FAILURE IS ALWAYS AN OPTION
East Palestine and the case for resilient design in the chemical industry
“Failure is not an option.” Ed Harris, playing NASA lead flight director Gene Kranz in the 1995 movie Apollo 13, thus inspired his team to bring the astronauts home after a seemingly minor flaw in an oxygen tank mushroomed to a near fatal disaster. Referring to the plan to restart the capsule using only reentry batteries, members of the team noted in surprise, “Never been tried before. Hell, we’ve never even simulated it before.”
This is the nature of many catastrophic failures: a series of seemingly minor, improbable events leads to havoc that catches everyone, even the experts, by surprise.
Last month, western Pennsylvania had a front-row seat for yet another example of this kind of cascading, systemic failure: a faulty axle led to a train derailment, then a tanker leak, and finally a “controlled burn” of dangerous chemicals that may require years, or decades, to remediate. While the proximate causes of this nightmare are related to the railroad industry in particular, one can trace the root causes to a hundred-year legacy manufacturing, shipping and using of dangerous chemicals in inherently unsafe ways.
Redesigning our chemical supply chains with resilience as a top priority would greatly aid both people and the planet. Resilience is the ability for a system to “bounce back” from disturbance — moving from fail-safe to “safefail.”
In traditional design, engineers try to guard against failure by hardening systems to resist disturbances. Despite this, systemic failures continue to occur, with massive environmental, medical and economic consequences:
A series of storms just happens to impact certain airports one after the other, and Southwest Airlines — a typically reliable carrier — nearly melts down over the Christmas travel season.
A virus emerges from China, and normally efficient supply chains collapse as countries shut their borders to try to prevent further spread. These very countries, in turn, cannot get the personal protective equipment they need to fight viral spread, exacerbating the crisis.
In 2010, four containment systems at the Macondo rig — Deepwater Horizon — improbably failed simultaneously, leading to unimpeded flow of crude oil into the Gulf of Mexico for nearly 3 months.
In 2014, 4-methylcyclohexane methanol, which is used to clean coal, leaked from a tank in West Virginia, overwhelmed the attached “leak containment system,” then flowed into the adjacent Elk River — directly upstream from a water treatment facility. Drinking water for 300,000 people was harmed.
While in hindsight the design flaws in all these systems were obvious, as they failed spectacularly the people responsible for them responded with surprise.
In resilient design, we engineer systems with the assumption that failures will occur — but that through effective design, the systems can avoid
catastrophe and recover quickly. For instance, increasing the height of levees and seawalls is the usual approach to flood prevention. Resilient systems, on the other hand, assume that nature can and will overcome any man-made flood barrier, and so focuses on providing paths for floodwater to go — besides through living rooms. The Netherlands, long a master at the art of flood prevention via dikes, is now taking the lead in resilient flood management of the Waal River. In the U.S., municipalities are finally looking to buy people out of floodplains, rather than trying to help them rebuild time and again.
While the response to the East Palestine incident has focused, not wrongly, on the culpability of Norfolk Southern, the derailment also shows why resilient design is desperately needed in the chemical industry. The derailed train contained five railcars, or over 100,000 gallons, of liquid vinyl chloride. This is particularly dangerous because vinyl chloride is acutely toxic, heavier than air and must be stored under immense pressure to transport as a liquid. The fear that one or more of the tankers would fail catastrophically, resulting in a boiling liquid expanding vapor explosion (BLEVE) that could propel shrapnel and chemicals at least a mile away, led to the decision to try the oxymoronic “controlledburn” of the material.
Nearly all vinyl chloride is used to generate the common plastic polyvinyl chloride (PVC). In a more resilient system, vinyl chloride would only be produced at sites that also generate PVC, meaning that vinyl chloride would simply not need to be transported. A more advanced resilient solution would be devising a way to manufacture PVC that uses safer intermediates than vinyl chloride. As T.A. Kletz once said in “Chemistry and Industry” (1978), “What you don’t have, can’t leak.”
Resilient design in the chemical industry could (and should) extend beyond manufacturing to the molecular structure of the chemicals themselves. Given that effectively all vinyl chloride goes to produce PVC, perhaps we could design an alternative to PVC itself. PVC is environmentally unfriendly in several respects beyond the use of vinyl chloride, such as its use of energy-intensive and dangerous chlorine gas and its inability to be recycled: Low thermal stability means that less than 1% of post-consumer PVC is recycled.
A lack of resilience presents the chemical industry with almost intractable environmental problems that quickly become financial millstones. Most chemicals are intended for a single use, where the end-of-life destination is meant to be an incinerator or a landfill. Despite efforts at containment, a wide variety of chemicals “leak” into the environment — a systems failure that can lead to human and ecological health problems that are extraordinarily difficult to correct. These chemicals often become the focus of news stories and lawsuits: perfluorinated compounds (known as PFAs or “forever chemicals”), brominated flame retardants, phthalates, Bisphenol A and commoditythermoplastics.
If commercial molecules were created with resiliency in mind, compounds might even be designed with a built-in “self-destruct” mechanism so they degrade into benign fragments once they leave their intended place of application. Design of chemical products that contain safer components, are less complex and are more inherently recyclable would also aid resilience.
One factor blocking a more resilient chemical industry is its own history. Most of the chemicals used in commerce today were not designed, but rather discovered and then adapted to a wide variety of applications to increase demand and profitability. The problematic chemical classes noted above, for example, were discovered and adapted between the 1930s and the 1970s, and recent innovation in the industry has been painfully slow.
Resilient design would benefit the chemical industry by decreasing legal and financial liability, and it would benefit people and the planet by increasing safety. But it will require a level of entrepreneurial drive that one finds in information technology or pharmaceutical development or energy storage — and that is lacking in the stodgier world of commercial chemicals.
This could be greatly aided by a move within academia to recognize resilient design at the molecular level as a key part of overall design, and hence to prepare the next generation of scientists and engineers to adjust the priorities of the industry. Just as climate change is a crisis we must address now to save the plant for our grandchildren and great-grandchildren, so too must we adopt resilient design in every part ofthe supply chain.