Flowline Detection/Inspection Technologies

Acknowledgements

Project Advisor: Emily Sievers

Technical Advisor: Emmanuel De Moore

Donations Made by: Mines Petroleum Engineering

Video

Elevator Pitch

Team Go With The Flow was asked to address the issue presented by unmapped or improperly abandoned subsurface flowlines that may be damaged and present a danger to their surroundings, such as the tragic accident that occured in Firestone, CO in 2017, where two people died from the gas leaking out of an improperly abandoned flowline which caused an explosion. The team’s goal was to develop a noninvasive methodology to detect these flowlines, obtain qualitative data about their condition, and incorporate this into a modular system that utilizes drone technology to inspect large sites.

The group designed a sensor system that utilizes a triaxial magnetometer in order to begin achieving these goals. Our simulation testing bed allows for easy system testing with different sensor positions, pipeline failures, pipeline depths, soil types, and pipelines with and without current. The group aims to test the magnetometer readings while changing these variables to validate its potential effectiveness. With this proof of concept, the group hopes to hand off this research to future Senior Design teams with the intent that this system might be implemented into a commercial product for real world applications.

Design Approach

Design Solution

Each of the experiments were conducted using the constructed testbed and the FG-33+ sensor. The plexiglass table top was designed to have 3-inch spacings that guide the position of the sensor as each experiment is carried out. We took slices of magnetometer readings with a set number of samples so that the readings can be plotted to their XY-coordinates after each experiment is done.

In order to place the pipe in the testbed and imitate the conditions of each flowline failure type, the steel pipes needed to be cut to be 2-feet long and machined to a respective failure mechanism mentioned below. A 2-foot section of control pipe was also used to compare with each experiment. A list of how each condition was satisfied is shown below:

• Pitting condition: drilled a ¾ inch hole in the 2-foot section of steel pipe. 
• Line strike condition:  milled a 4-inch groove in the middle of a section of 2-foot steel pipe. 
• Wall loss condition: lathed at 20%, 40%, and 80% wall loss in the middle of a section of 2-foot steel pipe.
• Current Test: Used a DC power supply to send current through the pipe at 50 mA, 500 mA, 1 A, and 5 A.
• Soil Test: Used samples of soil types found in Colorado in a smaller scale testbed to test the effects of soil type.

A total of 55 experiments were conducted, including the experiments that were conducted multiple times to verify results. 

After the testing with the 2-foot section of pipe was completed, it was determined that the results had some complicating effects due to the small nature of our experimental setup. The team could not claim results to be proof of concept, but the results seemed promising enough to encourage future experimentation. It was difficult to determine the effects of some failures on the magnetometer readings for most of the failure mechanisms, but the results for the pitting condition turned out to be very promising as the effects on the magnetic field of the pipe seems large and easily detectable. This in addition to the fact that companies are already using similar magnetic sensors to detect flowlines already leads to the team foreseeing positive next steps for the project.

Next Steps

In order to continue this project, large scale testing needs to be done using this team’s findings as a solid foundation from which future groups can work off of. We ended up finding a lot of useful information, and with extra time, we could find valuable information which could possibly be incorporated into a commercial system.

It is our hope that our client moves forward with continuing our research, as we feel there is a lot of potential value to be had in the system we worked towards. We may not have had enough time to do it ourselves, but we feel that future work would result in a valid proof of concept which could eventually be transformed into a successful commercial product.

Meet the Team