established. In actual fact, In addition to the measured wind drift of the 

 oil, a net surface current on the order of 0.6 knot in a southeasterly direc- 

 tion was observed during that period. In less than 50 m of water, it appeared 

 that the net currents were responding well to the wind. The pancakes of oil 

 emanating from the wreck were observed to build up in thickness as they moved 

 away from the wreck. After 1 or 2 weeks of movement, thick patches of oil 

 which were originally 1 1/2 - 2 inches became 5 to 10 inch thick patches. 

 The ethereal "3%" wind factor that has been tossed about freely for years is 

 now being pinned down. Patches of Avgo Merchant oil were measured moving 

 relative to the water at 0.7 to 1.1% of the wind speed, for wind speeds of 10 

 to 30 knots and oil thickness of 1 to 2 inches. The thinner sheens covering 

 much of the sea surface appear to be "fed" by the thick patches so that their 

 movement is limited to the 1% wind factor as well. This wind factor probably 

 represents the effect of energy transfer from waves interacting with the oil. 

 There is a wind-induced surface current, amounting to about 2% of the wind 

 speed, that also needs to be considered when predicting oil movement using 

 subsurface current information. However, when drift cards or bottles are the 

 source of current data, the "3%" figure is excessive as a wind factor and 

 should be replaced by a figure of about 1%. 



A large amount of information has been gathered which has been and will 

 be of value in improving both operational forecasting of real oil spill tra- 

 jectories and statistical models of oil spills from a "risk analysis" point 

 of view. It is inadequate to use look-up tables for currents combined with 

 statistical models of wind, or vice-versa. In the near-shore environment, 

 the winds and currents are too highly correlated for the above approach to be 

 adequate. Moreover, for short-term forecasting, tidal currents as well as 

 wind drift should be included for realistic output. For real-time forecast- 

 ing, the value of accurate slick maps cannot be understated. Accurate meas- 

 urements of oil/water differential velocities, the observation that pancakes 

 build up in thickness rather than disperse, and the underslick cinematography 

 will all serve to improve the state-of-the-art in oil spill modeling. 



5.2 Fate of the Oil 



The Avgo Mevohant No. 6 fuel is composed of about 80% straight-run No. 6 

 and about 20% light distillate fuel (which was used as a "cutter stock" to 

 make the oil easier to handle) . It now appears that some of the cutter stock 

 entered the water column, at maximum levels on the order of 250 parts per 

 billion. Highest concentrations were found under fresh oil slicks though not 

 in the near-surface samples, but at depths of 2-3 m. Concentrations decreased 

 after a few days to background levels through turbulent mixing of the homo- 

 geneous water column. More definitive chemical analyses are being conducted 

 to verify these findings. 



Severe artifical weathering of a cargo sample indicated that the 

 straight-run No. 6 component retains a positive buoyancy and will not sink 

 unless aided. The oil found in the vicinity of the wreckage is associated 

 with shell fragments in the sediments and would otherwise rise due to its 

 natural buoyancy. The U.S. Navy divers reported no visible oil on the bottom 



128 



