RISK ASSESSMENT OF A PETROLEUM OIL STORAGE TERMINAL

Abstract

In the last decade, major accidents have occurred in petroleum storage terminals around the world, e.g., Burchfield, UK (2005), Puerto Rico, USA (2009), and IOCL Jaipur, India (2009), which have drawn attention of expert groups as well as researchers working in the area of risk assessment. Irrespective of these accidents, many developed countries have already adopted legislations for risk assessment and consequence analysis taking risk as a criterion. Developing countries like India, however, couldn’t develop comprehensive guidelines due to lack of centralized mechanism to collect and maintain industrial accident database / information pertaining to the causes, consequences and preventive measures. Moreover, quantitative risk assessment studies related to accidents in petrochemical industry and storage facilities are sparse in India. It is necessary to study the mechanisms of risk assessment of petroleum industrial accidents to prevent or effectively tackle the emergencies in future. The consequences of major accidents provide only latent information as it is difficult to generate or simulate accident scenarios either through laboratory or field investigations. That is why a scientific analysis of the past accidents is necessary so that it can provide significant information to determine the most probable accident scenarios for new situations. Therefore, to fill the existing gape in the area of risk assessment of petrochemical storage facilities, the present research study has been carried out, which mainly focus on simulation and modeling of vapour cloud explosion (VCE) and large tank fire at a petroleum oil storage terminal. The study also includes the assessment of cumulative effects of explosion and fire on onsite and offsite population in terms of individual and societal risks. The results obtained from explosion fire modeling have been used to propose an appropriate emergency response plan based on electronic incident command system (e-ICS). To make the research more application oriented, the simulation and modeling of VCE and large tank fire have been applied for post risk assessment of Indian Oil Corporation Ltd. (IOCL) depot, Jaipur (India) accident. During this accident, high intensity blast pressure waves were generated across the terminal followed by multiple tank fires that destroyed several buildings and caused fatalities in the immediate surroundings. iv Significant learning has occurred from the IOCL, Jaipur accident. The typical methods and practices employed for the hazard identification and risk assessment of this site were unlikely to have identified the high overpressure generated at the time of accident. However, the research on recent accidents such as Burchfield, UK (2005) provides evidences on how the geometry of the terminal and vegetation in vicinity can generate the conditions for very high overpressure. The IOCL Jaipur accident resulted in release of approximately 2000 tonnes of gasoline when a 15 m high storage tank outlet valve leaked for approximately 80 minutes before the ignition of flammable mixture. A large and homogeneous flammable cloud covering an area of 180,000 m2 was formed due to very low wind speed and neutral or stable weather conditions. The overpressure in excess of 200 kPa (2 bar), due to one of the major VCEs, was generated across most of the site, which was not uniformly distributed throughout the terminal (Chapter 4). The directional indicators point to the source of the detonation being in the Pipeline Division area in the north east corner of the site. Subsequently, massive explosion immediately triggered the intense multiple tank fires. The large tank fire modeling reveals that the calculated flame height (H/D)max,calc lies between 0.9 and 1.5, which is within the observed range. The estimated surface emissive power ranges between 27 and 123 kW/m2, as evaluated by adopting various models (SFM, MSFM, TZM and TRSMFM) for large-scale tank fires (Chapter 5). The irradiances (Er) are calculated with a point source model and validated by the DNV Norway-based risk assessment PHAST 6.51 software. The modelled parameters of large tank fires showed a maximum percentage error of 25% with the observed values. The assessment of individual and societal risk associated with the effects of explosions and fires gives the maximum individual risk level as 10-4 per year at a distance of around 100 m from the release point. The next individual risk level (i.e.10-5 per year) has reached up to a distance of 280 m within the terminal boundary. The maximum individual risk to offsite people is 6.8 ×10-5 per year. The individual risk values for onsite and offsite people falls within the acceptable range. F/N curve indicating societal risk falls in the ALARP (“As Low As Reasonably Practicable”) region where risk can be controlled with additional precautionary measures (Chapter 6). Thus, the total risk at the terminal does not lie in the unacceptable region but in ALARP region where substantial measures for risk reduction were needed. The consequences in and around the terminal might be high due to improper implementation of essential precautionary measures. v Accidents such as studied in this thesis may result in huge loss of lives and property, along with widespread environmental damage due to improper coordination and communication in handing the emergency. To address this issue, an automated networking system, named as the electronic - Incident Command System (e-ICS) has been proposed taking into account the results obtained from simulation and modeling of the Indian Oil Corporation Limited (IOCL) Jaipur storage terminal accident (Chapter 7). At the end, recommendations based on the lessons learnt from this study for improving safety aspects of petroleum storage terminals have been made in terms of appropriate preventive and mitigation measures with respect to future accidents (Chapter 8). This is expected that overall methodology of simulation and modeling of vapour cloud explosion (VCE) and large tank fire proposed in this thesis followed by estimation of individual and societal risks can help the regulatory agencies to enhance and strengthen safety measures at petroleum oil storage terminals to prevent accidents and/or mitigate the consequences. The approach evolved and applied in this study can also be used for emergency response planning at petroleum oil storage terminals.