Testimony on Acoustic Technology for Determining Oil Spill Size
Richard Camilli, Ph.D., Associate Scientist, Applied Ocean Physics & Engineering, Woods Hole Oceanographic Institution
Good afternoon Chairman Markey and members of the Subcommittee. Thank you for the opportunity to speak with you today on marine acoustic measurement technologies. My name is Richard Camilli. I am a Scientist in the Woods Hole Oceanographic Institution’s Department of Applied Ocean Physics and Engineering, Deep Submergence Laboratory. My research focus is on developing advanced observation systems, including in-situ sensors for detecting subsurface pollution and studying greenhouse gas dynamics in marine environments. I have published more than two dozen peer-reviewed scientific papers and book chapters describing new oceanographic instrumentation and related robotic technologies. I have worked in the Gulf of Mexico’s Mississippi Canyon Block area and served as an expert consultant in numerous offshore oil cleanup operations, including surveys of BP oil platforms that were destroyed during Hurricanes Katrina and Rita.
For today’s hearing, you have asked me to describe acoustic technology that could be used for determining oil flow rate from the source, how it could also help to understand the recently reported oil plumes, and what would be needed in order to deploy these technologies and make these measurements.
A call for assistance
I would like to begin by providing a context describing how I became involved with assessment efforts in the aftermath of the Deep Water Horizon accident. On May 1, I received an email from my colleague Andy Bowen, who is Director of the National Deep Submergence Facility, describing a conversation that he had with BP representatives. He relayed that BP was anxious to receive any assistance for learning more about the internal status and workings of the failed blowout preventer. In particular, they needed information on the position of various internal valves and actuators.
A consensus emerged among Bowen and the BP representatives that a working group needed to be rapidly assembled to identify possible methods for achieving an internal view without risking further damage to the blowout preventer. Therefore, in Bowen’s email (dated 9:30PM May 1), he specifically requested that we consider “any technique/technology that could be used to ‘see’ inside a complex steel structure.” He also conveyed to the group, “if any of you have ideas (no matter how wild) on how to help, please pass them to this group. We should also add people who we think may have expertise... I have expanded the circulation already to include people who I think can help. No doubt there are more.” He included some suggested methods that included radiographics, ultrasonics, eddy currents, directed acoustics, and thermal imaging.
I replied via email at 1:30AM on Sunday May 2 with four questions that would help me assess viable options.
Bowen’s reply came at 9:13AM asking if I could meet him and others at the Laboratory for an 11:00AM conference call with BP representatives. Based on that teleconference the participants decided to expand the working group to include experts from areas beyond those typically associated with oceanography. Over the ensuing days there were numerous discussions and conference calls. Experts were invited to join from around the world, spanning at least 9 time zones.
By Tuesday May 3 BP had supplied blueprints of the blowout preventer and, based on its physical shape and size, the working group ruled out thermal imaging and eddy current analysis. During that afternoon’s telecon the working group agreed that teams should be formed to concentrate on high intensity x-ray imagery, direct contact acoustics, and plume imagery. At this point we did not have access to any still or video imagery of the plumes. Information provided by BP made it clear that the plumes were exiting the sources at significantly elevated temperatures, as high velocity jets. We also were informed that the product was approximately 50% gas.
Based on our experience studying hydrothermal vents, and the apparent similarities to the leaking oil plume, Bowen and I agreed to investigate possible methods for measuring flow rate from the leak sources. I would like to stress that this analysis was intended to establish an upper bound on the leakage rate. This was considered important, to help ascertain if there was at least a partial flow constriction through the blowout preventer, suggesting that at least one of the shear rams had actuated. As a part of our concept design we evaluated a range of flow measurement techniques.
The likely presence of acoustic backscattering particles within the plume and a high density contrast at the perimeter of the plume, were theoretically favorable for measuring the plumes’ cross sectional areas with an imaging multibeam sonar, and calculating vertical velocity using an acoustic Doppler current profiler (ADCP).
Multibeam sonars are routinely used by the oceanographic community to image the seafloor and features in the water column. Multibeam sonars provide very precise images of objects with different density than the surrounding seawater including fluids such as oil, gas, or hot water. We proposed to use the smallest, most precise multibeam sonar available to estimate the size and shape of the cross-sectional area of the rising plume.
A current profiler would measure the second necessary parameter – the plume velocity. The current profiler can measure velocities at the plume external boundary, but also within its core. Ideally, this technique could provide a complete cross-sectional velocity profile of the plume. Under optimum conditions, these measurements are capable of precision to better than one centimeter per second. Measurements of this precision could be obtained by mounting the instrument on the front of a remotely operated vehicle and facing the plume over a period of tens of minutes.
Based on the various performance tradeoffs of available systems, Bowen and I selected a 1,200kHz current profiler, and an imaging multibeam sonar operating at 1.8MHz. Our goal was to estimate the total plume flow (or leak) rate to within an order of magnitude and ideally to better than a factor of two.
We enlisted Teledyne-RDI, and Weatherford Pipeline and Specialty Services to help mobilize equipment and personnel as needed for this effort. We also enlisted the support of Professor Daniela Di Iorio from the University of Georgia’s Dept. of Marine Sciences. She is a leading expert in studying stratified shear flows and developer of high frequency acoustic systems for measuring fluid flows.
Twenty four hours later, on Tuesday May 4 Bowen and I drafted and submitted a proposal to BP specifying these methods for measuring flow rate at the leak sources. At 3:00AM May 5 BP responded favorably to the proposal and asked for specifications related to integrating the flow measurement system on a ROV. At 1pm the next day, BP sent a message putting the project on hold. We were informed that the containment structure had been completed ahead of schedule. Our team was thanked politely by BP representatives for our efforts. I have had no further communication with BP since May 6.
Other assessment applications
Although this acoustic flow measurement technique was intended for determining if obstructions were restricting flow within the blowout preventer, in principle, this technique may be useful for estimating total spill volume. By integrating the flow rate over time, an upper bound of the total spill volume can be calculated. By extension, this mass balance estimate can be used along with information as to the plume’s initial chemical composition (at the leak source) to estimate total gas/condensate spill size, and total expected slick area. Significant differences in the expected vs. observed slick area may indicate the presence of subsurface oil plumes. The accuracy of these estimates would be subject to several factors, including, but not limited to variability in source flow or leakage rate over time, chemical transformations such as weathering and biodegradation, as well as the inherent errors of the measurement processes.
I would like to close by stating that there are other advanced, field-proven technologies that can be applied to locate, characterize, and quantify subsurface petroleum plumes even after the source is stopped. I am prepared to expand on that point during the question and answer period if the Subcommittee so chooses.
Originally published: May 19, 2010