0
Research Papers: Gas Turbines: Controls, Diagnostics, and Instrumentation

A Data-Driven Approach for Condition Monitoring of Reciprocating Compressor Valves

[+] Author and Article Information
Christopher J. Guerra

Product Development Engineer
Research and Development,
Dresser-Rand Company,
Olean, NY 14760
e-mail: cguerra@dresser-rand.com

Jason R. Kolodziej

Assistant Professor
Mem. ASME
Department of Mechanical Engineering,
Rochester Institute of Technology,
Rochester, NY 14623
e-mail: jrkeme@rit.edu

1Corresponding author.

Contributed by the Controls, Diagnostics and Instrumentation Committee of ASME for publication in the JOURNAL OF ENGINEERING FOR GAS TURBINES AND POWER. Manuscript received August 9, 2013; final manuscript received November 4, 2013; published online December 10, 2013. Assoc. Editor: Klaus Brun.

J. Eng. Gas Turbines Power 136(4), 041601 (Dec 10, 2013) (13 pages) Paper No: GTP-13-1300; doi: 10.1115/1.4025944 History: Received August 09, 2013; Revised November 04, 2013

This paper focuses on condition-monitoring of three different valve failure modes common in reciprocating compressors. They are missing valve poppets, valve spring fatigue, and valve seat wear. First, a targeted instrumentation study is performed on a Dresser–Rand ESH-1 industrial reciprocating compressor to investigate detection methods for these failures. This is followed by the development of a novel health classification methodology based on frequency analysis and Bayes theorem. The method is shown to successfully classify the condition of the valves to a high degree of accuracy when applied to actual seeded valve faults in the compressor.

Copyright © 2014 by ASME
Your Session has timed out. Please sign back in to continue.

References

Fagundes Schirmer, A. G., Gernades, N. F., and De Caux, J. E., 2004, “On-Line Monitoring of Reciprocating Compressors,” NPRA Maintenance Conference, San Antonio, TX, May 25–28.
Motriuk, R. W., 1996, “Reciprocating Compressor Valve Failure—Digital Modelling and Analysis,” 1st International Pipeline Conference (IPC'96), Calgary, AB, Canada, June 9–13, pp. 993–1002.
Metcalf, J. R., and Woollatt, D., 1995, “Reciprocating Compressor Valve Reliability Improvements,” 24th Turbomachinery Symposium, College Station, TX, September 25–28, pp. 167–173.
Matsumura, M., Kato, M., and Hirata, T., 1992, “Behavior and Analysis of Reciprocating Compressor Valve,” KOBELCO Technology Review, 14, pp. 20–24.
Bloch, H. P., 2006, A Practical Guide to Compressor Technology, Wiley, New York.
Sela, U., 2000, “Reciprocating Compressor Condition Monitoring,” Hydrocarbon Process., 79(2), pp. 59–62.
Schultheis, S. M., Lickteig, C. A., and Parchewsky, R., 2007, “Reciprocating Compressor Condition Monitoring,” 36th Turbomachinery Symposium, College Station, TX, September 10–13, pp. 107–113.
Woollatt, D., 1993, “Factors Affecting Reciprocating Compressor Performance,” Hydrocarbon Process., 72(6), pp. 57–64.
Ahmed, M., Gu, F., and Ball, A., 2011, “Feature Selection and Fault Classification of Reciprocating Compressors Using a Genetic Algorithm and a Probabilistic Neural Network,” 9th International Conference on Damage Assessment of Structures (DAMAS 2011), Oxford, UK, July 11–13.
Ahmed, M., Gu, F., and Ball, A., 2011, “Fault Classification of Reciprocating Compressor Based on Neural Networks and Support Vector Machines,” 17th International Conference on Automation & Computing, Huddersfield, UK, September 10, pp. 213–218.
Manepatil, S., Yadava, G. S., and Nakra, B. C., 2000, “Modelling and Computer Simulation of Reciprocating Compressor With Faults,” J. Inst. Eng. (India): Mech. Eng. Division, 81, pp. 108–116.
Manepatil, S. S., and Tiwari, A., 2006, “Fault Diagnosis of Reciprocating Compressor Using Pressure Pulsations,” International Compressor Engineering Conference at Purdue, West Lafayette, IN, July 17–20.
Elhaj, M., Gu, F., Ball, A. D., Albarbar, A., Al-Qattan, M., and Naid, A., 2008, “Numerical Simulation and Experimental Study of a Two-Stage Reciprocating Compressor for Condition Monitoring,” Mech. Syst. Signal Process.22(2), pp. 374–389. [CrossRef]
Guerra, C. J., and Kolodziej, J. R., 2013, “A Validated System-Level Thermodynamic Model of a Reciprocating Compressor With Application to Valve Condition Monitoring,” ASME Dynamic Systems and Control Conference, Palo Alto, CA, October 21–23.
Holzenkamp, M., Kolodziej, J. R., Boedo, S., and Delmotte, S., 2013, “An Experimentally Validated Model for Reciprocating Compressor Main Bearings for Applications in Health Monitoring,” ASME Dynamic Systems and Control Conference, Palo Alto, CA, October 21–23.

Figures

Grahic Jump Location
Fig. 1

Dresser–Rand ESH-1 reciprocating compressor @ RIT (3,628–kg, 3 m × 1.8 m)

Grahic Jump Location
Fig. 2

(a) Typical cut-away of a reciprocating compressor; and (b) side view of the ESH-1 compressor at RIT

Grahic Jump Location
Fig. 3

(a) ESH-1 suction valve assembly; and (b) discharge valve assembly

Grahic Jump Location
Fig. 4

ESH-1 poppet valve with springs. Left to right: two poppets, nominal spring, and degraded spring.

Grahic Jump Location
Fig. 5

Cut-away view of the compressor's poppet valve body

Grahic Jump Location
Fig. 6

Thermocouple locations

Grahic Jump Location
Fig. 7

Temperature trend behind a discharge valve missing one poppet; inset: a zoomed view

Grahic Jump Location
Fig. 8

Comparison of machining the valve seat versus the poppet

Grahic Jump Location
Fig. 9

Three valve seat degradation conditions: nominal/healthy, 1/32″ shorter, and 1/16″ shorter

Grahic Jump Location
Fig. 10

Raw pressure—volume diagram for the CDV: (a) various spring conditions; (b) various valve seat wear conditions—100% loading

Grahic Jump Location
Fig. 11

Accelerometer mounted on the head suction valve; (a) with fingers; and (b) without fingers

Grahic Jump Location
Fig. 12

HSV for valve seat wear; (a) accelerometer signal; and (b) FFT of the signal

Grahic Jump Location
Fig. 13

HSV for valve spring fatigue; (a) accelerometer signal; and (b) FFT of the signal

Grahic Jump Location
Fig. 14

Installation location of dynamic pressure probe behind crank discharge valve

Grahic Jump Location
Fig. 15

Proposed classification methodology

Grahic Jump Location
Fig. 16

Raw measurement signals for the illustrative example

Grahic Jump Location
Fig. 17

Measurement following Hanning window application

Grahic Jump Location
Fig. 18

Hanning windowed FFT's for the 3 health classes with a frequency resolution of 0.61 Hz along with the 55 (6 Hz) bins between 20–350 Hz

Grahic Jump Location
Fig. 19

Classification results for the simulated example—0% misclassification

Grahic Jump Location
Fig. 20

CDV dynamic pressure for spring fatigue—50% loading. (a) nominal; (b) degraded; and (c) no springs

Grahic Jump Location
Fig. 21

Hanning windowed FFT with bins of the CDV dynamic pressure—50% loading. Bin range from 8 Hz to 350 Hz with 6 Hz bin size. (a) Nominal; (b) degraded; and (c) no springs.

Grahic Jump Location
Fig. 22

Classification results for the CDV spring fatigue at 50% loading. (a) Results from the PCA, p(z|wk), and the Bayesian classification bounds. (b) Percent variance from the PCA.

Grahic Jump Location
Fig. 23

Hanning windowed FFT with bins of the CDV dynamic pressure—100% loading. Bin range from 8 Hz to 350 Hz with 6 Hz bin size. (a) Nominal; (b) degraded; and (c) no springs.

Grahic Jump Location
Fig. 24

Classification results for the CDV spring fatigue at 100% loading: (a) results from the PCA, p(z|wk), and the Bayesian classification bounds; and (b) percent variance from the PCA

Grahic Jump Location
Fig. 25

CDV dynamic pressure for valve seat wear—50% loading. (a) Nominal, (b) degraded 1; and (c) degraded 2.

Grahic Jump Location
Fig. 26

Hanning windowed FFT with bins of the CDV dynamic pressure—50% loading. Bin range from 8 Hz to 350 Hz with 6 Hz bin size. (a) Nominal; (b) 1/32″; and (c) 1/16″ wear.

Grahic Jump Location
Fig. 27

Classification results for the CDV valve seat wear at 50% loading. (a) Results from the PCA, p(z|wk), and the Bayesian classification bounds. (b) Percent variance from the PCA.

Grahic Jump Location
Fig. 28

Hanning windowed FFT with bins of the CDV dynamic pressure—100% loading. Bin range from 8 Hz to 350 Hz with 6 Hz bin size. (a) Nominal; (b) 1/32″; and (c) 1/16″ wear.

Grahic Jump Location
Fig. 29

Classification results for the CDV valve seat wear at 100% loading. (a) Results from the PCA, p(z|wk), and the Bayesian classification bounds. (b) Percent variance from the PCA.

Grahic Jump Location
Fig. 30

PCA results for 3D case for various seat wear conditions—100% loading. Validation data (135 pts.)—0.0% quadratic 0.0% linear misclassification.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In