A combined telescopic fluorescence instrument and a wide band multibeam sonar (WBMS) were developed as dual sensors for detecting underwater oil leaks and plumes. The fluorescence instrument is a forward-looking fluorescence polarization instrument with motorized telescopic focusing that can detect oil at a standoff distance. The instrument responds only to materials showing fluorescence polarization, and therefore is able to distinguish oil from other fluorescing species in water. The WBMS is as an acoustic sensor that provides 2D as well as 3D topology for mapping and water column imagery. The operational integration of the two sensors result in a more definitive identification and mapping of oil in the water column co-registered in time and space. The sonar records a 3-dimensional map of a leak, and then the fluorescence sensor is use to confirm that the leak that the sonar detected is oil. A sensing scheme to integrate the operation and overlap the field of view of the 2 instruments was developed as well as the software algorithm to automate the detection process. Evaluations of the integrated sensors were done in a test pool and in open water. Testing results shows that the combined sensors are very effective in detecting and identifying oil plume in the water column.
Global demand for oil has resulted in an increase in oil production and its byproducts. The transportation of oil through pipelines and tankers increases the risk of polluting bodies of water. Even with strict rules and regulations of oil transportation and oil explorations, accidents leading to oil spills still frequently occur. Numerous accidents occur annually, with thousands of tons of oil being spilled into aquifers, resulting in the contamination of marine environments and endangering marine ecology. The Deepwater Horizon oil spill in the Gulf of Mexico clearly shows the magnitude of environmental problems that can occur with an oil spill (Natural Resources Defense Council Report, June 2015). Reliable oil spill and leak detection instruments are needed that can be used as early detection and warning systems for oil leaks and spills so that proper and effective remediation measures can be implemented immediately. Ideally, sensors that can operate at a standoff distance and interrogate a wide area are preferable and will be more effective in tracking oil leaks and plumes in the water column.
Several oil spill control methods based on radio wave reflection suppression, oil fluorescence and contact electrical sensors have been developed (Brown et. al. 1998). Radio wave radars operate well in open seas, where sea surface waves are stable, but in closed coastal areas they are much less effective. On the contrary, application of contact electrical or fluorescence sensors is limited to small high-risk areas. Remote laser induced oil fluorescence is probably the most reliable method allowing oil detection on any surface, and a number of airborne fluorescence lidars have been developed (Sato et. al. 1978, O’Neal et. al. 1980). Some of them currently operate to monitor spills in active sea traffic regions. However, airborne lidar fluorescence is very expensive, not applicable for long term continuous spill monitoring, and not able to detect deep underwater oil contaminants. It is also susceptible to interference from other fluorophors in water such as humic compounds and chlorophyll.
A more useful instrument form of a fluorescence sensor is a submersible system that can detect oil leaks or spills at a standoff distance. EIC laboratories, Inc. has developed a submersible fluorescence instrument with motorized telescopic focusing (Bello 2010). Figure 1A shows a photograph of the fluorescence instrument. Unlike traditional fluorimeters used for underwater measurements, the fluorescence instrument employs a laser projected telescopically outwards from the instrument body. As it is deployed near the seabed, the laser is automatically focused on the subsurface using feedback from a sonar altimeter. Fluorescence from the focused, polarized laser source is collected in 180° backscatter by the same telescopic optics and is separated optically into its vertically and horizontally polarized components. After passing through wavelength selective filters, the intensity of these components is measured using separate photomultiplier detectors. As currently configured, this fluorescence instrument is capable of detecting fluorescent objects at distances ranging from near contact to 25 meters away. This fluorescence sensor is a very effective tool for detecting and monitoring oil contamination in the water column and for oil deposited in the seafloor (Hansen et. al. 2009). Because oil main constituents are aromatic organic compounds, crude and refine oils are highly fluorescent when illuminated with ultraviolet and visible light. A limitation of the telescopic fluorescence sensor, however, is that it is a point sensor and therefore has limited detection coverage. Also, the detection range of the fluorescence sensor is not very far, especially in murky water conditions.
A wide band multibeam sonar (WBMS) platform is an acoustic sensor that provides 2D as well as 3D topology. Through the use of electronic beam forming, the WBMS has the advantage, of being able to look at multiple angles of incidence at the same time, which allows the acquisition of 3D imagery. Norbit has developed ultra compact WBMS designed specifically for use on all platforms. The wideband multibeam technology allows long-range real-time data collection, while simultaneously achieving high range resolution. The WBMS has previously been demonstrated in mapping oil plume, both in the water column as well as in water interfaces e.g. seafloor and ice (Eriksen et. al. December 2010, Eriksen et. al. November 2010). Although the WBMS can acquire an image of a possible leak, it cannot definitively identify the image as that emanating from oil since the sonar detects only differences in reflectivity and attenuation in the water column. Figure 1 shows a photograph of a WBMS sonar from Norbit.
By combining the operation of the telescopic fluorescence sensor and the WBMS and correlating the data from the 2 instruments, a more definitive identification of oil leaks in the water column can be achieved. The operation protocol for the 2 instruments can be setup in such a way that the WBMS is used to scan a wide range of the water column for possible leaks and pinpoint the position of the leaks for the fluorescence sensor to then interrogate and identify as hydrocarbon leaks.
TELESCOPIC FLUORESCENCE SENSOR AND WBMS SYSTEM INTEGRATION
To implement the operational integration of the 2 sensors, the instruments were mounted on an underwater platform that positions the field of view of the instruments into the same area. Figure 2 shows how the 2 sensors are mounted in the platform along with other supporting instruments. A high definition video camera is included in the instrument suite and used to obtain visual recording of the area that the instruments are interrogating. Two sonars are mounted into the platform, a 200kHz WBMS and a forward-looking sonar (FLS). The WBMS is used to obtain a wide view of an area to identify probable oil leaks. The WBMS is mounted at a fix position on the platform. From the WBMS sonar image of possible leaks, the coordinates of these are determined. Three sensors are mounted on a pan-tilt positioner: the FP instrument, the FLS and the HD video camera. The pan-tilt is a software controlled positioning device that provides an up and down and rotational motions, which can be precisely positioned to a spot and can be programmed to scan an area. Through software, the pan-tilt can be programmed to obtain several point measurements in the suspected leak areas using coordinates derived from the WBMS sonar image. The FLS is mounted on the pan-tilt so that it’s field of view overlaps with that of the fluorescence sensor. The FLS, which has a narrower field of view than the 200 kHz WBMS, provides a close-up image of the probable leak area that is being verified with the FP instrument. In addition to the sonar, the HD camera, with its field of view also directed to fluorescence instrument scanning range, provides an additional visual image of the interrogated area. All the instruments are connected to a processing bottle. The processing bottle contains an embedded computer and power supply that powers all the instruments. The embedded computer contains the software that controls the operation of all the instruments and connected to a top computer above via Ethernet. The top computer controls the operation of the embedded computer, can send commands to operate the control software in the embedded computer and process, display and store data.
OIL DETECTION AND VERIFICATION PROCESS
A Labview software (Leakview) was developed that implemented an automated oil detection and verification process using both the WBMS and the laser fluorescence sensor. Figure 3 shows a diagram of the oil detection and verification process that was implemented through the Leakview software. The process is started by acoustically sweeping a 3D space in the water column. The sonar steerable antenna transmits a narrow beam vertically while the receiver spatially filters the space in the horizontal direction. The process forms 3D matrices, which can cover the half-space dome. The space is then divided into a selected number of vertical slices. Inside each slice, a motion detection is performed. The motion detection data is then passed to the leakage detection, which is a temporal filter and exemplifies the persistent motion areas. If the persistency exceeds some predefined threshold, that area is considered as a leak and an alarm is generated. The alarm consists of five locations (bearing and range) of the largest leaks. The size in dm2 is also passed along with the alarm as a NMEA text message to use minimal bandwidth. This information is easily passed through the acoustic modems or slow links to the topside operation software. The topside software processes the alarms and decides if they should be verified with the fluorescence sensor. The fluorescence verification utilizes the underwater pan and tilt mechanism to position the fluorescence sensor in the proximity of the detected leak. The scanning is performed in the predefined sector, e.g. 5 degree around the bearing given an alarm message. At the same time video is recorded and still high-resolution camera images are recorded. Also the FLS, which is mounted on the pan/tilt mechanism, is constantly logging the imagery data for future processing and verification.
PERFORMANCE TESTING OF THE INTEGRATED LEAK DETECTION SENSORS
Underwater testing of the integrated sensors was conducted at the US Army Cold Regions Research and Engineering Laboratory (CRREL) Test Basin pool. Two different in tank detection tests were conducted to verify the performance of the oil leak detection system. The first experiment was conducted with oil placed inside a small plastic water bottle (16.9 oz.) that was tethered to a string and dangled in the water column. The second experiment was an actual release of oil leak into the water column using an oil pumping mechanism. The experiment with the bottled oil was to prove the automatic detection and verification mechanism of the system, while the oil leak experiment was to show more of the operational aspects.
When the bottled oil was dropped in front of the leak detection system, the movement of the bottle caused the WBMS to report it as a leak and flag it as yellow color in the sonar displayed image. The coordinate that was recorded by the WBMS was then sent to the pan-tilt positioner to position the fluorescence sensor to the spot. The pan-tilt was programed to scan a predetermined angle sector around the spot so that the fluorescence sensor could confirm that the detected leak was a hydrocarbon. If the fluorescence sensor confirms the scanned area as oil, Leakview changes the yellow marking into a red spot. Figures 4 and 5 show the results of one of the experiment that was conducted with the oil bottle. In Figure 4, the red spot in the WBMS sonar image is an area in the water column that was confirmed by the fluorescence sensor as an oil leak. Figure 4 also shows still image of the oil bottle that was recorded by the video camera. Figure 5 shows that the possible oil leak confirmation by Leakview. The figure on the left shows the oil bottle before being hit by the laser, and the figure on the right shows the oil bottle being illuminated by the laser. As shown in Figure 5 , before the laser beam is on the oil bottle, the Oil Detected button on Leakview is not illuminated. However, when the laser is on the bottle, fluorescence signal is detected and the Oil Detected button on Leakview turns to green. This indicates that oil is detected by the system.
OIL LEAK DETECTION EVALUATION OF THE INTEGRATED WBMS/FLUORESCENCE SENSOR
The oil release experiment was done using the CRREL oil pumping mechanism. A flexible hose was attached to a pump and the other end was attached to a pipe with a nozzle tip. The nozzle was positioned in the bottom of the tank and in the middle of a containment hoop that was placed in the middle of the tank. The pump speed could be regulated from 6Hz to 60Hz. Figures 6 shows a result of the oil release experiment showing that the WBMS detected a possible leak and then a confirmation by the fluorescence sensor as indicated by the red marking in the WBMS image. Figure 6 also shows the still image from the video camera showing that the laser was positioned in the plume during the detection and confirmation. Figure 6 shows another oil release detection experiment result. The Leakview display shows that the fluorescence intensity that is detected by the fluorescence sensor is quite strong. Figure 7 shows another oil release result showing that the fluorescence sensor was able to detect and confirm the oil plume several times as it scanned back and forth through the plume. Leakview shows the FP signals increased and decreased as the laser passed back and forth through the plume. Figure 8 shows an oil release experiment with the instruments positioned at ~4 meters distance from the oil leak. The water visibility in the water column was poor during the testing because of the presence of algae in the water. Even with the poor visibility in the water column, test the result in Figure 8A show that the fluorescence sensor was able to confirm the presence of oil. Figure 8B shows the fluorescence signal detected by the FP sensor at various positions in the plume, and showing quite strong fluorescence signals.
SUMMARY AND CONCLUSIONS
Data from a sonar and fluorescence sensor are more valuable if the 2 sensors are operated as a single, unified detection platform. The integration of a WBMS and a telescopic fluorescence sensor showed that the data fusion from the sensors is very effective in the detection and definitive confirmation of an oil leak in the water column. The operational scheme of the integrated sensors was configured so that the WBMS scans a wide area of the water column for possible leak sites and records the positions of these suspected areas. A verification and confirmation with the fluorescence sensor is then performed at these sites. If the fluorescence sensor records a signal then that particular spot is flagged as an oil leak. A Labview automation software was developed that automates the detections scheme steps. Testing of the integrated sensors was performed in a test pool to evaluate the performance of the combined sensors. A controlled oil leak release was setup in the test basin with the sensor platform positioned in the bottom of the tank. Test results show that detection scheme was very successful in detecting the plume as a leak and then verifying it as an oil leak. These results show the advantage in integrating multiple, complementary sensors into a single sensor system for oil spill detection. With further development of the combined WBMS and fluorescence sensor, the system can be deployed in underwater vehicles such as an AUV or an ROV that can then be used for detection of oil contamination in the water column, in the seabed or under ice.
This study was funded by the Bureau of Safety and Environmental Enforcement (BSEE), U.S. Department of the Interior, Washington, D.C., under Contract E14PC00033. The findings and opinions expressed in this article are solely those of the authors and do not necessarily reflect the views and policies of the BSEE, nor does mention of the trade names or commercial products constitute endorsement or recommendation for use.