【Haichuan Chemical Accident Updates】British Cases of Chemical Accidents That Occurred Abroad Over the Years
Thread Content
The explosion and fire at the Bunsfield oil depot in the UK on December 11, 2005: This incident was the largest explosion and fire disaster to occur in Europe to date. It resulted in the destruction of over 20 large oil storage tanks, 43 people were injured, no deaths occurred, and the direct economic loss amounted to 250 million pounds. Course of the incident: At 19:00 on December 10, 2005, tank No. 912 in area A of the HOSL western section of the Bunsfield oil depot in the UK began to receive lead-free gasoline from the T/K pipeline, with a flow rate of 550 m3/h (which was within the allowable range). In the early hours of December 11th (at midnight), Tank No. 912 stopped receiving oil, and staff conducted an inspection of the tank; the inspection was completed around 1:30 a.m. on the 11th, at which point no abnormalities were detected. Starting at 3 a.m. on December 11, the level gauge of tank No. 912 ceased to change, at which point the tank continued to receive lead-free gasoline at a flow rate of 550 m3/h. Tank No. 912 was completely filled at 5:20 on December 11. Since the protection system of this tank failed to activate automatically to shut off the oil inlet valve when the tank level reached the set maximum level, the T/K pipeline continued to supply oil to the tank, causing oil to overflow from the top of the tank and a cloud of oil vapor to form rapidly around it. As a tanker carrying oil passed by the Bonsfield refinery, sparks from the vehicle’s exhaust ignited the vapor cloud formed by the spilled oil, resulting in an explosion and fire. At 6:01, the first explosion occurred, followed by more explosions. The explosion triggered a huge fire, with over 20 oil storage tanks engulfed in flames. The explosion injured more than 40 people, but there were no deaths, and commercial and residential property in the vicinity was severely damaged. At 6:10, firefighters arrived at the scene. The fire burned for 3 days, destroying most of the facilities on site and releasing large plumes of black smoke into the air (Figure 1). On the evening of December 13, all the fires were extinguished except for 2 oil storage tanks. The damaged oil tank is shown in Figure 2.
Analysis of the accident cause: 1) The automatic level gauge system (ATG) of tank No. 912 failed to function; when the tank was filled, the level gauge stopped at 2/3 of the tank’s capacity. The ATG alarm system did not activate, nor did the tank’s independent high-level switch automatically shut off the oil inlet valve, resulting in oil overflowing from the top of the tank. The spilled oil evaporated, forming a vapor cloud that, upon encountering an open flame, caused an explosion and fire. 2) Although the Bunsfield oil depot was equipped with three levels of protection, defects in the first level of protection caused the spilled oil to flow out in a waterfall-like manner, accelerating the formation of vapor clouds. The second and third levels of protection were primarily intended to protect the environment; however, the spilled oil led to the formation of large-scale fires, and the high temperatures damaged the fire barriers, causing them to collapse and break, which in turn affected the third level of protection as well. A large amount of oil and firefighting foam flowed out of the depot area. 3) The electronic monitors of some storage tanks and pipeline systems, as well as the related alarm devices, are not operating properly. 4) The combustible gas detection instruments near the storage tanks and pipeline systems are not sensitive. 5) Inspections for certain devices that are not operating properly are not carried out in a timely manner, and the response to such issues is slow, such as the automatic shut-off valves at the tank inlets and the control valves at the pipeline inlets. 6) The structural design of the storage tank (such as the design of the tank roof) is not entirely reasonable, which to some extent increases the likelihood of the formation of oil vapor clouds. 7) The location selection and protective measures for emergency facilities in the tank farm (such as fire pump rooms) are unreasonable. Lessons from the accident: Strengthen inspections, strictly control sources of fire, and improve the safety and reliability of equipment and instruments. One of the lessons learned from the Bunsfield oil depot accident, in addition to design and operational factors, is that the location of the depot posed safety risks. There is a high level of commercial activity and residential living around the Bonsfield oil depot, which poses significant risks to the safe operation of the depot. How to handle the economic development of the areas surrounding oil depots is a fundamental issue that all parties need to consider. During the previous approval process, the Health and Safety Executive (HSE) in the UK did not give sufficient consideration to the location of oil storage facilities and the surrounding development. It only focused on the risk of pool fires arising from oil leaks within fire dikes, without fully addressing the potential danger of explosions caused by oil vapor clouds ； Only the risk of oil vapor cloud explosions at the loading and unloading platforms for road tankers is addressed with corresponding preparations in the emergency plan. The Bonsfield oil depot accident reminds us of the need to make new revisions to the safety assessment criteria for oil depot tanks and site selection. Furthermore, prior to the Bunsfield oil depot accident, the UK Health and Safety Executive (HSE) focused most of its land planning efforts on a certain type of economic development that met the planning requirements (such as the Bunsfield oil depot), without giving sufficient consideration to other matters. Therefore, when planning the land use for oil depots in the future, all factors that may be affected by major accidents at such depots should be taken into account (including demographic factors).
On June 1, 1974, an explosion occurred at a caprolactam plant in the town of Flixborough, UK. At around 4 p.m. on that day, an explosion took place at the Nypro company, resulting in 28 deaths inside the factory, 36 injuries among the workers there, and 53 injuries among people outside the factory. The economic loss amounted to $254.4 million. 1. Accident overview: Nypro in the UK is a factory that primarily produces caprolactam and sulfuric acid-based fertilizers. Its cyclohexane processing unit features 6 series-connected oxidation reactors, which use cyclohexane as raw material to produce caprolactam. On the evening of March 27, 1974, a crack 150 cm long was discovered in the carbon steel casing of Reactor No. 5 in the oxidation reactor system, resulting in a leak of cyclohexane; the cause was stress corrosion induced by nitric acid-based substances. The duty officer reported to the supervisor, and with his approval, began lowering the pressure and temperature of the reaction system in preparation for shutting it down to check for leaks. Upon inspection, it was found that both the lining and the outer shell of Reactor No. 5 had suffered considerable damage, so it was decided to remove Reactor No. 5 for maintenance. The next morning, after discussion in the plant management meeting, the plant manager and relevant technical staff concluded that the shutdown for maintenance would take 3 to 6 months. Given the urgent demand for caprolactam in the UK at that time, it was not advisable to halt production and reduce output. Therefore, it was decided to move Reactor No. 5 and to connect a pipeline between Reactors No. 4 and No. 6, so as to maintain production using the five oxidization reactors temporarily. Immediately after removing Reactor No. 5, repair work was carried out. When installing the connection pipeline between Oxidation Reactor No. 4 and No. 6, the workers did not carry out any preliminary planning or design; no proper design drawings were created, nor were any necessary calculations of engineering stresses performed. Instead, a simple sketch of the repair work was drawn on the ground using chalk, with the pipes supported by scaffolding at the site. Due to the height difference between the two oxidation reaction tanks (approximately 35.5 cm), a three-curve bypass pipe is used as the connecting pipe between them. It turns out that the diameter of the expansion joint at the outlet of the oxidation reaction tank was 720 mm, but at that time the plant only had pipelines with a diameter of 510 mm. After a rough calculation, the construction workers concluded that a pipeline with a diameter of 510 mm could provide the flow rate required for the reaction. Assuming it to be a straight pipe, they believed it could withstand the pressures during operation; therefore, they decided to use a 510 mm pipeline in place of the 720 mm one. The three-bend bypass pipe is supported only by an eagle frame divided into four points, and its structure is not robust (see Figure 1). All repair work was completed at 2 a.m. on March 30.
Figure 1: 510mm connection pipe and eagle frame connecting Reactor Vessels No. 4 and No. 6
On the afternoon of April 1, a pressure testing for leak detection was carried out on site. The first test was conducted using nitrogen at a pressure of 0.39 MPa, and leaks were detected. The staff reduced the pressure in order to locate the source of the leaks, but were unable to identify it; only by increasing the pressure again could the leak location be found. After repairing the leak through welding, the component was put back in its original position. At 4 p.m. that day, a second test was conducted, using nitrogen at a pressure of 0.88 MPa for leak detection; no leaks were detected, so the pressure was reduced and the nitrogen was released to resume production. On May 29, another leak occurred in the reaction system; the leak was located at the lower level gauge, which reduced the pressure throughout the reaction system to 0.15 MPa and caused the temperature to drop, leading to a 2-day shutdown for partial repairs. At 4 a.m. on June 1, work resumed. Soon, another leak occurred in the cyclohexane circulation section, so heating was stopped and repairs were carried out again; operations resumed only at 5 o’clock. During the shift change at 7 a.m. on June 1, the on-duty supervisor failed to clearly explain the status of the leak repair to the personnel taking over the shift, so they did not pay sufficient attention to it either. On the afternoon of June 1, a leak of flammable gas began, but it went unnoticed. Around 4 p.m., a large amount of flammable gas was present in the air and spreading outward. Two minutes later, an ignition source may have caused a fire in Hydrogen Workshop 2, which subsequently led to an explosion. The explosion of the cyclohexane vapor cloud caused 2 alternative expansion joints to crack and buckle due to excessive external forces, while the elbow connecting Reactor No. 4 and Reactor No. 6 was distorted by the explosion“
2 Causes of the accident: 1) No detailed planning during the maintenance process. After it was discovered that Reactor No. 5 needed repair due to a rupture, the design of the pipelines connecting Reactors No. 4 and No. 6 was not carried out by experienced engineers; instead, the entire design was roughly drawn on the ground using chalk. The flange at the outlet of the oxidation reaction tank was supposed to be equipped with a pipeline of 720 mm in diameter, but one with a diameter of 510 mm was used instead; furthermore, no strength calculations were carried out for this bypass pipe, nor were any pressure resistance tests conducted. After the accident, the bypass pipeline bent into“
The chemical industry is a high-risk industry. When putting the process into operation or updating it, a Hazop must be conducted
2 Causes of the accident: 1) No detailed planning during the maintenance process. After it was discovered that Reactor No. 5 needed repair due to a rupture, the design of the pipelines connecting Reactors No. 4 and No. 6 was not carried out by experienced engineers; instead, the entire design was roughly drawn on the ground using chalk. The flange at the outlet of the oxidation reaction tank was supposed to be equipped with a pipeline of 720 mm in diameter, but one with a diameter of 510 mm was used instead; furthermore, no strength calculations were carried out for this bypass pipe, nor were any pressure resistance tests conducted. After the accident, the bypass pipeline bent into“