Pollution
Humain
Environnement
Economique

At 10:30 am, a watchman at a Seveso-rated hydrocarbon depot noted a large flow of crude oil in 2 retention basins. The leak was located near a roof rainwater drain, running inside the tank. Initially spilled into the first basin, the majority of product then migrated (via a natural underground drain?) into a second basin downslope. The discharge alarm did not activate in either basin despite the presence of hydrocarbon detectors in both. The technician closed the drain’s bottom valve and sounded the alarm. After verifying the product’s lower flammable limit and the absence of H2S, internal fire-fighters recovered a portion of the product in the retention basin; this receptacle however was not perfectly sealed. Of the 100 m³ discharged, 3 m³ infiltrated into the ground.

Informed that morning, the inspection authorities for classified facilities visited the site the next day. The depot operator provided inspectors with a set of operating guidelines that had not appeared in the procedures submitted to administrative authorities.

Upon analysis of the hydrocarbon detectors, it was found that during winter, the existence of ice generated a current intensity similar to that observed when hydrocarbons were present. Since these detectors had been considered as instrumented risk control measures in the safety report, a technical study on their reliability along with the immediate compensatory measures required to remedy this deficiency were demanded.

The operator was served an injunction by way of Prefectural order to conduct a diagnostic assessment of environmental conditions and devise a pollution control plan incorporating a cost-benefit analysis within 3 months. The operator was also required to propose a timetable for basin sealing works.

This leak stemmed from a break in an articulated arm replaced in 2010 and fitted to the roof drain. All of the site’s articulated arms were replaced by a product of a different brand. New frost-resistant detectors with positive safety features were installed (in measuring electrical conductivity rather than electromagnetic energy absorption).

Three sampling campaigns, carried out in November and December 2012, using the depot’s piezometers indicated no pollution. The consultant assigned to conduct the environmental assessment concluded that the product’s vertical migration was blocked by the presence of various clayey layers under the damaged basin. A total of 920 m³ (1,500 tonnes) of fouled earth were excavated at the basin bottom between 12 March and 19 April 2013. The content of materials remaining at the pit bottom upon completion of the clean-up operation was less than 800 mg/kg of dry mass.