MYTH: Axial thrust does not matter

For the last ten years, we have only seen sporadic mention in publications about axial thrust calculations. Yet axial thrust management in centrifugal compressors has recently become a hot topic, driven by two distinct technology development trends.

One train of axial thrust discussion is related to the development of machines with magnetic bearings: Specifically, integrated hermetically sealed subsea compressors. The other is about increasing the allowed operating range of centrifugal compressors in highpressure machines.

But let us begin with a review of few basics: Axial thrust in compressors is a result of an imbalance in static pressure between the front and rear faces of each impeller, together with the reaction force of deflecting gas from the axial to the radial direction. The rear face of the compressor impeller sees impeller discharge pressure while the front face sees a mix of suction and discharge pressure.

The result is a net pressure differential axial force in the suction direction. Usually, this thrust is balanced by means of a balance piston (and in back-to-back machines by having some of the impellers face the opposite direction from other impellers). The remainder of the unbalanced thrust is the loads on the axial thrust bearing of the compressor.

The axial thrust contribution from each impeller, while well understood as a concept, is complicated by relatively complex process gas-flow patterns on the front and rear side of the impeller. This pattern is influenced by the operating point, seal clearances and the presence of swirl brakes.

“The axial thrust contribution from each impeller, while well understood as a concept, is complicated by relatively complex process gas flow patterns on the front and rear side of the impeller.”

The various seals (balance piston seals, impeller eye seals and division wall seals) are also a matter of concern. If any wears and clearances are increased, the thrust balance of the compressor may change. Thus the simple concept of centrifugal axial thrust is much more complex and cannot be easily predicted when it comes to the actual determination of net thrust and its long-term behavior.

For machines with multiple sections in a single casing, there is some evidence that back-to-back compressors see somewhat larger variations in axial thrust than in-line machines. They are also particularly sensitive if operating conditions in one of the sections changes at a different rate than another section. The thrust bearing then will have to carry the residual axial thrust but may not have originally been properly designed for these varying load conditions.

Also of importance is the fact that the magnitude and direction of axial thrust encountered by the thrust bearings changes with the operating point of the compressor, as well as with the level of aerodynamic impeller degradation and close clearance component wear. And for machines operating in gas or oil fields, liquid slugs are of major concern because the ingestion of large amounts of liquids will tax the thrust bearings.

The resurgence of concerns about axial thrust load is caused, at least to some extent, by the growing use of magnetic bearings. In fact, thrust bearing overload issues, and the resulting reduction in operating range, have been reported for subsea compressors. Magnetic bearings have a significant lower load bearing capacity than hydrodynamic bearings, and are thus less capable to cope with high levels of axial thrust.

Even for compressors with hydrodynamic bearings, many manufacturers will not allow the operation of their machines in the “choke,” “stonewall” or “overload” region. This can create severe limitations for machine operation, especially during transient conditions or during start-up, when the machine may see little pressure differential. Limitations in the overload operating range of a compressor, therefore, not only impact the operational flexibility, but also can require more complicated control systems.

Fortunately, axial thrust is well understood, but due diligence is necessary to avoid operational problems. The compressor itself is protected as it is usually possible to use increased thrust bearing temperatures as an early indicator of potential problems.

Written by: Klaus Brun & Rainer Kurz