GAS COMPRESSION SYSTEMS DESIGN

MODELLING & SIMULATION METHODS

JAYANTHI VIJAY SARATHY

IChemE Membership No.
November 2013
Technical Report IChemE Membership No.

TABLE OF CONTENTS
1. Introduction ............................................................................................................................................ 3
2. Scope of Work ........................................................................................................................................ 3
3. Process Description .............................................................................................................................. 3
3.1 Process Conditions ........................................................................................................................... 3
3.2 Compressor Performance Curves ..................................................................................................... 4
3.3 Vendor Electric Motor Details ............................................................................................................ 5
3.4 Gas Properties at Flange Conditions ................................................................................................ 5
4. Design Methodology & Preliminary Estimates ................................................................................... 6
4.1 Design Approach ............................................................................................................................... 6
4.2 Equipment Sizing Philosophy ............................................................................................................ 6
4.3 Case Studies ..................................................................................................................................... 7
4.3.1 Start-up Operations ................................................................................................................. 7
4.3.2 Shutdown Operations .............................................................................................................. 8
4.4 Process Conditions - Design Summary............................................................................................. 8
5. Equipment Sizing ................................................................................................................................... 9
5.1 Equipment Volumes .......................................................................................................................... 9
5.2 Suction Side Volume ......................................................................................................................... 9
5.3 Discharge Side Volume ..................................................................................................................... 9
5.4 Settle Out Conditions ...................................................................................................................... 11
5.5 Anti-Surge Valve Sizing................................................................................................................... 12
6. Simulation Results .............................................................................................................................. 14
6.1 Shutdown Operations ...................................................................................................................... 14
6.2 Start-up Operations ......................................................................................................................... 15
6.2.1 CASE 1: Induction EM+CC configuration with Fixed ASV position at Start up ........................ 15
6.2.2 CASE 2: Induction EM+CC Configuration with Variable ASV Position ....................................16
6.2.3 CASE 3: VFD EM+CC configuration with fixed ASV position .................................................. 18
6.2.4 CASE 4: VFD-EM+CC Configuration with Variable ASV Position ........................................... 19
6.2.5 CASE 5: Induction EM+CC configuration with Suction Throttling ............................................ 21
6.2.6 CASE 6: VFD EM+CC configuration with Suction Throttling ................................................... 23
6.3 Start-up Power ................................................................................................................................ 25
7. Results & Conclusions ........................................................................................................................ 27
Appendix A: Design Summary ................................................................................................................ 28
A1. Polytropic Head ............................................................................................................................... 28
A2. Gas Outlet Temperature .................................................................................................................. 29
A3. Adiabatic Efficiency ......................................................................................................................... 29
A4. Inlet Volumetric Flow Rate .............................................................................................................. 29
A5. Compressor Duty ............................................................................................................................ 30
A6. Driver Duty ......................................................................................................................................30
A7. Cooler Duty .....................................................................................................................................30
Appendix B: Centrifugal Compressor Coast down Characteristics .................................................... 31
Appendix C: Anti-Surge Valve Sizing .....................................................................................................32
Appendix D: Settle-Out Conditions......................................................................................................... 33
Appendix E: Inertia Number – Determination of Hot Gas Bypass ....................................................... 35
Appendix F: Compressor Head Calculations......................................................................................... 37
Appendix G: Compressor Discharge Temperature ............................................................................... 41
References ................................................................................................................................................ 42

Page 2 / 42
Technical Report IChemE Membership No.

1. Introduction
It has been proposed to install a centrifugal compressor for the gas company ABC Corporation’s gas
gathering station. The compressor operating capacity is 8 MMscfd and past experience indicated operational
problems due to surge related issues and power inadequacy. This often resulted in loss of production and
increasing overhead costs in terms of maintenance, economic losses, etc. Operational problems often result
due to mechanical issues such as dry gas seal failure, lube oil system failure, etc or due to process
parameters such as operation envelope being violated, under sizing of anti-surge valve thereby reducing the
recycle flow, etc. In the current undertaking, the process parameters are addressed as a starting point to
eliminate any process related issues that cause operational failures.

2. Scope of Work
To abate the operational concerns, the compressor loop design is addressed to tackle surge flow related
issues that induce mechanical vibrations followed by a trip of the compressor driver. Additionally, restarting
compressors after settle out conditions demand higher start-up power and hence the compressor loop
configuration is also attended to in order to minimize start-up requirements. Employing a sparing philosophy
of one compressor working and one compressor on standby (i.e., 1W + 1S), due avoid production losses, a
transient analysis is made to understand the compressor’s behaviour during situations such as emergency
shutdown, normal shutdown and start-up scenarios.

3. Process Description
3.1 Process Conditions
The compressors operating capacity is 8 MMscfd with the suction flange conditions of 0.2 bar(g) and 250C
and a discharge flange pressure of 2.7 bar(g). The gas composition is shown in Table 1.

Table 1. Gas Composition

Component Mass fraction

C1 0.7857

C2 0.0818

C3 0.0480

iC4, n-C4 0.0316

iC5, n-C5 0.0196

C6 0.0234

Water 0.0098

Total 1.0000

Page 3 / 42
Technical Report IChemE Membership No.

3.2 Compressor Performance Curves

To achieve these conditions, a proposal is put forward to a certain compressor manufacturer firm which
responded with the following compressor curves. The head vs. flow rate curve is indicated in Fig. 3.2.1.

Figure 3.2.1. Head vs. Flow rate Curve

Since the surge line is not indicated, it is assumed as shown in Fig. 3.2.1. The surge margin is taken as 10%
and the surge line is fitted to an equation to obtain the fitted surge line. The polytropic efficiency vs. flow rate
curves is shown in Fig. 3.2.2.

Figure 3.2.2. Polytropic Efficiency vs. Flow Rate

Page 4 / 42
Technical Report IChemE Membership No.

3.3 Vendor Electric Motor Details

The driver used to drive the centrifugal compressor (CC) is an electric motor & the details are shown in
Tables 3.3.1 & 3.3.2 and Figure 3.3.1.
Table 3.3.1. Electric Motor Parameters
Parameter Value Unit

Electric Motor Type Asynchronous Induction Type [-]

Number of Poles 4 [-]

Power Source Frequency 50 [Hz]

Motor Slip 1 [%]

Rated Speed 1500 [rpm]

Table 3.3.2 Speed vs.

Torque characteristics
% Speed % Torque
0 55
10 50
20 45
30 50
40 55
50 60
60 70
70 85
80 105
97 150
99 1

Figure 3.3.1. EM Speed vs. Torque characteristics

3.4 Gas Properties at Flange Conditions

The following properties exist at the flange conditions of the centrifugal compressor (CC) [Table 3.4].
Table 3.4. Gas composition after suction separator
Sr. No. Property Value Unit

1 Molecular weight of gas (MW) 18.38 kg/kgmol

2 Fluid density (r1) at suction flange 0.9153 kg/m3

3 Fluid density (r2) at discharge flange 2.03 kg/m3

4 Heat capacity at suction side 2.138 kJ/kg0C

5 Heat capacity at discharge side 2.474 kJ/kg0C

6 Compressibility factor (Z1) at suction flange 0.9964 -

7 Compressibility factor (Z2) at discharge flange 0.9964 -

8 Ratio of specific heats at suction side 1.2740 -

9 Ratio of specific heat at discharge side 1.2304 -

Page 5 / 42
Technical Report IChemE Membership No.

4. Design Methodology & Preliminary Estimates

4.1 Design Approach
The general design approach consists of performing steady state calculations, i.e., a heat & mass balance to
estimate the process conditions as well as sizing the equipments, valves & lines. In order to do so, prior
information about the plant layout is needed. However since in the current undertaking, no layout information
is available, the approach envisaged consists of performing a transient simulation for approximate volumes
to avoid a surge during start-up & shutdown followed by shaping these volumes into detailed piping &
equipment estimates. A general schematic of the compressor loop envisaged is shown in Fig 4.1.1.

Figure 4.1.1. Proposed CC System Configuration

The compressor loop design is analysed for two possible configurations which are,

Table 4.1.1. Compressor Operating Configuration

Sr. No. Compressor Configuration Remarks

1 Fixed Speed Asynchronous Induction Type Electric Motor

2 Variable Speed Variable Speed Frequency Drive (VFD-EM)

4.2 Equipment Sizing Philosophy

The equipment sizing philosophy employed as part of the design approach involves generating preliminary
estimates that are fine tuned further using transient simulation as a tool. This is shown as a high level flow
diagram below in Figure 4.2.1.

Figure 4.2.1. High Level Flow Diagram for Preliminary Compressor Loop Design Estimates
The driver proposed is an induction electric motor (EM) which is coupled to the compressor via a gearbox.
The upstream and downstream of the compressor suction & discharge flange respectively comprises of
equipment, piping & instrumentation elements which offer a certain amount of pressure drop that are fixed
speculatively. A summary of the pressure drops offered are shown in the table 4.2.1.
Table 4.2.1. Assumed Pressure Drops
From To Assumed Max DP [bar]
Suction Battery Limit (SBL) Compressor Suction Flange 0.6
Compressor Discharge Flange Suction Check Valve 0.5

Page 6 / 42
Technical Report IChemE Membership No.

In order to estimate the plant parameters, a steady state heat & mass balance simulation is initially made in
Aspen HYSYS (Trademark of Aspentech Technology Inc.) to determine the stream properties from the
boundary conditions to the flanges.

Figure 4.2.1. Proposed CC system Schematic

4.3 Case Studies

The compressor loop design is checked for stability in terms of surge and power adequacy based on the
situations where it experiences the maximum amount of uncertainty, i.e., during start-up and shutdown
operations. If the compressor loop with it’s anti-surge system and power driver capacity is designed
adequately, the surge phenomenon can be effectively abated and start-up power requirements are also met.

4.3.1 Start-up Operations

The compressor loop design is checked for stability in terms of surge and power adequacy based on the
centrifugal compressor start up modes can be varied to accomplish lower start-up power consumption. This
can be applied to both directly coupled EM-CC configuration as well as VFD-EM configurations. Though a
number of configurations are available, in the current undertaking, the following configurations are studied,

Table 4.3.1. Start-up Operations – Case Studies

Electric Motor Configuration Anti-surge Valve (ASV) Position
Start-up
Suction Throttling
Case No.
Fixed Speed Variable Speed Fixed ASV Variable ASV

1 Ö Ö
2 Ö Ö
3 Ö Ö
4 Ö Ö
5 Ö Ö
6 Ö Ö

Page 7 / 42
Technical Report IChemE Membership No.

4.3.2 Shutdown Operations

The two types of shutdown operations can be classified as normal shutdown (NSD) during which the plant
operator initiates a planned shutdown while an emergency shutdown (ESD) which occurs during an
emergency event such as a driver trip due to power failure or excessive compressor vibration. The cases
that are analysed here are as follows,
Table 4.3.2. Shutdown Operations – Case Studies

Case Event 1 Event 2

Normal Shutdown (NSD) Anti-Surge Valve (ASV) Opening Compressor Shutdown1

Emergency Shutdown (ESD) Compressor Shutdown1 Anti-Surge Valve (ASV) Opening

Note 1: Compressor Shutdown occurs due to Driver (Electric motor) tripping

4.4 Process Conditions - Design Summary

Based on the gathering station compressor data, for an inlet pressure of 0.2 bar(g) and 250C and a
discharge pressure of 2.7 bar(g), the design parameters are calculated as shown in Table 4.3.1. Detailed
calculations are shown in Appendix A.
Table 4.3.1. Design Summary

Parameter Value Unit

Inlet Flange Pressure 0.2 bar(g)

Outlet Flange Pressure 2.7 bar(g)

Polytropic Head 177.23 kJ/kg

0
Gas Outlet Temperature 138.3 C

Adiabatic Efficiency 67.10 %

Inlet Volume Flow 8100 Act_m3/hr

Compressor Duty 559 kW

Driver Rated Duty 600 kW

Cooler Duty 427 kW

Page 8 / 42
Technical Report IChemE Membership No.

5. Equipment Sizing
5.1 Equipment Volumes
The main equipment that contributes to the compressor loop volumes are suction scrubber and discharge
cooler. In situations where gas condensation occurs after the compressor discharge an additional discharge
scrubber is installed to knock out any liquids. The discharge side volumes particularly affect the response
time of anti-surge system. Excessive discharge side volumes tend to make the anti-surge system sluggish
and result in a delay in recycling the gas. Hence discharge side volumes are to be kept as minimum as
possible. It must be ensured that the anti-surge take-off point before the discharge side check valve and
compressor discharge flange is to be chosen such that it is neither too close to the compressor discharge
flange (which can cause the anti-surge valve to rattle) inducing noise related issues while too far increases
the anti-surge response time. The rate at which the compressor coast down, i.e., speed decay occurs, also
determines the size of the anti-surge valve that regulates the amount of the recycle flow. Based on the above
considerations, the equipment volumes are calculated as described in the next section.

5.2 Suction Side Volume

Based on the Author’s experience, the suction side volume is calculated initially for twice the rated volume
flow.
Vsuction , pre lim inary = Qsuction scrubber ´ Margin (1)

3
æ 8100 ö m
Or, Vsuction , pre lim inary = ç ÷ ´ 2 = 4.5m 3 (2)
è 3600 ø sec
Therefore on the suction side, an approximated volume of 4.5 m3 is taken. This is a preliminary estimate that
is subjected to change depending on the plant layout.

5.3 Discharge Side Volume

The discharge volume needs to be predicted for the worst situation and this happens to occur during an
emergency shutdown (ESD) which is dependent on the decay rate of the compressor speed. Therefore the
compressor coast down characteristics which is a function of the compressor and electric motor inertia &
power rating needs to be analysed to estimate the discharge side volume. From the electric motor data
which is a 4 pole model with a power source at the plant site existing at 50Hz, the synchronous speed of the
motor is,
120 ´ frequency 120 ´ 50
N synchronous = = = 1500 rpm (3)
Poles 4
From the steady calculations, a 600 kW induction type electric motor has been proposed to be installed. The
torque required to sustain the compressor at the rated conditions is,
2pNT
P= kW (4)
60 ´1000
2 ´ p ´ 1485 rpm ´ (T )N × m 559 ´ 60 ´ 1000
559kW = ÞT = » 3595 N × m (5)
60 ´ 1000 2 ´ p ´ 1485
Hence the electric motor has to provide a torque of 3595 kg-m2 to sustain the compressor at 3000 rpm &
meet the rated suction & discharge conditions. The maximum torque that can be provided by the electric
motor before it breaks down is calculated at the breakdown torque’s corresponding speed

Page 9 / 42
Technical Report IChemE Membership No.

N = 0.97 ´1485 = 1440.45 rpm (6)

600 ´ 60 ´ 1000
Therefore, Tbreak down = ´ 1.5 = 5966.45 N × m (7)
2 ´ p ´ 1440.45
Hence the power absorbed for 5966.45 N-m is 900 kW at a speed of ~1440 rpm during start-up. The inertia
offered by the electric motor is calculated by the empirical relationship as [Ref 7],
1.48
æPö
I = 0.0043ç ÷ (8)
èNø
Where, P = Power [kW]
I = Inertia [kg-m2]
N = Speed [rpm/1000]
Therefore, the inertia offered by the electric motor is,
1.48
é ù
ê 600 ú
I = 0.0043 ´ = 30.97 kg × m 2 (9)
ê1485 ú
ë 1000 û
Therefore the inertia offered by the electric motor (EM) is 30.97 kg-m2. In order to scale up the speed of the
compressor, a gearbox is installed. The gear ratio is,
3000
GR = = 2.0202 (10)
1485
Hence the gear ratio is 2.0202. Assuming that the rotor mass is about 25,000 kg with an average radius of
gyration of 100 mm, the rotor is taken to offer inertia of 250 kg-m2 (I=MR2). Hence the total inertia required to
be overcome by the EM is the sum of CC rotor inertia, EM inertia, Gear box Inertia & Gas inertia during start-
up & settling out conditions. The inertia offered by the gas is a variable since it changes with the
depressurization rate.
I Total = I CC rotor + I EM + I Gearbox + I Gas , startup / SOP Þ I Total = 250 + 30.97 + I Gearbox + I Gas , startup / SOP (11)

I Total = 280.97 kg × m 2 + I Gearbox + I Gas , startup / SOP (12)

Since the inertia offered by the gearbox is not available, it is neglected. However in real time applications this
must be considered. For calculation purposes, the equipment inertia is rounded-off to 300 kg-m2.

Based on the Author’s experience, the discharge side volume can be taken as approximately 1/3 times the
suction volume. Therefore taking the discharge volume that is less than suction volume = 1/3rd of 4.5 m3 is
1.5 m3. The decay rate of the rotor speed is calculated using Eq. 13 (Refer Appendix B)
60 P 60 ´ 496 ´ 1000
k= 3
N × m × min 2 Þ k = = 0.0011 022 N × m × min 2 (13)
N 3000 3
The calculation was performed with the power as 559 kW since this is the initial condition from where the
CC’s speed will decay. Therefore, the decay rate depends on the power consumed by the CC as well as the
inertia of the system at the time of ESD. Substituting them in Eq. (B.10) (Appendix B),
1 1
N (t ) = Þ N (t ) = (14)
1 216000 ´ 0.001102 ´ (t - 0 ) 0.000334 + 0.02t
+
3000 (2p ) ´ 300
2

Page 10 / 42
Technical Report IChemE Membership No.

Plotting Eq. (14), for a time period,

Figure 5.3.1. CC Speed Decay Rate Curve

From the above plot, it is seen that the compressor speed reaches zero speed at infinity as eq. 14 produces
an asymptotic curve. Hence a criterion is set based on the rate of change of speed, i.e., when the speed
gradient does not change appreciably, it is considered that the compressor comes to complete halt. The
speed gradient based on the author’s experience on various compression systems can be taken to be
between 0.1% to 0.15%. Considering a slope change criterion of 0.15%, the compressor comes to a halt in
about 80 seconds.

5.4 Settle Out Conditions

During coast down period, the suction & discharge pressures reach the settle out pressure. The settle out
temperature is estimated (Refer Appendix D) as,

(m1c p T1 + m2 c p T2 ) - QCooler [(V1 r1 )1 c p T1 + (V2 r 2 )2 c p T2 ] - QCooler

SOT = 1 2
= 1 2
(15)
m1c p1 + m 2 c p2 (V1 r1 )1 c p
1
+ (V2 r 2 )2 c p2

SOT =
[(4.5 ´ 0.9153) ´ 2.138 ´ 298] + [(1.5 ´ 2.03) ´ 2.474 ´ 411.5] - 426.9 = 324 K » 510 C (16)
[(4.5 ´ 0.9153) ´ 2.138] + (1.5 ´ 2.03) ´ 2.474
The settle out temperature calculation is made considering the cooler. In reality the discharge cooler
operation could differ since it can have variable speed facilities as well as ESD / NSD procedure during coast
down causes the cooling temperature to vary. The settle out pressure for 510C is therefore,

(4.5 ´ 0.9153) + (1.5 ´ 2.03) ´ 0.9964 ´ 0.08314 ´ 324

n ´ Z avg ´ R ´ SOT
SOP = = 18.38 = 1.74 bar (a ) (17)
VTotal (4.5 + 1.5)

Page 11 / 42
Technical Report IChemE Membership No.

5.5 Anti-Surge Valve Sizing

To estimate the anti-surge valve Cv, the ASV inlet pressure and outlet pressure for a flow rate must be
known when the valve is opened 100%. Normally a CC vendor documents would provide the compressor
performance curves as a plot of discharge pressure vs. flow rate from which the discharge pressure of the
compressor minus the inline equipment pressure drop is considered as the ASV’s upstream pressure & the
compressor suction conditions as the ASV downstream pressure.

Considering the maximum possible flow through the compressor is at the stonewall region, the anti-surge
valve can be sized for this flow. However in reality it is to be avoided to push the compressor to its limits and
hence a margin of 10% to 15% on the stonewall flow at 3000 rpm is taken to ensure that the anti-surge valve
does not recycle excess fluid back to the suction side.

Figure 5.5.1. Maximum CC suction flow considered for ASV sizing

Using HYSYS, for a maximum suction flow of 10,440 m3/hr the ASV conditions from HYSYS are

1. ASV Inlet Conditions à Pressure = 2.715 bar(a) ; Temperature = 80.980C = 354.13 K

2. ASV Outlet Conditions à Pressure = 2.521 bar(a); Temperature = 80.910C = 354.06 K

3. ASV Actual flow rate = 5766 m3/hr

4. Standard Flow rate (1.01325 bar(a) & 15.60C) =

5766 ´ 2.715 288.6

Q ASV ,in , Normal = ´ = 12591 S m 3 hr (18)
354.13 1.01325
Therefore the ASV Cv can be computed using Eq. (C.4) in a rearranged manner as,

é æ 1.4 X öù X
QV = 2250 ´ C v (t ) ´ p1 ´ ê1 - çç ÷÷ú ´ (19)
ë è 3kX T øû (MW ´ T1 ´ Z )
Taking X = XT in the expansion factor term,

Page 12 / 42
Technical Report IChemE Membership No.

é æ 1.4 öù X
QV = 2250 ´ C v (t ) ´ p1 ´ ê1 - ç ÷ú ´ (20)
ë è 3k øû (MW ´ T1 ´ Z )
Rearranging Eq. (20),

é ù
ê ú
CV = ê Qv ú é MW ´ T1 ´ Z ù (21)
ê é æ 1.4 öù ú êë X ú
û
ê 2250 p1ê1 - ç ÷ ú ú
ë ë è 3k øû û

é ù
12591 ê18.38 ´ 354.13 ´ 0.9964 ú
CV = ê ú (22)
é æ 1.4 öù ê æ 2.715 - 2.521 ö ú
2250 ´ 2.715 ´ ê1 - ç ÷ú ê ç ÷ ú
ë è 3 ´ 1.274 øû ë è 2.715 ø û

CV = 979.9 » 980 (23)

The valve chosen has a Cv of 980 USGPM.

Page 13 / 42
Technical Report IChemE Membership No.

6. Simulation Results
With the data consolidated based on the previous sections, Aspen HYSYS Dynamics is employed to perform
transient calculations to analyse the compressor performance for surge related issues during start-up and
shutdown scenarios. The start-up power for various start-up modes are also investigated and summarized in
this section.

6.1 Shutdown Operations

For the case of normal shutdown (NSD) and emergency shutdown (ESD), the operating point migration is
tracked to ensure that it does not cross the surge line. During a normal shutdown, the anti-surge valve is first
opened to allow gas from the discharge side to move towards the suction side and ensures that the
operating point moves away from the surge line. Once sufficient margin on the flow is available, the
compressor is tripped thereby ensuring that the gas flows into the suction side and avoids a flow reversal
into the compressor that causes a surge. The transient plots are indicated in Fig 6.1.1 which shows that the
compressor does not operate in the surge region at any point in time during the compressor coast down.

Figure 6.1.1. Compressor Operating Point Migration during NSD & ESD

Page 14 / 42
Technical Report IChemE Membership No.

6.2 Start-up Operations

For the various start-up modes described in section 4.3.1, the start-up power required and the operating
point migration are studied.

6.2.1 CASE 1: Induction EM+CC configuration with Fixed ASV position at Start up
Calculations performed previously reveal that surging does not occur during either an emergency shutdown
or a normal shutdown; therefore the next step is to check for surge during start-up. The start-up procedure
differs with compressor configurations. The start-up scenario for the current configuration involves,

1. CC Speed = 0 rpm at t=0 sec.

2. ASV is 100% open while suction block valves & discharge block valves are kept completely closed.
3. EM speed is ramped up that follows the curve as per Fig. 3.3.1 at t > 0.
4. Suction & Discharge block valves start to open at t = 58.75 sec and reach 100% opening at 98.75 sec.
5. ASV starts to close at t =78.75 sec (chosen iteratively), till t = 128.75 sec when it reaches 0% opening.
6. The simulation is run till t = 200 sec while the compressor reaches the speed of 3000 rpm at ~188 sec.
7. A plot of the operating point migration is shown in Fig. 6.2.1.1 which shows that no surge occurs during
start-up for the ASV size & operational philosophy selected.

Figure 6.2.1.1. Operating Point Migration during 100% ASV Start-up Mode

Page 15 / 42
Technical Report IChemE Membership No.

6.2.2 CASE 2: Induction EM+CC Configuration with Variable ASV Position

The variable ASV position is attained using an Anti-surge controller (ASC) which regulates the opening of the
anti-surge valve. This ensures that during start-up no surging occurs. Normally at about 70% of speed, the
anti-surge control (ASC) is put into automatic mode and does not require human intervention. For the current
analysis, 50% speed is chosen at which the CC operates in Auto mode. This is achieved as follows,

1. CC Speed = 0 rpm at t=0 sec

2. ASV is 60% open while suction block valves & discharge block valves are kept completely closed.
3. ASV is operated manually during start-up after which the surge controller is operated in Auto mode at
50% speed.
4. EM speed is ramped up that follows the curve as per Fig. 3.3.1 at t > 0.
5. Suction & Discharge block valves start to open at t = 180 sec and reach 100% opening at 200 sec.
6. ASV changes to Auto mode from manual operation at 1500 rpm & reaches 0% opening at gas export
condition.
7. The simulation is run till t = 210 sec since the CC reaches the rated rpm of 3000 rpm at about 200 sec. A
plot of the operating point migration is shown in Fig. 18 which shows that no surge occurs during start-up
for the ASV size & operational philosophy selected.

Figure 6.2.2.1. EM+CC Start up curve with Anti-surge Controller

The variation ASV opening as controlled by ASC is shown in Fig. 6.2.2.2

Page 16 / 42
Technical Report IChemE Membership No.

Figure 6.2.2.2. Variation in ASV Opening – Case 2

Page 17 / 42
Technical Report IChemE Membership No.

6.2.3 CASE 3: VFD EM+CC configuration with fixed ASV position

Variable frequency drives (VFD) are used when a variation in speed is required during CC operation. The
usage of this is dictated based on the application of the compression systems. In most occasions when a
compressor re-start is not feasible due to settling out conditions & requires depressurization, VFD’s prove to
be effective. This is achieved through a controlled start-up of the electric motor by altering the frequency of
the current passing through the EM which thereby raises the CC speed gradually. Additionally VFD’s
consume lower power during start-up. In the current study, a VFD is used to start the compressor in the
following way,

1. CC Speed = 0 rpm at t=0 sec

2. ASV is 60% open while suction block valves & discharge block valves are kept completely closed.
3. EM speed is ramped up by a predetermined time period. In this case, a start-up time period of 180 sec is
chosen.
4. Suction & Discharge block valves start to open at t = 150 sec and reach 100% opening at t = 170 sec.
5. ASV starts to close at t =170 sec (chosen iteratively), till t = 180 sec when it reaches 0% opening.
6. The simulation is run till t = 250 sec since the CC reaches the rated rpm of 3000 rpm at ~250 sec. A plot
of the operating point migration is shown in Fig. 6.2.3.1 which shows that no surge occurs during start-up
for the ASV size & operational philosophy selected.

Figure 6.2.3.1. VFD-EM+CC Start up with ASV at 60%

Page 18 / 42
Technical Report IChemE Membership No.

6.2.4 CASE 4: VFD-EM+CC Configuration with Variable ASV Position

The variable ASV position is achieved by using an Anti-surge controller (ASC). The start-up of a VFD-
EM+CC with variable ASV position configuration involves,

1. CC Speed = 0 rpm at t=0 sec

2. ASV is 60% open while suction block valves & discharge block valves are kept completely closed.
3. ASV is operated manually during start-up after which the surge controller is operated in Auto mode at
50% speed.
4. EM speed is ramped up by a predetermined time period. A start-up time period of 180 sec is chosen.
5. Suction & Discharge block valves start to open at t = 150 sec and reach 100% opening at 170 sec.
6. ASV changes to Auto mode from manual operation at 1500 rpm & reaches 0% opening at gas export
condition.
7. The simulation is run till t = 250 sec since the CC reaches the rated rpm of 3000 rpm at about 250 sec. A
plot of the operating point migration is shown in Fig. 6.2.4.1 which shows that no surge occurs during
start-up for the ASV size & operational philosophy selected.

Figure 6.2.4.1 VFD-EM+CC Start up with ASC Mode

The ASV opening varies as shown in Fig. 6.2.4.2.

Page 19 / 42
Technical Report IChemE Membership No.

Figure 6.2.4.2 Variation in ASV Opening – Case 4

Page 20 / 42
Technical Report IChemE Membership No.

6.2.5 CASE 5: Induction EM+CC configuration with Suction Throttling

Suction throttling methods involve the use of a globe valve or a butterfly valve at the suction line of the CC to
reduce the amount of fluid entering the compressor and simultaneously the start-up power is also decreased.
However a limitation exists on the operational range in terms of power reduction since reducing the flow rate
causes the operating point to move towards the surge line during start-up and hence closing of the throttling
valve must be regulated. This method of start-up is achieved as follows,

1. CC Speed = 0 rpm at t=0 sec

2. ASV is 100% open while suction block valves & discharge block valves are kept completely closed.
Throttle valve is kept at 15% opening.
3. EM speed is ramped up and follows as per Fig. 3.3.1 at t > 0.
4. Suction & Discharge block valves start to open at t = 58.75 sec and reach 100% opening at 98.75 sec.
5. ASV starts to close at t =78.75 sec (chosen iteratively), till t = 128.75 sec when it reaches 0% opening.
6. Throttle valve starts to open at time t = 100 sec and reaches 50% at t = 150 sec.
7. The simulation is run till t = 200 sec since the CC reaches the rated rpm of 3000 rpm at ~190 sec. A plot
of the operating point migration is shown in Fig. 6.2.5.1 which shows that no surge occurs during start-up
for the ASV size & operational philosophy selected.

Figure 6.2.5.1. EM+CC Start up with Suction Throttling

A comparative study can be made to understand the start-up patterns between case 1 and case 5 in which
both the start-up steps are similar with the difference being in use of suction throttling method absent in case
1 while it is present in case 5. Figure 6.2.5.2 shows this.

Page 21 / 42
Technical Report IChemE Membership No.

Figure 6.2.5.2. Comparative Study between Case 2 & Case 5 – Start-up Operating Point Migration

Page 22 / 42
Technical Report IChemE Membership No.

6.2.6 CASE 6: VFD EM+CC configuration with Suction Throttling

As performed in Case 5, the start-up can be performed for a VFD-EM system in a similar manner with
appropriate modifications. The start-up involves the following steps,

1. CC Speed = 0 rpm at t=0 sec

2. ASV is 100% open while suction block valves & discharge block valves are kept completely closed.
Throttle valve is kept at 15% opening.
3. ASV is operated manually during start-up after which the ASC is operated in Auto mode at 50% speed.
4. EM speed is ramped up by a predetermined time period. A start-up time period of 180 sec is chosen.
5. Suction & Discharge block valves start to open at t = 125 sec and reach 100% opening at 145 sec.
6. ASV changes to Auto mode from manual operation at 1500 rpm & reaches 0% opening at gas export
condition.
7. The simulation is run till t = 250 sec since the CC reaches the rated rpm of 3000 rpm at about 250 sec. A
plot of the operating point migration is shown in Fig. 6.2.6.1 which shows that no surge occurs during
start-up for the ASV size, piping volumes & operational philosophy selected.

Figure 6.2.6.1. VFD-EM+CC Start up with Suction Throttling

A comparative study is made between case 4 & case 6 to check for the effect of start-up throttling. This is
shown in Fig. 6.2.6.1.

Page 23 / 42
Technical Report IChemE Membership No.

Figure 6.2.6.2. Comparative Study between Case 4 & Case 6 – Start-up Operating Point Migration

Page 24 / 42
Technical Report IChemE Membership No.

6.3 Start-up Power

For the various cases performed the power absorbed during start-up is also plotted. Figure 6.3.1 shows a
summary of the power absorbed by the CC during start-up.

Figure 6.3.1. CC Start-up Absorbed Power – All Cases

The compressor speed plotted against time is shown in Fig. 6.3.2.

Figure 6.3.2. Compressor Speed vs. Time – All Cases

Page 25 / 42
Technical Report IChemE Membership No.

Table 6.3.1 provides a summary of the peak power absorbed in the 6 cases studied.

Table 6.3.1. Summary of CC Start-up Absorbed Power

Peak Absorbed % Power
Case Description Power Savings
[kW] [%]
1 (Base Case) Induction EM+CC - fixed ASV position at Start up 869 0
2 Induction EM+CC - Variable ASV position 866 0.3
3 VFD-EM+CC - fixed ASV position 711 18.2
4 VFD-EM+CC - variable ASV position 666 23.4
5 Induction EM+CC - suction throttling 865 0.5
6 VFD EM+CC - suction throttling 666 23.4

Figure 6.3.3. Compressor Start-up Power Comparison

From Table 6.3.1, it is seen that for cases 1, 2 & 5 which use an induction motor coupled with Gearbox
arrangement, the start-up power is higher while for cases 2, 3 & 6 which use a VFD, the start-up power is
lower by ~23%. Between cases 3, 4 & 6 the start-up power varies very marginally due to the mode of start-
up used (such as suction throttling or ASV regulation). In Fig. 6.3.2, it is to be noted that the speed vs. time
curves for cases 3 & 4 overlap with each other and hence similar. The same trend is also present between
cases 5 & 6 and hence the curves overlap with each other.

Though VFD’s help in saving start-up power, they also represent a capital cost. Hence a careful economic
study must be made to understand the capital & operating costs along with a transient simulation to
understand if VFD’s are economically feasible.

Page 26 / 42
Technical Report IChemE Membership No.

7. Results & Conclusions

From the calculations and simulations performed, it is seen that the methods employed to design a gas
compression system provide a robust measures against surge related issues during normal shutdown (NSD)
and Emergency Shutdown (ESD). The anti-surge valve size is sufficient enough to prevent the compressor
operating point on the compressor curves from crossing the surge limit line and indicates that a hot gas
bypass system is not required. Factors that contribute to the compressor surge such as compressor speed
coast down time, equipment volumes, anti-surge valve size are addressed and their contribution to the surge
phenomenon have been addressed and described in Section 6.1.

For start up operations, surge is abated for the anti-surge valve (ASV) size chosen. The start-up powers are
also estimated for various cases and is concluded that a variable frequency or variable speed drive
compressors with either suction throttling or variable anti-surge valve positioning during start-up is effective
to reduce start-up power by consuming about 666 kW. This is so since due to throttling of gas during start-up
the gas pressure across the suction throttle valve/anti-surge valve decreases causing the gas temperature to
drop. Subsequently the lowering of gas temperature causes the gas density to increase. For an increased
gas density, the polytropic head decreases which translates into less power for the same compression ratio.
To summarize a colder gas causes the centrifugal compressor to consume less power during start-up.

It is also seen that in the current undertaking, a compressor restart after ESD/NSD shutdown, does not
require the compressor loop to be depressurized as the 600 kW electric motor is sufficiently rated enough to
start the compressor after settle out conditions.

Although in the current undertaking a hot gas bypass (HGB) is not used to design the compressor loop, a
procedure based on dimensionless numbers is provided in Appendix E to qualitatively estimate the
requirement of a hot gas bypass.

Page 27 / 42
Technical Report IChemE Membership No.

Appendix A: Design Summary

A1. Polytropic Head
From the gathering station compressor data, the polytropic head produced is calculated from the basic
thermodynamic equation (Refer Appendix F),

é n -1
ù
Z avg RT1 æ n ö êæ P2 ö n
Hp = ç ÷ ç ÷÷ - 1ú (A.1)
MW è n - 1 ø êçè P1 ø ú
ëê ûú
Where,
H p = Polytropic head produced (ft)

Z avg = Average gas compressibility factor

T1 =Inlet temperature of gas at suction flange (K)

n = Polytropic volume exponent
P2
= Pressure ratio at flange conditions
P1
To estimate the polytropic volume exponent, the following calculation is made.
n k
= ´h p (A.2)
n -1 k -1
Where, h p = Polytropic Efficiency

The specific heat ratio considered is the average specific heat between suction & discharge flange
k1 + k 2 1.274 + 1.2304
k avg = = = 1.2522 (A.3)
2 2
The polytropic efficiency is taken from Fig. 3.2.2, where the efficiency at which the compressor operates is
~71.84%. Therefore substituting in Eq. (2),
n 1.2522
= ´ 0.7184 Þ n = 1.3896 (A.4)
n - 1 1.2522 - 1
An alternate way to calculate the polytropic exponent is by using the equation

æP ö
lnçç 2 ÷÷
P
n= è 1ø (A.5)
ær ö
lnçç 2 ÷÷
è r1 ø
æP ö
lnçç 2 ÷÷ lnæç 3.71325 ö÷
P 1.21325 ø
Therefore, n= è 1ø = è = 1.4043 (A.6)
æ r2 ö æ 2.030 ö
lnçç ÷÷ lnç ÷
è r1 ø è 0.9153 ø
It is seen that the value of ‘n’ calculated using Eq. (5) is almost equal to the value calculated using Eq.
(2). However, since the value of ‘n’ was estimated using a graphically calculated value of hp as in Eq. (4),
the value obtained using Eq. (6) is considered for calculations.

Page 28 / 42
Technical Report IChemE Membership No.

Using the values calculated above, the polytropic head produced is therefore,

é 1.4043-1
ù
=
[(0.9964 + 0.9964) 2]´ 1545.35 ´ 536.67 æ 1.4043 ö êæ 3.71325 ö 1.4043
- 1ú
Hp ç ÷ ç ÷ (A.7)
18.38 è 1.4043 - 1 ê
ø è 1.21325 ø ú
ë û
H p = 59334.33 ft » 18085m » 177.23 kJ / kg (A.8)

A2. Gas Outlet Temperature

The outlet temperature of the gas is calculated as (Refer Appendix G),
n -1
T2 æ P2 ö n æ Z1 ö
=ç ÷ çç ÷÷ (A.9)
T1 çè P1 ÷ø è Z2 ø
Therefore,
1.4043-1
æ 3.71325 ö æ 0.9964 ö
÷ ´ (273.15 + 25)
1.4043
T2 = ç ÷ ´ç (A.10)
è 1.21325 ø è 0.9964 ø
T2 = 411.5K » 138.30 C (A.11)

A3. Adiabatic Efficiency

The adiabatic efficiency is calculated as [Ref 7],

é n -1
ù
æ n ö êæ P2 ö n
ç ÷ êçç ÷÷ - 1úú
n - 1 ø è P1 ø
hp è êë úû
= (A.12)
hA é k -1
ù
æ k ö êæ P2 ö k
ç ÷ ç ÷ - 1úú
ç ÷
è k - 1 ø êè P1 ø
êë úû

é 1.4043-1
ù
æ 1.4043 ö êæ 3.71325 ö 1.4043
ç ÷ ç ÷ - 1ú
1.4043 - 1 ê 1.21325 ú
hp è ø è
ë
ø
û = 1.3197 = 1.0703
= (A.13)
hA é 1.2522 -1
ù 1.233
æ 1.2522 ö êæ 3.71325 ö 1.2522 ú
ç ÷ ç ÷ - 1
è 1.2522 - 1 ø êè 1.21325 ø ú
ë û
0.7182
hA = » 67.10% (A.14)
1.0703

A4. Inlet Volumetric Flow Rate

The inlet volumetric flow rate is calculated as,
m
Q1 = (A.15)
r1
7413
Q1 = » 8100 m 3 hr (A.16)
0.9153

Page 29 / 42
Technical Report IChemE Membership No.

A5. Compressor Duty

The power absorbed by the compressor is therefore [Ref 7],
Hp ´m
P= (A.17)
hp
177.23 ´ 7413
P= = 1829304 kJ hr (A.18)
0.7182
Or, P = 508.2kJ / sec » 508kW (A.19)

A6. Driver Duty

Assuming mechanical losses + margin of 10% of the absorbed power (Actual value to be confirmed by Vendor),
the estimated power requirements is
P = 508.2 ´ 1.1 (A.20)
P » 559kW at 3000 rpm (A.21)

In actual practice, the power value selected as shown above might not meet the required power to start the
compressor from settling out conditions. The selected driver is a 600 kW induction electric motor to round –off
the calculated value.

A7. Cooler Duty

In order to cool the gas from the compressor discharge to the required process temperature 500C, the cooler
duty is calculated as
Q = mC p ,avg (T2 - T3 ) (A.22)

7413 (2.473 + 2.220 )

Q= ´ (411.5 - 323.15) (A.23)
3600 2
Q » 426.9kW » 1.537 ´ 10 6 kJ hr (A.24)

Page 30 / 42
Technical Report IChemE Membership No.

Appendix B: Centrifugal Compressor Coast down Characteristics

The decay rate of driver speed is governed by the inertia of the system consisting of the compressor, coupling,
gearbox & driver, which are counteracted by the torque transferred to the fluid [Ref 1]. Also neglecting the
mechanical losses, we have,

æ dN ö
T = -(2p )J ç ÷ [N-m] (B.1)
è dt ø
éJ M ù
Where, J = System Inertia (Compressor + gearbox + driver) [kg-m2] = J = J C + ê 2ú
ë (Gear Ratio ) û
-1
N = Speed of Compressor Rotor [rpm] or [min ]
The rate of speed decay as well as the system inertia determines the torque decay of the compressor.
Therefore the power transferred to the gas, is

P = (2pNT ) N × m (B.2)
min
Substituting Eq. (C.1) in Eq. (C.2), the power transferred during emergency shutdown (ESD) is,

é æ dN öù
P = 2pN ´ ê- (2p )J ç ÷ú (B.3)
ë è dt øû
Applying fan power law as an approximation in which ‘k’ is relatively unvarying for a given curve,

3 60 P
P µ N Þ P = kN ; k = 3 N × m × min Þ P =
23 kN 3
60
( ) (B.4)
N
Substituting Eq. (C.4) in Eq. (C.3), we have

kN 3 é æ dN öù
= 2pN ´ ê- (2p )J ç ÷ú (B.5)
60 ë è dt øû

dN kN 2 æ N ö dN kN 2 æ kg × m ö
Rearranging, = ç
ç ÷
÷ Þ = ç ÷÷ (B.6)
dt - (2p ) J ´ 60 è kg × m ø
2
dt - (2p ) J ´ 60 çè kg × m × sec 2
2
ø
Integrating Eq. (C.6), and also multiplying by (602) to convert sec2 (rev/s) to min2 (rev/min)
N = N (t ) t =t
dN k ´ 60
ò
N = N0 N 2
=
- (2p ) J
2 ò dt
t0
(B.7)

N (t ) N (t )
é N - 2+1 ù k ´ 60 é1ù 60k (t - t 0 )
ê ú = ´ (t - t 0 ) Þ ê ú = (B.8)
ë - 2 + 1û N - (2p ) J (2p )2 J
2
0
ë N û N0

1 1 60k (t - t 0 ) 1 1 60k (t - t 0 ) 1
- = Þ = + Þ N (t ) = (B.9)
N (t ) N 0 (2p ) J
2
N (t ) N 0 (2p ) J
2
1 60k (t - t 0 )
+
N0 (2p )2 J
Where, N0 is the speed of the compressor before ESD. The second denominator term exists with the units
N.m.min/kg.m2 & this is converted to min-1 which gives,
1
N (t ) = (B.10)
1 216000k (t - t 0 )
+
N0 (2p )2 J

Page 31 / 42
Technical Report IChemE Membership No.

Appendix C: Anti-Surge Valve Sizing

The flow of gas through the anti-surge line is estimated using the ANSI/ISA S75.01 standard equation as,

é X ù
QV = 2250C v (t )F p p1Y ê ú (C.1)
ë MW ´ T1 ´ Z û
Where, Fp = Piping geometry factor (Usually taken to be 1)
X
Y = Gas expansion factor è Y = 1 -
3Fk X T
Dp
X = Pressure drop ratio è X = where, Limit X = Fk X T
p1
k
Fk = Ratio of specific heats è Fk =
1.4
X T = Pressure drop ratio factor
Rewriting the equations by substituting for Fp, X &Y in Eq. (C.1),

æ Dp ö Dp
ç ÷
p1 p1
QV = 2250 ´ C v (t ) ´ p1 ´ ç1 - ÷´ (C.2)
ç æ k ö ÷ MW ´ T1 ´ Z
ç 3´ ç ÷XT ÷
è è 1.4 ø ø
æ p2 - p1 ö p2 - p1
ç ÷
p1 p1 é æ p - p1 1.4 öù é p2 - p1 ù
QV = 2250 ´ Cv (t ) ´ p1 ´ ç1 - ÷´ Þ QV = 2250 ´ Cv (t ) ´ p1 ´ ê1 - çç 2 ´ ÷÷ú ´ ê ú
ë (MW ´ T1 ´ Z ) p1 û
ç æ 3k ö ÷ MW ´ T1 ´ Z êë è p1 3kX T
ç ç ÷ XT ÷ øúû
è è 1.4 ø ø (C.3)

é æ 1.4 ´ ( p 2 - p1 ) öù p 2 - p1
QV = 2250 ´ C v (t ) ´ p1 ´ ê1 - çç ÷÷ú ´ (C.4)
ë è 3kX T p1 øû (MW ´ T1 ´ Z ) p1
The units of measurement in Eq. (C.4) are shown in Table C.1
Table C.1. Units of Measurement for Eq. (C.4)

The value of Qv when calculated with the SI system using the Eq. (C.4), is obtained in Nm3/hr and this needs to
be converted to actual conditions. Fp is the piping geometry factor and since it is not known, a value of 1 can be
taken. Y is the expansion factor of the valve.

Page 32 / 42
Technical Report IChemE Membership No.

Appendix D: Settle-Out Conditions

The settle out conditions of a centrifugal gas compression system can be made from first principles which
consist of using a lumped parameter model in which an energy balance is made across the total volume in the
system with the effect of the compressor deceleration rate. Figure D.1 shows a schematic of the compressor &
the following assumptions are made,

Figure D.1. Schematic of Compression System

Assumptions
1. The settle out conditions are calculated by considering the CC system to be a closed loop & no gas has
escaped the system. The calculations are performed by applying an energy balance over this constant
volume as a series of steady state calculations when integrated yield transient results.

2. The rate of closure of the upstream loading valve & downstream check valve, which causes some process
gas to escape, is not considered for the settle-out conditions calculation.

3. The cooler is assumed to be running at constant duty during either NSD or ESD sequence. If ESD takes
place during cooler failure, then heat rejection (Qcooler = 0) is considered to stop instantaneously.

4. The piping is considered to be adiabatic & no heat escapes from the piping & equipment volumes. The
suction scrubber if considered to have accumulated liquid is subtracted from the suction volume & similarly
for discharge scrubber if present.

5. The time delay between the fully closed to fully open position of the ASV is not considered.

6. The driver coast down curve is estimated as shown in Appendix B since during coast down; some amount of
residual work is done on the gas thereby delaying the gas temperature depreciation rate & pressure to the
final settle out conditions.

Page 33 / 42
Technical Report IChemE Membership No.

Calculations
The settle out conditions is calculated by considering the suction & discharge volumes as,

A. Suction side gas mass = [(

m S = VSuction Side - %VSuction Scrubber Liquid Volume ´ r S ) ] (D.1)

B. Discharge side gas mass = [( )

m D = V Disch arg e Side - %V Disch arg e Scrubber Liquid Volume ´ r D ] (D.2)

Performing an heat balance over the closed loop CC system,

E In = EOut (D.3)

Or, QSuction + QDisch arg e = QCooler (D.4)

Taking that the energy reaching the gas through the compressor is acting only on the mass of gas enclosed &
calculating on a per second basis,
m S c p , S (T - TS ) + m D c p , D (T - TD ) = QCooler (D.5)

Taking m = (m S + m D ) & rearranging Eq. (D.5)

(m c
S p,S TS + m D c p , D TD ) - QCooler
Or, T = SOT = (D.6)
mS c p,S + mD c p,D
The mass specific heat for the cooler in Eq. (D.6) is taken to be an average value between the upstream &
downstream flow. The settle out pressure (SOP) is calculated using Ideal Gas equation as,

é (Z + Z 2 ) ù
n´ ê 1 ú ´ R ´ SOT
ë 2 û
P = SOP = (D.7)
VTotal
m ´ Z avg ´ R ´ SOT
Or, SOP = (D.8)
MW ´ (V1 + V2 )

Page 34 / 42
Technical Report IChemE Membership No.

Appendix E: Inertia Number – Determination of Hot Gas Bypass

A hot gas bypass is used when the anti-surge valve & line are inadequate in terms of recycling sufficient gas
back to the suction side thereby indicating poor surge action in a short period of time. This is used only during
an emergency shutdown or rarely during a normal shutdown. A hot gas bypass is also used when a surge
during ESD coast down occurs due to excessive discharge vessel and piping volumes that slow the rate of
recycle when the anti-surge valve is opened prior to a compressor trip. Often at the early stages of design
projects when minimal data such as only compressor curves is available, requirements for a hot gas bypass can
be done based on a dimensionless number called the inertia number (Ref [2]). This relationship is based on
industrial data obtained for 17 industrial centrifugal compression systems. The decision to employ a short
recycle around the compressor unit during ESD depends on,

1. Effective compressor/driver rotor inertia defined at the compressor end (I) (kg.m2)
2. The delay time before the recycle valve starts its opening stroke (t) (milliseconds)
3. The maximum fluid energy extracted from the compressor/driver power train & the compressor speed which
can be approximated by the product mH s where the subscript ‘s’ refers to the surge point at maximum
compressor speed.

The inertia number is calculated from the total energy balance as by neglecting the frictional losses,

æ dN C ö æ dN M ö
mH S = -ç (2p ) J C N C ÷ - ç (2p ) J M N M
2 2
÷ (E.1)
è dt ø è dt ø
Where,

m = Gas Mass Flow [kg/s] J M = EM Inertia [kg-m2] J C = CC Inertia [kg-m2]

N = Rotational speed [rpm] H S = Head at Surge [J/kg]
Using the gear ratio relationship between the electric motor & the compressor,
NC
Gear Ratio(GR ) = Þ N C = GR ´ N M (E.2)
NM
Substituting Eq. (E.2) in Eq. (E.1) & rewriting the left hand term of Eq. (E.1),

é
2 æ dN C ö æ N C dN C öù
mH s + (2p ) êç J C N C ÷ + çç J M ÷ú = 0 (N.3)
êëè dt ø è (GR )2 dt ÷øúû
æ J ö dN C
mH s + (2p )2 çç J C + M 2 ÷÷ N C =0 (E.4)
è (GR ) ø dt

æ J ö
Taking, ç JC + M ÷=J (E.5)
ç (GR )2 ÷
è ø
Where, J = Total Inertia [kg-m2]

Substituting Eq. (E.5) in Eq. (N.4) & taking N C = N,

dN
mH s + (2p ) JN
2
=0 (E.6)
dt

Page 35 / 42
Technical Report IChemE Membership No.

dN mH s
Or, mH s = -(2p )2 JN Þ- dt = NdN (E.7)
dt (2p )2 J
The operating point would reach the surge line before the ASV full stroke time period ( dt ) during ESD. During
this period, though the speed at the surge line would vary, it is assumed that the CC speed does not vary
appreciably since the recycle valve stroke time is of the order of a few milliseconds. Therefore,
( )
- dt = t recycle valve,initial - t recycle valve, final = t (E.8)

Since during a fast transient of compressor speed for a recycle valve stroke of a few milliseconds,
(N initial - N final ) » N initial (E.9)

Or, NdN = N initial ´ (N initial - N final ) » ( N initial )2 (E.10)

Substituting Eq. (E.8) & Eq. (E.10) in Eq. (E.7), & taking N initial = N
mH st
= N2 (E.11)
(2p ) 2
J

Rearranging, Eq. (E.11), NI =

(2p )2 JN 2 (E.12)
mH st
N I = Inertia number [-]
Converting ‘rpm’ to rev/sec by dividing Eq. (E.12) with 3600, we have

æ ö
ç ÷
Inertia Number ( N I ) =
(2p ) J ´ N ç kg.m 2 ´ (rpm )
2 2 2
÷ (E.13)
3600 ´ mH st ç kg J ÷
ç ´ ´ sec ÷
è sec kg ø
Based on industrial data for 17 compressors, an N I < ~30 would require a hot gas bypass (HGB) to prevent the

compressor unit from experiencing surge, while for N I > ~100 would mean the conventional recycle system

(Anti-Surge system) would be adequate. For 30 < N I < 100 , a detailed dynamic simulation would indicate the

need for a HGB system.

Table E.1. Inertia Numbers for Various Industrial Compression Systems

Page 36 / 42
Technical Report IChemE Membership No.

Appendix F: Compressor Head Calculations

The general energy balance for compressors may be written in differential form as [Ref 3]

æ c2 ö
dy + dq = dh + d çç ÷÷ + gdz (F.1)
è2ø
where,
y = Specific mass referenced compressor work input
q = Heat flow into the compressor through walls [kJ]
h = Enthalpy of Gas [kJ/kg (or) m]
c = Absolute velocity of gas [m/sec]
g = Gravitational acceleration [m/sec2]
z = Elevation [m]

From the above expression it is seen that the specific mass referenced work input and the heat flow to the
compressor is equal to the enthalpy change, change in kinetic energy & static heat difference. Neglecting the
velocity terms, static head contributions & the heat input through the walls of the compressor, we obtain,
dy = dh (F.2)
The change in enthalpy of gas is given as,
dp
dh = (F.3)
r
where,
r = Density of Gas (kg/m3)
p = Pressure of Gas (bar(a))
The specific compressor mass referenced work input is calculated as,
p2
dp
y= ò
p1
r
(F.4)

The actual work is found by dividing the mass referenced work input by efficiency
y
W= (F.5)
h
Where,
W = Actual work applied to the compressor

The integral (Eq. F.4) can be solved in different ways equivalent to different compression paths such as,

1. Isentropic compression (reversible & adiabatic) - Entropy is constant - DQ ¹ 0; DT ¹ 0; DS = 0

2. Isothermal compression (reversible & diathermic) - Temperature is constant - DT = 0
3. Polytropic Compression (irreversible & adiabatic) – Efficiency is constant

Page 37 / 42
Technical Report IChemE Membership No.

Isentropic Process
As an isothermal process is not feasible in real world applications, this is neglected. However an Isentropic
compression can be considered to be an idealistic situation as such a situation can exist when the process is
completely adiabatic & not heat transfer takes place. Therefore for isentropic compression,

p.v k = const (F.6)

Where,
Cp
k= = Ratio of Specific Heats
Cv
Rewriting eq. (F.6), we get
k
æ1ö p
pçç ÷÷ = const Þ k = const (F.7)
èrø r
1
p p1 æ r1k p ö r1k p k
= Þr = Þ r = çç ÷÷
k
(F.8)
r k r1k p1 è p1 ø
1
æ pö k
r = r1 çç ÷÷ (F.9)
è p1 ø
Therefore,
Applying Eq. (F.9) in Eq. (F.4), we get
p2 p2 1 p2
dp 1 p1 k dp
y= ò r
=ò 1
dp =
r1 òp 1
(F.10)
p1 p1 æ pö k
p1
k
r1 çç ÷÷
è p1 ø
p2
p1 k é p k ù
1 p2 1 - 1 +1
p1 k -1

r1 pò1
y= p dp Þ y =
k ê ú (F.11)
r1 ê - 1 + 1ú
ë k û p1
p2
é k -1 ù
k ê
pk ú
1
p
y= 1 ê ú (F.12)
r1 ê æ k - 1 ö ú
êë çè k ÷ø úû
p1

Taking the denominator inside the integrated term outside, we get

1 p2 1
p k æ k öé k ù
k -1
p1 k æ k ö é k k ù
k -1 k -1

y= 1 ç ÷ê p ú Þ y = ç ÷ê 2p - p1 ú (F.13)
r1 è k - 1 ø ë û p1 r1 è k - 1 ø ë û
Taking out P1 as a common term from the brackets, we get

é k ù
k -1 k -1
1
p1 k ´ p1 æ k ö ê p2
k

y= ç ÷ k -1 - 1ú (F.14)
r1 è k - 1 øê p k ú
ëê 1 úû

Page 38 / 42
Technical Report IChemE Membership No.

1 k -1
1
k -1
é k -1
ù + é k -1
ù
p1 k ´ p1 æ k ö êæ p2 ö k æ ö
k

ú p k k
æ k öê 2 kp
y= ç ÷ êçç ÷÷ - 1ú Þ y = 1
ç ÷ êçç ÷÷ - 1úú (F.15)
r1 è k - 1 ø è p1 ø r1 è k - 1 ø è p1 ø
ëê ûú ëê ûú
The exponent of p1 in the numerator equals to 1. Therefore, we get
1+ k -1
é k -1
ù
p k
æ k ö êæ p 2 ö k
y= 1
ç ÷ êç ÷ - 1úú
ç ÷ (F.16)
r1 è k - 1 ø è p1 ø
êë úû

é k -1
ù
p1 æ k ö êæ p 2 ö k
y= ç ÷ êçç ÷÷ - 1úú (F.17)
r1 è k - 1 ø è p1 ø
êë úû
Using Ideal gas equation the pressure & density terms are rewritten using the expression (F.18),
pv = nZRT (F.18)

n p
= (F.19)
v ZRT
The above expression gives the density in kgmol/m3 & is written in terms of kg/m3 by dividing with molecular
weight of the gas.
M p
= (F.20)
v ´ MW ZRT
Where M = mass of gas
M p ´ MW p ZRT
=r= Þ = (F.21)
v ZRT r MW
Hence for the inlet conditions of the gas into the compressor Eq. (A.21) becomes,
p1 ZRT1
= (F.22)
r1 MW
Substituting Eq. (F.22) into Eq. (F.17), we get,

é k -1
ù
ZRT1 æ k ö êæ p2 ö k
y= ç ÷ êçç ÷÷ - 1úú (F.23)
MWgas è k - 1 ø è p1 ø
êë úû
The above expression therefore gives the polytropic head produced by the compressor and hence can be
written as

y = Ha (F.24)

Where, Ha = Adiabatic Head [m]

é k -1
ù
ZRT1 æ k ö êæ p 2 ö k
Ha = ç ÷ êçç ÷÷ - 1úú (F.25)
MW gas è k - 1 ø è p1 ø
ëê úû

Page 39 / 42
Technical Report IChemE Membership No.

Polytropic Process
In all real world compression applications, the polytropic process is predominant & hence the exponent in the
ideal gas equation becomes ‘n’ (polytropic volume exponent) which is 1 < n < ¥ & Eq. A.6 becomes,

p.v n = const (F.26)

Performing the above set of calculations using the polytropic exponent, we get,

é n -1
ù
ZRT1 æ n ö êæ p 2 ö n
Hp = ç ÷ êçç ÷÷ - 1úú (F.27)
MW gas è n - 1 ø è p1 ø
ëê úû
The power absorbed by the compressor or the power that is needed at the compressor shaft is,

H p ´ min , gas
Pactual = (F.28)
hp
Where,
h p = Polytropic Efficiency
min , gas =Mass flow rate of gas at compressor inlet [kg/sec]

Pactual =Actual power required at compressor shaft [kW]

Page 40 / 42
Technical Report IChemE Membership No.

Appendix G: Compressor Discharge Temperature

The temperature rise at after compression is calculated as,
n
æ ZRT ö
p.v = const Þ p ´ çç
n
÷÷ = const (G.1)
è p ø
Taking the value of gas constant to the right hand side of Eq. (B.1), we get

p1- n (ZT ) = const

n
(G.2)

For inlet conditions (suffix 1) to outlet conditions (suffix 2), Eq. (B.2) can be written as,

p11- n (Z1T1 ) = p12- n (Z 2T2 )

n n
(G.3)

p12- n (Z1T1 )
1- n n
n
æp ö æ ZT ö
= Þ çç 2 ÷÷ = çç 1 1 ÷÷ (B.4)
1- n
p1 (Z 2T2 ) è p1 ø
n
è Z 2T2 ø
1- n 1- n
æ p2 ö n æ ZT ö T æ p ö n æ Z2 ö
çç ÷÷ = çç 1 1 ÷÷ Þ 1 = çç 2 ÷÷ çç ÷÷ (G.5)
è p1 ø è Z 2T2 ø T2 è p1 ø è Z1 ø
æ 1- n ö
-ç ÷
T2 1 æp ö è n ø æ Z1 ö
= 1- n
= çç 2 ÷÷ çç ÷÷ (G.6)
T1 p è Z2 ø
æ p2 ö n æ Z2 ö è 1 ø
çç ÷÷ çç ÷÷
è p1 ø è Z1 ø
æ n -1 ö
ç ÷
T2 æ p2 ö è n ø æ Z1 ö
=ç ÷ çç ÷÷ (G.7)
T1 çè p1 ÷ø è Z2 ø
Hence the discharge temperature from the compressor outlet is calculated from Eq. (G.7).

Page 41 / 42
Technical Report IChemE Membership No.

References
1. Kurz, R., White, R.C., “Surge Avoidance in Gas Compression Systems”, Transaction ASME Journal of
Turbo-machinery Vol.126, pp.501-506, 2004
2. Kamal K. Botros, Subramanian Ganesan .T., “Dynamic Instabilities in Industrial Compression Systems with
Centrifugal Compressors”, Proceedings of the Thirty-Seventh Turbo-machinery Symposium, 2008
3. Hansen C., Dynamic Simulation of Compressor Control Systems, Aalborg University, Esbjerg
4. Kurz R., Ohanian Sebouh, “Modelling Turbomachinery in Pipeline Simulations”, Pipeline Simulation Interest
Group, 2003
5. Patel V., Feng J., Dasgupta S., Ramdoss P., Wu J., “Application of Dynamic Simulation in the Design,
Operation, and Troubleshooting of Compressor Systems”, Proceedings of the Thirty-Sixth Turbomachinery
Symposium, 2007
6. http://www.machinerylubrication.com/Read/488/compressor-lubricants
7. AspenTech HYSYS 3.4 Documentation – Dynamic Modelling

Page 42 / 42