Oil spill trajectory simulation for the Saltpond oilfield, Ghana, West Africa

Abstract

The exploration and production of petroleum resources is not without environmental consequences and these remain an important factor in contemporary global oil industry. As an emerging oil producing nation, very little or nothing is known about the behaviour of Ghana's sweet light crude in water under varying environmental conditions. This is crucial for an efficient management of accidental spills. Therefore, this study designed a deterministic model for oil spill trajectory simulation using the Saltpond oilfield (Ghana) as study location. The main objectives were to: (a) design a diagnostic GNOME model to configure and generate localised oil spill scenarios; (b) model different oil spill scenarios and weathering profiles for varying environmental conditions e.g. wet/dry seasons, different currents and wind conditions; and (c) Predict the 'minimum regret' Splots as a precautionary principle. A combined Eulerian-Lagrangian approach involving a hydrodynamic model containing localised environmental parameters from the Saltpond oilfield was utilised. The General NOAA Operating Modelling Environment (GNOME) based on a Cϩϩ programming language was used to execute the analyses. In addition, the ADIOS2 oil weathering tool was used to estimate the oil budget for different spill quantities based on the physical and chemical properties of Ghana's sweet light crude at Saltpond field. The key findings are: 5000 and 10,000 barrels of spilled oil beached after a minimum of 12 hours in 97% of the scenarios under the influence of winds from the South. Wind speeds leading to beaching were 35.7% for 14 and 16 knots, 14.3% for 17 knots and 7.1% for 15 and 20 knots respectively. Weathering results showed a maximum of 43% of spilled oil evaporated after 116 hours with a minimum and maximum dispersal of 1% and 7% respectively. This study, for the first time, has provided hitherto unavailable analysis and information on oil spill transport forecasting and trajectory simulation for the Saltpond Oilfield in Ghana which environmental managers/engineers would find readily useful in responding to future oil spill disasters. The approach and methodology used in this study could also be adapted for other offshore regions in Ghana and West Africa.

An Analysis of the Fate and Trajectory Simulation of Bonny Light Crude From The Dangote Single Point Mooring(SPM) Terminal Off the South Atlantic Ocean, Lagos Nigeria

“…Aside GNOME, there are other modelling tools (Elizaryev et al 2018), however, GNOME is widely used because it is freely available, user friendly, simple and provides accurate and detailed analysis of oil spills in the marine environment (Ibarra -Mojica et al 2018;Marta-Almeida et al 2013). It also has the capacity to model different types of oil and provides the minimum regret (worst case scenario) of spilled oil (Korsah and Anifowose 2014). Despite all of this, GNOME has limited 3D capabilities (Marta-Almeida et al 2013), weathering characteristics and oil-shoreline interaction processes (Beegle-Krause 2011).…”

“…The assumed spill source that was selected in GNOME is the "point/line source". The hypothetical spill was also configured at 1000 splots to account for the map window display output (Korsah and Anifowose 2014). The choice of pollutant used determine the output of GNOME and ADIOS2 (Zelenke et al 2012).…”

“…According to the NOAA (1996), the "minimum regret" uses available methods to analyse the sensitivity of various estimates to errors in the input data, thereby exploring the implication of alternate plausible scenario. The "what if" function has the capacity to increase the predictive accuracy of the model by taking into consideration errors and uncertainties in the environmental conditions inputted when setting up the model (Korsah and Anifowose 2014). Taking this into consideration, the minimum regret function in GNOME represented by "red splots" (Cheng et al 2011), was activated and the spill scenarios created with the hydrodynamic data obtained for the periods shown in table 3.…”