ASA Lay Language Papers

2nd Pan-American/Iberian Meeting on Acoustics


Acoustic Observations in Support of the Response to the Deepwater Horizon Oil Spill

 

 

Thomas C. Weber – weber@ccom.unh.edu

Larry Mayer – larry@ccom.unh.edu

University of New Hampshire

Durham, NH 03824

 

Alex De Robertis - Alex.DeRobertis@noaa.gov

Christopher D. Wilson - chris.wilson@noaa.gov

NOAA-Alaska Fisheries Science Center

Seattle, WA 98116

 

Sam Greenaway - samuel.greenaway@noaa.gov

Shep Smith - co.thomas.jefferson@noaa.gov

Glen Rice – glen.rice@noaa.gov

NOAA-OCS

Silver Springs, Maryland 20852

 

Popular version of paper 3aUWa2

Presented Wednesday morning, November 17, 2010

2nd Pan-American/Iberian Meeting on Acoustics, Cancun, Mexico

 

 

On April 20 of 2010 an explosion on the Deepwater Horizon drilling rig killed 11 workers and precipitated an oil spill in which approximately 4.9 million barrels of crude oil were released from the damaged wellhead.   The rig was operating in 1500 meters of water, and this depth confounded our collective ability to readily observe what was happening to the oil as it exited the well.  By mid-May, observations of underwater oil plumes were reported, and later that month we embarked on the first of many cruises onboard NOAA research ships with the principal aim to use scientific echo sounders (i.e. sonars) to map subsurface gas and oil near the wellhead.   Over the next few months, we used these echo sounders to map the many natural methane gas seeps in the area, to directly observe the oil in the top few hundred meters of the water column, to examine some of the local effects of the oil plume on marine organisms throughout the water column, and finally to monitor the integrity of the well after it was capped in mid-July.

 

Each of these subsurface acoustic mapping cruises used a variant of the Simrad EK60 scientific echo sounders that have been developed for fisheries research.  These systems provide a quantitative, calibrated output and have a low noise floor and high dynamic range.  Multiple acoustic frequencies, ranging from 12-200 kHz were used during the research cruises.  No single frequency would have sufficed: the lower frequencies required to ‘sound’ the full ocean depth are particularly useful at identifying gas seeps in the water column.  By contrast, small oil droplets – some of which were thought to be only 10’s of microns in diameter- only weakly scatter acoustic waves  which are more easily observed using higher frequencies.  Unfortunately, higher frequencies (for example 200 kHz) do not travel as far in the ocean and could not sample the full water column when mounted on a surface vessel. 

 

At the outset, little was known about the size distribution or quantity of the subsurface oil droplets, and so we essentially embarked on missions of exploration with the express aim of learning anything we could to help aid the response to the oil spill.  Concerns about our echo sounders acoustically interfering with efforts to stem the oil flow limited our access to the area where surfacing oil was prevalent.  Thus, the initial cruises were mostly spent no closer than 10 km from the wellhead.  During this time we mapped a number of natural gas seeps while searching for changes to the deep scattering layer (a ubiquitous community of deep-living marine organisms that scatter acoustic waves) that might indicate the presence or effect of subsurface oil plumes (Figure 1).  Although natural seeps commonly occur in the Gulf of Mexico, the number of nearby seeps was not previously known.

 

weber01.gif

 

Figure 1.  An ‘acoustic curtain’ representing the raw 18 kHz acoustic echosounder data exhibiting both the Deep Scattering Layer and observations of natural seeps.  The acoustic data were processed to extract and correctly position the natural seeps (white point clouds).  The gray surface represents the seafloor topography.   

 

Despite coupling our acoustic observations with other direct (fluorometer) and indirect (dissolved oxygen) readings that indicated the presence of subsurface hydrocarbons, no unambiguous direct acoustic observations of the oil were made 10 km or more from the well head.  The scenario changed when the NOAA Ship Thomas Jefferson gained access to within 1.5 km of the well head.  During this time, the surfacing oil plume was detected using high frequency (200 kHz) echo sounders down to depths greater than 150 m (Figure 2).  Low frequency (12 and 38 kHz) echo sounder data showed disturbances throughout much of the water column, with a morphology that indicated that the source of the disturbance was the rising oil plume (Figure 3).  These observations indicated a high potential for acoustically mapping portions of the oil plume and its local effect on marine organisms, but no subsequent cruises were undertaken until the well was capped in mid-July.

 

weber02.gif

 

Figure 2.  Near-surface (top 150 m) 200 kHz acoustic observations of oil from the surfacing plume.  Top: locations of observations (stars) and corresponding acoustic oil plume images (arrows).  Bottom: schematic representation of the oil plume and six locations of the observations (left) with their corresponding volume scattering strength estimates of the plume (right).  Red indicates that the plume is more dense, blue indicates less dense.  Here, the plume density is observed to decrease with increasing distance from the well head.

 

weber03.gif

 

Figure 3.  Acoustic observations showing anomalies in the acoustic backscatter (38 kHz) from marine organisms.  Although these are not direct observations of the oil, the anomalies are thought to be associated with the rising oil plume.

 

After the well was capped, our focus shifted toward monitoring the integrity of the wellhead by acoustically searching for leaking gas directly over the well head and in the immediate vicinity.  After conducting a number of acoustic tests that indicated we would not interfere with other on-going operations, we were given relatively unfettered access to the site.  During this wellhead integrity monitoring stage, a small gas leak was detected on the flange prompting an extremely high level of scrutiny including continuous video monitoring from ROV’s and (nearly) daily acoustic observations to track changes to the leak over time (Figure 4).  This vigilant acoustic monitoring of the capped well continued for many weeks until the well was finally cemented in August.

 

weber04.gif

 

Figure 4.  Acoustic observations of a small gas leak at the Macondo wellhead several days after it was capped in mid-July.