Hydrocarbon Processing Copying and distributing are prohibited without permission of the publisher
Email a friend
  • Please enter a maximum of 5 recipients. Use ; to separate more than one email address.

Optimize compressor configurations for hydrocarbon applications

09.01.2011  |  Almasi, A.,  WorleyParsons Services, Brisbane, Australia

Extend the life of the compressor-driver package

Keywords: [compressors] [hydrocarbon] [WorleyParsons] [units] [motor driver] [synchronous motors] [induction motors] [gear systems] [flow stage]

One of the most important decisions a plant manager makes is the selection of the right centrifugal compressor train, especially given the context of emerging trends and challenges of modern compressor technology. As part of a comprehensive train optimization program, each compressor train component is examined to find ways to support goals of higher efficiency, lower cost, simpler maintenance and higher availability. Utilizing such a program can lead to higher total efficiency, higher train reliability and lower total cost of ownership. To optimize design and manufacturing, the compressor train concept should reduce the required manufacturing activities, simplify field assembly and extend the life of the compressor-driver package.

Compression unit arrangement.

Regarding optimum compressor configuration, in the basic design stage (early project stage), general questions about series or parallel arrangements, the optimum number of compressor units and standby requirements should be answered. The study of operating scenarios suggests certain requirements for the compression system. Beyond the quest for higher compressor peak efficiencies, the operating requirements usually require a compressor capable of operating over an operating range at reasonably high efficiency. Regarding compressor operation optimization, first consideration involves the capability to cope with changes in flow and head. Plant requirements determine the compressor operating conditions in terms of head and actual flow, and subsequently determine the required power from the driver. The second consideration deals with the fact that the nominal capacity may grow in revamp scenarios, future expansions, etc. The growth scenarios, if foreseeable, drive a compressor station layout to possibly allow additional train(s) or compression station(s) to be installed in the plant. The alternative scenario, where the consumption declines over the years (one example being that the gas supply from the field declines), is also a possibility.

Because the failure or unavailability of a compressor unit can cause significant loss in revenue, the installation of standby capacity may be considered. There is always great debate about the standby concept. In a majority of cases, standby unit are not required for API 617 centrifugal compressors. Modern centrifugal compressors can achieve an availability of 97% and higher. For a few very special processes, standby centrifugal compressor capacity may be considered. These standby capacities can be arranged such that each compression point has one standby unit (this is very rare), or that the standby function is covered by over-sizing all compressor trains and particularly drivers in each compression point. In this case, failure of a compressor does not mean that the entire system ceases to operate, but rather that the flow capacity of the process is reduced. Since some processes, like gas pipelines, have a significant inherent storage capability (“line pack”), a failure of one or more compressors does not have an immediate impact on the total throughput of those processes. Additionally, in a typical compressor station with multiple compressor trains, planned shutdowns due to maintenance can be implemented during times when lower capacities are required.

Compressor configuration.

Gas turbine drivers are used in very large compressor units, in remote areas and where cheap fuel is readily available. Gas turbines are very popular in remote pipeline stations, offshore units, LNG large trains and large compression units for various gas or petrochemical plants. The gas turbines allow for immediate starting capability if the need arises. This is a unique feature and offers great advantages in applications such as gas pipelines where variations in demand are expected and timely response is required. Gas turbines are relatively expensive and in some cases have demonstrated relatively higher maintenance costs.

Electric motors are less expensive than other drivers, such as gas or steam turbines. They are efficient and they also require reduced maintenance compared to other drivers. All of this contributes to lowering the operating costs and increasing the compressor reliability and availability. From an emissions standpoint, they produce very low noise emissions and no exhaust emissions. Electric motor drivers have been increasingly used in recent years with the availability of the latest generation of high-power variable speed drives (VSDs). Electric motor with VSD features a very dynamic power control.

Steam turbines can readily be speed matched to the compressor. Steam turbines have the ability to operate over a relatively wide speed range, which is ideal for the centrifugal compressor. Good examples include some petrochemical units such as urea and methanol plants or some refinery units in which steam-generated on a large scale is used as part of a process or for a specific use. For a steam turbine driven centrifugal compressor train, the mezzanine type arrangement is common. Steam condensers, as well as inter- and after-coolers and other auxiliaries, are located below the train operating floor. In this configuration, downward compressor nozzles are usually applied (downward piping branches are preferred). But, depending on plant design, upward compressor nozzles are also used. The steam turbine condenser is installed beneath the train.

  Fig. 1. An example of a compressor train.

Motor driver considerations.

Step-by-step, the electric motor and VSD power outputs are increasing, and an 80 MW VSD electric motor is no longer considered a new or leading-edge. The compression marketplace is quite conservative and no one wants to be the first to try new technology. The size of the motor is mainly tied to the size of the VSD, which in the past has been the limiting component of the electrical chain, especially in the application of large-size induction motors. Competent motor manufacturers now offer high-power VSDs for both synchronous motors and induction motors.

Induction motors. VSDs for induction motors are generally versatile units that can also be used for synchronous motors. The VSD for induction motors is usually based on pulse width modulation (PWM) technology and uses the latest generation of high-power, insulated-gate bipolar-transistor (IGBT) press packs. In combination with induction motors, these converters significantly enhance compressor performance in terms of efficiency, ease of maintenance, low noise and vibration levels. There is no need for harmonic filtering because these converters have been specifically designed to eliminate the most powerful harmonic currents. Extensive harmonic studies on these units have demonstrated that the total harmonic distortion is lower than 2% at every operating point.

The VSDs designed for induction motors are usually medium voltage frequency converters of the voltage source inverter (VSI) type connected to the high-voltage switchboard through transformers. The induction motor is fed by one or more converters according to the power in a fully redundant configuration. In case one converter has a maintenance outage, the motor can be operated at reduced torque by the other converter. VSDs for induction motors presently cover an approximate range of 1 MW to 40 MW.

Synchronous motors. VSDs for synchronous motors are usually based on load commutated inverter (LCI) technology and use thyristors. The system is tailored to drive synchronous motor with powers over 10 MW with practically no upper limit. Limits could be set presently around 100 MW. Based on vendor information, units above 100 MW can also be constructed based on market request. Two technologies are considerably different. The transistor switching power technology of induction motor VSDs allows it to turn the transistor on and off at high speed to provide very fast protection, even with high levels of overload current flowing, for improved control. With thyristor technology, the control can only turn the thyristor on, and the current in the thyristor has to fall to zero to turn the thyristor off. Another major difference between the two systems is that the induction motor VSD allows control of the system up to 300 Hz, corresponding to a nominal speed of two-pole electric motors of around 18,000 rpm.

Variable speed induction motors are available for the high-speed drive of compressors. The possibility of operating these electrical machines at high speed allows direct coupling to centrifugal compressors without the need for a step-up gearbox, which is beneficial in terms of installation cost, maintenance, reliability and availability. Induction motors, given their construction simplicity (especially in their rotating components) are more apt to rotate at high speed. VSD drives can also be used when the speed of a compressor, for process reasons, has to be maintained rigorously fixed. This, at first glance, may seem a contradiction in terms, but it is not. If the grid is unstable, small frequency variations may reflect on the motor and thus on the compressor speed. If a variable speed drive is placed between the grid and the motor, this can ensure constant frequency current supply and thus a perfectly even rotation of the motor independently from the grid disturbances. This technology can also be used in soft-start and soft-stop systems. Confidence in the latest generation of VSDs is also gaining momentum in the various critical industries, and an increasing number of applications are foreseen in the near future.

New vs. old. Most attention is usually directed to the relatively new technology of the frequency converters and harmonics correction. There are also reports of problems and issues in matured and old-fashioned equipment technology. Despite the fact that transformers have been in the market since the 1880s, these are failure reports on the old-fashioned transformer instead of the frightening power electronics. Often the transformers used are undersized for the application. In some cases, since the very beginning of operations, the transformers experienced overheating, even when running at partial load. Thermographics is a very useful condition monitoring tool in such cases. Thermographic pictures taken of the equipment (such as transformer’s coils) can show extreme temperatures. Temperatures in excess of expected ones show problems. For example, expected temperatures for class F electrical insulated systems would be in the range of 120°C. Temperatures in excess of 130–140°C can be a good indication for further inspection and monitoring. The overheating leads to early failure-caused short circuits in various electrical systems. There are reports of coil failures of undersized transformers in less than a year of operation. In order to mitigate the overheating problem, and therefore allow continued and trouble-free operation, specially designed forced ventilation can be considered and installed.

Controlling by means of varying the speed is by far the most flexible and efficient way of adjusting the operation of the centrifugal compression trains to the demand of the unit. Under steady-state conditions, the operating conditions that the plant imposes on the compressor show a roughly quadratic relationship between head and flow (since pressure drops are mainly due to friction). In transient conditions, flow may change without instantly changing head. Efficiently controlling the unit with multiple compressor trains requires careful studies. If the compressor trains are about the same size and efficiency (majority of cases), control the train so that the task is accomplished with the smallest number of running trains, with the load evenly shared. If the trains are different in size and efficiency (which is rare), it is often best to base-load the biggest and most efficient train and take the load swings with the smaller train. However, issues like different maintenance requirements, starting times and reliability have to be considered. Straightforward and simple answer cannot be expected. It is a complex issue.

  Fig. 2. An example of a gas turbine.

Gear systems.

Gear units are extensively used in centrifugal compressor trains that are driven by electric motors or gas turbines. For compressor train optimization, considerable knowledge and experience of gear unit operation and selection are necessary. Gear operation is a combination of rolling and sliding motions. At first contact between two teeth, the motion is mostly sliding, but, as the two pitch circles become closer and closer, more and more rolling occurs. When the pitch circles intersect, and the teeth are on the center-line between the two shafts, the contact is all rolling. Then, as the teeth go out of mesh, there is progressively more and more sliding.

Gear units are designed, manufactured and selected using some basic rules and some useful codes. The gear unit bearings are based on a certain hydrodynamic or rolling element design and calculations, while the teeth have to withstand the operating fatigue stresses. These stresses are complex to calculate and evaluate. Extensive simulations, studies and rigorous design reviews are required for gear units. Hertzian fatigue loading studies of the gear contact surfaces, similar to a rolling element bearing; plus sliding friction and the lubrication demand review for all pairs of surfaces that involves both sliding and rolling are required for a successful gear unit. Most industrial gear units are using hardened gears somewhere between HRC 54 and HRC 63, and they should never show measurable wear or pitting. It is necessary to understand the gear contact pattern. The ideal is to have a contact pattern completely across each tooth that is uniform all around each gear. This design is based on full contact, but sometimes there are machining or assembly errors and other times there is distortion of the housing. Consequently, tooth stress can increase tremendously.

Proper specifications will address these effects. Normal dedendum wear is fine pitting seen in the dedendum of teeth. It occurs after millions of load cycles, when a minimal oil film and sliding contact put the tooth surface into tension. The result is minor cracking and pitting, and slow removal of the dedendum surface.

Destructive pitting. Destructive pitting happens when the lubricant is grossly overloaded and large or sharp pits develop. The result is a noisy and rough gear in serious trouble with rapidly increasing damage. If the pits are relatively small and well rounded, they can support the lubrication film and the gear will last a long time. At the other extreme, large irregular pits destroy the lubrication film. Sharp and linear pits can cause formidable stress concentrations. Normal dedendum wear results in slow and measurable tooth deterioration which typically allows for a relatively long and predictable life.

Destructive pitting, though, will rapidly grow rougher and noisier, and may result in a catastrophic failure. Early in the gear’s life, it may be difficult to determine if the wear is corrective or destructive, but with corrective pitting the wear rate rapidly drops off. The most common other damage seen on gears occurs when the teeth are so heavily loaded that plastic deformation occurs. This is commonly called rolling, where metal is rolled or pushed up the active faces of the teeth, and peening, where the shape of the tooth is hammered irregularly until it is no longer an involutes curve. In both rolling and peening, the tooth form is slowly destroyed and both mechanisms show that either the gear is very heavily loaded or there is poor lubrication. The amount of allowable wear depends on the possible consequences of a failure.

With a critical application, the loss may be limited to only 15%. It shows the importance of routine inspection of gear units.

The best advice regarding special purpose gear units for critical compressor trains is the use of an API 613 gear system. API 613 in many places refers to superior selections of codes. Good collections of specifications, metallurgical, design, manufacturing, inspection and test requirements are noted in this code which allows optimum gear unit selection and proper train operation.

  Fig. 3. An example of a steam turbine.

Compressor considerations.

Competent vendors use state-of-the-art interactive design and prediction (simulation) tools to optimize aerodynamic, performance and increase efficiencies of compressor flow-path, particularly impellers. Compressor impellers and matched stationary flow-path components are developed using computational fluid dynamics (CFD) analyses and other modern design tools.

Three-dimensional (3D) blade profiles, diffuser flow angles, crossover bend curvatures, area ratios, and return channel vane shapes are optimized for each impeller stage to provide the best possible efficiency and reliability. Additional performance enhancements are achieved by improving the flow distribution channels at the inlet and discharge volutes. These enhancements allow providing some of the industry’s highest operational efficiencies in aerodynamic field for centrifugal compressor 3D impellers.

A key achievement of modern compressor technology is the ability to offer superior, 3D closed-type impeller designs which maximizing performance over a broad range of pressure and flow applications. To verify predicted performance, extensive single-stage testing is performed for each family of 3D impellers. These tests data and information are considered manufacturer proprietary and are not delivered to buyers. Based on experience only some data may be obtained in bidding stage before machine order. Remember: That after order, vendors only send the machine (each casing or let say combined impeller package) performance curve and data. Fig. 4 shows an example of modern closed-type impeller.


  Fig. 4. An example of a modern closed
  type impeller.

Flow stage ratings. Higher and lower flow stage ratings are derived from the prototype test data to form a family of impellers. Within each stage family, impeller geometry is fixed; blade heights are varied for higher or lower flows. Using this methodology, several stage families are used to span the desired flow coefficient range. Impellers and stationary components are then scaled up or down for different frame sizes.

For maximum flexibility, aerodynamic components are also scalable within each compressor frame size. Modern impeller manufacturing uses five-axis milling to ensure the quality of the impellers. Impellers are stress relieved, machine finished, balanced statically and dynamically, spin tested, and then mounted with an interference-fit onto the shaft.

Compressors use either fabricated steel diaphragms or a combination cast-and-fabricated steel design. Precision machining ensures dimensional accuracy and significantly improves the diaphragm surface finish. Nearly all diaphragms are horizontally split and finished at all horizontal and peripheral joints and on gas path surfaces.

Optimizing rotor-system design requires extensive researches in the fields of rotor-dynamic stability, aerodynamic cross-coupling stiffness, and rotor-bearing systems. Competent vendors usually have developed proprietary analytical tools for this purpose.

The general trend is to increase rotor stiffness by increasing shaft diameter, reducing impeller weight, and increasing journal bearing sizes. This allows higher torque transmission capabilities and higher-speed operation, with improved rotor stability characteristics, which are essential as gas densities and operating pressures increase. One of the most important factors influencing overall compressor reliability is the rotor-dynamic stability and its response to the unbalance forces.

API 617 covers requirements of rotor-dynamic stability and rotor-dynamic behavior. Compressor manufacturers should be regulated to satisfy all the requirements of this standard without exception. However, some additional requirements can be added on top of the requirements of API 617 to enhance reliability of the compressor. A good recommendation is that the unfiltered vibrations measured during mechanical run tests is limited to 80% of API 617 limits for balanced rotor (for example, a 20µm peak-to-peak unfiltered vibration compared to a 25µm API 617 limit for operation around 12,000 rpm).

Most mechanical running tests are carried out in a vacuum and, therefore, this vibration limit should be lowered at test conditions. This will require tight manufacturing tolerances and sound balancing of individual rotor components and rotor assembly during fabrication to ensure improved reliability of the machine in the field. For compressors with long strings (multi-casing compressors known as complex trains), train torsional analysis as per API 617 should always be specified. Reference and compressor manufacturer past experience of similar applications should be evaluated for aero-dynamics-induced “cross-coupling forces” encountered causing excessive vibrations and instabilities in the shop and in field tests.

Foundation and civil design. Designers should specify the required concrete tensile strength to ensure the foundation has sufficient strength to resist the dynamic forces. Standard concrete cannot receive epoxy grouts for approximately 28 days because of excess free water in the concrete. Steel fibers, placed in the concrete at the time of mixing and cast, help control plastic-shrinkage cracking and increase the tensile strength of the concrete. The increased tensile strength enables the foundation to better resist cracking caused by compressor dynamic forces.

Modern anchor bolts for compressors and drivers are typically made of high-strength steels such as B-7, with strength around 700 Mpa. In older compressor foundations bolts, tensile strength was typically 250 Mpa (comparable to St 37, with tensile strength around 37 ksi). When designed properly, the increased strength of the bolts enables greater clamping force of the machine to the foundation and a smaller bolt diameter. Research on compressor anchor-bolt design has shown that a termination point at the bottom section of the anchor bolt cast into the concrete will reduce the local tensile stresses in the concrete. J-bolts and L-bolts are not recommended. An industry standard practice is to use a heavy hex nut, economical and readily available, as the termination point for an anchor bolt. Spherical washers should be used whenever anchor bolt preload is critical, such as on vibrating rotating equipment. Spherical washers allow the anchor bolt clamping forces to be transferred uniformly across the bearing surface, rather than concentrating at one side of the washer and nut if the anchor bolt is slightly misaligned. Spherical washers offer an economical method of ensuring the anchor bolt loads are transmitted to the foundation.

A precision, non-shrink grout should be used on rotating equipment so that the void between the reinforced concrete foundation and the equipment bearing surface is completely and permanently filled. Concrete has a tendency to shrink as it cures and, therefore, will leave a small void or cause a misalignment of the equipment. Cementitious non-shrink precision grouts were used many years ago, but should not be applied now.

Epoxy grouts. Around 1955, epoxy grouts were developed for use in the petrochemical industry for applications such as gas compressors and for rotating equipment in caustic areas. Epoxy grout has a much higher compressive strength compared to its cementitious counterpart. Epoxy grouts range in compressive strengths from approximately 85 Mpa to 140 Mpa.

Its real value comes from it being impervious to lubricant oils and chemicals, and the ability to endure impact from vibrating equipment. The only acceptable grout for modern rotating machine is epoxy grout. Epoxy grouts provide the most effective transfer of static and dynamic loads from heavy rotating equipment to the foundation. Because the linear thermal expansion coefficient of epoxy grouts is typically two to four times that of standard concrete and steel, appropriate material properties must be designed into the expansion joints. Viscosity is another important property in choosing the appropriate epoxy grout, specifically viscosity, during installation.

A product that has desired properties during the final cure, such as a high compressive strength and low creep, may not have a viscosity that will enable straightforward installation. In this case, the epoxy will not make adequate contact with the entire bearing area of the equipment. Epoxy grouts offer a much greater dampening effect to damp vibrations. Oil-resistant sealants, such as a room temperature vulcanizing (RTV) silicone sealant, are a cost-effective solution for seal expansion joints, chock perimeters and the epoxy-chock interface with the compressor frame. RTV silicone will prevent oil and other contaminates from penetrating into the joints and interfaces of the foundation.


Modern centrifugal compressors and their drivers combine comprehensive modern processes, aerodynamics, mechanical and control knowledge and technologies. The latest manufacturing processes and state-of-the-art machine tools are used to produce them. As a result, modern multistage centrifugal compressor trains lead the compressor industry in both performance and reliability. Transient capabilities, growth capacity, flexibility, availability and total cost of ownership have been studied to offer optimized centrifugal compressor train design and arrangement. Various configurations of centrifugal compressor trains (including electric motor, gas turbine and steam turbine driven packages) are discussed. The VSDs are composed of complex equipment and many complicated electrical, electronic and mechanical systems. Usually, a complex combination of transformers, frequency converters, harmonic filters and cooling systems are required. Modern and latest technology components, as well as old-fashioned equipment such as transformers, should be covered for comprehensive design reviews and inspection programs to ensure the performance and long-term reliability of compressors and power systems. HP


1 Haight, B., “Dresser Rand sells 700th datum centrifugal compressor,” Compressor Tech Two, October 2009.
2 Bloch, H., Compressor and Modern Process Applications, John Wiley, New Jersey, 2006.
3 Chellini, R., “Synchronous and induction motors for compressor drive,” Compressor Tech Two, October 2007.


LNG     Liquefied natural gas
VSD     Variable speed drive
PWM    Pulse width modulation
IGBT    Insulated gate bipolar transistor
VSI       Voltage source inverter
AGMA   American Gear Manufacturer Association
API        American Petroleum Standard
RTV      Room temperature vulcanizing, a type of silicone sealant
CFD     Computational fluid dynamics

The author 

  Amin Almasi is lead rotating equipment engineer for Worley Parsons Services in Brisbane, Australia. He previously worked for Technicas Reunidas (Madrid, Spain) and Fluor (various offices). He holds a chartered professional engineer license from Engineers Australia (MIEAust CPEng–mechanical) and a chartered engineer certificate from IMechE (CEng MIMechE). He is also a registered professional engineer in Queensland and holds MS and BS degrees in mechanical engineering. He specializes in rotating machines, including centrifugal, screw and reciprocating compressors, gas and steam turbines, pumps, condition monitoring and reliability. He has authored more than 60 papers and articles dealing with rotating machines.  

Have your say
  • All comments are subject to editorial review.
    All fields are compulsory.


Interessting article.
Please is there a guide to select a driver (Gas turbine or Electric motor ) for compressor on pipelines

Related articles


Sign-up for the Free Daily HP Enewsletter!

Boxscore Database

A searchable database of project activity in the global hydrocarbon processing industry


Is 2016 the peak for US gasoline demand?




View previous results

Popular Searches

Please read our Term and Conditions and Privacy Policy before using the site. All material subject to strictly enforced copyright laws.
© 2016 Hydrocarbon Processing. © 2016 Gulf Publishing Company.