ExxonMobil Proprietary

COMPRESSORS Section Page

XI-K 1 of 8
DRIVER TYPE SELECTION
DESIGN PRACTICES December, 2001

Changes shown by ➧

CONTENTS
Section Page

SCOPE ............................................................................................................................................................ 2

REFERENCES ................................................................................................................................................ 2

BACKGROUND .............................................................................................................................................. 2

BASIC DESIGN CONSIDERATIONS ............................................................................................................. 2

DRIVER SIZING...................................................................................................................................... 2
ECONOMIC CONSIDERATIONS ........................................................................................................... 3
VERY LARGE DRIVERS ........................................................................................................................ 3
SPECIAL SITUATIONS .......................................................................................................................... 3
DRIVER TYPE SELECTION PROCEDURES......................................................................................... 4

TABLES
Table 1A Comparison Of Compressor Driver Types ..................................................................... 6
Table 1B Comparison Of Compressor Driver Types (Torque, Size, Weight) ................................ 7
Table 2 Summary Of Major Considerations In Driver Type Selection......................................... 8

FIGURE
Figure 1 Classification Of Compressor Driver Types ................................................................... 5

Revision Memo
12/01 Added VFD comment to special situations.

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page COMPRESSORS
XI-K 2 of 8
DRIVER TYPE SELECTION
December, 2001 DESIGN PRACTICES

SCOPE
This section presents comparative information on various types of compressor drivers commercially available and the basis for
driver type selection. Details on various driver types and construction styles and on mechanical design features are presented
in Sections XI-L, XI-M, and XI-N.

REFERENCES
GLOBAL PRACTICE
GP 10-11-1 Sizing of Drivers and Transmissions for Compressors, Fans, and Pumps

OTHER REFERENCES
Gibbs, C. W., Compressed Air and Gas Data, Ingersoll-Rand Company, 1969.
Ludwig, E. E., Applied Process Design for Chemical and Petrochemical Plants, Volume III, Gulf Publishing Company, 1965.

BACKGROUND
Selection of the optimum type of driver for a compressor unit involves a qualitative screening of the circumstances surrounding
the overall plant design, followed by an economic evaluation of the most likely choices. Driver type selection is sometimes so
obvious as not to require the detailed evaluation study. However, where significant competition exists between rival types, the
comparative study and selection must be made early in the process design stage because a great many aspects of the plant
utility system design and compression services design are dependent on this decision.
The basic types of drivers used for process plant compressors are classified in Figure 1. The driver classification system most
relevant to process plant design is based on the driver's source of energy. With electric motors, energy is supplied via power
transmission cable from an electric power distribution network. With combustion engines, liquid or gaseous fuel is supplied to
the machine for internal conversion to shaft work via combustion. With expanders, energy is added to a working fluid medium
within one discrete unit operation (steam generator, regenerator, nitric acid plant converter, etc.), and extracted via the
expander at some convenient location downstream of the energy source.
For a comparison of driver types see Table 1.

BASIC DESIGN CONSIDERATIONS

DRIVER SIZING
Actual Size - Selection of actual driver power rating or “sizing" is performed during detail engineering, following compressor
model selection. The basis for driver sizing is specified by GP 10-11-1. This basis should also be used to estimate the
required driver size during the process design phase, so that the utility supply system can be designed. Per GP 10-11-1:
Driver power rating shall be the larger of 1 and 2 below:

1. Driven Equipment Normal Power Requiremen t x Load Factor

Mechanical Efficiency of Separate Power Transmission

2. Driven Equipment Power Requirement of Specified Alternate Operating Condition(s) x Load Factor
Mechanical Efficiency of Separate Power Transmission

Load factors, specified in GP 10-11-1, vary between 1.0 and 1.15. Mechanical efficiency of a gear unit between compressor
and driver ranges between 95 and 98.5%, as shown in Section XI-D. When a gear unit is not used (driver and compressor
operating at same speed), the transmission efficiency is 100%.
Driven equipment normal operating point is the point at which normal operation is expected and optimum efficiency is desired.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
COMPRESSORS Section Page
XI-K 3 of 8
DRIVER TYPE SELECTION
DESIGN PRACTICES December, 2001

BASIC DESIGN CONSIDERATIONS (Cont)

Alternate operating points are often controlling in selection of driver size. The specific compressor operating conditions which
are likely to characterize the highest driver load operating point are the following:
1. Highest flow rate.
2. Highest pressure ratio.
3. Highest specific heat ratio of the gas mixtures specified.
4. Highest average molecular weight.
5. Highest inlet pressure.
6. Lowest inlet temperature.
The Design Specification should include the estimated driver rating required to handle the operating point with the highest
power requirement, including the load factor of GP 10-11-1.
Rerating - When driver types are used which are pre-designed to fixed nominal ratings (some gas turbine models, standard
reciprocating engines, and all but very large electric motors), the compressor capacity rating may be increased to fully utilize
the available driver rating, assuming that this can be done at nominal extra equipment cost.
This practice is not followed when the vendor's equipment price increment nullifies the economic advantage. For instance,
most gas turbine models are priced on the basis of the driven load power requirement (to be competitive), and the increment for
rerating the compressor to the full capability of the turbine model is frequently so large as to make rerating unattractive.

ECONOMIC CONSIDERATIONS
Driver type selection is always based on economic considerations. The most important factor is the operating cost of the
energy supply. The key factors determining the energy cost are the base fuel value, and the overall efficiency at which the
available fuel energy is converted to mechanical work.
Other critical factors in the economic evaluation are:
• Number of units required, as dictated by the continuous run length and availability potential of the practical driver choices.
• Cost of downtime for maintenance.
• Investment requirement for both the driver and its utility supply system: equipment prices plus installation costs.
• Cost of maintenance.
Of relatively little influence in driver type selection are:
• Process control requirements, since adequate control systems can be designed for all driver types.
• Operator monitoring requirements, since, with proper instrumentation, the differences are small in comparison to other
economic considerations.
• Noise level, since satisfactory attenuation can be engineered at low cost for all types.
• Size, weight and complexity, since the direct economic impact is often small compared to other factors. However, size and
weight are important considerations in offshore applications.
• Driven equipment speed, since gearing can be included to perform any speed change necessary.

VERY LARGE DRIVERS

Drivers in the 35,000 to 100,000 hp range are in a special category in that very few process plant applications have ever been
made and each potential application requires extensive engineering study.
MACHINERY SPECIALIST consultation is recommended whenever services in this power range are under consideration.

SPECIAL SITUATIONS
Steam turbines can be applied to a variety of inlet/exhaust pressure levels within the plant to balance process steam needs.
Optimum matches of individual turbine power requirements to expansion energy availability should be sought.
Gas turbines have been justified for process plant service only when the exhaust heat is utilized to generate steam or can be
recovered in a process furnace. Investments and operating credits for the exhaust heat recovery equipment must be included
in the overall evaluation.
Starting Drivers - The steam turbines normally employed as starting drivers for gas turbines can be used for continuous power
contribution, if designed for continuous operation. The starting turbine rating can be specified for more than the minimum
required for starting, if dual drive is the economic way to meet a high driver load requirement.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
Section Page COMPRESSORS
XI-K 4 of 8
DRIVER TYPE SELECTION
December, 2001 DESIGN PRACTICES

BASIC DESIGN CONSIDERATIONS (Cont)

Common Drivers - When the power rating of an attractive gas turbine model exceeds the needs of a single compressor
service, two services can sometimes be handled by a common driver, with the compressors connected in tandem.
Power recovery turboexpanders can be applied with supplemental drivers in tandem to drive a compressor, where the
expansion energy available is not sufficient for the driven machine load, e.g., nitric acid unit compressor drive.
Surplus Power - When the power available from a power recovery turboexpander is greater than that needed by a process
compressor, an electric generator can be placed in tandem to harness the surplus power economically. See Section XI-M.
➧ Variable Frequency Electric Motors - As costs for these drivers tends to decrease with time, they are becoming more
attractive. Detailed study in concert with machinery and electrical specialists is required prior to selection.

DRIVER TYPE SELECTION PROCEDURES

Screening Study - Before making a detailed driver selection study one should eliminate all the unrealistic possibilities by a
screening study. Table 1A is first used to determine the possible choices from a power level standpoint. Table 1B presents
comparative size and weight data for driver types. Table 2 is then used to compare a wide variety of design circumstances and
objectives with the characteristics and attractive features of the more commonly used driver types. This qualitative process is
often sufficient to make a logical choice without detailed economic evaluation. In most cases, the choices can be quickly
reduced for case evaluation.
Before each service is studied individually in a project with multiple drivers, a check should be made on the effects of using all
drivers of the same type, since this approach normally gives the lowest utility supply systems cost and is often encouraged by
energy cost and utility availability factors.
Preparation of Design Specification - Refer to Section XI-L, XI-M or XI-N, for the specific type of driver selected.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
COMPRESSORS Section Page
XI-K 5 of 8
DRIVER TYPE SELECTION
DESIGN PRACTICES December, 2001

FIGURE 1
CLASSIFICATION OF COMPRESSOR DRIVER TYPES

Compressor Driver

Electric Motors Combustion Engines Expanders

Squirrel Cage Synchronous Steam Gas

Induction

Wound Rotor
Induction (1)
Back Pressure Condensing Power Recovery Cryogenic
Turboexpander

Extraction/Induction (5)
Reciprocating Dynamic

Radial Axial Radial Axial

Reciprocating (3) Gas Turbine

Gas Dual Fuel Diesel Industrial (4) Aircraft Derivative

Engine/Expander

Separate Integral Gas Engine Single Two

Driver Compressor Shaft Shaft

Notes:
(1) Rarely used for compressor drive.
(2) Impulse blading, reaction blading, and combinations of the two are all commonly used.
(3) Almost all are supercharged, rather than naturally aspirated. Both 2- and 4-stroke cycle designs are used for all
three types.
(4) Almost all for compressor drive are open cycle, rather than closed cycle; and non-regenerative, rather than
regenerative.
(5) Usually condensing. DP11KFF01

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page COMPRESSORS
XI-K 6 of 8
DRIVER TYPE SELECTION
December, 2001 DESIGN PRACTICES

TABLE 1A
COMPARISON OF COMPRESSOR DRIVER TYPES(1)

APPROXIMATE COMPRESSOR
ORDER OF AVAIL. SPEED, Rpm SPEED TYPES TO WHICH
USAGE HORSEPOWER (60 CYCLE VARIATION COMMONLY
DRIVER TYPE FREQUENCY RANGE POWER)(2) POSSIBLE EFFICIENCY APPLIED

Induction 1 1 to 30,000 3,600/N*less 2% Constant speed 10 hp (7.5 kW) - 86% All types
Motor (N = 1 through 8) drive - 100% or 100 hp (75 kW) 91%
Variable Frequency 1,000 hp (750 kW) 94%
Drive – 100% down 10,000 hp (7500 kW)
to 70%(3) 95.5%

Synchronous 5 100 to 40,000 3,600/N* Constant speed 93 to 97% Reciprocating,

Motor (and larger) (N = 2 through 13) drive - 100% or centrifugal and axial
Variable Frequency flow
Drive – 100% down
to 70%(3)

Steam Turbine 2 10 to 100,000 1800 to 35000 100% down to 70(3) 50 to 82% adiabatic Centrifugal, axial
(larger for fixed expansion efficiency flow, high pressure
speed, power rotary screw
generation services)

Power 7 500 to 26,000 hp Steam Turbine 55 to 85% adiabatic Centrifugal, axial

Recovery speeds expansion efficiency flow, and rotary high
Turbo- pressure rotary
expanders screw.

Combustion 3 5,000 hp (3750 kW) 100 to 90% single 26 to 29% over all Centrifugal, axial
Gas Turbine @ 10,000 rpm to shaft thermal efficiency for flow, and high
35,000 hp (26,250 100 to 80% two shaft simple open cycle 29 pressure rotary
kW) @ 4700 rpm. to 36 with regenerator screw
Many commercial
models available
below
3,000 hp (2250 kW)
for miscellaneous
applications.

Integral Gas 4 From 1700 hp (1275 100 to 75(4) 32 to 40% over-all Reciprocating
Engine kW) to 10,000 hp thermal efficiency
(7500 kW) @ 300
rpm

Coupled Gas 6 100 hp @ 1000 rpm 100%(4) 31 to 36% Reciprocating

Engine to 6,000 hp @ 300 41%, if turbocharged
rpm

Coupled 8 100 hp @ 600 rpm to 100%(4) 40 to 50% Reciprocating

Diesel 6,000 hp @ 360 rpm
Engine (higher power
models applied to
marine propulsion)
(high speed models
in low power range
available for
miscellaneous
applications)

Notes:
(1) Tabulation limited to the commercially applied ranges of horsepower and speed. *N = Number of pairs of poles.
(2) For 50 cycle power, speed is 0.833 of 60 cycle speed.
(3) Larger speed variation possible except as limited by train critical speed locations.
(4) Consult with MACHINERY SPECIALIST before assuming wider speed range. Range limited by train torsional critical speed or engine turndown
considerations.

ExxonMobil Research and Engineering Company – Fairfax, VA

 ExxonMobil Proprietary
COMPRESSORS Section Page
XI-K 7 of 8
DRIVER TYPE SELECTION
DESIGN PRACTICES December, 2001

TABLE 1B
COMPARISON OF COMPRESSOR DRIVER TYPES
(TORQUE, SIZE, WEIGHT)

ROUGH WEIGHT
STARTING TORQUE & AMPERAGE, STALLING LEVEL, TYPICAL LENGTH
DRIVER TYPE % FULL LOAD TORQUE lb/hp (KG/kW) RANGE ft (m)

Induction Motor 100 to 150% torque, 400 to 800% amp 150 to 200% 10 (6) 5 - 15 (1.5 - 4.5)

Synchronous Motor 30 to 40% torque under 514 rpm, 40 to 50% — —

100% torque above 514 rpm; 300 to 500%
amp

Steam Turbine 175 to 300% 115 to 300% 4(2.5) 10 - 30(3 - 10)

Power Recovery Same as steam turbine — 3(1.8) 5 - 10 (1.5 - 3)

Turboexpanders

Combustion Gas Turbine Both single and two-shaft gas turbines — 6 (3.6)(1) 20 - 50 (6 - 15)(1)
require a sizeable starting motor or steam
turbine. The single shaft requires much
larger starting driver because of its poor
part load torque characteristic.

Integral Gas Engine Nil, started by compressed air About 120% 80(48) 20 - 40 (6 - 12)
(INC. COMP.)

Coupled Gas Engine Nil, started by compressed air About 120% 50 (30) 20 - 40 (6 - 12)

Coupled Diesel Engine Nil, started by compressed air About 120% 50 (30) 20 - 40 (6 - 12)

Note:
(1) Heavy duty gas turbine weight and length shown. Comparable aircraft derivative engine weight is 2 lb/hp (1.2 kg/kW), length
20 to 35 ft (6 to 10 m).

ExxonMobil Research and Engineering Company – Fairfax, VA

ExxonMobil Proprietary
Section Page COMPRESSORS
XI-K 8 of 8
DRIVER TYPE SELECTION
December, 2001 DESIGN PRACTICES

TABLE 2
SUMMARY OF MAJOR CONSIDERATIONS IN DRIVER TYPE SELECTION(1)

SEPARATE
INTEGRAL DIESEL OR POWER
ELECTRIC STEAM GAS GAS GAS RECOVERY
CONSIDERATIONS MOTOR TURBINE TURBINE ENGINE ENGINE TURBOEXPANDER

Utility Considerations
Purchased power available, reliable, cheap X
Low process requirement for steam X X X
High process requirement for steam X X(2)
Steam generation within process system X

Fuel Considerations
Low cost fuel gas available X X
Heavy fuel oil is principal fuel X

Process Integration
Hot exhaust gas available X(3)
Process furnace can utilize hot, oxygen rich X
exhaust gas

Load Level
Compressor load < 6,000 hp (4500 kW) X X X X X
Compressor load 6,000 to 20,000 hp (4500 - X X X X
15000 kW)
Compressor load 20,000 to 30,000 hp (15000 - X X X
23000 kW)
Compressor load 30,000 to 40,000 hp (23,000 - X X
30,000 kW)

Driven Machine Type

Reciprocating compressor service X X(4) X X(4)
Centrifugal compressor service X X X X
Power generation service X X X X

Design Objectives
Minimum energy or fuel cost (5) X X X X
Minimum utility supply system cost (6) (6) X X X
Minimum number of units required X X X X
Minimum driver cost X X
Minimum total plant investment X(7)
Minimum maintenance requirement, long runs X X X
Minimum operator manning X X (8)
Minimum driver physical size, weight X X

Notes:
(1) X indicates consideration tends to favor this driver type.
(2) Steam generated in exhaust heat recovery boiler.
(3) Supplemental tandem driver sometimes required.
(4) Drive train requires special engineering to assure operable torsional vibration characteristics.
(5) Usually highest energy cost.
(6) For low power compressors, incremental utility supply system costs are low.
(7) Assuming use of purchased power.
(8) Higher operator manning than motors and steam turbines, but much lower than reciprocating engines.

ExxonMobil Research and Engineering Company – Fairfax, VA