Professional Documents
Culture Documents
works well in reservoirs with excess injectivity and low ability of the gel to propagate very deep into the reservoir
salinity where the polymer can be injected at significant rates still exist.
with relatively high viscosity, typically 10 cP at about 7 sec-1. A system was still required which can reduce the
Problems exist that render it a less attractive option in most permeability of thief zones deep within the oil reservoir to
reservoirs. Polymer solutions are typically shear thinning, and achieve more efficient displacement of the oil to the producing
the point of injection is normally the highest shear rate wells. An essential feature was seen as having only one
environment. The shear rate falls rapidly as the solution enters injected component, so that no separation could occur. Also,
the reservoir. The viscosity of the injected polymer solution density close to that of an average injection brine was seen as
increases as the shear rate falls and a viscous bank is formed desirable, to minimize segregation. This paper reports some of
around the injection well. This reduces the injection rate the results from a joint oil industry research consortium
achievable. The majority of water floods operate with project aimed at providing such a system. Figure 2 below
minimum excess injection capacity, and reduction of illustrates conceptually how incremental oil could result from
injectivity can mean that new injection wells need to be drilled such a system.
to allow polymer flooding. This is often an undesirable
®
prospect. A further limit is imposed on the injection rate by Figure 2: Thief zone plugged with Bright Water treatment
the shear instability of some polymers. If the shear rate during
injection is too high, it can break the molecules apart and
significantly decrease the cost effectiveness of the process.
Near wellbore treatments have been used in attempts to
correct waterflood sweep profile. However, in general, the
depth of penetration, is typically no more than 15 feet (5
meters), and is too small to exert a controlling influence on
reservoir flow unless extensive barriers to fluid movement
exist orthogonal to the well. Even if vertical conformance is
corrected, areal conformance problems can still be significant. Field simulations within BP Exploration and
Alternating slug treatments using synthetic polymers Technology led to an initial estimate that over the life of a
cross-linked with aluminium ions supplied as Aluminium field improved volumetric conformance could increase
Citrate were proposed by Pye as a means of treating injection recovery up 10 percent in favorable cases. It can also
profile control problems.5 These were widely used in the contribute to a significant reduction in operating costs related
field, and with some variations have achieved a degree of to reduced water production.
success.6 The viscous polymer slugs tended to divert the
crosslinker into different zones, so it may be hard to get them Statement of Theory and Definition
to mix and the degree of mixing caused in the reservoir is not Concept of the reagent. In 1996 an industry and academic
controllable. Additionally, there are problems in propagating seminar, sponsored by BP, produced numerous potential
aluminium ions (as a citrate complex) deep into a reservoir options for achieving physical or chemical changes using
rock. Loss of aluminium to the rock matrix by adsorption or “triggers”, i.e. properties of the reservoir that differ between
precipitation often occurs which means that large quantities of thief zones and the bulk of the rock. The concept eventually
aluminium can be required to establish in – depth treatment.7,8 developed was that of a particle which could inject and
Deep Diverting Gels (DDG) have been proposed as a propagate with the water flood through the pores of the rock
means of achieving placement of a cross-linked polymer slug matrix, then after a temperature change in the thief zone(s) or
deep within the reservoir.9 An AcrylamideAcrylate polymer after a certain time, would increase in volume (Popping) to
was mixed with Aluminium Citrate under controlled block pore throats, diverting chase water into poorly swept
conditions. Association between the polymer and aluminium zones. To maintain injectivity, it was important that the
ions took place, but a gel was not formed for a considerable injected viscosity of the system should be as near as possible
time. A field trial of the process showed that the injection to that of water. Once popped, interactions with pore throats
profile was modified as expected but the improvement was were intended to be the means of delivering water
premature and did not last for long. Loss of the aluminium to resistance factor.
the porous medium was believed to be the major limiting It was envisaged that a large volume of such a reagent
factor. The permeability modification appeared to be might have to be injected to achieve a major influence on
primarily in the region of the injection well. reservoir sweep, which dictated that it should be low cost.
Colloidal Dispersion Gel (CDG) is related to DDG in Rock thermal properties together with the water
being composed of a mixture of polymer and aluminium injection and reservoir temperatures determine the conditions
crosslinker. Association between the polymer and the metal under which the particles must propagate and trigger. The
ion again lies at the heart of the process where intramolecular diameters and rheological properties of the particles before
crosslinking is believed to create cross-linked polymer popping must be compatible with the pore throat size
molecules or small aggregates in the formation. Willhite et al distribution of the target rock. A typical pore throat size
showed the injectivity of the mixture to be limited at low shear distribution for Minas sandstone, obtained by mercury
rate.10 Smith et al showed that the mixtures could be injected porosimetry, is shown below in Figure 3.
at higher shear rates.11 Given the rapid fall in shear rate with
distance away from an injection well, questions about the
SPE 84897 3
Figure 3: Pore throat size distribution for CPI Minas Core sand pack method used was very similar to that reported by
Morgan et al.12
The Bright Water® kernels are supplied as a 30% active
dispersion of polymer in light mineral oil. Test solutions were
made by high shear dispersion of the oleic concentrate into
water with an added surfactant, using a high shear mixer.
Before use, the kernels “solutions” were conditioned in an
ultrasonic bath. The ultrasonic bath was used to simulate the
individual particles dispersed in water by field equipment.
to use a shut-in period after injection to allow sufficient time wells have local injection-induced fractures. Therefore, these
for activation. stratified targets remain an option for Bright Water®,
particularly where areal and vertical sweep issues both exist.
Description and Application of Processes However, the preferred reservoir target type for the
The field trial was to demonstrate that the polymer solution initial Bright Water® version tested here had some
could be injected at normal pre-trial well rates and pressures, communication between swept and unswept zones. With this
would activate as planned deep within the formation, and structure, when the thief zone was blocked deep in the
would divert water to the bypassed oil zone and improve reservoir, pressure would build up between the injector and
waterflood sweep efficiency. the block, and water would exit the thief zone, being diverted
into lower pressure unswept zones.
Screening fields for sweep efficiency improvements. Many The material initially produced had a fixed particle size,
factors, some positive and some negative, need to be which meant that the thief zone permeability had to be within
considered when ranking candidates for in-depth sweep a required range. Also, the activation time was temperature
improvement. Most can be covered by these questions: dependent. The temperature encountered in the reservoir had
1. Is there sufficient target in terms of improved oil to be at least 80C for activation to occur in an acceptable time.
rate/reserves (or reduced water production)? The injection water temperature had to be no more than 70C,
2. Is there a place where water diversion can be activated to to avoid activation near the wellbore. Cold injected water will
mobilize and recover these reserves? cool the rock it encounters, an excessive volume of cold-water
3. Can the process be placed there without encountering or injection prior to the treatment could mean that the particles
causing technical problems? injected would not see reservoir temperature rock until far
4. Can the predicted waterflood improvement be observed from the well. Activation might not then occur in an
and quantified, and is it economic? acceptable time or distance. All these considerations came
5. Is there an easier, cheaper or lower risk option to achieve into the initial target screening for this project. Since then,
the same ends? these restrictions have been eased considerably by
The approach used was to eliminate targets with further development.
obvious negative factors, and then to rank remaining targets The following points should be considered when
based on a weighted multi-question risk assessment, based evaluating a potential candidate for Bright Water®. Preferred
primarily on the above questions. target properties:
Numerical simulation, and laboratory tests, can also be 1. Available movable oil reserves
used to optimize the treatment placement, and the best 2. Early water breakthrough to high water-cut
treatment size/strength for each candidate. This approach 3. Problem with high permeability contrast, (thief zone at
helps quantify the likely production gains as well as treatment least 5 times unswept zone)
costs. It can also compare in-depth diversion with alternative 4. Porosity of highest perm zone > 17%
well treatment options, and with alternative waterflood 5. Permeability in thief zone >100 mD
management options. In practice, the acquisition of the 6. Minimal reservoir fracturing
required local reservoir description and performance data, plus 7. Temperature from 50 C (122 F) to 150 C (302 F)
the simulation of many targets can cost more than the 8. Expected injector-producer transit time >30 days
treatment itself. Therefore, we usually defer this work until 9. Injection water salinity under 70,000ppm
candidates have been pre-selected based on a simple
ranking process. Pattern selection within the Minas Field. Many variables
Usually, reservoir engineers are well aware of the must be considered when selecting a good candidate for
efficiency of their waterfloods. When the localized recovery profile modification, including evidence of a thief zone,
factor in a given reservoir sector is predicted to be high (60% bypassed oil reserves target, and well pair connection.
of OOIP or better), and the amount of water required to After initial screening, the most attractive patterns were
achieve this is 1.5 sector pore volumes or less, then there will further evaluated using pressure pulse tests to verify injector
be no significant target for remedial treatments. /producer connections. The selection of Pattern 7E-12,
Generally, the worse the waterflood performs, the more targeting the A1 Sand, resulted from the pulse testing. The
likely it is that an economic remedial treatment can be applied, test showed offset producers 6E-97, 7E-13, and 7E-23 were
but there are exceptions. Examples include, poor sweep due hydraulically connected to injector 7E-12 while other wells
to excessive reservoir compartmentalization or lack of within the pattern were not connected.
reservoir continuity, and highly fractured reservoirs. The Finally, tracer was injected and monitored to determine
treatment presented in this paper is intended for diversion in the direction of flow, and verify the magnitude of the thief
matrix, not fractures. zone. Tracer tests confirmed fluid communication between
For reservoirs with vertical pay zones separated by these well pairs (Fig. 5), and showed the magnitude of those
impermeable shales, it should be simpler and less costly to connections (days to tracer breakthrough).
mechanically shut off individual watered-out layer(s) either at
the producer or injector. However, in some cases, the first
zones to water out still contain valuable reserves, and shut-off
at the producer is considered unacceptable. Also shutting off
an isolated zone at the injector is often difficult because the
SPE 84897 5
Figure 5: Tracer results in Minas 7E-12 pattern A pressure transient analysis was conducted in 7E-12 to
investigate k(mD) before the Bright Water® treatment.
Permeability was estimated in the range of 400 to 600 mD.
These were within the range of the 420 mD, the
log-derived permeability.
A spinner survey was conducted on 7E-12 to verify
fluid distribution at A1 sand after the A1 was isolated. The
log showed 96% of the injected water entered the lower
interval of the A1 sand, and the rest leaked below the packer.
The upper A1 showed no injection.
Post-treatment
Breakthrough (day) 18 115 43 108 108 115 122
Peak (concentration) 26 12 13 14 4 5 4
Difference
Days pre and post -12 72 -27 NA NA NA NA
SPE 84897 7
Figure 9: Production plot of 7E-12 pattern 1. The polymer was pumped at low viscosity, about twice
20.00
Pump
Incremental Oil ?
200
that of water at the same temperature.
18.00
Upsizes
(only pattern wells included) 180
2. The polymer was successfully displaced a significant
16.00 160 distance from the wellbore.
14.00 140 3. The material appeared to gel as designed, 125 feet
distance from the injector, and caused A1 Sand injectivity
M Bbls Water / CD
ST Bbls Oil / CD
12.00 120
10.00 100
to reduce by more than 60% from 13,000 BPD to
8.00
Bright Water
80
5,000 BPD.
Treatment
4. No polymer was produced in the offset wells and tracer
6.00 60
survey indicated flow directions were altered.
4.00
Avg. FIFO = .35
40
5. The popping time can be manipulated by pH control,
after BW Job
2.00 20
though we can now also manipulate it just by
0.00
Ma Ap Ma Ju Ju Au Se Oc No De Ja Fe Ma Ap Ma Ju Ju Au Se Oc No De Ja
0
molecular design.
r-
01
r-
01
y-
01
n-
01
l-
01
g-
01
p-
01
t-
01
v-
01
c-
01
n-
02
b-
02
r-
02
r-
02
y-
02
n-
02
l-
02
g-
02
p-
02
t-
02
v-
02
c-
02
n-
03 6. Pumping was trouble free in the Minas field trial, and in
FIFO M CDwater
Time
M CDwinj CDoil
the simplified North Sea field trial.
7. The volume of incremental oil production attributable to
Although this is monthly data it is interesting to note the project is uncertain. As discussed, a number of factors
that the response time is faster than expected. The transient likely contributed to this outcome, factors which may or
time from the tracer test, Table 1 referenced above, did may not be related to the effectiveness of this new thief-
indicate cycle time from 3 to 6 weeks for two of the offset blocking agent. Additional field trials are needed to
producers. The initial oil production response could have quantify the commercial viability of this concept.
resulted from oil being dislodged by the surfactant in the
leading edge of the polymer slug and being quickly swept Acknowledgements
ahead through the thief zone to the producers. The second We would like to acknowledge CPI engineering and support
bump in production, in the January-February 2001 time frame, functions, including Adly Guchi, Construction Coordinator
may be related to some diversion of the waterflood into the who saved the day for us logistically. Others included:
lower permeability zones. Fuaadi Ibrahim, Reservoir Engineering; Darmawan Boedi,
From the period leading up to the 7E-12 Bright Water® Facilities Engineering; Ginting Kasmir and Hadi Prijono,
treatment through the last attempts to gather data, a number of Facilities Electrical; Petrus Krisologus, Safety, Health &
operational factors come into play which make it difficult to Environmental; Pulangan, Operations Wells and Fac.; Andi
quantify exactly what volume of incremental oil production Bachtiar, Project Champion; Dike Fitriansyah Putra, Produce
resulted from the Bright Water® treatment. These factors Sub-Surface. Kudus Kurniawan was our Ondeo Nalco Energy
include pump changes just prior to the treatment as well as Services(Indonesia) representative.
injection water rate changes in and around the test pattern. We thank ChevronTexaco’s Les Munson for reservoir
The fluid-in fluid-out ratio for the pattern after the simulation, and candidate well selection, and Medhat Kamal
treatment was considerably less than one. A low voidage for the design and interpretation of the Pressure Fall-off test.
replacement ratio, coupled with not all of the wells in the
pattern being aligned with injection into the A1 sand, could Nomenclature
have contributed to the apparent low volume of OOIP, original oil in place
incremental oil. sec-1, reciprocal seconds, shear rate
It should be noted that given proper in-situ profile
modification candidates, with water cycling to offset References
producers in a thief zone(s), when the thief zone is plugged
and water injection into it is reduced, oil rates should improve 1. A. L. Benham, SPE 513, 1st November 1962. “A model
because of the re-distribution of areal support to the study of viscous fingering”.
offset producers. 2. E. Shirif, K. Elkadiffi & J. J. Hromek. SPE Reservoir
Evaluation and Engineering, February 2003. “Waterflood
Further development. A second field trial using the Bright performance under bottomwater conditions:
Experimental approach.”
Water® polymer was conducted in December 2002 on a North
3. A. L. Barnes, Journal of Petroleum Technology October
Sea (UK) production platform. A total of 88,360 bbls of 1962, 1147, Trans. AIME volume 225. “The use of a
injection seawater containing 1500 ppm of polymer and 750 viscous slug to improve waterflood efficiency in a reservoir
ppm of surfactant was pumped over a twenty (20) day period. partly invaded by bottom water.”
A much smaller and simplified pumping skid was used in this 4. D. J. Pye, Journal of Petroleum technology, August 1964,
second field trial. The per-barrel injected polymer cost for this 231, Trans. AIME volume 231. “Improved secondary
treatment was about $4.30/bbl. recovery by control of water mobility.”
5. R. B. Needham, C. B.Threlkeld & J. W. Gail, SPE 4747
Conclusions presented at the 1974 SPE Improved Oil Recovery
Symposium, Tulsa, OK, April 23 – 24. “Control of water
Several of the major objectives of the Minas field trial were
mobility using polymers and multivalent cations”
met while other important goals were not. 6. H. Surkalo, M. J. Pitts, B. Sloat and D. Larsen. SPE 14115
Presented at the 1986 SPE International meeting on
8 SPE 84897
Appendices