Sei sulla pagina 1di 7

Use of Laboratory and Field.

Testing
To Identify Potential Production
Problems in the Troll Field
R. Hartley, SPE, and M. Bin Jadld, NS Norske Shell
Summary. The areal extent of the oil found in Troll made it clear at a very early stage in the field's appraisal that subsea wells
would be required if the oil were developed. Owing to cooling in the subsea flowline, subsea wells can be expected to pose more
production chemistry problems than would be expected with conventional platform wells. Consequently, a number of laboratory
tests were carried out during the appraisal campaign to identify problems to be expected with scaling, foaming, emulsification, wax
deposition, and hydrates. Dehydration and wax deposition tests were also carried out offshore during appraisal-well testing. These
tests are described, together with the methods subsequently adopted to minimize future production problems.
Introduction
The Troll field is located in the Norwegian Trench some 80 km
[50 miles] west-northwest of Bergen, where the water depth varies
from 300 to 340 m [985 to 1,115 ft]. The field extends over four
license blocks-31/2, 31/3, 31/5, and 31/6-awarded to two
separate license groups. Norske Shell is operator for Production
License 054, which covers Block 31/2.
The field was discovered in 1979 with Well 31/2-1, a crestai well
in the large western gas structure. Subsequent exploration and ap-
praisal wells in the block confirmed the presence of a major gas
accumulation underlain by an oil rim with a 22- to 27-m [72- to
89-ft] thickness in the west and - lI-m [36-ft] thickness in the re-
mainder of the block. Reservoir depth is some 1500 m [4,920 ft]
subsea. By the end of 1984, a total of 15 wells had been drilled
in block 3112, completing the appraisal of that part of the field
(Pig. 1).
The 340-m [1, 115-ft] water depth and 15OO-m [4,920-ft] -subsea
reservoir depth imply limited reach during drilling of deviated wells
from a platform. This, together with the areal extent of the oil, in-
dicated at a very early stage that subsea wells would be required
if the oil were developed. Therefore, as part of the field appraisal,
a number of studies were carried out to identify potential production
problems associated with subsea oil wells.
Scaling Tendency
The scaling tendency of the Troll formation water was assessed from
the analysis of samples taken in Well 31/2-11.
Calculations based on the Stiff-Davies 1 stability index indicated
that, under normal producing temperatures, formation water is not
expected to form scale. The high skin temperature (up to 150C
[3OO
0
P]) of the crude heaters, however, will cause severe CaC0
3
scaling. This expectation was confirmed through laboratory tests
with synthetic formation water. The result indicates a requirement
to inject scale inhibitor upstream of the heaters. A polyphosphate
scale inhibitor was found to be effective in the laboratory tests.
Scaling treatment by injection of polymer scale inhibitor is
common to North Sea operators and is normally included in the
crude-heater design package.
Emulsion
Simultaneous production of oil and water in Troll subsea oil wells
is expected to result in the formation of emulsions. This will pose
crude dehydration and water de-oiling problems and will cause a
flowline capacity reduction because of the higher emulsion viscosity.
Extensive laboratory dehydration tests and an on-site dehydration
test were conducted to identify the emulsification tendency of Troll
crude. The results are summarized in Table 1 and discussed below.
Laboratory Tests. The laboratory investigations were aimed at as-
sessing the emulsification tendency of Troll crude when produced
simultaneously with formation water and the means of separating
Copyright 1989 Society of Petroleum Engineers
34
oil and water from the emulsion created; assessing the feasibility
of de-oiling the separated water; and measuring the emulsion vis-
cosities at different temperatures, shear rates, and water cuts.
Dehydration and De-Oiling Tests. Dehydration experiments were
performed with emulsions containing 20% water. The emulsions
were formed by mixing crude and formation water for 2 minutes
at the anticipated wellhead temperature of 66C [151 P] in a high-
speed blender. The demulsifier was added and dispersed in the
emulsion by an additional lO-second mixing before transfer into
dehydration bottles, which were heated in a water bath set at the
desired temperature.
It was found that effective dehydration required temperatures in
excess of 40C [104P] and demulsifier treatment of some 200 ppm
(by volume, see Pig. 2). A static settling time of about 10 minutes
was required to obtain residual water cuts of 2 to 4 % .
Sludge buildup was observed at the oil/water interface at settling
temperatures less than 20C [68F]. At 5C [41F], free water
did not separate out.
The separated water containing up to 1,000 ppm (by weight) oil
required de-oiling by gas flotation to achieve a disposal quality of
25-ppm (by weight) oil in water (Fig. 3). The use of flocculant
during gas flotation reduced the time required to achieve 25 ppm
from 2 or 3 minutes to 1 minute.
Laboratory de-oiling tests on a water sample collected during the
on-site test in Well 3112-5 indicated a much lower de-oiling rate
(Fig. 3). This sample is considered unrepresentative of Troll for-
mation water because it was aged and contaminated with iron rust.
Coalescence Tests. To simulate the effect of pipe coalescence,
the treated emulsion was sheared between two concentric cylinders
before transfer into settling bottles. A mixing intensity of 4800 Pa' s
[4.8 cp] for 15 minutes was applied at 20,40, and 60C [68, 104,
and 1400P]. After coalescence, the emulsion was transferred into
a calibrated tube and allowed to settle at 20,40, or 60C [68, 104,
or 1400P].
Simulated pipe coalescence at temperatures above 40C [104F],
combined with demulsifier injection, is beneficial for dehydration
performance (Table 2), giving less than 1 % residual water in oil
after 30 minutes of settling time. No pipe coalescence can be ex-
pected at 20C [68P]. In an uninsulated flowline, the temperature
(Pig. 4) will rapidly decrease to below 20C [68P] and the resi-
dence time in the hotter section of the flowline will not be suffi-
cient for effective coalescence.
If the flowline were insulated, a longer section of the flowline
would be favorable for demulsifier reaction and pipe coalescence.
Separation of water from the emulsion will thus be enhanced; con-
sequently, flowline hydraulics (see the following discussion) will
become more favorable.
Viscosity Measurements. Emulsion viscosities have been
measured as a function of water content (10, 20, and 40%), tem-
perature, and shear rate in a thermostatted rotating viscometer. The
shear rates were varied between 0.277 and 27.7 seconds - I, with
measurements taken at temperatures between 5 and 20C [41 and
68F]. Above 20C [68P], water separated from the emulsion,
rendering viscosity measurements unreliable. The apparent viscosity
SPE Production Engineering, February 1989
31/2
3115
KEY
-f-
Fig. 1-Troll field.
of the emulsion below 20C [68F] increases drastically with the
water cut in the emulsion and decreases with increasing shear rate
(Fig. 5). Emulsions containing more than 20% water were found
to behave as pseudoplastic fluids.
The representativeness of the laboratory-generated emulsions,
however, is doubtful. Field experience with emulsified light North
Sea oils over the last few years indicates that laboratory tests have
consistently been conservative. In the field, often only a fraction
of the liquid phases appeared to be present as emulsion and the emul-
sified parts were generally found easier to break than in the labo-
ratory.
Fig. 5 compares the laboratory-measured viscosity with the vis-
cosities derived from field correlation. The latter are considered
the more realistic.
Compatibility of Demulsifier With Other Chemicals. Addition
of up to 200-ppm concentration of wax modifier did not affect the
30
25
30
25
~
~ to
a:
30
25
#
5" 20 ........
15
J 15
~ 10
i
\
\
SETILING AT 20 DEG C
10 20
SETTLING TIME, MIN
SETILING AT 40 DEG C
10 20
SETTUNG TIME, M!N
SETILING AT 60 DEG C
DEMULSIFIER-USED
200 PPM LAB. TEST
400 PPM FIELD TEST
30
30
CURVE ll. OLD WATERIOlD OIL
CURVE -0 FORMATION WATEAlFRESH Oil
CURVE 0 SYNTH. FOAM. WATER/FRESH Oil
CURVE 0 FIELD EMULSION (3112-5)
\
-..;;::\
' - ~ - - - - - - - -
10 20 30
SETTLING TIME. MIN
Fig. 2-Dehydratlon performance of field and artificially made
emulsions.
dehydration performance of the treated emulsion, even when the
two treating chemicals were premixed and pumped through a pipe
loop to simulate transport in a single subsea injection line to the
subsea wellhead.
Foaming Tendency of Troll Crude. Laboratory experiments in-
dicated formation of unstable foams from Troll crude, suggesting
that foaming in the separation facilities will not be a problem.
On-Site Dehydration Test. The dehydration behavior of produced
emulsions was studied on site in Well 31/2-5 to confirm the de-
hydration results obtained in laboratory experiments. A simplified
TABLE 1-SUMMARY OF EMULSION TESTS
Emulsion
Formation
Severity
Water-in-oil droplet size, I'm
Dehydration
Temperature, C
Chemical dosage, ppm by volume
Residence time (minutes)
Residual water (vol%)
Oil-in-water, ppm by weight
De-oiling (to <25 ppm by weight)
Residence time with flocculant,
minutes
Sludge accumulation
at settling temperature, C
SPE Production Engineering, February 1989
Laboratory
Yes
Tight
10 to 20
<!40
200
<10
2 to 4
700 to 1,000
1 to 2
s20
Field
Yes
Moderate
10 to 20
<!40
200 to 400
<10
0.1 to 1
1,200 to 2,300
s20
35
OIL CONTENT
PPMW
1000
900
800
700
600
500
400
300
200
100
<D Separated water from laboratory
dehydration test no chemical flocculant
As with <D PPMV FRI 869
@ Aged field water sample (Ex. 31/25)
+ 10 PPMV FRI 869
Aged field water sample as In @
with no chemical flocculant
These performance curves are considered
not representative of Troll water due to
aging and contamination
10 11 12 13 14 15
MINUTES FLOTATION TIME ____
Fig. 3-Effect of dispersed air flotation on oil content of water.
sketch of the production test facilities and the sampling points is
shown in Fig. 6. Demulsifier can be injected into the production
stream just upstream of the choke manifold. In the heat exchanger,
the temperature of the production can be raised (if necessary) to
70 to 80C [158 to 176F].
The conclusions of the on-site test were as follows.
1. Troll oil readily emulsifies with formation water to form stable
emulsions. Emulsions were formed upstream of the choke, and si-
multaneous heat and chemical treatments were required for de-
hydration. It was observed that free water does separate out from
a sample of untreated emulsion. Demulsifier treatment at 400 ppm
(by volume) enhanced the free-water separation. The temperature
required for dehydration was at least 40C [104F]. At a settling
temperature of 20C [68F], a sludge layer was formed at the
oil/water interface. At settling temperatures above 40C [104 OF],
no sludge was observed and a sharp oil/water interface was obtained.
A static settling time of less than 10 minutes was required to
achieve residual water cuts between 0.1 and 1.0% (Fig. 2). This
is better than that obtained in the laboratory because of the somewhat
"softer" field emulsion and perhaps some pipe coalescence. The
required chemical dosage was 400 ppm (by volume), which is twice
that required in the laboratory. In general, the reverse would nor-
mally be expected; the higher chemical dosage required during the
on-site test might be a result of inadequate reaction time, because
the demulsifier was injected in the cold stream (about 25C [77F])
only some 20 m [66 ft] upstream of the sampling point.
2. The separated water contained 1,200 to 2,300 ppm (by weight)
oil in the form of an oil-in-water emulsion. This is greater than
the oil levels obtained in the laboratory and is attributable to the
higher dosage of demulsifier in combination with the presence of
suspended solids [2,000 ppm (by weight)].
The water sample from the on-site test was subsequently de-oiled
with a bench model flotation unit in the laboratory. A much slower
de-oiling performance was obtained (Fig. 3).
3. Some additional emulsification seemed to be taking place in
the test separator, as evidenced by the relatively slower separation
of free water from samples taken downstream of the separator. This
phenomenon suggests that breakout of free gas from the liquid phase
can hamper effective coalescence of the water droplets, affecting
dehydration efficiency.
4. With cold emulsion in the riser and flowline during a well shut-
in, no problems with restartability were experienced, indicating that
emulsion blocking will not be a problem.
Implication for Facilities. It has been demonstrated that Troll
emulsions can be dehydrated effectively by applying heat (temper-
atures in excess of 40C [104 OF] are required) and chemical [200-
ppm (by volume) dosage] treatments, plus a residence time of up
to 10 minutes, to allow adequate settling of water.
The crude separation facilities for a possible Troll oil development
are shown in Fig. 7. The crude stream from subsea wells with an
expected platform arrival temperature of about 5C [41F] will be
warmed by heat exchange with the export crude; free water is
knocked out in a first-stage separator and the crude is then heated
to the desired process temperature (>40C[> 104F]) before it
enters the second-stage separator. Provision for the injection of
demulsifier is required at the wellhead and upstream of the crude
heater.
An unfavorable effect of gas breakout on dehydration has been
observed in the field, and complete separation of water in the
primary separators cannot be expected. A dedicated liquid/liquid
separator downstream of gas/liquid separators has therefore been
included.
Gravity settling in CPI will remove only free oil; effective de-
oiling of water, however, can be achieved with gas flotation. Ad-
dition of chemical flocculant improved flotation performance, but
chemical alone was not effective.
Although severe foaming of Troll crude was not indicated in the
laboratory tests, provision to inject defoamer upstream of the sepa-
rators is provided.
Wax
The temperature of crude flowing through an uninsulated flowline
will be reduced to seabed temperature (-5C [-41F]) a short
distance from the wellhead (Fig. 4). The Troll crude, which contains
some 2.5% wax, may deposit part of its wax on the pipe walls when
the crude temperature has decreased to below the wax crystalli-
zation temperature of 27C [81F].
TABLE 2-EFFECT OF PIPE COALESCENCE ON THE DEHYDRATION OF
TROLL CRUDE OIL
36
Water-in-Oil After Water-in-Oil After
Coalescence Settling
30 Minutes of Settling 30 Minutes of Settling
Temperature Temperature
(%) (ppm)
(0G) (OC) 2 3 1 2 3
---
20 20 no free water <2,000
40 40 3.6 0.4 0.6 350 700 470
60 60 2.4 0.2 0.7 570 180 120
20 60 0.6 2.1 720 490
'Only for Nos. 2 and 3.
1 = Without pipe coalescence.
2 = With pipe coalescence and demulsifier injection (200 ppm by weight) before coalescence treatment.
3 = With pipe coalescence and demulsifier injection (200 ppm by weight) after coalescence treatment.
SPE Production Engineering, February 1 9 ~ 9
FLOWING
TEMP C
60
50
40 ... ____ - - . 3 0 ~ - - - - ~ ...... ;;;:"'=='--- TEMP. VERY FAVOURABLE FOR COALESCENCE
....... :!'---------""' ...... :!:----- WAX DISSOLUTION (35CJ
30
PLATFORM
20
10
NO INSULATION (4000 BID)
5km 10 k.m "5 km
10 20 30 40
DISTANCE IN THOUSAND FEET ALONG SUBSEA FLOWLINE
Fig. 4-Estlmated flowing temperature profile along 5-ln.-10 subsea oil flowlines.
10000
1000
100
10
BS & W, Shear
% rate,s'
I
20
028
596
KEY.
---- MEASURED (3112-5)
- - - - FIELD CORAELATION
NOTE- MEASURED ON AATIFICAlLY
MADE TIGHT EMULSION
BS&W,
%,
I
.0
I
60
40
20
g(DEAD OIL)
TEMPERATURE 0C
Fig. 5-Comparison of laboratory-measured and field-
correlation viscosities.
Wax deposition was indicated during several Troll oil appraisal-
well production tests. However, no serious operational problems
were experienced during the short tests.
Wax deposition after crystallization depends on the temperature
differential between the bulk crude stream and the pipe wall and
on the Reynolds number, The larger the temperature differential,
SPE ProductiM Engineering, February 1989
WELLHEAD --'---.--ff'I;.t)--'--{
DEMUL$IF!ER
WATER
DISCHARGE
= SAMPLE POINT
SKIM
TANK
GAUGe
TANK
GAS TO FLARE
OIL TO BURNERS
Fig. 6-Schematic of production test facilities.
the greater the chance of wax deposition because molecular diffusion
of dissolved wax to the pipe wall intensifies as a result of the in-
crease in wax concentration near the pipe wall brought about by
the temperature differential. Maximum wax accumulation is ex-
pected in the Reynolds number transition range (between 2,000 and
4,000). At low (laminar) flow, the net transport of wax to the wall
is reduced by a relatively thick laminar layer adjacent to the pipe
wall; during turbulent flow, wax buildup is limited by erosion.
The addition of wax modifier to crude disrupts the normal crys-
tallization mechanism of wax, thereby reducing wax deposition,
Because the chemical reacts during the onset of crystallization by
cocrystallization, it is effective only when introduced at tempera-
tures above the cloud point.
The wax modifier keeps the wax crystals as discrete small
structures with low cohesive properties and may lower the adhesive
forces between the wax crystals and other surfaces.
37
ass. gas ass. gas
sea level
water
C.1 (5)
CRUDE SEPARATOR FACILITIES
WATERTREATMENT FACILITIES
C.I. (1,2,6)
gas lift
sea bottom
WEMCO
C.I (5)
r L , ~ " ,
Note: C.I = chemical injection
1 = wax modifier
2 = demulsifier
3 = Scale inhibitor
4 = defoamer
5 = floculiant
6 = methanol
Fig. 7-Prellminary Troll crude 011 production facilities and produced-water treatment system.
Various laboratory tests (Table 3) were conducted to assess
waxing characteristics of the crude, restartability of cold crude in
a flowline, treatability of the wax deposition by chemicals, disso-
lution of deposited wax by heat treatment, and rate of wax depo-
sition in the subsea flowline. In addition, a wax-deposition test was
conducted in the field during the production test of Well 3112-14
to substantiate the laboratory test results.
Laboratory Tests. Cold Startup. This test was perfonned to define
the restartability of crude in the subsea flowline after a well shut-
in. It was carried out in a stainless steel coil with a 5-m [16-ft] length
and a 4-rrun [0.16-in.] !D.
The coil was filled with a bottomhole crude sample and immersed
in a thermostatted bath. After 1 hour at 46C [115P], the temper-
ature of the bath was gradually decreased at 1.5C/h [2.7P/hr]
to 4C [39P] and maintained at this temperature for 2 weeks. After
the test period, flow was obtained by applying 4 kPa [0.04 bar]
(equivalent to 207 kPa [30 psi] in the field with a 12.7-cm [5-in.]
flowline), indicating that cold startup would not be a problem in
the field.
Wax-Deposition Tendency. The objective of this test was to assess
the wax-deposition tendency of Troll crude in a cool subsea flowline.
The equipment consists of a thennostatted stirred vessel from which
crude oil is circulated by a gear pump through a thermostatically
controlled stainless steel capillary. The temperature in the vessel
was maintained at 65C [149P], and the capillary tube was held
at a constant temperature below the cloud point.
It was confinned that maximum wax deposition occurs when the
temperature differential between the bulk crude stream and the tube
wall is a maximum (Table 4). Por example, with a crude temper-
ature of65ClI49F] at the inlet of the capillary-i.e., well above
the sample cloud point of 30C [86P]-the wax deposition onto
a wall was approximately 100 times greater at 5C [41P] than at
10 or 20C [50 or 68P).
Wax Dissolution. This test was conducted primarily to establish
the temperature at which deposited wax would redissolve. The ex-
periment was conducted in the capillary test rig used for the wax-
deposition testing.
The crude was circulated through the test tube with a wall tem-
perature of 5C [41 P] and a crude temperature of 65C [149P].
After 2 hours, the test-tube temperature was increased continuously
at a rate of2C/h [3.6P/hr]. A plot of pressure vs. time and wall
temperatures (Pig. 8) shows that wax deposition occurs at wall tem-
peratures below 25C [77P]. Above 25C [77P], the pressure
TABLE 3-SUMMARY OF WAX TESTS
38
Crude properties
Density at 15C, g/cm3
Wax content, %
Cloud point, C
Pour point, C
Congealing point, C
Viscosity at 40C, cSt
Deposition criteria
Deposition rate, mm/month
Congealing point, C
Dissolution temperature, C
"Hardness"
Restartability
Type deposit
'NO = not determined.
Laboratory
0.882 to 0.886
2.5
27 to 31
-21 to -30
60 to 69
8.1
ND'
60 to 78
35
Soft
OK
ND
Field
2
68 to 85
Soft
OK
Paraffinic
(H/C=2)
SPE Production Engineering, February 1989
TABLE 4-CAPILLARY TEST RESULTS, WAX DEPOSITION
Storage-Vessel Temperature Varied; 5C Capillary-Tube Temperature
Vessel Circulation Deposit Wax Congealing
Temperature Time Build-Up Point""
(0C) (hours)

(OC)
65 5 276 76
40 2.5" 163 69
20 5 77 67
65 5 207 78
65C Storage-Vessel Temperature; Capillary-Tube Temperature Varied
Capillary-Tube Circulating Deposit Wax Congealing
Temperature Time Build-Up Point
(0C) (hours)

(OC)
20 24 9 NOt
10 24 8 NO
5 5 207 77
5 (repeat) 5 206 78
5 (+ 100 ppm by weight 5 83 76
wax modifier)
5 ( + 200 ppm by weight 5 2 NO
wax modifier
'Experiment shortened because of heavy
"Field wax sample congealing points 68-85
0
C,
t Not determined because of insufficient 'amount of deposit.
0.9
0.1
DISSOLUTION
ABOVE 25C
COMPLETE
DISSOLUTION
AT 35C
O.O+-------r-----"""T------,------,
5 10 15 Time hrs. 20
'5 5 i5 Js
ConstantWail temp.
Crude temperature 65C
Wall temperature C
Fig. 8-Characterlstlcs of Troll wax deposition and disso-
lution.
decreased, indicating dissolution of wax; at about 35C [95P],
the pressure equaled the initial pressure, indicating that all the
deposited wax had been dissolved.
Wax-Deposition Treatment. In the same capillary test loop, the
severity of wax deposition was found to be reduced by a factor of
100 when wax modifier was used (Table 4).
On-Site Wax Test. During June 1984, a wax-deposition test was
conducted in Appraisal Well 3112-14. The objectives were to assess
the extent of wax deposition on site and to compare it with the lab-
oratory wax-deposition behavior in the capillary test loop.
SPE Production Engineering, February 1989
PRESSURE
PSIA bara
1000 69
900 .2
800 552
700 "'S.3
600 41.4
500 34 5
/
"'" '76 /:
'I
, "
'1
I :
POSSIBLE I I
HYDRATE
FORMATION
I
I
WELL PRODUCING AT HIGH GOR
NO HYDRATE
FORMATION
/
i WELL PRODUCING AT SOLUTK>N GOR
/
I
/
/
/
ASSUMPTIONS:
5"'011 km UNINSULATED FLOWLINE
4W'OD TUBING
PI 20 BOFIPSI
STABILISED CONDITIONS FOR FLOWLINE
_ RISER
IITillTI MAJOR PORTION OF FLOWLINE /
/
300 20 7
IIIIIIIIIID FLOWLINE WITHIN 1200 m OF WELLHEAD
200 13.8
100 69
... PLATFORM INLET
AT STARTUP TEMPERATURE
THROUGHOUT WILL BE APPROX Soc
(SEA BED TEMP)
,'--,'--,[---------------
1 0 2 4 6 8 10 12 14 15 16 18 20
TEMPERATURE JOC)
Fig. 9-Flowllne operating temperature and pressure ranges
related to hydrate-formation region.
To measure wax deposition, 4.78-cm [1.88-in.] -ID, 3-m [lO-ft]
-long flow tubes were installed at three different locations in the
tubing riser during the production test. Nipples to hang off the flow
tubes were preinstalled in the tubing riser at locations that would
cover the predicted temperature profile.
The well was flowed for 5 hours at about 95 m
3
/d [600 BID],
corresponding to transition flow. The flow tubes were then replaced
39
and the well flowed at about 318 m
3
/d [2,000 B/D] for 5 hours
to show the erosional effect of turbulent flow on wax deposits.
Wax deposits were recovered from all flow tubes, and the depo-
sition rate at the wall under transition flow conditions was estimated
to be 0.003 mm/h [0.00012 in.lhr]. Turbulence appeared to erode
wax deposits, reducing accumulation on the pipe wall.
The wax deposits were predominantly paraffin (hydrogen-to-
carbon ratio of approximately two), with congealing points ranging
between 68 and 85C [154 and 185P]. The wax deposits generated
in the laboratory have congealing points in the range of 67 to 79C
[153 to 174P].
Implication for Facilities. Flow in an uninsulated subsea flowline
is cooled from 55C [13 lOP] to a seabed temperature of 5C [41 P]
(Pig. 4) some 1200 m [3,940 ft] from the wellhead. Beyond this
point, wax deposition will be minimal because there is no temper-
ature differential. Within 60 m [200 ft] of the wellhead, the stream
temperature is in excess of 35C [95P] and no accumulation of
wax will occur; any wax deposited during a shutdown would be
redissolved when flowing.
Untreated crude will deposit wax in the section where cooling
occurs (between 60 and 1200 m [200 and 3,940 ft] from the
wellhead), but this can be minimized by continuous doping with
wax modifier. Periodic pigging can be expected to remove the soft
wax deposits easily.
Pull insulation of the flowline to achieve crude arrival tempera-
tures in excess of 35C [95P] is prohibitively expensive and hard
to realize. Potential wax deposition must therefore be countered
by the provision of pigging/scraping facilities and by injection of
wax modifier at the wellhead.
Downhole wax-inhibitor injection will not be required because
the stabilized flowing temperature within the well tubing and at the
wellhead will be about 55C [131 P], which is well above the cloud
point and wax-dissolution point. If wax deposition occurs in the
cooler top section of the tubing during well shut-in, the deposits
should dissolve under flowing conditions. Moreover, they are likely
to be thin and soft and should not cause operational problems.
Hydrates
Hydrates are crystalline compounds consisting of hydrocarbon
molecules occupying voids within a framework of water molecules.
Hydrocarbon molecules, which are essential for the formation of
hydrates, are methane, ethane, propane, and isobutane. Other
molecules, such as CO
2
and H
2
S, are also suitable. The lattice
structure formed by the water molecules varies depending on the
guest molecules associated with the particular hydrate.
If present in sufficient quantities, hydrates can potentially plug
flow conduits. Pield experience of hydrate plugging in oil wells
is scarce. In the few instances of which we have evidence, hydrate
plugs tended to form at restrictions, dips, valves, and vertical
sections. Depressurization, sometimes followed by supplemental
flushing with methanol, overcame the problems. Recent work
2
in-
dicates that hydrate problems have been experienced in the Magnus
subsea wells. Depressurization was also effective in overcoming
these problems.
Laboratory Tests. A gas/oil/water mixture was introduced into
a windowed cell that could be agitated, cooled, or heated. Water
was maintained at about 25 wt% in the mixture. Tests were con-
ducted at normal (65 std m
3
/m
3
[360 scf/bbl]) and high (540 std
m
3
/m
3
[3,000 scflbbl]) GOR's.
Cooling resulted in spontaneous hydrate formation at tempera-
tures between -I and -2C [30 and 28P] for low GOR's and
between 6 and 7C [43 and 45P] for high GOR's, independent
of pressure. The celI was heated and the temperature and pressure
at which the hydrate sublimated were recorded. The spontaneous-
hydrate-formation temperature was much lower in both the high-
and low-GOR cases, indicating a degree of supersaturation with
40
respect to hydrate formation. A summary of the measured temper-
atures and pressures at which the hydrate sublimated is shown in
Pig. 9.
The effect of methanol on hydrate-formation temperature was
measured and a suppression power ofO.4C/wt% [O.72P/wt%]
methanol was derived. This is in line with the literature
(O.4IC/wt% [0.74P/wt%]).
Implications for Facilities. Pig. 9 relates the flowline and riser
conditions anticipated in Troll subsea oil welIs to the possible
hydrate-formation region, defined by the experimentally measured
hydrate decomposition temperature and pressure. The subsea
flowline and riser conditions can be expected to vary depending
on flowline length and diameter, welI PI, GOR, water cut, reservoir
pressure, etc. To illustrate the possible variation, two cases
representing high and solution GOR's are given in Pig. 9. At startup,
temperatures will be about 5C [41P] (seabed temperature) for
all cases.
It is clear that during startup, conditions conducive to hydrate
formation can exist at the wellhead and flowline and in the riser.
Once flow has stabilized, conditions suitable for hydrate formation
can exist in the subsea flowline, particularly at higher GOR's.
Hydrate formation in the subsea lines is therefore a recognized
risk, and various measures are required to overcome the potential
problem: (I) provision for glycol/methanol injection should be made
at the wellhead and (2) the flowline should be insulated to achieve
a platform arrival temperature of 15C [59P]. Achieving this tem-
perature avoids hydrate problems during stabilized flow and is
beneficial for treatment of emulsion and wax.
Conclusions
Owing to cooling in the subsea flowline, subsea wells can be ex-
pected to pose more production chemistry problems than would be
expected with platform wells. The laboratory and field tests de-
scribed provided an effective and inexpensive means of identifying
potential problems during the field appraisal. As a result, provisions
to overcome the problem could be incorporated into the facilities
design at an early stage.
Acknowledgments
We acknowledge the contribution of the personnel at the Shell Group
Research Laboratories, KSEPL and KSLA, who carried out the
laboratory studies described in this paper. We thank AIS Norske
Shell and the License 054 partners-Statoil, Norsk Hydro, Conoco
Norge, and Mobil-for permission to publish this paper.
References
I. Stiff, H.A. and Davies, L.E.: "Method for Predicting the Tendency
of Oil Field Water to Deposit Calcium Carbonate," Trans., AIME
(1952) 195,213-18.
2. Dawson, A.P. and Murray, M.V.: "Magnus Subsea Wells: Design,
Installation, and Early Operational Experience," SPEPE (Nov. 1987)
305-12; Trans., AIME, 289.
SI Metric Conversion Factors
bar x 1.0* E+02 kPa
bbl x 1.589 873 E-Ol m
3
cSt x 1.0* E-06 m
2
/s
ft x 3.048* E-Ol m
0p
(OP - 32)/1.8 C
in. x 2.54* E+OO cm
miles x 1.609 344* E+OO km
psi x 6.894757 E+OO kPa
* Conversion factor is exact. SPEPE
Original SPE manuscript received for review Sept. 18, 1986. Paper accepted for publication
Nov. 4,1987. Revised manuscript received June 10. 1988. Paper (SPE 15892) first presented
at the 1986 SPE European Petroleum Conference held in London, Oct. 20-22.
SPE Production Engineering, February 1989

Potrebbero piacerti anche