Very low flow rates pose a challenge to gas pipeline pigging due to lack of pushing force for pig movement. However, if a supply source with a high enough pressure is available, a process called "Dynamic Pushing" may be used to create sufficient velocities for the pigging operation. This paper explains the concept, demonstrates how to simulate the process using a transient hydraulic computer model and finally presents a real case of pigging a six mile 36" gas pipeline using this concept with the guidance of the computer model.
A normal pigging operation in a gas pipeline relies on the gas flow in the pipe to provide the pressure differential ("pushing") needed to move the pig. There is usually a required minimum speed for the "smart pig" (a computerized mapping tool) to work properly. Sometimes, due to various reasons the maximum flow rate through a pipeline is so low that it is impossible to run a "smart pig" through the line at the speed necessary to obtain valid data. In this case, if a sufficiently large source of high pressure gas is available, a "Dynamic Pushing" process may be used to create a higher flow rate to make the pigging possible. The basic concept is: first, lower the pressure in the pipeline segment to be pigged as far as possible to create a low linepack zone; second, launch the pig into this pipeline segment and apply higher pressure behind the pig to move it at a desired speed. With limited demand flow at the downstream end of the segment, moving of the pig will result in the pressure ahead of the pig to continue to increase. Overall the process will raise the pressure of the pipeline being pigged to rebuild the linepack which was purposely lowered at the first step. This concept is demonstrated in Figures 1 and 2.
The Dynamic Pushing process can be simulated using transient hydraulic computer software. In our case, commercially available simulation software has been used for this task. At the first glance the problem appears to be complex in that it involves a moving boundary of pressure across the pig. However if we assume this pressure differential across the pig is small (say, a few psig) and the pipeline is of a relatively large size (36" for example) and short (a few miles only) we can treat the average pressure of the pipeline through the entire length as a single variable, and build the model using the configuration shown in Figure 3. For steady state (initial state of the transient model) the outlet node is set to Known Flow with a flow value assigned. We are now ready to "launch the pig" and start the transient simulation. The minute-by-minute simulation results are presented in the following figures for pigging 6.2 miles of 36" pipeline starting at 255 psig. Figure 5 shows the simulated average pressure change with time.
