By: Luke Richardson
Subsea pipelines are an integral component to offshore oil and gas and a key discussion in the survey community. Inspection of these assets is critical to reducing environmental risk along with avoiding costly preventable maintenance repairs. Over the years, the techniques for pipeline inspection have evolved from the manual “click and scroll” of video or sonar data to offering data quality that can allow deep learning automation to efficiently offer eventing and damage assessments without the need for time consuming data review. This is the premise for moving towards the “Automated Pipeline Inspection.”
With the advent of resident vehicles and increased utilization of AUVs, the idea of further automating pipeline inspection has become more prevalent than ever. Newer technologies and sensor packages are providing levels of data quality that only a few years ago seemed to be infeasible, allowing companies to make better informed decisions on subsea inspections and to offer datasets that can easily be interpreted through machine learning. Systems that can support the automated pipeline inspection initiative are:
Laser Scanners – Extremely high-resolution 3D point cloud models that are quantifiable for automation and advanced machine learning. Capable of detecting small anomalies including dents, buckling – along with enabling predictive maintenance through modeling free-spans, nearby debris, concrete mat pitting, anode volume/depletion rates and pipeline movement. Industry leading resolution provided by laser scanning systems offers advanced pipeline tracking, eventing and structural referencing to unlock further automation capabilities.
Stills Camera – Clear and crisp stills images with vehicle position market within the meta-data for advanced reviews. While operating simultaneously with a capable laser scanning system, a stills camera is able to generate a photomosaic that can visually qualify 3D models, and can be draped over the laser model for further detail. Selecting a system with advanced parameters such as higher bit depths and sensitivity will allow users to get more out of image automation through real-time image enhancements/corrections for better photomosaics.
3D Mapping Sonar Systems – Including synthetic aperture, side scan, multi aperture and multibeam echosounders offer complimentary 3D point cloud models over larger areas with lower resolution for general inspection and qualification. Functional in most water conditions, these sonar systems can generate a broader model of the surrounding seabed and asset, while offering initial detection of potential concern areas to further validate with higher resolution sensors.
Sub-Bottom Profiler – A system capable of offering a quantifiable model of the asset below the surface. Combined with the surface modeling technology, the sub-bottom profiler offers the overall condition of the asset for locating and modeling buried lines. Allows users to make decisions on their assets without seeing the line and can support tracking algorithms.
Field Gradient CP System – Relatively new solution that offers remote Cathodic Protection Survey without need for reference measurements at anodes. This system is a contactless sensor using electric potential to model CP assessment and could reduce the need for ROVs in the field. Enhances the AUV potential and allows users to perform multiple scopes from one platform.
Positioning Systems – including inertial navigation, doppler velocity logs, USBL, help position the vehicle and feed into the other sensor data to offer accurate anomaly positioning. Higher accuracy positioning enables longer missions and precise event locations.
Utilizing each of these sensors for your pipeline inspection will enhance the data quality and accuracy of your survey, with the added benefit of simplifying the analysis through automation. The future of subsea inspections and the role automation plays in enhancing surveys will contribute to reducing project timelines through minimizing time to results and reducing the data analyst’s workload, while offering certainty in decisions based on quality of data from the sensor packages. Gaining confidence in machine learning can allow us to focus on reducing exposure to offshore personnel and drive towards a cloud-based infrastructure with remote operating stations and on-shore evaluations, ultimately offering a safer workplace along with cost beneficial project implementations.