The Process Industry’s Response to Natural Disasters
To be sure: there will be LOTS of second-guessing, “armchair quarterbacking” (or the rugby equivalent – if there is one!) and “Chicken Little Prophets” decrying the risk our communities are exposed to by the Process Industries (PI) because of the devastating effects of Hurricane Harvey. So many of the articles published in recent days are written by people with limited knowledge and understanding of PI, chemical processes, the controls and safeguards used to operate them. They understand less neither the risk assessment procedures, practices nor regulatory requirements associated with PI operations. Most of them do not actually live anywhere near the communities or work within the industries about which they write. Furthermore, the headlines for these articles sensationalize the events with misleading keywords and serve only to incite irrational alarm and fear with the unsuspecting public at large.
But let’s be clear about how process safety professionals assess the risks associated with these chemical processes: it’s based on probabilities; a reasonable and fundamentally sound probability of something to happen or fail. Safety Integrity Level (SIL) calculations, which are used to determine the redundancy level of process safeguards, are based on a “failure rate” or “probability of failure”. So, if an event or failure has NEVER happened before, how, then, can a probability for the same even be determined? How often has a region – ANY region – received over a year’s worth of rain in only TWO (2) DAYS? The last time such an event occurred witnesses reported seeing pairs of animals getting onto a very large boat! Even seasoned meteorology professionals were confounded by both the predictions and the reality of Hurricane Harvey. It was an epic and unprecedented event by all measures!
By analogy, the analysis of aviation accidents often reveals that there is a “cascade effect” of system failures, external events and decisions or actions that ultimately result in the accident. If one element of this cascade is removed then the accident never occurs. The events in SE Texas had no such causal trail: it was, effectively, an “act of God” which was not, and never could be, anticipated. Double Jeopardy, that is, a failure of two independent safeguards, is discounted from most Process Hazard Assessments. Harvey caused, not only double, triple systems failures at the Arkema site that no one could reasonably predict, much less, assign a non-zero probability.
In many ways, the process hazard assessment process is similar to what we, as individuals, do when assessing where to live: it’s a question of balancing risk versus cost; that is, how much risk we are willing to accept versus the cost to mitigate the risk. For example, we often choose where to live based on schools, distance from where we work as well as other subjective judgments or assessments of the community. The Meyerland area of Houston, for example, a fairly affluent area, was especially hard hit when Braes Bayou overflowed its banks. Some homeowners had incurred the expense (to say nothing of the inconvenience) of having their homes raised above grade, based on previous floods; others, however, did not. How many of us would be willing to pay for, much less be able to afford, similar mitigation for where we live and how would one go about determining how much mitigation was necessary? Who decides or defines what amount of mitigation is “reasonable” or required? This is where technically sound judgment – not fear-mongering or sensationalism – is needed. It also requires the acceptance of a certain level of risk even in the face of mitigating actions. This is just one of the many “balancing acts of life” that we deal with every day – whether we think about them explicitly or implicitly; the same is true of how the PI deals with the potential risks associated with their facilities.
The insurance Actuaries and Claims Adjusters will have their hands full for months to get to each claim. Insurance rates and deductibles will most certainly rise as insurers attempt to recoup their very significant losses. For the PI, Process Hazards Analysis procedures are also likely to be affected, but it will take some time – years perhaps – before Process Safety Professionals develop fundamentally sound assessment practices and procedures that strike an acceptable balance between risk assessment and mitigation costs for such unusual and unpredictable events.
While “bad news” sells, we can not afford to have, nor should we allow, fear-mongering articles and editorials, written by uninformed people outside our communities and not familiar with the PI, drive the narrative for situations and events that affect us where we live. The PI must, however, also be willing to have an open and honest discussion with the communities in which they operate so that the people that live in those communities understand the risks associated with the process and the efforts the facilities are making to mitigate them. Likewise, the people living in the surrounding communities must also be ready to listen and learn about the potential risks the community and PI Workers expose themselves to every day and the efforts they take to minimize or eliminate them. This is just about being Good Neighbors.