Condensate cripples compressor

Jan. 5, 2006
Readers offer tips on how to keep a standby compressor ready for service.


We have two parallel centrifugal compressors that deliver motive gas to an ejector system that has two parallel ejectors. The ejectors extract fines from a cyclone to prevent cyclone blockage. One compressor operates at all times, with the other in standby. We have been getting condensation in the standby compressor due to cooling some of the heavy material in the motive gas. The liquids cause the compressor to fail on start-up, defeating the entire point of having that unit. One idea we have is to allow some discharge gas from the on-line compressor recycle back through the standby compressor to keep it hot. This would lead to reverse flow in the standby machine. Is this okay? Or would some type of auto-drain, slow-roll operation, or aggressive heat tracing work better? Does anyone have experience with these ideas or with other solutions?

Install a drip leg drain
Putting discharge gas from the operating compressor through the standby compressor is not recommended. Reverse gas flow is likely to cause the standby compressor to rotate in the reverse direction without control. Most centrifugal compressors are not designed for reverse rotation and operating in reverse rotation could damage the compressor.

If the compressors have down-oriented nozzles and have slow roll capability, a drip leg drain could be added to the discharge line together with level monitoring instrumentation. The compressor could be drained either manually or automatically depending on the instrumentation. If the compressors have up-oriented nozzles and casing drains, the drains can be manifolded to a drip leg using the method described for the down nozzles. If none of the methods mentioned are feasible, the addition of aggressive heat tracing can be considered.

Royce Brown, principal,
RNB Engineering, Houston

Pipe in a recycle line
Not knowing the details of your system, I will make some assumptions:
1. Common suction and discharge lines.
2. Automatic valves on suction and discharge that open/close when the standby compressor is needed.
Also, since you said you could slow roll, I assume you have a steam-turbine driver or VSD motor.

1. Do not run compressor train in reverse.
2. Simply pipe a small recycle line from the discharge of both compressors to the suction of the other compressors.
3. Locate the tie-ins for the recycle lines between the automatic isolation valves and the compressors.
4. Size the line to keep a minimum flow to prevent condensation. Use a
restricting orifice or a valve that could be adjusted while running to achieve desired results. If you want to get fancy, add an automatic valve with a temperature sensor to minimize recycle.
5. If not already done, insulate both compressors. Insulation blankets and thermal shrouds can be easily obtained.
6. If this results in the standby compressor slow rolling, all the better for start up when needed; this also should help prevent a bow in the shaft from occurring. It will be turning in the correct direction for operation.
7. Have isolation valves in the recycle line so that maintenance can be performed on one compressor while the other compressor runs.
8. With the arrangement described above, either compressor that is running will be able to recycle a small flow back through the suction of the other compressor on standby and back to the suction of the compressor running.
9. This system will use some energy because of the recycle, but should provide much greater reliability and lower maintenance cost.
10. Either compressor could be operated with the other compressor on standby with this system.

George Hogg, maintenance/reliability manager
El Dorado Chemicals, El Dorado, Ark.

Try a condensate trap
Could the fix be as simple as a condensate trap — like one that would be used on a steam system?

Jason Englund, instrument tech II
Amgen, Bothell, Wash.

What about condensing heavies?
It would help to know what the “motive gas” is that the compressors boost to get the ejectors extracting. Guessing by your description, the motive gas might be a combination of solvent vapors with the “heavies” condensing at a cold spot in the system — the idle compressor. Are there any other factors such as environmental, product run material?

Online compressor recycle. We have used simple single-valve, manual kickbacks on spare, parallel liquid centrifugal pumps for many years as standard installations for freeze protection.
We have also considered an “on-line compressor recycle” as you noted in an effort to eliminate condensed ammonia gas at an idle, parallel reciprocating compressor. With our reciprocating compressor, a small-orifice solenoid valve would be installed on existing ports in the compressor head to allow discharge gases from the online compressor to back-flow across the idle compressor’s internal discharge reed valve and its internal suction reed valve. This hot-gas kickback solenoid valve would fail open when the compressor is off and timer-close a few seconds after the compressor starts (also serving as an unloading valve to lower startup costs). Our volume of kickback would be very small and within heat-traced and insulated piping.

There may be other considerations with your larger centrifugal compressor such as a much larger, inefficient kickback to achieve heating across all dead spots, possible seal damage due to condensation/seepage and safety issues relating to isolation and clearing. This approach would require a detailed “Management of Change” per PSM. Finally, any warranty might be voided on the compressor itself.

Heat tracing and insulation. If the original installation was heat-traced and insulated, then it seems logical to improve coverage to eliminate cold spots.

A different approach — condensing heavies. If the motive gas and ejectors are part of an overall closed loop (as opposed to makeup air going in and finally exhausted to atmosphere), then perhaps the heavies need to be condensed and “purged” before the inlet of both of the compressors. If the motive gas composition has changed over time and is at dew point in an idle compressor, maybe this aspect needs to be evaluated.

Hopefully the condensed liquids can be recycled to recover costs. If the heavies have to be eliminated at some point anyway, then perhaps the compressor inlet is the place? The inlet cooling and liquids separation should also improve overall compressor efficiency and maintenance.

Tobin D. Kueper, acting plant manager
CYANCO, Winnemucca, Nev.

Add a drip tank
A simple drip tank with an integral demister chamber and an automatic drain valve installed between the discharge of the standby compressor and the tie point of the discharge from the primary compressor should completely eliminate any condensation from entering the standby compressor. It should be sized so that pressure drop is not a concern.

Sean Taylor, mechanical installation manager
Mueller Field Operations Inc., Springfield, Mo.

Consider gas re-routing issues
Has the compressor failure on start-up been proven to be due to condensed gases, that is, to liquids in the casing? If not, some more work should be done to confirm why the compressor is failing. Is it possible the failure mode may be related to the fact that the compressor on standby is cold and then gets hit by hot gases on start up — particularly, the rotor thermally growing faster than the casing and losing clearance? This latter point may be a contributing factor of the compressor failing.

If the failure mode has been narrowed down to “condensation,” then simply have the field operators drain the compressor casing on their rounds a couple times per shift (or as needed). Of course, this assumes the compressor can get this type of attention on a daily basis.

If the failure mode is due to additional factors like cold start up and thermal issues, then the solution may be the re-routing of some of the discharge gases off the main compressor and back through the standby compressor; the benefit of the re-routed gas will be to keep the standby compressor hot and (hopefully) liquid free.

If hot gas is to be rerouted back through the standby compressor, the following issues need to be addressed:

1. Can fines from the reverse gas “lay down” in the standby compressor, creating a new problem? The re-routed gas should come off after the ejectors to minimize fines lay down.
2. Reverse rotation will (likely) need to be avoided. The equipment manufacturer should be contacted to confirm this, as should the seal manufacturer (many seals only rotate in one direction). Reverse rotation of the compressor can be prevented with either a locking device that permits rotation in the “allowable” direction or the reverse flow can be restricted so as not to overcome the machine stationary momentum, but sufficient to keep things hot.
3. I assume there is a check valve in the piping system of the standby compressor (if only to avoid reverse rotation); this will need to be bypassed (obviously) depending on location in order to get reverse flow.
4. Bearing lubrication may be an issue. Are the bearings force-fed? Do they receive oil only when the compressor is running? If so, they may need to be fed while on standby to prevent heat damage to the bearings.
5. Are there other ancillary systems on the compressor that don’t run when the standby compressor is not running? For instance, is buffer gas/oil to the seals off when the compressor is on standby? Do these systems now need to be running when the compressor is not?
6. A complete review of how the compressor is intended to operate versus how it will now operate should be completed, including a review of the start-up procedure for the standby compressor, which should include draining the casing and checking to ensure it is free to rotate; obviously the startup procedures are irrelevant if the standby compressor auto-starts, but can be made into checks (standard operating procedures) as part of the daily field operators’ rounds.

Conrad J. Horvath, P.E., rotating machinery engineer
Syncrude Canada Ltd., Fort McMurray, Alberta

Install a sump
Your best bet is to install a sump at the lowest point of the pipe with a drain valve located at the bottom of the sump actuated by a level switch. Interlock the valve to the compressor motor so that if the level in the sump is high or the drain valve is open, the compressor cannot start. Include a position switch on the valve so that the operator in the control room can see the on/off position of the valve.

A tee can create the sump: The compressor discharge line begins with a tee. The top of the tee goes to the destination. The bottom is extended to about 5-ft long. This section serves as the sump with a level switch and a drain valve. The switch opens the drain valve on high level and closes the valve on low level.

Dilip K. Das, P. E., principal engineer (pressure safety specialist)
Bayer CropScience, Kansas City, Mo.








A fuel-gas pipeline in a pipe trench went under a road bridge inside a plant. A flanged joint fitted with a compressed asbestos fiber (CAF) gasket was located only 2.5 m (8.2 ft) from the roadway (Figure 1). According to the area classification code used by the company, there was a Division 2 area for a radius of 3 m (9.8 ft) around a CAF joint and road vehicles should not be allowed unrestricted access to Division 2 areas. However, the plant wanted to open the bridge to unrestricted traffic. Two courses of action were suggested:

1. Leave the gasket alone. The operating staff wanted to ignore the code for once. The joint had been there for many years. It had never leaked. It was more likely to leak if it were disturbed rather than if it were left alone. Replacing the gasket with another type would be a hazardous operation because a long length of pipeline would have to be emptied and made safe. The code is for guidance; it is not law, they argued.

2. Replace the gasket. Others countered that leaving the gasket alone would give everyone the impression that the plant disregarded its safety codes as soon as they became inconvenient. These people recommended replacing the CAF gasket with a spiral-wound one. This would reduce the radius of the Division 2 area to 1 m (3.3 ft) and thus allow unrestricted traffic.

Would you go with either of these options or push for a different one?

Send us your comments, suggestions or solutions for this question by May 30. We’ll include as many of them as possible in the July 2005 issue. Send visuals, too — a sketch is fine. E-mail us at [email protected] or mail to ProcessPuzzler, Chemical Processing, 555 W. Pierce Rd., Suite 301, Itasca, IL 60143. Fax: (630) 467-1120. Please include your name, title, location and company affiliation in the response.

And, of course, if you have a process problem you’d like to pose to our readers, send it along and we’ll be pleased to consider it for publication.