You are on page 1of 10

SPE 112924

Identification of Wells with High CO2-Leakage Potential in Mature Oil Fields


Developed for CO2-Enhanced Oil Recovery
T.L. Watson, SPE, T.L. Watson & Associates Inc.; S. Bachu, Alberta Energy Resource Conservation Board
Copyright 2008, Society of Petroleum Engineers
This paper was prepared for presentation at the 2008 SPE/DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 1923 April 2008.
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
Previous work, presented in SPE 106817, Evaluation of the Potential for Gas and CO2 Leakage along Wellbores, described
a method to predict the potential for wellbore leakage which primarily occurs in the shallow areas of a wellbore.
The work presented here focuses on the potential for leakage to occur from the deep regions of a wellbore, particularly
from viable oil reservoirs. The potential for wellbore leakage where CO2 emhanced oil recovery (EOR) or sequestration is or
may be conducted is specifically investigated.
The Alberta Energy Resources Conservation Board collects and maintains data regarding cement types used in primary
cementing of casing strings, stimulation information and abandonment data. These data were used to determine the potential
for deep wellbore leakage in the presence of CO2. This deep wellbore leakage potential is then coupled with shallow wellbore
leakage potential to predict which wells may leak to other reservoirs, potable groundwater aquifers or to the atmosphere in a
CO2 sequestration or EOR project.
Cements with additives such as bentonite have been shown to be particularly susceptible to CO2 attack. Wells were
screened for cement blends placed in the deeper sections of the wellbore during primary cementing. This information is
useful in predicting if a wellbore will remain leak free if CO2 is placed in the reservoir and the wellbore is contacted.
The stimulation method was evaluated to determine if injected CO2 may break through to existing wellbores prior to full
reservoir sweep, thus decreasing the time that the wellbore could remain leak free. Stimulations such as hydraulic fracture,
perforating and acidizing with pressure were deemed to increase the likelihood that wellbores would leak due to cement
sheath cracking coupled with CO2 attack of cement and casing.
The abandonment method was also evaluated. In Alberta the primary form of zonal abandonment utilizes a mechanical
bridge plug capped with cement. The bridge plug material is expected to fail in the presence of CO2 due to CO2 attack on
elastomers and cast iron.
Two large field evaluations are presented as case studies.
Figure 1: Shallow and deep
Introduction
Several client studies have been conducted to date in Alberta, Canada, to determine factors which
may affect wellbore leakage. The findings of these studies were presented previously (Bachu and
Watson, 2006; Watson and Bachu, 2007). These papers, particularly SPE 106817, presented an
analysis of the factors that affect the leakage potential in the shallow part of a well, and a
decision-tree type model which enables ranking of a wells potential to leak based on factors
determined in the client studies.
The original ranking system generally focused on leaks from the shallower regions of the well
(Figure 1). These leaks typically manifest themselves at surface as annular pressure (surface
casing vent flow-SCVF, known also as sustained casing pressure) or soil gas migration. This
ranking system did not take into account the potential for a native or injected fluid or gas to
create or increase the potential for a well to leak.
In Alberta several reservoirs are currently being evaluated for CO2 enhanced oil recovery
(EOR), while more than 30 depleted oil and gas reservoirs and deep saline aquifers are used for
acid gas (CO2 and H2S) disposal. In an attempt to determine the potential for a well penetrating
one of these reservoirs or aquifers to leak CO2 or acid gas from the target reservoir, a computerbased tool was developed. This tool uses the data available in electronic form from the Alberta

areas of the wellbore.

SPE 112924

Energy Resources Conservation Board (ERCB) to create a score for each wellbore. This score indicates the leakage potential
from both the shallow and deep regions of the wellbore as depicted in Figure 1. Data were also compiled to allow for the
future evaluation of possible consequences of a leaking wellbore.
The shallow leakage factors have been determined by analysing known leakage in the field and determining the most
prevalent factors that lead to annular pressure and soil gas migration (Bachu and Watson, 2006; Watson and Bachu, 2007).
The deep leakage factors, however, have been determined from theoretical deduction and laboratory results.
Shallow Leakage
Factors previously determined to have an effect on shallow wellbore leakage (Bachu and Watson, 2006; Watson and Bachu,
2007) were assigned values which reflected the influence that a particular factor has on shallow wellbore leakage. Table 1
indicates these factors, the criteria and default values assigned based on these criteria. These values are then multiplied to
determine a wells Shallow Leakage Potential (SLP) score. Table 2 summarizes scores into general leakage-potential
categories.
Table 1: Shallow leakage factors.
Factor

Criterion

Spud Date
Abandonment Date
Surface Casing Size
Well Type
Geographic Location

1965-1990
<1995
244.5 mm
Cased
Special Test
Area
>2500 m
1.2-1.8
No
Unknown
No
Unknown

Well Total Depth


Well Deviation
Cement to Surface
Cement to Surface
Additional Plug
Additional Plug

Table 2: Shallow leak potential.


Meets
Criterion
Value
3
5
1.5
8
3

Default
Value
1
1
1
1
1

1.5
1.5
5
4
2
1.5

Shallow Leak Potential (SLP)


Low
Medium
High
Extreme

Score
<50
50-200
200-400
>400

1
1
1
1
1
1

In general the analysis is based on the potential for a well to have a leakage pathway along the wellbore and does not
depend on the fluid or gas source that may be leaking.
Deep Wellbore Leakage
Three factors were used to evaluate the Deep Leakage Potential (DLP) for a wellbore in CO2 sequestration, acid gas disposal
and CO2-EOR. Deep leakage was evaluated based on the presence of CO2 or acid gas and the impact this acidic environment
would have on wellbore construction and abandonment materials such as cement, steel and elastomers. Table 3 indicates the
factors, criteria and assigned values used to determine the possibility that deep leakage would occur. These factors are then
multiplied to obtain the wells overall DLP. Table 4 summarizes the product score values into general categories.
Table 4: Deep leak potential.

Table 3: Deep leakage factors.


Factor

Criterion

Fracture
Fracture
Acid
Acid
Acid
Perforations
Abandonment type
Abandonment type

count =1
count >1
count=1
count=2
count>2
count>1
Bridge Plug
Not abandoned

Meets
Criterion
Value
1.5
2
1.1
1.2
1.5
2
3
2

Default
Value
1
1
1
1
1
1
1
1

Deep Leak Potential (DLP)


Low
Medium
High
Extreme

Score
<2
2-6
6-10
>10

Cement Type
Several studies have been conducted to determine the effect of CO2 on the quality of cements used in wellbore
construction (Browning, 1984; Onan, 1984; Bruckdorfer, 1986; Krilov et al., 2000; Duguid et al., 2004; Kutchko et al.,
2007). Most of these studies indicate that cement will not withstand CO2 attack and will fail to provide a seal in the
casing/hole annulus when CO2 is introduced (Nelson and Guillot, 2006a). Recent laboratory work suggests, however, that
cements with low free water ratios may not be as susceptible to CO2 attack due to the formation of an impermeable barrier on
the cement sheath that halts further deterioration as shown in Figures 2 and 3 (Kutchko et al., 2007a). These studies also
suggest that the inclusion of additives, such as bentonite, which increase the free water ratio, increases the potential for

SPE 112924

cement break down in the presence of CO2 (Kutchko et al., 2007a). Since no such studies exist for the effect of acid gas on
well cements, it is assumed that the conclusions and results for CO2 would apply to acid gas as well since the reactions are
due to the acidic nature of the fluid (Kutchko et al., 2007a).
Figure 2: Kinetic results for the penetration
rate of CO2 -saturated brine of the cement.
(Kutchko et al., 2007b). (Graph courtesy of B.
Kutchko, U.S. Department of Energy.)

Bentonite, commonly referred to as gel, is the most


common additive to cement. This additive is used as an
extender to increase the yield of the cement, decrease the
cement density and decrease the fluid loss (Nelson and Guillot,
2006b). The addition of bentonite increases the set cement
porosity, increasing the cement susceptibility to corrosion due
to acidic formation fluids (Nelson and Guillot, 2006a). Cement
blends which include bentonite are often used as filler
cements to reduce the cost of cementing and the hydrostatic
pressure applied to the producing formation or weak
formations. Other additives, such as gypsum, which are highly
acid-soluble, are also expected to deteriorate rapidly in the
presence of CO2 or acid gas.
Inert additives such as fly ash, silicate (sand) and nitrogen
(foamed cement) do not increase the free water ratio and are
assumed for this study to have a lesser effect on the
deterioration of cement in the presence of CO2 (Nelson and
Guillot, 2006a).
The data, provided by the ERCB, include information
regarding the cement types used in the cementing of the
production casing. Table 5 itemizes the cement types found in
the ERCB electronic database and the values assigned for the
purpose of assessing a wells potential to be adversely affected
by CO2 or acid gas and potentially cause leakage. Often, several
different types of cement may be used to cement the wellbore.
Generally, better quality cement, designed for the well purpose,
depth, temperature and reservoir conditions, is run across the
zone or completion interval. Filler type cements are often run in
the shallower areas of the wellbore. The ERCB data indicate
these different cement types in a wellbore, as well as where
they are placed by providing the sequence that the cement was

Figure 3: SEM-BSE image of Class H neat cement cured


for 28 days. The CaCO3(s) is forming a barrier to further
CO2 infiltration (Kutchko et al., 2007a). (Photo courtesy of
B. Kutchko, U.S. Department of Energy).

Table 5: Cement types and values.


Cement Type
1:1 POZ MIX

Assigned
Value
1

Description

Cement, fly ash and various


quantities of bentonite
Unknown

1:1:# POZ
BLACKGOLD

1
1

CAP (NEAT)
CLASS X NEAT

1
1

FILL ECP
1
FOAMED
1
G + # PC SALT
1
G + # PC SAND
GPSL/GPCEM/THX

3
3

Cement and fly ash

Cap pumped on top of foam


cement, not applicable.
Various neat cements
Cement to fill annular packer,
not applicable
Cement foamed with
nitrogen
Cement with various percent
salt additive
Cement with various percent
silica sand additive
Gypsum and gel additives

SLOTTED LINER

Assumed gel additive to


reduce density
No cement, hole allowed to
slough in on casing
Blast furnace slag, reduces
cement porosity
No cement

SLURRY 6D

Unknown

TAPERED CASING
TH CEM/CEM
FNDU
UNCEM
CSG/LINER

No cement

Thermal cement, usually


sand or silica additive
No cement

LIGHT WEIGHT
3
SELF STRESS
1
SLAG

SPE 112924

pumped into the well. Cement data for wells spud prior to 1986 are sparse, however most wells spud after 1986 have detailed
information regarding cement type and placement. From this information it is possible to determine which cement type is in
the deepest portions of the wellbore. The cement type recorded as being at the bottom of the well is the cement type that this
evaluation is based on.
Stimulation and Perforations
The ERCB records information regarding most stimulation treatments to perforations or producing intervals in wellbores,
as well as the number of completed intervals in the wellbore. Two types of treatments and perforating are anticipated to have
potential effects on long term wellbore integrity due to excessive pressures (Watson et al., 2002), as shown in Figure 4, or
chemical reactions. In this study hydraulic fracturing and acidizing with pressure were classified as stimulation methods that
could have a negative impact on wellbore integrity. Carbon dioxide or acid gas
Figure 4: Cement sheath failure and
may migrate through cracks in the cement created by stimulation or
resulting cracks developed from pressure
perforating. These acidic fluids may increase and hasten the leakage by further
cycling the internal casing (Watson et al.,
2002). (Photo courtesy of Halliburton
deteriorating the cement sheath or corroding the steel casing.
Energy Services).
Hydraulic fracturing breaks the reservoir rock using a viscous fluid, high
pump rates and high pressures. The reservoir rock is fractured and the
fractures filled with a propping agent, usually sand, that creates a highly
porous and permeable flow path to the wellbore. Acidizing dissolves wellbore
scales, removes near wellbore damage and dissolves reservoir rock. These
stimulations can remove near-wellbore damage caused by drilling activity and
increase the flow potential near the wellbore.
In the situation where CO2, acid gas and/or water is used for reservoir
EOR, it is anticipated that CO2 or the acid gas may break through to these
reservoir fractures preferentially, bypassing unaltered reservoir rock due to the
higher permeability of the fracture (Marsters, 2007). As well, where wells
have been abandoned due to depletion or produced for long periods of time,
the localized reservoir pressure around these wells may facilitate the flow of
flood fluids to migrate towards these low pressure areas. Multiple completions
or perforated intervals provide a higher potential for cross flow between
discrete geologic zones within the wellbore itself.
The data were analysed to determine if a particular well was stimulated by
fracturing, acidizing or if it had multiple completions. Fracturing has been given a higher risk score due to higher treatment
pressures and deeper penetration into the reservoir, typically achieved by fracturing. In comparison, acidizing and perforating
are usually near-wellbore treatments. Both fracturing and acid treatments were counted for the individual wellbores. This
count information is used based on the assumption that the more times a well is stimulated, the higher the chance that casing,
cement and/or cap rock systems may be damaged. The number of completions is also counted and any well that has multiple
completed intervals was assigned a higher value due to the increased potential of crossflow.
Zonal Abandonment Method
In Alberta the ERCB allows three options for zonal abandonment of cased wells. The methods, as depicted in Figure 5
are: a cement plug that extends a minimum of 15 meters above and below the perforated interval, a cement squeeze through
the perforations with or without a retainer, or, the most commonly used method, a mechanical bridge plug capped with 8
meters of cement (ERCB 2007).
Figure 5: Regulatory approved zonal abandonment methods in Alberta, Canada.
Cement plug set across
perforations.

Cement squeeze with


retainer to perforations.

Bridge plug capped with 8


meters of cement.

SPE 112924

This factor identifies the abandonment method for a particular well and assigns a value based on these criteria. It is
anticipated that the bridge plug abandonment method will have a shorter life than other methods due to mechanical failure,
change of reservoir pressure due to the injection of CO2, acid gas and/or water, or a change in the gas/fluid chemistry below
the bridge plug (Schremp and Roberson, 1975). Bridge plugs are typically made of cast iron and an elastomer sealing unit as
shown in Figure 6. Both iron and the typical nitrile elastomers used are subject to CO2 or acid gas attack that can lead to seal
failure (Schremp and Roberson, 1975).
Cement caps placed above bridge plugs are typically dumpFigure 6: Bridge plug capped with 8 meters of cement
bailed into place. Cement caps evaluated in an earlier client study
with infiltrating CO2.
were found intact in only 50% of the wells investigated. This
corresponds to laboratory findings that indicate dump-bailing of
cement may be ineffective in providing adequate seals (White et al.,
1992). Based on these findings, the ability of dump-bailed cement
plugs to maintain an effective seal above the bridge plug is
considered negligible for this study.
Zonal abandonment type had originally been included in the
potential for shallow wellbore leakage. For this study the zonal
abandonment method has been evaluated as deep potential for
leakage, because it has a direct effect on the potential for wells to
leak from the CO2 sequestration or acid gas disposal formation up
the inside of the production casing where it may impact all horizons
in the wellbore.
Wells that have not been cased are called drilled and abandoned.
These wells have been left out of the analysis since records in
Alberta indicate that the leakage occurrence rate is 0.5% of drilled and abandoned wells compared to 13% of cased
abandoned wells (Bachu and Watson, 2006). Additionally, no data exist to determine the cement type in the cement plugs
used to abandon uncased wells. It is assumed, based on field knowledge, that cement plugs typically consist of neat type
cements, and these wells will therefore have a lower DLP.
Case Studies
The Pembina and Zama oil fields in Alberta (Figure 7) were evaluated using the previously described method. These fields
are being considered for CO2 or acid-gas EOR, and pilots are currently being run in both. The results of the evaluation can be
used to determine the overall leakage potential for existing wells in oil fields where CO2 or acid gas injection may be used for
EOR and/or future sequestration. The information provided by this
analysis can be used to target particular wells for further evaluation based
Figure 7: Map of Alberta, Canada, showing the
location of the Pembina and Zama Fields in red.
on their individual DLP and SLP scores. Table 4 provides general
information for comparison between the two oil fields and the potential
for well leakage based on the predetermined factors.
3
u4
3
u4
1
1
43
u
3
1
4
u
u
34
43
b
1
u
3
34
5
3
"54
4
"5
5
1
"uu4

3
u
3

u
u
"u
1
4
4
b
"5
"3
u
5
*4
43
u
"
M
"4
4
45
1
31
b4
5
1
3
13
*3
1
uu
G
35
4
1
G
G

4
1
IG
5
1
u
5
4
u
3
E
b
4

3
u
u4
*3
*b
E
"b

13
u
1
5
G
1
*3
E
"*3
3
5
u
3
b
3"5
*E
4
3
3
G

4
b
1
4
4
u1
5
u34u
I**5

u4
O

u
1
3
*3
G
1
4
5
1
*4
G
3
5
*G

4
*5
4
5
1
G
K
4
G
5
"1
b
5
4
b
3
b
u
4
b
u
"G
3
u
E
b
3
5
3

u
u
5
34
1"E
*5
4
5
1
*1
O
G
5

"*3
4
1*5
"b
4

*4
3
5
4
5
G
I
E
4
*1
b
"u
4

u
3
4
5
1
1
u
5
3
G
b
"u3
1
3G
4b
5
*bu
u
*G

Zama Field

Table 4: Field data and results summary.

ALBERTA

Edmonton
3*3
5
*u
4"G
3
4
u
D
G
5
1
1u4
1
"3IG
G
3
1
4
31
*5
u5
"3
43
4
"3
"*3
N
O
3
I5
5
1
34
4
3
5
3
u3
u
"5I4
3I5
14
3
5
1
4
"*"4
u

1
"u
4
3
G
"3

1
b

35
3
u5
5
4
"4
1
3
5
"4
u
G
3
"3
4
""E
4
3
33
u
5
u
1
5
u
4
3
4
b
G
*"5
3
D
5
*O
*E
*3
"5
35
45
"u
*
G
"*1
I4
IG
1
G
4
u
3
5
u
I33
b"E
G
1
u
G
5
5
5
4
u
G
u
3
b
*u
5
"I3
5
4
4
4
*5u
u

O
N
4
b
u
4
u
*4
u

"b
D
"G
D
G

*u

*"u
G
3
G
4
D
4
3
3
"u
4
O
"4
G
D
I3
**4u
G
*L*"G
u
3
O
G
"O
"O
3
5
u
1
3
5
3
4
1
D
I
5
3
5
1
3
1

u
b
*5
*u
5
3
3
5
4

45
3
D
G
1
O
G
D
"4
N
1
1
5
1
K
25
*
4

G
3
"G
D
"G
5
IN
N
5
5
I1
"G
4
41
5u
3
1
3
4
E

O
*1
5
4
D
*G
2
4
D
u
3
G
b
2
"*I4
G
5
O
4
u
"
L
N
*5
1
4
E
4
*5
u
u
G
O
1
4
G
u
"
5
G
*"3
u
5
3
2
1
4
**5
G
u
5
1u
3
1
5
*1
u
G

15
E
D
2
ID
3
*5
"u
3
3
I5
O
G
I3
I**O
4
"3
3
3
u
5
u
3
4u
15
IO
3
*"b
4
*"5
u*b
4
G
D
I3
5
3
u
5
b
3
O
5
3
"E
L
5
*N
G
"u
D
O
*G
3
"5

"u
G
*u
""4
b
3*O

4
5
3
G
*"L
u
4
3
5
*G
3G
4
"5
u
5
u4
"u
*3
""Iu
44
3
G
1
G

5
*G
4
4
3
u
"3
3
5
u
*G
I4
b

G
u
*5
1
4
1
5
"N
u
1
4
u

O
5
IO
O
I5
L

"
1
N
D
u

4
3N
4
4
3
L
I5
D
*D
13
5
G
1
5
"G
4
5
3
1"u
I3
O
u
4
3
u
K
3
5
O
N
O
1
3
I"1
u
4
3
3
1
4
3
4
1
4
3
b
"*N
D
5
u"3
*"*5
u5
4
u
3
"I5
*IG
u
N
3
5

""G
*
u
5
**u
G
G
5
N
44
5
5
"O
5
1
M
G
"N
1
3
E
O
3
1
G
*45
4
"L
O
*L
1
O
1
5
u
E
1
5
u
3
u
*3
4
3
4
*4L
5
u
*54
E

5
"G

I1
N
K
E
O
3
u1
*G
3

5
4
G
4

"5
G
D
3
5

O
5
"G
u
3
G

4
5
3
"G

b1
IG
G
4
*1
3
5
3
"G
G
*5
u
5
"u
G
O
*"3
I1
IN
b
*5
""u
5
3
4
4
u
I3
L
"G
*"N
u
I4
E
*
5
O
N

L5
3
"1
G
*Ib
*13
*I5
ID
5
u
I
*
5
G
u
*4
3
4
u
*4
3
1

5
3
Iu
*Iu
5
"O
2
u
O
N
*3
G
N
2
*u
O
u
I5
O
5I
D
G
E
14
b5
4
4

*5
4
"5
"5
4
4
3
4
31
u
13
u
44
u5
G
D
N
O
M
3
O
D
E
3
*5
5
N
53
1
D
O
u
G
**D
1
*
2
4
1
u
5
4
G
4
3
O
4
1G
O
IL
"
4
D
u
*4G
O
N
3
3
5
5
3
4
*Ib
G
3
41
5
I3
L
3
G
*5
3
u
N
4

4
1
"N
I*u
*3
u
"O
4
u
G
"I3
*E
4
5
3
"G
O
I*4
1D
I
5
*3
G
L
D
4
II"O
u
*3
G
"G
u
I3

u
5
5
5
1
D
5
*G
G
I
4
1
3
*D
N
1
3
4
O
D
"
3
3
N
5
u
u
1
D
G
u
"4
*5
3
"1u
u
4
O
1
3
"5
3

u
u
"*5
u
5
N
G
I*O
G
*E
u
G
3
5
1
1
*D
O
4
"4
G
L
4
IG
5
4
N
1
b
IG
*1
5
LG
*O
4
3
1"3
*u
""*5
E
O
ID
5
5
*
u
L
4
G
3
L
4
*I"
1
G
4
4
G
*4
D
3
"1
D
*L
3
1
3
5
4
3
"5
IG
3
u
O
I"1
5
I"uG
4
O

5
5
5
"4
3
5
u
*I
u
3
u
1
u
5
4
3
G

u
5
3
4
*"3
D
u
u
D
O
G
54
O
u

"u
3
1
b
"K
1
I5
b
3
u
5
3
4
1
5
L
I35
14
u
L*G
N
O
I
4
I3
*5
u
L"u
1
"1
5
u
41
5
5
G
4
*5
G
4
5
b

O
1
3
4
4
"*O
4
I3
3
b
5
"5
O

*G
u
5
u
*G
3
u
**
1
5
3
u
4
1
5
*3
u
I"4
u1
*3
3"*5
1
3
G

4
3
4
1
41
I4
*u"5
u3
1
G
u5
35
3
4
u
3u
G
3
3"*G
3
1
"D
u
5
5
b1
u
4
u
1
bG
N
4
1
"5
O
3E
D
1
3
u
4
"5
4
u
IG
5
1
D
G
13
"u

4
4
1
u

4
3
u
14

4
3
"u
5
3
u
4
4
u
*"5
4
u
b
*3
3
u5
4
3
5
u
4
*
*
5
*
u
3
4
3
5
G
1

1
4
3
1

u
1

5
*
"
5
u
5
1
5
u
1
4
"
3
b
3
u
4
4
*
"
3
4
4
u
4
u
5

4
5
u
5
3
3
b
1
4
u
1
3
4
3
D
1
*
G
u
1
5
u
*

3
1
u
I
4
5
4
4
1uu 33

4
u
*3" 543 33
3

Pembina Field

Calgary

Pembina

Zama

Number of cased wells

9860

607

Number of wells drilled and abandoned

1050

106

% of wells with cement data

40%

64%

% of wells with high DLP cement score

28%

20%

% of wells fractured

75%

2%

% of wells acidized

47%

80%

% of wells abandoned

12%

13%

% of wells with multiple completions

11%

55%

% of wells with extreme DLP

14%

28%

% of wells with extreme SLP

7%

18%

% of wells with extreme SLP and DLP

1.6%

4.3%

SPE 112924

Pembina Field Wellbore Leakage Potential


The Pembina Field, the largest oil field in Canada, was discovered in the early 1950s with drilling continuing on until
today. Wells in the field typically are drilled to a depth of approximately 1600-1800 m. The wells are generally completed
and produced from the Cardium Formation which is a sandstone with an overlaying conglomerate and a thick shale caprock.
In recent years some uphole potential has been realized. The wells are generally vertical with single completions. Wells that
have been cased and completed are typically abandoned with a bridge plug set within 15 m of the perforated interval with 8
m of cement dump-bailed on top. Generally, wells in the Pembina Field were stimulated by hydraulic fracturing and to a
lesser extent by acidizing the completed interval.
Historically, the primary cementing requirements have been 100 m above the Cardium Formation at 1400 m or the
overlying Belly River Formation at an average depth of about 800 m (ERCB, 1990). Surface casing setting depth is at
approximately 180 m. The base of groundwater protection varies between 200 and 600 m, with typical Base of Ground Water
Protection (BGWP) at 470 m (Alberta Environment, 1995). Although there are wells with potentially 1200 m of uncemented
casing in the Pembina Field, there are very few gas or oil bearing zones within the overlying horizons to leak to surface.
External casing corrosion does not seem to be problematic in the area, with 1.3% of Pembina wells reporting casing failure
compared to 1.1% province wide.
The distribution of wells, by either DLP or SLP scores, for the Pembina Field is presented in Figures 8 and 9,
respectively. This analysis gives an overall picture of the potential for well leakage and which area of the wellbores is most
prone to leakage. Figures 10 and 11 indicate the geographic location of wells with extreme deep and shallow leakage
potential in the Pembina Field, respectively. This information would be useful when determining the areas of an oil field that
would be most desirable for CO2 EOR or sequestration, and gives some indication of the potential economic impact of
wellbore remediation prior to injection of CO2 at the scoping stage. The well locations of a particular DLP or SLP score or
range of score may be used to determine potential consequences by overlaying population density, ground water, surface
water or other potential leak receptors. Individual factors, such as which wells are abandoned or have at risk cement types can
also be analysed based on the nature of the proposed project.
Figure 9: Pembina Field SLP Score distribution.

1000

20

Figure 10: Location of Pembina wells with extreme


DLP shown in red against all field wells in black.
The green grid indicates 6x6 miles/square.

25

30+

00
+

15
DLP Score

10

10

90

80

0
0

2000

70

200

3000

400

4000

600

Extreme

60

800

High

50

1000

Low
Medium

1200

5000

40

1400

6000

Extreme

30

High

Low Medium

20

1600

10

Number of Wells with DLP


Score

1800

Number of Wells with SLP


Score

Figure 8: Pembina Field DLP Score distribution.

SLP Score

Figure 11: Location of Pembina wells with extreme SLP shown in


red against all field wells in black.
The green grid indicates 6x6 miles/square.

SPE 112924

The analysis of shallow leakage potential in the Pembina Field indicates leakage potential similar to the reported
provincial average of 6.1% (Bachu and Watson 2006), with only 7% of wells falling in the extreme categories. The analysis
of the deep leakage potential indicates that wells are predominantly cemented with neat cements that pose the lowest risk for
deterioration when in contact with CO2 compared to other common wellbore cements. The deep leakage analysis also
indicates that fracture stimulation may create leakage pathways in the deep regions of the wellbore and that fracture
stimulation is the main factor in DLP for the Pembina Field. Figures 10 and 11 also indicate that the potential for shallow
well leakage (SLP) in the field may be of lesser concern and is not as widely distributed as the potential for deep well leakage
(DLP).
The low potential for leakage scores, however, may be underestimated if wellbore abandonment procedures do not
change in the future. With only 12% of cased wells currently abandoned in the Pembina Field, the leakage scores may not
reflect future leakage potential due to abandonment method.
Cement type data are electronically available only for 40% of the wells. It was assumed that the unknown well cement
types would be type neat and would, therefore, have a lower potential for deterioration due to CO2 attack. This assumption
may underestimate the potential for deep well leakage.
Based on the available information it appears as though wells in the Pembina Field could withstand project
implementation of CO2 EOR or CO2 sequestration if wells abandoned in the future use a more robust downhole abandonment
method than bridge plug capped with cement. Investigation to determine zonal isolation after stimulation in wells indicated as
having extreme DLP would be beneficial in determining the actual effect of the stimulation on zonal isolation. More study is
required to determine the durability of neat cement and bridge plugs to improve the confidence of these findings.
Zama Field Wellbore Leakage Potential
The Zama Field was discovered in the mid 1960s. The productive zone is the carbonate reef Keg River Formation at an
average depth of 1600 m. The wells are generally vertical with multiple completions. On average, each well has two
completions, or perforated intervals, with several wells having five or more completions. Wells that have been cased and
completed are typically abandoned with a bridge plug set within 15 m of the perforated interval and with 8 m of cement
dump-bailed on top. Wells in the Zama Field were stimulated by acidizing the completed interval. Very few wells are
hydraulically fractured.
Historically, the cementing requirements have been 100 m above the Slave Point Formation, which overlays the Keg
River formation, at an average depth of about 1200 m (ERCB, 1990). Surface casing setting depth is at approximately 275 m,
however older wells have setting depths less than 200 m. The Base of Ground Water Protection varies between 75 and 400
m, with the average BGWP at 250 m (Alberta Environment, 1995). Within the uncemented depths of the wellbores several
productive intervals have been discovered since many of the wells were drilled. Overlying gas-bearing formations such as the
Jean Marie and the Beaverhill Lake, may contribute to gas leaking to surface. The reported casing failures for the Zama Field
are 6% of all wells compared to failures of 1.1% of wells in Alberta. These high numbers of casing failures may be explained
by the exposure of casing to the Wabamum Group and Banff Formation which are limestones within the exposed casing
depths. The presence of CaCO3 has an adverse effect on external casing corrosion and ultimately casing failure (Caswell,
1988).
The distribution of wells, by either DLP or SLP scores, for the Zama Field is presented in Figures 12 and 13, respectively.
These score distributions indicate that deep leakage potential is quite prevalent within the Zama Field. Figure 14 and 15
indicate the geographical location of wells with extreme leakage potential in the field. Because the Zama Field is made up of
several hundred reefs, this information would be useful when determining which pool would be most desirable for CO2 EOR
or sequestration and gives some indication of the potential economic impact of wellbore remediation prior to injection of CO2
or acid gas at the scoping stage. Both deep and shallow leakage potentials are broadly distributed throughout the field. The
deep leakage potential appears to have a denser distribution across the field. Although Figures 14 and 15 focus on extreme
DLP and SLP scores, the distribution charts indicate that there are large numbers of wells in the high categories for both DLP
and SLP. The well locations of a particular DLP or SLP score or range of score may be used to determine potential
consequences by overlaying population density, ground water, surface water or other potential leak receptors.

SPE 112924

150
100
50

SLP Score

DLP Score
Figure 14: Location of Zama wells with extreme DLP
shown in red against all field wells in black.
The bold green grid indicates 6x6 miles/square.

41
3 u 3 4u u
4
u
3
3 1
4
u
3
1
3
4
3u
u 3
1
4
4
4
u
"
u
34 "
u
4
3 3
"
1 u
4
"4
3
5"5" " ""
1
5
4 5 3
5
4
3
5b
4
" 43
5
5"

u 4
3 u
31
55 3
4
3
3 4
5
4
4
5
5
4
b
3
3
3
5
3 33b
54
5*
5""5
4
uu
u4 3
5
"
1
3
5
4
3 *
4
u
451"
5
3
4
4" 3
55
4
33
u5 5
" 4 " 3
5
4
*3" 553
"15
u
3 5
4
3
u3 4 3
1
5
4
u
5
4
u
5
M
3
*
1
5
5
4
3
4
3
1
3
1 43 u 41*4 3
5
4 3
3
G 5 55
4

5
5
5b
4
b
1 b
4
43
15
4 41 5
"3
5 3 341
4 u 3
5
4
41
*35
1 3
b
33
u3
" u
4
1
1
5
4
54
5" 4
1
4
1
*4
55
4
I*
4
E
*5
G
54

5
4
3
4
5
E
5
4

4
u
5
u
5
4
u
G

5
1
4
5
4
b3
5
4 5G
1
34 4
1u
5
4
431
1
3
"
3
5
G
4
b
1
4
u
"
5
4
4
*
5
4
5
5
G
*
5
G
4
b
4
4
u
5
34
54
u
*5
E
u 4
415
1
5 514
G
4
u
4b
55
54
5
G
u33 3
153 3
5 "5
4
55"3

5
3
4
*53
G
534

333
55
4
51 u
334
5
E5
4
5G
5 5
1
"
*43
4
4
1
"4
u
4 35
54
1
5
4
54
54
**u
3 u
u 4
4
G4
4
u
5
5G*1
4
5
555 *
4
3u
555
5
4 331
54
55u
54G
E
5
u
3 3u
4
E
4
"
5
4
3
4
5
b 4
5
4
5
5
4
3
"
5
5
*
5
u
3
34
5 4" 4
u
1 15
3
5
4
E5"5
43
5
4
5
4
5
4
G
33
33
"
1
5
3 415
3
4
55
G

4
5
4
G
5
3
4
5
5
4
b
5
34
5
4
54
4
3u
5
3 34
5
4
3
4
1
555
4
31
1
*1G

5
55O
4
54
1u
*15
3 34
455 u
3
5
3 135
G
GI5 4
4
II
4
u
G 5
4
45
4 5
4

4
u3
4
5
G *54 4
4
5
3 4
*
1
*
u
5
5
*

4
5
4
5
G
G
5
*5
4 33 u13 334
5G
1 "4
5
4
*1
5 5

4
3
1 *4
4
5 34
1
53

K
5
4
1
*
4
5 "5
" *4
b" b
5 3
4
G3
4
u
u5
4
G
51
4
43
5
*
4
*45
4

5
G
b
1
4
5
3
5
4
b
"
4*u
G
5
4
u
53u5 4

u
5
3
b
1
u
4
5
3
5
u
4
G
5
5
4
"
5
3
E
5
4

b
5
* E
1
3 54
55
5
3
4
u
5
G
3
54 5
3
45
4
5
G
GE
G
43
3
u
5
*
35
"3 " *u
4
5
4
G
5
3 1
4
G5
3 5"4
35
* 5
5 5"5 5 5
4
u
53
E
4
"
1u
5 54
5 b
5 5
3
4
*53
4
1
u
5
G
4
*u 3 54
5
3
G
3
4
5
5
4
5
"

G
5
4
b 4
3
1 4
G
*O
14
3
5b
3
545u
3
"
5 5
G
u3
5
5 G
5
5
555 4
4
5
4
43
5 3 u
51
*b
5u
4
5 G4
4
5 G
4
5
*
4 5IE4
3b
53
4
G
ub
4
54
4
*1515
"4 u
5
4
5
1
5u
55
54
4
3 34 4
5
5
451 5
4 5
5
1 4
4
b5"E
15
3 54
4
u
5
1
5 5
55
5
4
u
5
511
55 * "
4
u
4
O
"u
4
bG 15
4
5
35
*1 u
4

30+

00
+

25

90

20

10

15

80

10

70

60

0
0

200

20

Extreme

50

40

High

60

Low
Medium

40

80

250

100

300

30

Extreme

High

20

120

Low Medium

10

N u m b e r o f W e lls w ith D L P
Sc o re

140

Figure 13: Zama Field SLP Score distribution.

N u m b e r o f W e lls w ith SL P
Sc o re

Figure 12: Zama Field DLP Score distribution.

3
4
b
4 u
3
u
4

u
4
4 3
3
4
4
u
4
u
4 3
1
u
4
u3 13

Figure 15: Location of Zama wells with extreme SLP


shown in red against all field wells in black.
The bold green grid indicates 6x6 miles/square.

3u

4
u1
u 4
4
u

3
3

u
1
3

4
u 3
b
4
"
4 u
u
34 "
u
4
3
3
"
3
1 u
4
"
"4
55 " "
1
3
4
5
4 5 3
4
5
" 433 " " 1
5b
4
u
4
5
5"
4 3
u 4
3
3 u
3
55 5
4
3
3
44
3 4
5
4
5
b5 4
3 5
4
5* 3 53 3 33b
uu
5""
4
u4 3
"5
1
3
"
4
5 5
4
u
555
4
3 *
4
4" 3
u
4
1
4
4
" 3 3
33 u
55 5
u
"
515
4
*3" 5 3
u u
43
3
5
4
u 31 4 3 *44 3
15 " u
51M
55
4
3
4* 5
1
3
3
5 b
4
3
4 3
1
43 u 41 3
5
G
4
4
5b
4
b
5
3
5
3
5
5
1
4
3
4
1
5 4
4
43 4
1
u
"3
5
4 u b
5
4
5 *3
14
1 3
33 3 41
" u
1
5
u3
5
u3 13
4
5 1
5" 4
1
*
4
1
*4
I*
4
5
4
E4
5 4

5G
*
5 5u
E
354
5
54
5
u
u
G
541 b
5" 4
G
54
4
5
4
u
34 44 1
43
"
33
5b
3u
5G
G
4
*4
4
5
4
4 1
5u
54 1
5
4
5b
4
5 *1
4
3 13 3G
5
E
*
5
G
4
u
5
u u
4
5 14
4
55
5
4
1
5
G
54
54
5 "34

4
43
u
4 ""5
b
3153 3 1
5534
G4
*5
3

54 u
5 5
4
5 4
E
*535
5
4
333
5 4
1
"
"5
4
1
5 3
3
u
5
45
1
5
4
553
54
5 1**u
54
54u
3 u
5Gu
5 4
G
4
5
5G
4
*1 335
3u
5
4
4 4
u
E
4
55
4G u
*E3
"*55
4555553
3
3 3 u
53
4 1
5
54
4
" 55
b 4
3
5"5
u
14
4
4
5
4
3
4 "u
E
3
55"31
5G
14
5 "3 3
4
3
5
5 4
4
3
4
4
5
5
G

1
55 534
4 5O
5
45
G
3
445
4
b
5
5
5
4
54
4
3u
5
1
3 34
3
4
55
4
1
554
4
*11G

4
5u
*1u
u
3 331
555
4
5I5 4
G
4
3 135
5
G
G 3
4
I5I
4
4 5

u34 4
*54
5
4
4 *44
5
G
4
34
*
u
5
1

5
*

4
5G
4
5
5
"
"4
G
4
5G
*5
4* 3 3 u13 33 4
1
1
5
4
5
3
5 *4
1
4
5 34
*

55
K
5
4
1
*1
4
" 4
5 "5
b" b
5 3
4
4 3
u
G
G3
u
"4
51
4
43 b 4
*45
4

5
*
4 3
3b
5
51uu
*
4
" 5 G
4 G
5u
4
5

u
4
5
5
bG
1 u
54
4
5 3
5u
4 5u
"
33
4

5
3
E
5
4
*
5
1
5
b
5 E
5
3
E
4
u
5
54
3
54
5
G
3
4
G G

G 44 5
5
55
*
43
3
u
35
"3 " *u
5
4
G
3 5"4
3 1
4
5
5
55
5 55
4
u
3G
53
* 5
E
4
15
"
u 54
35 5
5 "5
3
4 *5
1
*u 3 54
5 5G
G
4
u
5 4
4
b
3
5
G "5
3
4

5
G
1
G
*
4
1
3
O
b
3
5
3
u
5
3
4
51
"
5u
u
4
4
3
5 5
G
b

5
4
5
43
5
5 3
555 4
4
54 u
*b
5u
4
5 G4
4
5 G
4
*
5 4
4
3b
*1
55
G
55
4
" 4
4
ub
4
54
4
5 55
IE
4
3
1
u
5u
54
3 34 4
5
4
51 5
4 5
1 4
4
15
b5u
5
4 3 555
"E
u1 5 4
54
5
15 1
5
5

"
5 54
5
u
5
4
u
G 1"5
4
u 5
4
O
5
3
b
*
*1 5
4
3u

4
1
4

u
4

Within this analysis the wells in the Zama Field show significant potential for leakage, both shallow, which is
independent of fluid type, and deep, which may occur in the event that CO2 or acid gas is introduced to the historical
producing formations.
Previous work identified that approximately 6.1% of all cased wells in the province had shallow leakage in the form of
surface casing vent flow (annular pressure) reported (Bachu and Watson 2006). Based on the analysis of the Zama Field,
18% of the wells have extreme potential of shallow leakage. This result is expected since the reported incidence of surface
casing vent flow, determined from ERCB data, within the Zama Field is 30% of cased wells, well above the provincial
average.
The potential for deep leakage shows that 28% of the wells in the Zama Field have an extreme deep leakage potential.
This assessment for deep leakage potential is based on:
1. Approximately 30% (based on known and extrapolated data) of the wells have cement in the deeper regions of
the wellbore that may be susceptible to CO2 or acid gas attack because of the additives used.
2. Eighty percent of wells are stimulated by acidizing, with 60% of the wells having multiple acid stimulations.
These stimulations may reduce the near wellbore seal, thus affecting isolation, and may reduce the time for CO2
or acid gas breakthrough to the wellbore.
3. Fifty-five percent of the wells have multiple completions, increasing the potential for leakage between zones.
4. Thirteen percent of wells have been abandoned with a bridge plug. The remainder of the wells have not yet been
abandoned. The assigned zonal abandonment value of two for wells that have not yet been abandoned may be
too high. In the future, if zonal abandonment practices change to some other, more robust system the assigned
value for zonal abandonment could be reduced to one.

SPE 112924

Conclusion
The development of a computer-based tool to compile and mine available regulatory data to determine the potential of a well
to leak has created the ability to conduct a first-pass evaluation of large numbers of wellbores to determine their relative
potential for leakage, both shallow and deep. The use of this tool will speed up the wellbore assessment process by zeroing in
on potential problem wells which may require closer scrutiny prior to the implementation of a CO2 EOR, acid gas disposal or
CO2 sequestration scheme. The tool can also be used in risk assessment analysis if potential consequences of leakage are
overlain, such as population density, ground water information or the presence of H2S along the wellbore. The tool need not
be specifically used to determine leakage in the presence of acid gas or CO2, but can also be used to evaluate the potential for
other types of leakage such as annular pressure (surface casing vent flow, aka sustained casing pressure) or gas migration.
Based on the findings of this study, careful well evaluation should be conducted to determine if wells in a particular CO2
EOR or sequestration scheme are at risk for leakage. More study is required to determine the durability of cement and bridge
plugs to improve the confidence of this evaluation.
Acknowledgements
The authors would like to thank Django Dunn (Silent D Software) for the database programming.
References
Alberta Environment 1995. ST-55: Albertas Useable Groundwater Base of Groundwater Protection Information. Alberta Environment,
Edmonton, Canada.
Bachu, S., Watson, T.L., 2006. Possible Indicators for CO2 Leakage along Wells. Paper presented at the 8th International Conference on
Greenhouse Gas Control Technologies, June 19-22, Trondheim, Norway. CD ROM, Elsevier..
Browning, D. 1984. CO2 Corrosion in the Anadarko Basin. Paper SPE 12608, Presented at SPE Deep Drilling and Production Symposium,
Amarillo, USA 1-3 April.
Bruckdorfer, R.A. 1986. Carbon Dioxide Corrosion in Oilwell Cements. Paper SPE 15176, Presented at the Rocky Mountain Regional
Meeting of the Society of Petroleum Engineers, Billings, USA, 19-21 May.
Caswell, D., 1988. External Casing Corrosion Survey. Report prepared for the National Association of Corrosion Engineers. University of
Calgary. Calgary, Canada.
Duguid, A., Radonjic, M., Braunt, R., Mandecki, T., Scherer, G., Celia, M., 2005. The Effect of CO2 Sequestration on Oil Well Cements.
Proceedings of the 7th International Conference on Greenhouse Gas Control Technologies, September 5-9, 2004, Vancouver, Canada.
Wilson, M., Morris, T., Gale, J.J., and Thambimuthu, K., eds., Elsevier, Vol. II Part 2:1997-2001.
ERCB, 1990. Directive 009: Guide to Minimum Requirements for Cementing Intermediate or Production Casing, Multiple Editions.
Energy Resources Conservation Board, Calgary, Canada.
ERCB, 2007. Directive 20: Well Abandonment Guide. Energy Resources Conservation Board, Calgary, Canada.
Krilov, Z., Loncaric, B., Miksa, Z. 2000. Investigation of a Long Term Cement Deterioration Under a High-Temperature, Sour Gas
downhole Environment. Paper SPE 58771, Presented at the International Symposium on Formation Damage Control, Lafayette, USA, 2324 February.
Kutchko, B.G., Strazisar, B.R., Dzombak, D. A., Lowry, G.V., Thaulow, N., 2007a. Degradation of Well Cement by CO2 under Geologic
Sequestration Conditions. Environmental Science Technology, 41(12):4787-4792.
Kutchko, B.G., Strazisar, B.R., Dzombak, D. A., Lowry, G.V., Thaulow, N., 2007b. Degradation of Wellbore Cements: Results of Long
Term Experiments and Effect of Additives. Presented at the 3rd Wellbore Integrity Network Meeting of the IEA Greenhouse Gas R&D
Programme. Available at, www.co2captureandstorage.info/docs/wellbore/3rdWellboreMtg_FinalReportSummary.pdf. Santa Fe, USA, 1213 March, 2007.
Marsters, S. 2007. Enhanced Oil Recovery, Dont Fight Mother Nature, New Technology Magazine, June: 13-16.
Nelson, E. B., Guillot, D. 2006a. Well Cementing, second edition, 244. Schlumberger, Sugar Land, USA.
Nelson, E. B., Guillot, D. 2006b. Well Cementing, second edition, 58. Schlumberger, Sugar Land, USA.
Onan, D. 1984. Effects of Supercritical Carbon Dioxide on Well Cements. Paper SPE 12593, Presented at the Permian Basin Oil and Gas
Recovery Conference, Midland, USA, 8-9 March.
Schremp, F., Roberson, G., 1975. Effect of Supercritical Coarbon Dioxide (CO2) on Construction Materials. Paper SPE 4667, Presented at
the SPE-AIME 48th Annual Fall Meeting, Las Vegas, USA, 30 September- 3 October.

10

SPE 112924

Watson, T.L., Getzlaf, D., Griffith, J.E., 2002. Specialized Cement Design and Placement Procedures Prove Sucessful for Mitigating
Casing Vent Flows-Case Histories. Paper SPE 76333, Presented at the SPE Gas Technology Symposium, Calgary, Canada, 30 April-2
May.
Watson, T.L., Bachu, S. 2007. Evaluation of the Potential for Gas and CO2 Leakage Along Wellbores; Paper SPE 106817, Presented at the
SPE E&P Environmental and Safety Conference, Galveston, USA, 57 March.
White, W.S., Calvert, D.G., Barker, J.M., 1992. A Laboratory Study of Cement and Resin Plugs Placed With Thru-Tubing Dump Bailers.
Paper SPE 24574 presented at the SPE Annual Technical Conference and Exhibition, Washington, USA, 4-7 October.