[Music] Narrator: November 15th, 2014. The DuPont facility
in La Porte, Texas. Four workers were killed and three were injured
during a large release of highly toxic methyl mercaptan, within the plant's
insecticide processing unit. The sequence of events that led to the
incident at DuPont began on Monday, November 10th, when the plant's Lannate Unit was shut down,
due to a problem with a reactor. Within that unit, methyl mercaptan reacts with
other chemicals to create the insecticide Lannate. By noon on Wednesday, November 12th,
operators attempted to restart the unit. However, the startup was unsuccessful, because
piping within the reaction system was blocked. This blockage was common, following a shutdown,
and was caused by a slurry that formed in the pipes from a reaction
between methyl mercaptan and other chemicals. To clear this slurry, operators
flushed hot water through the blocked piping. By noon on Thursday, November 13th, operators
determined that the initial blockage was cleared. But during the operation to remove the blockage,
a valve was inadvertently left open, that should have prevented hot water from flowing into
other piping, known as the methyl mercaptan feedline. The feedline connected the reactor system
to a methyl mercaptan storage tank. Operators later discovered that approximately 2,000
pounds of water had escaped through the open valve, into the feedline
and back into the tank. And the water, mixed with
methyl mercaptan in the feedline, creating a new blockage
that would cause more problems. Since temperatures in the Houston area
had been consistently cold that week, the water and methyl mercaptan mixture formed a solid
material called a hydrate, which blocked the piping. By Friday, November 14th, DuPont personnel were
aware of the hydrate and a group of Operations management and technical personnel developed
a strategy to address the problem. They decided to run hot water directly on the blocked
piping, by placing hoses under the pipes' insulation, heating the methyl mercaptan feedline
to above 52 degrees Fahrenheit, enough to transform the solid hydrate
back to liquid methyl mercaptan and water. Along the methyl mercaptan feedline, there were
three locations where it was connected by valves to a vent header or piping intended to
remove excess or unwanted vapor from the process. DuPont personnel used
pressure gauges at those three valves to determine where the blockage was and
what progress they were making to clear it. They also knew that when heated, methyl mercaptan
can expand and would need a safe place to vent, to avoid overpressure
of the feedline. Because the unit was shut down, an additional valve
between the feedline and the reactor system was closed, preventing methyl mercaptan from entering
the reactors while they were not running. As a result, in this configuration, the
only place for liquid methyl mercaptan to potentially flow would be into the
vent header, where it was never meant to be. However, DuPont personnel did not
consider this hazard when forming their strategy. First, operators cleared the blockage from a section of
piping, leading from the methyl mercaptan storage tank to a pump that forces methyl mercaptan
into the reactor system. Then operators began working to clear a
section between the pump and the closest valve. They cracked open the valve and began to
heat the piping, to get rid of the hydrate. This caused some methyl mercaptan liquid to vaporize
and flow through the valve into the vent header. When this occurred, pressure
increased on the nearby gauge. Operators continued this process until the
pressure at the first valve stopped increasing, at which point they determined that the
blockage in that section of piping was cleared. Hoses were then moved to the
next section of the feedline. But as it was nearing the end of the Friday day shift,
the hoses were not turned on at that time. At around 6:00 p.m.,
the night shift came into work. The day shift supervisor verbally briefed the night
supervisor about the strategy developed earlier in the day. The operators turned on the hot water hoses
and continued efforts to clear the blockage. By this time,
the second valve was fully open. By approximately 1:30 a.m., the operators
believed they had succeeded in clearing the piping and attempted to
start up the unit. With the methyl mercaptan pump on, they opened the
valve between the feedline and the reactor system. However, they found methyl mercaptan
was still not flowing to the reactors. The blockage remained and they
once again closed that valve. After the failed startup attempt, the Operations
crew took a break and went to the control room. But the methyl mercaptan pump
was left running. The hot water hoses were
still heating the piping. And the second valve to the
vent header remained open. Unknown to the operators,
at approximately 2:45 a.m., the level in the methyl
mercaptan storage tank began to drop. The CSB concluded that at that time the
hot water removed the remaining hydrate and liquid methyl mercaptan
began to flow through the feedline. The methyl mercaptan fed to the system by
the pump followed the path of least resistance, through the open second valve,
into the vent header. The vent system quickly filled with liquid methyl
mercaptan, where DuPont never expected it to be. The vent system connects the methyl
mercaptan storage tank with process equipment, inside the Lannate
manufacturing building. As liquid flowed into the vent header,
pressure began to build within that system. However, high pressure events
within the vent system were not unusual, due to flawed equipment design that allowed
liquid to accumulate at low points in the system. These events occurred so frequently that DuPont
instructed operators to drain liquid from the vent system daily. After the control room operators separately told
two coworkers about the pressure problem, they each went to the third floor
of the Lannate building. The required response was to
drain the vent system of liquid. Two valves were opened and the unanticipated
liquid methyl mercaptan drained into the building, where it readily vaporized,
filling the room with the highly toxic gas. Although one of the operators was able to make a
distress call, both workers died, unable to escape the building. Four additional operators responded to the
distress call and entered the manufacturing building. Two of them were brothers; they died
together on the third floor of the Lannate building. The two other
responding operators survived. In total, four workers were
killed during the release. The release continued for
another hour and a half, before emergency responders with proper protective
gear were able to enter the building and close the valves. DuPont estimated that approximately 24,000
pounds of toxic methyl mercaptan was released.
If you found this interesting, the USCSB's channel is full of videos like this. I think all or most of their major investigations get a video.
I'm a former engineer at DuPont but not at this plant. I was involved in the fatality investigation at the Tonawanda NY. The CSB also has an excellent video on this incident as well. http://www.csb.gov/csb-releases-new-safety-video-on-fatal-hot-work-explosion-at-dupont-facility-in-buffalo-hot-work-hidden-hazards-shows-danger-of-inadequate-gas-monitoring-safety-video-follows-release-of-the-csb-s-investigative-report/
I can tell you it is the absolute worst being involved in a fatality at a site. DuPont takes safety very seriously and it was really the managers that failed here with negligence. We can all learn from these videos in our lives and careers.
I love these videos. The Texas City Disaster is really long but it's totally worth the watch.
I can't believe that they kept saying the never expected their chemicals to be in the vent lines? Like ever? How does one make a silly assumption like that especially when you have a known blockage and are increasing the pressure in the line with a pump and heat
These videos are spectacular, even down the the music! Bravo!
IANAE and I was not previously aware of this incident. I'm sure that after this investigation was finished, and the cause was determined, that DuPont had certain fixes to process/procedures and/or factory layout/equipment mandated to ensure that this mistake doesn't happen again. Would anyone be able to discuss those?
Everything that could go wrong, did go wrong. What a nasty way to die. And, my lord, that must have stunk.
Very heartbreaking, especially when the bothers died together.
Is there an easy way to make animations like seen in this video?Especially if you've already got a navisworks model of the unit? It seems cool to make some small ones for training and hazard studies. I don't have justification to spend lots of time or money on it.