Lessons from Harvey

K. Sahasranaman

Climate patterns are changing all over the world. Severity of rains, droughts, hur­ricanes and typhoons is becoming far unprecedented than before with its occur­rences ominously threatening to hit with regularity, leaving death and destruction in its trail.

When it strikes regions with chemical industry concentrations and when the in­dustry is not geared to face up to it, as happened in Texas when hurricane Harvey stuck, the consequent chemical plant leakages, plant & equipment collapses, fire and explosions, all prove that there is an underestimation of the risk.


With such disasters only set to increase with global warming, Sahasranaman says that it is time to revisit the entire gamut of risk assessment, preparedness and emergency management by the chemical industry.

Lessons from Harvey

Hurricane Harvey is on course to join the dubi­ous list of disasters – head­ed by Bhopal, Flixbourough and Seveso – in the Chemical Industry. Not only did the category 4 hurri­cane cripple the refining and pet­rochemical hub of the world in Texas, but it also led to unprece­dented spillage and release of tox­ic and carcinogenic chemicals in­to the atmosphere. According to a report in New York Times, more than 2000 additional tonnes of chemicals were released from 46 industrial facilities during the week Aug 23- Aug 30. In addition, 14 sites holding toxic wastes were inundated. It is not that Harvey ar­rived unannounced. It was known for a week and its lethal potential was assessed at least 48 hours be­fore it struck Texas. Yet the chemi­cal plants in arguably the most in­dustrially advanced country of the globe were caught napping and their responses turned out woeful­ly inadequate. How did this hap­pen? A silver lining of every acci­dent is that it mandates us to raise the bar for safe practices. But for Flixbourough we would not be having Hazop; Responsible Care emerged out of Bhopal’s grave­yard. So, what do we learn from Harvey? I can quickly think of fol­lowing three lessons.

Revisiting risk assessment

Hurricanes like Katrina, Harvey and Irma can no longer be con­sidered as “acts of God” or Black Swan events. They are now the new normal and risk matrices need to be reconstructed. Weather ex­perts believe that hurricanes will get more frequent and more pow­erful as a fallout of global warm­ing. Chemical Industry needs to have plans in place to deal with them. Mathematically speaking, risk is a product of severity and likelihood. Likelihood of monster hurricanes like Harvey and Irma need to be ratcheted up by one or even two levels. This will alter our risk perception completely and will call for new safeguards and di­saster management plans.

Hazop is another risk manage­ment tool that needs a serious re­think. While a lot of attention gets lavished on the Inside Battery Limit (ISBL) plant during Hazop study, the Outside Battery Limit (OSBL) plant gets a short shrift. It is the OSBL that has the biggest foot print, usually several multi­ples of the ISBL. Again, it is the OSBL that holds most of the in­ventory in the plant. OSBL comes across as a poor unglamorous cousin of ISBL during design, engi­neering, personnel training, opera­tion and maintenance of a process plant. Even in education curricula it hardly gets the importance it de­serves. This approach and attitude needs to change.

Plants, especially those on sea coasts, can be asked to redo Hazops considering flooding and power outage as a cause. The worst credible consequence and safe­guards for this scenario should be placed in public domain. This will go a long way to improve the trust and confidence of the larger com­munity around the plant.

Revisiting plant inventory

Many chemical plants, especial­ly those built in the last millenni­um, carry far too much inventory. This excess baggage harks back to the pre-computer and pre-Internet era when supply chain manage­ment practices, as we know them today, were not prevalent. Online procurement now has done away with time consuming paperwork and approvals. RFID tracking of shipment has reduced uncertain­ties. Logistics planning, optimised shipping methods and routes, syn­ergy with suppliers etc helps the industry to prune down the inven­tory. Inventory reduction not only brings down the operating expens­es significantly, but is a giant step towards improving plant safety.

Consider what happened dur­ing Harvey. Nearly half a million gallons of gasoline spilled out from just two tanks owned by one of the largest pipeline operators. The ex­act cause of the leak is still under investigation. The pounding rain­fall also reportedly sank floating roofs of at least a dozen large stor­age tanks leading to leaks. At least two dozen storage tanks holding various refinery materials have col­lapsed spewing out carcinogen­ic aromatics – Benzene, Toluene and Xylene. API standards man­date that floating roofs should be designed to withstand a rain­fall of 250 mm in 24 hours. Harvey brought more than double that rain. Perhaps on hindsight, we need to redesign and strength­en floating roof tanks for a high­er rate of rainfall. It will most cer­tainly add to the cost, but would be a small price to pay for protect­ing the environment. Increased fre­quency of inspection, maintenance and structural audit of large stor­age tanks should also help. Clearly storage tanks bore the brunt of the storm surge. The damage would have been less with reduced inven­tory.

It must be mentioned here that some refineries were able to in­crease the levels in their stor­age tanks during the build-up to Harvey. This made the tanks less buoyant and less vulnerable to floating when the water swamped the tank farm.

Revisiting plant siting

Reasons for siting refineries on the coast are obvious. But they are sitting ducks for hurricanes. And if hurricanes and typhoons are like­ly to be more frequent we need a radical rethink. New refineries, not that many new ones are likely to be built, should be located inland. Existing refineries should consider erecting multiple barriers to avoid loss of containment.

One of the most horrific acci­dents in the aftermath of Harvey was the fire and explosion in a per­oxides plant. Peroxides are very unstable compounds and need to be stored under refrigeration. When the hurricane knocked out main power supply to the plant, the backup generators failed to start because they were submerged in water. Locating the generators at grade level was clearly a bad idea. But it appears bad only on hind­sight. The company had refriger­ated trucks to move out the perox­ides to a safer inland location. But by the time they decided to act, the roads were overwhelmed with wa­ter. The accident could have been totally avoided had the backup generators been located at a height beyond reasonable access of flood waters.

Post Script

Seismic zones are taken into ac­count while designing structures in a chemical plant. Now that hurri­canes are expected to become even more frequent than killer earth­quakes, we need an improved sys­tem and response in place from de­sign techniques to disaster man­agement. Ironically, the response to Harvey was hampered by indus­tries taking shelter under the fig-leaf of an anti-terrorism act, un­der which they were not obligat­ed to disclose to authorities the na­ture and quantity of chemicals they stored in their premises.

Readers’ responses may be sent to:

k.sahasranaman@gmail.com or



Please enter your comment!
Please enter your name here