Feature ArticleRapidly Deployable SeaSonde For Modeling Oil Spill Response
By Dr. Anton G. Kjelaas
Technical Projects Manager
CODAR Ocean Sensors Ltd.
Mountain View, California
Recent spill events such as the 2010 Deepwater Horizon accident in the Gulf of Mexico and the 2007 MV COSCO Busan collision in San Francisco Bay have highlighted the benefits and need for surface current mapping for response operations. CODAR Ocean Sensors' SeaSonde high-frequency radar networks, which provided maps of ocean currents during these events, consisted of permanently installed stations at sites in developed areas that were easily accessible by road with access to grid power and high-speed data links. Most of the SeaSondes operating around the world are located at such sites, and those that are not generally require weeks or months of planning alternative power and communications solutions. Much of the world's coastlines are at risk from the production and transport of oil, but high-frequency radar current maps only cover a fraction of them. Resolving this problem requires a mobile rapid-response capability.
The concept of mobile high-frequency radar systems is not a new one. In fact, it was one of the driving concepts behind the development of the compact Coastal Ocean Dynamics Radar system at NOAA in the 1970s and early 1980s. Others have since created their own customized mobile solutions using SeaSondes integrated into automobile trailers at both Texas A&M University and NOAA's Center for Operational Oceanographic Products and Services. These solutions, while innovative in their ability to quickly deploy equipment once on site and for their utilization of wireless communications and self-contained power sources, were still limited to coastal sites accessible by roads. Areas such as the Louisiana coast, the Mississippi Delta and along most of the coast of Norway, require a solution that can utilize a number of modes of transportation, including final positioning by helicopter.
In 2009, on behalf of the companies operating on the Norway's continental shelf, the Norwegian Clean Seas Association for Operating Companies and the Norwegian Coastal Administration launched Oil Spill Response 2010, a multiyear development program for oil spill response technology. The goal of this program was to achieve significant improvements in providing continuous and effective oil spill response offshore and in coastal and shoreline areas during various weather, daylight and climatic conditions. In response to this call, CODARNOR AS, along with development partners CODAR Ocean Sensors, QUALITAS Remos (Madrid, Spain) and the Norwegian Meteorological Institute, developed a rapid-response SeaSonde for the rugged and remote Norwegian coastline.
Surface current fields reconstructed from open-boundary modal analysis. Note the existence of coastal tangent current.
The objectives of this project were to develop a mobile SeaSonde high-frequency radar unit that can be rapidly deployed to the coast of Norway to aid in effective and efficient oil spill response. The project aimed to accomplish this through developing a data service that provides high-quality SeaSonde-derived 2D current fields to the Norwegian Meteorological Institute in near real time. This would be used for spill drift model input and operations planning while also demonstrating that these current fields can improve operational oil spill drift model results.
Adapting the commercial SeaSonde to a rapid-response system required outfitting it in a protective housing that is easily transportable by a number of methods, including helicopter. The mobile unit's shell was also required to be large enough to contain all equipment and tools, provide enough room for one person to sit comfortably, and be lightweight and compact.
With this in mind, a shell was manufactured using a half-inch PVC foam core that was fiberglass-reinforced and polyester-laminated inside and out. The shell dimensions were 1.2 meters long by 1.2 meters wide and 1.5 meters high, mounted on skids that span a 1.2-by-two-meter area. For operating frequencies of 11 megahertz or higher, the unit uses a single transmit-receive antenna, which measures six to eight meters in total height with a four-meter-tall mast and whip antenna for the rest. For compactness during transport, the whip is quickly removable, and the mast was redesigned to be a two-piece hinged assembly that was mounted on the side of the shell. Once the rapid deployment unit is delivered, a single person can completely assemble and erect the antenna in about 10 minutes.
Power was supplied via redundant, off-the-shelf, two-kilowatt Honda generators with uninterruptible power supply backup and power conditioning. The prototype unit required 400 watts of continuous power, but a new transmit module design has since brought this number down to about 200 watts. Communications for data transfer were provided by the Ice.net CDMA 450-megahertz cellular service available along the Norwegian coast, and a redundant satellite communications system will be available in production models.
Rapid Data Products
While designing hardware for rapid-response deployments is important, perhaps more important is devising a plan for how to provide data products to operators and responders. Spatial gaps or shadows in surface-current coverage due to coastline geometry, radio interference, nearfield antenna distortions, etc., can often be present in real-time, high-frequency current maps. Moreover, when using a conventional local combination of radial data to obtain surface current data, it is not possible to have currents close to the coast where radial vectors from the two stations are almost parallel due to purely geometrical considerations. To overcome these difficulties, project participants have implemented and tested, in a real-time configuration, a robust and well-established method based on open-boundary modal analysis (OMA) to fill spatial gaps and expand the data to the coast.
OMA relies on the decomposition of the total velocity field as a sum of divergence-free and irrotational modes as well as modes describing flow through the open boundary. The radial data from the high-frequency radar stations are then fitted to an optimal linear combination of these modes through the minimization of a cost function. The modes depend exclusively on the domain selected, and the definition of the closed and open boundaries are computed only once and do not change in time. In addition to filling gaps and the baseline region, radial data from a single site can be used to reproduce an entire total vector field when data from another site are unavailable. Updated 2D current vector maps can be available with temporal resolutions of 10 to 60 minutes and delays due to data transfer and OMA processing of 10 to 20 minutes.
On their own, quality-controlled 2D vector maps available in near real time provide vessel pilots information on the currents in the area they are working. A real added value, however, would come from using the data to improve spill predictions and, therefore, being able to plan positioning and utilization with much more efficacy. This is specifically one of the goals of this effort. To continue this article please click here.
Dr. Anton G. Kjelaas is a special advisor to the'Research Council of Norway on development of new offshore technology. As president' and owner of CODARNOR AS, he has been focusing on new applications for seasoned high-frequency radar along the'coasts of Norway. A pioneer in the development of microwave radars for ocean wave measurements in Norway, Kjelaas has more than 25 years of experience working within the oil industry.
Chad Whelan has been technical projects manager and lead field engineer for CODAR Ocean Sensors since 2001, both planning and managing development projects and experiments related to the SeaSonde high-frequency radar.' He has more than 17 years of experience working with developmental and commercial high-frequency radar systems for ocean monitoring.