Economics: Strive To Suppress Short-Sightedness

Nov. 4, 2021
Ignoring the humdrum to chase after the next shiny thing can stifle success

A steam trap jets out periodically, as it has for perhaps a decade, slowly eroding an I-beam that supports pipe from our raw materials tanks. In the last five years, this problem has been written up four times that I know of. When the support finally collapses, it will shut down plant operations for at least two days, likely costing about $400,000. Changing the direction of the steam jet and repairs, as a rough estimate, only would cost $6,500. Yet, the work hasn’t gotten done and won’t be because it competes with hundreds of other repairs and projects deemed more critical to the plant.

Many companies prefer to spend money on expansions and equipment improvements rather than on such ho-hum repairs. Expansions and improvements look good on resumés — nobody’s going to promote you because you kept an old plant running!

Humans don’t see cause and effect well — especially if the effect is years in the future, as a column in Psychology Today emphasizes. Reflecting that, our entire economic evaluation process is geared to short-term tangible results. We prefer to invest to get smaller rewards sooner rather than larger ones later. The last capital projects meeting you were in undoubtedly underscores that point! When I worked in oil reservoir analysis, we could show an 85% recovery by implementing a project to maintain the pressure in the sands. However, economic theory favors accepting a 15% oil recovery with a huge cashflow for seven years, while leaving the rest of the oil behind, instead of a smaller cashflow for 30 years. Economics doesn’t take into account long-term intangibles like a vanishing resource or greater difficulty doing secondary recovery. This thinking also applies to acute versus chronic toxin exposure, customer product satisfaction, water resources in arid climates where aquifers are drained to irrigate crops, corrosion’s impact on reliability, and dozens of other situations.

As instruments have become less expensive, they increasingly are being deployed to monitor processes to reveal spikes in temperature or pressure and variations from norms. (They never will overcome the inherent problems with variabilities in ingredients, equipment and other conditions that occur in batch processes.) Moreover, the growing sophistication of devices gives us capabilities to measure in ways we could only dream about 50 years ago. Of course, some limitations still require human intervention. The classic one is pH control; another is automation of complex laboratory analysis, e.g., titration or liquid/liquid chromatography.


Meanwhile, artificial intelligence (AI) has helped make the intangibles more real. In addition, AI allows us to create remarkable models of chemicals and thermodynamics; we can predict what will happen in ways not imaginable in 1970. We can design equipment such as distillation columns from models where previously we estimated parameters and over-designed the tower. We can’t improve the human thought process, though.

As a propellent-development scientist in the U.S. Air Force in 1980, one of my projects was working with a company creating software to use quantum chemistry to identify useful chemical compounds for rocket research. That firm hoped someday to determine whole chemical processes by quickly establishing the best routes to a particular compound. The dream was to eliminate years of bench-scale development and the expenditure of millions of dollars. A side benefit would be saving lives by reducing laboratory accidents; six people lost their lives in propellent development during my 3½ years in research.

Recently, Pfizer, and other pharmaceutical companies used CRISPR to pinpoint weak points in the SARS-CoV-2 virus. Further, they developed software to identify pathways for making the vaccines. This type of software someday might replace chemical engineers by compiling decades of engineering and scientific knowledge, perhaps learning from and extending that knowledge. Still, decision-making often must remain with humans. As arbitrators, we only are limited by our incapacity for imagination.

AI can’t help much with our desire for short-term gains nor our biases. I suppose in a future generation, a computer calmly will chastise a manager for pushing a project just to spend money so that next year’s budget isn’t reduced. I likely won’t live long enough for that but enjoy daydreaming about it.

All we can do is aim to develop economics to account for the less tangible as well as better contend with our cognitive biases, as highlighted in a column in BetterUp.