Strong pressure pulsations into the suction or discharge of a centrifugal compressor can move its operating point into operational instability regions such as surge, rotating stall, or choke. This is of special operational and safety concern in mixed pipeline compressor stations where many centrifugal compressors operate in series or parallel with reciprocating compressors. Over the last 30 years, several authors have discussed the impact of piping flow pulsations on centrifugal compressor stability and specifically, on the impact on surge margin and performance. For example, Sparks (1983), Kurz et al., (2006), and Brun et al. (2014) provided analysis and numerical predictions on the impact of discrete and periodic pressure pulsation on the behavior of a centrifugal compressor. This interaction came to be known as the “Compressor Dynamic Response (CDR) theory.” CDR theory explains how pulsations are amplified or attenuated by a compression system’s acoustic response characteristic superimposed on the compressor head-flow map. Although the CDR Theory describes the impact of the nearby piping system on the compressor surge and pulsation amplification, it provides only limited usefulness as a quantitative analysis tool, primarily due to the lack of numerical prediction tools and test data for comparison. Recently, Brun et al. (2014) utilized an efficient 1-D transient Navier-Stokes flow solver to predict CDR in real life compression systems. Numerical results showed that acoustic resonances in the piping system can have a profound impact on a centrifugal compressor’s surge margin. However, although interesting, the fundamental problem with both Spark’s and Brun’s approach was that no experimental data was available to validate the analytical and numerical predictions.
In 2014, laboratory testing of reciprocating and centrifugal compressor mixed operation was performed in an air loop at Southwest Research Institute’s (SwRI®) compressor laboratory. The specific goal was to quantify the impact of periodic pressure and flow pulsation originating from a reciprocating compressor on the surge margin and performance of a centrifugal compressor in a series arrangement. This data was to be utilized to validate predictions from Sparks’ CDR theory and Brun’s numerical approach. For this testing, a 50 hp single-stage, double-acting reciprocating compressor provided inlet pulsations into a two-stage 700 hp centrifugal compressor operating inside a semi-open recycle loop which uses near atmospheric air as the process gas. Tests were performed over a range of pulsation excitation amplitudes, frequencies, and pipe geometry variations to determine the impact of piping impedance and resonance response. Detailed transient velocity and pressure measurements were taken by a hot wire anemometer and dynamic pressure transducers installed near the compressor’s suction and discharge flanges. Steady-state flow, pressure, and temperature data were also recorded with ASME PTC-10 compliant instrumentation. This paper describes the test facility and procedure, reports the reduced test results, and discusses comparisons to predictions. Results provided clear evidence that suction pulsations can significantly reduce the surge margin of a centrifugal compressors and that the geometry of the piping system immediately upstream and downstream of a centrifugal compressor will have an impact on the surge margin reduction. In severe cases, surge margin reductions of over 30% were observed for high centrifugal compressor inlet suction pulsation. Pulsation impact results are presented as both flow versus surge margin and operating map ellipses. Some basic design rules were developed from the test results to relate predicted flow pulsation amplitudes to corresponding reductions in surge margin.
A Contribution to the Study of the Design of an Industrial Centrifugal Compressor
