Purdue University

Purdue e-Pubs
International Compressor Engineering Conference School of Mechanical Engineering

1984

Performance Evaluation by Analog Study of a Two

Stage Reciprocating Compressor Handling
Complex Hydrocarbon Gas Mixtures
S. N. Dwivedi

Follow this and additional works at: http://docs.lib.purdue.edu/icec

Dwivedi, S. N., "Performance Evaluation by Analog Study of a Two Stage Reciprocating Compressor Handling Complex Hydrocarbon
Gas Mixtures" (1984). International Compressor Engineering Conference. Paper 442.
http://docs.lib.purdue.edu/icec/442

This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact epubs@purdue.edu for
additional information.
Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/
Herrick/Events/orderlit.html
 PBRFORMANCE EVALUATION BY ANALOG STUDY Of A TWO STAGE
RECIPROCATING C(J,U)RESSOR HANDLING CCNPLEX HYDROCARBON GAS MIXTURES

s. N. DWIVEDI, DY. MANAGER - CQ.1PRESSOR DESIGN

BHARAT PUMPS & <XUPRESSORS LTD., NAINI, ALLAHABAD (INDIA)

AQSTRACI

A two-stage reciprocating compressor was designed to handle two different compositions

of crude overhead gas (i.e. a complex hydrocarbon mixture containing heavier hydrocarbons
identified as pseudo-cuts) available from two different sources. The the:modynamic p%Qperties
and flash calculations were don& with the help of a computer program developed eapecially fer
this purpose on a DEC 10 Computer System.

Special care was taken during the design stage tD ensure that condensation does not
occur at the inlet of the cylincbrs or inside the cylinders, Coolers and Separation devices
were sized so as to remove the condensate effectively before the gas stream enters the
cylinders of respective stages. Piping layouts were kept such as to prevent any accumulation
and consequent carryover of condensate to compressor cylinders. Piston rings and sealing
rings we:re made of filled PTFE grades in view of the dry service.

Upon CO!q)letion of the designs of the main compressor dongwi th pulsation suppression
devices, heat exchangers, separators and finalization of the layouts, the whole syatem was
simulated on an Analog Simulator so as to assess the perfo:rmance of w compression plant.
The Analog study utilizing guide-lines in API 618 Design Approach 3 was done at Southwest
Research Institute, USA on the $A Analog Simulator. The study duplicated the compressor,
vessels, heat exchangers and piping system in between the first major vessel at cCJIIPl'Ossor
suction at the upstream side and the second stage separator on the dawn stream side. Various
test conditions were included in the study to represent various operating parameters of the
compression plant.
Some modifications to the original systeM were applied so as to minimise certain un-
desirable acoustic characteristic s, such as higher pressure pulsations and pressure drops
than that suggested by API, found in the original system. The resulting pressure pulsations
and acoustic shaking forces were found to cause stresses on piping and bottles Within
acceptable limits.
Perfo~ance analysis of the modified plant system indicated only marginal deviations fro.
ideal PV card.
INTRODUGIIQN
Analog study of a Reciprocating Compressor system at the design stage itself, by simulat-
ing the compressor a1ongwith the vessels, heat exchangers, piping etc. on to an active Analog
Simulator, has been widely accepted in the industr{ as a reliable means for ensul'ing the
optimum performance of the plant. Such Analog study considers the generation of flow and
pressure pulsations .inside the compressor and its dynamic interaction with the piping system
connected to the compressor. API standard 618-1974 Design Approach 3 gives the necessary
guidelines for such an Analog study. EXperience has proved the effectiveness of these guidelines
so that these are specified by most of the major users.

116
 THE AfPLIGATION
The co~ressor applica tion involved compres sing crude OVHD Gas
mixture contain ing C2, C3, ~. C5 and heavier hydroca rbons identifi i.e. a complex hydroca rbon
ed
their molecul ar weights , sp. gravity and boiling point. TWo differe as pseudo- cuts indicati ng
(see table I) were specifie d and the working pressur es also varied nt alterna tive composi tions
1.4 IIJ/cm2 abs. suction pressur e to 9.5 ra/cm2 abs. discharg epressu in the two casas viz. from
re
abs. suction pressur e to 5.5 ~cm2 abs. discharg e pressur e in case in case I and from 1.37 .VCIJI2
compres sors Unit A and Unit B, of which only one was to :run at II. There were 2 nos.
a time and the other was intended
to be a standby unit.

MIXTURE PR<PERTISS

A especia lly develop ed compute r program , based on Soave~a

was used for predicti ng the various thermodynamic and transpo rt dlick-Kwang and UNIFAG methods
mixture and for doing flash calcula tions to estimat e the amount propert ies of this hydroca rbon
of condens ate to be taken out
of the compres sed gas stream after it emerges from the cooler.
An accurate estimat ion is very
vital for sizing of compres sor plant as the amount of condens ate
may be substan tial.
THE COOPRESSOR

Based on the pressur e ratios involved and keeping in view the optimum
tempera ture conditio ns, a two stage compres sor, with suitably sized intersta ge pressur e
require d capacit y, was designe d. The intersta ge cooler and separat cylinde rs to handle the
or and the final stage cooler
and separat or were designe d to take care of the case I design conditio
time also ensuring reaSona bly good working in case II conditio ns. ns, while at the same

Separat ors were made of proprie tary vane.opack design to ensure

(under various load conditi ons). Piping layouts were made so as effectiv e condens ate removal
consequ ent carryov er of condens ate to COJil>ressor cylinde rs. to prevent any accumm ulation &

In view of the non-lub ricated applica tion piston rings and rider
rings in ca:rbon graphit e filled PTFE. grade were used. CoJil>ressor rings and rod packing
suction and discharg e valves
of proprie tary double da111ping (mechan ical) design were used to
ensure low pressur e drop through
valves.

THE ANALOG STUDY

The Analog study utilizin g API 618 Design Approach 3 guidelin

Researc h Inst'!. tute on the &;A Analog Simulat or. The study duplica e was done at SouthWest
ted the coJil>re ssor, vessels ,
heat exchang ers and piping system in between the first major vessel
Knock out drum and the final stage separat or. A schemat ic vieYJ at c~ressor suction viz.
of the system is shown in fig. 1.
Due to operatin g and gas conditio ns a 10 percent deviatio n in
included in the investig ation. This is easily accompl ished by the the velocit y of sound was
speed by the correct percent age. Thus frequen cies appear that Analog by varying compres sor
do not correspo nd directly to
integra l multipl es of compres sor speed, the amplitu des of which
are valid, however , and will
occur at compres sor excitati on frec:p.1encies for the analogo us operatin
~eration at various capacit y steps was studied
g and gas conditio ns studied .
at both case
lists the test conditio ns evaluate d dUring the study as these I conditio
and case II conditio ns. Table II
pulsatio n levels in the system. ns involved the highest

Acousti c charact eristics data was recorded on the origina l suction

systems for both stages of compres sor for unit A and unit B based and discharg e piping
upon case I and case II
operatin g conditio ns. Undesir able acoustic charact eristics found
in the origina l system were
pulsatio ns in the piping exceedin g API guidelin es and pressur e
drop through the origina lly
installe d orifice s in compres sor nozzles exceedin g API guideli nes
compres sor cylinde r perfo.nn ance. and thus adverse ly affectin g
For getting lower pulsatio n levels and reduced pressur e
drops so as to improve compres sor perfo~nce certain modific ations
were conside red during
the study which included increasi ng orifice diamete r, installa
tion of fresh orifice s, deletion
of a few orifice s and increasi ng the pipe diamete r at certain location
system offered a better compro'llise on pressur e drops, pulsatio ns s (Fig. 2). The modifie d
perform ance more or less meeting API guideli nes. and the coi1J)resaor cylinde r

117
 'II'

A mechanical evaluation, with the help of a digital coq:~uter, of the coqJressor volume
bottles was considered to determine the adequacy of supports with a view to have overall systena
vibration control. The overall shaking forces were found to be low in amplitude and the
supports were adequate so that no undesirable vibration of the vessels could be there.
The mechanical evaluation also included the analysis of the piping response. With the
addi tlon of a few pipe supportli the owrall vibration levels of the piping at different points
were brought to acceptable limits so as to cause cyclic stresses well within the limits specified
by API.
The performance of the modified compressor system as indicated by the P.V. diagrams obtained
from the simulator at different test conditions is indicated in Fig. 3 to 6.
1

CONCLUSIOO
The performance of the compressor system 'can be very well evaluated at the design stage
itself by simulating the system on to an Analog Simulator and doing the acoustical analysis
followed by a mechanical evaluation on a digital caq>uter. Modified versions can also be eval-
uated for the ultimate perfo:rmance so as to optimize the system at this stage itself, rather than
going for some change when the compressor system poses some problem while commissioning at site.
REPRENGSS
i) "Reliability and Performance Assurance in the Design of Reciprocatin g Compressors nd Pump
Installation s" by w. Von Nimitz, 1974 Purdue Compressor Technology Conference.
li) "Role of Analog study towards better per forman;; e of Reciprocatin g Conpressor Plantsn by
s.N. Dwivedi and W. Von Nimitz, Seminar on Compressor Performance and Improvements
(8th- 9th May 1983), Allahabad, India.
iii) nApplication Criteria for Compressors Handling Natural Gas containing heavier hydrocarbons
-by J.s. Pandey and s.N. l:Mivedi, Seminar on Compressor Perfoxmance and ImproV811ent .
(8th- 9th May 1983), Allahabad, India.

118
 TABLE I
GAS QJ.U)OS ITIONS

CASE I CASE: II
CoJI1)onent Mol.Fr n. Mol. Wt. API Boiling Mol.Fr n. Mol. Wt. API Boiling
(oF) point (OF) point

c1
C2 o.o82 30.1 o.ols 30.1
c3 o. 33 44.1 o.l86 44.1
G 4 0.223 58.1 o.331
C5 58.1
0.0337 72.1 o. 29 72.1
Pseudo 1 0.0319 68.8 105.9 69 0.0478 88.4
2 o.o446 73.8 157
74.8 85.4 102 o.o373 95.6 69.1
3 0.0589 81.6 185
71.7 137 0.0198 104 64.7 215
4 0.0618 89.1 62.2 172 0.0093 113
5 0.0349 6o.4 245
<n.2 55.0 207 o.oo36 123 57.6 275
6 0.0145 106.0 49.4 242 o. 001
7 133 54.7 305
0.0002 115.0 5ft--:S 256 o.ooo2 144 52.1
8 0.00022 116.0 355
44.7 277
9 0.0009 121.0 53.4 277
10 0.0001 125.0 41.0 309
11 0.0004 133.0 49.8 312
Water o. 083 18.0 o. 056 18
l.ooo 54.9 1.000 61.4

TABLE II
ANALOG STUDY TEST CCWITIONS

CASE: I CASE I CASE II CASE II

Test Number 1 2 1 2
Initia l Suction 1.4 1.4
press. (Ata.) 1.37 1.37
Final dischar ge 9.5 9.5
press. (Ata.) 5. 5 5. 5
Unit Flow (Nm3/hr) 4613.8 2306.9 4355 2177.45
Compressor loading 100% 50% 100% 5o%
(a) Pocket s Open None None l.Jone None
(b) Single active None HE-2
cylind ers None HE-2
CE-1 CE-1
No. of units in 1 1
paralle l 1 1
Piping Flow through Flow through
Config uration Flow bypass ing FloW bypass ing
intera> oler interco oler interco oler interco oler

119
~ .CHEM;lTIC !10QIFIW 5YSKM

120
 PERFORMANCE ANALYS IS
CASE 1 MODIFIED SUCTION
<1,- DISCHARGE

CYLINDEr\ N0.2 CYLINDER N0.1

HEAD END HEAD END

HORSEPOWER HORSEPOWER
DEVIATION DEVIATION
6. 6 % 3,2 ;;

FLOW DEVIATION FLOW DEVIATION

3.5 % UI

CRANK END CRANK END

HORSEPOWER HORSEPOWER
DEVIATION DEVIATION
7' 0 % 5. 6 %

FLOW DEVIATIDN FLOW DEVIATION

3. 9 ;; UI

FIG. 3

PERFORMANCE ANALYS IS
CASE 2 1001. LOAD MODI-
FIED SUCTION &! DISCHARGE

:YLINDER NO.2 CYLINDER N0.1

HEAD END ~EAD END

HilRSEPOWER HORSEPOWER
DEVIATION DEVIATION
6. 5 ;: 9. 2 "!.

FLOW DEVIATION FLOW DEVIATION

1.~ i. 1.6 !

CRANK END CRANK END

HORSEPOWER HORSEPOWER
DEVIAmN DEVIATION
5. 5 !. 7. 9 %
FLOW DEVIAT!O~
FLOW DEVIATION
1.b z

FIG. 4

121
 PRESSURE VOLUME CARDS
2 NO. STAGE ( 50 ,4) LOAD

a:
F r \

!/
w
0
~

8 8
...J
>
u
----

FIG. 5

PRESSURE VOLUME CARDS

1ST, STAGE (50 7.) LOAD

cr:
w
0
z
::J
>
u

FIG. 6

122