A Study of Oil Spill Contingency Planning Through Spill Data Analysis Narragansett Bay Case Study

The general objective of this thesis is study methods for the development of meaningful oil spill contingency plans and to develop a new oil spill contingency plan for Rhode Island coastal and inland waters. This objective has been pursued in the three parts of the thesis described below. In Part I, oil spill reports from the Rhode Island Health Department and U.S. Coast Guard for the 1967-1971 period were analyzed. The analysis was performed by dividing the Rhode Island Estuary into nine spill regions and by developing the following four categories to describe further the results: spill source, Path-to-Estuary, notification and response. The results are presented in terms of detailed statistics for each region and summary data for the Estuary. General conclusions and recommendations for a new plan are presented based on the oil spill data analysis. In Part II a new oil spill contingency plan for Rhode Island is formulated. It contains a summary of the results of Part I, information on Federal, State and local responsibilities in oil spill situations and lists of data for operational use on equipment and manpower. The conclusions and recommendations of Part I have been incorporated in the new plan. Sections of the. new plan which are of immediate use will be distributed to all plan users. Sections of the new plan requiring legislative action will be proposed for such action. In Part III, the use of a computer simulation to predict the spreading and drifting of an oil spill was examined. This simulation is a possible on-line aid in cases of oil spills in Narragansett Bay. The case of the P.W. Thirtle spill which occurred in 1960 was studied. Using existing data for oil spill spreading and drifting, and a numerical computer hydrodynamic model for Narragansett Bay the Thirtle spill motion was predicted. Other assumptions are given in Part III. The predictions from the simulation were compared with data from The Providence Journal-Bulletin and good agreement was found between the predictions and data.