Comparison of 5 oil-weathering models

Abstract

Within the project CALPREA at RIKZ application of reliable oil spill models and oil data bases is an important issue. An oil spill model contains in general a transport module, an oil weathering module, an oil data base, and possibly a user-friendly graphical user interface. Recently, de Jong (2004) has made a study of four particle models for transport of floating and dissolved substances (DemWaq, DREAM, GNOME and SIMPAR) within CALPREA. The present report gives a detailed comparison of five pre-selected oilweathering models. Oil weathering models were selected on the following criteria: 1. They can be coupled to the hydrodynamic programs WAQUA (2d) and TRIWAQ (3d) in SIMONA at RIKZ; or 2. They are at present available to RIKZ through the Internet or other through means and are user friendly The oil-weathering modules of category 1 for this study were: - SIMPAR oil module - MFMW-DREAM - GNOMF The oil-weathering modules of category 2 for this study were: - ADIOS - OWM CONCLUSIONS AND RECOMMENDATIONS Considering documentation: • Technical documentation of the oil-weathering models ADIOS-2 and DREAM/OWM is not very accurate. Especially, correlations between TBP-curves and vapour pressures (and molecular weights) for oil fractions cannot be found. • For GNOME a technical document does not exist. • For SIMPAR a first version of a technical document exists but it has not been accepted yet in SIMONA since the module is a beta-release. Considering test results: • Results are very sensitive to the input parameters used in the models (e.g. see differences between SIMPAR-O and SIMPARA results). Thus the existence of a reliable oil databases is a prerequisite for use of these oil-weathering modules. • OWM seems to be most reliable since its oil database is fully based on lab experiments. For ADIOS-2 this is not known and for SIMPAR there is no oil data base, whereas for GNOME it is limited to 6 oils; • However, the OWM model gives much larger oil-in-water dispersion than the other two models, nearly a factor 10 after 1 day of simulation. There is no good explanation for this difference and this needs further investigation. Recommendations: 1. With respect to oil weathering modules: • In general, more research is needed during real operational conditions during oil spill incidents. Especially, for strong wind conditions the models need to be validated. • The present testing of the oil-weathering models should be extended with more oil types. Also the effect of oil fractions on emulsification must be investigated in more detail. • Discussion with SINTEF is required, especially concerning the oil-in-water dispersion results obtained with OWM and DREAM. • Discussion with NOAA is required concerning the quality and openness of the oil data base of ADIOS-2. 2. With respect to coupling to particle tracking module of SIMONA: It would be interesting to couple the SIMPAR algorithm of particle tracking to the DREAM (or OWM) oil weathering module and its data base. SIMPAR has the following advantages: a) SIMPAR has drying and flooding. b) Particle tracks depend on water depth whereas this is neglected in DREAM. In shallow areas this leads to different results (see also de Jong, 2004). c) SIMPAR was validated successfully by Directorate Zeeland with tracer experiments and it is therefore well accepted within Rijkswaterstaat. d) End of 2005 SIMPAR will have a backtracking function. e) In a beta-release of SIMPAR there is 3-dimensional transport including an accurate numerical scheme for vertical diffusion. For simulations of dissolved matter in stratified waters like the Dutch coastal zone this is an important issue.