What Went Wrong Case #24: What Caused The Cylinder Gas Leak?

No company, product, or individual is immune from experiencing failures or shortcomings. The solution to these problems can often result in new “best practices.” By relating some of these real case histories, it is hoped that the lessons learned will be educational to the readers. In each case, a failure or incident is described, and the question is asked, “What went wrong?” In each case, the answer (or at least the best speculation) and the solution that was applied to the problem will be explained.

“What Went Wrong” is published monthly in Gas Compression Magazine.


Case #24: What Caused The Cylinder Gas Leak?


The first of several single-stage propane refrigeration compressors were being commissioned after installing new cylinders. As the compressor and piping inside the system between the suction and discharge isolation valves was being pressurized with nitrogen to check for leaks, a significant leak was observed from the piston rod packing on all four compressor cylinders. Although most piston rod packing is not leak tight when the compressor is not operating, the magnitude of the leak was excessive.

Maintenance personnel determined that most of the leakage was not coming out along the piston rod, but seemed to be from around the packing case itself, or, more likely, from the packing case gasket. The torque on the four nuts retaining the packing case was checked and found to be reasonably close to specification. Increasing the torque on the nuts by 20% slowed the leak only a small amount. With no chance that the packing case seal was safe for use with propane, the commissioning team and the operator asked the question, “What went wrong?”

The cylinder and packing manufacturers were both consulted and sent representatives to the site to investigate the gasket issue and the leakage. A CAD model of the 300 psig (20.7 bar) MAWP, 21.63 in. (549 mm) x 14 in. (356 mm) stroke, cylinder and a cross-section of the packing case assembly are shown in Figure 1. The packing case was secured by four 1 in. (25.4 mm) x 8 UNC studs with nuts that were tightened to provide 30,000 psi (206.8 MPa) prestress on the studs. The inner end of the packing case was sealed against the sealing face in a recess in the wall of the cylinder by a round wire metal gasket. The recess and gasket sealing faces are shown in the section of the crank end of the cylinder in Figure 2.

Removal of a packing case from one of the cylinders immediately revealed the first smoking gun that answered the question, “What went wrong?”! Some unusual deformation was discovered on the … Click Here To Continue Reading This Article In The December Issue.

PLEASE NOTE: There is no cost whatsoever to continue reading the article. By continuing to the digital issue, you might find even more articles that interests you. In addition, our visitor and pageview numbers go up, something we love to see! A win-win for both of us!

Lastly, if you’re not a subscriber, CLICK HERE to sign up for free monthly delivery. You choose print, digital, or both.

Previous article2022 Year In Review: Delivering On Promises
Next articleResults From The 2022 Gas Compression Industry Survey