Skip to main content
Clean Hydrogen Partnership

Hydrogen explosion in a chemical processing plant

Event

Event ID
1004
Quality
Description

The event occurred at manhole cover of a of a pilot plant reactor where hydrogenation had been performed in a hydrochloric aqueous solution.
The reactor had been flushed three times with nitrogen (i.e. nitrogen injecting at 3.2 bars and then relieving).
An employee opened the manhole cover of the reactor to take a sample of the reagent. . After opening it, a minor detonation occurred.

According to the DECHEMA database, the investigations found out these following causes:
1. The geometry of the reaction vessel was such that a hydrogen bubble remained in the area of the manhole despite three flushing operations with inert gas.
2. catalyst material was deposited in the area of the manhole. This probably occurred during insertion of platinum catalyst and/or during stirring. This material constituted an ignition source, once dried out..
3. On opening the manhole cover, air was able to penetrate into the upper part of the reactor and created an explosive mixture with the remaining hydrogen. This mixture was then ignited by the baked-on catalyst.

Event Initiating system
Classification of the physical effects
Hydrogen Release and Ignition
Nature of the consequences
Macro-region
Europe
Country
Germany
Date
Component involved
reactor / oven / furnace / test chamber
Failure mode
rupture & ignition
Initiating cause
inadequate or no purge
Root CAUSE analysis

The INITIATING CAUSE was the introduction of air into the reactor which was still containing hydrogen. The ROOT CAUSE was related to an inadequate purge, and the lack of a procedure for sampling reactor composition before opening it to air.

Root causes
Date entry in HIAD
01/01/2019

Facility

Application
Chemical industry
Sub-application
Hydrogenation process
Hydrogen supply chain stage
All components affected

hydrogenation reactor; manhole; platinum catalyst;

Location type
Confined
Location
industrial area
Operational condition
Pre-event occurrences

The operations causing the incident occurred at the end of the reaction process, when sampling the final product.

Emergency & Consequences

Number of fatalities
unknown
Number of injured persons
unknown
Post-event summary

The DECHEMA incident database, event no. 12/ 2003) reports only a 'small detonation'.

Lesson Learnt

Lesson Learnt

Conclusions taken from the DECHEMA database, event no. 12/ 2003:
1. For purging a plant, multiple evacuation and removal of the vacuum with inert gas is always to be preferred to flushing with inert gas or forcing in the inert gas under pressure followed by pressure relief. Especially in the case of hydrogen, forcing in the inert gas under pressure always involves a risk of inadequate mixing.
3. If the plant is not vacuum or pressure resistant and therefore only the flushing method can be used for purging, the plant geometry must be such that the gas phase can be completely displaced by inert gas. Attention must be paid to the direction of flushing (in this case from the bottom upwards).
4. In technical terms, sampling should preferably be carried out such that it is not necessary to open the manhole cover on the reactor (e.g. through sampling valves).
5. When handling flammable or highly flammable gases or liquids, all potential ignition sources are to be identified for all process steps in which the generation of explosive atmosphere is possible or conceivable and suitable measures are to be specified as part of an explosion prevention concept.

In-depth data

Release type
gas-solid mixture
Involved substances (% vol)
H2,
Pt
Probable IGNITION SOURCE
Explosion type

References

Sources categories
DECHEMA
Reference & weblink