Myth: The Surge Line Is Always The Surge Line


Surge is a global system instability caused when a centrifugal compressor’s process gas flow fully reverses, causing load swings on the rotating impeller. Every centrifugal compressor has a surge limit and operation in surge can damage close clearance components, such as seals, bearings, balance pistons and even cause impeller rub.

The fundamental reason for surge is that the compressor is forced into an operating region where the work input from the impeller is insufficient to overcome the head differential across it, thus causing flow to momentarily reverse to balance suction and discharge pressures.

The surge limit is usually drawn as a line near the left side of the compressor’s headflow map. Unfortunately, there are a couple of well-established myths about the surge line that we’d like to briefly address:

Myth #1.

Surge is always preceded by rotating stall at sub-synchronous frequencies; and vibration measurements can be used to detect the onset of surge. Many compressors experience rotating stall prior to full surge. Rotating stall starts with flow separation in the diffuser section and travels upstream into the impeller discharge as the compressor operating point approaches the surge line. However, this is purely an aerodynamic design characteristic of a given compressor geometry.

Many compressors never experience rotating stall. In these machines, surge is a sudden event, not preceded by any high axial or radial vibrations related to rotating stall.

Thus the claim that a surge detector can be designed based on vibration measurement has to be evaluated with great care. This only works for some machines. Even for those compressors, the margin between onset of rotating stall and full surge tends to be too small for the response of the surge control system to safely prevent the machine from going into full surge.

Myth #2.

The maximum head on the speed line corresponds to the surge point.

The surge line is usually placed at the peak of the head on the speed lines of the standard compressor head-flow map.

However, for many compressors when reducing flow, the head actually drops left of its peak before going into surge.

Since surge is a system issue, interaction between the compressor and the system may allow this in some circumstances. Although the compressor may not necessarily be in surge, any operating point to the left of the peak of the speed line is inherently unstable since there are two possible operating flow points for a given applied head.

A large flow fluctuation or pressure pulse can rapidly transition the machine between these two operating points and lead to operational instability and, if amplified, even result in periodic surge. In other cases, rotating stall occurs prior to maximum head, and many manufacturers will display a surge line that also avoids operation in rotating stall.

Myth #3.

The surge line can be determined by taking a few test points on the speed line, drawing a curve-fit through that and locating the highest head point.

This is a dangerous practice by users of older machines when they cannot find the original manufacturer compressor maps. As discussed above, the surge line does not necessarily cross the maximum of a speed line of a standard head-flow compressor map.

Furthermore, and more importantly, speed lines tend to be functionally complex. Assuming one can simply draw a higher order polynomial curve-fit to represent a speed line can introduce significant errors into the compressor map.

Myth #4.

Surge is purely a compressor characteristic, independent of the piping system or fluid properties.

This implies that the surge test result from the factory impeller test is universally applicable, which is not always the case. The onset of surge, its amplitude and frequency are strongly dependent on the compressor’s process piping system.

For example, surge cycles are dependent on the upstream/downstream piping volumes and surge limits are functions of the acoustic impedance and resonance frequencies of the piping system. In general, a control line is defined at a slightly higher flow than the surge line. Approaching and crossing the control line will activate the recycle valves.

Myth #5.

The surge line is straight and with a couple of surge points measured on a lower speed line, one can extrapolate the surge line to higher speeds.

The surge line is usually only a straight line if there is a single impeller in the compressor. For multi-impeller compressors, that is, for most compressors, the surge line is a composite of the individual impeller surge lines.

This surge line can be highly non-linear, especially in compressors with three or more impellers. Thus the assumption that one can simply measure a couple of low-energy surge points to define the surge limit at low speed, and then extrapolate the surge line to higher speeds, is false and can be dangerous for compressors where the slope of the surge line flattens at higher speeds.

Important note

A surge control system is only as accurate as the transducers that are used to measure the compressor’s operating point. These are primarily the process gas flow, pressure, and temperature transducers. If these sensors are not properly calibrated or are otherwise inaccurate, even the best and most complex surge control system cannot keep a compressor out of surge.

To our readers: When writing the Myth Busters column, our goal is to provide arguments that are coherent and sometimes even cohesive, but intentionally not abundantly forceful so as to encourage discussion, debate, and dissent. In this spirit we ask you to provide us with more feedback on our topics through letters to the editor or to us directly. We are also seeking ideas for future Myth Busters that can be dissected and disseminated.

Written by: Klaus Brun, the Machinery Program Director at Southwest Research Institute in San Antonio, Texas. He is also the past Chair of the Board of Directors of the ASME International Gas Turbine Institute and the IGTI Oil & Gas applications committee.
Written by: Rainer Kurz, the Manager for Systems Analysis at Solar Turbines Incorporated in San Diego, CA. He is an ASME Fellow since 2003 and the chair of the IGTI Oil and Gas Applications Committee.