Decarbonisation Technology May 2026 Issue

overall evacuation and mitigation performance. In addition, recognising that human behaviour during emergencies is inherently unpredictable – affected by factors such as location, movement speed, fatigue, and stress – ABS integrates Monte Carlo simulation techniques into the agent-based approach. Monte Carlo methods apply repeated random sampling to model the probability of different outcomes in processes with uncertainty. By doing so, the simulation can account for a wide range of possible human responses rather than

Figure 3 Agents model the dispersion behaviour of ammonia

rhythms, roles, and psychological factors. These agents can perceive their environment, respond to alarms, and act based on contextual factors such as location, stress, fatigue, or panic. By simulating how individuals interact with each other and the environment during emergencies, the model captures emergent behaviours that are difficult to predict through conventional methods. This level of behavioural realism helps to enable stakeholders, such as shipowners, port operators, and bunkering facility managers, to evaluate how individual differences and local interactions influence broader evacuation patterns, the formation of bottlenecks, and the effectiveness of mitigation strategies. Simulations run under different operational conditions and times of day help uncover critical variations in response effectiveness, providing insights into risk hotspots and potential failure points in emergency plans. Ultimately, agent-based modelling offers a powerful tool for evaluating and improving emergency response strategies by bridging the gap between technical risk assessments and real-world human behaviour. It helps planners visualise not only what might go wrong, but how and why it happens, helping to lead to more practical, adaptive, and human-centred emergency response planning. This approach allows for the examination of how individual differences and local interactions may lead to large-scale patterns in evacuation flows, bottleneck formation, and

relying on fixed or deterministic assumptions. This allows for the generation of statistical distributions that reflect the variability and randomness of real-life behaviour, enabling a more realistic and robust evaluation of emergency scenarios and their potential consequences. The integrated approach facilitates dynamic risk assessment, where event probabilities evolve as response operations progress, incorporating real-time changes in environmental conditions, resource availability and operational constraints. This temporal dimension is crucial for emergency response planning, as the effectiveness of interventions and the risks faced by responders and evacuees change continuously throughout an incident. The framework considers time constraints that are fundamental to emergency scenarios, where delayed decisions or actions can have chain effects on overall system performance and safety outcomes. Environmental variables such as weather conditions and hazard propagation are integrated into the models, allowing for more realistic assessments of how external factors influence both agent behaviour and system-wide emergency response effectiveness. The integration of traditional emergency response frameworks with advanced simulation, such as CFD and simulation-based quantitative risk analysis, creates a more robust, system- wide and time-dependent emergency response strategy. This comprehensive approach enables identification of risk hotspots, evaluation