Refinery accident pollution estimates badly need refining

By Benjamin Leger, The Lens contributing opinion writer |

Talk to anyone who lives near an oil refinery or chemical plant in Louisiana, and they’ll tell you they worry every night about what might happen.

“We go to sleep always ready to roll,” says Sonyja Thomas, a resident of north Baton Rouge near ExxonMobil, the second largest refinery in the nation. She’s had to evacuate before, when the toxic fumes from a late-night accident entered her home.

Our neighbors in Texas are currently facing a similar chemical emergency. In Texas City, a company town nestled in an industrial hub near Galveston, refineries owned by BP, Valero and Marathon Oil and a Dow Chemical facility all suffered massive power outages last week that are still causing problems days later. The outages  resulted in huge smoking flares all night long, forced residents to shelter in place until the following day, closed area schools and by week’s end had sent at least 25 residents to the emergency room with respiratory problems. The BP refinery, the only one of the four facilities with its own power-generating equipment, is the same site where, in 2005, an explosion killed 17 workers and injured another 170.

The cause of the recent outages? Salt and dust buildup on the facilities’ electrical equipment due to dry weather. Refineries in Louisiana often cite tropical storms and bad weather as the causes of their accidents, bringing to mind the tsunami-triggered outages in Japan that led to the ongoing Fukushima reactor catastrophe. But apparently dry conditions can wreak havoc, too.

Refinery officials in Texas City claimed Tuesday morning they had been continuously monitoring the flares after the power outage and publicly stated there was no indication of harmful chemicals in the air. Later that day, the Texas Commission on Environmental Quality (the counterpart of Louisiana’s Department of Environmental Quality) said the amount of chemicals spewing from the cluster of facilities was so off-the-charts that their monitoring equipment couldn’t properly read it.

Within two days of the outages, reports to the TCEQ showed BP had emitted 74,000 pounds of sulfur dioxide, and Valero had emitted 43,000 pounds, in addition to carbon monoxide, benzene and other chemicals.

These early figures are alarming, but sadly not the full picture. In Texas, as in Louisiana, industrial facilities are required to report any accident (they call them “upsets”) within 24 hours and provide a full report within a week. Those initial reports often lowball the amount of pollution leaving facilities, in those precious hours when communities are most urgently in need of accurate information.

During Hurricane Gustav in 2008, ExxonMobil Baton Rouge suffered a power outage that shut down sulfur units, caused flaring for days and, if we are to believe initial reports from the refiner, released 599,122 pounds of pollutants – mostly sulfur dioxide, with about 100,000 pounds of nitrogen oxide and volatile organic compounds.

The actual releases were possibly much higher. That’s because the initial reports sent to LDEQ are rarely based on actual monitoring of emissions – they are based on “engineering estimates,” a phrase that also comes up in the majority of TCEQ reports since the Texas City accident. These estimates may be just a fraction of the discharges that actually occur.

Last year, the Louisiana Bucket Brigade analyzed five years of refinery accident reports in the state for our Common Ground report. Here’s what we discovered: “Refineries routinely overestimate the efficiency of flare incineration rates. Refinery accident reports show (that) the plant environmental managers make assumptions that the flares will incinerate 98 or 99 percent of chemicals. … There is, however, no engineering proof to support these calculations. The evidence (specifically an EPA analysis of BP in Texas City last year) shows that actual combustion efficiency of the flare can be as low as 50 percent.”

Many reports reiterate the same industry mantra: that chemicals sent to the flare during an accident are 99 percent burned off, regardless of whether the facility is functioning optimally. In the Hurricane Gustav example above, the 599,122 pounds of pollutants reported represent what ExxonMobil claims is the 1 percent that escaped. And LDEQ relies on these reports without any follow-up investigations.

With the alarming frequency of refinery accidents in Louisiana and Texas, and the typically non-alarmist responses from officials in the aftermath, it’s important that we demand more accurate information. Our safety and our health depend on it.

When a refinery or chemical plant is flaring, we need to know if officials are monitoring the incineration rates at the source or just estimating based on “engineering judgment.” When officials claim they are monitoring air quality after an accident, we need to know what chemicals are being tested for and the detection limits of the equipment used. Industry samples are often set at higher levels of exposure and therefore rarely reflect the minimum health screening levels used by the EPA.

Furthermore, the trends and frequency of accidents are often shrugged off as “acts of God” or the price of doing business, when they are really a threat to public health and safety and need to be regulated as such.

All of these accidents could be prevented if industry would quit cutting corners and invest in updated equipment that would save them money in the long run and make our communities safer and healthier. While industry keeps forking out cash for preventable mistakes, we’re paying for it through health care costs and the ruin of our environment.

Benjamin Leger is a Monitoring and Evaluation Specialist with the Louisiana Bucket Brigade.

Help us report this story     Report an error    
The Lens' donors and partners may be mentioned or have a stake in the stories we cover.