Torstein Olsmo Sæbø¹, Hayden John Callow¹, Per Espen Hagen<sup>2

1</sup>Norwegian Defence Research Establishment, Kjeller, Norway
²Kongsberg Maritime, Horten, Norway

Proceedings of the 33rd Scandinavian Symposium on Physical Acoustics

Abstract

Inspection of subsea pipelines traditionally requires multiple sensors and a large number of passes. This makes pipeline inspection a time-consuming and costly process. Introducing interferometric high-resolution synthetic aperture sonar (SAS) can potentially increase the efficiency substantially. Mounted on an autonomous underwater vehicle (AUV), such a system can provide highly detailed data sets of a pipeline with just one or two passes - either fully autonomously, or with operator supervision as desired.

A likely concept for for pipeline inspection with a SAS is running parallel with the pipeline in two tracks, one to each side. This ensures full coverage of the pipeline itself and of a large swath around the pipeline. In addition, the shadow behind the pipeline can be removed by co-registering and merging imagery and bathymetry from each side, into a high-resolution three-dimensional view. Unfortunately, the geometry of the pipeline is such that, except for the strong specular reflection, very little energy is reflected directly back to the sonar. Also, some energy is reflected into the ground before it is received, causing an additional degradation of the data.

We suggest that it can be beneficial to inspect the pipeline from other aspects than parallel to the pipeline. By using tracks perpendicular to the pipeline, the specular reflection, the multipath and the shadow all disappear. Equally important, we receive echo from the entire upper part of the pipeline, which results in a much better three-dimensional representation.

In June 2009, a Kongsberg Maritime HUGIN 1000-MR AUV equipped with a HISAS-1030 interferometric SAS performed such a pipeline inspection mission in Norway. The HISAS data was processed to 2x2 cm imagery and fully co-registered bathymetry with a resolution of approximately 18x18 cm. In this paper we will show how to combine imagery and bathymetry observed from opposite sides of the pipeline into one merged dataset. This enables us to estimate the diameter of the pipeline, which in this dataset, is unknown. We also estimate the height of the pipeline relative to the seafloor from the shadow behind the pipeline. Furthermore, we use the bathymetry from perpendicular tracks to estimate the three-dimensional curvature of the pipeline and to estimate an average shape of the pipeline.