High Water, Bad Math and Explosions (Part 1)

Flooding along at Lake Houston after Hurricane Harvey, Sep. 2017
Photo of flooding around Lake Houston, courtesy of Brea Walton

Classes started at the end of August, but that first week of school my attention was moored firmly in Texas, pushed and pulled by the tides of the unfolding crisis in my hometown. I moved to Durham to start a Master’s degree only a handful of days before Harvey reorganized and ramped up in the Gulf of Mexico; before the move, I’d lived in Houston pretty much since birth. I found myself glued to my phone screen between every lecture and lab, passing along emergency relief info where I could, but mostly just holding my breath as spotty digital updates trickled in from friends and family a timezone away.

After punching through Rockport and Port Aransas to the south, Harvey’s roulette wheel spun in agonizingly slow motion over the western Gulf Coast for days, dealing out soggy carpet here, 6 feet of water there; boat rescues, midnight evacuations, death somewhere else. “What’ll it be?” Texans wondered each night as thunder rumbled. “Is this band of rain the one that will drive us from our home, wading into the filthy river raging through our street? Or will we slide by one more time, with water only lapping at our door? ”

Flooding in Central Houston after Hurricane Harvey, Sep. 2017
Flooding in the Houston Heights below I-10

By the end of the next week, Texas was still reeling, Irma was leveling entire Caribbean islands as she turned to gaze at Florida, and I clearly wasn’t the only one on campus fixated on storms. The following Monday (as wind and surge battered Tampa, but still a week before Maria would be christened), a trio of Nic School professors held an evening talk session, publicly answering some of the rush of questions they’d been receiving from the community and beyond. (If you have a Duke NetID, you can log in and watch a recording of the full event here.)

Dr. Susan Lozier started things off with some of the more immediate questions being tossed around in the news: she helped contextualize how these recent massive storms measure up with the historical hurricane data we have, and whether we can pin the strength of this year’s cyclones on climate change. (Short answer: No, because statistics don’t work like that. More on that topic in a later post.) Dr. Martin Doyle walked us through exactly how the National Flood Insurance Program came to amass its current quarter trillion dollars in debt, after Hurricane Katrina shattered our idea of how much storm recovery could cost. He also noted the strange political bedfellows created by the long-stagnant push to reform the country’s arguably unsustainable approach to funding flood relief. (Seriously, how often do the Natural Resource Defense Council and the Heritage Foundation have anything in common?)

Dr. Deborah Gallagher guided the event toward the sometimes-eyebrow-raising profitability of disaster relief, whether as a windfall to construction suppliers or as part of broader historical trends of community-level corporate disaster aid. Then the short presentations gave way to a less formal chat between the speakers and the audience members, who wove their own questions and diverse experiences into the resulting conversation.

The night’s discussion eventually touched on one of the strangest stories to come out of Harvey: the tense, multi-day wait for a set of explosions expected at a flooded chemical plant in Crosby, Texas. Before the rains rolled into the region, the Arkema plant northeast of Houston had been storing (and, more importantly, chilling) large quantities of organic peroxides — substances used to manufacture plastics, which have a nasty tendency to spontaneously catch fire if they heat up to, say …  room temperature. The storm swamped the plant with feet of rain, knocking out normal power and backup generators, and turning the storage trailers of peroxides into time bombs as they slowly warmed. Most people living in a close radius had lead time to evacuate before the first few trailers exploded (or “popped”, soon the preferred word of county officials managing public panic). Folks at the plant eventually went ahead and preemptively set the remaining stock ablaze, ending the suspense.

Arkema, now facing a lawsuit from nearby residents (and from some of the emergency workers hospitalized by exposure to the fumes), has claimed that Harvey-level flooding was a situation the company couldn’t have reasonably predicted or prepared for. And maybe the courts will find that they’re right: 40 inches of rain over a few days is a staggering volume of water, and both legal standards and cultural norms draw limits around what “reasonable” preparation for catastrophe looks like. (There’s plenty of room to argue about whether those limits are in the right place, whom they benefit, and whether we should aim to build larger of margins of safety into how we deal with chemicals in flood-prone Houston, where rules about what can be stored where are notoriously lax. This, too, will definitely be the subject of upcoming posts.)

Whatever you think of Arkema’s claim of unpredictability, however, a storm of Harvey’s magnitude hitting the beating heart of the Chemical Coast has long been one of the scenarios that keep emergency planners and environmental scientists across the region up at night.  Since the state’s first enormous oil boom over a century ago, a jungle of refinery complexes has grown along the coastal plain from southwest of Houston to Louisiana and beyond. The western Gulf Coast now hosts the largest concentration of refineries on the continent — though plans to build large-scale storm protection for this vulnerable economic engine have languished year after year without serious federal funding.

For context, the Arkema plant and its towering plumes of black smoke may have been only the most visible of the chemical goings-on around Houston during and after the storm. People in the city’s industry-heavy eastern neighborhoods reported weird odors and headaches for days after the rain stopped; high levels of benzene were measured in the air soon after. A dozen or so Superfund sites flooded, at least one of which definitely leaked toxic waste further into the San Jacinto River (a major waterway feeding into the nationally-important fishing hub of Galveston Bay). Millions of pounds of storm-related air pollution have been reported so far, released as area plants shut down, took on water, and sputtered back to life in the days after — though the state turned off its air quality monitors during the storm to prevent damage, making any full accounting of what was released hard to confirm.

Houston’s dense chemical infrastructure and past pollution, woven through and around residential neighborhoods on a bayou-laced, hurricane-prone coastal plain, make some contamination issues like the ones seen during Harvey not only predictable, but maybe inevitable. The question, during a catastrophic weather event, looks less like whether something will spill, and more like which particular tanks will leak; which plants will take “unpredictable” damage, which neighborhoods will find themselves flooded with what kind of fumes or residues. When Houston, as we’ve built it, faces nature, these kinds of incidents aren’t really a surprise except in their specifics (They’re likely not a surprise to some of the long-time residents of the most commonly impacted areas, anyway).

Harvey is shaping up to be one of the most costly natural disaster recovery efforts in national history, with damage estimates as high as $180 billion being tossed around. Economists may be able to point to a more refined number in a few years or a decade as the process winds down, summing up the cost of reconstruction and cleanup from the natural aspects of this disaster. But some costs (like the impacts of a surge in pollution on Houstonian health) may not reveal themselves for years — or ever (In fact, having only health impacts minor enough to go mostly unnoticed is probably the scenario to root for). The messy realities of epidemiology and exposure assessment make cause and effect even of known chemical releases maddeningly hard to trace and tease apart. Messier still to count are the long-term impacts that build up slowly over time to the surrounding landscape and environment: how do you put a dollar amount on the slow and subtle degradation of Houston’s coastal wetlands, for example, or other ecosystems whose value and critical functions we may not recognize until they disappear?

These are the sorts of costs that don’t typically show up on any company’s ledgers, or make it into business plans. We don’t always tally the costs of our smaller, quiet disasters — the ones, in a sense, made by hand. But someone, somewhere, is probably paying these costs, whether they know it or not.


The visible and invisible chemical landscapes of Houston, and the potential they hold for local and wide-spread catastrophe, played a major role in my decision to pursue grad school in environmental management. It’s a fitting topic from which to wade into this blog, which I plan to use to sift through large- and small-scale consequences of environmental policies and economic decisions we make in a complex, interconnected world.

I’ll likely spend several future posts on topics connected to the recent storms (the latest as of this writing being Ophelia, which looked like it was going to hit Ireland (!) at hurricane strength until as recently as Sunday). As the school year races along, I plan to dig into a range of other topics at the nexus of environment, public health, economics, science communication and ethics. How do we measure the impacts, good and bad, of environmental rules and practices? How complete are the pictures we have of how policy decisions play out in practice? What shapes conversations that influence those decisions?  What do we value, and how do we translate those values into laws, regulations and enforcement?

2 thoughts on “High Water, Bad Math and Explosions (Part 1)

  1. Al Gore’s presentation this past Monday at Rice University built an explanation attributing Harvey directly to global warming. The heat that is being trapped in the atmosphere is absorbed 7% by the land masses (creating very large heat waves especially in northern Africa and the Middle East) and 93% to the oceans. The heat trapped in the oceans is heating water to warmer temperatures and particularly raising temperatures at lower levels in the oceans. In the Gulf of Mexico, the water has been heated essentially to the bottom of the Gulf. This creates additional water vapor in the region which adds fuel to tropical storms and the heat intensifies the spin of the storms. Meanwhile the jet stream is being pushed lower into the North American continent by other factors, barring the storms from traveling northward across the United States. The storms are more intense, contain more water vapor, and remain stationary over the Gulf states for a longer period of time. This is precisely what made Hurricane Harvey so damaging in Houston and the surrounding areas.

    Additionally, storms are intensifying much more rapidly as they approach land instead of slowing down as was usual in the past. Maria set a record by intensifying from Category 1 to Category 5 in 18 hours.

  2. Hi Kathleen! Thanks for the comment. You’re absolutely right that we expect storms /like/ Maria and Harvey as a consequence of climate change, and the way I phrased my note in the text above may have been confusing in light of that. The point made during the talks I attended was that because of the way statistical analyses on storms are done, we can’t (yet) point at any specific storm and say things like “Climate Change caused /that/ storm to be as bad as it was, or X-percent worse than it would have otherwise been”. That’s definitely frustrating from the perspective of those trying to highlight the urgency of climate action (particularly in the face of years of irresponsible news stories claiming the earth couldn’t possibly be warming, just because New York had a particularly cold/snowy day one winter). There are plenty of other types of data that do make it very clear that the earth’s climate is changing, however. I’ll see if I can place a link in this post to my planned followup, when it’s out.

Comments are closed.