Spe 167274 Ms

SPE 167274
Downloaded from http://onepetro.org/SPEKOGS/proceedings-pdf/13KOGS/All-13KOGS/SPE-167274-MS/1537187/spe-167274-ms.pdf by Pandit Deendayal Petroleum University user on 04 November 2022
Strategies for Improving Sweep Efficiency and Waterflood Recovery in a
Complex Reservoir in the Greater Burgan Field; Learning from a Pilot Water
Injection Project and Evaluation Using Sector Model Simulation
S. Rajan, M. Al-Naqi, A.A. Ameen, H.H. Al-Hashash, S. Madhavan, M.N. Al-Qattan, N. Al-Enizi, A. Al-Qattan,
Kuwait Oil Company; A. D. Brooks, A. Beer, AAR Energy
Copyright 2013, Society of Petroleum Engineers
This paper was prepared for presentation at the SPE Kuwait Oil and Gas Show and Conference held in Mishref, Kuwait, 7-10 October 2013.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
Abstract
The Greater Burgan Field, operated by Kuwait Oil Company (KOC), is the largest clastic and overall the second largest
oilfield in the world. First discovered in 1938, with production from 1946, production to date has relied on primary recovery
methods. In recent years secondary and enhanced recovery techniques have been investigated with water flood now at an
advanced state of implementation. The first such water flood project is in the Upper Cretaceous (Cenomanian) Wara
Formation, one of the main producing reservoirs within the Greater Burgan complex where production has been accompanied
by falling reservoir pressure.
The Wara Formation comprises multiple sandstone units deposited in a fluvial-tidal coastal system. The reservoir exhibits a
considerable degree of permeability heterogeneity, which poses significant challenges to the implementation of waterflood.
The highly complex nature of the reservoir necessitate for implementation of suitable strategies from an early stage of
waterflood to ensure better volumetric sweep and ultimate recovery. Several initiatives/studies, including sector model
simulation, were undertaken to formulate strategies and evaluate the waterflood performance in this multilayered reservoir.
This paper summarizes results of the studies and strategies:
• Issues related to vertical sweep and lessons from pilot study.
• Strategies for improving vertical sweep.
• Evaluation of strategies by using sector model simulation to improve recovery.
• Well completion strategies and candidate selection criteria.
Introduction
The Wara sand, discovered in 1948 is the uppermost productive component of the Greater Burgan complex (Figure-1);
the largest sandstone reservoir in the world. The Wara reservoir is highly heterogeneous due to large variations in
depositional environment, cored wells having an average Dykstra parson coefficient of heterogeneity of 0.85. The reservoir is
divided in to 7 layers, starting from W1 at the top to W7, which is the lowermost layer, separated from the underlying Burgan
formation by the low permeability Mauddud limestone reservoir. Based on the varying depositional environments the Wara
layers are grouped into 3 units i.e. Lower Wara, Middle Wara and Upper Wara. The lower Wara (W5, W6 and W7), is
comprised mostly of marine shales with localized tidal and ridge / mouth bar / channel deposition occurring primarily in the
southern part of Burgan field area. Middle Wara (W4) sands were deposited in a relatively high energy environment, in a
progradational period with mostly lower delta plain and tidal channel deposition. Of the three units this sand is more
abundant and regionally continuous. The Upper Wara unit (W1, W2 and W3) is constituted of thin bars and distributary sand
deposits well developed in the Northern part of the field. From observations of pressure and sweep communication between
Wara reservoir units varies from isolated to tortuously connected and sometimes well connected (Adbul Rahman et al, 2012,
Ambastha et al, 2009, Al-Naqi et al, 2009).
To date, the dominant reservoir drive mechanism is solution gas drive; there being small degree of peripheral aquifer
support limited to several locations around the field. The field has been in under continuous production since 1948, the lack
of adequate aquifer support resulting in a significant drop in the reservoir pressure from original pressure of 2100 psi to 1500-
1600 psi resulting in the formation of a secondary gas cap in many areas.
2 SPE 167274
Figure-1: The location of the Greater Burgan reservoir
Several measures were initiated, ranging from short-term to long-term from 2004 onwards to arrest the decline in pressure
and to improve the recovery. Short-term strategies aimed at immediately arresting further decline in pressure by formulating
a suitable withdrawal strategy. Intermediate measures planned to augment the reservoir energy by means of step-in peripheral
water injection, known as Wara pressure maintenance project (Wara PMP), in along most of the periphery of the reservoir,
which is also expected to help in understanding the reservoir behaviour. In the longer term, full field water injection scheme
is planned to look after the entire Wara reservoir.
KOC quickly indentified that management of volumetric (vertical x areal) sweep efficiency would be a key challenge in
the implementation of a successful waterflood. In this paper we present work done on two important aspects of sweep
efficiency namely the conformance profile at the injection wells and completion strategy focusing mainly on the zones
completed in both production and injection wells. At the injection wells studies are based on experience in a large early water
injection pilot, known as the Early Wara Pressure Maintenance Project (or EWPMP). Since injection water breakthrough in
EWPMP has yet to be observed to any meaningful degree the thought process on completion strategy is guided mainly by
simulation studies based on a high-resolution geocellular model.
Injection Performance in the Early Wara Pressure Maintenance Project (EWPMP)

Matrix injection is a pre-requisite for layerwise control of injection, i.e. an attempt to create an ‘even’ distribution of inflow at
the injection wells. Under fractured injection water will preferentially flow into the zone in which the fracture is initated
(presuming the reservoir has the capacity to accept it with pressuring up). If a selective completion is employed using zonal
isolation packers then matrix injection is preferable, reducing the risk that a fracture might propagate vertically beyond the
packer depth. In the EWPMP the evidence from Hall plots, pressure transient analysis, production logs and performance
analysis of seven injection wells is that matrix injection conditions prevail.
Figure-2 shows the three year injection history of three EWPMP injection wells with low, medium and high injectivity
index. The average injection rate varies between 3000 bbl/day to 12000 bbl/day. For operational reasons the wellhead
injection pressure has varied from 1150 psi to 1650 psi. The Figure shows a clear linear relationship between wellhead
pressure and injection rate for the individual injection wells which is a characteristic of matrix injection well performance.
SPE 167274 3
Well 1 Well 2 Well 3
Figure-2: Injection pressure vs injection rate for three EWPMP injection wells
Hall plot analyses and injectivity indices

Hall plots show the summation of pressure times time against cumulative injection used for detecting changes in an injection
well’s flow capacity. Changes in the slope of the plot indicate a change in the flow capacity with a decrease in slope
indicating increased flow capacity due to stimulation or decreased skin/damage, additional perforations in a new zone, or a
change to fractured injection conditions. An increase in slope is an indication of plugging or other wellbore damage. The
relative change in slope is comparable to the change in flow capacity.
Figure 3 shows the Hall plot for three EWPMP wells, which exhibit a deflection towards the Y-axis indicating reduced
flow capacity through an increase of skin or wellbore damage. Surveillance engineers have been able to correlate these events
with excursions in the quality of injection water when the total suspended solids (TSS) and oil in water (OIW) exceeded the
recommended threshold limit of 5 mg/l (Rajan, 2013). Where the slope reduces slightly, for instance around 3200 Mbbl in
Well 2, this is a response to acidisation. As seen later at around 5000 Mbbl is improvement is often short-lived.
What is clear from the Figure 3 is there is no significant reduction in slope in any of the three wells. This can be taken as
further support for matrix injection conditions.
Figure-3: Hall plots for three wells in the EWPMP

4 SPE 167274
Analysis of injectivity indices through time indicates a continuous decrease in injectivity index (Figure-4) despite
attempts to maintain injectivity by repeated stimulation jobs. Interpretation of pressure fall-off studies confirms the presence
of damage in the near wellbore region evident from the high skin factor (>+5). Under fractured injection the injectivity index
would be much less sensitive to water quality, and the fracture would propagate rapidly with poor reservoir quality and more
slowly as water quality improves.
Figure-4: Injectivity indices in Well 2 and Well 3
Downhole injection profiles

Maintaining vertical conformance is one of the main steps towards ensuring effective waterflood performance. The injectivity
of the various reservoir layers is dynamic changing according to layer pressure, injection pressure and the condition of the
near wellbore zone. In the EWPMP pilot, conformance is thought largely to be governed by the quality of the injected water,
which is lower than that will be available for the full field project (see Rajan, 2013 for details of water quality studies).
Significant changes in inflow profile are clearly evident from ILT results as shown below (Figure-5).
In some cases, example; Well 3, injected water does not enter some of the perforated layers (W5), and in some layers
injectivity is continuously droping from over 60% to less than 30% of the total injection.
In Well 2, W3 is receiving more than 80 % of the injected water whereas other perforated zones receive very little water.
An acid stimulation job performed in Well 2 before the third ILT did help in improving the injectivity in W6&7 however, not
to the extent that was expected based on calculated layer permeabilities. The W3 zone continued to dominate, receiving most
of the injected water.
In Well 4, W3&4 layers received more than 80% of the injected water for several years until an effort to improve the
injectivity in W6 by re-perforation changed the scenario completely. Post re-perforation, the sixth ILT shows that 93% of
flow is diverted to the W6 layer leaving other layers starved of injection.
In these three examples we see the challenge of maintaining reasonable vertical conformance in commingled injection
wells. Left unchecked this could lead to reduced vertical sweep, premature water breakthrough and reduced recovery from
waterflood. Based on Hall plot, analysis of injectivity indices and ILT inflow profiles it is clear that formation damage
leading to ‘blocking’ of inflow to some layers is a significant problem. For the full field injection project higher levels of
water treatment are specified and expected to largely resolve this issue.
There remains a need to individually control injection and stimulation of individual layers, perhaps best illustrated by the
example of Well 4. Completion strategy largely depends on the methodology adopted for achieving conformance; whether at
injector well level or at producer well level. Since matrix injection is intended then conformance control at the injector was
decided as the best policy to optimise waterflood performance. For the full field water injection project this principle will be
tested in one third of the injection wells through installation of selective completions. The seven layers in Wara are to be split
into 3 groupings based on permeability, layer pressure and injectivity. Isolation will be achieved through hydraulic packers
set in the production casing and flow control devices will be used to apportion injection (and stimulation) fluids to individual
zones.
SPE 167274 5
Perforated
interval
Figure-5: Inflow profiles measured with a conventional Injection Logging Toolstring (ILT) superimposed on interpreted openhole log
data. Perforated intervals are marked in green. ILT are shown as coloured bars with percentage inflow per zone. Details of the ILT
acquisition are shown in the track header.
Introduction to the Simulation Study

The main objectives of the simulation study were to explore the relative benefits of different completion and workover
strategies for the Wara PMP wells.
1. Investigate the best strategy for comingling layers

2. Decide on waterflooding strategy; waterflood in low permeable or high permeable layers first or simultaneously to
achieve best Waterflood receovery.
3. Identify best strategy for shutting off water producing zones.
In a separate study of the EWPMP water injection pilot a 15km by 8km sector model was built with 50m x 50m cell sides and
1 ft. vertical resolution to capture the features (high permeability and sealing layers) already known to have influenced the
encroachment of edge water into the field. For this study a 25 hectare sector was cut from this model and used to build a 5-
spot element of symmetry simulation model (Figure-6) with a distance between injector and producers of 700m. The location
of the 25 hectare model was chosen to have cored well at one corner. In this way an accurate representation of the geology, at
least at this well, could be assured by cross check of log and core properties against model properties. The cell dimension of
approximately 17m x17m was selected and the 1ft vertical resolution of the geological model was retained for this study.
Despite such high grid resolution, in black-oil model the waterflood simulation run times were reasonable on a LINUX
cluster. The same model would be used to investigate EOR options for Wara, using cell dimension of approximately 6m x6m
and retaining the 1ft vertical resolution. In this paper we report the evaluation of completion and perforation strategy for
waterflood, the investigation of EOR is the subject of a future paper.
6 SPE 167274
Figure-6: The study used a sector model extracted from the Greater Burgan full field geological model.
The well spacing between injector and producer was around 700 m. Several layers were combined or ‘lumped’ together in
the injector and producer for the purposes of modeling selective completion intervals; the injector well with 5 and the
producer with 4 such intervals. It should be noted that even in such a simple model there is not an exact correspondence
between ‘layers’ at wells separated by 700m. This is a realistic feature of the Wara’s fluvial-tidal reservoir geology and thus
any recommendations drawn from modeling studies need to be sense-checked against the real world practicalities of
managing such a complex reservoir. The selection of completion intervals and the associated average permeabilities are
shown in Figure-7.
Figure-7: The study used a sector model extracted from the Greater Burgan full field geological model.
Based on an asset level analysis an upper limit of 98% watercut is considered. In the study seven completion cases were
considered as outlined in Table 1.
SPE 167274 7
Case Control of producer Control of injector completion

completion intervals intervals
1. Base Case All completions open throughout the run. All completions open throughout the run.
Well is produced to 98% wc
2. CE1 All completions open at t=0 All completions open throughout the run.
“Completion economic limit” Completion closed when the completion
watercut reaches 98%
3. WE1 All completions open at t=0 All completions open throughout the run.
“Well economic limit” The worst offending completion is closed
when well level watercut reaches 98%
4. LEINK-13 All completions open at t=0 At t=0 injection only into high permeability
“Selective injection – high K” The worst offending completion is closed completion zones; 2, 3 and 5.
when well level watercut reaches 98% Once the corresponding completions water
out at the producer inject into zones 1 & 4.
5. LEINK-15 All completions open at t=0 At t=0 injection only into low permeability
“Selective injection – low K” The worst offending completion is closed completion zones
when well level watercut reaches 98% Once the corresponding completions water
out at the producer inject into zones 2,3 & 5.
6. LEINK-13-PA At t=0 produce from high permeability At t=0 inject into high permeability completion
“Selective injection & production – completion zones only; 2, 3 &4. zones only; 2, 3 & 4.
high K” Low permeability zones added later Low permeability zones added later
7. LEINK-14-PA-S At t=0 produce from low permeability At t=0 inject into low permeability completion
“Selective injection & production – completion zones only zones only
low K” High permeability zones added later High permeability zones added later
Table-1: Scenarios evaluated in sector model simulation.
Discussion of Simulation Results

This discussion is made with reference to the cases described in Table 1 and Figures 8 – 14. Additionally, Figure 15 shows
cumulative water productions for all the cases. Where additional recovery is mentioned it must be noted that this is a
percentage increase or decrease in cumulative oil production, not to be confused with a change in recovery factor.
In the first three cases injection and production are across the entire reservoir interval. Compared to the base case the most
significant gain (5% additional recovery) comes from closing the worst offending completion only when the well level
watercut reaches the assumed economic limit of 98% (case 3, “well economic limit”). This gain is made by effectively
extending the life of the well. Whether this can be fully achieved in practice will depend on very careful well testing and
production logging, however, the principle is clear that maximum recovery is achieved by allowing each zone to produce for
as long as economically possible. Referring to Figure 15 it should be noted that this comes at a cost of significant additional
water production, compared to case 2 there is an additional 6% oil recovery but 100% additional water production.
In the next four cases varying degrees of selective completion are employed. In this discussion we will compare the
results of these runs with case 3.
Cases 4 and 5 retain fully completed intervals at the production well but selective completion at the injector, initially
targeting high and low permeability zones respectively. Cumulative oil is only slightly less (-1% compared to case 3) for both
cases. Life cycle water production is slightly higher (+6% to 8%). However consideration of the oil production profile
indicates a significant deferral of oil production for both case 4 and 5 when compared against case 3. Although economic
analysis is not presented in this paper it is obvious there is significant loss in project value. Coupled with slightly lower oil
production and higher water production neither case 4 or 5 are attractive completion options.
Cases 6 and 7 employ selective completion at both producer and injector, initially targeting high and low permeability
zones respectively. Cumulative oil is much lower compared to case 3; a 10% reduction for case 6 and 6% for case 7.
However, life-cycle water production is much reduced by 40-50%. Again there is also a significant deferral of oil production
and erosion of value. In both these cases the higher and lower permeability zones are effectively flooded sequentially. It
appears better do flood these zones concurrently taking advantage of stream lines crossing between higher and lower
permeability systems.
In summary, to ensure maximum recovery it must be ensured that water is injected across the full reservoir section.
Facilities must therefore be designed to handle eventual high levels of water production. Based on field experience from time
to time, and in line with prevailing field constraints, it may be necessary to close zones in production wells that have
moderate to high watercut. However simulation suggests these closures should be only temporary and completion zones
should be produced until the econimc limit is reached.
8 SPE 167274
Conclusions
The conclusions of the studies into injection well performance and completion strategy are as follows:
• In the EWPMP pilot there are indications of poor conformance at injection wells. This is related to the inherent
reservoir heterogeneity but also due to completion practices and injection water quality.
• Simulation modeling highlights the need for water to be injected across all zones. Injecting in one zone at a time is
not practical (reduced injection and production rates, much longer time to flood, longer time to re-pressure the field)
nor is it expected to provide any benefit in terms of improved recovey.
• In future water injection wells selective completions will provide a means of control of both injection and
stimulation, where this is required to correct or restore the inflow profile.
• The water quality for the full field water injection project is of a higher specification than the EWPMP pilot and it is
anticipated this will result in less problems related to injectivity decline at both well and completion level. Careful
reservoir monitoring will be required to verify this.
• At the production wells the intent is to produce from all oil bearing zones. Simulation indicates this results in
maximum recovery. Shut-offs should be done only at high watercut or reserves will be lost. In practice this may
require plugs, cement squeeze and straddle devices, however, in an onshore environment all these options are
practical.
• Future studies will consider additional models representative of other areas of the field. Evaluation of new
technology will include options such as mobility control (polymer) and deep conformance control chemical to
maximize sweep efficiency.
References
Adbul Rahman, B. et al, 2012, An Integrated Solution to Effective Waterflood Surveillance and Pressure Maintenance: A KOC Pilot
Project of Greater Burgan Field, Kuwait, SPE154009
Ambastha, A. et al, 2009, Observations from a Fieldwide Pressure Data Acquisition Campaign in the Wara Formation of the Greater
Burgan Field, Kuwait
Al-Naqi, M. et al, 2009, Lessons Learned From the First Water Flood Pilot Project in a Clastic Reservoir in the Greater Burgan Field in
Kuwait, SPE120427
Rajan, S. et. al. 2012, Implementing Water Flood in the Greater Burgan Field, Kuwait: Improved Development Planning Through Pilot
Testing and Field Appraisal Activities, SPE164216
SPE 167274 9
Figure-8: The base case simulation (CS1)
Figure-9: Producer completions closed at the economic limit (CE1)
Figure-10: Worst offending producer completions closed at the well economic limit (WE1)
10 SPE 167274
Figure-11: At t=0 injection only into high permeability completion zones (LEINK-13)
Figure-12: At t=0 injection only into low permeability completion zones (LEINK-15)
Figure-13: At t=0 inject & produce from high permeability completion zones only (LEINK-13-PA)
11
Figure-14: At t=0 inject & produce from low permeability completion zones only (LEINK-14-PA-S)
Figure-15: Cumulative water production for seven simulation cases

SPE 167274

Spe 167274 Ms

Uploaded by

Copyright:

Available Formats

Spe 167274 Ms

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Spe 167274 Ms

Uploaded by

Copyright:

Available Formats

SPE 167274

Copyright 2013, Society of Petroleum Engineers

Injection Performance in the Early Wara Pressure Maintenance Project (EWPMP)

Well 1 Well 2 Well 3

Hall plot analyses and injectivity indices

Well 1 Well 2 Well 3

Figure-3: Hall plots for three wells in the EWPMP

Figure-4: Injectivity indices in Well 2 and Well 3

Downhole injection profiles

Well 3 Well 2 Well 4

Introduction to the Simulation Study

1. Investigate the best strategy for comingling layers

Case Control of producer Control of injector completion

Table-1: Scenarios evaluated in sector model simulation.

Discussion of Simulation Results

Figure-9: Producer completions closed at the economic limit (CE1)

Figure-15: Cumulative water production for seven simulation cases

You might also like