Process Plant Machinery--Centrifugal Compressors (A-B)

Home | Using Industrial Hydraulics |

Applications of Computer-Aided Manufacturing

AMAZON multi-meters discounts AMAZON oscilloscope discounts

SUB-SECTION A: Compressor Design

The operating conditions of a plant determine the compressor design. The impellers are selected on the basis of the thermodynamic data. Selection is influenced by peripheral conditions such as materials data, aerodynamic and thermodynamic limits, or the operating characteristics.


Centrifugal compressors are designed to convert velocity energy into pressure energy. They thus change work introduced through the impellers into the required gas-side operating conditions; this primarily means the required discharge pressure.

The minimum driver size or rating needed is dictated by the energy required to accomplish this conversion, in which the losses should obviously be as small as possible.

The work performed to accomplish the change in volume results in an equivalent quantity of heat energy being transferred to the gas, causing an increase in temperature. Compressor design takes into account the fact that critical temperatures should not be exceeded if they might initiate polymerization or reaction of the medium with the compressor materials.


The compression cycle is designed to prevent the emergence of near-sonic velocities. These could initiate shock waves at the leading edges of the impeller blades, which then choke the flow cross sections and manifest themselves as irregular pressure rises. Near-sonic velocities may easily occur with compression media of high molecular mass, small isentropic exponent, and low temperature.


Optimized centrifugal compressors are designed so that their characteristics match the process requirements.

The total required energy is divided among the various stages in accordance with a number of criteria. Lightly loaded stages favor a stable working range that is as wide as possible, although the number of stages becomes larger and the work involved in building the compressor therefore more extensive. If a compressor is designed for a narrow volume flow range, highly loaded impellers are more economical. This enables the number of stages and hence the size of the compressor to be drastically reduced.

In addition to fluctuations in pressure and temperature, changes in the composition of the gas also occur. If the change is toward greater molar mass, a speed-controlled driver can be employed to reduce the drive speed and prevent surge in the final stages.

Design Diagrams

The principal characteristics of a compressor can be approximated from the diagram reproduced in Fig. A-1 for the power, number of stages, and speed in conjunction with the auxiliary diagram for quantity relationships ( Fig. A-2). The first of these diagrams is valid for isentropic, i.e., uncooled, compression. The specific energy Ys and compressor efficiency are not shown, but they are implicit as a function of their influences.

Compressors with pressure ratios of up to 7:1 are covered by the diagram, since cooling is almost always necessary for higher pressure ratios. In order to determine the compressor data in such cases, the compression process is split at the point where intermediate cooling takes place; the two separate sections of the compression process can then be summed after each has been determined separately.

Calculation Example

Using metric units, the prescribed operating data are as follows: Medium: a gas mixture consisting of inert gases and hydrocarbons Molar mass Mid = 36.5 Real Gas factor Z -- 1.0

FIG. A-1 Diagram for determining power, number of stages, and speed of centrifugal compressors.

(Mannesmann Demag, Duisburg, Germany.)

FIG. A-2 Quantity relationships for compressed gases. (Mannesmann Demag, Duisburg, Germany.)

Isentropic exponent r = Cp/Cv Mass flow rate rh Intake pressure Pl Intake temperature T degree Discharge pressure P2 Maximum tip speed (assumed) UEmax

The design procedure is as follows:

= 1.31

= 134,046 kg/h

= 5.14 bar

= 303 Kelvin (K)

= 19.45 bar

= 240 m/s

1. From the intake pressure Pl and discharge pressure P2, the following pressure ratio is obtained: Jr = P2/Pl = 19.45/5.14 = 3.784

2. Correction of the ideal molar mass Mid, using the real gas factor Z: M = Mid/Z = 36.5/1 = 36.5

3. The data coveting the volume flow rate ~' and the mass flow rate rh for any quantity can be taken from the adjacent auxiliary diagram for the quantity relationship for the given operating conditions. Taking M, p, and T into consideration, the following volume flow rate is obtained:

4. Starting from pressure ratio Jr -- 3.784, the arrow points vertically upward to the point of intersection with the curve for the isentropic tc -- 1.31; from there the next arrow leads horizontally toward the point of intersection with the curve for the same molar mass M -- 36.5. Vertically above that point, the temperature line T1 = 303 K is reached, after which the next arrow points horizontally toward the range of influence of the volume rate V. The heavy line then runs parallel to the family of curves until the specified volume flow rate V = 5 m3/s is reached, after which the heavy line runs horizontally again until it intersects with the corresponding mass flow rate line rh = 37.24 kg/s. Vertically above this point of intersection the power requirement P = 5150 kW can be read off.

5. The heavy line is now extended horizontally through the range of influence of the volume flow rate V until the line reaches the volume-dependent lines of opposite inclination. From this point it’s extended vertically downward as far as the stage number line z. For the assumed tip speed of u2 = 240 m/s we then obtain z = 4 as the number of stages.

6. Starting from the volume flow rate scale ~' = 5 m3/s, the heavy line in the direction of the arrow then impinges vertically on a specific size of compressor casing. From that point the heavy line is taken horizontally to the fight until it intersects the line u2 = 240 m/s established under step 5 (above). Vertically beneath this point of intersection, the speed n - 8200 RPM is read off.

Calculation Procedure

Influence of Intercooling

Energy saving is the main advantage and aim of cooled compression. Since the energy required is proportional to the intake temperature, the compression process is divided into a number of steps; the intake temperature of each of these steps is reduced in an intercooler.

The advantage of intercooling becomes apparent when the thermodynamically attainable energy savings are contrasted with expenditure on the cooling medium and coolers. The thermodynamic study must take into account the flow resistance of the cooler. This pressure loss is compensated by additional compression work.

The number of intercooling stages therefore depends on the overall pressure ratio Jr, the isentropic exponent K, which is determined by the temperature rise during compression, and the temperature differential.

At between the intake temperature of the first stage and the recooling temperature in the subsequent stages. The recooling temperature is determined by the temperature of the cooling medium and the heat exchange surface of the cooler.

The possible advantage of intercooling can be visualized with the help of Fig. A-3.

Example: Pressure ratio Jr = 5.35 : 1 Isentropic exponent K = 1.38 Number of intercooling stages j = 2 Result: Power factor f = 0.9, i.e., for the case in question, intercooling twice would result in a power saving at the compressor coupling of about 10% For further information on compressor calculations, see Sub-section B.

FIG. A-3 Influence of intercooling on gas compression efficiency. (Mannesmann Demag, Duisburg, Germany.)



High-Speed Centrifugal Compressors

As mentioned at the beginning of this section, dynamic gas compression is achieved by the mechanical action of a rotating impeller that imparts velocity energy to the gas. This velocity energy is then converted to pressure rise of the gas in a diffusion section that can be of a partial emission, vaneless, or vaned diffuser design. The rotating element or impeller of the centrifugal compressor was historically limited in rotational speed by the motor or turbine driver speed capability. With the use of separate speed-increasing gearboxes or high-speed drivers, the rotor speed was gradually increased to a maximum of 10,000 to 12,000 RPM. These are primarily multistage, medium-speed, high-flow centrifugal compressors. In the late 1950s, commercial work was compelled on centrifugal compressor designs that used an integral speed-increasing gearbox with the single impeller of the compressor rotating at speeds of 34,000 RPM and more. This provided the process industries with a design for low flows with high head capabilities.

Flow rates for high-speed centrifugal compressors can range from 10 ACFM to more than 100,000 ACFM. Power capability ranges from 15 to 2500 HP or greater.

See Fig. B- 1 for the composite performance envelope of high-speed centrifugal compressor designs.

Integrally geared high-speed compressors ( Fig. B-2) are available for use with any type of gas that a process design might involve. These gases can range from air or nitrogen to such exotic gases as hydrogen bromide.

To define where a high-speed centrifugal compressor is primarily used, the concept of specific speed and specific diameter should be understood and applied.

Specific speed, Ns, is a dimensionless index number for the impellers or rotors of various types of compressors and pumps. Using English/US values, the definition is the same for both pumps and compressors: N4- s-- (H)3/4

Another dimensionless quantity for impellers or rotors is termed the specific diameter, Ds, defined by: D(H) 1/4 Ds -- In both formulas: H = Head in ft-lbf/lbm, or foot-pounds (force) per pound (mass)

Q = Flow capacity in ft3/second at inlet conditions N -- Rotational speed in revolutions/minute D --- Diameter of impeller or rotor in feet

FIG. B-1 Typical head versus flow capability of high-speed centrifugal compressors. (Sundstrand Fluid Handling, Arvada, CO.)

FIG. B-2 Vertical high-speed centrifugal compressor with integral gearbox. (Sundstrand Fluid Handling, Arvada, CO.)

FIG. B-3 Specific speed versus specific diameter for initial selection of a type of single-stage compressor.

(A Study on Design Criteria and Matching of Turbomachines- Part B.)

The process or unit size determines the head (H) and capacity (Q) for a given system design. Components N and D are defined by the availability of mechanical designs from the various manufacturers. For high-speed centrifugal compressors, the normal range for these variables is N from 5,800 rev/min to 50,000 rev/min and D from 5 inches (127 mm) to 36 inches (914 mm). After the process design requirements and the need for the compression equipment have been defined, it’s necessary to determine the equipment best suited for the process by using the quantities of specific speed and specific diameter. Using Balje's chart ( Fig. B-3) with specific speed and specific diameter calculated for a particular application, it’s possible to determine if the application is a good fit for a single-stage, high-speed, centrifugal compressor. High-speed centrifugal compressors can be of partial emission, radial flow, or mixed flow design. By adding additional parallel stages or series stages of compression, the performance of a single-stage machine can be increased in flow or head. This can be done integrally within the same compressor frame, or it can be done by separate compressor units.

The detailed sizing of a high-speed centrifugal compressor is no different than sizing any centrifugal compressor (for further information see previous section and also Suggested Reading, Lapina). Virtually every compressor manufacturer has literature available that will assist in sizing the specific unit that would be required for a given process. The available literature usually contains performance envelopes showing the manufacturer's capabilities based on a single gas for a range of flows (Q) in ACFM or cubic meters per hour and heads (H) in foot-pounds per pound mass or newton-meters per kilogram. This enables the user to compare anticipated flow rates and heads with the machinery offered by a given manufacturer. To obtain a budgetary quotation or further information from the manufacturer, the user will have to provide the following data: inlet pressure, P1, either in gauge or absolute; discharge pressure, P2, either in gauge or absolute; inlet temperature, TI, ~ or ~ required capacity or flow rate, Q, preferably in pounds per hour or any weight flow unit; and the molecular weight, MW, of the total process gas or the individual components of the total process gas mixture. The specific heat ratio, K, is also required, as is the inlet compressibility factor, ZI. The compressor manufacturer can define compressibility factors and specific heat ratios for the majority of gases or gas mixtures. Proprietary or nonstandard gases must have the properties defined for proper selection.

FIG. B-4 High-speed centrifugal compressor estimating selection chart-procedure.

(Sundstrand Fluid Handling, Arvada, CO.)

Adiabatic head for single-stage compression is usually calculated from the following expression:

Dividing the discharge pressure (absolute) by the inlet pressure (absolute), one obtains the pressure ratio. Selection charts from the various manufacturers display either head or pressure ratio. The weight flow can be converted to volume flow rate by using the following formula:

With these two values, the head and flow, it’s normally possible to determine if the available compressor meets the user's needs. The design horsepower will be ...

The chart in Fig. B-4 is an approximating method for high-speed centrifugal single-stage compressor selection based on head and flow. The chart applies to two different impeller/diffuser designs from Balje ( Fig. B-3), the partial emission design for low flows, and the radial/mixed flow design for higher flows.

To use Fig. B-4, calculate the required head and flow for a given application, and enter the chart locating the intersecting point. This point will define the approximate speed, impeller diameter, and power for the compressor. The chart will also provide the discharge pressure (P2) and temperature (T2) by the formulas on the chart. The power curves in the chart are based on MW -- 4. To calculate power for other MW gases, use the correction formula as shown. When calculating the value of T2 for another MW gas, note the two different values based on each design type (efficiency). Power values are based on aerodynamic work only. Frictional losses are not included, but they won’t greatly affect the estimate.

For a given mechanical design of a high-speed centrifugal compressor (speed, impeller diameter, etc., are all fixed), there will exist only one performance curve of head versus flow based on the inlet conditions (P1, T1, MW, K, and Z1) of the process gas selected. The head/flow curve will relate discharge conditions to the given inlet condition. If inlet conditions change for the fixed mechanical design, the head/flow curve remains the same; only the discharge conditions will change. This head/flow curve is fixed by the mechanical design of the high-speed centrifugal compressor to accommodate the worst combination of operating conditions. Any other off-design conditions that don’t fall on the head/flow curve must be met by control methods as described later. The head generated by a single impeller stage of a high-speed compressor is quite high for a centrifugal design. The high-speed compressor design won’t match the head capability of a positive displacement design. However, unless high heads are required, the benefits of the centrifugal design, such as flow variation at constant pressure, often outweigh those of positive-displacement machines.

Typical applications for the high-speed centrifugal compressor involve those that require pressure ratios ranging from approximately 1.005 to 3.5. These are classified as recycle-, regeneration-, or booster-type applications. Pressure ratios in excess of 3.5 can be achieved by series arrangement of the high-speed centrifugal design. These arrangements can be either individual compressor units or multiple impeller stages mounted on the same speed-increasing element. Typical applications for the high-speed centrifugal compressor are molecular sieve absorption regeneration systems, vapor recovery systems, gas injection systems, gas recycle systems, booster compressors, chlorine transfer, and numerous other types of applications that require low pressure rise. High-speed centrifugal compressor designs are also used for inert gases such as air and nitrogen.

Available metallurgies for high-speed centrifugal compressors include carbon steel, 316 stainless steel, and 17-4 PH stainless steel, which are all quite compatible with normal gases. When the gases become more corrosive or erosive, as might be the case with chlorine, bromine, hydrogen sulfide, etc., special materials such as Hastelloy B or C, Inconel | or titanium may be required and are available from most manufacturers.

Any type of process gas can be effectively handled by the high-speed centrifugal compressor design, from hydrogen to Freon | because the components of the high-speed centrifugal compressor that come in contact with the process gas are relatively few and small. Corrosive or erosive gases can also be handled quite effectively through the use of special metallurgy. Similarly, these dynamic compressors can be built with gas containment seals suitable to meet environmental and loss prevention concerns.

A number of seal designs are available for high-speed centrifugal compressors.

They can range from simple labyrinth design for nonhazardous low-pressure applications to the typical mechanical contact face-type seals in either single, double, or tandem arrangements. The seals can be either gas- or liquid-type. Any type of seal arrangement must have some leakage across the seal face for it to function properly.

This leakage can be either the process gas or a buffer fluid (gas or liquid), depending on the environmental requirements.

All high-speed centrifugal compressors consist of three major components: the compressor, the speed-increasing element, and the main driver. The main driver can be an electric motor, steam turbine, air motor, gas turbine, combustion engine, or some other prime mover. Driver selection is generally related to the utility balance of the process plant installation. The compressor or process fluid element is fairly typical for all high-speed centrifugal compressors and consists of an impeller and diffuser contained within a pressure casing. The speed-increasing element between the driver and the compressor impeller is typically a unique integral gearbox that is provided by the same manufacturer ( Fig. B-2). The gearbox has been specifically designed to provide reliable and efficient operation of the unit.

High-speed centrifugal compressors are manufactured with either vertical or horizontal in-line, or horizontal axial inlet, and with top or side discharge. Other arrangements are possible and depend on physical size restrictions of the installation.

The following control methods are available for high-speed centrifugal compressors:

++Speed variation

++Suction pressure throttling

++Inlet guide vanes

++Flow control

++Variable diffuser

++Discharge pressure throttling

Control and operation of individual series and/or parallel units is essentially managed the same as with other dynamic compressors. Surge control must also be provided to protect the high-speed centrifugal compressor from serious mechanical damage. Further details on these subjects can be found in the section on axial compressors.

As with any mechanical device, routine maintenance of the high-speed centrifugal compressors is required. Depending on the process application and local environment where the high-speed centrifugal compressor is installed, yearly inspection and/or replacement of lubricating oil, oil filtration cartridges, ball bearings, and mechanical seals is recommended. Some applications may require lube oil analysis and/or replacement at a six-month interval. This can be determined by operating experience or discussion with the individual compressor manufacturer. The high-speed centrifugal compressor does not require any more maintenance than any other piece of mechanical equipment within the typical process plant. Experience shows that with proper selection, installation, and operation, the high-speed centrifugal compressor will provide many years of satisfactory operation and performance for any application in a gas compression system.

If a high-speed centrifugal compressor has been defined as a potential candidate for a given process, the design advantages may be quite numerous: oil-free operation (no contamination of the process gas), compact design (requires very small floor space and has minimal foundation requirements), simple process piping arrangements are allowed, smooth continuous flow (no pulsations into the downstream process), high inlet pressure capability, simple control systems are required, a wide range of metallurgies to handle various process gases is readily available, and there are no close clearances between rotating and stationary components impeding the passage of small contaminants.

Potential concerns with the high-speed centrifugal compressor design should be reviewed when these designs are compared with the conventional designs described earlier in the section. An inherently narrow, stable operating range must be considered, as must the fact that like all centrifugal compressors, high-speed machines must be designed for the worst combination of operating conditions. Also, depending on the specific process conditions, high-speed centrifugal compressors may exhibit lower efficiencies than comparable types of equipment that may be available for a particular application. The user must judge if these characteristics are significant to the potential installation.


Ingersoll Rand Co.: Compressed Air & Gas Data. Lapina, Ronald O.: Estimating Centrifugal Compressor Performance. Gulf Publishing Company, Houston TX, 1982.

Balje O.E.: A study on Design Criteria and Matching of Turbomachines- Part B. Trans.

ASME, J Engr Power, January 1962.

Prev. | Next

Home     top of page