API - REPORT 79-28
Review of the State-of-the-Art of Oil Spill Simulation Models Phase II
|Publication Date:||1 August 1981|
This report summarizes the results of the second of a three-phase study project Raytheon Ocean Systems Company (ROSC) is under taking for the American Petroleum Institute (API). The objective of the project is to review, evaluate and then provide a comprehensive view of the state-of-the-art, both theoretical and applicable, for modeling the fate of the spilled oil in the marine environment. The result of the project will be recommendations for the construction of an oil fate model framework with state-of-the-art components which is suitable for environmental impact assessment or oil spill risk analysis, exclusive of socioeconomic considerations.
Oils, especially crudes, are a mixture of many different petroleum
components that, upon release into the marine environment, will undergo
various physical, chemical, and biological processes before reaching
their ultimate destination. In general, these processes have been more
specifically categorized into the following processes: advection,
spreading, evaporation, dissolution, emulsification (water-in-oil
emulsion formation), dispersion (oil-in-water emulsion formation),
auto-oxidation, biodegradation, and sinking/sedimentatio
The ultimate aim in modeling the fate of spilled oil is to develop a comprehensive model incorporating sub-models, or algorithms, for each of these processes. Given the uncertainties in many of these processes, however, it may not be feasible or desirable to develop and use such a model since the uncertainties in each of the algorithms would be magnified in the final prediction. It may be more useful to model only the most significant processes leaving the less important ones out of the calculation and there by reducing uncertainty in the outcome. To date, most of the oil spill models simulate only one process - advection. A few models incorporate advection and spreading and a few further incorporate a number of weathering processes such as evaporation and dissolution to provide a more comprehensive oil spill fates modeling. Different techniques with varying degrees of sophistication have been applied in the existing oil spill models to simulate each of these processes. In order to facilitate the assembly of a comprehensive oil fate model with state-of-the-art components, there has to be an evaluation of individual process models. This allows for the selection of whole models or components of models which best represent the state-of-the-art technologies.
Thus, information on some thirty-five oil spill models (Table 1.0-1) was gathered and reviewed during Phase I of this study. There view included the identification in each oil spill model, the processes included, and the model (or modeling technology) used for each individual process (process model). A preliminary evaluation of the spectrum of process models employed by different oil spill models for a given process was then performed to identify process models which represent potential state-of-the-art technologies and which merit further evaluation in Phase II. The main criterion used for such identification is that the selected model has a unique treatment which makes the model more realistic, reliable, valid, theoretically sound, or applicable than others. These selected component process models are summarized in Table 1.0-2, which is taken from the Phase I addendum report.
The goal of the Phase II evaluation is to attempt to identify the model(s) which "best" represent(s) the state-of-the-art technology, for a given process so that a comprehensive oil spill fates model with state-of-the-art components can be assembled. The determination or the assessment of which model(s) "best"represent(s) the state-of-the-art technology for a particular process can be made by evaluating various models against the following general criteria:
• Theoretical soundness or realism
• Level of data input (amount, sophistication, availability)
• Application (applicability, range, ease, cost)
• Accuracy of output
In order for a model to be acceptable for application, it must be based on theoretically sound and/or realistic assumptions. In considering the soundness of the assumptions one must not over look the practical aspect of the problem, namely, such that the model could become overly complex to the extent that its outputs are practically unverifiable or that its requirements for input data are too large or too unrealistic to be practical. Of course there liability of a model is also related to its verifiability. The applicability of a given model is the next factor to be considered in the model evaluation. A model with applicability to a wide range of geographic and physical conditions is certainly a desirable one, but such a feature should not be overburdened by excessive cost or at the expense of ease of operation. Thus, in considering applicability one must also consider the associated cost and ease of application. Finally, the model should not only produce realistic results, but also accurate results. Accuracy is the ultimate goal toward which every modeler strives. Unfortunately, this last criterion is usually not attainable for most of the oil spill models. There is a general lack of laboratory or field data collected during actual or experimental spills that is useful for the assessment of model accuracy. There is even a lack of sufficient data to validate a model. Therefore, the evaluation of the various selected models is generally limited to assessing: the realism of the inherent or underlying assumptions made in deriving model formulations; model verifiability and input requirements; the extent to which the model may be applicable, including ease of application and the associated computational effort.
This report is organized into four sections. Following this introductory section, Section 2 provides details on the evaluation of the various models, on a process-by-process basis. Section 3 discusses the synthesis of various component processes into a comprehensive oil fates model. Finally, Section 4 briefly summarizes the results of this evaluation phase.