To help minimize the effects of oil spills on marine environments, chemical dispersants are used to disperse the oil in the water column so the oil can be consumed by naturally occurring bacteria. During the Deepwater Horizon incident, 1.1 million gallons of dispersant were injected directly into the flowing plume of oil and natural gas over 1500 meters deep. Dispersant's main effect is to decrease the surface tension at the oil-water interface causing the oil to form droplets smaller than ~70 microns so they can remain in the water column. Currently the efficacy of aerial applied dispersants on surface slicks is determined by measuring the droplet size decrease using a Laser In-Situ Scattering Transmissometer (LISST) or by detecting the oil in the water column using fluorometers. LISST instruments are limited to dilute mixtures, below ~500 ppm, because the LISST signal saturates for concentrated mixtures, and their windows can become occluded by oil and biofilms. Fluorometers only measure oil concentration; thus they cannot distinguish between naturally dispersed oil droplets, which can float back to the surface, from chemically dispersed oil droplets, which will remain in the water column to be naturally biodegraded. When gas is present as was the case in the Deepwater Horizon incident where it was estimated that the plume consisted of ~22% natural gas, the LISST cannot distinguish between oil droplets and gas bubbles and thus is not able to track the effectiveness of dispersants in the presence of gas. Acoustic measurements overcome the problems associated with the LISST and fluorometers and are ideal for applications subsurface near a blowout and for low ppm levels expected for surface slicks. One of the key features of the sound wave propagating through the water is the scattering at the interface between the water and object. In previous work we showed the proof of concept to measure the average oil droplet size using acoustic. We used the resonance behavior of the gas bubbles to identify them and separate their contribution to the measured acoustic scattering for various oil and dispersant combinations . We developed acoustic scattering and resonance measurements to track the size of oil droplets in the presence of gas during subsurface releases at SINTEF and in Ohmsett's large wave tank.

Dispersants play an important role in oil spill response by decreasing the spread and impact from surface oil spills and subsea blowouts. Subsea dispersants used during the Deepwater Horizon incident introduced the spill response community to dispersant injection directly into a flowing plume of oil and natural gas over 1500 meters below the ocean surface. While the application of the dispersant during the incident was intended to help the oil to naturally biodegrade by keeping oil from rising to the surface, there were no measurement tools to determine dispersant efficacy in-situ. The objective of this work was to develop the technology base to lead to such tools.

Dispersants are designed to keep the oil in suspension by decreasing the droplet size so that the natural turbulence in the water column overcomes the buoyant force pushing the oil droplet towards the surface of the water. This decrease in size is accomplished by decreasing the surface tension at the oil-water interface, causing the oil to form smaller droplets. When the droplet size decreases to less than ~70 microns in diameter, the droplets tend to remain suspended within the water column long enough to be consumed by naturally occurring bacteria [Lewis and Aurand].

Currently, the efficacy of dispersants is monitored by measuring the droplet size decrease using a Laser In-Situ Scattering Transmissometer (LISST) or by measuring the concentration of oil in the water column using fluorometers. Figure 1 shows a typical droplet size distribution of crude oil and chemically dispersed crude oil measured with a LISST during our experiments from a subsurface release at Ohmsett. The figure on the left is the droplet size distribution and the figure on the right is the cumulative volume concentration vs. droplet size. The addition of dispersant decreased the droplet size and the moved the volume distribution to lower values. Applicability of commercial LISST instruments is limited to dilute mixtures, below ~500 ppm. In addition, the LISST cannot distinguish oil droplets from gas bubbles, and the sensor windows quickly become occluded by oil and biofilms. In our testing, the windows of our LISST needed to be cleaned several times per day, while the response of the acoustic sensors was not affected by oil films. In addition, fluorometers only measure oil concentration; thus, they cannot distinguish between naturally dispersed oil, which will float back to the surface, from chemically dispersed oil droplets [Nedwed]. Clearly, there is a need for instrumentation that can directly measure the droplet size in a concentrated oil plume (e.g., subsea oil well blowout).

Figure 1.

The particle size distribution and cumulative concentration determined from the LISST for Endicott oil and Corexit dispersant at a dispersant to oil ratio of 1:50

Figure 1.

The particle size distribution and cumulative concentration determined from the LISST for Endicott oil and Corexit dispersant at a dispersant to oil ratio of 1:50

Close modal

During this work we expanded our previous work to measure the dispersibility of oil in in the presence of both air and methane for subsurface releases. We conducted experiments in the lab, during two field trips to Ohmsett and at the SINTEF Tower Basin tank (a 3 meter wide by 6 meter deep tank used to simulate subsurface releases). Details of these results are discussed in subsequent sections of this report.

We focused on measuring subsurface releases of dispersed oil and gas in our lab, at Ohmsett and SINTEF. In addition to the experimental activities we modeled the expected acoustic scattering and resonance responses from oil droplet and gas bubble based on accepted theories. We performed measurements on 4 different crude oils, Dorado, Endicott, Troll and OB for various dispersant to oil ratios (DOR) of Corexit 9500 and various amounts of gas including air and methane. In the remainder of this section we will describe the results from these measurements and modeling efforts.

Subsurface Oil Release Tank System

Photographs of our test tank setup for subsurface releases of oil and gas in our lab are shown in Error! Reference source not found.. The 150 gallon cylindrical white test tank held the water for the subsurface release of oil, dispersant and gas. Also shown are a 19.8 gallon tank for oil and a 2.8 gallon tank for dispersant. These are both connected to an air compressor to push the oil into the water tank. We also have air lines running to the bubble generator so that we can simultaneously insert oil, dispersant and air in to the water in test tank. The bubble generator will be described later in this section. The blue tube is the air hose to the bubble generator, and the white lines are for the oil and dispersant. The inside view on the right side of Error! Reference source not found. shows the bubble assembly, the oil and dispersant nozzles, the transmitting hydrophone, the receiving hydrophone, and the monostatic ultrasonic transducer.

Figure 2.

Photographs of experimental setup.

Figure 2.

Photographs of experimental setup.

Close modal

The nozzles for oil and dispersant range from 0.78 mm to 3.18 mm in diameter. The key to producing the appropriate size oil droplets is achieving the specific flow rates. We were able to achieve a 0.5 L/min flow rate with the 1.27 mm nozzle, which is in the range of flow rates and sizes used at SINTEF. The effect of premixing of the oil and dispersant on the effectiveness of the dispersant is an important parameter that we began to study. Images of gas and dispersed oil releases in our lab are shown in Figure 3.

Figure 3.

Photographs of releases of (a) air, (b) oil and air, and (c) oil, air and dispersant in our lab tank system.

Figure 3.

Photographs of releases of (a) air, (b) oil and air, and (c) oil, air and dispersant in our lab tank system.

Close modal

Bubble Generation

The underwater release of air typically generates very large bubbles (with diameters between 1 and 10 mm); however, turbulence on the scale of an uncontrolled oil blowout is likely to shear the large bubbles into numerous sub-millimeter-sized bubbles. Porous ceramics have been shown to produce bubble distributions with very few large bubbles [Commander and Prosperetti, 1989].

Figure 4) shows the bubbler we assembled using a 2 bar ceramic disc along with the steady stream of bubbles in our water tank.

Figure 4.

Porous ceramic enclosures for subsurface release of pressurized air.

Figure 4.

Porous ceramic enclosures for subsurface release of pressurized air.

Close modal

Acoustic Theory

We modeled the acoustic resonance of gas and the scattering from oil as well as the rise rates of oil and gas as part of this work to determine the expected behavior. The next subsections describe the physics and results for acoustic scattering from oil and resonance of gas.

Acoustic scattering from oil and resonance of gas

Based on our knowledge of acoustic interactions with bubbles and droplets and previous our experience, we expected the traveling acoustic wave to be scattered and attenuated as it traversed a plume of oil drops and gas bubbles. The waves scatter at the droplet-fluid and gas-fluid interfaces due to a mismatch of the acoustic properties at the boundary because of the density and viscosity differences at the oil-water and gas-water interfaces. The oil droplets and gas bubbles also oscillate back and forth in the traveling wave, and the gas can change shape through resonant vibrations and thus reradiate acoustic energy. The scattered signals create the reflections we see in sonar images, fish finders, and medical ultrasound images. In addition to providing an imaging modality, the scattering removes energy from the acoustic field, causing it to attenuate as it propagates through the plume. The droplet motion and the shape changes also remove energy from the acoustic field, further decreasing the acoustic field amplitude. In compressible particulates, like gas bubbles, energy is efficiently absorbed and reradiated when the acoustic frequency matches the resonance frequency of the bubble. This absorption and reradiation can be measured with attenuation and backscattering. Schematic representations of these mechanisms for an oil-gas-water mixture are shown in Figure 5.

Figure 5.

Schematics of acoustic measurements of oil droplets (black) and gas bubbles (green).

Figure 5.

Schematics of acoustic measurements of oil droplets (black) and gas bubbles (green).

Close modal

Both the oil droplet and the gas bubbles scatter sound and are moved by the acoustic wave. However, based on well-accepted theories, only the gas bubbles will resonate and they will dominate the reradiation acoustic energy [Medwin]. The gas bubbles are compressible, while the oil is a nearly incompressible fluid. This means we should expect to observe effects of the resonating bubbles. The resonant frequency of a gas bubble including thermal effects, the frequency dependent stiffness of the gas and effects of surface tension on the resonance frequency [Medwin] is given by:

Where, a = radius of the bubble, γ = ratio of specific heats of the gas bubble or oil droplet, PA is the ambient pressure at depth, ρw is the density of water, and β and b are dimensionless parameters that correct for the fluctuations in the effective specific heat ratio resulting from oscillation-induced heat transfer. β and b are given by the equations below.

where τ is the surface tension, Cpb is the specific heat at constant pressure of the gas inside the bubble, and κb is the thermal conductivity of the gas inside the bubble. Figure 6 shows the backscattering as a function of bubble diameter.

Figure 6.

Backscatter cross sections as a function of diameter for methane bubbles and oil droplets.

Figure 6.

Backscatter cross sections as a function of diameter for methane bubbles and oil droplets.

Close modal

Acoustic Measurement of Gas Bubble Size Distribution

The Trevorrow distribution displayed in Figure 7 is given by the following equation.

Figure 7.

Calculated backscatter from Trevorrow distributions of oil and gas at 5 depths. The droplet size distributions are plotted on log and semi-log scales.

Figure 7.

Calculated backscatter from Trevorrow distributions of oil and gas at 5 depths. The droplet size distributions are plotted on log and semi-log scales.

Close modal

Where n is bubble number density as a function of radius, a is bubble radius, in meters, N is a scaling factor to model different bubble concentration levels

To effectively measure the bubble size distribution we needed to insonify the plume with frequencies near the resonance frequencies of the bubbles which were between 5 kHz and 200 kHz. We measured the acoustic attenuation by transmitting the sound from the transmitting transducer to the receiving transducer placed on the opposite side of the tank. The bubble size distribution was calculated using the acoustic attenuation and methods described below.

The acoustic attenuation through an ensemble of bubbles is dependent on the acoustic frequency and the size distribution of the bubbles. A formula to calculate the attenuation, α(n,f), is given by the equation below ( Medwin, Urick, Caruthers):

where c is the ambient speed of sound, f is the insonfying frequency, a is the bubble radius, δ is the damping coefficient as a function of radius, n is the number density of bubbles as a function of radius, and fR is the resonance frequency as a function of radius.

The measured attenuation and corresponding bubble size distribution at three different pressures for the 2 bar bubbler are shown in Figure 8.

Figure 8.

The measured acoustic attenuation and calculated bubble size distribution from the 2 bar ceramic at 40, 50, and 60 psi based on our acoustic measurements.

Figure 8.

The measured acoustic attenuation and calculated bubble size distribution from the 2 bar ceramic at 40, 50, and 60 psi based on our acoustic measurements.

Close modal

Measurements of Subsurface Releases at Sintef

Pictures of our set up, the installation of our sensors into the SINTEF tank are shown in in Figure 9. We collected data at low frequencies between 5 kHz and 200 kHz to resonate the air bubbles and at high frequencies between 0.5 MHz and 5 MHz.

Figure 9.

Our acoustic instruments in the tower basin tank.

Figure 9.

Our acoustic instruments in the tower basin tank.

Close modal

For these measurements the oil was flowing at 1.5 Liters per minute and the DOR ranged from 1:1000 to 1:25 when premixed and 1:25 to 1:100 suing the SIT. The oil was Troll B and the dispersant was Corexit 9500. There were two different mixing scenarios, one premixed upstream from the nozzle and a supplemental insertion tool (SIT) where dispersant was released next to the oil nozzle.

The particulate size distributions produced by the SINTEF LISST for the experiments on June 27 are shown in Figure 10. The experiment on June 27 with oil and dispersant was performed at the targeted dispersant-to-oil ratio (DOR) and the droplet size did indeed decrease for most applications of dispersant as can be seen in Figure 10. The droplet size slightly decreased with the application of dispersant even though it was under dosed.

Figure 10.

The resultant particulate size distribution for the June 27 experiments from the SINTEF LISST.

Figure 10.

The resultant particulate size distribution for the June 27 experiments from the SINTEF LISST.

Close modal

Measurements of Subsurface Releases at Ohmsett

In addition to our measurements at SINTEF, we performed measurements at Ohmsett.

Figure 11 shows the frame we built and the instruments attached for deployment in the Ohmsett wave tank. We were the first team to perform a subsurface oil and dispersant release at Ohmsett. The tank was sufficiently deep to perform numerous experiments, and in a given day and we were able to work for several hours straight without stopping to clean the tank.

Figure 11.

Reassembling the frame and installing the equipment at Ohmsett Figure 12. A close up picture to identify the equipment on the frame at Ohmsett and a close up of the nozzles surrounding the 2 bar bubbler for gas, oil and dispersant releases.

Figure 11.

Reassembling the frame and installing the equipment at Ohmsett Figure 12. A close up picture to identify the equipment on the frame at Ohmsett and a close up of the nozzles surrounding the 2 bar bubbler for gas, oil and dispersant releases.

Close modal

Images of the oil, gas, and dispersant releases in calm water in the Ohmsett tank are shown in Figure 13. The dispersed oil is evident by the light brown or characteristic “café au lait” color. Acoustics images from our high frequency (2.25 MHz) transducers are shown in Figure 14 for air and a pre-mixture of oil and Corexit 9500.

Figure 13.

Pictures of subsurface releases of oil, gas and Corexit 9500 dispersant at Ohmsett.

Figure 13.

Pictures of subsurface releases of oil, gas and Corexit 9500 dispersant at Ohmsett.

Close modal
Figure 14,

Acoustic image at 2.25 MHz of subsurface release of air and premixed oil and Corexit 9500 at Ohmsett.

Figure 14,

Acoustic image at 2.25 MHz of subsurface release of air and premixed oil and Corexit 9500 at Ohmsett.

Close modal

These plumes moved from side to side due to the natural currents in the wave tank and looked more diffuse than ones seen at SINTEF. These experiments are very dynamic and point to the need to perform additional measurements in the lab and at Ohmsett. We performed measurements on Dorado and Endicott mixed with both air and methane with Corexit 9500 mixed at DOR of 1:50, 1:100, and 1:200.

Acoustic Measurement of Oil Dispersibility

While sonar are very good for visualizing oil in the water column, sonar produces an image based only on the amplitude of the received signal. We have gone beyond acoustic imaging and developed methods to directly measure dispersibility and oil droplet size in plumes of oil based on the attenuation and backscattering of the acoustic waves over a frequency range. This section describes the methods and results used to determine the gas bubble size distribution, oil dispersibility and oil droplet size from our acoustic data. While at Ohmsett and SINTEF we collected data on several different oils, for different DOR and with different gas types. We used the LISST to determine the droplet size distribution and to calculate the cumulative volume of oil that was less than 70 microns in order to benchmark our acoustic measurements. When oil droplets are smaller than 70 microns they tend to stay dispersed in the water column due to natural turbulence. Thus the cumulative volume of oil smaller than 70 microns represents the volume percent of oil that will likely stay dispersed.

One of the important outcomes of this project was the ability to measure the gas bubble size distribution in a subsurface release of oil and gas. To achieve this goal we measured a plume of air release at Ohmsett using acoustic attenuation. Figure 15 shows the acoustic attenuation as a function of frequency for various mixtures of air and oil, including an air only release, two mixtures of Dorado oil when the oil and dispersant (Corexit 9500) were premixed, and 4 releases of oil and Corexit 9500 through separate nozzles. The right graph in Figure 15 shows the bubble size distribution using our acoustic attenuation method described earlier. The air bubble size distribution was not significantly changed by the presence of oil and dispersant.

Figure 15.

The acoustic attenuation as a function of frequency for various mixtures of air and oil and Corexit 9500.

Figure 15.

The acoustic attenuation as a function of frequency for various mixtures of air and oil and Corexit 9500.

Close modal

The average droplet size distributions shown in Figure 16 were calculated during the times before (green) and after (blue) the dispersant was injected at a DOR of 1:50 into a plume of Dorado oil mixed with air. The cumulative concentrations on the right plot in Figure 16 were calculated by cumulatively summing the PPM's in each diameter bin divided by the total concentration. The first value is therefore the smallest bin's PPM over the total PPM, and the final bin is always 100% (the sum of all bins divided by the total concentration). Figure 16 shows the cumulative concentration of oil with droplet sizes below 70 microns which was calculated from each list record. This value is the percent of the oil that is less than 70 microns and thus likely to remain dispersed in the water column. For this run, the percent of oil with droplet sizes less than 70 microns is nearly zero until ~15:15, when the dispersant was injected into the plume. After the dispersant was injected, the percent of the oil with droplet sizes less than 70 microns shot up dramatically and trended around 60 percent, implying that ~60 percent of the oil has a size below 70 microns and will likely stay in the water column.

Figure 16.

Averaged droplet size distributions from the LISST for undispersed (green) and dispersed (blue), plotted in PPM (left) and cumulative concentration (middle) and the volume percent of oil with a size less than 70 microns (right).

Figure 16.

Averaged droplet size distributions from the LISST for undispersed (green) and dispersed (blue), plotted in PPM (left) and cumulative concentration (middle) and the volume percent of oil with a size less than 70 microns (right).

Close modal

By comparing our acoustic signals to the percent of the oil less than 70 microns we developed a method to measure the amount of dispersed oil. We measured the frequency response of the acoustic signals that traversed the plume as well as the backscattered signal that reflected off of the droplets and bubbles in the plume. We then calculated the attenuation and backscattering as a function of frequency as described below.

Acoustic images of the backscattering over 20 second periods before and after the dispersant was injected into the plume of Dorado oil are shown in Figure 17. The backscattering image was produced by measuring the amplitude from a series of pings as the plume flowed up through the field of view of the transducer with red indicating high scattering and dark blue low scattering. The images before and after dispersant injection look quite different and could be used as a visual indicator of a change to the plume. However, the visual change is not quantitative and would be an inaccurate means to measure the dispersibility.

Figure 17.

Acoustic images of many consecutive pings, showing backscatter from a rising oil plume of Dorado oil (top) and the corresponding frequency responses (bottom) before and after dispersant is injected into the plume.

Figure 17.

Acoustic images of many consecutive pings, showing backscatter from a rising oil plume of Dorado oil (top) and the corresponding frequency responses (bottom) before and after dispersant is injected into the plume.

Close modal

For comparison, the droplet size distribution, cumulative volume percent less than 70 microns, and the percent volume less than 70 microns for Endicott oil are shown Figure 16 and Figure 18 respectively. The addition of the dispersant reduced the size of the oil droplets dramatically and decreased the size of ~70% of the oil to below 70 microns.

Figure 18.

The droplet size distribution and cumulative concentration for Endicott oil at Ohmsett and the volume percent of oil with a size less than 70 microns

Figure 18.

The droplet size distribution and cumulative concentration for Endicott oil at Ohmsett and the volume percent of oil with a size less than 70 microns

Close modal

We also measured the signal which transmitted through the plume to calculate the attenuation. The attenuation from the through transmitted signals as a function of frequency is shown in Figure 19. In this case the bandwidth of the transducer was ~1.5 MHz to ~3 MHz. For Run 33 using Dorado and a DOR of 1:50 the attenuation before and after are only slightly different. However for Run 45 using Endicott and a DOR or 1:50 the attenuations are quite different in amplitude and in frequency response.

Figure 19.

Average through-transmitted attenuation for undispersed (blue) and dispersed (green) oil from Dorado runs 33 (left) and Endicott 45 (right).

Figure 19.

Average through-transmitted attenuation for undispersed (blue) and dispersed (green) oil from Dorado runs 33 (left) and Endicott 45 (right).

Close modal

It is important to understand that the backscattering and the attenuation are affected by both the number of particulates in the water as well as the particulate size. In general, the height or amplitude of the backscattering and attenuation are related to the concentration to first order and droplet size to second order, while the frequency dependence is related to droplet size in the first order and is minimally affected by concentration. This behavior is because the attenuation and scattering are intimately related to the ratio of the scatterer size to the wavelength of the acoustic signal, while the concentration is not. We used these two effects to isolate the concentration effects, which varies quite a bit from run to run.

We correlated the attenuation before and after adding dispersant with the median droplet size and the change in the percentage of oil below 70 microns. The plumes with oil only are plotted as black squares, the one oil and dispersant run is the purple diamond and the plumes with oil, dispersant and gas (air or methane) are blue circles.

Figure 20.

Change in attenuation response after dispersant addition as a function of the volume percentage of oil with droplet size less than 70 microns. The amount of oil less than 70 microns is the oil that will likely be dispersed due to the addition of Corexit 9500.

Figure 20.

Change in attenuation response after dispersant addition as a function of the volume percentage of oil with droplet size less than 70 microns. The amount of oil less than 70 microns is the oil that will likely be dispersed due to the addition of Corexit 9500.

Close modal

Our goal in this work was to continue developing practical tools to measure oil droplet size in-situ using acoustic scattering methods in the presence of gas. We conducted measurements in our laboratory, at Ohmsett and at the SINTEF Tower Basin on subsurface releases of oil, dispersant, and gas. In total, we made measurements on 4 different oils, some mixed with air and others with methane. From the data collected during these tests, we developed methods to measure the gas bubble size distribution in the presence of oil and, more importantly, track the amount of oil that will likely be dispersed in the water column.

We were the first team to perform subsurface releases of oil and dispersant at Ohmsett. The wave tank at Ohmsett provided a very good facility for our work and allowed us to perform many hours of measurements without needing to clean the tank as was the case at SINTEF. The large volume and general flow kept the area where we were operating clear between runs. In addition, the ability to test with waves was an added benefit.

During our work we developed acoustic measurements utilizing the complete waveform and frequency response of the attenuation and backscattering. Overall, the results are very encouraging and are a significant step in the development of in-situ field methods for measuring oil droplet size using acoustic scattering.

This study was funded by the Bureau of Safety and Environmental Enforcement (BSEE), U.S. Department of the Interior, Washington, D.C., under Contract Number E12PC00048.

Arden
Ahnell
,
Personal communication, 2012
Agrawal
,
Y.C.
and
Pottsmith
H.C.
,
2000
.
Instruments for particle size and settling velocity observations in sediment transport
.
Marine Geology
,
168
:
89
114
.
J. W.
Caruthers
,
P. A.
Elmore
,
J. C.
Novarini
, and
R. R.
Goodman
,
1999
,
“An iterative approach for approximating bubble distributions from attenuation measurements,”
J. Acoust. Soc. Amer.
,
vol. 106
,
pp
.
185
189
.
Fan
,
L.-S.
, and
K.
Tsuchiya
.
1990
.
Bubble wake dynamics in Liquids and Liquid Solid Suspensions
.
Stoneham, MA
:
Butterworth-Heinemann
.
Kerry
W.
Commander and Andrea Prosperetti, 1989, “Linear pressure waves in bubbly liquids: Comparison between theory and experiments,”
J. Acoust. Soc. Am.
,
Vol. 85
,
No. 2, February
.
Leifer
,
I.
, and
R. K.
Patro
.
2002
.
“The bubble mechanism for methane transport from the shallow sea bed to the surface: A review and sensitivity study.”
Continental Shelf Research
22
:
2409
28
.
Alun
Lewis
and
D.
Aurand
,
1997 Putting Dispersants to Work
.
Overcoming Obstacles, API Publ. 4652A
,
American Petroleum Institute
,
Washington, D.C.
Lohrmann
,
A.
,
2001
.
Monitoring sediment concentration with acoustic backscattering instruments
.
Nortek Technical note no 3. www.nortek-as.com
.
5
p
.
Medwin
,
H.
2005
Sounds in the Sea: From Ocean Acoustics to Acoustical Oceanography
.
Cambridge University Press
.
Report to the President, 2011, National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, “Deepwater The Gulf Oil Disaster and the Future of Offshore Drilling”
.
Staff Working Paper No. 4, Draft, 2010, National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, “The Use of Surface and Subsea Dispersants During the BP Deepwater Horizon Oil Spill”
,
Tim
Nedwed
,
Personal Communication, 2012
Paul D.
Panetta
,
Leslie
Bland
,
Grace
Cartwright
, and
Carl T.
Friedrichs
,
2012
“Acoustic scattering to measure dispersed oil droplet size and sediment particle size”
,
Proceedings of the Oceans 12 Conference
.
Traykovski
,
P.
,
Latter
,
R.J.
, and
Irish
,
J.D.
,
1999
.
A laboratory evaluation of the laser in situ scattering and transmissometry instrument using natural sediments
.
Marine Geology
159
:
355
Robert J.
Urick
,
1983
“Principles of underwater sound”, 3e
,
Mc Graw Hill
M. V.
Trevorrow
,
2002
.
“Measurements of near-surface bubble plumes in the open ocean with implications for high-frequency sonar performance.”
J. Acoust. Soc. Amer.
,
vol. 114
(
5
)