ABSTRACT

The basic function of the centrifugal compressor rotor is to impart the

required compression energy to the gas.

The impeller is main part of the rotor. When the impeller is rotating at high speed, air is
drawn through the eye of the impeller. The absolute velocity of the inflow air is axial. The
magnitude and the direction of the entering relative velocity depend upon the linear velocity
of the impeller at the radial position of the eye considered, as well as the magnitude and the
direction of the entering absolute velocity. The impeller vanes at the eye are bent to provide
shock less entrance for the entering flow at its relative entrance angle. The air then flows
radially through the impeller passages due to centrifugal force. All the mechanical energy
driving the compressor is converted into kinetic energy, pressure and heat due to friction.

Impellers are most stressed components of the compressors demanding highly precise
manufacturing methods. The impellers are finish turned on CNC lathe to get the accurate
profiles and required the surface finish. The dimensional accuracy of the impeller and the
radius joining at the intersections of two surfaces should be smooth without under cuts.

Absolute programs are being used along with the machining cycles while finish turning
on CNC lathe. The programmer is putting lot of efforts in calculating the co-ordinates and
writing the programs. The efforts are continuous.

The finish machining of impeller is taken up as a project and efforts were put in
learning the CNC programs and no of exercises are practiced to standardize the machining
activity utilizing the CNC system features to the possible extent. Graphic simulation for the
tool paths are verified and carried out the machining operations on regular piece and found
OK. Profiles also checked with the templates given.

Introduction

Centrifugal compressor is one of the most critical equipments in a process industry.

Keeping in view of the customer's requirements of increased reliability and availability of
the equipments, optimization of compressor and its auxiliaries is aimed at. Thus the
experience gained in operating and maintaining of compressor Plants is the basis for
specifying, selecting and designing of these equipment’s.

Many serious problems during commissioning and subsequent operation can be avoided
by considering some of the minor but vital points during contract stage. Effort is made to
discuss some of the aspects, which need more emphasis during selection/design of the
equipment and need to be explicitly specified in the enquiry.

All impellers used in GE Oil & Gas centrifugal

Basic Data

It is needless to mention that adequate basic data for carrying out proper selection and
design of compressors must be furnished. The data includes mass flow, inlet and outlet
pressures, inlet temperature, type of gas, its analysis, and related information like
availability of utilities etc.

Specifying special process requirements and presence of certain constituents in gas that
are dangerous to machine or process will ensure proper selection of materials and sizing
of various equipments. Presences of potentially dangerous impurities like dust, H2S,
HCL, H2SO4, hydroxides, oil vapour etc. even in ppm levels must be specified.

Moisture Separators

No compressor specification is complete without mentioning the need of moisture

separators downstream intercoolers or at the inlet of compressor when liquids are
present in the gas stream.

If separators are not present or inadequately sized, it may lead to ingress of liquid
particles inside compressor causing high rotor vibrations, pitting and corrosion/ erosion
of impellers and seals, clogged gas passages etc.
Suction Filters

Centrifugal compressors are basically handling gases and effective removal of foreign
particles like dust, ash, tar etc. ahead of compressor will contribute to minimum
maintenance and long trouble free operation. Deposits of solids inside compressor tend
to block the flow and cause unbalance in rotor.

Degree of filtration and type of filter, particle size and analysis of dust should be mutually
discussed to ensure machine safety and process compatibility.

Excess Capability

Due to peculiar drooping characteristic curve of centrifugal compressor at any speed,

virtually no possibility exists for increasing the discharge pressure for additional volume
of gas over designed value. In view of this, the compressor must be built with excess
capability for achieving higher plant loads beyond rated one. This implies that the driver
also must be suitably sized over and above the ten percent reserve specified in API
Standards.

Parallel Operation

In some cases additional compressors are envisaged for augmenting existing capacity.
These compressors are required to operate in parallel with existing machines. Since no
two compressors can have identical characteristics, it is quite likely that these machines
will operate at different loads. Hence, following points need careful consideration, while
deciding parallel operation of compressor.

System Resistance

In many cases, the variation in system resistance with plant load is not properly defined
which will make the stable parallel operation difficult. It is therefore extremely important
that all aspects of system resistance shall be thoroughly discussed before
recommending parallel operation.

Characteristics of Existing Machines

The actual characteristics of existing machines under all operating conditions should be
furnished to enable selection of a compressor with optimum matching characteristics.

Control System

Generally, system resistance curve does not follow compressor characteristics. This
implies that the selection of proper control system is necessary for successful parallel
operation.
Operation at off/ design Point

Sometimes, compressor is required to operate at conditions, which are entirely different

from those specified in the contract. This may involve operation at different suction and
discharge conditions and/ or with different type of gas resulting in lower compressor
efficiency and higher operational cost. If such off design conditions are envisaged during
contract stage, optimum sizing can be made.

Third Party Inspection

It is a common practice to specify inspection of components during manufacturing stage

by an independent agency. However, when the scope of inspection is unclear or not
defined completely, this may result in delays in execution of project and hence, It is
desirable that various stages of inspection are clearly defined in the contract in respect
of major equipments.

Information on utilities

The availability of utilities like steam, water, electric power, power factor, nitrogen should
be clearly specified in the contract with their parameters to enable proper selection and
sizing of auxiliary equipment.

The actual quality of cooling water available at site generally defers from the quality of
cooling water specified in contract. This affects the cooler performance and reduces tube
life. Hence correct cooling water analysis and fouling factors should be furnished for
correct sizing of coolers, and material selection etc.

Noise control
In order to reduce total noise, contributions from various individual equipments must be
identified and evaluated. The ideal method for reducing noise is to prevent generation of
noise by better design such as
- Use of acoustic abatement in compressor hall
- Proper dimensioning of compressor hall to eliminate noise due to resonance
- Better ventilation in compressor hall with preferably sides open at least up to
working level.
- Use of special low noise valves
- Isolation of background plant noise.
- Proper location of high noise auxiliary equipment.
Practical approach would be to try to reduce noise level after the compressor and
auxiliary equipment’s are installed by use of acoustic abatement.
Instrumentation & Controls

The objective of any compressor is to achieve smooth capacity regulation and to prevent
the compressor from surging, even though the process flow reduces below the surge
limit of compressor. In order to achieve this objective effectively, it is advisable to include
compressor control in the scope of machine manufacturer, who has better knowledge
about the performance characteristic of machine and provide correct type of instruments
necessary for the service.

Splitting the scope of instrumentation and controls between the user and the vendor can
cause mismatch due to enormous work involved in the inter-phase, approval of drawings
and specifications and defining scope of supply in every case. Any communication gap
can lead to delays in commissioning of units particularly when special instruments are
involved which are not readily available.

Electric motor for compressor drive

The choice of electric motor as a drive for compressor is being favoured due to lower
installation cost and where cheap and reliable electric power supply is available.
Although electric motors are considered more reliable, following points need special
attention

 Motor voltage

During start up, motor draws high currents resulting in line voltage dip, the magnitude of
which depends upon size of electrical system. In cases where motor driving the
compressor is the largest in the plant, voltage dip can be 20% or more.

Since reduced applied voltage reduces motor torque by square of applied voltage, the
matching of motor torque under worst voltage condition to that of compressor torque
should be ensured.

It is a common practice to have a minimum of ten percent margin between torque

capability of motor and the torque required by the compressor from zero full speed.

 Compressor torque

Compressor torque comprises of friction, inertia and thermodynamic power. The often-
neglected factor is the inertia. To ensure safe acceleration, the inertia of total train
referred to motor speed should be correctly evaluated.

The starting of motor with load inertia larger than the inertia for which the motor was
designed may cause increase in starting time and it will be detrimental to the motor.

Sealing system

Taking into account the type of gas and process limitation, customer should select
simple sealing system.
Labyrinth sealing

This leads to leakage of considerable amount of process gas to the operating floor. The
limitation should be considered before the selection

 Injection system

If the process gas accepts the contamination by the sealing gas then this type of sealing
is best advisable.

 Injection and Ejection System

With this system, a little amount of process gas is thrown to atmosphere. Therefore,
customer should design his plant for this continuous leakage of the gas.

 Oil Seals (floating seals)

Seal oil system is slightly complex to the one referred above. During maloperation, the
sealing of process gas is impaired. However, during oil failure the trapped gas should be
(evacuated) vented.

 Mechanical Seals

In this type of sealing system a continuous contact will be there between stationary &
rotating parts of equipment. During seal oil failure these seals seal the trapped gas. Thus
loss of parent gas is kept very minimum.

 Dry gas Seals

For compressors to be both economical and operationally reliable, the sealing systems
used in their construction have to meet very high operational reliability. A critical role is
played in this connection by the sealing system. It used to be normal to use a variety of
sealing systems such as labyrinths or maintenance-intensive oil seals for this purpose,
but today the preferred choice is the dry-running, gas-lubricated seal. A gas-lubricated
mechanical seal necessarily needs a gas supply system to ensure pressurisation of the
seal with clean process gas or external gas.

The gas seal is a state of the art and a service-proven, operationally reliable component
of compressor technology in BHEL today. The design of a gas-lubricated mechanical
seal, mostly in the form of a cartridge, is essentially similar to conventional seal. It
consists, on the one hand, stationary seal face and, on the other hand, of a rotating seat.
The sealing faces of the seal face and seat collar maintain each other without contact,
resulting in a high level of operational reliability and long useful life. The non-contact gap
is maintained by a pressurised buffer gas or process system.
 COMPRESSOR SELECTION
Following paragraphs give general understanding of a compressor selection for a set of
specified operating conditions.

1. Compression characteristics of the gas are determined. For single component

gases like N2, AIR, O2 the characteristics are determined from the gas properties
available both for suction and discharge conditions.

The physical constants to be determined are

(a) CP/CV (b) Molecular Wt. (c) Compressibility factors.

For mixtures of gasses like hydrocarbons, the above constants are to be evaluated from
BWR equation of state.

2. Total compression ratio is determined. This is the ratio of final absolute to initial
suction pressure. It must now be decided whether the compressor is to have one, two,
three or more phases of compressions. Experience will indicate quite readily how many
compression phases are needed. The primary consideration is the expected discharge
temperature. A sufficient number of stages, with cooling between them, must be used to
prevent the discharge temperature of any of them from exceeding practical limits, such
as material limitations and oil seals limitations.

3. The overall efficiency of the machine depends on the optimum sizing of the phases
and intercooling and other factors like number of casings, number of stages, sizing of
each stage, impeller configuration and finally on the peripheral speeds.

4. It is customary in the refinery and chemical process industries for the project engineer
to establish the number of stages of compression, and for each suction pressures,
suction temperatures and discharge pressure with due consideration given to pressure
drops and cooling between stages.

DATA REQUIRED FOR SELECTION OF A COMPRESSOR

A) Thermodynamic data (properties, constituents, composition).

B) Suction/discharge conditions (pressure & temperature).
C) Capacity to be handled (kg/hr/nm3/hr).
D) Special requirements like inj/extn and other process requirements.
E) Type of gas - material stresses.
F) Type of driver.
Preliminary selection

No. Of phases No of No. Of Characteristics

Speed driver casings impellers in of impellers
Rating each phase
If this information is not contained in the enquiry, the manufacturer ordinarily will
estimate appropriate inter-cooler and piping pressure losses and pressures and
temperatures for various stages. He will likewise estimate stage suction pressures and
temperatures most optimum for the particular selection made by them.

A check is made to evaluate the total power requirement for the machine or total train
and with a suitable reserve driver is sized.

Quite often an enquiry will list one or more alternate process requirements such as
varying suction capacities, varying discharge pressures or both, or perhaps operation on
alternate gas stream may be needed. Where the capacities and pressures may not be
identical. In such applications, the requirements are analyzed and the machine is
designed for the most severe conditions giving best efficiency. In practice this rated
capacity is located at or just beyond the peak efficiency point to provide as wide and
stable operating range as possible.

Centrifugal compressor performance map showing relationship of

compressor speed and compression ratio with actual volume.
 YPES OF COMPRESSOR

Not too many years ago, it was common practice to use reciprocating
compressors when high pressures were required. Dynamic-type machines were used
only where larger volumes and lower pressures were involved. Dynamic compressors
were usually called blowers when air or gas was compressed to about 40 psig. The term
compressor was applied to any such machine where air or gas was compressed to a
final pressure over 50 psig. In recent years, this differentiation has become meaningless.
Industry now uses the term compressor for all types of machines compressing air or gas.

Two basic types of compressors exist: Dynamic type and Positive

displacement type.

Dynamic-Type Compressors

Dynamic type compressors are machines in which air or gas is compressed by the
dynamic action of rotating vanes or impellers imparting velocity and pressure to the
flowing gas. Two general variations, centrifugal and axial compressors, are available.

1. Centrifugal compressors are machines in which velocity and pressure are given
to the air or gas in a radial direction by one or more impeller-diffuser
combinations (Figure 1-4).

2. Axial compressors are the machines in which velocity and pressure are given to
the air or gas in an axial direction by one or more sets of moving and stationary
rows of blades (Figure 1-5).

Positive Displacement Type Compressors

Positive-displacement types of Compressors are machines in which successive volumes

of air or gas are confined within a closed space. The pressure is increased as the
volume of the closed space is decreased. Four general types, broken down according to
the constructional method used to carry out compression, are available.

1. Reciprocating compressors are machines in which the compressing element is a

piston following a reciprocating motion in a cylinder.

2. Rotary sliding-vane compressors are machines in which longitudinal vanes slide

radially in a rotor mounted eccentrically in cylinder (Fig. 1.6)

3. Rotary lobe compressors are machines in which two mating lobed impellers
revolve within a cylinder and are prevented from making a contact with each
other by timing gears mounted outside the cylinder (Figure 1.7).

4. Rotary liquid-piston compressors are machines in which water or other liquids are
used, usually in a single rotating element, to displace the air or gas handled.
 Although each of the positive-displacement types will vary from the other
somewhat, they can be grouped together as single class of machines for the purpose of
comparison with dynamic-type compressors. Figure 1.2 shows the basic pressure-
capacity constant-speed characteristics of positive displacement, dynamic-type and
centrifugal compressors. The centrifugal compressor is essentially a variable-capacity,
constant-pressure machine, the axial compressor and the positive-displacement
compressors are essentially constant-capacity, variable-pressure machines.

These basic characteristics, however, represent only part of the problem of

choosing the type of compressor best suited to a specific application. Equally important
is the capacity range that can be built into a single machine. As a general rule, positive-
displacement machines are for small capacities, centrifugals are for medium capacities
and axial machines are for large capacities. As in all other general statements, however,
it must be realized that there is considerable overlap of capacity range between these
different types of compressors.

Although flow limits for centrifugal are dictated somewhat by the specific
conditions involved and design philosophy of different manufacturers, general range of
compressor selection goes as below:

Positive displacement machines below 3000 M³ / Hr.

Axial compressors above 30,000 M³ / Hr.

Flow regions between 3000 - 12,000 M³ / Hr. can be considered over-lapping between
positive displacement and centrifugal compressors

Flow regions between 30,000 - 1,20,000 M³ / Hr. can be considered over-lapping

between centrifugal and axial compressors.
 105

104
DISCHARGE PRESSURE Psia

103
C E N T R IF U G A L

102
A X IA L

101
R E C IP R O C A T IN G

1 02 1 03 1 04 105 106
IN L E T F L O W A cfm

A P P L IC A T IO N R A N G E

FIGURE 1.1
Ranges for application of reciprocating, centrifugal, and axial compressors
Fig. 1-2 Comparison of Constant Speed Characteristic of various types of
Compressors.

16000
15000
14000
13000
12000
11000
10000
9000
8000
7000
6000

5000
4000
3000

2000
1000
0 20 40 60 80 100 120 140 160 180 200
 Fig. 1-3 Approximate Rated Speed comparison of Axial and Centrifugal
Compressors

Centrifugal Compressors

Centrifugal compressors are simple in construction and can be explained by relatively

simple means. A single-stage compressor consists essentially of the following basic
elements: (1) inlet connection, (2) impeller (3) diffuser, (4) casing (or volute), (5) shaft
and bearings and (6) shaft seal.

Gas enters the compressor through the inlet connection, which is proportioned to
minimize shock or turbulence as the gas enters the impeller. The impeller imparts
energy to the gas and delivers it at higher pressure and velocity to the diffuser where the
remaining velocity energy is gradually converted to pressure energy. The volute
completely surrounds the diffuser and serves to collect the gas and to recover additional
velocity energy. The impeller may be mounted directly on an extension of the drive shaft,
or the compressor may have its own shaft and bearings. In either instance there must be
a shaft seal where the shaft passes through the gas being handled. The type of seal
depends on the pressure level and the nature of the gas being handled. When the
pressure level exceeds the limits of a single-stage, a multistage compressor must be
used. This construction required a return passage to take the gas leaving each diffuser
and deliver it to the inlet of each succeeding stage.

Rotary Compressors

Rotary compressors have been proposed as the answer to many compressor problems.
The machine is suited for variable compression ratios and a wide capacity range,
operates at electric motor speeds, and has a maintenance potential equal to that of a
centrifugal compressor. The rotary compressor has no valves. It operates by confining a
volume of gas in a rotating pocket. The confined gas pockets converge into the
discharge chamber.

Straight Lobe Compressors

The straight lobe rotary compressor operates at the relatively low tip speeds of 50 fps.
This machine can accommodate a considerable quantity of liquid. The public utilities use
straight lobe compressors as displacement meters low-pressure pipeline boosters,
vacuum gas well gathering system and as super charger for gas and diesel engines.
(Fig.1.6)

Slide-Vane Compressors
In a slide-vane compressor the rotor runs eccentrically within a cylinder. Radial slots in
the rotor carry sliding vanes, which form a series of longitudinal cells. The centrifugal
spin of the rotor holds the vanes against the cylinder wall. Cell volume diminishes as the
rotor approaches the discharge chamber. A single-stage

slide-vane compressor can pull a 28" Hg. vacuum or pump 5 psig. A two-stage unit can
compress air at 250 psig. The unit is not suited for handling saturated and super-
saturated vapour, and is particularly inapt for cold jackets.

Helical Screw Compressors

A helical screw compressor is the most versatile form of rotary compressor. The most
successful helical screw compressor applications are delivering air for general
construction and road building purposes and operating pneumatic tools.

Spiral Axial Compressors

The spiral axial compressors have a top speed equal to the slowest helical screw
machine. The machine has a timing gear drive and runs dry, without the rotors making
contact. These machines provide 20-30 psig air for manufacturing operations.

DESIGN FEATURES

Introduction

Centrifugal compressors find their application in a variety of the industries associated

with the manufacturing of fertilizers, petrochemicals, oil refineries and steel plants for
various services covering the range from a suction capacity of 1000m3/hr to 450000m3/hr
and for a pressure ratio as high as 10.

Based on the type of casing design, compressors are classified into two types

1. Horizontally split casing

 2. Vertically split casing design

Horizontally split casing design (MCL type) (Fig. 4.1)

This design is used for low working pressure below 40 kg/cm2. These casings are in two halves with

horizontal parting plane. Main nozzles and auxiliary connections are provided in the lower half of casing

and the upper half serves as a cover and house the diaphragms, which may be lifted by removing the bolts

on the parting plane giving free access to the internals of the compressor for inspection and maintenance.

Vertically split casing design (BCL type) (Fig. 4.2)

This design is made of barrel construction closed on the sides by end covers with the
help of studs. This type of construction is suitable for high pressure operations upto 750
Kg/cm2. Sealing is provided between the casing and closing flanges with the help of
endless 'O' rings. In case of two compression stages arranged in series in case of 2BCL
Type , the outlet gas from the first phase is cooled in a heat exchanger and the cooled
gas is further compressed in the second stage. In these types of compressors, the
compression stages are arranged back to back.

Another type of compressor is the PCL, which has a cup shaped casing with a single
closing flange in the vertical plane instead of two as with the BCL. It is used exclusively
for the transfer of gas in gas pipeline service and it is for this reason easily recognized
by the two-flanged nozzles projecting from the casing suitable for connecting directly into
the line. The flexibility of being able to be used from one to three impellers in the same
forged casing, together with the robust, simple construction, render this type of machine
easily adaptable and excellent for continuous "heavy duty" service.
The SR and VC compressors have a single stage and are suitable for low pressures. The first have
normally a cast-iron case, whilst the others are fabricated. The Overhang impeller can be of the
open or closed type. The suction flange is, in general, on the axis of the machine, whilst the
delivery is in the vertical plane at the termination of the volute. These Compressors are used as
blowers or "boosters" in Industry, refineries and in petrochemical process when large
volumes of low
pressure gases are involved.

The DMCL compressors have two stages of compression in parallel in a single casing.
The two stages are identical and their delivery is on the vertical center line. The solution
is most balanced. The last impeller of each stage, i.e., the one nearest the center-line of
the compressor is a double flow type. A machine designed in this way enables the large
quantities of gases to be compressed with impellers of lower diameters. The other
aspects of construction are the same as for the MCL Compressor.

The 2MCL is a type of Compression with compressor stages arranged back to back.
This arrangement is used when intermediate cooling of the compressed gas is required.
For specific requirements, it is possible to provide supplementary volutes on in take and
delivery.

The 3MCL compressors have 2 or more stages in line with extraction or injection of side
loads. The number of extractions and injections varies from case to case (the number 3
before the frame designation is conventional). The weight flow through the machine is
not constant during the compression.

There remains finally the 2BCL compressor, a barrel type machine suitable for high
pressures. This type also has the compression stages arranged in back to back.

The designation that distinguishes BHEL compressors is formed by two groups, one of
letters and one of numbers.

1. The groups of letters define characteristics of the stationary part with particular
reference to the case ( i.e. BCL indicates the compressor is barrel type suitable
for high pressures whilst MCL indicates compressor suitable for relatively low
pressures with horizontal split casings).
2. A letter or a number before this group defines the arrangement of the
compressor stages or alternatively refers to intermediate extraction or injection.

For example: 2BCL 407 designates barrel type compressor with counter opposed
compression stages in series with impellers of 400 mm diameter having seven stages
compression. In addition to the normal BCL suitable for pressures upto 70 kg/cm2 ,
classes BCL/a, b, c, d have been developed for pressures upto 200 kg/cm2, 350 kg/cm2,
500 kg/cm2 and 700 kg/cm2, respectively.

Compressors are also designed with intermediate extraction or with injection quantity in
which case the flow through the compressor is not the same. However the constructional
features are basically same as those of other types of compressors indicated above
except for the additional nozzles required for the purpose of injection or extraction.
Design features

The performance of the compressor is assumed on the following criteria.

1. Thermodynamic performance the get required pressure ratio, is achieved with

the minimum power consumption and by optimum sizing of impellers.

2. Trouble free run of the unit.

This is achieved by proper dimensioning of the mechanical parts followed by the study of
flexural and torsional behavior of the rotors and finally by a proper assembly of the unit.

The important elements of the centrifugal compressors are casings, diaphragms, rotors,
seals, bearings, and couplings.

MCL COMPRESSOR

1. CASING UPPER HALF

2. CASING LOWER HALF
3. SUCTION DIAPHRAGM
4. INTERMEDIATE DIAPHRAGM
5. DISCHARGE DIAPHRAGM
6. SEAL ON BALANCING DRUM
7. SEAL IMPELLER
8. END GAS SEAL
9. OIL SEPERATION SEAL
10. THRUST BEARING
11. JOURNAL BEARING
12. SHAFT
13. IMPELLER
14. BALANCING DRUM
15. COUPLING
16. THERMO ELEMENT

BCL COMPRESSOR

A.) CASING BARREL

C.) SHAFT
D.) DIAPHRAGM
E.) BALANCING DRUM
F.) THRUST BEARING
G.) COUPLING
H.) JOURNAL BEARING
I.) THRUST BEARING PADS
Fig.4-1 CENTRIFUGAL COMPRESSOR (TYPE MCL)
BCL COMPRESSOR

FIG. 4-2

Fig 4-2
 MATERIAL & MANUFACTURING METHOD OF COMPONENTS
The present chapter deals with the materials used for manufacture of the principal
components of a compressor.

The materials used for the various applications differ, mainly due to the type of service.
The materials used for the rotor and the other internal parts differs more on account of
the temperature, the corrosive nature of the gas or the stress levels involved.

Casings

The casing of the BCL type is of barrel shape with nozzles and supports welded to it.
The barrel is made out of forging which gives high strength and renders the material
more homogeneous and therefore more suitable for high pressures.

Normally, carbon steel is used for the barrel, nozzles and supports and for the covers.
The material used is ASTM A 266 Gr. II (with carbon content limited to 0.2
-0.25%.), which is sufficient to obtain good mechanical characteristics and at the same
time suitable for welding. Stainless steel overlay is provided on the casing internals for
certain applications like CO2

The PCL also has a similar forged casing, its end cover is made out of forged steel
ASTM A266. The nozzles however are of cast steel because they have a complicated
shape.

The MCL compressor casing is made from cast iron or steel. For low temperature
operation one of the four grades of the steel ASTM A 352 is used, depending on the
actual operating temperatures. Finally for highly corrosive environments, stainless steel
castings such as ASTM A 351 Gr. CA15 or alternatively Grade CF8 (13% Cr) are
employed.

Diaphragms

The diaphragms constitute the dynamic profile of the stator parts of the compressor.
They are divided into three types viz., suction, intermediate and delivery. Suction
diaphragms guide the gas to the first impeller. Intermediate diaphragms have the
diffusers built in which serve the function of transforming the kinetic energy into
pressure, return channel serves to guide the gas to the next impeller with shock less
entry. Delivery diaphragm forms the diffuser for the last impeller and collecting chamber
and delivery spiral.

Diffuser

Diffusers can be free vortex design or bladed type. Vaneless diffusers are generally
used as they provide a good operating range and compressor characteristics. Vaned
diffusers are employed where a high degree of diffusion is required ,However they
restrict the operating range and are very sensitive to change in volume flows.

Diaphragms are generally cast, on account of their complicated form. Normally,

diaphragms are made out of cast steel, sometimes with additions of Nickel to improve
the impact properties for low temperature applications (1- 1,5% Ni). If the operating
temperatures are lower, one of the six grades available in ASTM A 352 is used or
alternatively austenitic stainless steel castings of grades ASTM A 331 Gr.CF8. are
employed.

For the intermediate diaphragms of the 2MCL and 2BCL compressors for low flow
application with narrow channels for all other services and to withstand high ∆ p, a more
suitable material such as high tensile carbon steel plate or forging is required which have
higher mechanical strength compared to castings.

End Covers

These components are used for the purpose of closing either sides of the barrel, which
also house the bearings, end gas seals and oils seals. Various oil and gas connections
are made on the end covers. Generally the end covers are made out of solid forgings.
Depending upon the gas medium either carbon steels or stainless steel are selected.

Rotor

The heart of the compressor is the rotor consisting of the shaft, impellers, balancing
drum, shaft sleeves, thrust collar. The shaft is made of heat-treated low alloy steel on
which the impellers are hot shrunk and keyed. Sleeves of special material such as
stainless steel are mounted on the shaft to protect the shaft from the contact of the gas
medium.

Shafts

The material used for the shafts for many services is 40 NiCrMO7 Steel. In reality, this
material has better mechanical characteristics than is generally required for normal
usage by centrifugal compressors. For low temperature applications special grades such
as ASTM A320 of appropriate grade , which give low temperature impact properties,
corresponding to compressor operating conditions are used

Impellers

The impellers are the most stressed components, in fact there is a tendency to exploit
them to a maximum so as to reduce the number of stages. They are made of low alloy
steel which give high mechanical characteristics but with low carbon content to ensure
satisfactory welding. This material has (also good for low temperature) having 2%
chrome 1% molybdenum and 0.13 to 0.17% carbon. Given that the weld and surround
zone are the same, a phenomenon of intergranular corrosion would appear to occur with
higher carbon content. It is for this reason that there is a need to limit the carbon content.
The intergranular corrosion weakens the metallurgical bond between grains, which leads
to the mechanical degeneration of the material.

The presence of carbon in the steel greater than that soluble in the austenitic matrix
determines the potential sensitivity of the material to intergranular corrosion. This is
prominent because carbon is principally responsible for the precipitation of the carbides
and impoverishment in chromium renders the material sensitive to corrosion.

When the impellers are to be used in corrosive fluids, steel with higher chromium content
is used such as X12Cr13 (13% of Cr), in particular corrosive conditions requiring very
high strength, steels with still high percentage of chromium as high as 15 - 19% may be
required,

Where high degree of corrosion and higher stresses are present, recourse can be made
to steel of 17-4 PH grade, with 17% of Chromium, and 4% of Nickel (precipitation
hardened at lower temperature). The steel with 9% Ni (N9 COGNE) is used for impellers
required to operate at very low temperatures upto -196º C.

Methods of Manufacture

The most common method of manufacture of the impellers is by welding.

1) Internal welding

Impellers are fabricated from solid forged discs. Depending on the size of
impellers, the vanes are formed by either milling process or by welding separate
vanes, which are preformed to required shape. The internal conventional manual
welding is employed to join the hub with vanes and shroud. As the welding is
carried out by manual process, special care is taken in the following areas.

1. To have homogeneity of the properties of both hub and shroud, both the
forgings are drawn from the same melt.
2. The heat treatment of the impellers in an inert atmosphere to control the
scale formation due to the oxidation.
3. Proper sequence of welding and uniformity of the weld fillet deposited.
4. Continuous monitoring of the weld quality during the process of
manufacture by liquid dye penetrant test at various stages.
Spark Erosion

Spark erosion method was used when the conventional welding cannot be
carried out because of narrower passages of the impeller channels. The
procedure allows manufacture of impellers for compressor handling capacities as
low as 100 m3 /hr. The process essentially consists of forming the shape of the
channel of impellers by removing the material by spark erosion by electro
discharge process. The electrode is made of pure copper machined to the exact
shape of the channel. Impellers with channels as small as 3 mm can be
manufactured with this method.

This method is now obsolete and not used any more for making compressor
Impellers.

External Welding

For handling very low volumetric flows and higher heads, new series of impellers are
developed and the characteristics of such impellers are very low outlet vane angles (15-to
18 degrees) and lower radii of curvature. These characteristics result in a very long
channel passage and the conventional welding becomes impossible. External welding
process has been specially developed to overcome this manufacturing constraint.
Following are the salient features of this manufacturing process.

 Manufacture hub with blind grooves with corresponding vanes on shroud.

 Match the grooves with vanes.
 Welding of root of groove by TIG automatic process with pulsed current.
The Material at the bottom of the groove is fused to the vane located
below without filler material.
 Subsequent layers of welding by TIG welding with filler wire.
 Suitable heat treatment process to achieve desired mechanical
properties.
 Finish machining.

Advantages

1) The external welding process does not have the weld deposit in the
internal gas passages and hence smoother gas passages are obtained.
2) The entire process is automatic and the quality of welding is consistent.

Balancing Drum
Across an impeller stage in a centrifugal compressor, there exists an axial thrust equal to
the difference of thrust between suction and delivery sides of the impeller. This thrust
adds up in a compressor for the total number of impellers provided in a stage and it is
too high to be taken by the thrust bearing all by itself. A balancing drum is fitted on the
shaft both sides of which are subjected to differential pressure. The size selected, such
that, a large amount of the thrust load due to the gas is compensated by the drum, and
the remaining thrust along with the coupling thrust is supported by the thrust bearing.

Spacer Sleeves

Spacer sleeves positioned in between the impellers have two functions:

1. To protect the shaft from corrosive fluids.

2. To establish the relative position of the impellers on the rotor. The spacer outer
profile also guides the gas into the suction of the impeller.

Generally sleeves are made from 13% Cr steels with less carbon content. The Spacers
are mounted under interference to ensure their positioning during running.

Bushes under Oil seals

The bushes positioned on the shaft under oil seals protect the shaft from the wearing
action of high flow of seal oil. These bushes are made by an overlay of hardened
material of 600 BHN. For high-pressure compressors the hard coating is unsuitable
because of very high interference required for shrinking on shafts where hardened
bushes are employed.

Bearings

The journal and thrust bearings are of white metal type with force feed lubrication. These
are housed outside the compressor in casing and they can be assembled without
dismantling the complete machine.

a) Lemon type or elliptical types of bearings are generally employed for low speed
and heavily loaded rotors. For high-speed machines, to have better damping
characteristics and minimize oil-whip phenomenon, tilting pad bearings are used.
The selection of the bearing mainly depends on the rotor lateral behavior.
b) Thrust bearings are of double acting tilting pad type with self-equalizing features.
Thrust bearings are lubricated by forced feed lubrication. Oil film between the
collar and shaft supports the rotor.
The fundamental parameter while choosing these bearings is the product of specific load
and peripheral speed. This parameter has to be contained within certain limits based on
method of lubrication and materials of bearing and shaft.

Improvements in the Design

Periodical review of the design is made for achieving more reliable, efficient and trouble free
operation of compressors at site. Extensive feed back from operating sites and the R&D at BHEL
enables modification existing design or adoption of new design philosophy. Also based on the
feedback from various sites, improvements in design are carried out in the existing design.
PRINCIPLES OF OPERATION OF A CENTRIFUGAL COMPRESSOR

As already mentioned, the centrifugal compressor impeller imparts pressure energy and delivers a
quantity of fluid depending on its size blade geometry and thermodynamic characteristics of the
gas.

In general, the fluid is directed to the inlet flange of compressor with a velocity of about
30m/sec., after an initial expansion (to about 60m/sec.), enters the eye of the first
impeller.

Let the initial conditions of gas at the nozzle inlet are indicated by V 0, P0 and T0 Volume,
pressure and temperatures respectively.

The conditions of gas at the eye of the impeller are determined on the basis of the
reduction in enthalpy and temperatures during the above-mentioned initial expansion

The following equation gives the reduction in enthalpy

C12 − C 02 h
∆H = and ∆T = ∆
2 g 427 Cp
C1 = Velocity at inlet eye of 1st impeller
C0 = Velocity at suction flange
g = 9.81 m/s2

At this point the fluid enters the first impeller, which transfers a certain quantity of energy
to it, bringing it to the final condition V 1, P1, and T1. Volume, pressure, temperature at
inlet.

According to the Euler’s theory, the energy given to a unit weight of fluid passing through the
channel between the two vanes of the impeller is equal to the change of its momentum.

1
∆hu = ( u 2 C 2 cos α2 − u1C1Cos α1 )
g
Velocity Triangles of Centrifugal Impeller
α1

φ
Since in general α 1= 90 (radial inlet) the expression becomes:
1
∆hu = C 2 u 2 Cos α2
g
For convenience the coefficient of head may be defined:
c 2 Cos α2 c2u
τ 2 = =
u2 u2
Therefore
1 Πnd e
∆hu = τ2 u 22 Where u 2 =
g 60

The quantity of energy per unit weight is called effective head (H eff) and does not
take into account the losses in the suction and discharge passages. The coefficient of
head τ 2, is a function of the geometry of the vanes and of the capacity as can be seen
from the diagram:

τ τ

η
η

In the diagram, trend of the polytropic efficiency η p is shown qualitatively which takes
account of the friction, the impact and the turbulence, of the real fluid on the channels
and vanes.

The polytropic head that the gas can utilize is only a portion of the effective head, which is given
by:

2
u
H poly = ηp × H eff = τ2 ×ηp ×
g

 K −1

K  P2 K ×ηΡ
= Z m × R ×T1 × ×ηp 


P 
 −1

K −1  1  
 
 Where
848
R= = Universal gas constant
µ
µ = molecular weight of the gas
ηp = polytropic efficiency
Zm = average compressibility of the gas
CP
k = = Where CP is Specific heat of constant
CV
pressure
and Cv is a specific heat of constant
volume
P2
= ρ = Pressure ratio P,
P1
T1, = Suction Temp (° K)

The polytropic head takes account of the thermodynamic properties of the gas. It can be
seen that the thermodynamic characteristics of the gas gets changed at the outlet as a
function of the inlet conditions.

From the head equation it is observed that with constant volumetric capacity and speed,
the polytropic head increases with µ m  olecular weight. For given head, increase in
µ ( molecular weight) reduces the constant R and hence increases the pressure ratio.
It follows from the equation that denser gases are easier to compress.

If instead, when the suction temperature is increased the pressure ratio is reduced, in
order to maintain the head. Conversely, for a given fluid, with the same volumetric
capacity and the same speed (RPM), the pressure ratio increases with reduction in the
suction temperature and vice versa.

On the contrary, change in suction pressure does not affect the pressure ratio as the
product
Po * Vo = R * To remains constant. The variation in the average compressibility of gas (Zm)
influences the compression ratio more in particular, as an increase of Zm corresponds to
a decrease of pressure ratio and vice versa.

If the number of impellers increases, to see how the thermodynamic characteristics of

the gas at suction influences the discharge, consider two stages of compression carried
out with impellers having the same characteristics. In the case of a gas having lower
molecular weight the intake and delivery volumes differ slightly and are:

Qo = GVo,

Qm = GVm
(G= weight capacity, Qm = volumetric capacity at the delivery of the first impeller and
therefore to the intake of the second)

From the diagram η p and τ 2 relative to the 2 equal impellers, it is seen that the head of
the second stage is higher because of the reduced volumetric capacity, the product
η pτ 2 increases but also the temperature after the first stage is increased, for this
reason the pressure ratio varies little.

τ τ
η

τ
η η

If a global head capacity curve for both of the impellers is required the heads for each
section capacity are added together as shown below:
Fig. 6-4

In general for a compression of one or more stages the capacity head curve at a given
speed has a shape similar to the one shown below:

The dotted curve is the surge limit, which as previously mentioned coincides with the
points of maximum head. If the gas to be compressed has a high molecular weight a
higher-pressure ratio will be generated and therefore the delivery volume will be smaller
than that at suction, Qm < Qo.

It will be clear therefore that in the second stage, if the impellers are the same (T2.η p
being greater) it will be operating close to the surge line.

Frequently, it is necessary to draw the characteristics of a compressor for rotational speeds

different from the design. As the effect of variation in speed on the velocity triangles is to
increase the dimensions without changing the angles and velocity relationship (similar triangles)
it is possible to define the relationship existing between capacity head, power and rotational
speed.

The capacity Q is proportional to the rotational speed, the following relationship may be written:

Q2 n 2
=
Q1 n1

from the formula of Euler the effective theoretical head is:

u2C2u
H eff =
g
and since u2 and C2u are both proportional to "n" it follows"

2
H 2  n2 
= 
H 1  n1 

From the preceding relationships given the formula;

 γQH
NPower =
η

It is simple to derive that the power is proportional to the cube of rpm:

3
N 2  n2 
= 
N 1  n1 

Analysis of the losses

a) Friction losses are proportional to the square of capacity These friction losses are
distributed and localized, and are caused by the movement of the fluid against the fixed surfaces
and also due to to rapid changes of direction and changing in cross sections,

b) Losses due to the impact of the fluid against the moving blades depend on the
capacity and also on the rotational speeds.

c) Volumetric losses are essentially due to leakage of compressed fluid from the
delivery to the suction of each stage , more generally, from a high pressure zone to a
low pressure zone. These losses are approximately proportional to the cube of the
speed and to the fifth power of the diameter.

To obtain a good efficiency in a subsonic compressor, it is necessary to operate at a

Mach No. lower than 0.8

V
(MACH NO = ; V = Velocity of the fluid the ducts a = mean speed of sound)
a

This is to ensure that the velocity of fluid at any point does not exceed velocity of sound due to
non-uniform distribution of velocity in channels, which would result in the formation of shock
wave with attendant increase in entropy giving rise to severe losses of efficiency.

DRIVERS FOR COMPRESSOR

A compressor may be driven by means of an electric motor, a steam turbine or a gas

turbine. The choice is always dictated by an economic balance, between initial
installation cost, operating cost for the depreciation period of the plant and consideration
of the forms of energy available.
Steam Turbine

For the process in which exothermic chemical reactions takes place (that is, for all the
processes in which there is an availability of steam, such as ammonia synthesis,
ethylene production plants, etc.), the compressor drive is normally by a steam turbine, in
this way the steam produced in the process is utilized . For example, in an ammonia
synthesis plant the steam is used in two networks, the first supplies steam at a pressure
of about 100ATA and 500°C to the turbine that drives the compression group, the
second supplies steam at about 35 ATA which is used in the turbine drives to the
refrigeration compressor and to the process air compressor for the second reformer, A
part of the steam from this second system is delivered to the first reformer.

From the point of view of the compression group, steam turbine drivers have three
advantages:-

1) Possibility of direct drive, which avoids the introduction of a gearbox,

which is always a critical item and available only up to a certain power
and within certain gear ratios.

2) Better regulating facilities, the governor of the turbine can be interlocked to a

process signal (intake pressure, intake or delivery capacity) so by simply varying
the turbine rpm, the operating level can be maintained at the desired point.

3) As a general example, it would be interesting to consider a synthesis plant

(fig.7.1).

In a condition of equilibrium, the plant upstream produces a certain quantity of gas, which the
compressors, rotating at a certain speed, increase from a pressure level Pi to a level Pf

If the production at the upstream plant of the group is reduced, the pressure on the suction
line tends to reduce because the compressors continue to run at the same speed, tending to
suck a larger quantity of gas than that is available. A pressure transmitter, set to the desired
pressure level, reduces the speed of the group until P=constant, an adjustment in the opposite
way gives an increase in production when the gas production is increased.

4) Start up Facility

After the turbine has been warmed up, the group can be started gradually and
easily, the torque/speed arrived at start/up is as shown in fig. 7.2.

From this diagram, it can be seen, how the rpm can be increased gradually,
without rapid acceleration, that is, without high inertia torque.
In conclusion, the turbine is an ideal driver for the centrifugal compressor. The
substantial disadvantage is the high initial cost of investment compared to electric motor.

Compressors group
Turbine

PRESSURE

TRANSMITTER

GAS
SYNTHETIC
PRODUCTION
REACTOR

Fig: 7-1

FIGURE 7-2
Electric Motor

Electric motor drivers present the following advantages:

Low initial investment cost.

The compression train always operates at the same speed which reduces, notably, the
risk of mechanical stability problems, naturally it is easier to determine a stable condition
for a single speed than for a wide operating speed range. Extreme reliability, the electric
motor is one of the oldest and most tried amongst modern machines,

The disadvantages of motor drivers are:

Regulation: - Having to work with fixed speed, the compressor must be designed for a
"rated" operating point, an operating point which, in the case of a variable speed drive,
would require the highest speed.

If the plant is required to operate at point needing lower capacity, two systems can be
used, the first requires the introduction of variable pre-rotation blades able to lower the
operating curve of the compressor, the second involves partially closing of a globe or
butterfly valve on the intake piping.

SEALS AND SEALING SYSTEM

The sealing of compressor ends is accomplished by either of the following methods.

Selection of a particular sealing system for a compressor depends basically on the gas
medium and leakage quantity, cost of the gas and nature of the gas like flammability and
environmental hazard etc.

The following sealing systems are generally used in compressors

1. Liquid film shaft seals

2. Labyrinth type seals
3. Injection extraction sealing
4. Mechanical seals
5. Dry gas seals

Liquid film shaft seals (Fig. 8.1 & Fig. 8.2)

This system is used when the gas medium is poisonous, inflammable and no external leakage is
permitted. In this type sealing system, high-pressure oil is injected in between a set of close
clearance seal rings mounted on shaft ends with a positive pressure of about 0.5 kg/cm2. The
seal rings are white metal led, free-floating type. The oil leakage through LP side is re-circulated
through the oil system and HP leakage is recovered by automatic traps is re-used after separating
dissolved gases.

Labyrinth seals

This system is employed for gases like Air, Nitrogen. The amount of gas leaking to ambient is reduced by selecting adequate number
of labyrinth fins. Some of the developments in a labyrinth type seals are stepped type seals, honeycomb type and turbine type. These
special types are employed for higher differential pressures and space restrictions in the machine.

Mechanical Seals
In this type of seals the carbon-rings forming part of stator part is kept in continuous contact with a rotating collar. The contact
surfaces are lubricated either by a cool buffer gas or lubricating oil. These seals are very effective means of providing shaft sealings.

Injection and extraction sealing

For gases like oxygen, raw gas and other explosive, corrosive, media the sealing has to be perfect
and where sealing medium like oil cannot be used, the above sealing system is selected. The
principle of this sealing is that vacuum is created by an ejector at the shaft ends which sucks the
gases and thus prevents any outward leakage of the contained gas.
Fig-8.1
 Fig-8.2

There are certain aspects of compressor design, which need careful consideration during design stage, and subsequent running of the
unit. Those include Leakage across Labyrinth seals, sizing of balancing drum. Thrust bearing sizing, thrust generated by coupling.

A. Leakages:

Worn out / damaged labyrinth seals cause higher leakage of gas and contribute
to loss of energy. The gas leaking out from high-pressure side joins the low-pressure
side and recirculates which cause loss of power with a consequential drop in efficiency.
The quantity of leakage depends on the differential pressure across the seal. The type of
seal is decided depending upon the seriousness of leakage (i.e., type of gas, differential
pressure across seal, to minimize the leakage for optimization etc). The clearances are
specified considering the centrifugal dilatation of the rotating element, type of seal and
material of seal.

Various locations in a compressor, which need a careful consideration to minimize

leakages, are:

1. From discharge of each impeller to its suction side through the impeller eye seal.

2. From high-pressure side of the balancing drum to the low-pressure side.

3. Inward leakages at suction, in case of compressors with sub atmospheric suction

pressure and leakage to atmosphere at high pressure end.

The types of seals generally used are

1. Conventional Lab seal (Fig. 8.4a)

2. Turbine type seal (Fig. 8.4b)

Fig 8-4a Fig 8-4b

B) Balancing drum sizing and thrust bearing sizing.

There exists some amount of thrust generated at each impeller because of the differential pressure prevailing across it. This
is schematically explained in figure shown in Fig. 8.5a.
 The cumulative thrust generated across all the impellers is huge and it needs to
be supported by the thrust bearing. The thrust bearing size becomes huge if all the gas
thrust has to act on the bearing. For this purpose a rotating element called "Balancing
Drum" is incorporated on the rotor. The drum is acted upon by a pressure differential,
which pushes the rotor in a direction opposite to that of the gas thrust. The thrust so
generated balances the thrust produced by the impellers. Amount of balancing depends
upon the size of the drum and pressure differential created across it. The pressure
differential across the drum is maintained by an efficient seal placed with optimum
clearances (See fig. 8.5b).

i. Fig 8-5b ii. Fig 8-5a

C. Thrust generated at Coupling:

The gear type couplings, while transmitting the required torque, also transmits an axial
force which is a function of the pitch circle diameter of the coupling, transmitted power and a
factor called friction co-efficient. The axial force so generated at each coupling gets transmitted
to the thrust bearing. The thrust generated at coupling increases with the friction between mating
gear tooth, (i.e., cleanliness of lubrication oil, condition of gear teeth, alignment of two shafts
etc.). The total thrust to be taken care for the sizing of the thrust bearing is as follows

TTotal =Tgas +Tcoupling −TBalacingdr um

LUBE AND SEAL OIL SYSTEMS

Every centrifugal compressor has a lube oil system. Seal oil systems, however, are not
always provided, depending on the type of gas such as air where simple labyrinth seals
are used .

The other sealing systems includes injection of an external medium such as Nitrogen
and / or extraction of compressed gas (e.g. CO 2 compressors). Inert gas injection is
provided to prevent leakage of process gas.

LUBE OIL SYSTEM

Lube oil is used to create a rotor lift by forming a hydrodynamic film of oil between the shaft and bearing,
and to cool the bearings.

Twin oil cooler is provided in the lube oil console to cool the return lube and seal oil in
order to maintain a constant temperature, good viscosity. Oil viscosity is a very important
parameter, which along with the geometrical characteristics of the bearings considerably
influences rotor dynamic behaviour and consequently vibrations. A lubricant may have
sufficient damping to allow operation at near critical speeds.

Low viscosity due to high bearing temperatures would make operation unsafe, if the weight of the rotor is
not supported by sufficient lift of oil film , the film could break and make the shaft and working surfaces
come in to contact, thus increasing temperatures and the risk of the bearing damages.

It should be stressed that the oil flow has to be sufficient to dispose of the heat generated by the bearings.
Hence, there is need for the bearing temperature to be always kept under control, it is regulated by varying
the inlet oil pressure with the pressure control valve (PCV). Also the drain oil temperature should be kept
under control to ensure a proper exchange of heat between the bearings and oil.

A schematic diagram of a lube oil system is shown in figure 9.1.

The following are main components:

a) Tank
b) Pumps
c) Coolers
d) Filters
e) Pressure Control Valve
 t o s e a l o il
P .C .V L u b e o i l h e a d be ro o s t e r p u m p s

F IL T E R S

C O M PR E SSO R

COOLERS

to tu r b in e c o n tr o l o il

PUM PS

O IL T A N K
N2
STEA M H EATER

C E N T R IF U G A L S E P E R A T IO N
 Fig. 9.1 A Schematic Diagram of Lube Oil System

Tank

The tank is always fitted with an electric or steam heater to give the oil, the right temperature and the right
viscosity for start-up. It has a sufficient capacity to ensure a good retention time (feeding the oil pumps
even without the drain oil returning to the tank) and to prevent foaming.

An inert gas (N2) blanketing is provided on the vapour space in the tank to prevent
formation of explosive mixtures of oil vapours and process gas, which may enter the
tank, along with the return oil.

Centrifugal compressors driven by steam turbines have a centrifugal separator, which takes oil from the
bottom of the tank and centrifuges it to separate the water accumulated in the tank, due to steam leakages
joining the drain oil from the steam turbine.
The separator is normally sized to handle the total quantity oil in the tank in a day. It may be designed also
to separate sludge and dirt, which is called clarifying, centrifuges.

Pumps

The oil is drawn from the tank and sent into the system by a suitably sized pumps Either positive

displacement or centrifugal pumps may be used; let's briefly look at the difference between these

two types of pumps.

A positive displacement pump gives more or less the same oil flow at varying pressures depending on the
system's resistance to pressure. Its characteristic curve is therefore as shown in Figure 9.2.

The operating point is where the characteristic curve and system's resistance curve (A in this case) cross.

Fig. 9.2 Characteristic Curve

Theoretically, a positive displacement pump's characteristic curve is parallel to the Y-axis but in practice
flow gradually decreases with a rise in pressure because of internal leakages (fluid leaks towards suction),
which increase, with pressure.
The only way to change the operating point is to alter the system's resistance curve since the speed of these
pumps is generally fixed. It may be easily altered, by adding a by-pass valve to direct the oil back to the
tank. By varying the opening of the valve, it is possible to move the operating point along the characteristic
curve for example, curve A represents the system's resistance with the by-pass valve partly open and curve
C the system's resistance with the by-pass more closed and B with wide open.

This valve normally opens automatically operated by the oil pressure that is to be kept constant (see figure
9.3).

PSV

PCV

Fig. 9.3
There are always two lube oil pumps provided on every lube oil system, one main pump and another
auxiliary pump. The latter starts automatically when the lube oil header pressure drops to a set point, which
is normally operated by an electric motor.

Downstream of the pumps, the discharge (see fig. 9.1) the piping branches to the turbine control oil system
(for compressors with steam turbine drivers) one line leads to the coolers, filters, lube and seal oil systems
and the other leads to the control oil line. Turbine control oil is not cooled.

Coolers

As already mentioned, oil has to be cooled in order to maintain good viscosity. A cooling system is
therefore provided consisting of two water or air exchangers in parallel which are
made to operate alternately by three - way valves. By controlling cooling water supply through a valve, the
oil temperature is controlled.
Filters

Clean oil should be kept circulating for optimum lubrication of bearings and seals and to prevent wear;
Two filter units are provided in parallel, one of them is working at a time with a provision to changeover
online with the help of a 3 way valve.

The filters are critical system components, which have to be kept under control with a differential pressure
gauge.

In short, a by-pass is necessary to regulate a positive displacement pump's delivery pressure. In addition, a
relief valve is always required to prevent uncontrolled pressure rise caused by accidental increases in
system resistance.

A centrifugal pump, however, has a delivery pressure, which varies little on changing the flow established
by the system's resistance.

The characteristic curve is as shown in figure 9.4

Fig. 9.4
It is necessary to select the pump to obtain required delivery pressure. In this case neither a relief valve nor
a by-pass valve is required, the delivery pressure is regulated by varying the speed. It is therefore,
understood that centrifugal pumps would be the most appropriate but centrifugal pumps have lower
efficiencies than positive displacement pumps for low flows.
When there is too much pressure drop due to dirty and clogged filter elements, the filter elements should be
changed to prevent damages to elements and rupture of paper cartridges.

When there are double filters and double coolers, it is necessary that flow changes over smoothly in the
three-way valves so that it does not interrupt the oil flow causing the pressure to rise upstream (and
therefore the relief valves to open) and a temporary interruption in flow (resulting in a minimum pressure
trip) to machines.

Pressure Control Valve

After the filters, there is a self-actuated pressure control valve (PCV), which keeps the oil pressure to the
bearings constant between 1.5 and 2.5 kg/cm2.

All the Oil inlet lines to the bearings are fitted with adjustable orifices and pressure gauges for regulating
the pressure to the individual bearings.

The oil should be able to easily drain without foaming or evaporating, which would make lubrication
difficult. It is therefore drained by gravity by providing adequate size drain chambers and oil return lines
with a suitable slope and to the main oil tank.

INSTRUMENTATION ON THE LUBE OIL SYSTEM

The instrumentation normally installed on a lube oil system comprises the following:

a) Level switch

b) Level gauge

c) Thermometer

d) Auxiliary pump start-up pressure switch

e) Differential pressure gauge or pressure switches on filters

f) Temperature switch for high temperature in the bearing header

g) Pressure switches for low oil pressure to bearings alarm and trip.

SEAL OIL SYSTEM

The seal oil systems used for compressors are either separate or combined. In combined seal oil system the
oil for the seals branch off from lubrication system oil header whereas for a separate seal oli system there
will be an independent seal oil tank , and seal oil pumps filters and coolers which is different from the
lube oil system.
To illustrate the sealing system and its components, let's first see how the gas is buffered at the compressor
shaft ends. The system adopted in 90% of the cases is with the help of seal oil system. The system consists
of creating an oil film between the floating rings and shaft. Very low clearances are provided (a few
hundredths of a milli meter) for the high pressure ring and relatively higher clearances of approximately
0.15 to 0.25 mm, for the low pressure ring (see figure 9.5)

The oil is injected between the two rings (HP and LP) at a pressure maintained constantly above the gas
pressure by approximately 0.5 kg/cm2 so that there is no gas leakage towards the exterior and only a small
quantity of oil passes through the H.P. ring and comes into contact with the gas (few litters per day). The
oil mixed with the gas reaches the traps, thereafter it is separated from the gas in degassing tank, after an
additional separation, it is returned to the main tank.

Fig 9.5

As the oil passing through HP seal is not adequate to totally dissipate the heat generated owing to smaller
clearances and smaller differential pressures, higher LP seal clearances are provided to ensure good oil
flow to prevent damage to HP and LP seal rings.

This LP seal oil, not being in contact with the gas, is drained straight into the tank through the lube Oil
return line (in combined systems).

To maintain a constant differential pressure of approx. 0.4 to 0.5 kg/cm2 between the Oil and gas, two
types of system are normally used to suit different compressor plant requirements and Operating
conditions.

The first system consists in maintaining the overhead tank level constant about seven meters above the
compressor axis so that the seal oil pressure (Po) is the same as the gas pressure (Pr) plus the static head
(H) (sec fig. 9.6).

The second system is essentially based on automatic control of differential Pressure between oil/gas which
maintains the desired differential pressure of approx. 0.5 by keeping the Overhead tank completely full and
under pressure (see fig. 9.7). In this case the tank position does not effect operation of the system so that it
may be installed directly on the compressor.
First System (See Fig. 9.6)

Fig. 9.6
The positive - displacement pump delivers the oil at approx. 10ata higher than reference gas pressure
established by the differential pressure control valve (DPCV) so that the level control valve (LCV) can
maintain constant tank level under the control of the level transmitter.

Second system (see fig 9.7)

As previously mentioned, the oil pressure is maintained at approximately 0.7 to 1 kg/cm2

higher than the gas by valve DPCV for any gas pressure.

The tank always works under pressure, its volume depends on the required retention time for the
compressor shutdown due to the minimum oil/gas differential pressure.

Fig. 9.7
When the process gas contains H2S, it could contaminate Seal oil, in such cases
accumulators are used. The accumulators prevent mixing of clean lubricating oil with
sour oil. Number of accumulators to be employed is decided by the specified retention
time.
Fig. 9.8
INSTRUMENTATION ON THE SEAL OIL SYSTEM

The instruments normally installed on the overhead for compressor sealing systems

a. First system (fig. 9.6)

PSL : pump suction low-pressure alarm

LSH : high tank oil level alarm

LSL : low tank oil level alarm

LSL : low level alarm and auxiliary pump start up

LSLL : minimum level trip

b. Second system (fig. 9.7)

PSL : pump suction low-pressure alarm

LSH : pressurization sequence permissive start

PDSL : Oil/gas low differential pressure alarm and auxiliary pump

start up.

LSLL : minimum level trip

The HP seal oil passes into the reference gas chamber and is drained by gravity to the traps, where it is
separated (sec fig. 9.9)

A float valve sends the oil to the gas-separating tank, where it is kept at a temperature of approx. 85 to 90
degrees C. A mechanical mixer or an inert gas (Nitrogen) scrubber may be used, if necessary for complete
stripping of any gas left in the oil.

The oil is then analysed and returned to the main tank or further regenerated if the gas is
not free from dissolved gases.

The gas coming out the top of the trap is released through a calibrated orifice of a specified size thus
creating a downward flow at the shaft ends of the compressor towards the traps collecting the seal oil.

Since the reference gas pressure is practically the same as the compressor suction pressure, there is a need
to increase the same, if the process gas leaking at the traps is to be returned to the compressor suction.
This is done by pressurising the balance gas with a buffer gas by taking a tap off from an intermediate
compressor stage. This always ensures a positive differential pressure at the orifice.

Seal oil drainage systems normally have the following instrumentation:

LG : trap level gauge
LAL : low trap level alarm
LSH : high trap level alarm
TSLH : gas separating tank heater on-off temperature switch.

re f.g a s
S e a l o il e q u a li z in g g a s

LP H P

B u ffe r g a s ( fr o m t h e d is c h a r g e o f c o m p r e s s o r
to ta n k

to a tm .

D r a in
Tap co m p re s so r
s u c tio n

LG

M e c h a n ic a l
m ix e r

V ent

N 2

EH

to ta n k TS LH
CENTRIFUGAL COMPRESSOR TESTING
I. INTRODUCTION
Every centrifugal compressor, after assembly, is subjected to several stringent tests to assess the mechanical performance. The scope
and extent of tests is detailed as per API standards. A thermodynamic performance test is also conducted whenever specified.

The above tests are recommended to be conducted at manufacturer's works.

These tests are very important and vital, as they reveal manufacturing and design defects, if any, and help in the rectification and
improvement of the performance of a compressor at manufacturer's works itself. This will avoid unnecessary wastage of time at the
process plant on such rectification efforts and minimises the commissioning time for these units.

II. TESTING - TYPES

The compressor testing can be broadly classified into

- Mechanical run tests

- Performance tests or partial load tests

Mechanical run tests as per API 617 can further be elaborated as follows.

- Mechanical run test (of individual barrel)

- Tandem test (of all the barrels arranged in a train).

- Combined unit test (All the compressors arranged in train and driven by
its
actual driver, using actual couplings, gear box, Lube oil piping, Base plate
etc.

- Thermodynamic performance tests are, generally conducted as per ASME

PTC - 10. These are carried out in either close loop or open loop
depending
on the gas used for test.
III FULL LOAD, FULL SPEED, FULL PRESSURE TEST

Many times full load tests are prescribed for certain critical applications
like hydrogen recycle compressors for Hydro-cracker where the
compressors are required to be run by creating operating conditions
close to the site. This is achieved by using a mixture of helium and
Nitrogen.

iV. MECHANICAL RUN TEST

Mechanical run test is carried out to verify the overall condition of the unit assembly and helps in

maintaining a high quality of the product. They generally reveal the manufacturing defects, if any and

demonstrate the mechanical performance of the machine.

Broadly, the following are verified during the shop inspection and testing

- Rotor balance.

- Bearing alignment, clearances and adequacy of lubrication.

- Clearances at interstage labyrinth seals and liquid film shaft seals.

- Freedom from internal rubs.

- Absence of oil leakage and restrictions in the lube and seal oil channels in the
compressor casing.

- Trouble free and smooth running of the unit, at maximum continuous speed for a
period of 4 hours.

- Avoidance of abnormal conditions at over speed or trip speed (110% of M.C.S.)

- Location of first critical speed of the machine, in respect of flexible shaft units.

- Shaft relative vibrations and casing absolute vibrations.

- Run outs on the shaft at low speeds.

- Noise level.

- Rotor response to unbalance

V TESTING PROCEDURE

The compressor can be directly driven by a variable speed steam turbine, or alternatively by a

drive arrangement consisting of an electric motor, hydraulic coupling and gear box. By varying

the oil let into the hydraulic coupling, it is possible to vary the speed of the secondary shaft from

about 500 to 1000 rpm to 3,000 rpm. Further speed increase is obtained from the step-up gearbox,

which in turn is connected to the compressor through a flexible coupling.

The compressor is tested under vacuum with the help of high capacity vacuum pumps. This will reduce the friction and windage losses
and hence the power consumption and temperature rise during the test run. The compressor discharge hood temperature will shoot up,
if connected to atmosphere instead of the vacuum line.

During the test run, the bearings are supplied with oil from a centralized lube oil console, which will pump oil at the required pressure
and temperature. The system consists of twin coolers, high-grade duplex filters and control valves.

The compressor is started from stand still and the speed is increased in 10% increments
till the maximum continuous speed is reached. The compressor is kept at this speed for
4 hours. After this period, the compressor is made to over speed at 110% of M.C.S. for
15 minutes. During the deceleration, the first critical speed is located and vibrations due
to shaft run-outs are recorded.

VI TANDEM MECHANICAL RUN TEST

During this test, shop driver, shop lube oil console and shop pedestals are used. All the
compressors in a train are arranged in a string. The mechanical run test procedure is the
same as above. The train component compatibility is checked during this test.

VII COMBINED UNIT TEST

In this type of test, the actual job turbine or motor is used for the test. Generally, all the active components, such as couplings, base
plate, oil piping are used during the test.

VIII STATIC SEALS TEST

After completing the mechanical run test, liquid film (shaft end) seals are assembled in
position. The compressor is pressurized to the operating pressure by an inert gas such
as Nitrogen (see fig. 3). A differential pressure of 0.5 kg/cm2 is maintained between the
seal oil inlet and the gas from inside (i.e. reference gas). Since the seal oil is at higher
pressure, it will prevent the gas from leakage from the compressor. The shaft is turned
slowly and leakage seal oil across the HP seals is measured. With this, it is possible to
judge the fitness of the seals.

Similar test is performed with dry gas seals and the assembly and performance at
standstill condition is verified.

IX THERMODYNAMIC PERFORMANCE TEST

This is carried out to predict the thermodynamic performance of the machine with a specified gas of known properties and under
specified conditions.

A compressor can be tested with a suitable test gas, whose physical and thermodynamic properties are well known and the results are
converted into expected performance of the compressor for its design or specified gas under its design operating conditions.

`ASME' power test codes and BS standards stipulate various types of tests for different
types of gases.

When a compressor is tested with the specified gas at the specified (Guaranteed) operating conditions, (whether treated as a perfect or
real gas), the test is termed as class I test. Class II and Class III tests are used when the compressor cannot be tested on the specified
gas at specified operating conditions. In class II tests, the test results are to be converted to performance data expected on the specified
gas under specified conditions, using the perfect gas laws.

Power test codes specify the limits of compressibility functions X and Y for the test and specified gases. If these limits are exceeded,
class III calculations are made. The conversion of test results to specified conditions is made using real gas equations in case of class
III tests.

X PERFORMANCE TEST PROCEDURE

The prescribed test gas is admitted at low parameters and purged for 3 times. After the gas analysis is satisfactory, the compressor is
started at low running after obtaining their test parameters, such as inlet pressure, temperature, gas flow and speed the discharge
pressure and temperature etc., are recorded.

From the recorded values, the computations of results are made. During the test run,
every effort is made to locate the incipient surge. The performance characteristics, such
as pressure ratio vs. flow, power vs., flow, polytropic head, vs., flow etc. are plotted.

SURGE & ITS CONTROL

Surge can be defined as the capacity below which the centrifugal compressor performance becomes
unstable. The surging conditions occur when the inlet capacity approaches the surge capacity. At a given
speed, as the inlet capacity is reduced, the discharge pressure of the machine goes on increasing. This
happens because of the collapse of the pressure developing capability of the machine at surge point. The
percentage flow at which the surge occurs in a centrifugal compressor depends upon the aerodynamic
design, gas properties, stability of flow, system characteristics, stage components design etc.

When surging occurs, the high-pressure gas at the discharge of the machine flows back through the
compressor in a complete reversal of the normal direction of flow. Once the discharge pressure is dissipated
sufficiently, the conditions causing the collapse of pressure no longer exist and the compressor resumes
normal flow and discharge pressure. If the discharge conditions on the machine have not changed, the
collapse of pressure producing ability will repeat and cycle through the same sequence.

In certain occasions, the flow components of stage may have a mild surging with the flow reversals within
its geometric domains without causing the overall surging of the machine, or without hampering the overall
pressure producing capability of the machine. Such mild surging cannot have audible noise or pressure
fluctuations, which cannot be felt externally. Most often, the mild surging of individual components creates
enough flow fluctuations that the adjacent components break down due to instability and the overall
machine goes in to surge.

All necessary precautions during installation of the machine in a system shall be taken to prevent surging.
Surging condition imposes transient stresses of cyclic nature on the machine. Certain safety margins are
built within the machine to absorb mild surging without damaging the machine. The effect of surging can
be excessive vibrations with a resulting rub between rotor and stationary parts, drastic over heating of the
machine, create large shaft torque impulses, over speeding, severe stresses in the suction and discharge
piping, pressure vessels etc.

It is impossible for the compressor manufacturer to quantify the magnitude of forces involved during
various surging cycles and to design the machine and related equipment and system to withstand such
forces under all possible circumstances. Hence, the best design practice would be to install proper piping,
controls, and liquid knockouts etc., to prevent surging and protect the machine and to avoid such a
potentially damaging situation.

The protection of compressor from surge is to provide an anti-surge control system. The basic idea in the
antisurge control system is to provide the minimum flow through the compressor suction at all speeds
which is higher than the surge limit either by re-circulating back the required quantity from the discharge of
the machine after cooler or by venting. Generally the anti surge control is effected based on the flow
differential at suction or discharge. But for certain applications, this is not an adequate function and hence
the control is obtained by modulating the control with pressure, temperature flow and combination of those
parameters.

On any control system, the system shall be also designed to prevent surging during starting and stopping.
The trip out of the driver shall be interlocked to immediately open the antisurge valve and allow the
machine to coast to a stop with antisurge line open. Otherwise the machine could be surging constantly
while coasting down.

One of the most potentially damaging occurrences for compressor is the ingestion of liquids with the
process gas. The following precautions are to be taken.

a) Trim cooling water or other process conditions to keep the compressor inlet
conditions above the liquifation points for any gas constituent.
b) Heat trace or purge normally stagnant lines when liquids are collected due to
stagnant gas cooling down to ambient temperatures. This is normally the case
with the Urea synthesis machines. This avoids the problems of erosion of valve
stems.
c) Recycle lines should re-enter main gas stream upstream or at inlet of knockout
drums.
d) If there is any possibility of liquid formation exists upstream or downstream of
compressor, drains and level indications should be provided at all low spots of
piping, vessels, etc. This allows routine checking for liquids and draining as
required.
e) During a start-up after a shut down, ensure that all the liquids formed by
cooling of the stagnant process gas are drained away. Heavy flow
variations during surging can pick up large slugs of liquid and can damage
the compressor.
f) Prevent sub-cooling of gases to temperatures below liquid line, after expansions
across valve openings. Some liquids or even solids may be formed from this
effect.

Apart from the design of process piping and vessels, the support structure of the filter
elements shall be so designed that they will not be torn and fragments enter the
compressor during heavy fouling.

In double flow compressors, another factor that must be considered is the design of the
inlet piping to achieve a well-balanced, distortion free flow into each side of compressor.
Otherwise, the compressor will not handle equal flow rates on each side and premature
surging may occur before anti-surge control is activated. If any uncertainity exists on
achieving balanced inlet flow a splitter is arranged in the piping ahead of compressor
inlets.

On all compressor inlets, the piping should be such that the flow is free of distortions at
the compressor flange. Butterfly valves, elbows, other obstructions etc. should be kept
sufficiently upstream of the compressor flange or flow-equalizing provisions shall be
built into the piping and within the allowable pressure drops.

INSTRUMENTATION & CONTROL SYSTEM

For the safe operation of the plant as well as the machines themselves a sizable number
of instruments and controls are necessary. The selection and sizing of these instruments
and controls are in turn co-related with the driver and gas circuits.

Control System:

The following methods are employed for maintaining the capacity, the suction or the discharge pressures

constant under varying process conditions.

1. Variation of compressor speed - This method is employed when steam

turbine or any other variable speed driver drives the compressor.
2. Suction throttling - This method is frequently used and is achieved by providing a
throttling valve in the suction piping. The minimum obtainable capacity is dictated
by the surge limit of the compressor.

3. Adjustable inlet guide vanes - This vanes induces a pre-rotation of the gas by
altering the direction of flow there by varying the compression ratio and capacity
of the compressor. This feature is not available on BHEL machines.

An inherent feature faced by centrifugal compressor is SURGE. A centrifugal

compressor working at a fixed speed transfers to the gas through the impellers, an
amount of energy, which can be considered, roughly, to vary linearly with suction
capacity of the compressor. This energy will be transformed with higher efficiency at
near the design capacity.

The energy lost due to friction and impact on the impeller vanes is, in fact, a minimum at
the design capacity, above and below which the loss increases rapidly. This follows in
the capacity control diagram on Fig. 9.1.
C A Pcharacteristic
In the centrifugal compressor A C I T Y curve
C O Ndrawn
T R Ofor Lpolytropic
D I A Ghead
R Aversus
M the
volumetric suction capacity, the surge capacity corresponds to the highest head and
lowest flow. This capacity assumes greater importance, since, it defines the limit below
which the operation of the compressor becomes unstable and gives rise to a pulsation
phenomenon of pressure and capacity, called surge.

Surging causes abnormal stresses on the compressor seals and bearings and may
result in serious damages. Simple means of preventing surge is to provide a valve,
which returns to the suction a volume of gas, which is at every instant at least equal to
the difference in volume between the surge limit line and the capacity available at
suction. To achieve this, the designers follow five algorithms to sense the surge and
provide anti surge control under varying process conditions.

T Y P IC A L P E R F O R M A N C E C U R V E

TIPS FOR EFFECTIVE & MAINTAINENCE

Centrifugal Compressors play a very important role in any modern fertilizer plant and are
rightly considered as the heart of such plants. The trouble free service assured by these
machines optimises plant utilisation. Uninterrupted trouble free service can be expected
from such machines as they are systematically operated with adequate care and
maintained properly. The amount of sophistication incorporated, while designing these
machines, enables uninterrupted operation over very long periods, thereby reducing the
down time to a minimum. Still regular maintenance, periodic overhauls accompanied by
regular tests on governing system, protective, safety and monitoring equipment enable
the plant operating personnel to get the best performance from these machines.

Some of the forced outages of turbo-machinery are reduced by systematic quality

control method followed, with good service experience, during the stages of :

a) Equipment design and selection.

b) Manufacturing and installation.

c) Operation and control of the equipment.

d) Preventive maintenance and routine plant maintenance.

Problems and failures can arise mainly because of

a) Non-conformance to design specifications.

b) Manufacturing defects.

c) Operational deviations and deficiencies.

d) Defects in installation.

e) Poor preventive maintenance.

f) Malfunctioning or failure of instrumentation and controls and protective devices

g) Handling by in-experienced operating and maintenance crew.

The overall reliability and optimum utilisation of these machines depend on many
factors. While, it is not possible to discuss all the factors, a brief reference to them is
made under the following headings:

1) Overall reliability

2) Operation and maintenance of compressors.

3) Training needs of operation and maintenance personnel.

OVERALL RELIABILITY

The associated drive equipment auxiliaries and systems such as the lube oil system,
steam and cooling water system, inter-connecting piping, valves and instrumentation
should be considered integral with the turbo-machinery train and not in isolation.
Operating experience shows that auxiliaries have been and continue to be the main
source of many frequent troubles. To obviate this, the systems are selected and
engineered by a single agency . The overall reliability of turbo-compressor unit does not
depend on main machinery alone but also on the auxiliary equipment’s and
instrumentation and controls, which are mostly bought out items. The machine can be
damaged not merely by incorrect operation of lube oil/seal oil system but also due to
sudden variation of load combined with an erroneous operation and by operation at
abnormally deviated design parameters that generate tremendous dynamic loads.
Instrumentation and protective devices are of vital importance. A failure of single
element or a breakdown of critical equipment can shut down the entire unit. History
sheets of failures and breakdowns for each and every machine should be maintained so
that breakdowns or failures can be analysed and thereby minimised in the long run by
preventive maintenance.