Explosion at a pipe in a nuclear power plant

An explosion occurred inside a pipe of a boiling water nuclear power reactor. The pipe was part of a core spray system, which was providing extra cooling when shutting down the reactor. The system did not have a safety function during operation. The explosion tore apart the pipe, with loss of radioactive steamThe 10 cm diameter pipe had totally disintegrated over a length of 2 to 3 meters. About 25 pieces of debris scattered around and 2 to 3 m of pipe was completely gone. The leak was stopped by closing an undamaged shutoff valve.

An explosion detector was triggered during the accident and other monitoring devices briefly delivered readings at the scale limit. However, the nature of the event was not fully recognised by the operators, who isolated the leak and continued operations.

The investigations suggested that the hydrogen (formed by radiolysis of the water) accumulated in a part of the pipe between catalytic recombiners.
The explosion was close to a valve connected to the reactor pressure vessel, but the valve was closed at the time and was not affected. A subsequent investigation of all German BWRs indicated that the design had a unique vulnerability to this type of event.

The INITIATING CAUSE was the formation of an explosive hydrogen-oxygen mixture in a section of a pipe.

The cause of the explosion was the formation of an explosive hydrogen-oxygen mixture, produced by radiolysis in all boiling water reactors during regular operation. The risk that the gas could collect in the pipe leading from the reactor lid and explode was overlooked in the safety assessments.

The ROOT CAUSE was the lack of knowledge of the phenomenon and its related hazards. When designing and operating a system containing water steam, it was assumed that no non-condensable gases (hydrogen, oxygen, nitrogen, etc.) could exist in its pipes.
The fact that the operators considering the event a non-safety-related failure and continued the operation without reporting, while the formal investigation found it a case requiring operative actions and reporting, indicates a misunderstanding or non-clarity of the legal procedures.

Pipe, core spray system, radiolythic hydrogen

DESCRIPTION OF THE FACILITY
The explosion took place in the so-called reactor core spray system. That system sprays cold water into the reactor vessel to cool the reactor down and remove residual heat in case of shutdown. It is not an emergency cooling system but a tool to reduce by about two hours the time needed to cool the vessel in normal maintenance shut downs. The spray system situated inside the containment vessel, a steel vessel of 27 meters in diameter in which the reactor vessel itself is located. The system is connected by an isolation valve to the reactor vessel. The valve is opened to allow injection of fresh coolant into the vessel during shutdowns.

Property losses are unknown but were probably limited to the replacement of cooling system affected by the explosion. Much more relevant was the loss of incom due to the long period of shut down of 2 years.

The Federal Ministry for Environment, Nature Protection and Reactor Safety was only informed by the Schleswig-Holstein authority on 18 February, the day that the operator agreed to reduce power output of the reactor. The federal ministry ordered further investigations and to be informed on measures to be taken. The ministry will only allow the reopening of the reactor if the cause of the explosion is clear and a repeat is excluded, all damage is repaired and the competence of the operator is proven. That competence is in doubt as the operator decided to continue operation after the explosion instead of starting inspections after the observed leakage.
Sources:
Bericht des Bundesministeriums, 25 February 2002
Press release, State of Schleswig-Holstein, 5 March 2002
Nucleonics Week, 5 March 2002

Despite the loss of radioactive water, this incident had no impact on external areas of the station, and involved a system which did not have any safety function (its goal was to accelerate the cooling done of the reactor after shutdown).

However, the incident had its root in a design vulnerability which could have triggered worse accidental scenario. This was not recognised at that time, due to the lack of knowledge on the possibility of formation of hydrogen-oxygen mixture in the steam pipelines of the reactor.
Since the explosion had damaged a non-safety-relevant part of the plant, the operators interpreted the incident as a "spontaneous leak" without reporting requirement to the authority. The reactor continued to operate at full capacity for several weeks without further investigation.
Following the conclusion of an official investigation, the event was classified as INES 1 (i.e. Anomaly, Impact on defence-in-depth).

See similar event HIAD_728occurred one month earlier in a Japanese BWR, for a summary and refences on the problem of formation of explosive atmosphere in BWR pipelines.

In addition to removing this pipeline and eliminating weaknesses in radiolysis gas control, extensive personnel and organisational measures were implemented to address deficiencies in safety management. The plant was restarted on March 25, 2003.
All other boiling water reactors in Germany were investigated for the possibility of similar explosions.