Non-intrusive pipeline assessment for unpiggables, combining conventional approaches with data analytics

Author: Lews Barton, Service Manager – Non-Intrusive Pipeline Assessment, ROSEN Group

Corrosion and other time-dependent anomalies are some of the major threats operators face, costing millions annually in identification, mitigation, and repair. The situation is often even more complicated for “unpiggable” pipelines, which pose their own unique challenges.

ROSEN is known for its industry-leading inspection capabilities. Where In-Line Inspection (ILI) is not possible or impractical, ROSEN has developed the Non-Intrusive Pipeline Assessment (NIPA). The NIPA approach is built on industry-recognized Direct Assessment (DA) methodologies, enhanced by the addition of Advanced Data analytics and Large Standoff Magnetometry (LSM).

Developed and implemented by ROSEN since 2019, NIPA has been constantly evolving and improving to meet the challenges posed by pipelines operating in differing environments around the world. Now, it is further evolving to be deployed by Unmanned Aerial Vehicles (UAV) for inspection in the most challenging terrains and distances. The NIPA methodology integrates and overlays multiple pipeline datasets, such as construction and operational records, in conjunction with data obtained from multiple above-ground surveys to gain a holistic picture of the pipeline integrity condition. This process involves an assessment of pipeline condition along with a review of data collected to support the critical elements of an integrity management plan. The information can be cathodic protection performance, corrosion control measures, and the potential for pipe deformation due to ground movement or external interference, amongst others. The data overlay follows the classic “Reason’s Swiss Cheese model,” which reduces risk and uncertainty, limiting missed defects by overlaying many different data sources. Crucially, this allows operators to screen for and monitor critical integrity threats and prioritize locations to centimeter-scale GPS accuracy for field verification or remediation and repair, all without upsetting pipeline operation.

The combination of multiple datasets and ROSEN’s proprietary alignment and prioritization algorithms have been proven by direct field experience and verification. When combined with ROSEN’s comprehensive warehouse of pipeline inspection data, the Integrity Data Warehouse (IDW), this significantly improves confidence in anomaly identification and monitoring of problem areas. ROSEN is well-positioned to perform insightful data exploration and develop powerful predictive models relating to pipeline integrity. This is made possible due to ROSEN’s comprehensive IDW of pipeline inspection data. The IDW contains over 26,000 historical inspections across pipelines located globally and contains detailed pipeline information, including routes, product, manufacturer details, and pipeline defect features obtained through in-line inspection (ILI). The IDW has also been enriched with additional geospatial features that enhance the data by introducing contextual information that can be integrated into predictive models. Understanding the relationship between geo-enrichment variables and known sections of a pipeline that do or do not contain defects can further enhance our understanding of defect risk factors. We have introduced the following datasets into the IDW to create a data-driven external corrosion risk model:

• Geology: Geological period and mineral composition

• Soil: Type, material content, moisture and pH

• Groundwater: Depth and distance to the nearest water table

• Precipitation: Hourly precipitation data

• Elevation: Digital elevation maps (DEM) constructed from RADAR and LIDAR point cloud data

• Terrain Classification: Clustering of terrain into discrete definitions

• Historical Landslide Data: Distance and duration since nearby landslides

• Land Use Classification: Clustering of land use into discrete definitions

• Open Street Maps: Intersections with roads, railways, or waterways offer a truly unique proposition in the industry.

These variables expand the data horizon by considering not only the local survey results of the pipeline in question but also how every other pipeline identified in the IDW has behaved – which provides valuable insights for condition prediction. In summary, integrating a data-driven external corrosion risk model and NIPA into a DA process provides data to back up the expertise and opinions of pipeline integrity/corrosion subject matter experts. This strengthens the experts’ position and provides them with additional input that can be used when historical inspection or survey data is sparse or questioned. This is especially true in the case of pipelines where minimal anomalies may be present, as proving the absence of defects can often be more challenging than proving their existence!