Depth Accuracy of Data Points in Marine Soil Investigation-Paper

Paper

Pdf Summary

The paper presents a probabilistic method for assessing the vertical depth uncertainty of marine soil investigation data relative to the seafloor, following ISO 19901-8. It evaluates uncertainty components such as sensor accuracy, seabed-frame settlement, tool and drill-string inclination and compression, water-level variation, vessel movement, and—in sampling—uncertainty in the recovered sample’s in-situ position. Estimates use a 95% confidence interval and are compared with ISO depth-accuracy classes Z1–Z5.<br /><br />Two examples demonstrate the approach. In a non-drilling cone penetration test (CPT) to 30 m below seafloor, estimated uncertainty met class Z1 (0.1 m) to about 26 m and class Z2 (0.5 m) at greater depths. Correcting for push-rod compression and measured seabed-frame settlement could bring the full profile within Z1. In a seafloor-controlled drilling example, both a downhole CPT and a soil sample met class Z2: their estimated uncertainties were 0.19 m and 0.41 m, respectively. The sample estimate includes substantial uncertainty from partial recovery.<br /><br />The analysis indicates that measuring seabed-frame settlement can materially improve depth accuracy, and that integrating existing measurements—such as corrections for rod compression and vessel excursion—offers further gains. Additional systems, including borehole-inclination monitoring and continuous drill-string control, may also help. The authors emphasize that achievable accuracy depends on deployment method, equipment, site conditions, and the chosen vertical datum. Seafloor referencing is especially relevant to subsea infrastructure, while GNSS-based referencing may suit fixed structures extending above sea level. Overall, the paper provides a detailed, probabilistic framework for documenting and improving depth accuracy in marine investigations.

Keywords

marine soil investigation

vertical depth uncertainty

ISO 19901-8

seafloor reference

cone penetration test

seabed-frame settlement

drill-string inclination

sensor accuracy

depth accuracy classes

probabilistic method