Você está na página 1de 15

SPE 89745

Water Injection Completion Philosophy in A Deepwater Subsea Environment Requiring


Sand Control: A Case Study of 29 Injection Wells West Of Shetland.
Alistair Roy, David Thrasher, Allan Twynam, Allan Wilson SPE, BP;

Copyright 2004, Society of Petroleum Engineers Inc.


Introduction
This paper was prepared for presentation at the SPE Annual Technical Conference and
Exhibition held in Houston, Texas, U.S.A., 26–29 September 2004.
Water Injection (WI) completions in a sand control
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as
environment have recently gained a higher profile within the
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to West of Shetlands, in particular, and the industry in general.
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at No clear concensus would appear to have been reached as to
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
what the optimum completion design should be and a variety
for commercial purposes without the written consent of the Society of Petroleum Engineers is of completions have been installed throughout the industry.
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous Areas causing particular concern are well longevity and near
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
wellbore injection conformance, both of which are directly
linked to producer performance and maximizing reserves
recovery. This study presents the range of injection
Abstract completions that have been installed in the 29 injection wells
drilled to date in the WoS fields. Areas that will be examined
Water Injection completion selection in a deepwater are completion design parameters and how they are derived,
subsea environment requiring sand control is currently an completion selection, impact of drilling fluid conditioning,
inexact science with lack of a proven track record for life of formation damage, well clean-up and completion performance
field solutions. The West of Shetlands (WoS) fields Foinaven, in terms of completion efficiency, injection conformance and
Schiehallion and Loyal have been on injection since 1997 and life of well.
are starting to develop a substantial track record of injection
performance. This case study will provide a performance Background to WoS Fields and Area
comparison of the 2 main completion types installed to date,
namely Cased and Perforated (C&P) and standalone Wire The Foinaven, Schiehallion and Loyal fields are located in
Wrap Screens (WWS), along with a subset of a three zone blocks 204/19 and 204/24a (Foinaven) and 204/20, 204/25a,
Smart well completed with premium sand screens. The 204/25b, 205/16 and 205/21b (Schiehallion and Loyal) which
comparison focuses on 4 main areas; lie approximately 190kms West of Shetlands, in water depths
- Well Longevity ranging from 350 – 510m.
- Completion Efficiency (skin)
- Injection Conformance (Near Wellbore) Figure 1: Location Map
- Performance of single and multi-zone injectors

The WWS screen performance is further subdivided by


clean-up technique of which there are currently 5;
- Backflow Oil Based Mud (OBM)
- Bullhead Oil Based Mud
- Bullhead Water Based Mud after spotting Breaker
- Toe Fracture using External Casing Packer
- Displace OBM to Solids Free OBM and bullhead

A variety of data will be presented as part of the


comparison including, PLT Data, MDT Data, PFO Results,
Hall Plots, UCS Data. A simple set of criteria is presented to
select when C&P or sand screens are viable completion
solutions.
2 SPE 89745

The developments are based upon 2 Floating Production All WoS reservoirs are prone to sand production and sand
Storage and Offtake (FPSO) vessels receiving oil from clusters control is required in both producer and injector wells.
of subsea wells via rigid flowlines and flexible risers. Shuttle Average rock and fluid properties across these unconsolidated
tankers are required to offload oil from Foinaven to the Flotta sandstone reservoirs are as follows;
terminal and Sullom Voe for Schiehallion. Both fields inject Reservoir Depth : 1,700 – 2,200m TVDss
seawater, and either re-inject produced gas or export it to Reservoir Pressure : 2,800 – 3,500psi
Magnus for use in an Enhanced Oil Recovery (EOR) scheme. Reservoir Temperature : 140 – 160degF
Schiehallion has also had periods of produced water Bubble Point pressure : 2,500 – 3,100psi
reinjection to reduce the environmental impact from produced Gas Oil Ratio : 350 – 530scf/bbl
water discharge to sea. Oil viscosity : 1.05 – 3.5cp
Permeability : 100 – 2000md
The current Foinaven subsea layout comprises two drilling Porosity : 22 – 27%
centres, DC1 and DC2, each based upon a manifold and well Rock UCS (core) : 100 – 800psi (Loyal up to 2000psi
cluster arrangement (Fig. 2). A remote water injection well depending on depth)
204/24a-8 (W22) is also located to the east of the field. The
East Foinaven satellite field was tied back in 3Q 2001. East The WoS fields are completed across a number of layers
Foinaven water injection is provided by flexible flowline from ranging from the T35 (stratigraphically highest) to the T25
the remote W22 site. (stratigraphically oldest). Figure 3 shows a N-S cross-section
across the Foinaven Field and the relative positions of these
Figure 2: Foinaven Subsea Architecture units. The variation in rock properties is a reflection of
completing wells in multiple horizons.

Figure 3: Cross-Section Across Foinaven Field

South DC1 DC2 North

P21

P11 P23, P29 P22,P25,P27,P24 P28a


P14 P210 P26
P12 P16
P18
P19
P13 P15
P17

P110

Panel 4 Panel 2 Panel 1 Panel 0 “ Wing"


W15 Panel 3

Reservoir
T35 0
Horizontal
T34 Vertical
The current Schiehallion subsea layout is similar to T32
0 Scale 1Km
Scale 50

100m

Foinaven but more extensive consisting of 4 Drill Centres - T31


T25

Schiehallion Central, Schiehallion West, Schiehallion North Foinaven Subsurface Team, Feb. 2003

and Loyal.
Current Well Stock and Role of Injection
The distribution of wells between the Drill Centres across
the West of Shetlands fields and split by completion type is For simplicity Loyal has been grouped with Schiehallion
shown in Table 1 below. as both fields are tied back to the same FPSO. There are a total
of 73 production and injection wells installed WoS comprising
Table 1: WoS Wells by Drill Centre & Completion Type 42 production wells, 29 water injection wells and 2 gas
Production Injection Completion Types disposal wells. Table 2 shows the distribution of wells
Wells 12 Gauge WWS 10 Gauge WWS Premium Screens C&P between Foinaven and Schiehallion, the well name shown is
Foinaven the slot number, for either the first or second letter (depending
Drill Centre 1 10 W11, W14 N/A W15 W12, W13
W22, W24, on which field) a P denotes a production well and a W denotes
Drill Centre 2 11 N/A N/A N/A W25 a water injection well.
East Foinaven 1 W41 N/A N/A N/A

Schiehallion There is a marked contrast in the approach to injection


CW10, CW11, between Foinaven and Schiehallion, the producer to injector
Central CW12, CW13,
13 CW16, CW15 CW19 N/A CW17, CW18 ratio in Foinaven is 3:1; in Schiehallion it is 1:1. To clarify the
WW04, WW05, nomenclature, well names beginning with an L denotes one of
West 4 N/A WW06, WW08 N/A WW09 the six Loyal wells, Loyal also has a 1:1 producer/injector
NW01, NW02,
North N/A N/A NW03 N/A N/A relationship. The main reason for the different water injection
LW04, LW05, strategies is that the Schiehallion faults cross cut the channels
Loyal 3 N/A N N/A LW06
SPE 89745 3

making it more compartmentalised than Foinaven where the W13 - A production log indicated fill in the well up to a
faulting is parallel to the channel axis. Foinaven are currently depth that corresponds to where the initial CT perforating guns
investigating the value of reducing their dependance on this were stuck for 12hours prior to being freed. One possibly
limited injector well stock in light of two failed WI wells. hypothesis is that the liner was damaged when perforated. The
liner is 25% Chrome therefore it is almost certainly not a
Table 2: West of Shetlands Well Stock (Sept 2003) corrosion issue

Key Risks to Injectors Requiring Sand Control

In the West of Shetlands sand prone environment a number


of risks related to sand control have been identified which are
incumbent to the current injection well designs.

Cased and Perforated;


- Sand control effectiveness is directly related to volume
of sump available below perforations to collect solids
prior to well performance being affected
- Perforating multi-zone wells leads to cross-flow within
the wellbore and potential sand ingress accelerating
filling of available sump
- Perforating what are perceived to be single zone wells
may lead to cross-flow due to variations not obvious on
Table 2 also highlights the wells in which production or the RFT data leading to significant cross-flow down-
injection capability has been lost (shown in red) and it can be hole. If a well shows an undisturbed pressure gradient
seen that to date 4 wells have lost functionality; prior to completion then it may vary significantly with
time from the initial assessment of single zone.
P27 – leaking SCSSV control line, loss of PI due to
undiagnosed sandface completion issues, high levels of sand Screen Injection Wells
production. - Sand control effectiveness is directly related to the
W11 – Sand fill encountered 70m above top screen, screen effectiveness of the screen as a filtration medium and the
failure obvious from coarse PSD of sand from CT clean-out. speed of build-up of formation sand in the annulus
Some conditioned mud was found in returns however it is not leading to depth filtration and improved sand control.
clear if this indicates that bullheading displacement was not Water injection can stiffen the formation and collapse of
100% efficient or if the mud was merely recovered from the the formation onto the screen will not always occur.
blank pipe below the screens. - Screen wells are typically completed with very little
W14 – Sand fill encountered 60m above top screen, fine sand sump volume to avoid exposing unwanted formations.
recovered indicating screens functioned correctly but large - Completing multi-zone wells leads to annular cross-
quantities of fines were still passing the filtration medium. flow (unless isolated with ECPs) potentially mobilizing
WP01 – combination of high water cut, sand production and and eroding any intra-channel shales, leading to fines
emulsions make the well currently unproducable to facilities. production which the screens are not designed for.
Demulsifier trials may change this situation.
Common risks to both completion types
The 2 Foinaven injector Completions that have failed were - Particle Size Distribution (PSD) and Unconfined
both screen completions, which probably have poor design Compressive Strength (UCS) data is inferred from the
features incorporated. The three Foinaven injection wells with log and core data available from the offset, mainly
the longest performance to date are 3 C&P wells, W22, W24 appraisal, wells.
and W25. In Schiehallion the injection wells with the longest - “Water hammer” on well shut-in leads to sand ingress
injection history are 2 screen completions. To date all cased - 2-3minute shut-in times for tree valves may lead to
and perforate completions are on-line and functional. significant cross-flow in the subsea system, which may
cause sand production down-hole. Figure 4 shows the
Well Interventions have encountered problems in 2 of the sequence of valve closure for the Foinaven injection
Foinaven C&P injectors; wells and the valve closure time.
W12 – Well can inject via ported debris sub however - Clean-up impact on injection performance /
reservoir access is not possible with any toolstrings. The TCP conformance / fracture location
guns have not dropped due to a component being run above - Erosion of lower completion components is dependant
the muleshoe which has a larger OD than the muleshoe ID, on metallurgy and fracturing geometry
hence while the guns released they could not drop below the - Corrosion of lower completion, adequate corrosion
Wireline Entry Guide (WEG). resistance depends on utilizing the correct metallurgy and
maintaing oxygen control below the target of 10 parts per
4 SPE 89745

billion (ppb) which has not always been successful, Table 3: Selected Schiehallion PSDs
excursions above 200ppb have been recorded. Core Depth
Well (m) D10 D40 D50 D90 D95 D10/95 D40/D90 Formation
Well 204/20-2 2007.5 670.7 299.7 243.5 32.1 11.2 59.9 9.3 T31
Figure 4: Foinaven Valve Closure Sequence and Times Well
Well
204/20-2
204/20a-7
2015.4
2141.5
701.1
460.1
347.7
228.8
291.3
186.0
39.8
16.7
14.2
5.5
49.4
83.7
8.7
13.7
T31
T34
Well 204/20a-7 2143.6 553.1 271.8 221.9 32.7 9.6 57.6 8.3 T34
Well 204/20a-7 2190.1 579.3 339.7 291.4 17.3 6.3 92.0 19.6 T34
Injector Tree Valve Closure Sequence Well 204/20a-7 2192.5 635.1 333.7 281.1 14.9 5.6 113.4 22.4 T34
Well 204/20a-7 Corex 465.5 265.4 228.5 59.4 13.3 35.0 4.7 T34
Close Time Well 204/20-W01 3353.3 482.9 248.4 206.3 17.2 5.0 96.6 14.4 T31
Well 205/16-2 1917.1 473.4 270.4 233.2 72.1 37.1 12.8 3.8 T31
Well 205/16-2 1919.9 433.4 231.9 194.7 38.4 22.4 19.3 6.0 T31
W14 Higher Pressure Well 204/20-N01Z 4632.9 414.2 254.3 219.8 52.6 16.1 25.7 4.8 T34
W13 Well 204/20-N01Z 4697.3 704.2 399.3 338.8 56.8 17.9 39.3 7.0 T34
W11 Well 204/20-N01Z 4643.4 460.0 300.0 270.0 50.0 10.0 40.0 6.0 T34
Well 204/20-N01Z 4646.1 290.0 162.5 137.5 17.0 5.0 58.0 9.6 T34
W41 Well 204/20-N01Z 4646.4 375.0 225.0 175.0 30.0 10.0 75.0 7.5 T34
W15 Well 204/20-N01Z 4646.9 350.0 187.5 150.0 6.0 2.0 175.0 31.3 T34
Well 204/20-N01Z 4647.0 395.0 220.0 205.0 40.0 10.0 39.5 5.5 T34
W22
Well 204/20-N01Z 4652.3 360.0 187.5 150.0 6.0 3.0 120.0 2.9 T34
W24 Well 204/20-N01Z 4701.7 420.0 225.0 205.0 30.0 7.0 60.0 7.5 T34
Well 204/20-WW06 3340.1 549.1 303.5 259.8 69.4 21.9 25.1 4.4 T31
W12
Well 204/20-WW06 3352.6 542.4 266.3 218.3 19.1 6.7 81.0 13.9 T31
W25
Lower Pressure Well 204/20-LW04 2549.2 595.9 345.1 296.4 85.9 40.0 14.9 4.0 T35

0 30 60 90 120 150 180 210


Time after trip (seconds) Figure 6 shows a similar PSD data set across the Foinaven
field, however on this occasion the data presented is derived
from sieve analysis, apart from the T25 data which has been
Source and Summary of Available PSD and Rock derived from LPSA. Again there is significant variation in the
Mechanics Data D50, ranging from 180 – 340microns (excluding the T25 layer
which is a deeper sand with only a single production well and
Whole Core data is generally available from appraisal wells, one injector completed to date).
only 5 development wells across the entire WoS have been
cored. PSD data for screen design is therefore limited to the Figure 6: Selected Foinaven PSDs
appraisal wells and this has to be extrapolated to the relevant
injection well. This lack of directly applicable data may have a 100

significant influence on the W14 failure. There is a significant Perforation Strategy and Effectiveness
90

variation in PSD across the fields and Reservoir Sands (T25, 80

T31, T34 etc). Figure 5 shows a sample of the PSD data Core data is
70
available for Schiehallion across different wells and reservoir
zones. 60
Cumulative %

50 T31 204/24a-5 T31 204/24a-7


T32 204/24a-2 T32 204/24a-3
Figure 5: Selected Schiehallion PSDs 40 T32 204/24a-5 T34l 204/19-3A
T34l 204/24a-3 T34l 204/24a-7
T35l 204/24a-6 T35l 204/24a-6
100 30
T35l 204/24a-6 T36 204/19-3A
T25 204/24a-2
90 20

80
10

70
Cumulative Percentage

0
60 1.00 10.00 100.00 1000.00
204/20-2; 2007.46m Particle Diameter (micron)
204/20-2; 2015.37m
50 204/20a-7; 2141.46m
204/20a-7; 2143.59m
204/20a-7; 2190.05m
40 204/20a-7; 2192.54m

30
204/20-W01; 3353.33m
205/16-2; 1917.06
205/16-2; 1919.92m
The difference in the fines tail described by the LPSA is
20
204/20-N01Z; 4632.88m
204/20-N01Z; 4697.29m
204/20-W03; 3340.10m
striking when compared to the sieve data – the sieve data
10
204/20-W02; 3352.63m
204/20-L02A; 2549.22m misses the fines detail. A comparison of the PSD results from
have been available the East Foinaven T34 sand is shown in Figure 7. This has a
0
10000 1000 100
Particle Size (microns)
10 1 significant impact on the estimates of D40/D90 and D10/D95,
which are commonly used for sand control design purposes as
per table 4.
Table 3 shows the same data numerically. The data
presented for Schiehallion is derived from Laser Particle Size
Analysis (LPSA). Figure 4 and Table 2 combine to highlight
the variations both vertically and areally across the WoS fields

The D50 ranges from 150 – 340 microns, which shows the
importance of identifying the correct analogue sand and well
for sizing screens, although significant variation is observed
over very short depth intervals in the same interval e.g. the
N01z results. Alternatively the completion design must be
able to cope with the range of PSDs given the uncertainty in
vertical and areal distribution of sand types.
SPE 89745 5

Figure 7: Comparison of Laser vs Sieve PSD Analysis which it is available. Note these are log derived values rather
than core (see below for correlation)
East Foinaven Composite PSD Plot

100
Figure 8: UCS Distribution across Foinaven T31 Layer
90

Gas Cap
80
Pane Water
T34L
l0
& T32 204/19-3A
Injector
70 B02Z Producer
1022 P 21 DC2
B0
1 E&A well
60
Cumulative %

B08 W25 B09Z Planned

owc
3 P26
B0 2
well
205/25b-5 - 2158.17m Sieve W
50
205/25b-5 - 2203.95m Sieve P2 P2
4
B10Z T3 ing
Pa 7 P28a
1L
n

B04 W24
205/25b-5 - 2158.17m Laser T3 e l 1
4L B0
40 1158

W 0
13
Pan

A1
204/24a-7
Z B05
205/25b-5 - 2203.95m Laser 7
Depth (m): 2131.6 2203.9 204/24a-6 el 2
T34 A04 11 P23 A12 P2
30 972 P 2 1Z
W1
L
B0

5
822

B11

P2
P29
Seive D40/D90 3 4

A06Z
P16
204/24a-3

6Y
20 owc 4
P A0
an

B0
T 34
U, T el 3 A09
Laser D40/D90 9.8 13.8 DC1 204/24-1A
2
W2
32
&T W14 o wc
10 31
14
204/24a-4

8P

41

Ea T34
Pan A0
204/24a-2

577

W
P1 5
0 e A0

st L
l4

A0
P1 1

41
A2
T35
12

Fo
1 10 100 1000 , T3 204/24a-2Z
429 5

zP
2 P
1z

ina
W1 7
1
ow
1210

A0
Particle Diameter (micron) A1

A1
c 204/24a-2Y

ve
owc

n
204/25b-5
204/24a-5

496

A0
P1

A0 7
Table 4 is a commonly used indicator for deciding which

2
P1
3

3
sand control technique may be appropriate based on PSD 0 1 Kilometer
A G Carruth, Aug. 2001

derived from Sieve analysis. Commonly observed figures In Schiehallion a number of the appraisal wells have both
West of Shetlands for D40/D90 from Sieve analysis will be in core and sonic log data available allowing a correlation to be
the range of 2.5 to 3.5 and D10/D95 in the range of 10 – 15 developed for estimating UCS values from the sonic P and S
placing us on the borders of where a Premium Screen would waves. This correlation can then be applied to those
theoretically be applied in a Production Well. (Ref 1. Bennett development wells where no core has been taken but a sonic
et al SPE 65140). log has been recorded
Table 4: Design Criteria for Sand control selection Figure 9 below shows the ratio of log-derived strength to
Sorting Uniformity Fines Content Sand Control Medium core derived UCS, which appears to increase in the weaker
Coefficient Coefficient (<44 microns) rocks. That is, the logs over-estimate the UCS strength of the
d10/d95 d40/d90 weaker rocks. The ratio (Log-Derived Sonic) LDS/UCS is
< 10 <3 < 2 wt% Wire Wrapped Screens plotted against UCS below. It shows a significant divergence
< 10 <5 < 5 wt% Premium Screens
below about 300psi where the shift from log to core increases
> 10 >5 > 5 wt% Gravel Pack or ESS
quickly over a factor of 2.
Corresponding numbers for LPSA analysis from exactly
Figure 9: UCS–Log Derived Strength (P+S) Correlation
the same core sample may have a D40/D90 in the range of 9 –
14 placing us firmly in the “Hole Stabilisation” region (e.g. 6305 WoS UCS - Log Derived Strength (P and S) Correlation (ALL DATA)
Gravel Packs or ESS). While the LPSA is useful as an 20
indicator of the fines content, the preliminary sand control
18
selection in the West of Shetlands fields is currently based on
P+S Log Derived Strength (psi)/UCS (psi)

16
PSD derived from Sieve analysis, with the actual sand control
design being based on sand retention testing. 14

12

Source / Summary of Available Rock Mechanics Data 10

8
y = 336.98x-0.8857
As for the PSD, direct measurement of UCS values is 6 R2 = 0.9096

limited to the appraisal wells where whole core is available. 4

The value of the UCS data will be shown in the following 2

sections on the decision to run Screens or Perforate. The UCS 0


(from core) values in the fields range from 80 – 800psi, with 0 200 400 600 800 1000 1200 1400 1600
Unconfined Compressive Strength (psi)
some of the deeper parts of Loyal having UCS values up to
2,000psi. Most of the core UCS values lie in the 200 – 400psi
range, where resolution of UCS calculated from logs begins to The log-derived UCS is not a true UCS, it should be
be less robust. considered a pseudo-strength index. This index has been
correlated to a laboratory derived strength measurement for a
Both fields show an increasing trend in UCS from the given formation. In this way, the log-derived strength has
South west of the field to the North East, which corresponds to become a predictive tool. The equations below show the
the dip direction. This is consistent for all horizons. A map of process for estimating the log derived UCS, which has been
the UCS data in the Foinaven T31 is shown in Figure 8 with based on a BP revision of an existing correlation (Ref 2).
the Mean UCS value in the T31 indicated for each well for
6 SPE 89745

8.7aE 2 .10 3 Figure 10: Injector Philosophy for Screens vs C&P


UCS = {0.008VCLAY+0.0045(1- VCLAY)} (1)
(1 − 2ν )
Injectors - Philosophy for Screens vs Case and Perforate

γ − γ MIN
Clay Volume Fraction, VCLAY = (2)
γ MAX − γ MIN

γ = Gamma Ray Value (API units)


2 cos φ
Cohesion (psi) a =
1 − sin φ

Poisson’s Ratio, ν = 0.1 + (VCLAY/4) (3)


1/ 2
Shear Wave Transit Time (µs/ft) DTSH = DT ⎡ 2(1 − v) ⎤ (4)
⎢ 1 − 2v ⎥
⎣ ⎦

Dynamic Elastic Modulus, E (Mpsi) = 26900ρ


(1 +ν ) (5)
(DTSH )2
The decision tree indicates that below a core derived UCS
Friction Angle, Ф (degrees) = 20 + 10(1 – VCLAY) (6) value of 250psi then a well is not suitable for perforation and
sand screens are required. If 80% of the UCS values are above
Decision to Run Screens or Perforate 250psi a single zone may be successfully perforated. If 80% of
the UCS values are above 600psi multiple zones may be
The logic for deciding whether to run sand screens or to successfully perforated. The UCS values referred to are those
perforate the water injection wells is difficult to turn into a derived from core or corrected log UCS not raw log UCS
hard and fast decision making process; it is instead based on values. These UCS cut-offs are based on past experience from
the following information; the Forties field, not any estimate of drawdown during cross-
- Number of zones to be completed flow, back-flow or water hammer.
- UCS data available from offset wells in the same layer(s)
- Required angle through reservoir The decision to perforate must be made prior to the well
- Reserves (value of well) being drilled as it affects the well design i.e. 12-1/4” hole is
- Required Life of Well drilled to TD 250m below the reservoir rather than 10m into
- Current Injection Performance of each completion type in the reservoir in the screen completions where the reservoir
relevant portions of the reservoir/zone section is drilled in 8-1/2” hole. A clear subsurface statement
regarding the validity of the reservoir being a single zone is
If the required well trajectory through the reservoir is required before deciding to perforate, this by necessity must be
above 60degrees then the well must be completed with sand based on offset well and production data prior to drilling.
screens - C&P is not appropriate in this case as any sand
ingress into the well will not fall downhole, and therefore An example of where C&P was applied was Schiehallion
bridges will form leading to loss of injection. CW17. Figure 11 indicates that the log derived UCS values
from the best analogue well are in the region of 400 – 500psi.
If the required well trajectory is less than 60degrees then Applying Equation (1) and the correlation from Figure 9
there is an opportunity to either C&P or run screens - it does would suggest the core values would be in the region of 210 –
not automatically mean C&P should be selected. The decision 350psi, with the bulk of the log indicating a UCS > 250psi.
tree in Figure 10 combines the well trajectory and the UCS
data to provide a screening tool to decide if C&P is a viable
option. Again, the ability to C&P does not automatically mean
it should be the selected completion.
SPE 89745 7

Figure 11: Log derived UCS for Schiehallion C09 General Water Injection Completion Objectives
(Analogue for CW17)
In the generic field Basis of Design (BoD) for water
injectors and individual well BoD, some non-well specific
NET_1
20 METRES 0
0.2
AHO90
OHMM 200 1
BVW
V/V 0
completion objectives are required to be met by the proposed
100
0
UCS_CORE_1
METERS 2000
0.2
AHO60
OHMM 200
ZONESmddbrt 1
PHI
V/V 0
well design;
40 0
UCS_PS_NEW_1
METERS 2000 0.2
AHO10
OHMM 200 0
VSH
V/V 1
- 20 year life of well
- No interventions planned through life of field
- Water injection rate between 20,000 – 50,000bwpd
- Mechanical skin < 5

3872.0
- 25%Cr or higher spec metallurgy for sandface
completion below Formation Isolation Valve (FIV) for
screen wells.
- L80 1%Cr metallurgy for 9-5/8” casing in perforated
T31R1
29.0

wells

Perforation Completion Design


3901.0

The main driver for C&P completions in this deepwater


subsea environment West of Shetlands is to reduce rig time
and costs by drilling and completing in a 12-1/4” hole and
omitting the 8.5” hole section which is required to install
screens. Obviously a major goal is that life of well and
injectivity achieved are not compromised by selecting a cased
and perforated completion in a reservoir environment with
The MDT data from CW17 (labelled as C19) is shown risks from relatively unconsolidated sands. To date, given the
below in Figure 12 and indicates that the single T31a sand data presented in Table 2, C&P outperforms screen only
prognosed prior to drilling is actually 3 distinct T31a sands at completions in terms of life of well. However, this view may
different pressures, with the bottom sand also being water wet change as further performance data becomes available with
as well as 55psi higher than the middle sand body. The time.
maximum differential pressure across the three sand bodies is
at least 190psi. A schematic of a typical Tubing Conveyed Perforation
(TCP) completion around the sandface area is shown in Figure
Figure 12: RFT Data for Schiehallion Well CW17(C19) 13. The C&P completions are very simple with only the
following components;
1940
Pressure data from Schiehallion Wells – CW17
20-1. 20-1z 20-2. C16

1960 7” or 5-1/2” TRSCSSV


7” or 5-1/2” Fibreglass Lined Tubing (Duoline)
20-1 Wet 20-1z Wet C10 C19

204/20-C10

Hydrostatic Set Production Packer


1980 Over pressured
Depleted by 540psi
by 265 psia
204/20-C16
2000 Nipple for depth control for cutting tailpipe with CT or E-
2020 Depleted by 650psi
line (contingency against gun failure)
204/20-C19 Perforation Assembly including Pressure Cycle Activated
2040
Gun Release System and Firing Head
2060 Swept Zone
Depleted by 510psi
204/20-C19
OWC 2064m tvdss Drilling the reservoir in 12-1/4” hole essentially commits
2080
the completion to C&P unless the casing was set high and
2100 the reservoir section was re-drilled in a smaller hole size.
2000 2100 2200 2300 2400 2500 2600 2700
Pressure psia
2800 2900 3000 3100 3200 3300
Another possibility would be to cement the casing off-
bottom. To date neither of these options has been
This potential multiple sand T31a scenario is not suitable implemented WoS.
for C&P where the UCS is at the lower end of that
recommended for perforating a single zone. As a
compromise, the water-wet sand was not perforated, for
reservoir reasons as well as the sand control issues.
8 SPE 89745

Figure 13: TCP Sandface Completion Schematic changes unless a vessel is in the area with an ROV capable of
performing this operation.

Table 5: Perforated Injector Performance


Sand Perforation
Production Packer Over / Under Mechanical Thickness Gun Size and
Well Field Layer balance (Total) Skin (m) Method
Landing Nipple W12 Foinaven T34 117psi 3.8 (9) 47m 4-1/2" TCP
W13 Foinaven T34 205psi 7 47m 3-3/8" Coil
W22 Foinaven T34/T32 206psi (7) 30m 3-3/8" Coil
Ported Debris sub
W24 Foinaven T34 -200psi (10) 61m 3-3/8" Coil
W25 Foinaven T34/T32 209psi (4.2) 89m 4-1/2" TCP
Wireline Entry guide LO2a Loyal T35 121psi -0.8 (0.0) 30.5m 2-7/8" E-line
LO5 Loyal T35 178psi -0.25 (0.1) 30.5m 4-1/2" TCP
LO6 Loyal T35 -161psi 0.2 (1.4) 30.3m 4-1/2" TCP
Logging clearance between WWO9 Schiehallion T31 347psi -0.2 (+3.8) 17m 4-1/2" TCP
WEG and top shot CW17 Schiehallion T31 550psi 16 (21) 44.6m 4-1/2" TCP

Perforation Interval Given the low post fracture skins achieved to date and the
lack of completion constraints on WI targets there is seen to be
Logging Clearance below little benefit in attempting a high degree of initial
Bottom Shot underbalance (a dynamic underbalance being achieved via
TCP anyway) perforating in these fractured WI wells;
TCP Guns dropped below - Perforating underbalance may lead to sand influx
Perforated Interval - Costly process in terms of rig time
- HSE implications of performing unnecessary operations

Screen Completion Design


As shown by Table 5 several types of perforation have been
performed from underbalance with guns run on coiled tubing As previously described, in multi-zone cases where the
(CT) to the current practice of TCP guns run on the end of the UCS is insufficient to support a C&P completion, or where
upper completion tubing. Table 5 shows the performance of multiple targets are required which cannot be accessed by a
the perforated injection wells in terms of post fracture skin. low angle (<60 degrees) trajectory through the reservoir, then
The skin recorded as mechanical skin is the global skin after a screen completion design must be implemented.
backing out the frictional pressure drop calculated from
Prosper from the THP gauge to the top shot (or top screen in A typical Schiehallion screen injector completion is shown
other well types). This is required as the Pressure Fall-Off in Figure 14. The main completion components in the screen
(PFO) is recorded from the Tubing Head Pressure (THP) injector are as follows:
Gauge. No correction for well deviation has been applied to 7” or 5” tubing hanger (Schiehallion and Foinaven
the skin factor and without PLT data it is not possible to infer respectively)
if other factors such as limited entry have any effect. 7” or 5 ½” TRSCSSV
7” or 5 ½” Fibreglass lined tubing (Duoline)
As can be seen perforating on CT with smaller 3-3/8” guns Hydrostatic set packer
appears to deliver higher skins than perforating with 4.5” TCP 5 ½” flush-joint tubing to stab into 7” lower completion
guns - which is not unexpected given the greater penetration 9 5/8” x 7” Liner hanger and packer
achieved with larger guns, reduced stand-off etc. The marginal Isolation valve (remotely cycled open after landing Xmas
over balances do not appear to cause significant damage, Tree)
although the most recent well perforated shows an abnormally 5 ½” Wire-wrap screens
high skin and in this case the overbalance was over 500psi
compared with normal overbalances prior to this of around Across Foinaven and Schiehallion, there have historically
200psi. This would possibly indicate that with an initial static been a number of differences in the screen injector design
overbalance of 200psi a dynamic underbalance is created once philosophy. The main differences being around the screen slot
the TCP guns are initiated, a situation that is not re-created sizing and the use of External Casing Packers (ECP’s) for
with CT conveyed perforation. When on long-term injection, minimising crossflow and shale induced damage. Typically
all wells have been capable of delivering the target injection Foinaven screen injectors have utilised 12 gauge wire-wrap
rate specified prior to completing the well. With the current screens and no ECP’s. Schiehallion screen injectors have
total well potential being approximately 1.5 times greater than typically used 10 gauge wire-wrap screens and multiple
the FPSO WI capacity, the main constraint on WI is therefore ECP’s.
facilities related. Water injection capacity is therefore
allocated across the fields with a view to maximizing
production potential, other constraints which are applied are
the Bottom Hole Injection Pressure (BHIP) to remain below
the cap rock fracture gradient and the inability to make choke
SPE 89745 9

Figure 14: Screen Sandface Completion Schematic (all completed with WWS). This consists of initial (attempt at)
injection, and on ceasing injection, pressure appears to be
locked into the well, with no leak-off taking place. Pressure
Liner Hanger then has to be bled down manually from surface (e.g. by
Wireline Entry Guide bleeding off 10-70bbls).
7” x 5-1/2” Crossover
There may be a possible formation permeability and/or mud
Formation Isolation Valve connection. The table below shows that four out of the five
wells have high average permeability (greater than 500md)
and/or permeability-thickness (greater than 100,000md-ft).
Screen Interval The four Schiehallion wells are all completed in the T31a,
which was also noted as being a possible ‘problem’ formation
External Casing Packer in relation to the observed skin values in some wells.

Screen Interval Table 6 Lock-Up Well Observations

External Casing Packer Well Zones Flow/Bhd Test Perm Kh (mD-ft) Observation
(mD)
Screen Interval W41 T34L Bullhead 3.8e+3 Post-inj
1700psi
CW11 T31a Flow 850 4.3e+4 Post-inj
Bullnose 1600psi
WW06 T31a&b Bullhead 200* (core 1.2e+5 Post-inj
1000) 1880psi
WW05 T31a Flow/CT 850 1.7e+4 (2.0e+5 Pre-inj
Sizing of sand screens has evolved considerably over the Diesel expected) 2000psi
field lives to date. Initial screen selection for Foinaven in 1994 CW16 T31a Bullhead/CT 1.3e+5 Pre-inj
Mud Breaker expected 2250psi
was based on using the results of sieve analyses from a range
of cores across the field. Applying the guidelines developed by The suggested lock-up mechanism (Laurence Murray, pers.
Coberly, Saucier et al. for a wide formation sand distribution comm.) is that fracturing initiated through mud cake/formation
(D10 range 100 – 700 microns), 12 gauge screens were damage carries debris from mud and/or sloughing unstable
selected to maximise sand retention while minimising the risk shale formations along the fracture, plugging the fracture face.
of plugging from fines and/or drilled solids. Injection water (or mud) leak-off then only occurs at the very
tip of the fracture. When injection is stopped, the fracture
More recently, sieve and laser particle size analysis have closes slightly, leaving only a highly damaged fracture face
been used as a guideline to select a range of potentially exposed, severely limiting leak-off and resulting in pressure
suitable screen sizes. Sand retention testing is then carried out ‘locked’ in. An alternative mechanism also suggested is that
to determine the optimum screen size for the particular mud-induced formation damage prevents fracturing in the
formation sand. sandstone formation, and fractures are actually induced in
shales/caprock. Some injection occurs, the shales collapse as
Coiled tubing interventions conducted on Foinaven screen they are water-sensitive, and no leak-off can take place.
injectors W11 and W14 during 2002 indicated that the screens
might have been over-sized. Subsequent retention testing with Figure 15: Measured Post-Fracture Skin against
core from relevant offset appraisal wells would also tend to Average Shaker Screen Size during Drilling
support this theory. Although there is considerable variation in
PSD across the fields, retention testing consistently shows that Post-Frac Skin vs Reservoir Mud Conditioning

10 gauge wire wrap screens are most appropriate for WoS.


20
ECP’s have also been used in six Schiehallion injectors, Skin higher than this - WWS10 WWS10
compared to only the recent W15 on Foinaven. Following the 15 not taking water

WWS12 WWS10
C&P
W14 intervention where the majority of the fill material was
Post-Frac Skin

10 FOI Oil Leg


C&P
found to be fines (<44 microns), the theory of shale induced 5
C&P FOI Water Leg
WWS10 WWS12
damage has been proven. WWS10
WWS12
WWS12
SCH Oil Leg
0 WWS10 SCH Water Leg
WWS10 WWS12
C&P WWS10
ECP’s also give benefits for future interventions in the -5 WWS10
WWS10 WWS12
C&P
form of zonal isolation by either plugging or straddles and also -10
for stimulation operations. 120 140 160 180 200 220 240
WWS10 10gauge Wire-Wrapped Screen
Average Shaker Screen (m esh) WWS12 12gauge Wire-Wrapped Screen
C&P Cased and Perforated
Drilling Fluid Conditioning Practices

‘Lock-ups’ have been noted in a number of wells at start-up – Injector completions with either 10 or 12 gauge WWS have
Foinaven W41, Schiehallion CW11, WW06, W07 and CW16 typically had mud conditioned to 230mesh (270mesh if
10 SPE 89745

running Premium screens) prior to running the lower formation damage encountered in both high and low perm
completion in order to minimise the risk of screen plugging. core. “Lock-up” was encountered in T31 injectors that had
Shaker screens are fined up as early as possible during drilling been both back-flowed and bull-headed. The test sequence for
of the formation to minimize circulation time required after these results had the mud flowed from the core first prior to
drilling, prior to running the lower completion. There are no the core being turned around to water injection, i.e. simulated
observations to suggest any revision to this specification for backflow.
10- or 12-gauge screens. One well with a post-frac skin of 18
and shaker screen mesh of 270 is Schiehallion WW08 (aka The damage in some of the higher permeability core is striking
WS1), the well drilled with Water Based Mud incorporating and would appear to explain the “lock-up” observed.
Calcium Carbonate solids. Finer specification conditioning to Unfortunately these results were obtained prior to the
270mesh shows no evidence of improvement in Flow Schiehallion T31 layer being identified as a particular concern
Efficiency. and no high permeability T31 core was tested. The concept of
lock-up in the high permeability injectors, while not a
C&P wells seem to be affected by damage and foregone conclusion, is obviously a real probability based on
conformance issues as well as WWS wells (Figure 15). This the results below.
indicates that whatever damage is found in the wells is more
likely to have been induced during drilling of the formation, Table 7: Foinaven Formation Damage Results
rather than being directly related to running the completion, or
reflects the lack of clean up in C&P wells. Sample Depth Fluid Applied Base Kw Kw after mud
No. (m) application /
@ Sro
A possible mechanism is as follows. The coarser particles drawdown /
(mD) injection (% change
(which would have acted as bridging agents during drilling) on base Kw @ Sro)
are actually being screened out as the mesh is fined up. Their
presence would otherwise give a better mud cake during 1B 2369.30 OBM 393 13.3
(Carbosea 1)
drilling, reduce losses/invasion, and prevent more damaging T32 (-96.6%)
fine particles from entering the formation. 2B 2369.47 Field OBM 331 2.77
(Carbosea 2)
T32 (-99.8%)
Most of the wells with high PFO-derived permeability or 5B 2079.15 Field OBM 29.5 14.2
(Carbosea 2)
kh have relatively low Flow Efficiency, possibly suggesting a T34 (-51.9%)
mud invasion/leak-off issue (Figure 16). Subsequent formation
damage tests confirm the higher permeability rock is more
susceptible to Formation Damage. Table 8: Schiehallion Formation Damage Results

Figure 16: Post-Frac flow Efficiency vs Permeability Sample Depth Fluid Applied Base Kw Kw after mud
Thickness No. (m) application /
@ Sro
drawdown /
(mD) injection (% change
Post-Frac Flow Efficiency vs Permeability-Thickness on base Kw @ Sro)

1A 1999.16 OBM 18.5 30.2


Test kh (m D-ft)
(Carbosea 1)
0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05 T31 (+63.2%)
1000% 2A 2016.28 Field OBM 21.7 16.7
Flow Efficiency (7/(7+Skin)) (%)

T31 (Carbosea 2) (-23.0%)


W07
3A 4629.59 Field OBM 241 210
C02 FOI Oil Leg
FOI Water Leg
T34 (Carbosea 2) (-12.9%)
100% 5A 4629.87 Field OBM 329 132
SCH Oil Leg
W03
SCH Water Leg
T34 (Carbosea 2) (-59.9%)

10% Once on injection the wells are observed to clean-up quickly


with thermal effects assisting this greatly, fracture behaviour
generally being observed in the step rate injection start-ups
Formation Damage Results performed by the rig prior to suspension for tie-in etc.

Formation damage testing has been performed on a variety Clean-Up Methods for Screen Completions
of core from different zones in both Foinaven and
Schiehallion. The focus of the work has been on the Oil Based Standard WoS practice is to initiate injection prior to the
Mud (OBM) with which the vast majority of both producers rig leaving location to establish the FIV is open, or that the
and injectors have been drilled. TCP guns have fired, and that the injection rate requirements
have been met. A debris sub is included above the TCP guns
Tables 7 and 8 present a sample of formation damage therefore injection can be established regardless of the guns
results from both Foinaven and Schiehallion comparing the dropping. In screen completions the proposed method of
SPE 89745 11

cleaning up the well prior to initiating injection has evolved technique in achieving good injection conformance, initial
through several iterations. The clean-up techniques will be assessment of production data from the offset producters
outlined below and several performance indicators evaluated would suggest that wells completed in the T31 and T32 have
to show their corresponding effectiveness. responded but the well completed in the deepest T25 zones has
not. The well will not have full “Smart” capability until later
The first 4 screen injection wells (CW10, CW11, CW12 in the well life when a CT deployed straddle will be deployed
and WW05) in Schiehallion and W14 in Foinaven were to give the required zonal isolation to make the valves,
cleaned up by backflowing the solids laden mud and installed during the initial completion phase, capable of
hydrocarbons to the semi-submersible rig requiring a well test controlling the WI flow distribution.
package, extended rig time and environmental impact due to
flaring. After backflow, injection was then initiated to confirm Completion Efficiency and Injection Performance
injectivity. CW11 and WW05 were single zone completions in
the T31, CW12 was a multi-zone injector completed in the In addition to life of well issues and the robustness of the
T34, T35 and T31, CW10 was a dual T31 zone injector. sand control, the completion efficiency and injection
performance are crucial to the performance of the injection
To minimize the cost and time associated with well testing, wells. 2 measures of the success of the completion are
injector clean-ups in screen wells were switched to presented below;
bullheading the completion fluid followed by approximately
2000bbls of treated seawater. Initial assessment showed no - Completion efficiency based on skin
difference in performance between bullheaded and backflowed - Completion efficiency based on %age of mobility (kh/µ )
clean-ups. The completion fluid below the FIV is solids laden thickness accessed.
OBM with clear completion brine above the FIV. 7
Schiehallion and 3 Foinaven Screen Injection Wells have been Well injection rate for steady-state radial flow is given by
cleaned up in this fashion. These 10 bullheaded wells were a the following equation:
combination of single and multi-zone completions.
Q = k h krw (Pinj-Pres)/{141.2 µ [ln(re/rw) – 3/4 + S]}
In multi-zone injection wells some enhanced clean-up (7)
techniques have been successfully applied to meet specific
well objectives. Where
Q = injection rate (bwpd)
In Schiehallion WW08 the well was drilled with a Water k = formation permeability (md)
Based Mud (WBM) and a breaker used to attack the filter h = formation thickness (ft)
cake. After running the Upper Completion and Xmas Tree the krw = relative permeability to water
well was then bullhead to initiate injection. This approach was Pinj = bottomhole injection pressure (psi)
taken in WW08 as it was a 4 zone injector where the primary Pres = reservoir pressure (psi)
target (70%) of the reserves was at the toe of the well below µ = water viscosity (cP)
the Oil Water Contact (OWC) and the three secondary zones re = effective drainage radius (ft)
lay above the OWC. A “Smart” completion was proposed but rw = wellbore radius (ft)
rejected for the well. A primary risk to the now non-smart well S = skin
was poor injection conformance and given the criticality of
achieving good sweep a water based mud and filter cake Typically, injection wells are hydraulically and/or
clean-up treatment was designed at short notice. As shown thermally fractured, having skins (S) of around –2 (or more
below the WBM/breaker was successful in achieving good negative), and with {ln(re/rw)-3/4} of around 6-7. An
conformance (kh distribution). undamaged, unfractured well has a skin of 0. The flow
efficiency of a completion for a given permeability-thickness
In horizontal Schiehallion injector NW01, a “Toe (kh) is therefore here defined by the term
Fracture” was performed which simply means the ECP in the
well was used to initiate injection into the sand package at the Flow Efficiency = 7 / (7+S) (8)
toe to ensure this zone took water when the well was put on
injection. This “Toe Fracture” involved a small pumping although it would be expected that water injectors (being
period (30bbls) as part of the lower completion program to fractured) should have a flow efficiency of around 140% for a
initiate injection with mud, keeping mud in the hole mitigates skin of –2. Where the data is available, the calculated
against any well control issues. Again the conformance data mechanical skin value has been used.
shown below indicates this process was successful.
Reservoir conformance will be determined by whether the
In the Foinaven multi-zone “Smart” Injector W15 the entire permeability thickness of the formation is contacted by
solids laden OBM was displaced to a solids free mud prior to the induced fracturing, or whether these fractures grow out-of-
the FIV being closed and ECPs set. The well was then zone and contact more formation than intended. The
bullheaded to initiate injection. No RFT or PLT data is conformance efficiency or kh Efficiency term is therefore
available from the well to assess the effectiveness of this
12 SPE 89745

given by the term (with the measured kh being derived from Figure 18: Water Leg Flow Efficiency Ranking
PFO analysis using a fluid viscosity of 1cP)
200%

Flow Efficiency (7/(7+S)) (%)


180%
kh Efficiency = kh / (Expected kh/µ) (9) 160%
140%
120%
with the Expected kh/µ values being determined by 100%
allocation of RFT/MDT mobility test measurements for the 80%
60%
WoS fields. The downhole temperature of injection water 40%
would typically be expected to result in a water viscosity in 20%
0%
the range 0.5-1cP.

CW15 WWS12 TFF

NW01 WWS10 TFS

CW17 C&P SZ
LW05 C&P

W22 C&P
W13 C&P SZ

W25 C&P SZ
W24 C&P SZ
W12 C&P SZ

CW16 WWS12P SZ
LW04 C&P SZ
WW09 C&P SZ
NW03 WWS10

WW08 WWS10 WBM

CW13 WWS10

NW02 WWS10
WW04 WWS10
WW06 WWS10

W11 WWS12

CW12 WWS10 Flow

CW10 WWS10 Flow


W41 WWS12 SZ

CW11 WWS10 Flow SZ


W14 WWS12 Flow SZ

WW05 WWS10 Flow SZ


Completions Efficiency Based on Skin

The following plots (17, 18 and 20) showing rankings of


skin and injection conformance have the following key;
Where;
WWS = Wire Wrapped Screen Some general observations can be made regarding the
C&P = Cased and Perforated effectiveness of some of the completion types;
WBM = Water Based Mud
SBM = Synthetic Oil Based Mud - The Foinaven injectors completed with 3-3/8” guns
conveyed with CT have higher skin values than the
WWS SBM Bullhead Schiehallion/Loyal wells completed with 4.5” guns
C&P SBM conveyed on the Upper Completion Tailpipe
- Schiehallion CW17 perforated with 4.5” guns performs
WWS SBM Bullhead Toe Frac
WWS SBM Flow poorly which may be related to the comparatively high
static overbalance of 500psi (see Table 5) compared to the
WWS WBM Bullhead
other wells completed with approximately 200psi static
overbalance with 4.5” guns
Figures 17 and 18 below shows there is evidence of - WWS completions would appear to be the hardest to
significant near-well damage in WW08 (drilled with Water predict, the completion efficiency showing the extremes of
Based Mud (WBM)), W11 (Water zone only) (possibly performance.
sandfill and/or undisplaced mud), W13 (possibly fill/debris) - There is possible evidence from well C04 that mud
and CW17 (relatively high over-balance). flowback to the rig may assist performance in a high kh
well vs other high kh wells which were bullheaded.
Figure 17: Oil Leg Flow Efficiency Ranking
Completions Effeciency Based on Near Wellbore
Flow Efficiency (7/(7+S)) (%)

400% Conformance
350%
300%
250% Very little PLT data is available on the injection wells - 3
200% PLTs from the initial start-up of W24, WW06 and CW12 are
150%
the only data available. As a substitute the RFT mobility data
100%
50% (where available) has been used to create an estimate of the
0% total kh/µ by zoning the wells based on mobility and summing
CW15 WWS12 TFF
NW01 WWS10 TFS

CW17 C&P SZ
LW05 C&P

W22 C&P
LW04 C&P SZ

WW09 C&P SZ
W25 C&P SZ

W12 C&P SZ

W13 C&P SZ
W24 C&P SZ

CW16 WWS12P SZ
NW03 WWS10

CW13 WWS10

NW02 WWS10
WW08 WWS10 WBM
W11 WWS12

WW04 WWS10
WW06 WWS10

CW12 WWS10 Flow

CW10 WWS10 Flow


CW11 WWS10 Flow SZ

W41 WWS12 SZ
WW05 WWS10 Flow SZ

W14 WWS12 Flow SZ

the average k/µ*TVD thickness for each of the zones. Figure


19 shows an example of this for the CW17 well.

The total Mobility Thickness derived from the PFO


analysis is then compared with the theoretical Mobility
Thickness to give an estimate of the near well bore
conformance. Nearwellbore conformance is a potentially
critical factor in maximizing sweep and reserves recovery,
especially in multi-zone injectors where failure to initiate
injection in one zone or more could have a significant impact.
SPE 89745 13

Figure 19: CW17 Zonation of RFT Mobility the shale at the top of the reservoir and subsequent injection
into a water-bearing sand. This has been known to occur in
well CO2 (CW11) (where an offset well encountered
significant over-pressure in thin sands above the reservoir),
and possibly in C04 (CW12), C09 (CW13) and N02z (NW02).

LW05 is believed to be injecting out-of-zone into the T35c


due to a poor cement job, the out-of-zone injection being
inferred from voidage considerations and over-pressuring
observed in other wells without any zonal offtake. This is
consistent with the PFO kh Efficiency interpretation.

Many/most wells are now choked back to avoid injecting


above caprock fracture pressure limits.

While PLT data has been observed to be in short supply,


the data that is available does allow some relevant
observations to be made. Figure 21 shows the PLT data from
Schiehallion WW06, which was cleaned up by bullheading the
drilling mud through the completion. The two passes
presented at different injection rates clearly show that the
injection front is moving down the well with time/increased
As discussed above WW08 reservoir section was drilled WI rate (this effect is seen on more passes than the two
with water based mud (WBM) and a breaker spotted prior to presented here). This PLT data logged inside the sand screens
closing the FIV to dissolve the filter cake. WW08 was a 4 clearly shows that the mud can be bullheaded from inside the
zone injector with the primary target containing 70% of the completion, however logging inside the screens with spinner,
reserves located at the toe of the well. As shown in Figure 20 density and temperature may not yield much information on
this well has achieved around 70% injection conformance the injection performance at the sandface.
compared to the theoretical value. However, the measured skin
value is high as formation damage testing had not then been 70 – 80% of the predicted mobility of the well
carried out as part of the filter cake removal fluid design, this 201,000md.ft/cp, using the RFT data, lies in the completed
skin value has remained high throughout 18months of interval 3420 – 3502m where the PLT clearly shows there is
subsequent injection. no injection (for presentation purposes the plots above have
been truncated just below the bottom active injection interval,
As previously outlined the “Toe Fracture” in NW01 was log data was recovered down to 3502m).
used to initiate injection into the sand package at the toe to
ensure this zone took water when the well was put on Figure 21: WW06 PLT Data
injection. The conformance data shown below indicates this
process was successful in achieving 90% of the estimated First Pass at 10,000bwpd
theoretical value.

Figure 20: Estimate of Conformance based on kh


Efficiency

350%
300%
kh Efficiency (%)

250%
200%
150%
100%
50%
0%
CW15 WWS12 TFF
NW01 WWS10 TFS

CW17 C&P SZ
LW05 C&P

W22 C&P
WW09 C&P SZ

W13 C&P SZ
W25 C&P SZ

W12 C&P SZ
LW04 C&P SZ

CW16 WWS12P SZ

W24 C&P SZ
WW08 WWS10 WBM
N02z WWS10

CW13 WWS10

NW03 WWS10
W11 WWS12

WW04 WWS10

WW06 WWS10
CW12 WWS10 Flow

CW10 WWS10 Flow


CW11 WWS10 Flow SZ

W41 WWS12 SZ
W14 WWS12 Flow SZ

WW05 WWS10 Flow SZ

Out-of-Zone injection is particularly an issue in


Schiehallion, where there is evidence (from kh Efficiency,
well modelling and seismic survey) of fracture growth through
14 SPE 89745

Third Pass at 29,000bwpd sump beneath this can only be a matter of conjecture. The
increase in permeability observed on the WW06 PFO data
once the WHP is increased is consistent with the performance
observed on the PLT data.

Figure 22: Schiehallion WW06 Hall Plot (Screen Inj.)

5000 WW06 HALL PLOT 500

4000 400

3000 300

WHIP
2000 200

1000 100

0 0
0 10000 20000 30000 40000 50000 60000
CUMULATIVE INJECTION (MBBLS)

While not an exact mathematical match, the PLT data would The W11 Hall plot in figure 24 clearly shows several
therefore likely appear to confirm the poor conformance declines in injection performance (slope change upwards) that
observed on WW06 (Fig 20) by comparing the PFO and RFT are not related to changes in reservoir pressure i.e. they are
mobility data. It also indicates that bullheading mud in early warning indicators that the well is at risk.
multizones or long wells can potentially lead to poor
conformance. Figure 23: Foinaven W25 Hall Plot (C&P Inj)

3500 W25 HALL PLOT 4000


Life of Well and Water Injection Failures
3000

Cum WHIP (x1000


3000
Some focus was placed on the life expectancy of the 29 2500
water injection wells currently in place WoS when 2 Foinaven
WHIP

psi)
2000
2000
screen injectors suffered catastrophic loss of injectivity. Well 1500
interventions encountered sand fill in the wells 60 – 70m 1000
1000
above the top screen depth. Change in slope related
500
to choking well back
0 0
Comparison of the Hall Plots for three wells is shown
0 10000 20000 30000 40000 50000
below; CUMULATIVE INJECTION (MBBLS)
Schiehallion WW06, a screen injector that has been on
injection for 5 years and has currently injected 55mmbbls
Foinaven W25, a C&P injector that has been on injection Figure 24: Foinaven W11 Hall Plot (Screen Inj)
for 6 years and has currently injected 42mmbbls.
3500 W11 HALL PLOT 5000
Foinaven W11, a screen injector that lost injection
capability after 3 years and 7mmbbls of injection. 3000
4000
Cum WHIP (x1000

2500
The WW06 and W25 Hall plots in figures 22 and 23 show
WHIP

2000 3000
psi)

that the well performance has essentially remained unchanged 1500 2000
apart from several minor fluctuations related to changes in 1000
WHP. Changes in WHP will result in changes in slope on the 1000
500 Injectivity decreasing
Hall Plot in WI wells as fracture behaviour changes and zonal with Increasing WHP
injection splits alter, both of which potentially cause changes 0 0
0 2000 4000 6000 8000
in Kh, skin and reservoir connectivity. The horizontal line on
CUMULATIVE INJECTION (MBBLS)
each plot indicates the theoretical cap rock fracture gradient.

The WW06 PFO analyses confirm that the Kh has Foinaven wells W11 and W14 had a number of potential
increased and skin has increased after the WHP has been reasons for their poor performance, which may be related to
increased. This would appear to give good confidence that some or all of the following factors;
these completions are currently unaffected by sand fill across - Low specification metallurgy in the sandface completion:
the zone currently on injection. What is happening in the 25%Cr screens were run with 13%Cr base pipe between
screen joints (Schiehallion practice is for all 25%Cr or
SPE 89745 15

higher specification below the FIV which is placed a screens too early in the drilling process can lead to problems
minimum of 50m above the casing shoe. initiating injection.
- Oxygen control in the injection water from the FPSO has
been variable with frequent excursions above the target of Acknowledgements
10ppb oxygen content. This exacerbates the effect of
running metallurgy of less than 25%Cr specification in the The authors of this paper wish to thank BP, along with the
lower completion. Foinaven co-venturers (Marathon Oil UK Ltd and Energy
- Larger 12gauge screens run in Foinaven wells W11 and North Sea Ltd) and Schiehallion co-venturers (Shell UK Ltd,
W14, while the bulk of the Schiehallion injection well Amerada Hess Ltd, Statoil Exploration UK Ltd, OMV UK
stock have 10gauge screens. Foinaven W41 and Ltd, and Murphy Petroleum Ltd.) for permitting the
Schiehallion CW15 also have 12gauge wire wrap screens publication of this material.
in place.
- Dual zones completed without annular isolation (ECPs) A variety of people have recorded completion installation
installed where only single zone targets were planned and performance data across the West of Shetlands Fields
- Relatively poor reservoir quality compared to other WoS since 1997 and the case study presented here would not be
wells with low Net to Gross increasing potential for fines possible without this data.
migration.
- Both wells had problems remotely opening the FIV The authors would also like to thank Laurence Murray of
requiring CT interventions. The current FIV has a high BP and other members of the WoS team for their contributions
success rate of opening remotely. and review.

Conclusions References

1. For subsea wells properly designed screen only and 1. C. Bennett, J.M. Gilchrist, E. Pitoni, R.C Burton, R.M.
Cased and Perforated completions have a demonstrable life of Hodge, J. Tronsco, S.A. Ali, R. Dickerson, C. Price-
well in excess of 5 years in relatively poorly sorted reservoirs Smith and M. Parlar: Design Methodology for Selection
requiring sand control. The maximum life of well that can be of Horizontal Open-Hole Sand Control Completions
expected from these completions is still to be determined Supported by Field Case Histories. SPE 65140
however the current reservoir surveillance data indicates that
the 27 injection wells still on-line do not currently show any 2. S. Bruce: A Mechanical Stability Log. IADC/SPE 19942.
warning indicators.

2. In terms of injection conformance screen completions


give the extremes of performance, especially where
conformance is poor. On the other hand, where specific
attempts have been made to improve conformance e.g. using
WBM plus breaker or performing a “Toe-Frac”, the screen
wells show good injection conformance, although the data-set
of enhanced clean-up techniques is a limited.

3. In terms of injection conformance C&P with 4.5”


TCP guns performs well, typically showing greater than 50%
of the theoretical value of conformance.

4. A procedure for selecting the appropriate completion


technique for water injectors in an unconsolidated sand prone
reservoir has been presented. This is based primarily on well
trajectory, UCS data from offset wells and prognosed reservoir
zonation.

5. Adequate injection performance in thermally


fractured reservoirs can be achieved with nominal static
overbalances up to 200psi without requiring the operational
complexity and risk involved in underbalance perforating. A
dynamic underbalance may be being generated as part of the
TCP operation, however this has not been proven.

6. Drilling fluid/formation damage impacts are at least


as important for injectors as they are for producers. In high
permeability formations conditioning the mud over fine shaker

Você também pode gostar