For a particular problem, the modeler has to decide the proper length scale and a suitable modeling methodology. For example, in assessing the thermal radiation hazard from jet or pool fire, a computational fluid dynamics–based approach is more appropriate. However, also, if one is interested in how that heat from radiation affects the structural integrity of occupied buildings, a combination of computational fluid dynamics and finite element should serve the purpose. Then, the modeler needs to focus on selecting an appropriate model within the computational fluid dynamics–based approach such as, will two-parameter k-epsilon turbulence model be suffice to model the cloud dynamics of, say, the released flammable gas cloud? Deciding the meshing size, typically variable across the domain, is another challenge a modeler faces in defining a problem at this scale. Often a facility performing a full-blown consequence analysis will require one to evaluate many potential scenarios. It is advisable to perform consequence analysis and quantitative risk assessment as initial screening studies, thereby helping in narrowing down the number of specific scenarios to be modeled further, such as CFD studies, by applying the Pareto principle (i.e., to select scenarios that cumulatively account for 80% or more of the plant's total risk).

While developing novel materials and processes, especially with chemicals and materials whose hazardous properties need to be determined, the application of molecular modeling–based approaches can help answer some questions. Molecular modeling–based approaches are powerful tools for the characterization of hazardous material properties through assessing safety-related concerns of novel chemicals under various process conditions. Molecular modeling–based approaches are promising tools for understanding the role of process chemistry in fires and explosion (Mannan et al., 2006). Several safety-related information including material reactivity, decomposition, and flammability can be estimated using molecular modeling approaches. In scenarios involving uncertainties, at the minimum, modeling approaches can reduce the requirement of much expensive experimental analysis for assessment of hazardous processes, such as understanding and estimating consequences of a thermal runaway reaction.

Several challenges, as mentioned in various chapters, remain in the context of implementing multi-scale modeling for process safety applications. However, in process safety applications, the lack of understanding of a phenomenon often introduces conservatism in a safety model. In some situations, this might even compromise safety. Additionally, owing to the nature of the problems, it is not always feasible to determine the underlying root behind a phenomenon through experiments. It is only appropriate to extend the advancement in theory of materials at different time and length scales to address less understood or coarsely defined safety concerns through exploitation of significant increase in computational power in recent times.

There are currently many examples where definitions of safe modes of operation are identified through empirical means such as safe-operating conditions for materials used in oil and gas production system codified in ISO15156. These safe zones are not rigorously established and hence are misleading while using changed process conditions or when novel materials are introduced in the process (Taylor, 2015). Ideally, a comprehensive integrated multi-scale model will lie at the intersection of the following as illustrated in Figure 10.1:

These activities working in synchrony have tremendous potential to improve safety, environment, and sustainability.

We hope that this book is able introduce to its readers to the effective application of multi-scale modeling approaches in process safety applications. Similar to any active area of science, the application of multi-scale modeling methodologies in process safety discussed here is incomplete with further development in the field expected. As one becomes adept at understanding of the strength and weaknesses of all these approaches, one will be in a powerful position to develop mapping between various modeling methodologies and its applicability to various process safety concerns. Hence, this book is expected to serve as an introduction toward the application of different modeling methodologies from the atomistic scale to the process level for addressing process safety concerns. Multi-scale modeling approaches are established powerful tools as proven in several areas of science, and process safety challenges are too big a concern for the industry to simply miss the boat.