Increased dispersion of oil from a deep water seabed release by energetic mesoscale eddies
Graphical abstract
Introduction
Oil spills from seabed releases are a worldwide risk to the marine environment, and drilling is gradually pushing into deeper waters as a result of resource discovery and technological innovation (Burgherr, 2007; Gallego et al., 2018). Deepwater Horizon is the highest profile example of a deep seabed release, costing BP $145 billion (Lee et al., 2018) and spilling 7–8 × 105 m3 of oil over 86 days (Camilli et al., 2010; Crone and Tolstoy, 2010; McNutt et al., 2012; Joye et al., 2016). Deep water drilling also occurs in the Faroe-Shetland Channel (FSC; Fig. 1a) in the North Atlantic, which has been a region of UK oil and gas development since the early 1990s (Smallwood and Kirk, 2005). As of August 2018, there were 162 active well heads in the FSC (source: UK Oil and Gas Authority), three-quarters of which were deeper than 200 m. A spill in the FSC lasting for a typical response time of 30 days could result in an oil release of up to 3.4 × 105 m3 (Gallego et al., 2018).
The FSC is a hydrodynamically complex and energetic environment. A strong slope current (up to 1 m/s) along the West Shetland slope transports warm, saline surface water north-eastwards towards the Norwegian Sea and eventually into the Arctic Ocean. Near the seabed at >1000 m depth, a bottom current transports relatively cold, fresh deep water south-westwards, which either follows bathymetry through the Faroe Bank Channel into the open North Atlantic, or overflows the Wyville Thomson Ridge into the Rockall Trough (Turrell et al., 1999; Sherwin et al., 2008). Mesoscale eddies can extend across the width of the FSC and south of the Faroe Islands near the Faroe Bank Channel (Sherwin et al., 1999, Sherwin et al., 2006; Darelius et al., 2011). Large internal tides and non-linear internal waves have been observed in the region, which can act to increase turbulent mixing rates (Sherwin, 1991; Hosegood and van Haren, 2004; Hall et al., 2011, Hall et al., 2019). A unique stratification structure is also present, where the main thermocline typically resides at several hundred meters below the sea surface and separates the exchanging water masses (Berx et al., 2013; Fig. 1c, d, e).
The behavior of oil from a seabed release depends on ocean currents and stratification, in addition to properties of the oil such as viscosity, temperature, gas-oil ratio (GOR), flow rate and orifice diameter (Yapa and Chen, 2004). A previous plume modelling study suggested that oil will be trapped at 650–800 m depth from a 1000 m release in the FSC, depending on the release rate and ambient ocean conditions (Johansen, 2000b). Main et al. (2017) used a global ocean circulation model based on the Nucleus of European Modelling of the Oceans (NEMO; Madec, 2016) to predict the transport of oil from the FSC, and found that far-field oil transport was dependent on its depth. Oil near the surface travelled north-eastwards towards the Arctic Circle, whereas oil trapped at depth reached as far west as Greenland. However, they did not consider the influence plume dynamics might have on the vertical distribution of pollutant, or the role of surface weathering processes such as evaporation and emulsification. Additionally, the horizontal resolution of the model (1/12°) was coarser than required to explicitly resolve mesoscale eddies in the FSC region (2 km or less; Oey, 1998).
Oil Spill Contingency and Response (OSCAR) is a state-of-the-art modelling system that can be used to predict the fate and trajectory of an oil release during emergency response. OSCAR comprises of a 3-D fates model (Reed et al., 1995, Reed et al., 2000), near-field plume model (Johansen, 2000a) and droplet breakup model (Johansen et al., 2013). It is typically forced with horizontal current velocities from an operational hydrodynamic ocean model. OSCAR has been well-validated against historical emulsion observations (Abascal et al., 2010), synthetic aperture radar (SAR) satellite observations (Pan et al., 2020), and against the DeepSpill field experiment (Johansen et al., 2003).
In this study, OSCAR is used to consider how oil from a seabed release in the FSC could be transported by hydrodynamic processes, and how an increase in hydrodynamic model resolution influences the predicted dispersion. To simplify our diagnostics, biodegradation and direct wind forcing are not included. We demonstrate that enhanced mesoscale variability in fine-resolution hydrodynamic models leads to a dramatic increase in horizontal dispersion, and that stratification influences the depth of trapping and subsequent far-field transport. These results will help guide the choice of hydrodynamic forcing for emergency spill forecasting.
Section snippets
Hydrodynamic forcing
To force OSCAR, both Oil Spill Response (OSRL) and the UK Centre for Environment, Fisheries and Aquaculture Science (Cefas) currently use operational ocean forecasts based on the UK Met Office 7 km horizontal resolution Atlantic Margin Model of the North-West European Shelf (FOAM AMM7 NWS, hereafter referred to as AMM7; O'Dea et al., 2012, O'Dea et al., 2017) In November 2018, an updated version of this model became available to use operationally (FOAM AMM15 NWS, hereafter referred to as AMM15;
Modelling the oil spill
Oil is released from the seabed (1122 m depth) on 1 February 2017 for nine days. The release location (61.07°N, 3.705°W) was chosen because it is in an area with several active well heads (Fig. 1a). The simulation is run for 30 days, which accounts for the release period plus three weeks of further dispersion. A total of 100,700 m3 of oil is released at a constant rate of 0.130 m3/s, guided by the estimate of Gallego et al. (2018). The oil exits from an orifice with diameter 0.1 m, similar to
Surface transport
Pollutant at the surface for the AMM7 release resides close to, and slightly west of, the release location for approximately two weeks (Fig. 3a, c). The surface emulsion is then transported by the slope current in a continuous band north-eastwards, parallel to the 600 m isobath (Fig. 3e, g). For the AMM15 release, initial surface transport is north-eastwards, and the emulsion has already begun to diverge into two distinct branches by the end of the release period (Fig. 3b). The emulsion
Discussion
Oil transport from a deep water release in the central FSC will divide into two main pathways. Oil that has reached the surface, in addition to oil trapped in the upper portion of the water column, will predominantly travel north-eastwards along the continental slope towards the Norwegian Sea. Deeper oil will be transported westwards, advected by deep currents and guided by bathymetry through the Faroe Bank Channel and eventually out into the open North Atlantic. No oil overflows the Wyville
Summary
This study provides an insight into how a hydrodynamic model with resolution fine enough to resolve mesoscale processes influences the predicted dispersion of oil from a deep seabed release. The modelling systems used here are currently in use by spill responders; this study therefore serves to directly inform industry of what is missed by coarser resolution hydrodynamic models, and how that may impact real-world predictions. Additional hazards that have been uncovered include the potential for
Funding
RMG was funded by the UK Natural Environment Research Council (NERC) industrial CASE studentship grant NE/M010023/1, awarded to the University of East Anglia and Cefas.
CRediT authorship contribution statement
Ryan M. Gilchrist: Conceptualization, Methodology, Software, Formal analysis, Investigation, Writing - original draft, Visualization. Rob A. Hall: Conceptualization, Methodology, Writing - review & editing, Supervision, Project administration, Funding acquisition. John C. Bacon: Conceptualization, Methodology, Supervision, Funding acquisition. Jon M. Rees: Conceptualization, Methodology, Supervision, Funding acquisition. Jennifer A. Graham: Methodology, Software.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
RMG was funded by the UK Natural Environment Research Council (NERC) industrial CASE studentship grant NE/M010023/1, awarded to the University of East Anglia and Cefas, and an associate member of the NERC EnvEast Doctoral Training Program. We particularly thank SINTEF for support with OSCAR. Well head data (latitude, longitude, depth and status) are available from https://www.ogauthority.co.uk/data-centre/data-downloads-and-publications/well-data/. This study has been conducted using EU
References (45)
- et al.
Analysis of the reliability of a statistical oil spill response model
Mar. Pollut. Bull.
(2010) - et al.
Environmental effects of the deepwater horizon oil spill: a review
Mar. Pollut. Bull.
(2016) In-depth analysis of accidental oil spills from tankers in the context of global spill trends from all sources
J. Hazard. Mater.
(2007)- et al.
Current status of deepwater oil spill modelling in the Faroe-Shetland Channel, Northeast Atlantic, and future challenges
Marine Pollution Bulletin
(2018) - et al.
Near-bed solibores over the continental slope in the Faeroe-Shetland Channel
Deep-Sea Res. II Top. Stud. Oceanogr.
(2004) DeepBlow – a Lagrangian plume model for deep water blowouts
Spill Sci. Technol. Bull.
(2000)Development and verification of deep-water blowout models
Mar. Pollut. Bull.
(2003)- et al.
DeepSpill – field study of a simulated oil and gas blowout in deep water
Spill Sci. Technol. Bull.
(2003) - et al.
Droplet breakup in subsea oil releases – part 2: predictions of droplet size distributions with and without injection of chemical dispersants
Mar. Pollut. Bull.
(2013) - et al.
The Gulf of Mexico ecosystem, six years after the Macondo oil well blowout
Deep Sea Research Part II: Topical Studies in Oceanography
(2016)