10 - Talking shop - Castets

 November 28, 2023 - Castets

It occurs to me that I've been fairly tight lipped about the equipment and project so here's a nutshell (maybe a coco de mer nutshell) summary:

Coco-fesses?  🤣  I suppose it's an apt name

Coconut meet cocobutt?

• The project here was to stick closely to what exists in Newark and build it in Castets to expand our global production capacity for this perfumery ingredient.  This molecule is projected to increase in sales so we're trying to stay ahead of demand.
• The new equipment consists of batch reactors, storage tanks, and distillation.  

• Batch reactors are basically big kitchen pots: you add stuff and stir while heating or cooling until it's cooked and seasoned to taste (maybe a slight oversimplification).  When you're done cooking, empty the pot out and you can start the next batch.

• When done cooking, you have your crude material (they call it "brut" here).  It's crude because it's not cognizant of social norms.  Jk, jk, it's crude because the stuff we're interested in is there but is mixed in with a bunch of stuff we don't want.  

• Example: laundry.  Your clothes go into the washer along with water and detergent, they all get mixed up so that the stank comes out of the clothes and we drain it with the water.  Ta-da!  Your clothes are clean and ready to wear... or are they?   
• To get the finished product, we have to separate the stuff we want from the stuff we don't.  Separations operations makes up a huge amount of the Chem Eng field since isolating your product is very important at the end of the day i.e. nobody wants a bottle of whole smashed oranges, they want the juice they paid for.  (Or out in the woods, nobody wants a jug of spoiled grain water; they want moonshine)
• Separations come in all different flavors based on physical properties, you've likely done several separations in your daily routine.  We use density, particle size, boiling point, solubility, etc. the list goes on but becomes more niche and tedious.  
• Density: oil and water have different densities so the oil floats on top.  Floating rubber duckies are way more fun than waterlogged ones.
• Particle size: Some things are bigger than the holes available, which is good otherwise we'd all be drinking roasted bean sludge instead of coffee.  
• Boiling point: for a certain pressure, everything boils at a certain temperature; mixtures boil as a compromise (and function) of what's inside.  Seawater evaporates to leave behind salt, but crank up the heat (a lot) and you could theoretically boil the salt.  

Hooray for café and not bean sludge!  Also, the water dispenser can do carbonated!

So here's where things stand:

• The equipment to convert brut to finished product was delivered as a pre-assembled module (roughly 12ft x 12ft x 40ft) and dropped into place.  Unfortunately, there's been mechanical and operability issues discovered - instruments facing the wrong way, air leaking through loose connections, etc.
• We've yet to test the system by trying to pump water around or verify the program is set up correctly.  
• The rest of the plant has been facing issues but they've been making slow progress in testing one reactor with water - ~2/3 through all the heating/cooling and moving the water around between equipment.  

As the process engineer (I think we used to be called industrial engineers before and maybe still in non-chemical fields), my wheelhouse is in understanding how the parts work individually and together to make the "process" work.  In Newark, I would focus on how long each step takes to identify bottlenecks and work with teams to increase production capacity or reduce waste.  Here, I'm the maritime pilot coming aboard to navigate the tricky waters of the first few batches; or maybe the guest conductor for a band's first few performances.  

Based on the white-knuckle ride that was this morning's meeting (I was able to follow the conversation and have a strong idea of what was being said, but couldn't divert my attention for a second to make notes), my focus in the next week or so will be:

1. Assisting in assigning limits (les seuilles) for the various instruments and program steps (what's too high and what's too low and how should we react).  
1. There's a lot of these little buggers, some more important than others.  We use alarm limits for safety (don't let the temperature get too high!) but setting them is a balancing game: too much margin and you're no longer protecting stuff, too little margin and we might be stopping every single batch.   
2. Providing feedback for operating procedures and protocols.
1. The production manager asked if I could provide a sanity check on a few Job Safety Analyses (JSA), which is a document that lays out how we safely perform the manual tasks of the process.  
2. It states what the task is (charge drums of stuff) and then breaks it down into individual steps/instructions; each step has the possible risks identified (risk of cutting or slipping or chemical exposure) and the safety barriers in place (wearing gloves / ventilation hood / faceshield/ etc.).  Of course, my focus was just on whether the steps of the task made sense or were clear, which for the most part I thought they did/were.
3. Testing/demonstrating the safety barriers we said we would have in the safety analysis.
1. In the safety analysis, we try to question every scenario of something going wrong and then develop barriers that either reduce the severity of the consequence or make it less likely to occur.  
2. For example: if we don't charge enough of chemical A, then when chemical B is charged we might form a solution that will readily explode.  Explosions are bad, so that's a severe consequence.  To reduce the liklihood, we can verify that we charge enough of A by comparing changes in fill levels of the associated tanks, we can check the total mass the flowmeter calculated, and we can also verify that the charge wasn't done abnormally fast.
1. Now I just have to prove that all of those barriers are actually being used in our program.