| Case Name |
Explosion and fire caused due to accumulation of methyl hydroperoxide at a methanol rectification column of a surfactant manufacturing plant |
| Pictograph |
|
| Date |
June 26, 1991 |
| Place |
Ichihara, Chiba, Japan |
| Location |
Chemical factory |
| Overview |
An accident occurred at a new in-house development plant. A pH meter, which controlled and adjusted pH in a neutralization process at downstream of a bleaching process, failed. However the operation was continued without noticing it. Therefore, the operation for neutralization was conducted in acidic condition and the peroxide which was generated in a bleaching process was not decomposed. So, the liquid charged to the methanol rectifying column contained the peroxide without being decomposed. Due to a shutdown procedure of the column, the peroxide condensed on a certain tray locally and the peroxide exploded.
Generation of peroxide had not been known, and the pH control was used only for quality control and there was no idea that pH control was effective for decomposition of the peroxide.
There were oversights of the self-development process and inevitable aspects due to insufficient information, but it could be also said that the accident was caused due to underestimating of the malfunction of the pH meter. |
| Incident |
An explosion occurred at No.1 plant of a company's own development process at the initial stage of commercial production, which followed a trial run. A methanol rectification column was pressurized abnormally and an explosion occurred in the column which recovered solvent. One third of the upper part of the column burst and debris of the column was scattered. |
| Processing |
Manufacture |
| Individual Process |
Distillation |
| Process Flow |
Fig2.Unit process flow
|
| Substance |
Methanol, Fig3 |
| Hydrogen peroxide, Fig4 |
| Methyl hydroperoxide, Fig5 |
| Type of Accident |
Explosion, fire |
| Sequence |
In January, 1991: The plant was completed. Trial runs were repeated several times. The production schedule was from June 19th to June 26th.
On June 25th: The pH meter of the neutralization process did not function well. As charge rate of alkali for neutralization was controlled automatically, pH was kept in acidic condition. From that night, the quantity of sodium hydroxide for neutralization was adjusted by analytical values of sampling.
On June 26th about 08:09: A sulfonation reaction finished and the plant was going to shut down.
08:41: The neutralization process was stopped to check the pH meter.
08:50: The bleaching process, which was prior to the neutralization process, was stopped.
09:06: Supply to the methanol recovery column was stopped, because there was no recovery methanol to the methanol rectification column. After that, with refluxing, switching to a holding operation for drawing out inventory in the tower and total reflux operations were conducted.
10:10: The operation was shifted to a boil-up operation for recovering purified methanol.
10:15: An explosion occurred at the methanol rectification column. |
| Cause |
According to the results of a study after the accident, methyl hydroperoxide was produced in a bleaching process, which was two-process-upstream from the process where the explosion occurred. The peroxide was produced by a reaction between methyl sulfuric acid and hydrogen peroxide in acidic condition, and methyl sulfuric acid was produced from methanol and sulfuric acid anhydride which remained in bleaching acid. In usual operations, the peroxide is kept to a small quantity with injecting sodium hydroxide, however on the accident day the peroxide was not decomposed and charged to the column. The rich peroxide was condensed through total reflux and total rundown operations at the middle of the column, and its concentration became 30-40% in a certain tray. According to a report, the peroxide decomposes instantly and generates much heat if the concentration becomes above 25%. According to another report, if the concentration becomes above 40%, a detonation occurs. The peroxide concentration of the explosion point at that time might be excessively high locally. Considering above, the cause was judged to be the locally concentrated peroxide and it induced a thermal explosion. |
| Countermeasures |
Measures to remove peroxide in methanol supplied to the rectification column should be taken, for instance, adding reducing agents such as sodium sulfite to the methanol.
Monitoring of the quantity of peroxide, etc. should be strengthened.
These two measures are incorporated into the operation manual, and the operators are reeducated. |
| Knowledge Comment |
1. At a design stage, the hazard of a process should be sufficiently investigated, and it is important to take suitable countermeasures.
2. As distillation is an extremely efficient separation process, minor components might be concentrated to be an unexpected high concentration. This phenomenon requires caution. |
| Background |
The trigger of the accident was a pH meter failure and operations continued without noticing the pH meter failure from underestimating of neutralization. However, it is required to think about other factors of accident causes.
At first, pH control in a neutralization process is very important in quality control, but sufficient facilities and management might have lacked. Next, production and decomposition of peroxide was not found at the R&D stage. However it might be difficult to find the phenomena at that time because the analysis method of the peroxide was established 2months after the accident.
The third point is how to operate the column on stopping. Distillation is an effective method for separation, and there have been many accidents caused due to condensation of impurities. Total reflux and boiled-up are more effective for condensation than usual distilling operation. In a stopping operation of the column, if the following operation was taken, the accident might not have occurred; all inventories of the column are run down to a tank and at start-up all inventories are charged in the column. |
| Incidental Discussion |
It was a newly developed manufacturing process at that time and it attracted the attention of the industry. The hazardous peroxide was just formed at the newly developed bleaching process, but the generation and influence of the peroxide were not understood. In the follow-up after starting the process, it was very difficult to find it out. The explosion might be avoidable where the pH control was done suitably and operations were executed correctly. |
| Reason for Adding to DB |
Explosion caused due to difficulty in finding out process hazards at an in-house development process |
| Scenario |
| Primary Scenario
|
Unknown Cause, Occurrence of Unknown Phenomenon, Not Trapped at Study and Development, Misjudgment, Misjudgment of Situation, Planning and Design, Poor Planning, Poor Poor Design, Bad Event, Chemical Phenomenon, Concentration, Secondary Damage, External Damage, Explosion/Fire, Bodily Harm, Death, 2 person died, Bodily Harm, Injury, 13 person injured, Loss to Organization, Economic Loss, Manetary Damage 1000 million yen
|
|
| Sources |
Fire and Disaster Management Agency, Fire caused by explosion of methanol rectifying tower on-stream. Accident cases of hazardous materials. 1991, pp.64-67
Tadao Yoshida, Masayoshi Nakamura, Kazutoshi Hasegawa, Methanol rectifying tower explosion with organic peroxide. Safety engineering, No.194, pp.370-378(1996)
Hazardous substance safety research association of Kawasaki City. Fire caused by explosion in methanol rectifying tower of synthetic detergent production plant. Useful cases of accidents at hazardous chemical facilities (with FTA) pp.8-10(1997).
|
| Number of Deaths |
2 |
| Number of Injuries |
13 |
| Physical Damage |
Debris from the explosion was mainly scattered within an 800 m radius of the west side of the plant, damaging 18 companies in the neighborhood. A methanol rectification tower and attached tank piping were heavily damaged, an attached pump was destroyed by fire, a building's walls and windows were damaged, and methanol was lost. |
| Financial Cost |
¥ 850 million. (Fire and Disaster Management Agency). |
| Economic Loss |
¥ 10 hundred million. |
| Multimedia Files |
Fig3.Chemical formula
|
|
Fig4.Chemical formula
|
|
Fig5.Chemical formula
|
| Field |
Chemicals and Plants
|
| Author |
ARAI, Mitsuru (Environmental Science Center, The University of Tokyo)
TAMURA, Masamitsu (Center for Risk Management and Safety Sciences, Yokohama National University)
|
|
