Clear Up a Compressor Conundrum

Readers decipher a compressor surge.

1 of 2 < 1 | 2 View on one page

Our unit superintendent is concerned about a multistage centrifugal gas compressor. It’s fed by a separator and discharges to an absorber; there’s a pressure control valve at the compressor suction and a flow control valve in a recycle line from the compressor discharge. Ever since the first startup after the last turnaround, the motor current unexpectedly surges during startups and other upset conditions before settling into generally steady service. The packing in the absorber was replaced during that turnaround. What could be causing the problem? Should the superintendent be concerned? Can we do anything so we can safely limp through until next year’s turnaround?

The following steps may be taken to avoid unwanted surging: 1) before starting the compressor close the suction and discharge valve and open the bypass/recycle valve so that there will be the same load; 2) purge the system before starting the compressor, unloading it to minimize motor current; 3) start the motor ? it will draw maximum current at the time of starting [compared to regular operation], which is a normal phenomenon; 4) after proper lubrication, the gas compressor should be loaded slowly in stages, i.e., 25%, 50% and 100%, allowing some stabilization time. This procedure will take care of unwanted surging and frequent tripping of motor. The motor may be diagnosed for desired current output and leakages, etc.
C. M. Pakhale, superintendent
Oil India Ltd., Duliajan, India

I suspect the compressor is running off its curve and surging. The new packing in the absorber has less pressure drop than the old packing. I would consider adding pressure drop between the compressor and absorber via a manual valve partially closed or installing a flow orifice. There are no clues in the problem to suggest a dynamics/tuning problem with the control valve on the suction.
Mike Gentilcore, principal research engineer
Covidien, Hazelwood, Mo.

This surge may be because of high resistance on the discharge side. While starting the compressor keep flow valve on discharge line throttled full open — more than what you normally would. As far as column packing is concerned, I suggest checking for the column pressure profile before and after the packing is changed. If the pressure drop is more than it should be, record it for future changes. But as I understand the problem was there even before the change of packing, so this particular the problem may be the opening of the valve. Maybe PID setting change can be accommodated. It may happen the pressure control valve is undersized.
Surender Khtri, process engineer
L&T Engineering - E&C Division, Faridabad, India

There are two solution based on following two conditions:
Condition 1: The packing replaced is not permeable [e.g., too small, wrong type], creating a vacuum on the suction side of the compressor. In turn, this increases the motor current draw. Eventually, the packing will be destroyed, thus allowing normal air flow, bringing motor current within limits after some time.
Solution 1: Check the differential pressure between absorber chamber and the discharge line connecting to compressor. Pressure inside the chamber should be high when compressor is “off” and should reduce little when the compressor is “on.” Open the absorber to check the packing.

Condition 2: The initial starting torque current of motor is much higher than running-condition current.
Solution 2: Consider using a variable frequency drive; look at the settings available in drives at low frequency and low voltage. The torque can be increased and it will conserve energy. Also, the smooth operation of drive will absorb the sudden surge current. Do one check, though. Verify that the outlet valve of discharge is open 100% and inlet valve [the pressure regulator] opens gradually.

1 of 2 < 1 | 2 View on one page
Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.


No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments