I. MANUAL PURPOSE (Revision 2)
I. MANUAL PURPOSE (Revision 2)
I. MANUAL PURPOSE (Revision 2)
To be used for selection, application into the system, power and cooling water estimation. This
manual does not for designing centrifugal compressor and those parts.
Pressure p bar A
Temperature t C
Absolute Temperature T K
Capacity (volume flow) Q m3 / hr
Power P kW
Brake horse power BHP kW
Gas horse power GHP kW
Speed N RPM
Head H m
Gas Constant R kJ/kg.K
Molecular Mass MW kg/kg (=lb/lb )
mole mole
Mole MM kgmole ( kgmole/h or kmol/h )
Density DS kg/m3
Specific Gravity SG
Specific volume v m3/kg
Specific Heat Cp kJ/kg.K
Mass Flowrate G kg / hr
Adiabatic Exponent k -
Polytropic Exponent n -
Compressibility Factor Z -
Efficiency E -
Gravity g m/s2 (9.81)
Heat Capacity MCp kJ/kgmole
Enthalpy h kJ/kg
Enthalpy different dh kJ/kg
Entropy s kJ/kg.K
Impeller Diameter D mm
Tip speed (tangential) U m/s
Number of impeller i -
Mach Number Ma
Flow Coefficient CQ -
Head Coefficient Y -
Mechanical power loss Pml kW
Subscript
Length ft 304.8 mm
inch 25.4 mm
Power HP 0.7457 kW
Head ft 0.3048 m
Note : American Standard State at 1.013 bar A and 15.5 C. In volume common written as SCF.
Normal condition at 1.0132 bar A and 0 C. In volume common written as Nm 3
IV. OPERATING RANGE OF CENTRIFUGAL COMPRESSOR
Fig.3 presents operating range of Centrifugal Compressor based on suction flow and discharge
pressure and fig. 4 presents operating range of centrifugal compressor compare with other type
of compressor based on suction volume flow and speed.
Fig. 4. Operating range of centrifugal compressor compare with other type of compressors
V. GAS COMPRESSION
Gases to be handled by compressor are both single component (pure gas) and mixed gas. This
manual also describes physical properties of mixed gas.
In the next equations and calculations, gas is assumed as ideal gas but then corrected by
correction factors and so ever is assumed equal to actual physical properties of the gas. By
any reason, for some cases, compressed gas is also assumed as non ideal gas.
dh
Gas compression process is presented in enthalpy versus entropy chart. Gas enter compressor
through suction nozzle (1) at ps = p1 measured as total pressure and become p2 as static
pressure in isentropic process. Gas goes to 1’st impeller eye (3) with little losses and then
compressed to condition (4) and then in diffuser (6). Gas flows through vane (redirected) until
condition (7) then come into 2’nd impeller eye (9). Next compression is in 2’nd impeller through
condition (10) until (13). Gas goes out through discharge nozzle at condition (14) or at p2=pd.
To determine adiabatic and polytropic efficiency, use the following chart and equations.
VI. INTERCOOLER
After compression, gas temperature will rise up but it is limited before entering to the next
compression. Temperature limitation is depending to what sealing material to be used and gas
properties. To decrease temperature before entering to the next compression, compressor
needs intercooler.
VII. AFTERCOOLER
Aftercooler is used when discharge gas temperature leaving compressor shall be decreased
before entering to other equipment or system.
Antisurge control shall be installed to centrifugal compressor because at low flow, compressor
will surge. Antisurge control is instruments to detect pressure and flow at where compressor will
surge. Antisurge control is completed with control valve to by pass discharge gas back to
suction or vented to atmosphere. See Fig. 6.
Fig. 6. Typical antisurge control for centrifugal compressor
IX. EQUATIONS
This manual uses the following simple equations. All symbols and unit are according to symbols
and units described in chapter II.
G.H .g.106
GHP = (kW) (3) r2
3.6
Where G is mass flow = DSs.Qs (kg/h), DSs is density (kg/m3), Qs is vol. flow (m3/hr) (4)
Ts.Qn.Zs Td .Qn.Zd
Qs = and Qd = (m3/hr) (5)
269.69( ps.Zn ) 269.69( pd.Zn )
Where Qn is volume flow at normal condition ( 0 C and 1.013 bar A)
n 1
1000( Zs )( R)(Ts ) n pd ( )
Hp = { }{( ) n
1} (m) (7)
g n 1 ps
Where (pd/ps) is compression ratio. Pd and ps are in absolute pressure (bar A) and
n k
( EFp ) (8)
n 1 k 1
R = Ro / MW (kJ/kg.K)
Ro = 8.314 (kJ/kgmole.K)
GHP
BHP = (kW) when total polytropic efficiency (EFpt) is known where (9) r2
EFpt
bearing and seal losses are included in this total efficiency. If mechanical losses are
calculated separately, than this EFpt is not necessary and BHP = GHP/EFp+Pml (see
equation 1). Data of efficiency in this manual is EFp (Efficiency of impeller) r2
Discharge Temperature
n 1
pd ( )
Td = Ts . ( ) n
(10)
ps
If discharge temperature is limited at Tdmax, then maximum pressure ratio become
n
pd Td max ( n 1 )
( ) MAX = ( ) (11)
ps Ts
3.14( D)( N )
U (m/s) (12)
60,000
Flow coefficient,
353.68(Qs )
CQ in the range 0.01 up to 0.15 see Appendix B (13)
(U )(D 2 )
Head coefficient,
19.62( Hp )
Y (14)
U2
Y values in the range 0.80 up to 1.1 for impeller with "backward leaning blades"
1.30 up to 1.45 ----- ,, --------- "90 degree exit blades "
X. PERFORMANCE CURVE
Fig. 7. Typical performance curve of centrifugal compressor (for variable speed driver)
APPENDIX A. GAS PROPERTIES
The following table presents single gas properties. There are MW (molecular weight), k
(adiabatic exponent), pcr (critical pressure ), Tcr (critical temperature) and MCp (=MW x Cp).
Z determined by gas compressibility chart using reduction temperature (Tred) and pressure
(pred) as the variables. Tred = T/Tcr and pred = p/pcr. See following Fig. 8.
Density of gas,
100( p)
DS = (A.3)
R.T .Z
Fig. 8 presents Z factor for pred = 1 and lower. For pred higher than 1 see Fig. 9.
Fig. 9. Gas compressibility chart for pred higher than 1.
i
MW = {0.01(%Mi)(MWi)} (A.4)
100( MMi)
% Mi = (A.5)
MMi
Where MMi is molal mass of each gas in kgmole or mols
Mgi
MMi = (A.6)
MWi
Z factor determined using Fig. 8 and 9 at above pcr and Tcr of mixed gas.
100( p)
DS = (kg/m3, all unit shall be as listed in chapter II) (A.10)
R.T .Z
Gas shall be dry in centrifugal compressor to prevent internal parts and impeller from erosion
due to liquid particles. Gas condition shall be kept at little far from wet condition. Following
table presents vapor pressure for some gases.
1. Relative humidity RH in %
2. Dry bulb temperature tdb in C and then Tdb = 273 + tdb in K
3. Atmospheric pressure patm at bar A
4. From psychometric chart, determine wet bulb temperature twb and Twb = 273 + twb
5. From H2O saturated pressure table, determine saturated pressure at twb, pg
6. Partial pressure of H2O pw = 0.01 (%RH)(pg)
7. Partial pressure of dry air pa = patm – pg
8. Mole fraction of dry air Xa = pa / patm
9. Mole fraction of H2O Xw = pw / patm
10. Molal mass of wet air MW = (MWdry air)(Xa) + (MWH2O )(Xw)
11. MCp of wet air MCp = (MCp dry air)(Xa) + (MCp H2O)(Xw)
12. Gas constant R = 8.314 / MW
13. k k = MCp / (MCp-8.314)
14. Density DS = 100.patm / (R.Tdb.Z)
From % RH and tdb determine twb from following typical psychometric chart
Following figure is simplified figure of impeller in relation with their performances. tip speed (U)
and exit angle. Tip speed is limited due to material strength and sound velocity in compressed
gas.
Fig. 12. Impeller geometry and their characteristic.
Shrouded impeller is equal with enclosed impeller where there is disc in the front integral with
vanes or blades (casting, welded or riveted) .
Backward curve impeller is when exit angle smaller than 90 degree and radial curve impeller
is when exit angle is equal to 90 degree.
To convert head from J/kg to m, divide J/kg unit by g (gravity in m/s2). Example 55 kJ/kg =
55,000 J/kg = 55,000/9.81 = 5606.5 m
Tip speed is limited due to material strength and sound velocity in compressed gas.
Below 35 310
Below 45 250
Below 65 200
Below 120 150
Even MW below 35 but gas contain corrosive matter or will be operated at low temperature
below -50 C, maximum tip speed is 250 m/s.
Maximum tip speed shall not higher than sound velocity. For approaching, using Umax = 0.9 x
a, where a is sound velocity = (1000 x k x T x Z x R)0.5 or = (8314 x k x T x Z /MW)0.5 . U= 3.14 x D x
N /60,000, or D = 60,000 x U / (3.14 x N) or D = 60,000 X 0.9 x (8314 x k x T x Z /MW)0.5 / (3.14 x N)
or D = 1568000 x (kxTxZ/MW)0.5/ N. And “0.9” is factor for incorrect approaching for all
assumption and calculation related to sound velocity and tip speed. This equation can be
plot at several (kxTxZ/MW)0.5 , see Fig. 13 blue dot line. D and N pair under blue dot line is
accepted because tip speed is lower than 0.9 time sound velocity.
Fig. 13. Impeller performance in head
Example : Air at 30 C (303 K) and 1.013 bar A to be compressed 6 barA. Suction flow 2000 m 3
/hr. From table 1, k =1.4, pcr = 37.7 bar A, Tcr =132.8 K, MW =28.97. Determine pred = 1.013/37.7
= 0.027 and Tred = 303/132.8 = 2.28. From Fig. 8, Z near = 1. Ratio value of (k.T.Z/MW) 0.5 = 3.8.
From Fig. 13, this line is near head per stage 4000 m line. At this line, if N=10,000 RPM can be
determined diameter of 550 m and maximum tip speed U become = 3.14.D.N/60,000 = 3.14 x
550 x 10,000 / 60 ,000= 287 m/s.
Compressor size is designed according to suction flow of gas. Compressor size is includes
impeller diameter, impeller width and number of impeller. Manufacturer usually indicates
impeller diameter in each compressor model. The following Fig. 14 presents nominal size
indicated by average impeller diameter. From suction flow can be seen approximately the
nominal size and also speed in RPM.
Example : Continued from above example. At suction flow = 2000 m 3 /hr, draw vertical line until
cross the range of each nominal size, provide nominal size 300 mm with impeller geometry “C”.
Cross point between flow and maximum tip speed 287 m/s is at maximum about 20,000 RPM. In
C impeller geometry, at flow 2000 m3 /hr speed in the range of 11,000 up to 20,000 RPM or tip
speed from 200 up to 287 m/s.
From rectangular bloc at the bottom of chart, geometry C impeller has flow coefficient CQ
small = 0.01 up to 0.05.
Number of impeller in one casing is limited. Maximum number of impeller can be estimated
from Fig. 13 as function of Mau and impeller geometry.
Example, continued from above example. If geometry C and nominal size 300 mm is selected,
approximate diameter is 300 mm. And if tip speed is selected at 280 m/s, then Mau become =
280 / (1000.k.R.T)0.5 = 280 /{1000 x 1.4 x (8.314/28.97) x 303}0.5 = 0.8. From Fig. 13 maximum
number of impeller Imax = 10.
Number of impeller will decrease if there are additional nozzle is installed such as for inter
cooler, admission or extraction. Each 1 nozzle will reduce 1 impeller.
Instead of preliminary efficiency that determined from Fig. 11, efficiency of compressor can also
be determined after preliminary diameter and tip speed is selected. Fig. 14 shows efficiency as
function of head coefficient and flow coefficient.
Example, continued from above example. D = 300 mm, U = 280 m/s have been selected at
above example. N = 17,800 RPM. Assumed impeller is backward 2 dimension shrouded impeller.
To determine head coefficient, preliminary efficiency shall be determine from Fig. 11, EFp=
0.695.
n
pd Td max ( n 1 )
pressure ratio max. at section 1 = ( ) MAX = ( ) = 2.95
ps Ts
extract air to 1’st intercooler at pressure pd1 = 2.95 x 1.013 = 2.99 barA
by the same way and assumed pressure drop is neglected aross intercooler,
n 1
1000( Zs )( R)(Ts ) n pd ( )
Hp1 = { }{( ) n
1} = 12,260 (m)
g n 1 ps
Plot again N=15,000 and D=300 mm on Fig. 12. Cross point is at head per stage = 3250 m
Hp2 = 10,600 m
From Fig. 14 with number of impeller = 8 and 2 dim. shrouded impeller, uncorrected efficiency =
0.72. Corrected efficiency become
EFp = 0.72 x Co = 0.72 x 0.915 = 0.66 or 66 % . This result is smaller than first assumption = 0.695 or
69.5 %
APPENDIX C. CALCULATION SHEET AND SYSTEM (AS ATTACHMENTS)
Typical calculation sheets as attachment 1 of this article are presented in excel file, named
“Cal_sheet_c_comp.xls” which are included :
Typical system around centrifugal compressor as attachment 2 are presented in PDF file, named
“Sys_c_comp,pdf” which are included :