The seabed on which the wreck lay, about 30 miles south of Penmarc'h and 50 miles west of Belle-Ile, was a vast expanse of muddy sand.
Currents did not exceed 0.8 knots. The bow lay at a depth of 114 metres and the stern at a depth of 128 metres. In January, the sea temperature, a key element, at the bottom was measured at 9 °C. It does not exceed 12°C in the summer.
The bow lay upside down and straight on the sea floor. Its external general condition appeared to be fairly good. The buckling it suffered seemed superficial, except in the fractured zone at the back of the section. The stern lay on its keel. The site was strewn with metal sheets and debris torn off the tanker when she broke up. The hull and deck's external structure was on the whole intact, except in the ruptured zone of the hull, at the front of the section. The afterdeck looked in rather good condition, 30 metres above the sea floor.
An initial mission to locate the various parts of the wreck was carried out by the French Navy mine hunter the Pégase, only a short time after the sinking. The exact position of the two wrecks was thus stated accurately. A SPF (Self-Propelled Fish = remote-operated robot equipped with a camera) captured the first images of the bow. The visibility conditions as well as the size of the wreck compared with the field of view of the robot made the interpretation of the images tricky. The supply ship the Abeille-Supporter implemented its exploration ROV (Remote Operated Vehicle = submarine robot) Abyssub on the back section of the wreck on the evening of 31 December. A preliminary investigation took place on 1 and 2 January 2000, formally identifying the wreck and showing its position on the sea bottom. A detailed investigation was then carried out by the research vessel Marianos, chartered by TotalFinaElf. It provided accurate information for setting up a method to neutralise the oil remaining in the wreck.
TotalFinaElf committed to neutralising the danger represented by the cargo and bunker trapped in the wreck of the Erika, under the control of the French Government. A survey showed that pumping offered the best advantages, particularly as regarded security and environmental protection when implemented.
Once pumping had been chosen as the best solution, various technical options were proposed by companies specialising in underwater response. The technique selected by TotalFina was validated by the Government's steering committee and group of experts. It consisted in drawing off the hydrocarbons by difference in hydrostatic pressure between the wreck and an intermediate tank lying on the seabed and maintained in controlled relative depression. In order to facilitate their transfer, the oil was to be fluxed by mixing with a thinner (methyl ester of rapeseed), harmless for the environment, as soon as it flowed out of the wreck and before reaching the intermediate tank. Once it was in the tank, the oil-thinner mixture was to be sent towards the surface in a continuous flow through a double-screw pump. A phase of fine cleaning of the wreck, comprising the elimination of the maximum amount of residual oil remaining after pumping, was envisaged by injecting rapeseed methyl ester.
The French-Norwegian consortium Coflexip/Stena Offshore/Stolt Offshore was given the responsibilty for carrying out the operations. The preparation of the two parts of the wreck, 10 km apart, began in mid-May by a reconnaissance phase and by the contractors setting up the work site. Hull drilling and the installation of the connecting systems to the pumping device started at the beginning of June. This was warried out by remote-controlled robots (ROV) and by divers who worked more particularly on the areas that were difficult to access. This phase started with the bow before beginning on the stern, which was more complex because of the obstruction of the deck. A little over 10,000 tonnes of heavy fuel were recovered during the main pumping operations. Fine cleaning added a further 1,200 tonnes.