| Case Name |
Leakage of raw material oil from a hydrocarbon resin reactor at the beginning of a reaction |
| Pictograph |
|
| Date |
July 6, 1989 |
| Place |
Kawasaki, Kanagawa, Japan |
| Location |
Chemical factory |
| Overview |
By anonymous information from an employee of the factory in Kawasaki, the fireman did an on-the-spot inspection. The result showed that there was a leak from a polymerization reactor a few days before the information. The reactor was a vertical vessel with an agitator and an external circulation cooler, and the reaction was carried out in batches. Stock oil was charged, heated to the initially set temperature, and a catalyst was injected. The quantity of the catalyst was increased because there was not a temperature rise which indicates a start of a reaction. The operator left the control room to do other work without observing the results of the increased catalyst. The reaction temperature had exceeded 100 °C when he returned after two minutes. The external cooler was hurriedly started. The reaction was completed safely, and the shift changed. Another operator of the next shift found the stock oil leaking from insulation on the reactor during a routine check, the reaction of the next batch was stopped, and the plant was shut down and checked.
The causes of the leakage were as follows: There was corrosion in the weld line of the SUS304 lining, and the reactor was kept in operation with a temporary repair. There were holes at the reactor base material caused due to a decrease in thickness due to external corrosion, and an improper operating method was used to start the batch reaction.
The factory changed the material of the reactor to SUS304 solid material. |
| Incident |
A leak occurred at a hydrocarbon resin plant that produced HC resin of C9 hydrocarbons from by-products of an ethylene plant. Stock oil oozed out through pin-holes of the reactor and leaked out through hot insulation material. |
| Processing |
Manufacture |
| Individual Process |
Reaction |
| Chemical Reaction |
Polymerization |
| Substance |
Pyrolysis gasoline residue (kerosene fractions) |
| Type of Accident |
Leakage |
| Sequence |
At about 03:30 on July 6th, 1989, 10 kL of stock oil was fed to the first polymerization reactor (A series).
At 04:10, circulation of an external heat exchanger and a reactor for heating was started and heating of internal fluid of the reactor was also started by supplying steam to the heat exchanger.
At 05:00, as the temperature of the reactor had reached 45 °C, circulation for heating was stopped. Next, 20 L of a polymerization catalyst was charged. In addition, 5 L of a catalyst was charged six times, a total of 30 L was charged because the temperature rise was small.
At 05:10, after adding catalysts, a person in charge left the site to perform a filtration process for the B series.
At 05:12, the person, who had returned to the control room, noticed that the temperature in the reactor had risen over 100 °C. Immediately, he started circulation for cooling between the reactor and the other external heat exchanger in which the brine was used.
At 05:30, as the temperature in the reactor had dropped to the normal 60 °C, the temperature in the reactor was kept and polymerization was continued.
At 08:00, there was a shift change.
At 09:00, the reactant in the reactor was transferred to a neutralization vessel because polymerization had been completed.
At 12:00, the next batch started.
At 12:40, an operator found that oil had leaked from the surface insulation of the reactor. |
| Cause |
As described later, the reactor was made of carbon steel and lined inside with SUS304 stainless steel. The thickness of the carbon steel had already decreased due to outside surface corrosion, and there were already pinholes on the weld line. The pinholes were temporarily repaired with epoxy resin. Stock oil etc. had entered into the gap between the main body and linings. The temperature rapidly rose and reached over 100 °C, because a large amount of catalysts were injected into the reactor in a short period on starting up polymerization on the day. Therefore, stock oil from pinholes at the lining weld had entered between the main body and the lining, evaporated, and the lining material expanded due to internal pressure. Therefore, the weld part of the main body which had become thin due to corrosion broke. |
| Response |
The plant was shut down, and liquid in the reactor was transferred to the neutralization vessel at the next stage. The leaked oil was recovered. The location of the leak was confirmed by a water tightness test, and a temporary repair was carried out with epoxy resin. |
| Countermeasures |
The material of the first reactor was changed to SUS304 stainless steel instead of lining structure, as well as the similar second reactor, The revision of the operation manual and retraining of operator were carried out. |
| Knowledge Comment |
1. The failure might have occurred due to long-term use without repairing the defect.
2. In a polymerization reaction in the batch, the reaction shows the maximum rate at the maximum monomer concentration when the reaction just started. In other words, it is in a state of transition. Generally, full attention needs to be paid at the transition state. It is necessary to make careful judgments and actions at the transition state.
Transition: A condition in which the situation of the plant changes greatly. For example, changes of grades and raw materials, startup of a batch reaction, etc.
3. Generally, the probability of an error occurring increases when one person carries out two or more works at the same time. The primary work for reaction control was delayed because the operator carried out field work which was different from the original work. |
| Background |
Although this accident occurred in July, 1989, there was already allover external corrosion of the reactor main body in 1983, and the wall thickness was 3.6 mm, as against the original one was 6 mm. In addition, 20 pinholes (0.5 - 3 mm in representative diameter) that penetrated the lining in the weld line were found in 1986. The repair to the pinholes was only temporary measures, such as filling epoxy resin without welding, because raw material oil had already entered the gap between the main body and linings. Therefore, the main cause was leaving the reactor in an imperfect condition; pinholes in linings and reduced body thickness.
The next important problem was that an operator increased the catalyst supply by 2.5 times as much as the first scheduled quantity when he had to carefully start the batch reaction. This reaction is exothermic, and most of the feed liquid had a polymerization capability. The reaction was controlled only by adjusting the quantity of catalysts and the reaction temperature. Naturally, the reaction quantity and the reaction rate are also large at the early stage of the reaction.
It is predictable that an abnormality would arise if the catalyst quantity is increased at that time. Ignoring of the operation manual and inadequate education about the reaction are second factors.
The third factor is that the operator stopped monitoring without confirming the results of increased catalysts to do something else.
At the most delicate moment on starting a batch reaction, he stopped monitoring, which is the most important task at that time, because
he had to immediately start the cooling operation by circulation when the reactor temperature rose.
If he had decided his actions by himself, it was a deviation from operation duties. If the work was regulated by an operation manual, inadequate supervision of managers was the third factor. |
| Sequel |
The A series was remodeled to be a continuous reaction plant like the B series. There would be good results in terms of safety because irregular work has decreased. |
| Reason for Adding to DB |
Example of aggravation of problem due to temporary repair performed only for a reactor |
| Scenario |
| Primary Scenario
|
Poor Value Perception, Poor Safety Awareness, Inadequate Risk Recognition, Organizational Problems, Poor Management, Slackness of Management, Carelessness, Insufficient Understanding, Insufficient Recognition of Risk, Polymerization Reaction, Malicious Act, Rule Violation, Safety Rule Violation, Bad Event, Chemical Phenomenon, Abnormal Reaction, Secondary Damage, External Damage, Leakage
|
|
| Sources |
Kawasaki City, Fire fighting station. Prevention division. Peace section. N resin Co., Ltd. Outline of dangerous material leak accident at hydrocarbon resin manufacturing plant. Material of the Kawasaki City Complex safety countermeasure committee
|
| Financial Cost |
247,000 yen (Material of the Kawasaki City Complex safety countermeasure committee) |
| Field |
Chemicals and Plants
|
| Author |
KOBAYASHI, Mitsuo (Office K)
TAMURA, Masamitsu (Center for Risk Management and Safety Sciences, Yokohama National University)
|
|
