A domino is a flat thumb-sized, rectangular block, the face of which is divided into two parts, each bearing from one to six pips or dots. It is normally twice as long as wide and can be stacked vertically or horizontally. A domino is used to play various games of chance or skill, usually by matching the ends of the pieces and laying them down in lines and angular patterns. The term is also applied to any of the many games played with such blocks, including chess, backgammon and poker.
A traditional double-six set of dominoes has 28 pieces. It is commonly made from bone, silver lip ocean pearl oyster shell (mother of pearl), ivory or a dark hardwood such as ebony, with contrasting black or white pips. Some sets also incorporate different materials such as metals, ceramic clay or frosted glass, giving the pieces a more novel look and increasing their weight.
The problem of domino effects has gained worldwide concern, since they can cause catastrophic accidents with multiple escalation vectors. For example, a fire at a process plant may trigger an explosion at neighboring equipment. Hence, modeling the evolution of domino effects, especially higher-level and temporal, is an important issue.
Several methods have been developed to analyze the probability of the occurrence of domino effect events and to assess their risk. The available assessment methods include analytical methods, graphical methods and simulation methods. In addition to assessing the risk, these methods can provide valuable information for the development of prevention strategies.
To address the complexity of the problems, Khakzad (2015) proposed a dynamic Bayesian network (DBN) model to take into account spatial and temporal escalation of domino effects. The model allows for a more accurate identification of the most probable sequence of accidents than the one offered by ordinary BN analysis. Moreover, it can deal with the uncertainty of the evolution sequence.
Another approach to the problem of evaluating the probability of domino effect events involves using graph metrics such as betweenness, out-closeness and in-closeness in directed graphs, and closeness in undirected graphs. These metrics can help to identify critical installations that are highly vulnerable to domino effect occurrences and support prevention decision-making.
Lastly, the behavior of chemical clusters can be analyzed with an optimization algorithm that takes into account the probability of an accidental or deliberate occurrence of a domino effect and the impact on the safety of other plants within the cluster. This approach enables the determination of an emergency level for each company in the cluster based on its vulnerability to the domino effect and supports preventive measures by reducing the attractiveness of the chemical plant to adversaries. This type of management can ensure that the safety and security of all involved companies is guaranteed. However, this method requires cooperation among all plants within the cluster to achieve complete success.