Research

To ensure process safety in the Oil & Gas and Petrochemical industries, we employ advanced CFD tools in conjunction with comprehensive Process Hazard Analysis (PHA) methods, including checklists, QRA, HAZOP, STAMP, and STPA, among others. This powerful combination enables us to simulate and thoroughly analyze potential hazards, such as the dispersion of flammable or toxic materials, the progression of explosions, and the performance of safety barriers and ventilation systems. Moreover, we harness the power of AI and ML algorithms to extract deeper insights from our numerical data, enabling advanced hazard identification, predictive risk modeling, and sophisticated scenario analysis. By meticulously modeling the fluid dynamics of these critical events, we contribute directly to the design of inherently safer processes and the development of effective risk mitigation strategies. The strategic integration of AI and ML significantly elevates our process safety capabilities, resulting in demonstrably safer operations, substantial risk reduction, and enhanced environmental stewardship. This data-driven approach paves the way for a more proactive and effective safety management system.

A) Potential tank fire scenarios (Argyropoulos et al. 2012) and B) Enthalpy contours around the tank on fire obtained using the commercial CFD software PHOENICS (Argyropoulos et al. 2013).

A critical consideration, often posing a higher risk, is the potential for toxi releases occuring outdoors to impact occupants withing nearby indoor environments. The ingress and subsequent entrapment of hazardous gases indoors can lead to significantly accelerated accumulation, potentially reaching lethal dosages more readily than in open-air conditions. Despite this pronounced risk, there remains a notable lack of data in empirical data and comparative studies regarding appropriate modelling techniques and effective mitigation strategies for indoor exposure. Consequently, we utilize advanced CFD techniqeus to specifically investigate phenomena such as toxic building ingress and to perform rigorous risk assessments based on dose-response methodologies. The proposed methodology will serve as a guide for the improvement of relevant risk assessment tools and future studies.

A) Distribution of the AEGL-3 values, for all building rooms, at different wind directions and for the CD-HVAC off-scenario, Radius corresponds to the AEGL value and the colour represents the number of rooms with this value for: (a)ASHRAE model, (b) QUIC-CFD model and (c) QUIC-URB model and B) Surface streamlines and Cp for “D5”and 8 different wind directions calculated by QUIC-CFD (Argyropoulos et al. 2017).

Furthermore, we harness the power of AI and ML algorithms to extract deeper insights from our numerical data, enabling advanced hazard identification, predictive risk modeling, and sophisticated scenario analysis. By modelling the fluid dynamics of critical events, we contribute directly to the design of inherently safer processes and developing effective risk mitigation strategies. The strategic integration of AI and ML significantly elevates our process safety capabilities, resulting in safer operations, substantial risk reduction, and enhanced environmental stewardship. This data-driven approach paves the way for a more proactive and effective safety management system.