Most sites can then be left to recover naturally since this type of treatment protocol has been demonstyated to minimise long term environmental impact (Baker 1995). In this paper we will discuss the easibility of additional bioremediation treatments to enhance the natural remediation process in sheltered environments.
In Alaska, following the EXXON VALDEZ incident, bioremediation was used to treat residual oil remaining within the sediments after the initial cleanup operations (Pritchard & Costa 1991 ; Bragg et al. 1994 ; Swannell et al. 1994) . The addition of sources of nitrogen and phosphorus by scientists from both the US Environmental Pro-tection Agency (US EPA) and Exxon was found to stimulate the biodegradation rate of the oil and hence promote a more rapid decontamination of the oiled shorelines (Pritchard et al. 1992 ; Bragg et al 1994).
A randomised block design was used to test the effect of 2 treatments at the site selected, Bullwell Bay. 3 blocks of 3 plots (each 9 m long by 0.9 m wide) were placed perpendicular to the sea 1.25m apart. An initial lithium tracer experiment had shown that this was adequate to prevent migration of the treatments between the plots. In each block one plot was left as a control, one was treated with, a weekly application of fertiliser (sodium nitrate and potassium dihydrogen phosphate) dissolved in seawater, and one was treated with a slow release inorganic fertiliser. At the beginning of the experiment the Total Petroleum Hydrocarbon (TPH) measurements, made colorimetrically, showed that the contamination was fairly consistent across all plots, with the highest oil concentration persisting at the landward end of the plots.
The conclusions of the bioremediation treatment at Bullwell Bay were :
・ The in situ rate of biodegradation of a mixture of HFO and Forties Blend was increased.
・ The treatments had no effect on the nutrient content of the seawater in Bullwell Bay.
・ There were no detectable toxic effects on the environment as determined by the sensitive oyster embryo bioassay
・ The slow release and weekly application of fertiliser were equally effective implying that the former may prove a cost effective technique for enhancing the natural recovery of shorelines.
SUMMARY OF SHORELINE CLEAN-UP ACTIVITIES
Despite the release of around 72,000 t of Forties crude oil and 480 t of HFO we estimate that only around 10,000-15,000 t of Forties emulsion and around 430-460 t of HFO impacted the shoreline. Around 2% of the oil released was recovered by the 29 February, 3-7% remained stranded on the shoreline in areas where it was difficult to remove further bulk oil.
Therefore, shoreline cleanup operations focusing on the stranded oil, which in fact comprises only around 3-7% of the total oil released, has been the main on-going activity since March 1996. The techniques employed on the high energy shorelines have been natural remediation, surf washing and dispersant application. The techniques used on the low energy shorelines have been natural remediation, and flushing. Bioremediation as a final polishing step has been shown to be appropriate for Bullwell Bay.
CONCLUSIONS
As a result of the grounding of the SEA EMPRESS, 72,000 t of Forties Blend oil was released into the envi-ronment making this incident among the 20 largest oil spills of all time. With 35-45% evaporating the potential was for 40,000-47,000 t of oil to come ashore. Since Forties Blend oil rapidly emulsifies to produce a 70% water-in-oil emulsion (Walker et al. 1993 ; Walker et al. 1995 ; Walker & Lunel 1995) this could have translated into 120,000-140,000 t of emulsion impacting the South Wales coastline if dispersants and mechanical recovery had not been used.
Fortunately, the result of the combined dispersant and mechanical recovery operation was that only around 10,000-15,000 t of emulsion impacted the shoreline. The mechanical recovery operation accounted for around 2,000 t of oil while it is estimated that 21,000-32,000 t of oil was dispersed.
The value of having a dispersant operation in place as a first response (along with monitoring of the spill) as part of the UK national contingency plan was clearly demonstrated. It was only because the MPCU control structure, surveillance aircraft, dispersant spray planes, trained pilots and monitoring teams were all in place in advance of the spill that a targeted and successful dispersant operation was feasible at this incident. Through NETCEN'S monitoring programme we have shown that the dispersant operation significantly enhanced the
