Science, technology, engineering and math (STEM) skills underpin the chemical industry. We believe fundamentals beat guessing 100% of the time. Sure, examples that defy initial analysis arise — but we eventually discover a new or misused property to include in future applications. However, until we pinpoint that parameter, part of process design often relies on art — turning STEM into STEAM.
Finding all the factors that predict the desired outcome is time consuming, especially when working with solids. I’m often reminded that it’s possible to spend countless hours researching various conditions without ever uncovering the one scenario that causes the specific issue at a plant. Art is selecting different screen sizes for various products that are physically the same size and shape. Art is why we install a device whose operation we can’t explain scientifically but which has performed successfully in the past. The inventor may not even know how it works. Whether it’s advisable to wait until its technical basis is proven depends on how much time and money are involved.
The practice of medicine is called a “practice” for good reason; so is a portion of solids process engineering. Crystallization primarily is science but do we truly understand nucleation and growth? We recognize that our crystallizer needs to be defrosted. Do we know how long or how to prevent flash-nucleation? Because time is money, we take shortcuts that speed up the process — such as rapping on the crystallizer head at just the right moment during cooling. This is something an operator has discovered over the years and is an art.
Scale formation is a form of undesirable crystallization/precipitation. Magnets have been used to minimize this problem. So, I asked: “What makes them work?” We commissioned a research project to study this type of device and attempted to understand how a magnet placed upstream from where the scale forms could prevent the problem. After several years of study, the researchers found no supporting evidence other than that a magnet seemed to work in several industrial settings but never in the lab. I’m convinced some science can explain this behavior. However, it could be the artist/scientist at work.
Cyclone design is an art. Each manufacturer has its own design, which has proven to work in most cases. However, each design looks different: some are short and wide while others are narrow and tall; some use a volute inlet and others have secondary bottom receivers. Hundreds of articles delve into the advantages of these nuances. A research consortium of industrial leaders developed a universal design program but found that it didn’t predict specific particle collection very well. However, cyclones of similar geometry performed in a very predictable manner, which probably is why every manufacturer uses a different shape in its design. Don’t get me wrong — designers still look at inlet velocity, space limitations, abrasion issues and a lot of technical aspects of their product to ensure they propose a robust design. When I was in the equipment manufacturing business, my boss always reminded me that we don’t buy a bad job, which means keeping any artistic features the same.
While I can offer an explanation for the drop-off in drying rate due to diffusion or capillary action, I end up with a “fudge factor” in the equations for the drying kinetics. When examining a new material, I often feel it between my fingers and relate that back to some other solid to select the best dryer configuration and fudge factor. It’s an artistic start in selection of a technology. While I’d like to do all the drying curves, the cost can outweigh the benefit; such art can shorten those studies.
You’ve probably had similar experiences where some innovation lacked a sound scientific explanation. We’re not taking anything away from STEM but rather benefiting from an additional view. As I look around to other engineers and academics, I see a movement that is picking up steam.
TOM BLACKWOOD is a Chemical Processing Contributing Editor. You can email him at TBlackwood@putman.net