Layered seabed, layered risk: North Sea ground models and cable burial

What the BGS study puts on the table

The British Geological Survey, working with Durham University and the University of Dundee, has published an open-access report mapping the shallow soils in the upper two metres of the North Sea seabed – the exact depth band that governs cable burial and protection. The work sits under the EPSRC-funded project “Offshore cable burial: how deep is deep enough?” and pulls together more than 12,000 geological records from BGS archives and the Crown Estate’s Marine Data Exchange. It is the most detailed regional picture of shallow subsurface conditions across the UK North Sea assembled to date.

The headline finding is straightforward and consequential. Layered soils are the norm, not the exception. Most sites contain multiple soil types stacked within that top two metres – commonly sand over clay, gravel, peat or shallow bedrock. The southern North Sea is dominated by surficial sands, while the northern North Sea is more variable. Thin gravel layers are widespread, and thicker gravel and shallow bedrock appear locally, most awkwardly in nearshore areas where burial is already hardest. Organic-rich soils and peat show up too, though less often.

The study’s direct challenge is to the simplified, single-soil assumptions that standard cable burial risk assessment commonly applies. It also restates a point that the installation community knows but that budgets often ignore: burying deeper does not automatically buy more protection. Deeper trenches cost more, add installation risk, and can trap heat around the conductor. That combination – real geological complexity against a modelling shortcut – is what puts route and burial decisions back on the table.

Where single-soil assumptions cost money and expose the asset

Cable Burial Risk Assessment, as formalised under the Carbon Trust Offshore Wind Accelerator, balances external aggression against burial protection to arrive at a target depth of lowering. The external side considers anchor penetration from merchant and fishing vessels and the reach of fishing gear – otter boards, beam trawls, hydraulic dredges. The protective side is the soil above the cable. When that soil is entered as a single representative type over long route sections, both sides of the balance are wrong at once.

The error is not academic. Anchor and fishing-gear penetration depend heavily on soil strength. A stockless anchor bites far deeper into soft clay than into dense sand, and a trawl door drags differently across gravel than across silt. Model the seabed as uniform sand when there is a soft clay layer beneath, and the assessed aggression depth is optimistic. Set a depth-of-lowering target against that optimistic figure and the delivered protection is thinner than the risk case assumes.

The cost exposure runs the other way as well. A blanket deep-burial target applied across a route that is genuinely benign in places spends money and vessel time cutting depth that the risk case never demanded. Trenching production rates collapse when a jetting tool meets stiff clay or gravel it was never specified for, and standby against weather while a second pass or a different tool is mobilised is where day-rate campaigns bleed. A layered ground model lets you set depth-of-lowering per segment, which is where both the saving and the assurance live.

Then there is thermal exposure, which single-soil assessments handle poorly. Cable ampacity under IEC 60287 is sensitive to the thermal resistivity of the surrounding soil. Saturated fine sediment might sit near 0.5–1.0 K·m/W, while dry or poorly graded sand can exceed 2.5 K·m/W. Bury a power cable deeper into a high-resistivity dry sand horizon and heat dissipation gets worse, not better – the conductor runs hotter for the same load. The BGS finding that sand frequently overlies materials with very different properties means the thermal environment along a route is layered too, and a rating derived from one assumed backfill resistivity can be unconservative in exactly the segments where sand dominates.

What actually governs the burial decision

Strip the problem back and three engineering questions decide burial, and all three depend on knowing the stratigraphy in the top two metres rather than a single averaged soil.

The first is protection against external hazards – the CBRA balance above. This needs soil strength as a function of depth, not a label. A cone penetration test profile, or a well-characterised vibrocore with particle-size analysis and shear-strength data, tells you how far an anchor or trawl board penetrates at that location. Marine soil investigation to ISO 19901-8 and CPT to the ISO 22476 series exist precisely to produce that profile.

The second is installability, which is a tool-versus-soil match. Jet trenching moves sand efficiently and struggles in stiff clay, gravel and rock. Mechanical trenching and chain cutters handle firm and cemented materials but are slower and heavier on the deck. Ploughs perform well in uniform soils and can be defeated by boulders, gravel lags and shallow bedrock. The BGS observation that thin gravel is widespread and shallow bedrock appears locally nearshore is a direct warning about trenching refusal – the point where the tool cannot achieve target depth and the cable is left proud, needing rock berm, concrete mattress or articulated pipe as secondary protection. Those are the most expensive metres on any cable route.

The third is thermal performance over the design life, which ties depth of lowering to the soil’s thermal resistivity and to how stable the seabed is above the cable. This is where the southern North Sea sand fields carry a particular trap. Mobile bedforms migrate; a sandwave crest recorded at installation can become a trough later, so the depth of cover measured on as-built can reduce over time even without any external strike. Depth of lowering has to be set against the deepest predicted trough over the design life, not against the seabed level on the day of the survey – which means the ground model has to be paired with a sediment-mobility assessment, not read in isolation.

In our experience, DNV-RP-0360 on subsea power cables in shallow water frames these same considerations, and we treat it as the natural reference point when rebuilding a burial case on the back of better ground data. The point the BGS work reinforces is that the inputs to that assessment have to be spatially resolved. A layered seabed demands a layered decision.

Where route and burial decisions go wrong

1. Averaging the soil along the route

The most common failure is exactly the one the study targets: taking a single representative soil for a long section because the geophysical and geotechnical data are sparse and the schedule is tight. Averaging masks the soft layer under the sand and the gravel lens under the silt. The fix is not more caution in the average – it is resolving the layers. Sub-bottom profiling that can pick out horizons at 0.25–0.5 m resolution, tied to geotechnical ground-truth, is what converts a smeared cross-section into a usable stratigraphic model.

2. Treating burial depth as the only protection lever

Budgets gravitate to a single number – a route-wide depth of lowering – because it is easy to specify and easy to check. That number can be simultaneously too deep for thermal comfort in dry sand and too shallow for protection over soft clay. Depth is one variable in the CBRA balance, alongside soil strength, hazard exposure and secondary protection. Managing the risk sometimes means shallower burial with a rock berm over a bedrock outcrop, not a deeper trench the tool cannot cut.

3. Divorcing the burial case from the thermal rating

Cable rating and cable burial are frequently owned by different teams – electrical engineering sets ampacity, installation engineering sets depth – and the soil thermal resistivity assumption falls between them. When burial depth is increased late in design for protection reasons without re-running the IEC 60287 rating against the actual backfill resistivity in the sand-dominated segments, the cable can end up thermally constrained in service. The two decisions have to share one ground model.

4. Surveying for bathymetry and calling it a ground model

A clean multibeam surface to IHO S-44 tells you the seabed shape, not what lies beneath it. Backscatter adds surficial sediment character, and there is real value in pulling seabed classification out of the multibeam returns you already collect before committing to expensive coring. But neither resolves the layering that decides trenching and thermal performance. Route-clearance decisions built on bathymetry alone routinely miss the shallow gravel and rock that later stops the plough.

How to decide: turning ground models into route-clearance calls

The BGS report changes the default. The reasonable starting assumption for a North Sea route is now a layered seabed, and the burden is on the survey and design programme to characterise those layers rather than to justify a single-soil shortcut. The following steps make that concrete.

• Build a layered ground model to at least two metres, segment by segment. Integrate multibeam bathymetry, side-scan, and a sub-bottom profiler capable of resolving 0.25–0.5 m horizons, and tie every geophysical interpretation to geotechnical ground-truth – CPTs and vibrocores with particle-size analysis and shear strength per ISO 19901-8. Set the sampling density by geological variability, not by even spacing: the northern North Sea and all nearshore approaches warrant tighter coring than the southern sand fields. Where the campaign feeds early engineering, treat it as a properly scoped site-characterisation exercise aimed at a FEED-ready ground model rather than a route ribbon of isolated points.

• Run CBRA on layered inputs and set depth of lowering per segment. Follow the Carbon Trust methodology but replace the single-soil input with the resolved stratigraphy so anchor and fishing-gear penetration are assessed against the actual soil strength profile. Report depth of lowering as a segmented target with the governing hazard named for each segment, not as one route-wide figure.

• Commission soil thermal resistivity testing and re-run the rating against it. Measure thermal resistivity (needle-probe method per IEEE 442 / ASTM D5334, feeding the value into the IEC 60287 rating) on samples from the sand-dominated segments and use the worst-case dry-state value in the ampacity calculation. Plan to validate in service with distributed temperature sensing so the thermal assumption is checked against reality rather than trusted indefinitely.

• Map trenching-refusal risk before selecting the tool. Flag every segment with gravel thickness, shallow bedrock or dense material that threatens the primary burial method, and pre-plan the secondary protection – rock berm, mattress or articulated pipe – with quantities and vessel scope, so a refusal in the field is a planned contingency rather than a standby event. Match the trenching spread to the dominant soils along the route, and carry a documented alternative for the flagged zones.

• Set burial against the mobile seabed, not the survey-day surface. Where sandwaves are present, quantify migration and lowest expected trough over the design life from repeat bathymetry or regional bedform studies, and specify depth of lowering relative to that deepest predicted level. A depth-of-cover figure that ignores bedform migration is an as-built number with no service meaning.

• Front-load nearshore and landfall investigation. The BGS work places the thickest gravel and shallowest bedrock nearshore, where burial constraints and hazard exposure both peak. Investigate these zones first, because they carry the highest chance of a route change and the most expensive protection, and a late discovery there disrupts the whole installation sequence.

The engineering value of this study is that it removes the excuse for the single-soil shortcut on the UK Continental Shelf. The data now exist to expect layering and to design for it. The decision facing cable owners and their survey contractors is whether to keep buying route-wide burial targets against averaged soils, or to spend the survey and interpretation effort that turns a layered seabed into a segmented, defensible protection case. On a multi-hundred-kilometre export route, the second approach is where both the cost control and the asset assurance are found.

---

Based on: Geological research in North Sea helping to safeguard subsea cables