SPE 102498

Stressed Shale Drilling Strategy—Water Activity Design Improves Drilling Performance

J.C. Rojas, SPE, BP; D.E. Clark, SPE, Baker Hughes Drilling Fluids; and J. Zhang, SPE, Baker Atlas/Baker Hughes
Drilling Fluids

Copyright 2006, Society of Petroleum Engineers

Introduction
This paper was prepared for presentation at the 2006 SPE Annual Technical Conference and It is generally accepted that non-aqueous drilling fluids are
Exhibition held in San Antonio, Texas, U.S.A., 24–27 September 2006.
superior to water-based muds in improving wellbore stability
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, as
in shale formations [1-5]. However, some field cases have
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to demonstrated that wellbore instability problems still occur
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 even when oil- or synthetic-based fluids are used, especially in
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
fractured formations [6, 7].
for commercial purposes 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 It is well-known that water fluxes into or out of the shale
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.
formation during drilling is a key factor in controlling
wellbore stability. Studies showed that water absorption by
Abstract shale formation altered the stress distribution, reduced the
Non-aqueous drilling fluids are often chosen to drill strength, and at the same time changed the Young’s modulus
troublesome shale formations in an effort to minimize of the near wellbore formation, which may potentially
wellbore instability problems. However, Gulf of Mexico destabilize the wellbore [3, 8, 9]. On the other hand, dehydration
(GoM) experience has indicated that when drilling in highly causing pore pressure decrease and strength increase is
faulted areas, oil- and synthetic-based fluids do not always beneficial to wellbore stability [3, 10-12]. However, over-
prevent wellbore destabilization. This is evidenced by dehydration results in fractures in the near wellbore formation
wellbore collapse, and the resulting difficulty with hole- and may also disturb wellbore stability [13-15]. Kelly argued that
cleaning, tripping, logging and casing running. water desorption in fractured shales may widen the fractures
and destabilize the wellbore. He suggested that slight water
It is known that the chemical, physical, and mechanical adsorption may, in fact, soften the fracture surface and is
effects resulting from the interaction between the drilling fluid necessary to stabilize wellbore [16, 17].
and the formation may degrade the stability of the borehole in
the already weakened and stressed fault interval. Commonly, In fractured stressed-shale formations, it is particularly
the practice has been to increase the drilling fluid salt content important to control water movement because the in-situ
to enhance the borehole stability. The perception that low stresses are in a critical state. Any disturbance of the formation
drilling fluid water activity is beneficial to wellbore stability is by chemical and/or mechanical means could result in shale
one that lends itself to a needed revision. breaking and sliding into the hole. Once the wellbore
instability is initiated, it becomes difficult to stop [16, 18].
A detailed laboratory investigation using preserved shale Santarelli (1992) and Labenski (2003) discussed how to
core and drilling information have confirmed that the water improve wellbore stability in fractured formations by keeping
activity of drilling fluids is often much lower than necessary. mechanical balance [6, 8]. In this paper, we will focus on
This study has shown that when drilling faulted or fractured maintaining chemical balance between shale formations and
shale, the correct, not higher salt content in drilling fluids will non-aqueous drilling fluids to minimize wellbore instability
reduce wellbore collapse problems and improve drilling problems.
performance.
It is very challenging to determine the chemical potential
A laboratory method, which allows the quantitative balance between shale formations and drilling fluids due to the
measurement of water and ion movement during shale/mud difficulty in directly measuring the chemical potential.
interactions, combined with geological information, optimizes Generally, we balance the water activity of the non-aqueous
the salinity design of drilling fluid, which controls water and fluids with the water activity of the formation to control water
ion movement. Laboratory data and field cases from GoM movement so as to improve wellbore stability [1, 10]. Although
drilling support the concept of optimum salinity to enhance it is difficult to balance water activity of shale formations with
borehole stability in naturally fractured formations as part of drilling fluids exactly everywhere in a well, because shale
the stressed shale drilling strategy to improve drilling water activity varies with depth and mineralogy [19], we try to
performance.
2 SPE 102498

balance the water activities of shale formations and drilling where σv, σv', Pp are total overburden stress, effective
fluids at the problematic locations. overburden stress, and pore pressure respectively. Based on
Mondshine’s theory, the water activity of shale formations can
In the following discussion, we will first review how to be calculated as
predict and measure the in-situ water activity of both shale ⎡V ⎤
formations and non-aqueous drilling fluids. It is found that a w , shale = exp⎢ w × σ′v ⎥ (3)
water activity is both temperature and pressure dependent. ⎣ RT ⎦
Experimental studies on determining water and ion movement where aw, shale is water activity of shale formations; and Vw is
during shale/mud interaction will be presented and the molar volume of water.
application of the results to determine the water activity of
non-aqueous drilling fluids will be examined. Finally, case Method of effective mean stress. It is generally accepted
studies will be demonstrated in support of the arguments and that in addition to the overburden stress, horizontal stresses
recommendations. The main results and conclusions from this also contribute to the dehydration of shale formation during
study will be given. compaction [13]. Mody and Hale (1993) suggested that the
effective mean stress, not the effective overburden stress,
In-situ Water Activity of Shale Formation and Non- causes water to be expelled from shale formation. The
aqueous Drilling Fluids effective mean stress is defined as [13]:
One of the fundamental driving forces for the movement of
water into or out of shale formations is the chemical potential σ v + σ H + σ h − 3Pp
difference between shale formations and drilling fluids [4, 13]. It σ m' = (4)
3
is difficult to directly measure the chemical potential of a
system; however, the chemical potential of the water phase
where σH, σh are total maximum and minimum horizontal
can be closely estimated through its water activity. By
stresses. Based on Mody and Hale’s theory, the water activity
definition, the water activity is related to the chemical
of the shale formation can be calculated as [13]:
potential by the equation:

μ i = μ i0 + RT ln(ai ) ⎡V ⎤
(1) a w, shale = exp⎢ w × σ m' ⎥ (5)
⎣ RT ⎦
where μ i is chemical potential of ith component; μ i0 is the
where σ’m is the effective mean stress.
chemical potential of the pure liquid of that component at
standard condition; a i is the water activity of that component; Modified method. Compared with the method of effective
R is gas constant; and T is absolute temperature. overburden stress, the method of effective mean stress is
improved because it considered the effects of horizontal
Water activity is an important parameter in controlling stresses on water expulsion. However, at least two issues are
shale/mud interactions. In order to optimize the salinity of the still not considered in the method of effective mean stress.
non-aqueous fluids, we need to be familiar with the water First, it did not take into account the deposition environment.
activity of both shale formations and drilling fluids firstly. From Equation (5), we can see that the water activity of the
outcrop shales or the shales at sea bottom is always 1 (fresh
Water activity of shale formations water) because the effective mean stress at surface or at the
sea bottom is zero. In fact, in the marine environment, the
Water activity of shale formations is an excellent indicator water activity at surface or sea bottom is close to the water
of the shale’s state of hydration and its potential to absorb or activity of sea water (about 0.98 aw), not 1.0 aw of fresh water.
desorb water. It is affected by changes in pressure, Secondly, it ignored the salinity alteration with depth.
temperature, mineralogy, c-spacing, pore fluid composition, Engelhardr and Gaida (1963) argued that the pore fluids differ
etc. The methods to predict and to measure water activity of greatly from sea water both in concentration and composition.
shale formations will be discussed as follows. For example, the salinity changes with depth from the St.
Genevieve Limestone are shown in Figure 1 [21]. It is easily
1) Shale water activity prediction seen that the salinity increases with an increase in depth.
After considering these two issues, the actual water
Method of effective overburden stress. Mondshine activity should be less than the predicted value using
(1969) did a pioneering work on determining the salinity of Equation (5). This agrees very well with the data from Mody
oil-based muds [20]. He suggested water-absorption force of and Hale’s paper, which is shown in Table 1 [13].
shale formations to be equal to the effective compact force
that expelled water from the shale. The effective compaction
force is the difference between the total overburden stress and
the pore pressure, which can be expressed as

σ v′ = σ v − Pp (2)
SPE 102498 3

0
Based on the uniaxial strain boundary condition, the
500 minimum horizontal stress (σh) is commonly estimated using
the poro-elastic equation, which is expressed as [22]:
1000 S
e ν
1500 a σh = (σ v - αPp ) + αPp (8)
(1 − ν )
Depth, ft

w
2000 a
t where υ is Poisson’s ratio; and α is Biot’s constant. After the
2500 e minimum horizontal stress was calculated by using the above
r equation, it is calibrated with the results from leak-off or
3000
extended leak-off tests to compensate the effects of lateral
3500 tectonic compression, temperature variation, and subsequent
relaxation [22].
4000
0 2 4 6 8 10 12 14 16 The maximum horizontal stress (σH) is determined from
breakdown pressures obtained from leak-off tests (where
Salinity, % available) and through a calibration process. The maximum
horizontal stress is first estimated from the breakdown
Figure 1 – Changes of salinity with depth from the St. Genevieve pressure using following equation (neglecting the tensile
Limestone (after Engelhardr and Gaida, 1963). strength):

Table 1 – Comparison between measured and predicted σ H = 3σ h − Pb − αPp (9)

shale water activity for four different deepwater GoM
shales (after Mody and Hale, 1993) [13]. where Pb is the breakdown pressure. The resulting ratio of the
Sample Average Predicted aw maximum to minimum horizontal stress at the leak-off test
experimental aw (Equation (5)) depth is used to multiply the entire known σh curve to
A 0.87±0.02 0.87 determine the maximum horizontal stress.
B 0.92±0.02 0.96
C 0.92±0.02 0.95 It is also important to note that another reason can be
D 0.93±0.02 0.95 applied to explain why the measured water activity is
generally lower than the predicted one, as shown in Table 1.
It is seen from Table 1 that the predicted water activity is This is due to the limitation of the techniques used to measure
normally higher than the experimental measured value, just as water activity of shale formations, which will be discussed
we discussed previously. Therefore, we suggest that the later.
method of effective mean stress can be modified as:
2) Shale water activity measurement
In the laboratory, water activity of shale is usually
⎡V ⎤
a w, shale = a w, c ⋅ exp⎢ w × σ m' ⎥ (6) measured using hygrometer or adsorption isotherm [1, 23, 24].
⎣ RT ⎦ Recently, an improved method using chilled mirror
technology was developed to measure water activity of shales
[25]
In the above equation, a w , c ( a w , c ≤ 1 ) is a water activity . All the above methods measure water activity at ambient
correction coefficient. It is affected by the deposition pressure and temperature. However, it was found that the
environment, depth, mineralogy, clay content, cation exchange water activity of shales changes with pressure and temperature
[10, 26]
capacity etc. For example, the activity correction factor in . Part of the laboratory results from Refernce 26 is shown
marine deposition environment is less than that in fresh water in Figure 2.
environment.
It is seen from Figure 2 that water activity increases with
It is seen from Equations (2), (3), and (4), the three in-situ an increase in confining pressure. Furthermore, it was found
stresses: overburden, maximum and minimum horizontal that the water activity of shale formations also increased with
stresses are required for predicting the water activity of increasing temperature [26]. This means that the water activity
formation. The total overburden stress can be obtained by should be higher if it is measured at high temperature and high
integrating the bulk density log: pressure than that measured at ambient temperature and
z pressure conditions. Therefore, in order to calibrate the
σ v = ∫ ρ b gdz (7) predicted water activity using the surface measured value, we
0
need to consider the pressure and temperature effects.
where ρb, g and z are bulk density, gravitational acceleration,
and vertical depth respectively.
4 SPE 102498

0.95
a w,mud = C1 exp(C 2T ) (10)
0.9

0.85 where aw, mud is water activity of drilling fluids; T is

Shale Water Activity

temperature in oC; C1 and C2 are constants.

0.8

0.75 Based on the data from Reference 10, it was found that the
constants C1 and C2 in Equation (10) strongly depend on the
0.7 concentration of CaCl2 solutions. The relations can be
expressed as:
0.65

0.6 Gum bo, 4100 ft C1 = −0.0153C + 1.1211 (11)

Speeton, 5400 ft
0.55 Cretaceous, 8000 ft
Kim m eridge, 12500 ft C 2 = 8.451 × 10 −5 exp(0.07114692 ⋅ C ) (12)
0.5
0 1000 2000 3000 4000 5000 6000 7000 8000 where C is the concentration of CaCl2 solution in the internal
Confining Pressure, psi
phase of oil-based muds. It is important to note that the above
two equations are only applicable in certain range of the
concentrations of CaCl2 solutions.
Figure 2 - Effects of confining pressure on water activity
(After Fonseca and Chenevert 1998).
The constants C1 and C2 can be determined using
Equations (11) and (12). Then the change of water activity of
Water activity of drilling fluids drilling fluids with temperature can be calculated using
Equation (10).
After the determination of the water activity of shale
formations, it is also required to obtain the water activity of In addition to temperature, the water activity of non-
drilling fluids so as to balance these two water activities to aqueous drilling fluids is also influenced by pressure.
minimize wellbore instability problems. As with the water However, pressure effects are ignored in this paper due to lack
activity of shale formations, the water activity of drilling of proper data.
fluids is also temperature and pressure dependent [10, 25].
Chenevert and Strassner (1975) found that the water activity After the in-situ water activities for both muds and
of the non-aqueous muds increase with an increase in formation are determined, the osmotic pressure can be
temperature [10]. Some of their results are shown in Figure 3. calculated as;

1 RT ⎛⎜ a w,shale ⎞⎟
Pπ = −α m ln (13)
10 wt% CaCl 2 Vw ⎜⎝ a w,mud ⎟⎠
0.9
20 wt% CaCl 2 where Pπ, αm, are osmotic pressure and membrane efficiency
0.8 respectively.
Water Activity

30 wt% CaCl2
Experimental Study
0.7
Gravimetric-Swelling Test (GST) was run for two shales to
quantitatively determine water and ion movement when a
0.6 shale sample interacts with a solution [27]. During GST, a
40 wt% CaCl 2
sample with size of 0.5″x0.75″x1″ was placed in a small
plastic bag and was positioned between the movable anvil and
0.5
stationary anvil of a swelling transducer. A 50 ml solution was
then poured into the plastic bag and the bag was sealed
0.4 immediately. The swelling or shrinkage was recorded
0 20 40 60 80 100 continuously until equilibrium was reached. In this test, it
Temperature, C
o
seems that the equilibrium took about 48 hours.

Figure 3 - Effects of temperature on water activity of Shales and test fluids

CaCl2 brines. Two preserved shales, outcrop Pierre II and a core from a
GoM deepwater well, were used in this study. The GoM shale
Based on the definition of the water activity (Equation was cored from a depth of 17,055.1 to 17,055.9 ft. It has
(1)), it is assumed that the relationship between water activity
and temperature can be expressed as:
SPE 102498 5

fractures and cracks extended in different directions; some Results and analysis
fractures are intersected, just as the stressed shale exhibits [7]. The effects of water activities of CaCl2 solutions and
SBMs on water and ion movement when they were interacted
The preserved GoM shale core was removed from its wax- with Pierre II and GoM shales are discussed as following.
sealed protective metal sampling. The complete shale core
was then submersed in mineral oil to prevent dehydration prior 1) Effects of water activity on water/ion movement for
to collecting the actual shale samples for water activity Pierre II shale
measurement.
The effect of water activity on water/ion movement for
Using oriven shale preservation techniques, an Pierre II exposed to CaCl2 solutions for 48 hours is shown in
approximately one inch thick slice was cut from the core at Figure 4. It is seen that when the water activity decreases,
sampled from 17,055.30 ft. with a chop saw. Three samples water is removed due to osmotic effects and ions are added
were then manually collected from the center of the slice to because of ionic diffusion.
ensure undisturbed representative shale samples. Sample #1
was collected near the surface of the core, sample #2 was
collected at an intermediate depth between the core surface 5.0
and the core center, and sample #3 was collected at the center
of the core. Each shale sample was placed in an Aqua Lab 4.0
disposable sample cup with lid in preparation for water Ions

Water/Ion Movement, wt %
activity measurement [25]. The water activity measurements 3.0
indicated the water activity of GoM shale to be about 0.82.
The mineralogical composition, water content and water 2.0
activity of the shales are listed in Table 2.
1.0
Table 2 - Mineralogical composition, water content, and Net
0.0
water activity of Pierre II, GoM shale, and Gumbo shale
from Reference 26.
[26] -1.0
Pierre II GoM shale Gumbo shale
Constituent ( wt% ) (wt %) Water
(wt %) -2.0
Quartz 15-20 15-20 20.8
1-2 5.4 -3.0
Feldspar
0.75 0.80 0.85 0.90 0.95 1.00
Chlorite - - 1.5
Kaolinite 5-10 10-15 10.7 Water Activity of CaCl2 Brine
Illite 20-25 25-30 26.2
Clay Smectite - 2.5 Figure 4 - Effect of water activity on water/ion movement
35-40 26.2 when Pierre II shale was immersed into CaCl2 brines.
Mixed layer 50-55
Total 75-90 70-85 67.1 As shown in Figure 4, the net transport including both
Expandable Layers water and ion, increases with decreasing CaCl2 water activity.
70-80 > 95 -
in Mixed Layer This result agrees with immersion test by Santos et al. (1997),
Water Content 17.7 8.25 12 who observed that the weight increase for shale in a salt
solution is higher than that in fresh water [28]. However, they
Water activity @ 0.98
atm conditions
0.82 0.86 accredited the weight change only to water movement and
probably reached the wrong conclusion: “Even with a very
low activity, the calcium chloride solution was not able to
CaCl2 brine and synthetic-based invert muds (SBMs) were
drive water from the shale”. When measurements of osmosis
used for testing. The SBMs consisted of synthetic-base oil
and diffusion are considered, there can be a net weight gain,
(16-18 IO), emulsifier and brine. Their compositions and
even when the shale is dehydrated. For example, when the
equivalent water activity for the internal brine phase at room
Pierre II shale was immersed into 0.85 aw CaCl2 solution for
temperature are listed in Table 3.
48 hours, the net weight gain was 1.6 wt%; while it lost 1.2
Table 3 - Composition of SBMs. wt% water.
Emulsifier SBR Organic clay Wt % of Water
CaCl2 activity
Figure 5 shows the effects of water activity of SBMs on
(lb/bbl) (lb/bbl) water/ion movement of Pierre II shale. It is seen that when the
6 80/20 2 0 1 water activity of the SBM is low enough (less than 0.98 aw),
6 80/20 2 10 0.95 water was removed from Pierre II shale. This perfectly
6 80/20 2 18.8 0.85
matches “balanced-activity theory” because the water activity
of Pierre II is also 0.98 [1].
6 80/20 2 24.8 0.75
6 SPE 102498

0.3 6

0.0 5
Water/Ion Movement, wt %

Water /Ion Movement, wt%

Net
Ions 4
-0.3
3
-0.6
Net Water
2
-0.9
1
Water Ions
-1.2
0

-1.5 -1
0.75 0.80 0.85 0.90 0.95 1.00 0.75 0.80 0.85 0.90 0.95 1.00
Water Activity of SBM Water Activity of CaCl2 Solution

Figure 5 - Effect of CaCl2 water activity on water and ion

Figure 6 - Effect of water activity on water/ion movement
movement when Pierre II shale immersed into SBM
when GoM shale immersed into CaCl2 brines.
(SWR: 80/20).
The effects of water activity of SBMs on water/ion
Compared with exposure to the CaCl2 solution (Figure 4),
movement for GoM shale are shown in Figure 7. It is shown
Figure 5 also exhibits that the amount of ion movement is
that water began to be removed when the CaCl2 water activity
very small when the Pierre II shale was immersed into SBMs,
was decreased to about 0.85 aw. This result perfectly matches
even at very low water activity (0.75 aw). For example, when
the results from fracture generation test [7]. Figure 8 shows the
Pierre II shale was immersed into 0.75 aw CaCl2 solution for
effects of water activity of SBMs on the generation of
48 hours (Figure 4), it gained 4.4 wt% ions. However, when it
fractures in GoM shale cores. It is easily observed that
was exposed the SBM with the same water activity (Figure 5),
fractures began to occur as the shale was exposed to fluids
it only gained 0.1 wt% ions. This clearly demonstrates that the
with internal phase salinity at 20 wt% (0.84 aw). Numerous
SBMs used in our tests act as a near-perfect semi-permeable
cracks were generated after the salinity was increased to 25
membrane, which allows the movement of water and restricts
wt% CaCl2 (0.75 aw) due to dehydration of the shale, as shown
the movement of ions.
in Figure 7.
2) Effects of water activity on water/ion movement of
GoM shale 0.5

Figure 6 demonstrates the effects of water activity of 0.4

CaCl2 brines on water/ion movement of GoM shale. It is seen
Water/Ion Movement, wt%

that water influx decreases, while ion adsorption increases 0.3

with decreasing water activity of CaCl2 solutions. Different Water
from the results of Pierre II shale, GoM shale still absorbs 0.2
water, even when the water activity of the water-based CaCl2
solution (0.75 aw) is lower than that for the GoM shale (0.82 0.1
aw). This may be caused by the de-saturation of the GoM shale Ions
0.0
during coring and processing [29]. The de-saturation caused
capillary suction when it immersed into water-based solutions -0.1
[9, 28, 30]
.
-0.2
Net
-0.3
0.75 0.80 0.85 0.90 0.95 1.00
Water Activity of SBM

Figure 7- Effect of water activity of SBMs on water/ion

movement for GoM shale.
SPE 102498 7

Stress, psi
0 2000 4000 6000 8000
3000

4000
Total overburden stress
5000

6000

TVD, ft
7000
Pore pressure
8000

9000
Figure 8 – Gulf of Mexico preserved core pieces immersed Total minimum horizontal stress
in synthetic-based drilling fluid exhibit increasing 10000
shrinkage and cracking as the fluid salinity is increased
(after Rajos et al., 2006) [7]. Figure 9- Pore pressure, overburden and minimum
horizontal stress profiles of Well A.
In summary, two important phenomena can be observed
from the above experimental tests. First, the non-aqueous mud
can indeed act as a semi-permeable membrane. Therefore, we Water Activity
assume the membrane efficiency of non-aqueous fluids to be 1 0.88 0.9 0.92 0.94 0.96 0.98
in the following analyses. Secondly, water desorption of 3000
formation may cause fractures and reduce wellbore stability. Effective overburden
4000 stress method
Case Study
5000
Well A
TVD, ft

6000
Well A is located in the Mississippi Canyon of GoM in
1,030 feet of water. It was sidetracked at 12,150 ft. Effective mean
7000
(MD)/6,771 ft. (TVD) to a total of 18,263 ft. (MD)/9,329 ft. stress method
(TVD).
8000
The in-situ stress profiles for Well A are shown in Figure
9. Applying these results and using Equations (3)-(6), we 9000
obtain the water activity profiles from different methods, Modified method
which are shown in Figure 10. Note that the water activity 10000
correction coefficient (aw, c) in our analysis is selected to be
0.98. This parameter can be determined using the results of Figure 10 - Water activity profiles of Well A using
water activity measured at in-situ stress and temperature different methods.
conditions [26].

It is seen from Figure 10 that the highest predicted water Based on the comparison between GoM shale used in this
activity is the one that used the method of effective mean study and Gumbo shale in Reference 26 on deposition
stress and the lowest one used the method of effective environment and mineralogical composition, as shown in
overburden stress. The predicted water activity from our Table 2, it is reasonable to assume that the GoM shale has
modified method is between the results from the above two similar water activity response on confining pressure as the
methods. Gumbo shale in Reference 26. If this is true, then the effects of
confining pressure on water activity of GoM shale can be
expressed as:

a w , shale = 0.82 + 3 × 10 −5 Pc (14)

From Figure 9, we find that the confining pressure (σh-Pp)

at 17,055 ft. (MD)/8,824 ft. (TVD) is about 1,950 psi.
8 SPE 102498

Therefore, the water activity was corrected to be aw = 0.88

using Equation (14). In addition to pressure, water activity Stress, psi
0 1000 2000 3000 4000 5000 6000
also increased with an increase in temperature [26].
Unfortunately, we do not have enough data for the 3000
temperature correction. We believe that the in-situ water
activity is very close to the predicted water activity (aw = 0.9) 4000
after corrected for temperature.
5000
Well A was drilled with a synthetic-based drilling fluid
with a 72/25 to 80/20 synthetic-brine ratio (SBR), and 25 wt% 6000

TVD, ft
CaCl2 brine internal phase. After considering the temperature Static mud pressure
effects, we obtain the water activity profle of the drilling fluid, 7000
as shown in Figure 11.
8000
Water Activity
Pore pressure
0.75 0.80 0.85 0.90 0.95 1.00 9000
3000
Osmotic pressure
10000
Formation
4000
Figure 12 - Pore pressure, osmotic pressure and mud static
5000 pressure profiles of Well A (25 wt% CaCl2 internal phase).
It is seen from Figure 12 that the mud hydrostatic pressure
TVD, ft

6000 SBM with 25 wt% is greater than the pore pressure. If the pressure difference
CaCl 2 internal phase between mud pressure and pore pressure overcomes the
7000 threshold pressure, the synthetic based fluids might invade the
micro-fractures which may reduce the internal friction angle
8000 and destabilize the wellbore [3, 18, 31, 32]. Therefore, in stressed
shale formations, the mud weight increase would have
9000 negative consequence on wellbore stability [6]. Meanwhile,
very high osmotic pressure was generated between the drilling
10000 fluid and formation. This osmotic pressure extracted water out
of the formation and caused fractures as shown in Figure 8.
Figure 11 - Water activity profiles for formation and These two phenomena act together and cause the wellbore lose
of support and wellbore instability problems.
drilling fluids for Well A.
It is seen from Figure 11 that there is a difference in water The original plan for Well A was to drill the well
activity between the drilling fluid and the formation. This averaging 22.8 days per 10K feet, as shown in Figure 13. As
difference is larger in the upper formation than that in the the well reached TD, wellbore instability problems occurred.
deeper formation. Even at the increased formation depths, the Prior to the instability problems, the well was drilled close to
water activity of the formation is higher than the water activity plan, with minimal drilling fluid losses. As instability occurred
of the drilling fluids by a factor of 0.12 aw. This high and another sidetrack was drilled, in excess of 2,500 bbls of
difference results in increased chemical potential difference synthetic-based fluid was lost into the formations and drilling
between the drilling fluid and formation. By using Equation time was increased from the target of 22.8 days to 70 days /
(13), the osmotic pressure can be calculated. In the 10K feet. The wellbore instability problems resulted in an
calculation, we assumed the membrane efficiency is 1 because increased drilling fluid cost which was 75 percent above the
it was found from our experiment tests that the SBM acted as target cost.
a near perfect semi-permeable membrane. The profiles of pore
pressure, hydrostatic mud pressure and osmotic pressure are
shown in Figure 12.
SPE 102498 9

12000 Stress, psi

0 1000 2000 3000
3000
14000
Measured Depth, ft

Actual 4000
Planned
5000
16000
6000

TVD, ft
Osmotic Pressure
Pore pressure
7000
18000

8000

20000 9000
0 15 30 45 60
Days 10000

Figure 13 – Planned and actual days for Well A. Figure 15 - Pore pressure, osmotic pressure and mud static
pressure profiles of Well B (18 wt% CaCl2 internal phase).
The profiles of pore pressure and osmotic pressure for
Well B Well B are shown in Figure 15. Compared with Well A, the
Well B was also located in Mississippi Canyon with a total osmotic pressure was significantly decreased due the small
depth of 18,610 ft. in 6,828 ft. of water. Using the same decrease of salinity. For example, at 17,055 ft. (MD)/8,824 ft.
geomechanical parameters as in Well A, the water activity (TVD), the osmotic pressure in Well A is about 3,600 psi,
profile of Well B is shown in Figure 14. while it is only about 750 psi in Well B. This decrease in
osmotic pressure may help to improve wellbore stability, as
Similar to Well A, Well B was drilled using a synthetic- discussed previously.
based fluid with an average 76/24 SBR. Different from Well
A, the average CaCl2 concentration in internal phase was Although there were some tight spots observed and a
reduced and maintained in a range of 18 - 19.3 wt%. After logging tool was stuck and recovered, no severe wellbore
considering the temperature effects, we obtain the water instability problems occurred. The well was drilled close the
activity profile of drilling fluids, as shown in Figure 14. planned 28 days / 10K ft., at 29.9 days / 10K ft., as shown in
Compared with Well A, the activity difference in Well B is Figure 16. The drilling fluid costs were 27 % below the target
reduced, as is the chemical potentials difference. cost.

Water Activity 7000

0.75 0.80 0.85 0.90 0.95 1.00
3000 9000
Actual
Measured Depth, ft

4000 11000

5000 13000
TVD, ft

6000 15000
Formation Planned

7000 SBM with 18 wt% 17000

CaCl2 internal
8000 phase 19000

9000 21000
0 5 10 15 20 25 30 35 40
10000 Days

Figure 14 - Water activity profiles of formation and Figure 16 – Planned and actual days for Well B.
drilling fluids in Well B.
10 SPE 102498

Conclusions ν = Poisson ratio;

The following conclusions can be derived from this study: ui = Chemical potential of ith component;
1) In fractured formations, stresses are in a critical state.
Therefore, any disruption, including chemical and u i0 = Chemical potential of ith component under standard
mechanical, may cause wellbore instability problems. condition;
2) A modified method, based on the effective mean Vw = Molar volume of water, 0.018 m3/mol;
stress concept, was developed to predict the in-situ ρb = Bulk density [=] m/L3;
water activity of shale formations. Results show that σH = Total maximum horizontal stress [=] m/t2;
the predicted value closely matches laboratory
measurements. σh = Total minimum horizontal stress [=] m/t2;
3) The SBMs in the test acts as a near-perfect semi- σ′m = Mean effective overburden stress [=] m/t2;
permeable membrane, which allows the flow of water
and restricts the flow of ions in shale. σt = Mean effective overburden stress [=] m/t2;
4) Field cases and lab results showed that down-hole σv = Total overburden stress [=] m/t2;
water activity of the formation is higher than that σ′v = Effective overburden stress [=] m/t2;
measured at surface.
5) Gulf of Mexico field experiences demonstrate that by α = Biot's constant;
properly increasing the water activity of oil-based αm = Membrane efficiency, dimensionless;
mud it can be expected to improve wellbore stability wt% = Percent by weight.
in naturally fractured shale formations. * [=] means has unit of, L, m, t and T are length, mass, time
6) The osmotic effect is a valuable tool to help maintain and temperature respectively.
wellbore stability. However, overused it may also
cause wellbore instability problems. References
1. Chenevert, M. E: “Shale Control with Balanced-Activity
Acknowledgements Oil-Continuous Muds”, Journal of Petroleum Technology
Special thanks are extended to Dr. Stephen Edwards at BP for (Oct. 1970).
his assistance to obtain the information used in the case 2. Salisbury, D.P. and Deem, C.K.: “Test Show How Oil Muds
studies, which can not be finished without his help. The Increase Shale Stability”, World Oil (Oct. 1990), pp. 57-65.
3. Hale, A. H.; Mody, F. K. and Salisbury, D. P.:
authors would also like to thank Mr. Bob Greene at Baker “Experimental Investigation of the Influence of Chemical
Hughes Drilling Fluids for his efforts to obtain the data used in Potential on Wellbore Stability”, IADC/SPE Paper 23885,
case study, Mr. Marvin Pless at Baker Hughes Drilling Fluids presented at the 1992 IADC/SPE Drilling Conference in
for his arrangement of the lab works. The assistances of New Orleans, Louisiana, Feb. 18-21.
Tommy Rush, Patricia Ann Potts, Bill Gusler, Tao Xiang and 4. Chenevert, M.E. and Sharma, A.K.: “Permeability and
Jim Norfleet at Baker Hughes Drilling Fluids are greatly Effective Pore Pressure of Shales”, SPE Drilling &
appreciated. We also wish to thank the management of BP and Completion (March 1993).
Baker Hughes Drilling Fluids to publish this paper. Finally, 5. Simpson, J.P. and Dearing, H.L.: “Diffusion Osmosis-An
Jianguo Zhang, one of the coauthors, extends his special Unrecognized Cause of Shale Instability”, IADC/SPE
59190 presented at the 2000 IADC/SPE Drilling
gratitude to Dr. Martin Chenevert at the University of Texas at Conference held in New Orleans, LA, February 23-25.
Austin for his generous donations of invaluable references, 6. Santarelli, F.J., Dardeau, C. and Zurdo, C.: “ Drilling
and Dr. Mengjiao Yu at the University of Tulsa for his Through Highly Fractured Formations: “ A Problem, a
constructive suggestions, and Ms. Yasmine Aziez at Baker Model and a Cure”, SPE 24592 presented at the 67th
Atlas for her proof reading. Annual Technical conference and Exhibition of the Society
of Petroleum Engineering held in Washington, DC,
Nomenclatures October 4-7, 1992.
ai = Water activity of ith component, dimensionless; 7. Rojas, J.C; Clark, D.E.; Greene, B. and Zhang, Jianguo:
“Optimized salinity delivers improved drilling
a w , mud = Water activity of non-aqueous mud, dimensionless; performance”, presented at the 2006 AADE Fluids
Technology Conference & Exhibition, Wyndham
a w , shale = Water activity of shale, dimensionless; Greenspoint Hotel, Houston, TX, April 11-12.
8. Yew, C.H.; Chenevert, M.E.; Wang, C.L., and Osisanya,
a w, c = Water activity correction factor, dimensionless; S.O.: “Wellbore Stress Distribution Produced by Moisture
C = Concentration of CaCl2 by weight percent; Adsorption”, SPE Drilling Engineering (Dec. 1990).
C1, C2 = Constants; 9. Zhang, Jianguo; Al-Bazali, Talal M.; Chenevert, M. E.;
Sharma, M. M.; Clark, D.E.; Benaissa, S.; and Ong,
g = Gravitational acceleration [=] L/t2;
Seehong: “Compressive Strength and Acoustic Properties
Pb = Breakdown pressure [=] m/L-t2;
Changes in Shale with Exposure to Water-Based
Pc = Confining pressure [=] m/L-t2; Solutions,” presented at the 41st U.S. Rock Mechanics
Pp = Pore pressure [=] m/L-t2; Symposium held in Golden, Colorado, June 17-21, 2006.
Pπ = Osmotic pressure [=] m/L-t2; 10. Chenevert, M.E., and Strassner, J.E.: “Temperature Effects
R = Gas law constant; on Water Activities of Argillaceous Shales and Oil Mud
T = Absolute temperature [=] T; Systems,” fifteenth Oil and Gas Conference, Balatonfured,
z = Vertical depth [=] L; Hungary, September 14-19, 1975.
SPE 102498 11

11. Yu, M.; Chen, G.; Chenevert, M. E. and Sharma, M. M.: Rock Mechanics in Mining, Petroleum and Civil Works,”
“Chemical and Thermal Effects on Wellbore Stability of Cancun, Quintana Roo, Mexico, June 3-5, 1998.
Shale Formations”, SPE71366 presented at the 2001 SPE 27. Zhang, Jianguo; Chenevert, M. M.; Talal, AL-Bazali and
Annual Technical Conference and exhibition held in New Sharma, M. M.: “A New Gravimetric – Swelling Test for
Orleans, USA, Sept. 30~Oct. 3. Evaluating Water and Ion Uptake of Shales”, SPE 89831
12. Van Oort, Eric: “On the physical and chemical stability of presented at the 2004 SPE Annual Technical Conference
shales”, Journal of Petroleum Science & Engineering, 38 and Exhibition held in Houston, Texas, U.S.A., September
(2003) 213-235. 26–29.
13. Mody, F.K. and Hale, A.H.: “A Borehole Stability Model to 28. Santos, H. and Fontoura, S.A.B. da: “Concepts and
Couple the Mechanics and Chemistry of Drilling Fluid Misconceptions of Mud Selection Criteria: How to
Shale Interaction”, SPE/IADC 25728 presented at the 1993 Minimize Borehole Stability Problems?” SPE 38644
SPE/IADC Drilling Conference held in Amsterdam presented at the 1997 SPE Annual Technical Conference
February 23-25. and Exhibition Held in San Antonia, Texas, October 5-8.
14. Horsrud, P.; Bostrom, B.; Sonstebo, E.F.; Holt, R.M.: “ 29. Chenevert, M. E and Amanullah, Md: “Shale Preservation
Interaction Between Shale and Water – Based Drilling and Testing Techniques for Borehole Stability Studies”,
Fluids: Laboratory Exposure Tests Give New Insight into SPE 37672 presented at the 1997 SPE/IADC Drilling
Mechanisms and Field Consequences of KCl Contents”, Conference held in Amsterdam, The Netherlands, 4-6,
SPE 48986 presented at the 1998 SPE Annual Technical March, 1997.
Conference and Exhibition held in New Orleans, Louisiana, 30. Zhang Jianguo; Clark, D.E.; Al-Bazali, Talal M.;
September 27-30. Chenevert, Martin E.; Rojas J.C. and Ong, Seehong: “Ion
15. Gomez, Sandra and He, Wenwu: “Laboratory Method to Movement and Laboratory Technique to Control Wellbore
Evaluate Fracture Development in Hard Shale Formations Stability”, AADE-06-DF-HO-37 presented at the 2006
Exposed to Drilling Fluids”, AADE-06-DF-HO-38 AADE Fluids Technology Conference & Exhibition,
presented at the 2006 AADE Fluids Technology Wyndham Greenspoint Hotel, Houston, TX, April 11-12.
Conference & Exhibition, Wyndham Greenspoint Hotel, 31. Chen, X.; Tan, C.P. and Detournay, C.: “The impact of
Houston, TX, April 11-12. Mud Infiltration on Wellbore Stability in Fractured Rock
16. Kelly, J.: “Drilling Problem Shales”, Oil and Gas J. (June Masses”, SPE/ISRM 78241 presented at the 2002
3, 1968 a) 67-60. SPE/ISRM Rock Mechanics Conference held in Irving,
17. Kelly, J.: “A New Look at Troublesome Shales”, Oil and Texas, October 20-23.
Gas J. (June 10, 1968 b) 94-112. 32. Al-Bazali, Talal; Zhang, Jianguo; Chenevert, M. E. and
18. Labenski, F.; Reid, P. and Santos, H.: “Drilling Fluids Sharma, M.M.: “Measurement of Sealing Capacity of Shale
Approaches for Control of Wellbore Instability in Caprocks”, SPE 96100 presented at 2005 SPE Annual
Fractured Formations”, SPE/IADC 85304 presented at the Technical Conference and Exhibition held in Dallas, Texas,
2003 SPE/IADC Middle East Drilling Technology U.S.A., October 9-12.
Conference & Exhibition held in Abu Dhabi, U.A.E.,
October 20-22.
19. Lal, M.: “Shale Stability: Drilling Fluid Interaction and
Shale Strength,” SPE 54356 presented at the 1999 Latin
American and Caribbean Petroleum Engineering
Conference held in Caracas, Venezuela, April 21-23.
20. Mondshine, T.C.: “New Technique Determines Oil-Mud
Salinity Needs in Shale Drilling”, Oil and Gas J. (July 14,
1969) 70-75.
21. Engelhardt, W.V. and Gaida, K.H.: “Concentration
Changes of Pore Solutions during the Compaction of Clay
Sediments”, Journal of Sedimentary Petrology (1963) v.33,
919-930.
22. Wolfe, C.; Russell, C.; Luise, N.; and Chhajlani, R.: “Log-
based Pore Volume Compressibility Prediction – A
Deepwater GoM Case Study”, SPE 95545 presented at the
2005 SPE Annual Technical Conference and Exhibition
held in Dallas, Texas, U.S.A., October 9-12.
23. Richard, V. B. and Peter, W. G.: “Saturated Solutions for
the Control of Humidity in Biological Research”, Ecology,
January 1960, Vol.41 No.1, 232-236.
24. Carminati, S.; Brignoli, M.; Marco, Di A. and Santarelli, F.
J.: “The Activity Concept Applied to Shales: Consequences
for Oil, Tunneling and Civil Engineering Operations”, Int.
J. Rock Mech. & min. Sci. 34, paper No. 038. 1997.
25. Jarrett, M. A., Gusler, B., Xiang, T. and Clapper,
“Improved Competence in Water Activity Measurement”,
AADE-04-DF-HO-31presented at the AADE 2004
Technology Conference, Houston, Texas, April 6 – 7.
26. Fonseca, C.F. and Chenevert, M.E.: “ The Effects of Stress
and Temperature on Water Activity of Shales”, presented at
the 3rd North American Rock Mechanic Symposium, “