The ever-increasing need for seabed infrastructure enabling offshore energy extraction and storage requires accurate prediction of erosion and deposition rates of the surrounding seafloor sediments. Seabed erosion and deposition (scour) can become catastrophic for the infrastructure itself but also for the surrounding habitats. Scour predictors largely assume that the bed is composed of unimodal sediments and objects have simple shapes. In mixed non-cohesive beds, however, complexity in the sediment transport process, including the hiding exposure (HE) effect, will likely impact the seabed erosion and deposition. Better understanding of scour dynamics in more complex settings can improve the ability to predict scour development around differently shaped objects sitting on mixed and coarse beds, which are ubiquitous in palaeo-glaciated environments, for instance.
In this work, the scour development around the 135 m long wreck of the ship SS Apapa was investigated through time-lapse analyses of multi-beam echosounder surveys over nine years (co-registered bathymetry and backscatter intensity), hydrodynamic measurements and sediment samples. Due to the interaction between the tidal flow and the wreck, flow velocities increased by about ~2.3 times downstream of the wreck, with the highest amplifications where an undisturbed flow first encounters the wreck. The extent of flow disturbance in the wake of the wreck measured between 0.76 and 2.3 times the length of the wreck, with the highest disturbance at slack tides. Vertical flow disturbance of up to 1.66 times the height of the wreck (17.5 m) was also identified directly over the wreck, with nearly a doubling of flow velocities. Laterally, a flow diversion of 90° was identified directly over the wreck. Two consistent ‘zones of bed mobility’ exist at either side of the wreck, at distances between 0.27 and 1 times the length of the wreck. The extent of these zones was identified by a drop in the bed mobility accompanied by a reduction of the flow speed between 30% and 35% when compared to the undisturbed background flow. The seafloor at the SS Apapa area was composed of mixed coarse sediment, with coarser material present in the deepest parts of the scour mark and finer material present at the depositional features. MBES datasets (bathymetric and backscatter) showed that the depositional feature remained fine over the years but was the most variable in bathymetry. Large variations in bed composition were also observed at the deepest points of the scour mark, where bathymetry varied less over time. A disintegration and shifting of SS Apapa between March 2018 and June 2019 changed the exposure of the wreck to the flow at the north-east and north-west sides of the wreck. This altered the hydrodynamics of the area and subsequently changed the erosional and depositional trends observed until 2018.
To investigate the role of the sediment mixture in scour dynamics in more detail, flume laboratory experiments were conducted with two flow speeds over a 9.4 cm long cylinder on a bed, using six sand and gravel mixtures, pure sand, and pure gravel. One lower flow speed mobilised just the sand fraction, whilst a higher flow speed mobilised both the sand and gravel fractions. The bed was acoustically scanned three times for each run and sediment cores were taken to analyse changes to bed composition laterally and with depth. For the lower flow, mobilising only the sand fraction, the scour mark was 66.4% longer, 12.1% deeper and 4.8% wider in a bed composed of 20% gravel and 80% sand, when compared to the pure sand bed. At the higher flow speed, mobilising both fractions, the scour mark was 43.6% longer, 40.9% wider and 13% deeper in beds containing 12.5%, 5% and 20% gravel respectively, when compared to the pure sand bed. Two ‘zones of bed mobility’ emerged in the wake of the object. At the lower flow velocity, they sat at a mean distance of 2.6 and 9 times the object’s length, with the first and second zones being longer for the bed consisting of 15% gravel (~4 times the object’s length) and 12.5% gravel (10.7 times the object’s length) respectively. At the higher flow velocity, the two zones sat at a mean distance of 2.2 and 8.6 times the length of the object with the first and second zones being longer for the bed consisting of 20% gravel (2.5 times the object’s length) and 12.5% gravel (10.8 times the object’s length) respectively.
To up-scale the application of all these offshore and laboratory findings, a model is needed where the parameters can be changed and scour in complex settings (in term of object shape and bed composition) can be predicted with better confidence. A coupled numerical hydrodynamic and sediment transport model (TELEMAC3D – GAIA) was therefore modified and used to replicate the SS Apapa site and the laboratory environment. An existing hiding-exposure correction was used in the sediment transport component of the model. The outputs of the model include hydrodynamic, bathymetric, bedload and suspended load information, and these were compared with the real-world analyses of flow modification and bed mobility, both offshore and in the laboratory. The modelled flow modifications around the wreck, deviated between 3% and 15% from the observations, while the modelled flow modifications around the cylinder in the lab experiments deviated less than 4% from the observations. The modelled scour extent for both the wreck and the cylinder measured withing 0-27% of the observations. The numerical model appeared to have difficulties in predicting the scour formation in the cases of the bed consisting of less than 10% gravel in the higher current speed but predicted well the scour mark formation for over 93% of the modelling work using different sediment mixtures. This gave confidence in the model to model the impact of changes to the object’s exposure and the bed’s composition. The numerical model has produced a 66% longer scour mark when increasing the object’s exposure to the flow by 10% in the case of the laboratory work. The numerical model has also identified an increase in the produced turbulent kinetic energy of 33% when doubling the exposure of SS Apapa to the flow.
This knowledge and increased model capability should enable developers to find solutions where the impact on the bed is minimised. The results should also aid the design and extent for scour for scour protection, again minimising enhanced bed mobility and the associated costs from habitat loss and damage to the infrastructure.