This post describes the last of the large volume chemicals I made when I worked at the Memphis plant in the 1970’s and early 1980’s. It is methyl methacrylate, which is used in many of the plastics that are known as acrylics. You may know them through their trade names like Plexiglas®, or Lucite®. You encounter them on every airplane flight you take, since they are used in the windows that let you see the clouds and the ground.
The process to make methyl methacrylate is complex, and large in scale. Our plant made several hundred million pounds per year. Some of the product was used on the plant in an acrylic sheet plant, that made both clear and colored, often marbled colored sheet. There are five steps to make methyl methacrylate. First, acetone (good old nail polish remover) is reacted with hydrogen cyanide (discussed in my first post on chemicals) to make something called acetone cyanohydrin, or ACN. As with anything involving cyanide, the material is toxic and great care was taken to prevent release of the chemical. The next step takes the ACN and mixes it with extra-strong sulfuric acid called oleum. Oleum is basically 100% sulfuric acid (one of the strongest and worst acids to deal with), with extra sulfur trioxide gas dissolved in the acid. When it hits anything containing water, it instantly reacts with it and sucks the water out of what it hits. This oleum is tweaked by adding tiny amounts of water to make the mix right at 100% acid when it hits the ACN.
The reaction process is very energetic, and produces an intermediate chemical called methacrylamide (I know, too many unpronounceable names). This intermediate chemical in a sulfuric acid solution was then reacted with methanol, and the resulting chemical was separated out and purified. The sulfuric acid solution contained a bit of organics, including some polymer. It was allowed to settle in a large tank so that the polymer could float up to the top and be removed in what we called skim tubs. The sulfuric acid tails were then fed into a sulfuric acid manufacturing process, where extra sulfur was added to make up for process losses, and new extra-strong sulfuric acid was stored and fed back into the reactors.
As you may have realized, these chemicals were all very nasty, and either toxic or corrosive or very hot, and I used to walk around miles of piping and vessels carrying these fluids under pressure. Only the product methyl methacrylate, was relatively non-toxic and non-corrosive, but it was at the end of a long process to make it.
In the few years I worked on this process, there were two main tasks I had. First, I was working with our staff of PhD chemists to improve the yield of the process. One very intriguing possibility was replacing the water that we used to mix with the sulfuric acid with methanol. Lab data showed a significant yield increase by introducing methanol in the first step. Since the reaction of sulfuric acid with methanol releases water, it solves the problem of controlling the acid strength when it is mixed with the ACN. The main difference between methanol and water was that it took a lot more methanol than water to provide an equivalent amount of water content. For every gallon of water, it took almost 1.9 gallons of methanol to substitute. But everything looked good in the lab, so we began work on a full-scale plant test. We went through an extensive process hazards review process to try to see if there were new hazards introduced, but could not come up with a reason to halt the test.
So I was the engineer in charge for the plant test when we got ready to swap out our water feed with a new methanol feed. The way we injected water into the sulfuric acid was through a mixer, where the acid was twisted through fixed barriers in the pipe to ensure complete mixing. We closed the valve for the water, opened it for the methanol, and watched to see what would happen. Almost instantly we became aware that despite all of our planning, something was going very, very wrong. The water injection line now holding methanol started to jerk around severely, and one thing you never want in a chemical plant is to have piping moving back and forth. I gave the order to turn the methanol off, and turn the water back on, and the piping stopped shaking. We probably were on methanol for no longer than 10 minutes before I halted the test.
What we had overlooked was that when we substituted water for methanol, we were adding a larger volume of a liquid that boiled at a much lower temperature. Methanol boils at about 149ºF vs. 212°F for water. It also takes a lot less energy to boil methanol. And when we started swapping out the water for methanol, some of the sulfuric acid would go partway up into the methanol pipe and induce boiling where we had never had boiling before. That was what caused the piping to jerk about. Fortunately I stopped the test before anything broke, but that was one of the scariest experiences I ever had in that plant. We never did go back to that test, since it would have taken a significant redesign to come up with a mixing system that could handle the differences between the two fluids.
The second project I had during this time was one I had inherited. I mentioned the skim tubs where polymer floated above the spent sulfuric acid as it cooled. That polymer had been skimmed off, packaged into metal drums and sent out as hazardous waste. Now there was an old incinerator down at the bottom of the plant, that someone had the bright idea to re-commission as a hazardous waste incinerator, depending upon its ability to meet hazardous waste disposal regulations.
One of the advantages of working for a world-wide company was that we had a wealth of technical expertise. There was a whole cadre of folks at the Engineering Services Division, or ESD, who had PhD credentials. They concocted the idea of putting the polymer into 30 gallon cardboard drums with plastic liners, and then burning them in the incinerator. But since this was an operation that needed to operate automatically without human intervention, they had created a Rube Goldberg contraption to make it work. They designed a conveyor system where drums would be placed on rollers. When the time came for a new drum to be inserted into the incinerator, alarm bells would go off, warning lights would flash, the knife valve they had installed on the top of the incinerator would open up, and the next drum in line would advance up the conveyor’s slope till it teetered at the end, and then would plummet headfirst through the top of the incinerator. Imagine an automated system to throw virgins into the maw of a volcano, and that’s what this thing looked like.
Well, I oversaw the construction work to install the conveyor and all of the equipment. We got ready to test the system, but there was one really little itsy-bitsy problem we encountered. See, during the time between when the scope was prepared for this incineration process and the design was installed, there had been another change made to the chemistry of the process, in order to improve yield. This chemistry change converted the polymer from being hard chunks that didn’t hold much acid, into a soupy mix that held a lot of the spent sulfuric acid. We had problems with drums leaking since the plastic liner was not intended to hold hot sulfuric acid, but worse than that, when the drums were consumed in the incinerator, a plume of sulfur dioxide came out of the stack and came down all over the place on the plant.
During the process of trying to get this incinerator to work, I had been transferred from Memphis to our Belle plant in West Virginia. The last thing I did at Memphis was to try to conduct a trial to see if this setup would meet the environmental requirements. We were successful in incinerating a liquid stream from our Lucite® sheet plant, but the attempts to incinerate the polymer drums was an abject failure.
Both of these efforts showed me how small and subtle things could cause a huge unforeseen problem. It was the effect of unintended consequences that got us in both cases. Once I went to Belle, I was working in a sister plant of the Memphis methyl methacrylate plant, only it was a plant that used water instead of methanol in order to create an organic acid. But the statistics I was exposed to in Memphis, proved crucial to me in the next phase of my career where I used statistical techniques to extend my working career well beyond many of my peers who weren’t as adept at math as I was.
Even A Broken Clock 