical energy during a collision, the transfer of impulse and momentum, and the critical conditions and thresholds for the occurrence of a chain reaction after a collision. A typical example of domino effect mechanic models applied in the safety science research field to solve practical problems is in the study of the crushing and pushing behavior of people in dense crowds. By combining with the social-force model, which is widely used in pedestrian dynamic research, domino effect mechanic analysis can effectively guide the analysis of mechanism and accident early-warning of stampede events. Unlike the dense crowd study, the escalation process of domino effect in process industries is not a mechanics transfer through direct contact, but a complex mechanics through the damage caused by thermal radiation, explosion shock wave, fragment, and other physical effects. The primary accident and the secondary accident in the domino effect can be abstracted to adjacent rigid bodies in the domino mechanic model. They can form a cascading chain accident through releasing and bearing accident physical effects, just like the incline and collision process in the domino model. In addition, the conception of resilience and safety barrier can also be discussed by applying the domino mechanic model. How to improve the robustness and functional redundancy of industrial facilities in order to enhance safety resilience, and how to set up safety barriers to interrupt the domino process development, those problems can be conceptualized and integrated in the domino mechanic model.

The core objective of this research is to propose the correspondence between the domino mechanic model and the elements of domino effect in the process industry, so as to construct a full-cycle and full-element framework for domino effect analysis. Based on that, this study hopes to obtain some interesting and useful conclusions about the process of accident escalation, risk reduction strategies and other important research targets of the domino effect, which can inspire and guide the subsequent research of domino effect. It should be mentioned that the transition from domino mechanic model to domino accident analysis is a transition from physical concept to actual industrial problem, from abstraction to concretion, which requires the proper usage of logical induction and deduction. It is not only a challenge faced by this research, but also is the practical value and novelty of this study.

The first part of this study introduces the domino model and its mechanic principles; The second part introduces the correspondence between the domino mechanic model and the elements of domino effect in the process industry, and constructs the framework of domino effect research; The third part focuses on the two main research targets: the accident escalation process in domino effect and loss prevention strategy; The fourth part explains the practical value of the application of the domino mechanic model in the process industry by taking the domino accident occurred in Xiangshui, China as a case study; The final part summarizes the unified physics model of domino effect in process industries proposed in this study, and puts forward an outlook for future research.