 When you go to work, you expect to be able to go home at the end of your shift. After all, most take up employment to work to live. However, at almost any workplace, tragedy can strike. Some industries tragedies can strike more frequently than others. I would say working in retail is probably more safe than say in a factory, give or take. Say a teddy bear shop is on one end of the death at work scale and in the army in a war zone is on the other end. Today's subject would probably be around three-fifths along the scale towards the war zone. Our subject today is a chemical factory. A chemical factory that would turn out to have some serious shortcomings in its safety, protections and protocols. It probably goes without saying my little scale, like all of my scales, is very subjective and essentially useless. But it kind of puts across my point for this video. Today's disaster would result in four workers not returning home from their shift. I should say that this video will be leaning on the CSB interim report and the final investigation report for its facts and figures. So as always, check them out for further information and reading. Insecticide Our story begins with this little chemical, methylmal. It's an insecticide and is rather good at killing stuff. It's even on the EPA's toxicity category one list, i.e. very toxic. It's sold under a number of different commercial names, but for this video's purpose it's sold under the brand name of Lanate, sold by the totally not controversial DuPont company. And one such place DuPont made this insecticide was at its La Porte site in Texas. The site dates back to the Second World War and over the years was lauded for its safety. It was one of the largest chemical plants in the area and several different products were made there, including another insecticide sold under the name Vidate. Both products made use of the chemical methylmocaptin as an intermediate raw material. It's a very deadly chemical, but although cutlery, it does have an odor similar to cabbage, thus it can give you some warning of when it's around. In 2011 DuPont installed a nitrogen oxides reduced scrubber incinerator at the La Porte site. This cost $20 million and was quite a large investment. It was aimed at reducing the insecticide's plant's pollution emissions. The system also was hoped to help increase production levels, but often with these things the project had some drawbacks. The new scrubber experienced high pressure events in the pipework for the waste methylmocaptin from its Lanate and Vidate processes. This was caused by the collection of condensate in the pipework, which created blockages. Now it was the system's pipework's design which caused the blockages, as the CSB would later say, that the pipework had low points which liquid could collect. In order to combat high pressure events staff were instructed to open manual drain valves and any liquid collected would drain out and trap gases would then be able to be vented to atmosphere via the scrubber stack. The building's control panel would alert operators to a leak of methylmocaptin, and although sensors were placed all around the site, alarms were only in one place. The process building was split in half and was named a wet end and dry end. Both parts had ventilation which pulled air out of its side via fans. However, both sides had been switched off due to defects earlier in 2014, thus eliminating any proper ventilation for the building. Over the years parts of the site would be sold off to third party companies, which reduced the amount of actual DuPont staff working at La Porte chemical plant. So this rather neatly leads us on to the disaster. Okay, so before I start the disaster outline, I'm going to show you this patented, well not really patented, plainly difficult disaster bingo card. Take a picture of it and cross it off as we get going along this disaster outline. It's Monday the 10th of November 2014, and a truck is pulling up on site to offload some raw product for the insecticide plant. The product is acetyl dehyde oxane, but for the ease of your ears and me pronouncing it dodgily, it can also be called AAO. Usually it is delivered as a load of 100% AAO and diluted down to 50% on site. However, on the 10th of November, the lorry had a shipment of pre diluted AAO down to 50%. The site's crew weren't trained for this, and during the offloading, the water dilution system was accidentally turned on. This diluted the AAO down to roughly 24%, making it out of spec for the insecticide production operations. On top of that, it also overfilled the AAO storage tank, as water had carried on being pumped after the delivery had been completed. Due to this, little materials faux pas, the lanate and vidate insecticide processes had to be shut down. Obviously this is not very good, but shut down would last two days, as management tried to figure out how to deal with the diluted AAO. But the delay would also bugger up another part of the production process. Well, with the system shut down, the methyl macatin in the reaction system used to produce lanate had formed a salt slurry. This stopped an initial restart after the reaction process. This scenario isn't actually particularly rare. However, Dupont never really officially devised an unblocking process. Instead, operators had figured out over the years that adding warm water to the reactor system pipe work usually cleared any slurries up. During the pumping of hot water, a valve connected to the feed line from the methyl macatin storage tanks should be closed. This was to stop any water running into the tanks, and thus was meant to isolate the feed line from the reaction system. This valve, for whatever reason, was inadvertently left open, and water flowed back towards the tanks. Interestingly, this scenario had not actually happened before, and the tanks were now overflowing back into the feed line, raising the tank level by 2%, a fact that was missed by the operators at the time. This water, now in the feed line, and was estimated to be around 2,000 pounds worth of liquid, began to mix with the methyl macatin, and in the cold November temperatures formed a solid hydrate that now created another blockage. So, another start-up of the reactors was attempted after they thought they had cleared the slurry, but again it failed due to no supply of methyl macatin. But operators couldn't figure out how. Well, we know it was because of the newly formed hydrate in the feed line, but the operators every time didn't know. DuPont management and operators would take until Friday to establish the course of the blockage. In order to return the hydrate back to liquid methyl macatin and water, management instructed operators to run hot water over the feed line's pipework underneath its insulation. This would warm up the pipe's contents, but at the same time, it was realised that the methyl macatin would likely expand as it was heated up. Thus, the pathway for over-pressure to go was needed. This brings back into our story the vent line to the scrubber incinerator, and its piping intended to remove excess or unwanted vapour from the system. The system had three valves that connected the feed line to the vent header. Pressure gauges were included on the three valves to monitor the progress of the unblocking activities. So, the hot water pumped over the pipework took place during Friday's day shift. I should say that during this, the valve for the feed line to the reaction system was closed, leaving only one potential route for liquid methyl macatin, which was out to the vent header. During the day, the hydrate was unblocked between the tank and the first vent valve header. By the time the night shift had come along, work to start and clearing up to the second valve was nearly ready. The day shift supervisor handed over with a verbal report on the day's clearing process. This was the first time that the night shift supervisor had been aware of the issues. So, the evening shift carried on, and by one in the morning, the team had cleared up to the second valve. They thought it would be a good idea to attempt a reaction system start-up. The valve between the feed line and the reactor was open, and the feed pump was started. But to their disappointment, start-up failed due to, once again, no flow of methyl macatin. Time for a break. The operators closed the feed line valve, but left the pump running. This was around 1.30 in the morning, on Saturday 15th November. The one water running over the pipe work was also left on, and the hydrate was slowly returning to liquid. This left the system in a dangerous state. With the pump running and the valve of the reactor shut, this only meant there was only one path for the liquid methyl macatin, and that was to the vent header. Remember, the vent header was only intended for vapours, not large amounts of liquid. At approximately 2.45 am, the methyl macatin storage tank level began to drop. This was a big hint that the blockage was now clear, and the deadly methyl macatin was now being pumped out of the tank. Some operators could smell the foul odour of the chemical, but it had become a bit of a regular experience, and was thus just ignored. Fluid was now pumping into the vent header, and the control room started receiving alarms of high pressure in the waste gas vent header piping to the NRS incinerator. As I mentioned earlier, condensation in the vent header pipe work was pretty common in this poorly routed pipe work, which collected in bends and low points in the system. The shift supervisor and an operator were sent to the third floor of the main building to carry out the regular operation of opening the drain valves to vent the scrubber system of any liquid. When they opened the valves instead of the regular brownish liquid, methyl macatin liquid poured out of the drain valve and quickly turned into vapor, filling the floor. Remember how the air ventilation system had not worked since October 2014? Well, the vapor was able to fill out the building unhindered. The operator, who had opened the drain valve, radioed a distress call. Sadly, though, they didn't express the nature and location of the emergency, thus more operators would attempt to offer assistance without any PPE. One of the assisting personnel went up to the fourth floor by mistake, found no one and made a PA announcement. He was then told to check floor three. He made his way down but passed out, hitting his head in the process. Another operator reached the second floor and upon feeling sick also retreated down the stairway, where he was met by one more operator who had come to help with air bottles. The operator with the air bottles made his way up to the third floor, where he would succumb to the gas. The one who had hit his head amazingly would gain consciousness 45 minutes after his fall and also managed to escape. In all, four had been overwhelmed by the fumes, which were three operators and a shift supervisor. Two of the operators, sadly were brothers and all of them would pass away. About an hour and a half after the opening of the drain valves, personnel with the correct PPE were able to enter the building and close the valves and recover the dead. Aftermath. So needless to say, this disaster was not good for DuPont and investigations would begin, which would be undertaken by both OSHA, the EPA and the CSB. So during the investigation, the surviving operators were interviewed and DuPont's processes were scrutinized and the cause was pretty clear. Just poor management of the site and an improper response to the chemical release. On top of that, a poorly designed venting and scrubbing system allowed for the foundations of the disaster to form. The CSB report would list its key findings. In its investigation of the November 15, 2014 DuPont-LaPorte incident, the CSB found that, one, DuPont did not effectively respond to a toxic chemical release. Two, DuPont's corporate process safety management system did not ensure that DuPont-LaPorte implemented and maintained an effective process safety management system. And, number three, DuPont-LaPorte did not assess its culture for process safety in the site's safety process surveys or any other formal assessment program, allowing serious process safety deficiencies to exist at the site. In total, approximately £24,000 of the very toxic methyl macaptain was released into the environment, which took the lives of four and injured many more. As reported by CBS, DuPont was ordered to pay $16 million in fines and Kenneth Sandel, the guy in charge of the unit, was sentenced to a year of probation, with the company itself being given a sentence of two years probation. Okay, I know it sounds a little bit odd of giving a company probation, but it just means that federal officials will have full access to all of the company's operating location. The site at LaPorte would actually get shut down by DuPont, and I suppose that is one way to fix any potential issues. This is a plain difficult production. All videos on the channel are creative commons actuation share light licence. Plain difficult videos are produced by me, John, in the currently wet and windy corner of southern London, UK. I have Instagram and a second YouTube channel to check them out for other bits and pieces I get up to. I'd also like to say thank you to my YouTube members and Patreon members for your financial support. And all that's left to say is thank you for watching and Mr Music, play us out please.