Finding buried offshore wind power cables with AUV magnetometry.

The baseline model uses a 220 kV HVAC export cable carrying 300 MW, buried 1.5 m below the seabed. That is representative of a mid-scale wind farm, or one ordinary export circuit in a larger project.[1][2][3]

The inter-array comparison uses a 66 kV cable carrying a modelled 50 MW mid-string load, roughly equivalent to three large offshore turbines. The source basis is the 66 kV string architecture and 15 MW turbine framing, not a stated 50 MW cable rating.[4]

Both scenarios are considered detectable at low AUV altitudes, with the export cable showing the greatest potential; robust burial-depth estimation is the harder part.

In the straight-conductor screening model, the 300 MW export cable gives about 659 nT at 5 m altitude, 210 nT at 10 m, 60 nT at 20 m, and 10.5 nT at 50 m. The modelled 50 MW inter-array section gives about 244 nT at 5 m, 73 nT at 10 m, and 20 nT at 20 m.

Measured offshore-wind examples sit in the same order of magnitude. OFG's MARELEC 2023 case study shows 50 Hz export-cable crossing amplitudes of a few hundred nT, inter-array responses that are smaller but still clear, and wind-farm-loop peaks approaching roughly 600 nT in the plotted survey example.[5]

Helical core lay and sheath/armour currents can strongly change the residual external field. The straight model is therefore an upper-bound screening case, not a final cable-physics claim.[6][7]

With near-perpendicular transects over a 300 m corridor, a single AUV magnetometer can locate the horizontal cable trace if the cable produces a measurable residual field. Line spacings of 25-50 m give useful repeat crossings and route confidence; 100 m is workable for a straight route; 250 m is sparse.

The cross-track anomaly shape is the useful information. Lower altitude gives a sharper peak and better localization. At 50 m altitude the anomaly is broad and burial-depth sensitivity becomes weak.

Magnetic field curvature can estimate an effective magnetic source depth because the profile width encodes total distance from sensor to source. In the straight model, the profile half-width is close to AUV altitude plus burial depth.

That is promising at 5-10 m altitude and potentially useful around 20 m with repeat transects. It is not robust at 50 m. More importantly, true burial depth remains model-limited: helical lay, phase imbalance, sheath bonding, armour currents and cable construction can mimic burial-depth changes.[6][8]