Pipeline &
Gas Journal
March issue 2000:
DMR Intelligent Pig
Measuring Internal Corrosion
In Small Diameter, Heavy Wall Pipelines
By Peter A. J. van der Veer
and Sicco F. Jager,
Shell Global Solutions International B.V.,
Shell Research and Technology Centre Amsterdam,
The Netherlands,
Rainer Schmidt and Frits Bukman, 3P Services, Lingen, Germany
At the end of 1997, failures were observed in subsea flowlines in the central North Sea which raised concern as to the condition of other flowlines, in particular those lines with similar design and operating conditions. It was decided that an investigation of these lines should be conducted to assess their condition as soon as practically possible.
This posed a number of challenges in that the flowlines in question were constructed in a bundled design, making external access impossible, except at the towheads where the lines emerged. Internal access to the lines would be possible since pig traps were installed. The flowlines are between 10-12 kilometers long, have an internal diameter of 4 inches and a wall thickness of 12.7 millimeters. These dimensions posed a problem in that inline inspection using MFL or ultrasonic pig technology was not at a stage of development to measure such a large wall thickness in a small diameter line.
A failure within one of the flowlines however, would not be acceptable. Even a small hydrocarbon leak within the bundle could lead to an eventual over-pressurization and rupture of the bundle. Therefore it was decided to start an inspection program to reveal the current status of the flowlines.
Operator’s options
One of the inspection options available was the use of a tethered tool which consists of a self-propelled pig connected to a 6,000-meter umbilical which would transmit ultrasonic wall thickness measurements back to the platform. Although this method would give the required information on the pipeline condition, it was felt that this option was not viable on the grounds that it would require some major modifications to the topsides pig launching/receiving facilities and loss of production. It would also require extensive cleaning and temporary stationing of one or more containers on the platform.
A second option was to install alarming devices on the bundles that could give an indication of when the corrosion allowance was reached. This method was unacceptable as it would not give information on the current condition of the lines.
The final and preferred option was to run an intelligent pig. This would require no major topsides modification and would give a quick indication of the condition of the flowlines within the bundle. This option, however, posed problems in that current in-line inspection technology was not at a stage where it could measure the large wall thickness within small diameter flowlines. Three pipeline inspection companies were approached with the proposal to develop an inspection device that would be capable of supplying the desired information. All three indicated that they felt it possible to either enhance existing technology or to look at possible new methods.
Two companies approached Shell with devices that would supply the desired inspection parameters. A 12-meter section of spare flowline material was then shipped to Shell Global Solutions International (SGSI) in Amsterdam where predetermined corrosion pits with a ”typical” width of 25 mm and various lengths and depths were etched internally. The flowline was cut in two and given to both companies for pull testing. The third company approached Shell with a plan to re-design the magnetic component of their 4-inch MFL tool. If successful, the inspection would result in a normal MFL run, giving all necessary information for pipeline assessment.
Based on the quick response from 3P Services (one of the three companies) and the promising results from their pull tests, it was decided to perform additional pull testing at Shell’s premises in Amsterdam and, if successful, use their tool for the inspection.
Contractor’s Approach
3P Services has positive experience inspecting 4-inch offshore pipelines having a wall thickness of up to 8.6 mm (schedule 80), using MFL pig equipment. For the specific situation, this was not considered feasible, taking ID, WT and the bend radii into account.
Basic experiments using a different inspection technique were under way in June 1998, when Shell first contacted 3P Services regarding the requirements. It was quickly decided to complete a prototype inspection module to be applied in Shell’s test pipe joint. This prototype inspection module was combined with the well-proven 4-inch pig components (data logger, odometer, power supply, etc). The pull tests were carried out in early August 1998, immediately after the delivery of the test pipe to 3P Services’ facilities in Germany.
Shell wanted to go ahead. At a meeting in mid August, 3P Services pointed out that the sensor technology was far from being mature. Due to the urgency of the project, Shell accepted this fact, placed a contract with 3P Services and an ambitious time schedule was agreed upon.
Applied Technology
The inspection pigs apply a magnetic measuring technique called DMR (Direct Magnetic Response), which is sensitive to internal local metal loss only. The pigs as designed for the project carry 32 sensors each, distributed over the circumference at 10 mm spacing. The sensors are Hall probes that measure the response of pig-mounted magnets to an increase of its distance to the internal pipe surface, as it occurs in case of internal local corrosion. Each sensor has a rather limited angle of view and has - in this respect - some similarity to an ultrasonic probe. The measurement does not penetrate the pipe wall and is therefore independent from the wall thickness. The measurement is also independent of pig speed.
The sensors of the applied pigs are incorporated into a pig cup. The measurement therefore relates to the distance between sensor cup and pipe wall. If an area of internal metal loss is large enough, then the cup will simply follow the surface and - if sufficiently smooth - measure no change. Therefore, the applied 4-inch pig cannot detect gradual or even loss of wall thickness due to general corrosion. If necessary, the DMR-inspection may be supported by a geometry inspection that will detect an increase of the internal pipe diameter due to general corrosion.
Performance Test
Within less than four weeks, 3P Services completed two identical units of the inspection pig, ready to inspect approximately 12 km of offshore flowline. Meanwhile, the test pipe was shipped back to SGSI and additional defects were etched internally. Pull tests at SGSI were carried out in mid September, using both identical pigs. The tests proved that the inspection tool would certainly detect pitting type of corrosion with a diameter >10 mm. Depth sizing was not regarded to be very accurate, but it was expected that the approximate depth in a range from 1-5 mm could be established. Previous modeling and analysis had indicated that the maximum defect depth to be expected in the pipe to be less than 5 mm. Given this, the decision was made to continue and carry out a full progressive pigging program on one of the flowlines.
Inspection
Directly after the testing program, the pigs were shipped to the platform and a pipeline cleaning and testing program was performed. The pipeline was prepared by cooling it down to an acceptable level and running a number of mechanical pigs, starting with two foam pigs with different density. These pigs, with a cover of solid polyurethane, arrived disintegrated with large ruptures and parts of the cover lost. The runs were followed by a run with a long, multi-disc scraper pig which took out several parts of the previous foam pigs. Other abnormalities were not observed and the inspection runs were started.
The first pig run by 3P Services was the profile pig. This pig is multi-segmented, bi-directional and carried three aluminum gauge plates of different diameters (see Fig. 1). If the pig would get stuck, it could be pumped back. The profile pig arrived with only minor damage and 3P Services approved the pipeline for inspection runs.
Both inspection pigs were run (see Fig. 1) so that comparison between the data deriving from the two runs could be obtained. The pigs were launched and received without any problem. During the pig runs, the pumping rate was to be controlled at 0.5 m/s, which, however, could not be properly maintained by the available equipment. Attempts to adjust the speed during the profile pig run were unsuccessful and finally stopped. The two inspection runs were therefore carried out at a velocity of approximately 1 m/s, which did not create a problem. The only effect would be the axial sampling interval to be larger and therefore the axial resolution to be lower.
Results
Within two days after the inspection runs, 3P Services presented the first results of the inspection in the client’s office. At this stage, printed outputs of both pigs’ inspection data were submitted at different scales. The repeatability of the measurements could be demonstrated at different enlargements. The typical corrosion pattern downstream of the producing wells was shown. A preliminary analysis, done manually on several significant signals, indicated that the depth of the pittings was not serious. Some larger features were observed, but detailed analysis and comparison between the data of the two runs proved that the features originated from chips from the shell of the foam pigs. This conclusion was based on the ”depth” of the feature which corresponded with a retrieved part of the foam pig and the fact that most of the features were found further downstream during the second inspection run.
The detailed interpretation of the inspection data was performed later at 3P Services’ base in Germany. Due to the lack of experience in analyzing such data, the interpretation involved considerable development and parallel testing. This work—usually performed ahead of a field inspection—concluded in the description of the measuring range of the technology and the details regarding its capabilities and limitations, see also Fig. 2.
In the final report, several hundred indications for small diameter corrosion (length and width mostly <10 mm) with a depth less than 2 mm were reported.
Validation Of Inspection Results
Although the pull tests were manually evaluated in Amsterdam just after the testing, 3P Services reported the pull test data in more detail later, based on the experience gained during the field inspection analysis. From this analysis, it appeared that the tool detected and sized all defects, also the defects with an actual depth larger than 5 mm. Even pit depths up to 12 mm were reported, which indicates that the tool is capable of detecting and sizing larger defects. However, the preliminary specifications of the tool allowed for sizing of defects up to a depth of 5 mm. Sizing of defects with an actual depth > 5 mm should therefore not be used for statistical analysis of the results. It clearly indicated, however, that the tool will find and measure defects with a depth larger than 5 mm if present, but with a rather low accuracy. Since from the inspection of the lines no defects with a depth larger than 2 mm were reported, it was concluded that certainly no defects exceeding 5 mm would be present and that the reported inspection results could be accepted.
More data need to be generated for determining probability of detection and detection confidence levels. It is anticipated that through additional research and technological optimization, the tool measuring capabilities can be enhanced.
Conclusion
The desired targets were achieved. The inspection revealed spots with metal loss damage in the flowline which was well within the corrosion allowance of 5 mm and the line was regarded fit for operation.
The development program was successful in that the new tool worked properly and the reported results could be validated, based on pull tests and the correlation of data collected by two identical tools. The new intelligent pig was found to be capable of detecting internal, localized type of corrosion defects. P&GJ