Fire in a ammonia production plant

The incident occurred at an ammonia production plant. The plant was starting, when a section of pipe between the secondary reformer and a waste heat boiler (“transition pipe”) ruptured. Process gas was released and ignited.
The ruptured transition pipe was a special component designed to contain process gas above 600 PSI (40 bar) at more than 1,600 degrees Fahrenheit (870 degrees centigrade). When the pipe ruptured, 4900 pounds (2.2 metric tons) of hot process gas were released into the air and ignited by auto-ignition, resulting in a jet fire. The intense heat from the jet fire caused the transition piping to expand and rupture further culminating in a large fire.

The composition of the process gas was a non-quantified mixture of hydrogen, nitrogen, methane, carbon monoxide, and carbon dioxide.
The transition pipe comprised four layers: an inner metal liner, an insulation layer, and an outer metal shell, all surrounded by a metal jacket containing water to maintain a consistent temperature of the shell’s outer surface.

The company’s investigation concluded that:
(1) The initial failure resulted from creep damage due to prolonged exposure to stress at elevated temperatures along a circumferential weld in the transition pipe’s shell. A lack of post-weld heat treatment contributed to the transition pipe’s short life.
(2) The larger failure of the transition pipe’s shell base metal resulted from short-term overheating from the initial jet fire.
(3) Damage to the pipe insulation liner and to the insulation layer, as well as the water jacket operating (at times) with only a partially filled liquid level, all contributed to the higher shell operating temperatures that led to the creep damage.

The INITIATING CAUSE was cracking of the pipe due to creep.
The ROOT CAUSE was low material quality, probably lack of proper post-treatment of a weld. CONTRIBUTING Factors to too high temperatures were some damages to the pipe and the partial inefficacy of its cooling system.

reformer, piping

The plant was being restarted.

The CSB report does not mention any lesson learned; it proposes nevertheless two types of causes for the high-temperature creep-driven failure of the piping:
(1) A material quality issue: a microstructural shortcoming which made part of the pipe more sensitive than expected to high-temperature creep. This shortcoming could be possible attributed to lack of post-weld heat treatment, which shortened pipe’s life.
(2) An operative issue: contribution to a shorter-than-designed life came from damages to the pipe insulation liner and to its insulation layer and the fact that the water cooling sometimes not at the nominal level.

For a plant operator it is impossible to detect micro-structural defects of the component installed. Due to the complexity of the pipe structure (4 layers: an inner metal liner, an insulation layer, an outer metal shell, all surrounded by a metal jacket containing cooling water), it was probably very difficult to detect degradation during inspection and maintenance procedures.
The only possible improvement suggestion which could be drawn is regarding the improvement of the active cooling of the pipe, and an improved recording of the operative parameters, above all local temperatures.