Professional Documents
Culture Documents
of PetroieurnEndneers
SPE 35585
CopyrlgM IsSt,
-ml, papa w
Inc
Ttchnobgy
cOllruWm
W
of H2S in the acid gas mixture, the Claus plant has difficulty in
achieving a high aulfhr reeovery. If the total sulfur rate is small,
say less than 5 tonneaklay (t/d), it maybe more economical to
recover the sulfur by some other process. Such other processes,
however, have many drawbacks of their own.
An alternative to recovering *
on a small scale is to
compress and injeet the acid gases into a suitable underground
zone, in a manner similar to the disposal of produced water. This
is envkmma@y deskeable as it eliminates the emission of mlii.u
mmpounda and CO1 to the atmosphere.
This paper discusses the technical considerations for acid gas
compression and injeetion into an underground zone.
In Calgary,
solodd
COntalnd
subn41td
ty UN mmm.
foilwlrq
MVW
cOnkl#d ttm~,
a,
Sox 833S3S,
Abrntract
When sournaturalgases are sweetenedwitk a regenerative solvent,
the extracted acid gases I-I.# and C02 have to be fhrther handled to
dispose of the sulfur in the ~S. The sulfur can be recovered by
various processes, which are generally quite expensive. For
dkprwd of small rates of sulfkr in the form of H.#, compression of
the acid gas stream and injection into an underground zone maybe
a suitable option. This paper diseuaaes the unique properties of
acid gases upon compression and cooling, and the problems that
arise in the handling of high pressure acid gas mixtures saturated
with water. Methods of overcoming the problems are also
reviewed. Process and metallurgical choices are discussed and
opportunities for research to minimim costs in acid gas
compression and inj=tion arc presented.
Introduction
Sournatural gas contains hy&ogen sulfide (H#), which has to be
removed to meet specifications fw sales gas. Sour natural gas also
contains carbon dioxide (COJ The removal of I-I.+3and C02,
usually called acid gases, ti-om sour natural goa is accomplished by
mean3 of a regenerative solvent. There are several amine solvents
used for this putpose. Upon regeneration of the solvent, the acid
gases am liberated, and are usually sent to a modifkd Claus plant,
where most of the H.#i is converted to ekmerkd sulfur. The acid
gas stream to the modifkd Claus plant consists of H#, C02, water
vapor and minor amounts of hydrocarbon gas. When the
concentration of C02 is cmsiderably greater than the concentration
193
SPE 35565
side of each lie, the compound is in the vapor phase, and at the
pressure and temperatures to the left of each line each respective
~*
as a liquid At the pressures and temperatures on
the lines, the pure compounds exist as liquid and vapor. Methane
(CHJ also exhibits this behaviour, but at much lower temperatures,
as shown in Figure 1 also. From this it can be seen that COZ and
H# can readily be liquefied at elevated pressures and at relatively
high temperatures, whereas Cm can be liquefied only at very low
tmpembs.
Mixturm of the acid gases are also readily liquefied
at elevated pressures, but not mixtures containing substantial
amounts of CH4 in their makeup.
The lines in Figure 1 terminate at the upper end at the critical
temperature and pressure for each of the compounds, Above the
critical points, the compounds exist in a dense phase.
194
SPE 35585
Disposal Zone
Most porous zones in the western Canada sedimentary basin
contain water, and at depths greater thao 1000 m, the water is salt
water, of no value. It therefore should not be difficult to identi@ a
zone that would be suitable for acid gas disposal. The operating
company would have to owm the righta to the zone, so that it could
best judge whether injection of acid gas into the zone would be
suitable. Acid gas injection into any zone is subject to the approval
of the Energy and Utilities Board (EUB) in Alberta, and the
195
SPE 35585
removed during cuoling after the fourth stage is about four times
that removed atler the third stage.
Metallurgy. The compressor cylinders will compress gas that is
at its water dewpoint on the suction side, and undersaturated in the
compressor cylinder and on the discharge side, due to the
temperature rise as a result of compression. Thus carbon steel
meeting NACE MRO175 standard requirements for sour gas
should be installed as a minimum The line to the aerial cooler can
be carbon steel meeting the NACE standard. In the interstate
cool- water will condense, and could pose a corrosion problem
Downstream of the coolers, the lines to the interstate scrubbers
and the scrubbers themselves will be exposed to the corrosive acid
gas and condensed water mixlure. Carbon steel meeting NACE
specifications for sour gas, coupled with an appropriate cmosion
inhibition program, should be all that is required to handle these
corrosive fluids, Atk the third stage of compression, the pressure
will be in the order of about 3000 to 4000 IcPa. This is the final
stage at which water will condense in the cooler. Upon boosting
the pressme substaotiaUyabove 6000 kpx the acid gas mixtm can
hold more water in solution than at the previous separation stage.
Thus upm carding aftcrthc fourth compression stage, there should
theoretically be no water of condensation dropping out. If this is
so, then there would be no need to dehydrate the gas, and the
downstream facilities could be constructed out of carbon steel
meeting NACE specifications, which would save the gas industry
a lot of money. This aspect of acid gas phase behaviour should be
experimentally investigated, as stated earlier.
Ifdehydrstion is chosen to ensure that no water of condensation
drops out at the high pressures, the dehydration should take place
after the second or third stage of compression. When dehydration
occurs after Uiesecond stage, the pressure is relatively low, but the
acid gas mixture contains more water than after the third stage of
compressicm and cooling. With dehy&ation occurring tier the
third stage, the opemting pressure of the dehydrator would be about
3 times the pressure after the second stage of compression. This
would require less glycol circulation, smaller vessel dhmeter, but
the amount of acid gases absorbed per gallon of glycol circulated
would be higher. However, glycol circulation would need to be
about one half of the required circulation for dehydration after the
second stage of compression, since there is less water to remove
from the acid gas stream at the higher pressure.
In general, the selected metallur~ would depend on whether
the gas is dehydrated or not. If stainless steel is selected, then the
stairdess steel equipment includes the suction scmbbers, pulsation
bottles, interconnecting piping and coder tubes.
The metallurgy of valves and instrumentation such as block
valves, dump valves, PSVS, gauge glasses, etc., should generally
match the material of the vessel and piping.
Compressor cylinder material can be carbon steel that meets
sour gas Specflcations. Compressor suction and discharge valves
should be stainless steel. The piston rod should be stainless steel,
and may be coated with a tungsten carbide overlay having a
maximum hardness of RC22.
Distance pieces can be single compartment with a vacuum
pump, or a two piece compartment with a sweet gas purge system.
196
SW 35505
simple.
A meter should be installed to record the flow. This
would provide infbrmaticmto the plant SCADA system concerning
the injection pressure, temperature, rate and the fluid density. The
wellhead would be equipped with a check valve as well as an ESD
valve.
dehydrated.
For dehydrated gas, sour service carbon steel
materials can be used, such as CSA-Z.662 Grade 359 Cat. II sour
service material. The addition of corrosion inhibitor should he
considered.
For nondehydrated acid gas, the line can be of carbon steel
with an internal epoxy coating, or a polyethylene liner. Another
pipeline material can be 316L stainless steel. One of the key
considerations is the length of the line. Stainless steel is more
expensive, and if the coat of the line exceeds the cost of a
dehydration unit, then it maybe more economical to dehy&ate the
acid gas ~am.
The proposed pipeline should be fidly evaluated
with the vendor of the materials to ensure compatibility with
anticipated process conditions of pressure, temperature and acid
gas composition.
The line diameter should be sized for liquid phase fluids if the
acid gas mixture contains no more than about 1 /o to 3 !o
hydrocarbons. This means that the injection rates would be quite
low in terms of liquid quantities. An injeetion rate of 28 10m3/d
(1 MMscfld) of gaseous C02 reduces to 44 litrea per minute
(thin) (1 1.8 US galhin) in the liquid phase. Thus the pressure
drop in a short, 50 mm diameter line would be quite small, at a
liquid velceity of about 0.37 mkcc.
If the diqmsal well is near the plant site, a 50 mm disposal line
out of 3 16L SS would cost about $40,000.
Under normal operating conditions, the line between the plant
and the injection well would contain the acid gas mixture in the
liquid phase or the dense phase. The temperature of the fluid in the
line would beat ground temperature, which could be as low as O
C in the winter time. If this line had to b dcpressurized,
however, Vezylow tmpmtmcs would result in the line. Since the
fluid
would be blown down to atmospheric pressure, the
temperature of the fluid in the line would decrease to the boiling
point temperature of the acid gas mixture at atmospheric pressure.
This could be as low as -50 C, or lower if the line were
insulated. Lines that are not insulated would benefit from heat
transfcrthlrnthc sumundm g soil. The pipeline material has to be
designed for low temperatures, even though under normal
conditions such low temperature would not be reached. A stie
blow-down design and procedure should be incorporated in such
a scheme.
The extremely low temperatures can be avoided by installing a
25 mm carbon steel line to the wellsite in the trench with the 50
mm injection line. The design pressure of this sweet gas line
would be 7 NIP%and under normal circumstances, this line would
be left idle at a low pressure. The line would be connected at the
wellsite to the injecticmline upstream of the wellhead. If it became
necesswy to depressurix the injection line, the acid gas would be
flowed back to the plant and into the flare system through a standby
heat exchanger or a rental line heater. The liquid acid gas mixture
would be heated and vaporized prior to the pressure reduction into
the plant flare system. To prevent the acid gases ti-om vaporizing
in the injection line, sweet gas fhmn the plant would be flowed to
the wellsite through the 25 mm line, from where the sweet gas
would push the acid gas fluids back to the plant at fill pressure.
Wellsite Facilities.
197
SPE 35565
conclusions
1. The compression and injection of acid gas into an
underground reservoir is technically feasible.
2. Research should be carried out to determine the water of
satmtmn of acid gaa mixtures containing 1 A to 3 Yemethane, in
the pressure range of 3 MPa to 10 MP~ and at temperatures
between OC and 40 C, The results would establish whether the
acid gas would need to be dehydrated to avoid the fbrmation of
hydrates under operating conditions of acid gas injection systems.
3. Injecting acid gas into an underground formation releases the
minimum ofgmnhouw gasea into the atmosphere when compared
with the options for small scale sulphur reemery.
4. Compression and injection of acid gas into an underground
zone provides great flexibility in operation with changing
throughput rate or fluctuating acid gas canposition.
References
1. Canjar, L.N. md Manning, F.S.: TYwrrnodytamic properties and
RcducedCcwela~fbr~
TX (1967) 126-133.
2,
3.
4.
5.
6.
7.
8.
Stage No.
1
2
3
4
198
SPE 35585
Io,mo
Fly&ah
aow
em
,-
1
1;
1,
q
:
4
H,s
.1
7,0C0
1,
1,
am
aom
,
I
.
t
I
.:;
4,000
co,:
C+t,
H$
:
.1:
am
:!
Zm
,*
1
I,ow
,
,$
-.
(12q
(150)
(*
F@
1-
for COi
Vapor
pro6suro
wrous
40
(q
Temperatum
1!20
(W)
tempmlure
and
hydrato condttione
and HIS.
% H#
20,000
1O,an
5,0M
i
i
2,0ao
1 ,m
1,Ioa
2,0M
aow
!Qcm
20,COQ
Premure(kPd
10152025W35
Tempwatwe
(Z)
Fig,
199
SPE 35585
10,000
moo
.. .
,
#..
$ ,$
....:
(.
,..
,..
8*WJ
8.:
# ...
!!
1:: ,
#.. ,
, ..,t
,4 :t.
/ . p, .
,. ,
t ... ,~
,. ,
,
4,000
,
t
..
,#
#
2,000
.......4....
?160)
(140)
.yJ;i;*-
(120) (loo)
(60)
(60)
(40)
(20)
20
40
60
Tomporaiwo (C)
nvolopoofor varlow g= mixturoo.
fig. 4- Prooouro- Tomporatwo phaao
Acid Ga9
from Roflux
Orum
--0
to Injwtlon
Wdl
If
;8
.:
4
1!
3 :
. . . . . . ..-
13s
..
..
150
..
,;
.
.
.
.
0
.
.
.,
.
,.
;
.
*
,
.
0
*
.
,*
.
~:
1
.
. . ...
.
.
7s
.
.
.
%. ..*
*-
60
.
~ 46
W*
- 15
0~o
F@ 5- Four-etago acid gas oomprwelon ahowlng proesuro rlso and tsmporaturo variation..
!/