Executive Summary
Sustained, natural oil seepage from the seafloor is common off southern California, and is of great interest to resource managers, who are tasked with distinguishing natural from anthropogenic oil sources. The major purpose of this study was to build upon the work previously funded by the Bureau of Ocean Energy Management (BOEM) and the U.S. Geological Survey (USGS) that has refined the oil-fingerprinting process to enable differentiation of the highly similar Monterey Formation oils from Outer Continental Shelf (OCS) production and adjacent natural seeps.
In these initial studies, biomarker and stable carbon isotope ratios were used to infer the age, lithology, organic-matter input, and depositional environment of the source rocks for 388 samples of produced crude oil, seep oil, and tarballs mainly from coastal California. The analysis resulted in a predictive model of oil source families that could be applied to samples of unknown origin.
Results of the original model identified three distinct types, herein called “tribes”, of 13C-rich oil samples that were inferred to originate from thermally mature equivalents of the upper siliceous, middle shale, and lower calcareous units of the Monterey Formation. Tribe 1 contains four oil families that have geochemical traits of clay-rich, marine-shale source rock deposited under suboxic conditions with substantial higher-plant input. Tribe 2 contains four oil families that have intermediate traits, except for abundant 28,30-bisnorhopane, indicating suboxic to anoxic marine-marl source rock with hemipelagic input. Tribe 3 contains five oil families that have traits of distal marine-carbonate source rock, deposited under anoxic conditions with pelagic but little or no higher-plant input. Tribes 1 and 2 occur mainly south of Point Conception in paleogeographic settings (shelf, slope, and basin), where deep burial of the Monterey Formation source rock favored generation from all three units or their equivalents. In this area, oil from the upper siliceous unit and middle shale unit (tribes 1 and 2) may overwhelm that from the lower calcareous unit member (tribe 3), because the latter is thinner and less oil-prone than the overlying units. Tribe 3 oils occur mainly north of Point Conception, where shallow burial caused preferential generation from the underlying lower calcareous unit member or another unit with similar characteristics. Samples that do not fit within the model are classified as family 0. These samples can be identified as Monterey Formation oils and likely are derived from natural seeps, but they cannot be classified strictly due to advanced biodegradation of the biomarkers used in the model. Each of these samples retains the bulk carbon-isotopic composition derived from oil and bitumen in the Monterey Formation.
Specific objectives of this study follow:
1. Identify new areas of hydrocarbon seepage that are known to occur near OCS platforms not sampled during the previous study;
2. Geochemically fingerprint new representative oils from the OCS platforms;
3. Geochemically fingerprint select coastal tar residues associated with unusual coastal oiling events;
4. Sample additional submarine seeps to strengthen correlations between offshore active seeps and coastal residues; and
5. Quantify the discharge rates of select natural seeps and attempt to scale such results into a regional perspective of natural oil and gas seepage rates.
A total of 106 new oil samples were collected and analyzed for biomarkers: 28 samples from seeps (27 submarine, one from a sea cliff); 47 samples from representative production zones and depths from OCS oil and gas platforms; and 31 samples from random tarballs that were deposited during a storm event in February 2008. In cooperation with the U.S. Coast Guard and the California Office of Spill Prevention and Response, we collected and analyzed selected tarballs from this event. The model results showed that the tarballs most likely originated from natural seeps and that they likely were driven northward from central and southern California by ocean currents. Other random tarball analyses of samples collected in southern California, at the request of government agencies, also were shown to originate from natural seepage.
The 106 additional sample data were added to the model in an attempt to more exhaustively identify all platform-produced oil samples within known oil families. All platform oils, most tarballs, and about half of the seep samples collected from the seafloor were sucessfully classified by the model. Seafloor seep samples are often viscous, asphaltic hydrocarbon residues owing to biodegradation. In many asphaltic seafloor seeps (46%), the biomarkers were so significantly biodegraded that the sample could not be classified. Tarballs, resulting mainly from less-biodegraded oil reaching the sea surface, were positively identified by the model about 97 percent of the time. We conclude that the original model is robust for determining oil or tarball samples originating from southern California.
Regulators wish to clearly distinguish naturally occurring seep oils from anthropogenically derived platform-produced oils. The biomarker parameters are sometimes sufficient to allow unique discrimination of individual platform oils. However, platform samples and seep samples from sources geographically close to each other have biomarker parameters too similar to definitively differentiate them on that basis alone. In some cases, the degree of biogeochemical degradation or weathering that the oils or tars have experienced can be utilized. Nonweathered and nonbiodegraded oils contain n-alkane hydrocarbons and pristane and phytane isoprenoids. All of the platform oils in our sample set contain these components. In contrast, the seep oils or asphaltic residues have been exposed to significant biodegradation in the reservoir, resulting in the loss of at least the n-alkanes and isoprenoids. Therefore, the combination of chemometric fingerprinting and the presence or absence of n-alkanes and isoprenoids help to differentiate anthropogenic production oils from natural seep oils and tars. As biodegradation proceeds, biomarker compounds are sequentially attacked, resulting in nonclassification within our model.
The tarballs that we collected during our previous surveys are thought to be of very recent deposition based on circumstantial evidence. They were often found to be lying on top of the sand without any extraneous matter, and they were near the previous tidal-cycle high-swash line. We assumed tarballs with dull reflectance were qualitatively older. Sand- impregnated tar clasts that were occasionally seen on the beach were assumed to be the oldest tarballs with a density greater than water. Based on these observations and the knowledge that natural oil seepage becomes much denser with weathering, we conclude that most of the tarballs we sampled had been deposited on the beach for a few days at most. The fate of older tarballs was assumed to be mixing with beach sand and further weathering resulting in removal from the beach, offshore sinking, and deposition of ever smaller tarball fragments until they become physically indistinguishable from the environment.We surmise that an oil spill from nonnatural sources (produced or tankered oil) would be subject to a short lifespan on the beach just as natural tarballs and could be recognized by the less biodegraded chemical nature of the oil relative to natural oil seepage.
Oils representing all OCS platforms with the exception of Habitat (producing gas and condensate) and Grace (off production) have been included in this updated study. The platform oils sampled, east to west, are Eureka, Ellen, Edith, Gina, Gail, Gilda, Grace, Hogan, Houchin, Henry, Hillhouse, Platform B, and Platform C. The remaining platform oils previously sampled and analyzed are from Platforms Irene, Hildago, Harvest, Hermosa, Heritage, Harmony, Hondo, and Platform A. Two platforms in state waters, Holly (active) and Hilda (decommissioned), were included in the original studies due to their proximity to many natural oil seeps.
Platform-produced oils are only classified in tribes 1 and 2, within families 11, 12, 13, 14, 211, 212, 213, and 22. Tribe 1 oils are restricted to the Los Angeles Basin and the eastern Santa Barbara-Ventura Basin. Family 13 is the most common family from this area and the most common oil family overall (~40%). The western Santa Barbara-Ventura Basin (west of Platform Holly) and the Southern Santa Maria Basin oils are mainly from tribe 2. Tribe 3 is not represented by any oil in southern California and, therefore, must have sources outside of this area, likely in little-explored basins offshore from the central coast of California.
The use of newly acquired, high-resolution seafloor maps has significantly boosted our ability to find and then sample seafloor oil seeps. Often, sustained oil seeps build extrusive, coalescing asphalt accumulations on the seafloor—the largest as thick as 18 m and as wide as 1 km in the Santa Barbara Channel. These mapping surveys have shown that the seeps most commonly occur just west of Point Conception to Coal Oil Point and generally within the 3-nautical-mile (5.56-km) limit of California State waters. Other active seeps in southern California were observed south and west of Point Arguello and south of Santa Barbara, Carpinteria, and Summerland. Our natural seep samples came from water depths ranging from 10 to 72 m below sea level. Unexplored areas where seafloor mapping reveal mound-like structures occur north and west of Point Conception, south of St. Augustin Creek, within the deeper regions of the Santa Barbara Channel, south of Santa Barbara, and just west of Carpinteria. These areas are likely locations of persistent oil seepage and warrant future attention.
Several trial experiments were conducted to explore new ways to quantify total emission rates of natural seeps using both single-beam and multibeam sonar systems. These intercalibration experiments revealed that the single-beam sonar return was generally insensitive to bubble-flow rates and that very small bubble plumes were invisible to the sonar. In contrast, multibeam sonar had far greater sensitivity than the single-beam sonar and allowed for correction of geometric uncertainties. We experimented with the Submetrix SwathPlus-L, a 117-kHz sidescan sonar used by the USGS from 2007–2009 to map nearshore seafloor bathymetry. Results indicate that, while better at imaging gas, reproducible calibration was not possible using this system. In 2010, the USGS mapped select actively discharging seep areas within the Coal Oil Point seep field with the Reson Seabat 7111 multibeam system, which produced impressive 3-D video visualizations of gas plumes in the water column. The relative intensity of the sonar returns could be quantified; however, we are still in need of a controlled calibration experiment to relate intensity to gas and oil volume. We conclude that future efforts in remotely quantifying seep emissions should focus on multibeam sonar technology.
Our studies support the hypothesis that natural oil seepage from seafloor vents are responsible for the majority of tarball accumulation on southern California beaches. Oil fingerprinting provides the crucial tool to verify the origin of this deposited oil. While our study results are persuasive, they are not conclusive, because they depend on the assertion that beached or floating tarballs, by their inherent characteristics, are very recently deposited. We found three primary areas of seepage currently active in the Santa Barbara Channel: Point Conception, Sacate and Gaviota beaches, and Coal Oil Point. We also found that only a small fraction of tarballs did not correlate with California derived oils and are most likely from unknown ship or land-based discharges into the ocean. Produced oil from offshore platforms can often be ruled out as the origin of tarballs through the fingerprinting process, because platform oil is not significantly biodegraded. The ability to distinguish between biodegraded oils diminishes with time, and, under typical conditions, most spilled platform oil could resemble seep oil residues and seep-derived tarballs in about one month. The ability, however, to distinguish between seep-derived oil residues and platform oils within this time span is extremely valuable to regulators responding to an oil spill incident. The four platforms north of Point Conception produce oil that can be fingerprinted on the basis of chemistry alone without the need to consider biodegradation, and can thus be distinguished from known natural oil seeps in and offshore California. Finely calibrated multibeam sonar techniques can produce detailed images of discharging plumes that could possibly be modeled to obtain volumes and discharge rates.
U.S. Department of the Interior |
U.S. Geological Survey