Pollution
Humain
Environnement
Economique

On a tubular exchanger, a disc broke over ¼ of its cross-section at 4:50 am during a pressure surge in the liquid ammonia (NH3) circuit connecting medium-pressure NH3 storage cells to a urea workshop operating under stable conditions. A valve calibrated at 31 bar was also protecting the exchanger; its exhaust as well as that of the disc was collected in a pipe linked to a (12″-diameter) tube shared by all valves on the circuit and then routed to the workshop’s low-pressure chimney. NH3 was partially led to a 100-m high degassing chimney.

Given stable weather conditions, a foul-smelling cloud drifted towards the city; residents notified local fire-fighters and public administrations between 7:40 and 9:30 am and were requested to remain indoors for the morning; 5 ppm of NH3 were measured in the environment, with a peak of 3 mg/m³ at a monitoring station.

The discharge occurred unbeknownst to technicians, who had incorrectly interpreted several alarms that had tripped. Once the diagnosis rendered, the device was isolated at 6:25 am. The operator only became aware of the severity of the event at 8 am; two and a half hours were then needed to fully determine the origin and likely causes.

Announcing 1 tonne of NH3 and then 10 tonnes a few days later, this leak whose media impact was strong was due to a succession of physical, organisational and human malfunctions:

  • The former disc (built 1982) installed in 1995 was removed; an assessment of its condition revealed 0.05 mm less thickness than that on subsequent discs delivered, with a calculated 22.3 bar rupture pressure (i.e. closer to the typical operating pressure than the expected 33.4 bar);
  • Lack of anomaly detection and automatic safety systems: disc rupture sensors, NH3 circuit pressure sensors in the urea workshop, low temperature detection sensors and alarm (frost), hence information made available to control room technicians was inadequate;
  • Circuit/disc design (using a shared collector pipe), with poor valve/gate settings?;
  • Despite unfavourable dispersion, inefficiency of peripheral NH3 sensors (discharge at over 100 m);
  • One of the chimneys not equipped with an NH3 sensor;
  • Insufficient safety report on the liquid NH3 circuits;
  • Poor diagnosis / decision-making process lacking adequate verifications despite several precursors: closing of bottom gates on the medium-pressure NH3 tank, fluttering of the ammonia water tank protection cover, a 15″ NH3 alarm, frost downstream of a valve (its malfunction was wrongly accused), and inefficiency of tank reheating;
  • Incomplete safety recommendations: valve closure, tracking of environmental discharge detectors – insufficient monitoring procedures and inspection plans.

This poor diagnosis would explain: the failure to incorporate ammonia water tank variations, the delay required to isolate the deficient circuit, and the potential impact of this discharge. Long periods elapsed between the onset of the accident, the alarm and activation of the internal emergency plan, source identification, causes and circumstances of the discharge, and then a definitive quantification. Moreover, no alarm procedure was in place between the operator and the local air quality measurement network.

Download the detailed report in .pdf format (97 Kb)