Oil spill risk assessment and management for the Coast of Algarve, Southern Portugalan operational oceanography approach

Resumen

Oil Spill Risk Assessments (OSRAs) have been widely employed to support decision makers in managing conflicts of interests associated to the oil industry. On the one hand, oil production, transportation and storage attend a social demand but, concomitantly, represent a source of risk to the coastal and marine environment. On the other hand, it is desirable to preserve the coasts and to keep them profitable in monetary (e.g.tourism and fisheries) and non-monetary terms (e.g. nutrient cycling, gas regulation). In the first part of this thesis, it was demonstrated that the available literature/methodologies in OSRA often fails in fulfilling basic requirements necessary to support the decision making: (1) uncertainties in the risk estimates have been neglected, (2) operational oil spills (i.e. intentional small, but frequent, spills associated to vessel operations such as tank washing) have not been addressed and (3) the risk analysis outputs are not appropriate for the their communication to the stakeholders. Relying on these conclusions, a general OSRA framework was proposed based on a critical analysis of the ISO 31000:2009 on risk management principles and guidelines, and addressing the limitations observed in the reviewed literature. The methodology, that employs ensemble numerical oil spill spill simulations to quantify the risk and its uncertainties, was applied to a real oil spill case, the explosion of the Jiyeh power station in Lebanon in 2006, being able to identify the most impacted areas and to visually communicate the risk, its components and its uncertainties. The framework developed in the first part of the thesis was later applied to compute the oil spill risk in the Algarve, southern Portugal. Improvements in the methodology, now called Information Technology OSRA (IT OSRA), were necessary in order to address dispersed sources of risk (i.e. maritime traffic), more sources of uncertainty (i.e. where and when the spill will happen, and oil spill model configuration) and the possibility of both accidental (rare but usually involving large volumes of oil) and operational spills to occur. Over 50,000 oil spill simulations were performed and the results obtained confirmed that oil spills associated to the maritime traffic represent a risk to the Algarve in ecological and socioeconomic terms. Significant seasonal variability of the risk was observed and quantified. High frequency variability (by the order of days) of the meteo-oceanographic variables was also found to play an important role in the oil spill risk. Finally, priority areas were identified based on the risk maps and the most likely sources of potentially impacting spills were mapped. The huge number of simulations performed allowed to discover that the distribution of concentrations of oil on the coast due to marine spills follow a Poisson curve. Such finding challenges the available literature that assumed a Gaussian distribution of the variable. The discovery will demand new approaches to deal with the oil spill risk.