Illustrated Case Studies in the Maintenance Reliability Engineering World of Failure Analysis, Predictive Maintenance, and Non Destructive Evaluation |
When designing gas compression systems, if the lubricating oil comes into contact with the gas at an elevated temperature and pressure then gas entrainment must be considered. When designing gas compression systems, if the lubricating oil comes into contact with the gas at an elevated temperature and pressure then gas entrainment must be considered. The system shown in the figure on the left (i.e. figure 1) compresses a gas. The gas compressor does not have a closed loop lubrication system that is isolated from the gas. Instead, the oil mixes with the gas as it is being compressed. Therefore, the oil is separated from the gas in the large separator vessel, shown in the middle of figure 1, immediately after it leaves the compressor. The separated oil is accumulated in the bottom of this vessel from where the lube oil pumps take suction. The gas compressed in this compressor was considered to be an inert gas with respect to the lubricant oil in that it neither would dilute the oil, like a hydrocarbon gas would do, nor react with the oil, like a chemically active gas would do. If it did react like a hydrocarbon gas then the oil viscosity might be reduced and a heavier oil would have to be used. Therefore, chemically speaking, the gas system shown in figure 1 should not have adversely affected the lubricant, or so one would have thought One might have thought that simply separating the oil from the gas would be the only thing of concern here. It was the only thing of concern, but it was taken more lightly than it should have been taken. The key factor behind separating oil from gas was not just the act of the separation itself, but also the time allotted to the process so that any gas entrained during the separation process was allowed to also separate from the oil. This can be the Achilles heel of any oil-gas separation system, as was the case here. One of the key phrases in separation systems is "residence time." Simply put, this is the time required for the gas bubbles to work their way out of the oil before the oil pump sucks the gas bubbles in with the oil. |
A close up view of the separator vessel is shown in the figure on the left (i.e. figure 2). The discharge from the gas compressor entered the separator at "A". Most of the bulk oil in the gas stream separated immediately as it entered the vessel at this location. A coalescing filter located at point "B" separated any finer droplets that were carried upwards by the gas stream. The cleaned gas stream entered the upper chamber of the vessel, exited at the top, and was carried downstream to its final destination. In the meantime, the coalesced oil droplets accumulated on the filter and dropped back into the bottom chamber of the vessel. The bottom of the separator also acted as the oil reservoir for the lubrication system. The pressure on this system was equivalent to the discharge pressure of the compressor. Therefore, any entrained gas bubbles that remained in the oil were at discharge pressure. The working level of the reservoir was between points "C" and "D". The oil pump suction was at the level marked "E". The theoretical minimum residence time of the oil was the time it took an imaginary oil "droplet" to go from point "D" to point "E". In actuality, the real residence time was less than that because the vortex action created by the pump suction grabbed the oil somewhere between "D" and "E". Depending upon the pipe design and placement, that distance could have been anywhere from one-third to one-half the distance above the suction inlet. It is plain to see that if the oil level was low at point "D" then there would be very little residence time before the pump suction pulled the entrained gas into the suction line at "E". The danger behind having an entrained pressurized gas in the oil is that it can and will influence the oil film. As an analogy, consider what happens when you open a bottle of soda pop. Before opening the bottle there are no visible signs of bubbles, but after opening the bottle the bubbles appear. Such a simple event brings joy to the drinker, but in a compressor it wreaks havoc. The bubbles appear in the soda pop for the same reason they might appear in the oil. The bubbles appear as the gas-liquid mixture is decompressed with the opening of the bottle. The volume of the bubbles increase and the bubbles expand. In a compressor, if the oil-gas mixture is fed to a bearing at a pressure lower than the discharge pressure of the compressor then the entrained gas will expand and interfere with the hydrodynamic properties of the oil film. In essence the film of oil will be sponge-like. |
Oil film thickness is designed for the loads encountered. The thickness is based upon a solid film of oil and not a spongy perforated film of oil. In the case of the gas compressor, shown in the figure to the lower left (i.e. figure 3), the oil entered the compressor at the two locations marked "A" and "B". The low-pressure side was at point "A". The pressure was less than one atmosphere. The high-pressure side was at point "B". The pressure here was equal to the discharge pressure of the compressor. The compressor was a double axial-screw design. Each end of each screw was supported by sleeve bearings. The oil to the bearings came in at points "A" and "B". In turn, the oil from the bearings drained into the pressurized gas chamber. The gas chamber was at less then one atmosphere at point "A" and was at discharge pressure at point "B". Very early in the life of this compressor the sleeve bearings at point "A" were consistently failing and the shafts were scoring. The bearings at point "B" were doing fine. The correct viscosity oil was being used, as recommended by the compressor OEM. The OEM insisted that the gas was not dissolved in the oil and creating the problem. They were probably correct. However, the OEM never considered gas entrainment. To investigate such a possibility, a 1/4" clear plastic bleed line was put on the supply header that fed points "A" and "B" shown in Figure 3. The bleed line spilled into an open bucket. Opening the bleed valve for the 1/4" line revealed gas bubbles entrained in the oil going to the compressor bearings. The gas bubbles can be seen in Figure 4. When the oil was momentarily shut off, the bubbles expanded and the result can be seen in Figure 5. There was no longer any doubt that entrained compressed gas was in the oil supply going to the compressor bearings. |
![]() |
Figure 4 |
![]() |
Figure 5 |
But if gas entrainment was the issue, why then weren't both sets of bearings at "A" and "B" failing? The answer was the soda pop analogy. The entrained gas in the oil which was compressed to discharge pressure entered the bearings at "A"; it decompressed and expanded. The oil film became spongy and could not support the bearing loads. The entrained gas in the oil entering the bearings at "B" remained compressedbecause it was essentially at the same pressure; it did not expand, and the oil film remained relatively solid and stable.
The solution to this problem was to go to a heavier viscosity oil. This action was taken 12 years ago and the compressor has run fine without incident ever since. Speaking in today's terms, is there still entrained gas in the system? Yes. Is the oil film on the low-pressure side bearings spongy? Yes. If we could not have gone to a heavier oil then what could have been done to eliminate the entrained gas? Designing an oil-gas separator with a significantly longer residence time for the oil would have reduced and/or eliminated the entrained gas problem. |
All Pictures and Text Copyright © 2002 - 2016 Contact Mr. Adler Adler Engineering LLC of Wyoming USA
Great care has been taken in the compilation of this article. However, no warranty, expressed or implied, including without limitation, warranties of merchantability or fitness for a particular purpose, are given in connection with this article or any article archived on this website. Although this information is believed to be accurate by the author, the author cannot guarantee favorable results will be obtained from the use of this article alone. This article is intended for use by persons at their sole discretion and risk. Since the conditions of product or material use are outside of the author's control, the author assumes no liability or obligation in connection with any use of this information. The author is not liable for special, indirect or consequential damages resulting from the use of this material.
No part of this article or any article archived in this website, or any part thereof, may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, without the prior written permission of the copyright holder R. H. Adler. Nothing contained in this article or any article archived in this website shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters of patent, copyright, or trademark, and nothing contained in this article or any article archived in this website, shall be construed as a defense against any alleged infringement of letters of patent, copyright, or trademark, or as a defense against any liability for such infringement. |