Robotic Inspection System for Unpiggable Pipelines

Abstract

Much of the existing natural gas infrastructure was designed and built without pigging as an operational consideration. There are many physical obstacles in pipelines that make the passage of SMART pigs impossible. The most intractable obstacles include: • Elbows with tight bend radius. • Back to back combinations of elbows. • Partially ported values. • Reductions/expansions greater than two pipe sizes. The use of pigs is totally dependent on the availability of pressure to “push” the pig through the pipeline. Unfortunately, the operation of many utility owned transmission pipelines is at a pressure too low to support the operation of a conventional pig. Although most interstate pipelines are many miles long, many high consequence areas along transmission pipelines are usually extremely short. Many of these pipeline segments are only one to two miles in length with no installed local traps. With the advances in robotics and sensor technology, the Office of Pipeline Safety has recently endorsed the concept that all oil and gas transmission pipelines should be capable of 100 percent inspection. The cost to replace just unpiggable valves and sharp bends has been estimated at over $1.5 billion (gas only). Therefore, the ability to inspect unpiggable pipelines presents a formidable technical and financial challenge. The inspection of unpiggable pipelines requires the marriage of a highly agile robotic platform with NDE sensor technology operating as an autonomous system. Foster-Miller and PII are developing a robot that is essentially a battery powered, train-like platform. Both front and rear tractors propel the train in either the downstream or upstream direction. Like a train, the platform includes additional “cars” to carry the required payloads. The cars are used for various purposes including the NDE sensor module(s), the power supply, and data acquisition/storage components. The onboard distributed intelligence gives the platform the capability of an engineer steering the train through the complex pipe geometry. The robot is designed with a slender aspect ratio and the ability to change shape as required by the physical obstacle presenting itself. The MFL sensor module must also morph itself through the physical obstacles, and thus, will require some level of segmentation. The system requires a very simple launch and retrieval station that is significantly less expensive to deploy.