Fire at a chlorine electrolyser

Two successive fires broke out in an interval of few minutes in four cells in the electrolysis room of a chlorine production plant.
The technicians brought the fire under control using the fire-fighting equipment located nearby. In both cases presence of hydrogen, ignited by a short circuit spark, was responsible for the two accidents.
Case (1): the hydrogen leaked when a rubber-plastic pipe was manually disconnected from a hydrogen trap. Despite the fact that that this operation is covered by a specific procedure, it was performed when hydrogen quality was not good enough to be collected.
Case (2): the hydrogen leaked when a sleeve failed locally by corrosion.

The two sub-events have the same INITATING CAUSE, a hydrogen fires during operation.

The two fires have different two ROOT CAUSES. In case (1) was a wrong operation (in presence of a formal procedure). In case (2) it was probably a design or inspection shortcoming (corrosion). The following investigation identified FACTORS related to design

The IGNITION SOURCE of the released hydrogen was an electrical spark caused by a short-circuit. Different potential causes of the short-circuit were identified:
- the compatibility of insulating material has not been checked
- lack of wet-proofing between anode and cathode
- not enough distance between anode and hydrogen pipe due to a modification of the connection and a miss-alignment of the cells

Electrolyser's cells, sleeve, pipe

In both cases, the fire must have been of a very low itnensinty, and rhe technicians brought it under control using the fire-fighting equipment located nearby.

The fire department was called to the scene and participated in the final shutdown of the hydrogen system as the fire was extinguished prior to their arrival. The partially filled tube trailer was disconnected and moved away from the incident area to a safe location.

A Notice of Violation was issued by the fire department that prevents the start-up of the liquid hydrogen system until all necessary safeguards are in place to re-start the system.

A lack of information on the system, components and the operations hinder the understanding of this event. Which electrolysis technology was used? What does it mean that "Too low hydrogen quality" was the reason for the first fire? Which was the operation foreseen on the hydrogen trap? Where was the failed sleeve located? Was the system operating normally or was it under maintenance?

The only possible general lesson emerging is the need for a quality check and control when reporting an incident. If the original event description provided by an operator or personnel is badly written, all the following up reporting cannot be better. Too often, the direct involved reporting person insider terms and assume the context known to the reader. Part of the safety training of personnel should focus on the important of a well written Operative Incident Report.

The analysis of the accident and the corrective actions required were presented to the company Committee “hygiene, safety and working conditions” before operations could resume. These included:
(a) choice of a better suited insulating material,
(b) improving process orifice to reduce the presence of chlorine in the hydrogen and avoid disconnection of the pipe from the trap,
(c) improving quality control and supervision by team leader,
(d) re-designing hydrogen flanges to space them out from the anode,
(e) communicating the from the accident to the staff,
(f) improving procedures and guidelines,
(g) taking stock of all insulation malfunctions of all cells.