Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

SPE/IADC 67740

Fracture Gradients in Depleted Reservoirs - Drilling Wells in Late Reservoir Life


M.W. Alberty and M. R. McLean, BP

Copyright 2001, SPE/IADC Drilling Conference


Fracture Gradient Determination
This paper was prepared for presentation at the SPE/IADC Drilling Conference held in The traditional method to measure the fracture gradient of a
Amsterdam, The Netherlands, 27 February–1 March 2001.
subsurface formation is through the use of leak-off tests. A
This paper was selected for presentation by an SPE/IADC Program Committee following
review of information contained in an abstract submitted by the author(s). Contents of the
leak-off test is conducted when casing is set immediately
paper, as presented, have not been reviewed by the Society of Petroleum Engineers or the above the interval to be measured and approximately three
International Association of Drilling Contractors and are subject to correction by the author(s).
The material, as presented, does not necessarily reflect any position of the SPE or IADC, their meters (ten feet) of fresh formation is drilled below the casing
officers, or members. Papers presented at the SPE/IADC meetings are subject to publication
review by Editorial Committees of the SPE and IADC. Electronic reproduction, distribution, or
shoe in the formation to be tested. Pressure is applied to the
storage of any part of this paper for commercial purposes without the written consent of the casing and the freshly drilled open hole to measure 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 response of the formation. See Figure 1.
contain conspicuous 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.
Initially, the pressure builds in a linear manner. The initial
slope of the pressure versus time plot is determined by the
pump rate and the compressibility of the system. The majority
Abstract of the system compressibility is a result of the fluid
A key requirement to successful drilling is the selection of a compression, with the casing and short length of open hole
mud weight that provides sufficient pressure to prevent the contributing only a minor amount to the overall
influx of formation pore fluids, while at the same time not compressibility. The pressure continues to build until a
exceeding the fracturing resistance of formations exposed in fracture is induced in the formation wall. Once a fracture is
open hole. The correct prediction of how pore pressure and induced the slope of the pressure build up curve decreases in
fracture resistance varies through the intervals to be drilled is response to the increased volume associated with the fracture.
critical to designing an appropriate casing program. The point where the slope changes is traditionally known as
In many instances fracture gradient profiles will be well the leak-off point and is often taken to represent minimum
defined from experiences of drilling offset wells. However, in horizontal stress.
cases where casing programs are modified or sub-subsurface As pressure continues to increase the fracture extends
conditions have altered, for example as a result of pressure through a ‘disturbed’ near well bore zone where stresses are
depletion, predictive methods are required to extrapolate the typically greater than in situ conditions due to stress
previous measurements and inferences of the fracture gradient. redistribution around the well bore, osmotic effects,
Conventional wisdom on fracture gradients suggests that temperature disturbances, clay reactivity with the mud,
pressure depletion can significantly reduce fracture resistance interfacial tension etc.
and seriously increase drilling difficulty. However, an The pressure will typically reach a peak and then decline
alternative theory on fracture gradients, coupled with growing rapidly, ultimately settling at a propagation pressure that is
evidence of its applicability, indicates that sands are not in somewhat lower than the peak pressure. There are a number of
general the cause of losses associated with induced fractures. plausible explanations for the processes associated with
Despite sands often being under lower in situ stress than extending the fracture beyond the peak pressure. A common
adjacent shale layers, it is in fact the shale that are more likely view is that the drop from peak pressure is associated with
to host an induced fracture responsible for large-scale mud extending the fracture beyond the disturbed zone and/or
losses. In general, shale has a higher Poisson’s ratio than getting sufficient fracture width to transmit pressure easily
sands. In addition, the extent of pressure depletion will be less down the fracture. Once the fracture has been extended
in intra-reservoir shale than in the sand layers. Therefore, the beyond this region the pressure needs only exceed the
predicted effect of reservoir depletion on the fracture gradient minimum horizontal stress for the fracture to propagate. If the
is much less if based on the shale intervals than if based on the pump rate is low, which is the case for leak-off tests, the
sands. excess pressure will be small and the propagation pressure will
Here the theories on fracture gradients and the reasons for be close to the minimum in situ stress.
their differences are discussed. Field examples supporting the If the test is immediately repeated, the high pressures seen
more optimistic approach are presented. as the fracture once again breaks through the disturbed zone is
significantly decreased but the fracture propagation pressure
2 M.W. ALBERTY AND M.R. MCLEAN SPE/IADC 67740

or minimum horizontal stress is essentially identical. The The concept of effective (or matrix) stress as applied to the
process of taking a leak-off test well beyond the initial overburden is equally applicable in the horizontal direction.
departure from linearity, generating a fracture well into the The total horizontal stress is equal to the sum of the pore
formation and then reopening it is often termed an extended pressure and the horizontal effective/matrix stress as expressed
leak-off test (Kunze and Steiger1). in equation (1) below.
In many hydrocarbon basins, the maximum stress is in the
vertical axis and the minimum stress is in the horizontal axis. Sh = P + σ h …(1)
In passive basins, or as a simplifying assumption, the
orthogonal principal stresses in the horizontal plane may be
considered isotropic, but in reality they will never be exactly where S h is the (total) horizontal stress,
equal and in some regions they will be significantly unequal and σh is the effective/matrix horizontal stress.
(Plumb et al2). A fracture will take the path of least resistance,
which in general will mean opening up against the least stress,
which leads to propagation orthogonal to the minimum in situ The magnitude of the effective horizontal stress is
stress. This is represented in Figure 2. dependent on the lateral conditions. In a passive environment
The peak pressure required to generate a fracture is usually formations are assumed restrained laterally while undergoing
greater than the propagation pressure. However, the vertical compaction through deposition. Therefore, the
propagation pressure should serve as the basis for safe well horizontal effective stress is generated to counteract the
design as the peak or initiation pressure cannot always be natural tendency of a formation to dilate laterally in response
relied upon. The magnitude and extent of the higher stressed to vertical loading. See Figure 4. A material’s Poisson’s ratio
environment surrounding a well bore may vary as a function measures its tendency to dilate laterally in response to vertical
of mud properties and well trajectory. It can be significantly loading. A high Poisson’s ratio is a measure of high tendency
lower if a pre-existing fracture or fault is encountered, and will to dilate and therefore a high effective horizontal stress is built
reduce if a surge pressure should temporarily induce a fracture up in response to vertical loading.
through this zone It can be shown that the build up of effective horizontal
Based on the discussed above a “model” can be formulated stress in a passive environment is given by
for selecting a mud and casing program that does not
compromise the fracture gradient. The basic principles of this µ
model are
σh = ( S v − P) …(2)
1− µ
• A ‘safe’ design fracture gradient is equivalent to the
propagation pressure, which in turn is equivalent to the
minimum in situ stress,
where µ is the formation Poisson’s ratio.
• Therefore, formations with the lowest in situ stress
gradient are the most prone to fracturing. (The above neglects temperature effects and non-linear
• Therefore, well design fracture gradient should be based behaviour, both of which may have considerable impact on the
on the formations with the lowest in situ stress gradient. accuracy of a computed horizontal stress from a laboratory or
log determined Poisson’s ratio.)
Principles of Fracture Gradient Prediction
Overburden stress is borne in a formation by both the fluid and Combining (1) and (2) gives
the rock frame. See Figure 3. Following Terzaghi3, the
overburden, pore pressure and effective (or matrix) stress can µ
Sh = P + ( S v − P) …(3)
be simply expressed as 1− µ
Sv = P + σ v There are many variations on this relationship. In the more
common variations k , the ratio of horizontal to vertical
where Sv is the overburden stress, effective stress, is substituted for the µ /(1 − µ ) term; in
P is the formation pore pressure, others a third term is added to describe tectonic stress.
and σv is the effective overburden stress. Regardless, the principles are common to each other.
For naturally occurring materials Poisson’s ratio lies
On application of overburden a formation attempts to between 0 and 0.5. From equation (3) it can be seen that as
compact. The extent of any overpressure built up is dependent pore pressure drops, so will the horizontal stress (except for
on how quickly the fluid can be expelled from the pores (i.e. the extreme case of µ =0.5). Also, it can be seen that as
drain away) relative to the speed of loading. Any over-
Poisson’s ratio decreases the horizontal stress decreases.
pressure generated results in an equivalent under-stressing of
Below are listed the typical ranges of observed Poisson’s
the rock matrix.
ratios for common lithologies.
SPE/IADC 67740 FRACTURE GRADIENTS IN DEPLETED RESERVOIRS - DRILLING WELLS IN LATE RESERVOIR LIFE 3

drilling induced fractures. The real time measurement


Sands - 0.10 to 0.22. establishes a base line resistivity profile that can be compared
Silts - 0.15 to 0.30 against MAD logs to determine where the highly resistive mud
Carbonates - 0.20 to 0.35 has entered fractures in the formation. Repeated use of these
Shale - 0.22 to 0.48 logs over the past ten years have demonstrated that most lost
Salt - 0.45 to 0.50 circulation events occur in shale. So perhaps the conventional
model does not hold up under drilling conditions.
A flaw in applying the conventional model is the
Sand has an average Poisson’s ratio that is significantly assumption that the generation of a fracture necessarily leads
lower than typical shale. Therefore, sand should be expected to major drilling losses. Figure 6 depicts the propagation of a
to have a lower fracture gradient than shale under similar fracture in a porous, permeable formation. As the fracture
conditions of pore pressure and overburden. opens and whole mud moves along the fracture face, the
When these principles are applied to a reservoir with filtrate of the mud invades into the adjacent formation
declining pressure, the problem of drilling these reservoirs is dehydrating the mud as it travels down the fracture.
easily recognized. Figure 5 shows how sand pressure is Ultimately, the mud slurry becomes so dehydrated that it
expected to decline as a function of the decline in reservoir becomes immobile, blocking the fracture and preventing
pressure. The initial sand fracture pressure is lower than the further fracture growth. Additionally, the dehydrated mud in
initial shale fracture pressure and the sand fracture pressure the fracture creates a displacement between the two fracture
declines at an increased rate relative to the shale fracture faces that raises the stress field in the near well bore
pressure. This is the root of the problem with drilling wells environment. In this manner the near well bore stress
late in a field’s life. As a function of reservoir depletion, sand environment is incrementally raised relative to the original
fracture gradient may have declined below the value necessary conditions. This process is well documented in the analogous
to maintain shale stability in adjacent shale and below the problem of propped fracture stimulation and known as fracture
value of un-drained nearby sands. “screen-out” (e.g. Smith et al4).
The increase in near well bore stress state as a result of a
Evidence for the Convention Model fracture screen-out will raise the pressure necessary to re-
The “conventional model” to fracture gradient prediction is fracture the formation. Successive re-fracturing and screen-out
founded on equation (3) or a derivative of it. This leads to the will further increase the local stress state (Moschovides et al5).
concept that (a) formations with a low Poisson’s ratio are at Ultimately, the adjacent shale will become easier to fracture
greatest risk and that (b) reservoir depletion can significantly than the sand. At this point further fracturing can lead to
reduce the fracture gradient, particularly in formations with a massive mud losses as the dehydration of the mud is
low Poisson’s ratio (based on the assumption that the insufficient in shale to screen-out and block the fracture.
reduction in fracture gradient is synonymous with the It is important to recognize that the far field minimum
reduction in the minimum horizontal stress). horizontal stress in the sand is still lower than the far field
The evidence for the conventional theory is built upon minimum horizontal stress in the adjacent shale. However, the
rock/soil mechanics principles and decades of experience and pressure in the borehole is unable to penetrate through the high
has been demonstrated to work well in many cases. near well stress zone, so substantial losses do not occur in the
Compelling evidence can be observed in water injector wells. sand.
The pressure to inject the water is well recorded and It is also important to recognize that shale inter-beds
temperature logs provide evidence that fractures have occurred within the sand and immediately adjacent to the sand will
in sands. Additionally, fracture pressure is observed to drop dewater into the sand slowly if sand pressure is reduced below
with decline in reservoir pressure. the virgin shale pressure. This reduction in shale pore pressure
In propped fracture treatments, fractures are contained will reduce the fracture pressure of the shale as discussed
within sand layers demonstrating that shale have higher stress. earlier. Therefore, the expectation should be that losses would
As in the injector case, fracture pressure drops with pressure most likely occur in shale inter-beds and shale immediately
depletion in accordance with sand Poisson’s ratio. adjacent to sand where pressure depletion is occurring. If
Finally, laboratory measurements and logs both confirm inter-bedded shale is the point where losses occur, it may be
the lower Poisson’s ratio in sand relative to shale. difficult to recognize those loss zones on MWD resistivity
logs because of the masking effect of the highly resistive pay
An Alternative Model sands that are being pressure depleted.
Given the conventional model described above, the
expectations would be that drilling losses would Evidence For the Alternative Model
predominantly occur in sands near the last casing shoe. The The potential for the screen-out effect in association with
use of measurement while drilling (MWD) resistivity devices drilling sands was recognized in the early 1990’s. Morita et al6
in real time and in measurement after drilling (MAD) in oil published evidence from laboratory results that showed that
base and synthetic mud provide the ability to recognize fracture resistance increased in sands when particle-laden mud
4 M.W. ALBERTY AND M.R. MCLEAN SPE/IADC 67740

was used as the fracture invading fluid. Field evidence was is maintained as the hole is deepened into the pressure
presented (Fuh et al7) and a patent secured (Fuh et al8) for regression below 4500 meters (14,764 feet) and into the low-
optimizing fluid properties to enhance the screen-out process. pressure sands below. Once drilling into these sands, returns
Further evidence for the process is found in analogous are lost at exactly the pressure predicted by the new theory and
technologies of (a) water injection where a loss of injectivity well above the fracture pressure predicted by the conventional
can be noted if particulates in the water are not removed theory.
before injection, (b) screen-out in propped fracture treatments This example clearly shows that the apparent fracture
(Smith et al4), and (c) cuttings re-injection where preferential gradient in sands is substantially higher than the conventional
fracture propagation occurs in low permeability strata when theory would indicate.
juxtaposed to permeable sand (Willson et al9).
Common drilling practices indicate that predictions based Drilling Example 2
on sands being the weakest link under-estimate the appropriate The second example is taken from the Judge Digby field in
fracture gradient for drilling a section. Hole sections with south Louisiana Shaughnessy and Fuqua10). The “A” sands are
sands open are routinely withstanding borehole hydrostatic highly pressured (1.92 sg, 16 ppg) water bearing sands as
pressures that should induce fractures under conventional shown in Figure 8. The “B” sands are gas-bearing sands that
wisdom. In the high pressure environment of the Gulf of have been depleted by production. Subsequent to their
Mexico deep water, mud weights are routinely used that are depletion, Amoco purchased the lease and pursued deeper
within 0.5 PPG of the leak-off at casing shoes without targets. The highly pressured “C” sands were discovered some
breaking down the sands near those shoes and when losses are 183 meters (600 feet) below the depleted “B” sands.
occurring, they are almost always in the shale rather than the Considerable savings could be made if the depleted “B” sands
sands. However, lack of major losses in over-pressured sands could be isolated behind the same casing string used to isolate
is not conclusive evidence that losses have been prevented by the “A” sands. A study was conducted and the alternative
screen-out of any induced fractures. model applied.
The examples that follow are associated with sands that are The alternative model suggests that the sands could
depleted or under-pressured relative to overlying formations. withstand slightly above 1.92 sg (16 ppg) mud although the
Some losses did occur, but they occur at a fracture gradient drilling margin would be very narrow. The zone was drilled
that is completely consistent with the alternative model and with 1.92 sg mud with some minor losses while circulating
are strong support for the theory that the sand fracture gradient and no losses with the mud pumps off.
has been elevated through screen-out processes. The conventional theory predicts that these sands would
have a fracture gradient on the order of 1.02 sg (8.5 ppg). The
Drilling Example 1. sands were subjected to 1.92 sg (16 ppg) mud with no losses.
In the southern Caspian Sea the Balakhany sands are Clearly the conventional theory significantly under estimates
associated with a major pressure regression immediately fracture resistance in sands when drilling with a conventional
below the regional high-pressure seal. The presence of the seal mud system.
above these sands and the occurrence of a high pressure sand
below create a drilling environment where high mud weights Limitations
are required to drill these low pressure sands cost effectively. The process of screening out sands requires a mud that loses
Figure 7 is a display of the lithology, pore pressure, actual filtrate to the fracture face to accommodate dehydration and
mud weight, leak-off, and formation integrity test data. The ultimately blockage of the fracture. In many instances the
calculated fracture gradient for both sand and shale under the process occurs effectively without any particular changes to
traditional theory is also displayed. conventional mud design. However, in clay rich or silty sands
The high formation integrity test (FIT) at 3600 meters getting early dehydration of the mud close to well bore may be
(11,812 feet) is a result of regional tectonic stress observed in more difficult. Similarly, the proposed fracture blocking
this and offset wells at this level. This test exceeds mechanism would be difficult to achieve with high density,
overburden, shows minimal fluid compressibility, and displays low solids concentration drilling fluids such as formates. In
a strictly linear pressure build-up curve. No leak-off or these cases greater effort in mud design may be required.
induced fractures have occurred at the reported pressure. Currently there are some JIP’s in place to try to determine the
There is no doubt that shale is exposed during this test. This relationship between mud properties and fracture induced fluid
tectonic stress is significantly reduced once the fault at 3900 losses.
meters (12,796 feet) is crossed as displayed on the subsequent As a result of the high differential and the multiple
measures of formation strength. This reduction is substantiated fractures that will be induced in the near well bore
by a lost circulation event that occurs immediately below the environment, significant formation damage should be
fault. expected. In all the empirical cases observed by the authors,
The mud weight is increased to cover the high-pressure none were in potential productive zones, so no measurement
sands, denoted with their measured pore pressures, in the of damage was ever observed. The use of acid soluble
pressure ramp at 4000 meters (13,124 feet). This mud weight particulates may be required if the principle is to be applied in
SPE/IADC 67740 FRACTURE GRADIENTS IN DEPLETED RESERVOIRS - DRILLING WELLS IN LATE RESERVOIR LIFE 5

potential productive horizons. 4. Smith, M.B., Miller, W.K. II, and Haga, J. (1987). Tip screenout
fracturing: A technique for soft unstable formations. SPE Pet.
Conclusions Eng., May, pp.95-103.
Conventional fracture theory predicts that sands have a lower 5. Moschovidis, Z. A., Gardner, D. C., Sund, G. and Veatch, R. W.
(1993). Disposal of Oily Cuttings by Downhole Periodic
horizontal stress than shale at similar overburden and pore Fracturing in Valhall, North Sea: A Case Study and Modelling
pressure. This theory is demonstrated as correct when clear Concepts. SPE/IADC 25757.
non-particulate fluids are used as the fracturing fluid. 6. Morita, N., Black, A.D., and Fuh, G-F. (1990). Theory of lost
However, when particulate laden fluids are used fractures will circulation pressure. SPE 20409.
screen-out or block the fluid progress along the fracture, 7. Fuh, G-F., Morita, N. Boyd, P.A. and McGoffin, S.J. (1992). A
significantly increasing the fracture resistance of sands. new approach to preventing lost circulation while drilling.
The increased fracture resistance in sands results in shale 8. Fuh, G-F., Morita, N., Whitfield, D.L. and Strah, D.A. (1993).
being the more likely loss zone associated with low pressure Method for inhibiting and propagation of formation fractures
or depleted sands. while drilling and casing a well. US Patent 5,207,282.
9. Willson, S.M., Steiger, R.P., Moschovidis, Z.A., Abou-Sayed,
The conventional model takes no account of the ‘healing’ A,.De Bree, Ph., and Sirevag, G. (1990). Laboratory
properties associated with sands resulting from the screen-out investigation of drill cuttings reinjection disposal by downhole
effect. As a result, the conventional model will lead to overly injection. 37th US Rock Mech. Symp., Vale, Co., June.
pessimistic assessments of drill ability, particularly in depleted 10. Shaughnessy, J. and Fuqua, R. (2000). Successfully drilling
reservoirs. Here, the alternative model advocated in this paper highly depleted sands. SPE 67744
clearly distinguishes between an unchanged fracture pressure
at the borehole wall, compared with the reduction in minimum
horizontal far-field stress (as predicted by the conventional
model). There is significant field evidence to support the
alternative model, first suggested in the early 1990’s.
Significant formation damage should be expected when the
screen-out principle is applied to highly overbalanced sands.
Mud formulation may be important to get the screen-out
effect, particularly in silt zones.

Nomenclature
µ = formation Poisson’s ratio
Sv = overburden stress, m/Lt2, MPa
Sh = (total) horizontal stress, m/Lt2, MPa
σv = effective overburden stress, m/Lt2, MPa
k = ratio of horizontal to vertical effective stress
P = formation pressure, m/Lt2, MPa

Acknowledgements
The authors wish to thank BP for allow them to present this
material. The authors would also like to thank BP’s partners in
Shah Deniz and Judge Digby field for allowing us to present
the enclosed well examples. The authors would also like to
recognize Nigel Last, Joe Hagan, Mark Aston, John
Shaughnessy and Ron Fuqua for their contributions in
substantiating the principles presented here and Steve Willson
for checking the technical content and providing useful
contributions.

References
1. Kunze, K.R. and Steiger, R.P. (1992). Accurate in-situ stress
measurements during drilling operations. SPE 245935.
2. Plumb, R., Papanastasiou, P. and Last, N.C. (1995). Constraining
the state of stress in tectonically active settings. SPE 47240.
3. Terzaghi, K. and Peck, R.B. (1967). Soil Mechanics in Engineering
Practice., June, pp.676-684.
Overburden

Peak

Bottom Hole Pressure Shut-in


Leak-
Min. in situ
off
stress
Bleed-off

Min horizontal
stress
Static mud
column pressure
Max horizontal Static mud pressure
Time stress
Figure 1. Leak-off test pressure profile. Figure 2. Effect of stress on fracturing.

Total Overburden Pore Pressure Effective Overburden

Sv P σv

= +

Figure 3. The relationship between pore pressure, effective stress and total stress.
Vert. stress

Unloaded Horiz. stress


Vert. stress

Reservoir Pressure
Fracture Pressure

Shale frac pressure

Sand frac pressure

Uniaxial stress Horiz. stress Reservoir pressure

Time
Vert. stress

Figure 5. Fracture pressure as a function of declining pore pressure.

Horiz. stress
Uniaxial strain
Figure 4. The Poisson’s effect.
SPE/IADC 67740 FRACTURE GRADIENTS IN DEPLETED RESERVOIRS - DRILLING WELLS IN LATE RESERVOIR LIFE 7

Figure 6. Fracture screen out in permeable sands.

Drilling Example - Shah Deniz 2


Mud Wgt Predicted Sand FG
Predicted PP Predicted Shale FG
3000

3500
Leak-off
Depth (m TVDBRT)

Induced losses
4000 Leak-off

4500

5000
Losses
FIT
5500

6000
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
Pore Pressure & Frac Gradients (SG)

Figure 7. The SDX-2 drilling example.


8 M.W. ALBERTY AND M.R. MCLEAN SPE/IADC 67740

Drilling Example - Tuscaloosa Parlange No. 11


Depth (m TVDBRT) Mud Wgt
4,000

FG (Shales)
4,500
Pore Pressure

5,000

5,500

Sand FG
6,000 ‘A’ Sands
‘B’ Sands

6,500 ‘C’ Sands


0 0.5 1.0 1.5 2.0 2.5
Pore Pressure & Frac Gradients (s.g.)

Figure 8. The Tuscalosa drilling example.

You might also like