[StBernard] Katrina's residential oil slick

[Westley Annis]

When Hurricane Katrina made landfall at 7.10 am on August 29th, 2005, it had
fallen from a category 5 storm to category 3. All the same, the winds were
whipping through at a battering 125 mph at their peak, causing billions of
dollars worth of damage. The most visible and widespread effects were
flooding and structural damage but there were other serious consequences. 

The floodwater from Lake Pontchartrain that infiltrated much of New Orleans
is known to be highly polluted, acting for many years as a sink for
pesticides, priority pollutants and heavy metals. Fortunately, as reported
in spectroscopyNOW.com earlier this year, the levels of pesticides and
gasoline components were at or below their detection limits in the freshly
deposited lake sediments around town. Only the levels of lead and arsenic
were found to be dangerously high, being well above the safe drinking water
recommendations. 

More localised pockets of pollution also arose as a direct result of
Katrina. Just 7 miles east of New Orleans at an oil refinery, the 17-foot
storm surge lifted a crude oil storage tank off its concrete foundation and
shifted it about 35 feet. The tank was ruptured when it collided with a
concrete sump and about 25,000 barrels of oil (about 1 million gallons)
escaped. Some was trapped in dikes at the site but a large fraction swept
into a nearby residential neighbourhood of 1700 homes in the town of
Chalmette, along with the floodwaters. 

As the waters receded, a dark tidemark was clearly visible on buildings,
inside and out, marking the extent of the crude oil pollution. In order to
initiate a properly managed cleanup, the spread of the oil had to be
assessed accurately. Moreover, the same data were required to help with a
class action litigation and the settlement program offered by the owners of
the oil refinery. 

This operation was not as easy as first thought, because oil had also leaked
from cars, garages, auto shops and boats in the vicinity (collectively
termed lube oils). Natural organic matter plant debris and peat transported
from bayous in the floods also complicated the picture. These extra
hydrocarbon sources had to be distinguished from the refinery crude oil. 

This immense task was undertaken by scientists from the NewFields
Environmental Forensics Practice, Rockland, MA, and the Center for
Toxicology and Environmental Health, North Little Rock, AR, with senior
reporter Scott Stout. Over 9 months, they collected 8696 soil samples from
the area and 6016 wipe samples from indoor and outdoor surfaces. 

The team embraced chemical fingerprinting of the extracted hydrocarbons
using a two-tier approach. In the first instance, GC/FID chromatograms were
compared to those of unweathered and weathered crude oil from the refinery.
Weathering affected the hydrocarbon composition in a predictable manner.
Samples with positive matches to the refinery crude oil plus those that were
equivocal were then analysed by GC/MS for further clarification. 

The distribution of certain biomarkers, especially the triterpene group, was
particularly useful. The crude oil was a Nigerian blend with a predominance
of the terpenoid oleanane, even in weathered samples. It proved to be a
consistent marker for distinguishing the crude oil from other oils and for
recognising the presence of crude oil in mixed oil samples. The lube oils
and the natural organic matter had different hydrocarbon distributions, with
other compounds dominating their profiles. 

These fingerprints were analysed using a pair of two-component mixing models
for crude oil-lube oil and crude oil-natural organic matter. Even when low
concentrations of the spilled crude oil had been mixed with other
hydrocarbon sources, the presence of the spilled oil could be confirmed. 

The mapped spatial distribution of the spilled oil showed an uneven
distribution over the polluted area. The soil and wipe analyses largely
confirmed the boundaries established by visual inspection but negative
samples were detected within the spilled area. The heterogeneity was
attributed to various factors, including the physical discontinuity of the
slick as it spread, the effects of storm drains and canals, and local wind
conditions.

Article by Steve Down