17 Compaction GeomechanicsAA

MBDCI
Compaction Geomechanics:
Mechanisms, Screening
Compaction Geomechanics
Maurice B. Dusseault
MBDCI
Compaction as a Drive
Compaction occurs whenever the net stress
increases ()
Magnitude depends of the rock stiffness,
fabric (some rocks have a quasi-stable )
Important in high-porosity sandstones
Important in North Sea Chalk (e.g. Ekofisk)
Important in Diatomite (California mainly)
Recompaction is important in cyclic steam
stimulation (porosity cycling)

Compaction geomechanics is fundamental
MBDCI
The Case of Ekofisk

3000 m deep
7 km wide
Very thick reservoir
High porosity chalks
from 25 to 50%
Overpressured by 1.7
EKOFISK
Large drawdowns are

feasible large v
Large compactions
Would we plan for these
nowadays?
MBDCI
Reservoir Compaction
Triggered by reduction in pore pressure
Important drive mechanism in high cases
(Maracaibo, Ekofisk, Wilmington, ...)
But, problems develop with compaction ...
Casing collapse in the reservoir
Surface subsidence from deep compaction

Casing shears above the reservoir (Ekofisk)
Reservoir simulator predictions are contentious
Large stress redistributions, microseismicity
MBDCI
Major Design Steps

Identify physical compaction mechanisms
Identify susceptible reservoirs
Based on experience in other reservoirs
Based on geophysical data (logs, seismics)
Based on core examination and lab tests
Based on geomechanics analysis

Based on monitoring information
Study the cost-benefit gain to the company

Implement mitigation measures in advance
MBDCI
Screening Strategies
Screening and Analysis
First-order screening
(geology + cases)
Geomechanics assessment Impact assessment Mitigation options

-geophysical logs -well modeling -pressure maintenance
-petrophysical evaluation -reservoir modeling -facilities redesign
-stress history -production strategies
Risk assessment Cost-benefit analysis Cost of options

-second-order screening $ - - Bs
-production predictions
Decision making Is compaction

-Monitoring cases -experience base
-learning and teaching beneficial?
MBDCI
Compaction Effects I
Compaction = (ij, E, , ...)
Is p always = ij? z
No, compaction is not uniform
ij, p
p is not uniform in reservoir
Overburden arching takes place
Cc (E, )
Thus, compaction moves out

from production wells, arching z ~ Ccv
delays full z development

This can be modeled quite well
MBDCI
Compaction Effects II
Compaction and depletion can change both
normal stresses and shear stresses
If sharp gradient of compaction occurs,
(shear stress) can be quite large
This can cause shearing, grain crushing,
loss of cohesion, liquefaction (Chalk), etc.
These factors affect the permeability, often

negatively, but shearing can also increase k
MBDCI
Delay of compaction always occurs in early

time, before p zones intersect (aspect ratio)
initially after some Q

p region
no p yet
MBDCI
z Delay Through Arching

full subsidence response delayed
arching occurs
in this phase, v until drawdown zones
is not equal to p interact at the
reservoir scale
drawdown
zones
compaction impeded
overburden stresses flow around the p, V zone
MBDCI
Full Compaction
full when zones meet,

subsidence arching is destroyed,
develops full
compaction full compaction occurs
triggered
stresses now flow without arching around zones

MBDCI
Compaction sustains production, and can
change the production profile substantially
oil lost?
predicted, assuming v = p
actual Q with
delayed z
Q - total field
predicted life
actual life
economic
cutoff Q
time
Negative Effects of ij
MBDCI
Productivity can decline with an increase in

ij , both in normal and shear stresses
Fracture aperture diminution (+++)
Pore throat constriction (+)
The relative importance depends on the

rock mechanics properties of the reservoir
Casing shear or buckling can occur

Surface subsidence can take place
These can be costly if unexpected
Positive Effects of ij
MBDCI
Changes in the effective stress can trigger

changes in the porosity
Compaction can be substantial (UCSS, chalk)
Compaction serves to sustain the pressure
Much more oil is driven to the wellbores
In high Chalk, a volume change as much

as 10% can take place, all oil production

Compaction of high shales can also expel
water into the reservoir, displacing more oil
MBDCI
Mechanisms I (Sandstones)
Pore pressure is reduced by production (p)
The vertical effective stress, v, rises
A sand of high compressibility will begin to
compact to a lower porosity
This maintains drive pressures in liquid-
dominated reservoirs, may cause subsidence
Also, a direct fluid expulsion occurs (V)

Water can be expelled from adjacent shales
MBDCI
Elastic Compression
leads to increased contactforces, fn
The contact area increases, porosity drops
This is a function of compressibility
If elastic, V is recoverable (-V = +V)
fi
fn
E,
Ap
Ap - A
p
fn
MBDCI
Inelastic Compression
leads to increased contactforces, fn
Grain rearrangement takes place, drops
Perhaps a bit of grain contact crushing
In high sands, this is an irrecoverable V
fi
fn
E,
Ap
Ap - A
p
fn
MBDCI
Elastic and Inelastic Strain

porosity Depletion = +
elastic behavior
Injection = -
inelastic compaction curve

behavior
irrecoverable
compaction elastic behavior
rebound curve
1 MPa 10 MPa 100 MPa
log(v)
MBDCI
Mechanisms II (Sandstones)
For small p, grain rearrangement is most
important; V not recoverable. Also,
Contacts compress elastically, recoverable
At intermediate p, grain contacts deform
elastoplastically, strain is not recoverable
At high p (high ), grain splitting,

crushing, and even creep occurs, especially
in lithic and arkosic sandstones
MBDCI
Processes and Compaction

porosity
elastic compression at low
elastic grain rearrangement

0.30 recovery at intermediate
irrecoverable
compaction elastoplastic grain
contact behavior
0.25 rebound
curve grain crushing

at high

log(v)
0.20 1 MPa 10 MPa 100 MPa
low high
MBDCI
Mechanisms, Chalk (I)

North Sea Chalks (and a few other materials)
exist in a high-porosity state (>35-40%)
This state is quasi-stable and exists
because of cementation between grains
If grains crush or cement ruptures, massive
compaction occurs (>10 m at Ekofisk)
This is triggered by the increase in v, and

also by increased stress difference (1 - 3)
Shear destroys cement, triggers compaction
MBDCI
Why Does Chalk Collapse?
Hollow, weak grains (coccoliths)
Weak cementation
(dog-tooth calcite)
Weak, cleavable grain

mineral (CaCO3)
MBDCI
Mechanisms, Chalk (II)

Threshold stress from cementation in Chalk
Grains are also hollow and weak
Once collapse happens, the Chalk can even
become liquefied locally
The stresses are transferred to adjacent rock
and the process can propagate far
The whole reservoir compacts when the

collapse is at the interwell scale
MBDCI
North Sea Chalk Collapse

porosity
threshold stress
irrecoverable strain collapse of fabric

rebound curves compression curve
1 MPa 10 MPa 100 MPa

log(v)
MBDCI
Geological History!
Diagenesis = pressure solution,
densification, grain-to-grain cementation
Cementation can preserve a rock at a very
high porosity (collapsible, like Chalk)
Densification and pressure solution can
make the rock stiffer at the same value
Overcompaction (deep burial history) can

make the rock stiffer, little compaction
Geological history is a vital factor
MBDCI
Cementation, Compaction
porosity apparent threshold
normal collapse if cement

densification is ruptured
cementation effect
virgin
compression curve
stiff
response
log(v)
MBDCI
Diagenetic Densification
diagenetic
porosity loss
@ constant virgin
compression curve
present state
stiff
response
log(v)
MBDCI
Precompaction Effect
virgin
compression curve
present state
stiff
response
log(v)
Deep burial followed by uplift and erosion lead to precompaction
MBDCI
Threshold Drawdown
Usually, some threshold drawdown must
occur before significant compaction starts
There are three effects responsible:
The sand may be geologically pre-densified
There may be a cementation to overcome (Chalk)
The p may not yet be at the reservoir scale (arching)
These are hard to quantify without careful

geological studies and laboratory testing
Short-term well testing can be misleading!
MBDCI
Cementation, Diagenesis
stresses
pressure
time solution,
temperature 25-32%
chemistry
porosity
initial state,
reduction cementation,
35% porosity
25-32%
Both solution and cementation reduce porosity, increase stiffness
MBDCI
Additional Mechanisms
Compaction can lead to loss of some
permeability in natural fractures
Grain crushing can occur as well, k
Depletion loss of lateral stress, increase
in mean , increase in shear stress
Increase in shear stress usually causes k
Compaction can release fines from strata

Are there other effects in your reservoirs?
MBDCI
Fracture Aperture, k
Fracture aperture is sensitive to n
Permeability is highly sensitive to aperture
Shear displacement and asperity crushing
can develop with
n
Itappears that a homo-

effective aperture
p p + p geneous constitutive macro-
scopic law is required for good
asperities
predictions in analysis
MBDCI
Fracture Permeability Loss

In many cases, fracture permeability
decreases by a factor of 1.5 to 3
This is to a degree a compaction effect (loss
of aperture as net stress increases)
It also retards compaction (e.g. fractures in
Chalk)
It is important in coal, North Sea Chalk, but

not in sandstones
MBDCI
Grain Crushing and k

Depletion or differential volumetric strain
causes high high fn on individual grains
Weak (lithic) or cleavable (felspathic)
grains crush and fragment
Pore throats then become smaller or
blocked by fragments, k drops
feldspar crushed, quartz intact
feldspar quartz
Depletion Effect on h
MBDCI
h stress trajectories wellbore h concentration
far-field stresses
h along
final
h
Zone after production (p)
wellbore
Operational consequences:
-low pfrac in reservoir
initial
h
-higher pfrac above reservoir
Z
MBDCI
Shear, Fracture Opening
v
Zone of pressure
decline, -p
h
Pressure decline leads to an increase in the shear stress

This leads to shearing, which causes fractures and fissures to open
This leads to increases in permeability, better reservoir drainage

MBDCI
Pore Blockage Mechanisms
Geochemical effects!
Fines migration
can block pores
Mineral
deposition
MBDCI
Most Sensitive Cases

Highly fractured reservoirs
Reservoirs with asphaltene precipitation,
scale deposition, or fines migration
potential
Tectonically stressed reservoirs (high )
Reservoirs with crushable grains or
collapsible fabric (North Sea chalk, coal)

Thermal shock in unconsolidated sands (?)
Other cases?
MBDCI
Porosity vs Depth
0 0.25 0.50 0.75 1.0
porosity
sands & mud
sandstones
clay
clay & shale,
normal line
mud-
stone
shale The specific details of

effect of
overpressures these relationships are
on porosity a function of basin age,
diagenesis, heat flow ...
4-6 km
depth
MBDCI
Subsidence MacroMechanics
MBDCI
Subsidence Bowl
(Vertical scale is greatly exaggerated)
subsidence bowl, L
compression
extension
zmax

compaction, T
depth, Z
width, W
MBDCI
Subsidence Magnitude
If W < Z, arching, little subsidence (<25% T)

If Z < W < 2Z, partial arching (25-75% of T)
If W > 2Z, minimal arching (>75% of T)
Bowl width: L = W + 2Zsin
If W > 2Z, zmax approaches T
(angle of draw) usually 25 to 45 degrees

In cases of complex geometry and stacked
reservoirs, numerical approaches required
MBDCI
Casing Impairment
Either loss of pressure integrity, or excessive
deformations (dogleg, buckling)
Problem in massively compacting reservoirs
Compaction can distort and even buckle casing
Threads can pop open, casing can be ovalized
Triggering of faults shear casing which pass through
Overburden flexure causes shear planes to develop
Casings cannot withstand much shear

These are vexing and difficult problems
More compliant casing and cement?
MBDCI
Casing Shearing!!
It is a major problem in all compacting
reservoir cases
Will deal with this in greater detail in
another presentation, as it is important for
many new production technologies:
CHOPS
Thermal methods
Etc.
MBDCI
Measuring Compaction
MBDCI
Measuring Compaction
In the reservoir
Radioactive bullets, casing collar logs,
gravimeter logs, behind-the-casing precision
logs, magnetic devices, extensometers, strain
gauges on casing, and other devices
At the surface (subsidence)
Precision surveys, aerial photos, differential

GPS, InSAR, depth gauges (offshore sea floor
subsidence), tiltmeters, and other methods
MBDCI
Radioactive Bullets
The zone of interest is selected
Before casing, radioactive bullets are fired
into the adjacent strata
Casing is placed, garbage removed
A baseline gamma log is run (slowly!)
At intervals, logging is repeated, and the

difference in gamma peaks is measured
Strain = L/L, accuracy of about 1-2 cm
MBDCI
Casing Collar Logs

Casing moves with the cement and the rock
The casing collar makes a thicker steel zone
This can be detected accurately on a log
sensitive to the effect of steel (magnetic)
Logs are run repeatedly, strain = L/L
Short casing joints can be used for detail

If casing slips, results not reliable
If doglegged, cant run the log
MBDCI
Borehole Extensometers
Wires anchored in the casing
Brought to surface, tensioned wire 3
wire 1
Attached to a transducer or to wire 2 sheaves
a mechanical measuring tool

W
Reading taken repeatedly anchor 3
Resistant to doglegging casing

anchor 2
Logs cant be run in the hole
Other instruments installed anchor 1
MBDCI
Other Borehole Methods

Strong magnets outside fibreglass casing
are used (fibreglass just over the interest
zone)
Strain gauges bonded to the casing, inside
or outside (best), leads to surface
Gravity logs (downhole gravimeter)
Other behind-the-casing logs which are

sensitive to the lithology changes
Tilmeters can be placed in boreholes
MBDCI
Surface z Measurements
Differential GPS can give accuracies about
one cm on land, not as good offshore
Precision aerial photos with stable targets
give down to perhaps one cm, a bit less
Surface monument array with surveying can
give precisions of less than a millimetre

Tiltmeters measure inclination extremely
precisely, give electronic readout
Other methods?
MBDCI
GPS - Fixed Monuments Visits
Antenna
Monument
MBDCI
InSAR - IOL - Cold Lake
+285 mm mega-row
+200 CSS
-210
+100 Vertical
displacements
+260
(mm)
+130 mm over 86 days
-165
heave
km subsidence
mod. Stancliffe & van der Kooij, AAPG 2001
MBDCI
Belridge Field, CA - Subsidence

30-40 cm per year
MBDCI
Belridge Rate - z/t
0.0 in./yr
12.5
25.0
over 18 months
MBDCI
Shell Oil Canada Peace River

Surface Multi-
lateral
uplift / tilt CSS
data
reservoir inversion grid

with 50x50m grid cells
ref. Nickles New Technology Magazine, Jan-Feb 2005

MBDCI
NAM, Netherlands - Ameland
- Gas Field
- 3350m reservoir depth
- 22cm subsidence
MBDCI
Measurement Parameters
Precision must be acceptable (5% of zmax)

No systematic errors if possible (random only)
The number of measurement stations must be
chosen carefully, depending on goals
If inversion needed, array designed rigorously
Array must extend beyond reservoir limits to

capture the subsidence bowl
Stable remote benchmark needed, etc.
MBDCI
Compaction Analysis
Prediction, measurement, and analysis is
almost a solved problem nowadays
Good data remain essential
Better coring and lab work needed
Screening criteria should always be applied
Can use subsidence to monitor processes

Casing/cement design to resist compaction
and shear collapse can be greatly improved
MBDCI
Discussion of Some Case Histories

MBDCI
Case Histories
Maracaibo in Venezuela
Groeningen in Netherlands
Niigata in Japan (gas)
Ekofisk in the North Sea (Norway Sector)
Ravenna in Italy (gas)
Many examples elsewhere as well

Good examples in the hydrogeological and
geotechnical literature are interesting
MBDCI
Ekofisk (I)
3000 m deep Chalk reservoir, very thick

Exceptionally high porosities, 48-49% at the
top, 30-35% at the base
Overpressured, v = 65 MPa, po = 54 MPa
Moderate lateral stresses, extensional regime
Chalk slightly cemented

Overlying shales overpressured
Large width with respect to depth (W > 3D)
MBDCI
Ekofisk (II)
Lengthy well tests failed to detect compaction
Subsidence assumed minor because of depth
Casing shearing became a problem in 1980s
Wells had to be redrilled, some twice
Subsidence first noted from platform legs
4.2 m in 1987, predicted max of 6.2 m

Platforms raised, 1987 ($US485,000,000.00)
Subsidence exceeded 6.0 m in early 90s
MBDCI
Ekofisk (III)
2.3 billion $
Redevelopment decision in 1994, S = 6.4 m

Pressure maintenance tried in 1980s, but it
seemed quite ineffective, in use now
More casing shear, most wells redrilled twice
Numerical analysis showed 80-85% of

compaction was appearing as subsidence
Microseismic activity in overburden, along
zones where casing was shearing regularly
MBDCI
Ekofisk (IV)
However, it is a fabulous reservoir!

100% of initial predicted production was
surpassed in early 1990s!
Life predicted to 2011, extended 30 years
Good compaction drive continues (z > 9 m)
Max z now thought to be greater than 15 m

Field may produce more than twice as much
oil as initially thought!
Ekofisk has been a major learning experience
MBDCI
Ekofisk Continues ....

Casing shearing not fully ceased or cured
Will high-angle flank faulting develop?
Redrilling wastes injection (Where? How?)
Surface strains and subsea pipelines: will
there be impairment of these facilities
Oil storage facilities relocated?
Can we reasonable predict these events?

I believe petroleum geomechanics has
advanced enough so that we can predict
MBDCI
Maracaibo, Venezuela (I)

Moderate depth UCSS, thick sequence
30-35% in situ
Lithic to arkosic strata
Geologically quite young
In a monotonically sedimenting basin, no
tectonic compression, no unloading

po slightly above hydrostatic
No cementation, no pre-compaction
MBDCI
Location
MARACAIBO
N
Lago de
Maracaibo
MBDCI
Maracaibo Setting
Sandstone reservoirs
MARACAIBO Late Cretaceous to
Subsidence area
N CABIMAS
Tertiary
Lago de TIA JUANA
Normally pressured
Maracaibo
I LAGUNILLAS High porosity for the
IX
XIV XII
X
II BACHAQUERO
present burial depth
MENE
III
VI
V
VIII
XI
IV
GRANDE Clay cement usually
V
Asphaltene present in
VII
XIII
Central
heavier oils (<30API)
development
areas
MBDCI
Maracaibo, Venezuela (II)

Subsidence up to 6.5 m, broad bowl
Adjacent to the coast, extensive dykes had
to be constructed
Some visible tension cracks developed at
the surface, on subsidence bowl crest
Thermal recovery methods seem not to
have triggered new subsidence of

consequence
Casing loss has been moderate
MBDCI
Ravenna, Niigata
Italy, Japan, some other places
Intermediate depth gas sands, only water
present as a second phase, no oil
High porosity (>30%), arkosic sands,
several stacked reservoirs, water influx
Compaction in the reservoirs, plus water
was expelled from bounding silts and clays

Serious problem was subsidence, as these
are both coastal cities
MBDCI
Wilmington, California (I)
Intermediate depth, many stacked reservoirs

Great aggregate producing thickness
UCSS, porosity > 30%, arkosic
Extensional tectonics (LA Basin)
No cementation, no geological history of

deeper burial followed by erosive unloading
Medium weight oil, liquid reservoir drive
Large area, but edges relatively smooth
MBDCI
Wilmington. California
Bowl shaped
Shear of casings
occurred mainly on
the shoulders of the
subsidence bowl
Few shears in the
middle, where z
greatest
Few on flanks
Associated
earthquakes
MBDCI
Wilmington, California (II)
Sudsidence reached 9.5 m

Minor earthquakes triggered, and in one
case, > 100 casings simultaneously sheared
On the sea coast = great problems with
naval shipyards, inundation
Railway tracks buckled, fissures opened,

buildings cracked sometimes, etc.
Pressure maintenance in 1960s
MBDCI
Little-Compacting Cases
Groeningen, Holland - competent rock

Deeper oil sands, Alberta - low overall
stresses + geological pre-compaction and
mild diagenesis, but no cementation
Faja del Orinoco, Venezuela - thick
quartzose sands, similar to Alberta, so

compaction will not be substantial
We can also learn from these cases
MBDCI
Surface Heave from T & p

Surface heave z
above a SAGD project
320 mm +z
m
1k
Surface heaves cannot be explained by T & p alone: there must be shear dilation
taking place. Therefore, there are massive changes in the reservoir properties k, C c, ,
MBDCI
How Much Compaction?

Depends on compressibility, Z, p, p,
Qualitative screening criteria (geology!)
If porosity > 25% (> 35%is virtually certain)
If the reservoir is geologically young (little diagenesis)
If it is at its maximum burial depth (no over-compaction)
If the mineralogy is arkosic or lithic (weak grains)
If p will be large, and particularly if overpressured

Mainly in extensional regimes and continent margin basins
If largely uncemented by SiO2 or CaCO3
If reservoir width > depth to reservoir (no arching)
Other criteria are probably of little relevance
MBDCI
Will the Reservoir Compact?
All reservoirs compact, but how much?

Best is to test truly undisturbed core samples
in the laboratory under representative
uniaxial and triaxial loading conditions
Failing this, a detailed comparison to other
cases of compaction is carried out (logs, etc.)

Predictions of compaction can be expected
only to be +/-25% at best (sampling
problems, long-term creep, etc.)
MBDCI
Prediction by Comparison
Other case histories are carefully studied

Quantitative comparisons are made:
Geological setting, thickness, etc.
Porosity from cores and density logs
Comparison of seismic velocities (vP, vS)
Study of diagenetic fabric and stress history
Geometry and scale of the reservoir wrt depth

Mineralogy and lithology of the sediment
Stresses, pressures, drawdowns, timing
Other factors?
A probability estimate is made
MBDCI
Reservoir & Overburden
Compaction delay due to reservoir arching

Later, arching destroyed, subsidence starts
If W>Z, 85-90% h transmitted to surface
Strain transmittal to the surface is essentially
instantaneous (if there is no arching)
Geometry is very important (next slides)

Overburden distortion leads to massive
and shear potential (next slides)
MBDCI
Geometry Effects
Everything depends on aspect ratios (W,L,Z)
A deep narrow sand will cause no z
A wide reservoir (W > 1.5Z) will always
transmit compaction to the surface as z
The subsidence bowl is wider than the width
of the compacting reservoir

If very wide, zmax approaches hmax
Simple models OK, but complex geometries
and stacked reservoirs: numerical models
MBDCI
Modeling Compaction
Best approach is a fully coupled flow-

geomechanics simulation (FEM or DD),
giving all stresses and strains directly
Next best approach is a reservoir simulator
coupled to a stress-strain FEM or DD
model, iterating between them to solve z

A simple but limited approach is to get p
from a simulator, calculate V, then project
the V to surface using nucleus-of-strain
MBDCI
Coupling Stresses, Flow
The assumption v = p is usually wrong

It ignores redistribution of stresses in the
reservoir and through overburden stiffness
Thus, a full stress-flow solution is needed:
Calculate p, use in a model (one step)
A model iteratively coupled to flow model

A fully-coupled finite element approach
Use of DD + flow model for is most efficient
Also gives overburden shear stresses changes
MBDCI
Stress Trajectories, -V Case

MBDCI
Overburden Arching
Delay of compaction always occurs in early

time, before p zones intersect (aspect ratio*)
stress arching
initially after some Q
p region
no p yet
*aspect ratio is W/H; if W>~3H, arching is disappearing

MBDCI
Reduced Lateral Stresses

wellbore
h stress trajectories h concentration
far-field stresses
final
h along Zone of high drawdown
h wellbore
Operational consequences:
low pfrac in reservoir
initial
h
higher pfrac above reservoir
Z
Prediction of ij
MBDCI
A flow-coupled geomechanics model is

required to correctly solve for ij and p}
FEM, FEM + FD, DD + FD, Hybrid models
using analytical solutions + FEM, DEM
Material constitutive behavior is critical
Non-linear E (granular and fractured media)?
Potential for shear of weak rocks, fractures?
Fabric changes & yield (grains, shearing )?
Boundary conditions and initial conditions!

MBDCI
Coupled Modeling
Coupling requires that the volume changes
from be analyzed along with p
Only limited closed-form solutions exist
Coupling can be achieved numerically by
(at least) two different approaches:
The complete coupled differential equations are
written and solved, usually by FEM

Or, an iterative approach can be used
Latter is instructive, as it shows principles...

MBDCI
Iterative Coupled Models

Pressures are solved for a single time step
{p} = {pi+1 - pi} calculated in flow model
Assume = p, solve a FEM model
Calculate {V} for all reservoir point
Use {V} as flow model source-sink terms
Get new {p} and iterate until error is small

Take another time step and continue
(Robust and rapid convergence)
MBDCI
Mitigating Casing Shear

Stronger cement and casing are not useful
There are three possible approaches
Avoid placing wells in zones of high shear
Manage reservoir development to reduce
incidence
Create a more compliant casing-rock system
Avoidance & management require modeling

Under-reaming & no cement delays distress
Better sealing cements to reduce p migration
MBDCI
Under-Reaming to Reduce Shear
sand stratum casing cemented, but not

in the under-reamed zone
interface slip under-reamed zone
g
casin
bedding plane slip
shale stratum
MBDCI
Under-Reaming of Hole
100
Unprotected Wells Wilmington
90 (155 total)
80 15" Under-ream
Percent of Total
(5 total)
70 26" Under-ream
(147 total)
60
50
40
30
20
10
0
Undamaged Damaged Failed
MBDCI
Risk Mitigation Approaches

Pressure maintenance
Water injection
Gas injection
CO sequestration, and use as an enhanced oil

2
recovery approach
Structural design of platforms
Judicious placement of wellbore to reduce the

incidence of casing shear
Special completion techniques
Monitor, monitor, monitor
MBDCI
Modeling Compaction
Best approach is a fully coupled flow-
geomechanics simulation (FEM or DD),
giving all stresses and strains directly
Next best approach is a reservoir simulator
coupled to a stress-strain FEM or DD
model, iterating between them to solve z
A simple but limited approach is to get p

from a simulator, calculate V, then project
the V to surface using nucleus-of-strain
Prediction of ij
MBDCI
A flow-coupled geomechanics model is

required to correctly solve for ij and p}
FEM, FEM + FD, DD + FD, Hybrid models
using analytical solutions + FEM, DEM
Material constitutive behavior is critical
Non-linear E (granular and fractured media)?
Potential for shear of weak rocks, fractures?
Fabric changes & yield (grains, shearing )?
Boundary conditions and initial conditions!

MBDCI
Coupled Modeling
Coupling requires that the volume changes
from be analyzed along with p
Only limited closed-form solutions exist
Coupling can be achieved numerically by
(at least) two different approaches:
The complete coupled differential equations are
written and solved, usually by FEM

Or, an iterative approach can be used
Latter is instructive, as it shows principles...

MBDCI
Iterative Coupled Models

Pressures are solved for a single time step
{p} = {pi+1 - pi} calculated in flow model
Assume = p, solve a FEM model
Calculate {V} for all reservoir point
Use {V} as flow model source-sink terms
Get new {p} and iterate until error is small

Take another time step and continue
(Robust and rapid convergence)
Analyzing Special Compaction MBDCI
Joints
FracPacPacker 9,743
9,818
Telescoping joint
Telescoping Joint
450 ft
9980
Screen, basepipe, for = 0.2,
210 ft
Screen, basepipe, couplings
couplings Cp = 910-6 psi-1
p = 2600 psi
H/H = 1%
10,192
Sump packer
Sump Packer
MBDCI
Mathematical Modeling of Strains

Modeling a 600 Compacting MBDCI
Section
F L A C 3 D 2 .1 0 J o b T it le : K W 1 : 6 0 0 ' W e llb o r e M o d e lin g
S t e p 4 7 0 0 0 M o d e l P e r s p e c t iv e V ie w T it le :
0 4 :2 4 :5 6 S u n O c t 2 0 2 0 0 2
C e n te r: R o t a t io n :
X : 9 .0 6 2 e + 0 0 0 X : 3 .8 6 2
Y : 4 .6 1 0 e + 0 0 1 Y : 3 5 9 .9 8 9
Z : - 1 .2 0 1 e + 0 0 5 Z : 3 5 9 .8 2 8
D is t : 1 . 9 9 1 e + 0 0 4 M a g .: 1 3 .7
A n g .: 2 2 .5 0 0
B lo c k G r o u p
B a s e _ P ip e
G ra v e l
C a s in g
cem ent
S h a le
S and
S c re e n
C o u p lin g
T e r r a lo g T e c h n o lo g ie s U S A , I n c .
A r c a d ia , C A 9 1 0 0 6
MBDCI
Elastoplastic Zone Generation

F L A C 3 D 2 .1 0 J o b T it le : K W 1 : 6 0 0 ' W e llb o r e M o d e lin g
S t e p 2 0 5 6 1 6 M o d e l P e rs p e c t iv e V ie w T it le :
1 1 :4 0 :4 6 F ri N o v 0 1 2 0 0 2
C e n te r: R o t a t io n :
X : 1 .2 4 1 e + 0 0 1 X : 0 .0 0 0
Y : 4 .6 1 0 e + 0 0 1 Y : 0 .0 0 0
Z : -1 .2 0 2 e + 0 0 5 Z : 0 .0 0 0
D is t : 1 . 9 9 1 e + 0 0 4 M a g .: 369
A n g .: 2 2 .5 0 0
C o n to u r o f e s _ p la s tic
M a g fa c = 0 .0 0 0 e + 0 0 0
G ra d ie n t C a lc u la t io n
7 .5 9 8 8 e -0 1 0 to 5 .0 0 0 0 e -0 0 3
5 .0 0 0 0 e -0 0 3 to 1 .0 0 0 0 e -0 0 2
1 .0 0 0 0 e -0 0 2 to 1 .5 0 0 0 e -0 0 2
1 .5 0 0 0 e -0 0 2 to 2 .0 0 0 0 e -0 0 2
2 .0 0 0 0 e -0 0 2 to 2 .5 0 0 0 e -0 0 2
2 .5 0 0 0 e -0 0 2 to 3 .0 0 0 0 e -0 0 2
3 .0 0 0 0 e -0 0 2 to 3 .5 0 0 0 e -0 0 2
3 .5 0 0 0 e -0 0 2 to 4 .0 0 0 0 e -0 0 2
4 .0 0 0 0 e -0 0 2 to 4 .5 0 0 0 e -0 0 2
4 .5 0 0 0 e -0 0 2 to 5 .0 0 0 0 e -0 0 2
5 .0 0 0 0 e -0 0 2 to 5 .5 0 0 0 e -0 0 2
5 .5 0 0 0 e -0 0 2 to 6 .0 0 0 0 e -0 0 2
6 .0 0 0 0 e -0 0 2 to 6 .5 0 0 0 e -0 0 2
6 .5 0 0 0 e -0 0 2 to 6 .7 3 6 9 e -0 0 2
In te rv a l = 5 .0 e -0 0 3
T e rra lo g T e c h n o lo g ie s U S A , I n c .
A rc a d ia , C A 9 1 0 0 6
MBDCI
The Design Paths
Reservoir Well
Deformation Well Deformation Common
Analytical Well Common
Analytical Estimate Performance
Performance Limits Design
Analysis Design
Tool Comparison
Comparison Analysis
Analysis
Tool Tool Database
Tool
Simple
Simple
Decision
Decision
Analysis Tool
Proprietary Proprietary Proprietary

Proprietary
Proprietary Proprietary
Reservoir Decision
Decision Well
WellDamage
Damage
Analysis Analysis
Analysis Analysis
Analysis
Analysis
Optimum Well Design

MBDCI
Combining the Elements
Well Damage Analysis

and Design Process
Proprietary Proprietary
Reservoir Well Damage
Analysis Analysis
Optimum Well Design

MBDCI
Decision Analysis Techniques

An economic decision tree model can be applied to compare the costs and
benefits of alternative well designs, while taking into account the inherent
uncertainties in geomechanical model input data, well damage location, and
effectiveness of various mitigation strategies.
In some instances the appropriate action is not to change completion design
and simply accept and account for damage risk in economic projections.
Input Parameters P10 P50 P90

Formation Depth (ft below mudline) 1.40E+04 1.60E+04 1.80E+04 Net Sand/Gross ratio
Areal Extent (acres) 2.00E+03 4.00E+03 6.00E+03

Gross Formation Thk. (ft) 3.00E+02 4.00E+02 6.00E+02 Sand Compaction (1/psi)
Net Sand/Gross ratio 6.50E-01 7.50E-01 9.00E-01

Sand Compaction (1/psi) 1.10E-06 3.30E-06 6.60E-06 Shale Compaction (1/psi)
Shale Compaction (1/psi) 1.00E-07 3.30E-07 1.00E-06
Poisson's Ratio 1.50E-01 2.50E-01 3.50E-01 Pressure Drawdown (psi)
Pressure Drawdown (psi) 2.00E+03 2.50E+03 3.00E+03
Well Inclination (deg from vertical) 1.00E+01 2.00E+01 4.50E+01 Well Inclination (deg from
vertical)
Overburden Young's Modulus (psi) 5.00E+05 1.00E+06 1.50E+06
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Percent Maximum Axial Strain

MBDCI
Example of Decision Analysis

Model Input Parameters
Completion #1 Cost 180000 dollars Initial damage risk 0.10
Completion #2 Cost 195000 dollars Risk reduction 0.50
Completion #3 Cost 220000 dollars Risk reduction 0.75
Avg Daily Production 30 bbls/day
Price per barrel
Replacement Time
20
60
dollars
days
Probability x Consequences =
Abandonment Cost 20,000 dollars Risk Cost
Damage Cost Risk Cost
Replacement Well 180000 $18,000

Well Damage Risk 0.1 Lost Production 36000 $3,600
Abandonment Cost 20,000 $2,000
Completion #1 Added Cost 0 $0
Total Risk Cost $23,600


Well Damage Risk 0.05 Lost Production 36000 $1,800
Abandonment Cost 20,000 $1,000
Completion #2 Added Cost 15000 $15,000

Well Damage Risk 0.025 Lost Production 36000 $900
Abandonment Cost 20,000 $500
Completion #2 Added Cost 40000 $40,000
Simple Decision Analysis Example

17 Compaction GeomechanicsAA

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

17 Compaction GeomechanicsAA

Diunggah oleh

Hak Cipta:

Format Tersedia

MBDCI

Important in Diatomite (California mainly)

Recompaction is important in cyclic steam

stimulation (porosity cycling)

The Case of Ekofisk

Large drawdowns are

Surface subsidence from deep compaction

Major Design Steps

Based on core examination and lab tests

Based on geomechanics analysis

Based on monitoring information

Study the cost-benefit gain to the company

Geomechanics assessment Impact assessment Mitigation options

Risk assessment Cost-benefit analysis Cost of options

Decision making Is compaction

Thus, compaction moves out

delays full z development

These factors affect the permeability, often

Delay of compaction always occurs in early

initially after some Q

z Delay Through Arching

full when zones meet,

stresses now flow without arching around zones

Productivity can decline with an increase in

The relative importance depends on the

Casing shear or buckling can occur

Changes in the effective stress can trigger

Much more oil is driven to the wellbores

In high Chalk, a volume change as much

as 10% can take place, all oil production

Also, a direct fluid expulsion occurs (V)

Elastic and Inelastic Strain

inelastic compaction curve

1 MPa 10 MPa 100 MPa

At high p (high ), grain splitting,

Processes and Compaction

elastic grain rearrangement

curve grain crushing

Mechanisms, Chalk (I)

This is triggered by the increase in v, and

Why Does Chalk Collapse?

Hollow, weak grains (coccoliths)

Weak, cleavable grain

Mechanisms, Chalk (II)

The whole reservoir compacts when the

North Sea Chalk Collapse

irrecoverable strain collapse of fabric

rebound curves compression curve

1 MPa 10 MPa 100 MPa

Overcompaction (deep burial history) can

normal collapse if cement

These are hard to quantify without careful

Compaction can release fines from strata

Itappears that a homo-

Fracture Permeability Loss

It is important in coal, North Sea Chalk, but

Grain Crushing and k

feldspar crushed, quartz intact

h stress trajectories wellbore h concentration

Shear, Fracture Opening

Pressure decline leads to an increase in the shear stress

This leads to increases in permeability, better reservoir drainage

Pore Blockage Mechanisms

Most Sensitive Cases