Skip to main content
Clean Hydrogen Partnership

Release of hydrocarbons mixture from a pressure vessel

Event

Event ID
567
Quality
Description
A release of contents of a pressure vessel occurred when a longitudinal weld tore open. The vessel had been shut down and had just been put into use again when the incident occurred.
The gas mixture contained hydrocarbons with 30% hydrogen. It had reached its working pressure of 31 bar, but was only at -26 degrees C, instead of the working temperature of -73 degrees C, no liquid was present.
An investigation found a crack, 1.6 metres long, had formed near the upper end of the weld.
Event Initiating system
Classification of the physical effects
Unignited Hydrogen Release
Nature of the consequences
Leak No Ignition (No additional details provided)
Macro-region
Europe
Country
Germany
Date
Main component involved?
Gas Storage Tank (Weld)
How was it involved?
Rupture & Formation Of A Flammable H2-Hc-Air Mixture
Initiating cause
Material Degradation (Generic)
Root causes
Root CAUSE analysis
The recommendations issued after the incident, suggest that the ROOT CAUSE could be a lack of proper inspection, and/or inspection performed with inadequate tools

Facility

Application
Chemical Industry
Sub-application
unspecified
Hydrogen supply chain stage
All components affected
weld, vessel
Location type
Unknown
Location description
Industrial Area
Operational condition
Pre-event occurrences
The vessel had been shut down and had just been put into use again when the incident occurred.
Description of the facility/unit/process/substances
PROBABLE PROCESS
Although not specified by the source, the event occurred very probably at a vessel used to distillation and gas separation processes of the refinery, for example in what is commonly called a ‘cold box’. This assumption is based on temperature data provided. Propane, for example, liquifies at -42 C, it is therefore still gaseous at 31 C (the temperature at which the release occurred) and liquid at -73C, corresponding to the liquid nitrogen temperature, which probably was used as coolant vector,

Emergency & Consequences

Number of injured persons
0
Number of fatalities
0

Lesson Learnt

Lesson Learnt
Inspection procedures had not been developed for many operations, including measuring temperature from outside field panels underneath the reactors. This hints at a poor process hazard assessement, that did not even reflect the actual equipment degradation and instrumentation used in the process.
The following recommendations were made:
1. All vessels of similar construction to be tested for incipient cracks on the inner surface by using a dye penetration test.
2. If the interior is inaccessible, welds and impact zones are to be tested by ultrasonic methods.

Event Nature

Release type
gas mixture
Involved substances (% vol)
hydrocarbons 70%
H2 30%
Actual pressure (MPa)
3.1
Design pressure (MPa)
3.1
Presumed ignition source
No ignition

References

Reference & weblink

Event nr 12735 of the UK database ICHEME in PDF format.<br />
https://www.icheme.org/knowledge/safety-centre/resources/accident-data/… />
(accessed October 2020)<br />

Primary source in Chemical Hazards in Industry, No. 1 January 1998.<br />
(unavailable)

JRC assessment