Vous êtes sur la page 1sur 31

Chapter 3

Well Planning

Chapter overview
Risk analysis in well construction
The choice of drilling fluids to minimize
reservoir damage
Predicting wellbore stability
Geosteering

WELL PLANNING

CHAPTER OVERVIEW

Optimizing well production and


reservoir understanding from well data (the
theme of this book) depends heavily on the
quality of the wellsquality in the sense of
being fit for the purpose, which in this case
is to give an expected rate of production
and/or amount of data from the reservoir.
These data must be acquired at a minimum
cost consistent with safe practices and
minimum impact on the environment.
Quality undoubtedly depends on good
planning. This chapter discusses some
aspects of this planning, in particular the
need to look at the whole picture, and the
particular case of horizontal wells.
Well construction is increasingly focused
on ensuring an optimum return on each well
by taking into account the needs of the many
stakeholdersdrillers, production engineers,
geologists, geophysicists, and petrophysicists,
as well as investors and managers. A key
issue is how to reconcile these needs and,
when conflicts arise, how to judge one
against another. One method is risk analysis.
The first section describes the process of risk
analysis and points out how geoscience
needs can be included along with the
traditional drilling needs.
Drilling fluids can have a strong effect on
production rate and data acquisition. The
problem of choosing the right fluid is a very
good example of how the requirements of
different groups need to be considered and
balanced against one another. The choice is
not always obvious; for example, using a
fluid that causes formation damage may or
may not be important, depending on the
completion and stimulation plans. This article
discusses the factors involved and shows
results from the sort of laboratory studies that
can help quantify these factors better.

Horizontal wells have spurred much of


the recent focus on well planning. The
placement of the well, stability of the
borehole, and effects of a very long reservoir
section are some of the issues that require
more precise and integrated planning. The
next section shows how data from a pilot
hole were used to plan the drilling and
completion of a horizontal well. In particular,
the pilot hole showed that if the well was
drilled parallel to the minimum stress it could
be completed openhole, thereby considerably reducing the investment. The well was
successfully drilled and completed,
producing at double the rate for a typical
vertical well in the area.
The final section discusses the technique
of steering a horizontal well geologically
rather than geometrically. This technique
depends on good planning, good wellsite
software and coordination and on having
formation measurements as close as possible
to the bit. An important part of planning is to
model the response of the measurements for
various possible scenarios so that if these
occur while drilling, the correct decisions can
be made more easily. The examples show
how wells can be steered successfully within
sands as thin as 5 ft for more than a thousand
feet of section.

R I S K A N A LY S I S I N W E L L C O N S T R U C T I O N

A change in general philosophy


concerning drilling or well construction has
taken place in the past several years. Along
with industry right-sizing, integrated services,
and increased global competitiveness has
come a way of thinking that keeps the end
in mind. Historically, the drilling profession
has been segregated not only functionally
from geoscience, reservoir, and production
counterparts, but, more importantly, in their
goals and objectives. Drillers have been
primarily interested in well designs that
resulted in getting the well down quickly,
with problem-free operations and the lowest
possible cost. Geoscientists, on the other
hand, were interested in obtaining as much
information as possible, reducing damage to
the reservoir, and having numerous testing
and completion options available to them.
With the changes that have taken place
in the last several years, both relating to
economics and the work force, we must now
concentrate on the big picture. Drillers must
concentrate not only on drilling performance,
but the effect their designs have on data
acquisition, reservoir characteristics and
ultimate well performance and economics.
Geoscientists, reservoir engineers, and
production engineers must now concentrate
not only on optimum reservoir performance,
but operational risks and economic/
operational tradeoffs in well design. Clear
and concise goals and objectives for each
project must be in place and all disciplines
must work together to economically achieve
them. These goals and objectives must be
focused to produce the optimum return on
each well or project undertaken. This requires
an understanding of how each disciplines
problems and/or concerns affect the projects
ultimate economics.
This section starts by describing the
overall objectives of well construction, the
priorities in well construction and the concept
of risk/reward. Example A then introduces
decision trees and shows how design

decisions have typically been made in the


past by drilling engineers with cost
information only. Next, one particular area of
well construction designthe wellbore
geometryis described and used as the basis
for Example B. In this example, the decision
tree employs information from various
disciplines and reveals the value of this
information. The designs are based on more
than just cost data and result in designs with
the end in mind. This technique allows the
melding of each disciplines expertise into a
more accurate decision-making process.
Well construction
Design hierarchy

In any operation, a hierarchy of


importance must be placed on all assets or
resources so that the decision-making process
is consistent with critical items such as safety
and technical integrity. Prioritizing these
resources can best be described by a simple
statement that has been used for quite some
time in drilling organizations: the PEOPLE,
the RIG, the WELL. Well design, operations,
and the decision-making process must all
follow this hierarchy (Fig. 3.1).
The safety of personnel is obviously the
number one goal in any operation we
undertake. Personnel safety should always be
the number one concern in well design and
rig-site operations.
The drilling rig represents a resource that
has a fixed value and thus is considered
higher in priority than the well under
construction. Additionally, rig safety policies
are, for the most part, consistent with
personnel safety. Any attempt to keep
problems away from the surface, and thus the
rig and the environment, aid in the ultimate
goal of personnel safety. Any risk of
personnel safety while attempting to salvage
the physical equipment of the rig would be
counter to the hierarchy.

WELL PLANNING

Last in our hierarchy is the well under


construction. No fixed value can be assigned
to the well and, therefore, it is placed under
the drilling rig in the hierarchy. Many safety
hazards can be eliminated prior to reaching
the surface (thereby keeping them away from
the rig, the personnel and the environment)
by forfeiting the wellbore. In many instances,
forfeiture of the wellbore allows salvaging the
drilling rig and ensures the safety of
personnel.
Therefore, the well must be designed
both technically and operationally consistent
to the hierarchy described. Getting the well to
the depth objective, evaluating the wellbore,
and finally completing the wellbore then
become the secondary hierarchy. This
grouping is fairly obvious. For example,
being able to evaluate or complete a well is
of no consequence if the well cannot reach
the objective formation. Each of these
(reaching depth objective, evaluating, and
completing) must be planned for with
reasonable assurance.

Figure 3.1

PEOPLE
RIG
RIG
WELL
WELL
The hierarchy of importance
used in drilling design.

Figure 3.2

Estimated well costs


$1,450,000

(0

.1
5)

Risked well costs

$1,000,000

M
(0.65)

$1,037,500

Reasonable assurance

A well plan that provides a 99% chance


that all objectives will be accomplished is
usually both an uneconomic and unrealistic
plan. This is where engineering, risk
management, and operational experience
play an invaluable role in well construction.
Options relating to mud systems, rig
equipment, borehole geometry, completion
and testing methods, cement design, extent of
evaluation, and others must be examined and
a reasonable level of assurance obtained in
the final design to meet the goals and
objectives of the well. In most cases,
operational experience is the most valuable
resource for success in this endeavor.
Obviously, reasonable assurance can
take on various definitions as we apply
different goals, objectives, potential
economic returns, and well types. For
instance, the goals and objectives for a
$100,000 shallow development well are
somewhat different from those of a
$20,000,000 exploration well in the jungle.
Therefore, reasonable assurance will
undoubtedly take on a different meaning.
For instance, a more expensive and complex
drilling fluid is justified in the latter case due
to the increased exposure versus that of a
shallow development well. Although this
example is extreme, others fall within a wide
spectrum of possibilities.

er

as

0)

us

.2

ua
l

(0

$850,000

Operational risk assessment:

Se

tl

in

Well construction design: Using


risk/reward/economics

El

$1,000,000

at

(0

in

.2
0)

im
H

(0
.1
5)
.2
5)

(0.3
5)
$1,847,500

M
(0.60)

$1,750,000

(0

.1
5)

A typical decision tree analysis for eliminating a casing point.

$775,000
$2,200,000

Fai

$825,000

(0

er

ed
c ce
Su
5)
6
.
(0

lin

$1,200,750

M
(0.65)

$852,500

$1,650,000

Various preliminary designs that meet the


goals and objectives of the well, as well as
health, safety, and environment (HSE) and
technical requirements must be reviewed. All
of the preliminary designs will have inherent
advantages and disadvantages. Each design
must be evaluated on the basis of operational
merit, risk/reward, and economics. Historically, operational merit has been used to
select the final design without much input
from other disciplines or extensive economics.

R I S K A N A LY S I S I N W E L L C O N S T R U C T I O N

Figure 3.3

20 in. @ 300 ft

20 in. @ 300 ft

1712 -in. hole

1338 -in. @ 4000 ft

958 -in. @ 4000 ft

1214 -in. hole

Lost circulation
zone
958 -in. @ 8500 ft

Reservoir

7-in. @ 10,500 ft

Two well construction options.

7-in. @ 10,500 ft

For years, decision trees much like Example A


have been employed by drilling engineers to
aid in operational decision-making. Although
very simple in nature, this example provides
the basics of decision tree analysis.
Example A: Eliminating
a casing point
Figure 3.2 shows a common decision tree
analysis. In this simple example, the option to
eliminate an intermediate liner is being evaluated. Typically, this liner is set to cover a lost
circulation zone above the target reservoir (Fig.
3.3). The engineer is evaluating the use of a
new lost circulation material in conjunction
with improved drilling practices in attempts to
eliminate the string. First, lets examine the
lower branch that relates to eliminating the
liner.

The possible outcomes (success and


failure) are identified and the appropriate
decision tree branches drawn. Next, an
estimated cost is calculated for each possible
outcome (high, medium, and low). Notice
that successful elimination of the liner results
in a lower cost relative to the same cost
outcome (high, median, or low) of setting the
liner as usual (upper branch). This
improvement is what the engineer is looking
for and is the result of reduced casing,
cementing, and time costs. Once these steps
are taken, the engineer assigns probabilities
for each of the possible outcomes and cost
scenarios. In this case, the engineer has
assigned a 65% probability of success for
eliminating the liner. Probabilities for high
well cost, median well costs, and low well
costs have also been assigned. Notice that the
probability for an above-average well cost is
more in the case of failure than in the case of
success. This is normal as the risk/unknown
is greater in this case. (Note: this method
requires probability inputs that are somewhat
arbitrary. This does not render the technique
invalid, as a sensitivity analysis is typically
done and the results incorporated into the
decision-making process.)
The top branch describes the decision to
set the liner across the lost circulation zone as
normal operations dictate. Again, an
estimated cost is calculated and probabilities
assigned to them. Notice that these values are
somewhat different than those assigned to the
high, median, and low cases in the bottom
branch. This is because of uncertainty in
eliminating the casing string. When
uncertainty exists, the probability for the
actual cost being higher than average should
be greater.

WELL PLANNING

The engineer now applies the probability


factors to each of the costs and arrives at a
risked cost for eliminating the liner as well
as setting the liner as usual. If necessary, a
sensitivity analysis is done to more completely evaluate the data. Sensitivity analysis
is usually reserved for more complex trees
where multiple options are present. In this
case, setting the casing as usual has resulted
in a lower risked cost as compared to
attempting to eliminate the liner.
In past years, designs with little regard to
the overall picture were sometimes selected
because of cost or operational advantages. In
todays environment, the reservoir and
economics in general must play a crucial role
in final design selection. Decision trees, risk
analysis concepts, and input from the various
disciplines can be used to select the design
that yields the maximum return on the
resource.
The challenge of well construction is
finding the optimum design for each
operation, a design that within reasonable
assurance accomplishes the goals and
objectives, is economic, and ensures the
safety of both rig and personnel. This
responsibility now must be shared by all
appropriate disciplines instead of by the
drilling engineer alone.
Discussed below is one particular design
the wellbore geometry configuration. An
example decision tree shows how a particular
design can benefit from the input of all the
appropriate disciplines and the resulting
economic analysis.
Wellbore geometry
Numerous factors influence the selection
of wellbore geometry. As a result, many
options are usually evaluated to select the
configuration that economically achieves the
stated goals of the well. Following is a partial
list of factors usually considered when
designing wellbore geometry.

problem zones (lost circulation, pore


pressure, instability, etc.)
requested completion (size of tubulars
and other equipment)
equipment availability
geological and reservoir data (reservoir
structure and composition, targets, etc.).
The geometry of the borehole is always
designed from the bottom up. Consideration
is given to the number of strings needed to
reach the final objective within reasonable
assurance. Depending on the type of well
and, in some cases, economics, a contingency
string may be allowed for in the preliminary
design. This particular portion of the design is
usually done when all of the information
from the above designs are gathered, as all
influence the configuration of the borehole.
Below is an example of how risk analysis can
be used to select wellbore geometry. In this
example, the accuracy of assessing the
position of a geological target is being
evaluated in a horizontal well application.
The optimum wellbore geometry will be
selected using drilling, geological, and
directional expertise combined in a risk
analysis approach.
Example B: Deciding to
drill a pilot hole
Example A was one of an operational
decision that has typically been made by well
construction (drilling) personnel solely on
cost information. Example B, on the other
hand, uses cost information in conjunction
with information provided by the geologist,
directional personnel, and others. The
additional input results in more accurate
decision making.
In this example, the team is attempting to
decide whether or not a vertical pilot hole
should be drilled prior to constructing a
lateral in a horizontal prospect (Fig. 3.4).

R I S K A N A LY S I S I N W E L L C O N S T R U C T I O N

Figure 3.4

Construct the build


and lateral sections
without a pilot hole.

Reservoir target

Construct a vertical
pilot hole, log, then
plug back and construct
the lateral.

Construct a pilot hole at


45 , log, then plug back
and construct the lateral.

Three possible options for drilling a horizontal well.

Figure 3.5

Estimated well costs


.4
0)

$3,500,000

(0

Risked costs

M
(0.35)

$3,075,000

ol
e
th

.2
0)
(0
H
M
(0.50)

lo

0)
.3

pi

$2,500,000
$2,500,000

(0

$3,000,000

$2,300,000

L
$2,150,000

.1
5)

im

in

at

$3,100,000

Drill 45 pilot hole

(0

El
$2,842,500

5)
.2

$2,888,000

(0

ed
eed
nn
c ti o
e
)
r
5
Cor
( 0 .7
No
c or
r ec
tion
nee
(0.2
ded
5)
$2,295,000

M
(0.65)

$2,850,000

Dr

.
(0

ill

)
20

ve

$2,650,000

r ti
ca
lp
e
ol
th

ilo

(0

.1
0

$3,300,000

$2,970,000

M
(0.60)

$3,000,000

L
(0
0)
.3

$2,800,000

A decision tree analysis using input from geoscientists, directional personnel and
others, as well as drilling engineers.

The reservoir is 8 ft thick and requires a


significant amount of geologic knowledge to
enter the reservoir target. Historically, these
teams have used a pilot hole drilled through
the reservoir to identify geologic tops so that
the curve could be designed and built with a
high degree of accuracy. The information
obtained by drilling and the subsequent
logging of the pilot hole has enabled the
construction of the lateral as desired with little
need for significant course correction.
From experience in the area and by using
measurements at the bit, the geologist
believes that the reservoir top could be picked
with reasonable certainty and the well cost
potentially reduced by eliminating the pilot
hole. The alternative of constructing a pilot
hole at 45, logging, and then correcting the
build rate is also being evaluated. Each of
these three alternatives have both monetary
and operational advantages and disadvantages
that need to be evaluated prior to selecting a
wellbore configuration (Fig. 3.5).
The top branch has been dedicated to
eliminating the pilot hole. The drilling engineer, with consultation from directional personnel and offset data, feels that without a
pilot hole the need for a significant course
correction is 75%. Course correction is being
defined as a probable plugback and redrill for
a portion of the hole. Note that the probability
for a higher-than-average cost is greater (40%
versus 20%) on the correction needed
branch than on the no correction needed
branch. This, as in the other examples shown,
is because of the uncertainty involved in unfamiliar activities. The middle branch relates to
penetrating the reservoir at a 45 angle,
logging, then constructing the final portion of
the curve. The bottom branch represents
constructing the vertical pilot hole, plugging
back, then building the curve (the conventional plan). The probability in the bottom
branch for high cost (10%) is lower than that
of the other options because this method has
been used before, and the cost estimates have
more basis.
3

WELL PLANNING

Figure 3.6

Reservoir
Reservoir characterization

Production
Production rates and
flowstream data

Geophysics
Seismic information
and interpretation

The boardroom of well


construction.

Geology
Structure and targets

WELL
CONSTRUCTION

Completion
Subsurface equipment
requirements

Petrophysics
Formation properties

After the branches were constructed, the


drilling engineer calculated the estimated cost
for each outcome and placed them in the tree
accordingly. Costs and risks associated with
the following were included:
geologic uncertainty for each case
setting/sidetracking off openhole cement
plug(s) in case of course correction
rig time associated with each possible
outcome
and others.
Once risked, the decision tree revealed
that the lowest risked cost is the case of
penetrating the reservoir at 45, logging, then

constructing the remainder of the curve.


Again, the decision (as well as the cost estimates) has been oversimplified for presentation purposes. In this case, the pilot hole
might well be eliminated as there is little difference between its cost and the 45 pilot
hole. However simplified, it demonstrates the
need for decision evaluation with input from
all disciplines.
Conclusion
In reality, net present value (NPV) is used
more often than cost data alone. Cash flowstream calculations are performed with
reserves, expected revenue, cost estimates,
taxes, and others to arrive at NPV for each
option. Once NPVs are assigned to the various
options, they are risked with the probability
factors assigned. With this method, the option
with the largest NPV is the optimum choice. In
either case the important point is that modern
well construction design takes into account
the goals of all disciplines (Fig. 3.6). The
decision tree is one method by which these
different goals can be quantified and put
together to arrive at a reasonable assurance
that they are met.

THE CHOICE OF DRILLING FLUIDS TO MINIMIZE RESERVOIR DAMAGE

Introduction
The current nearly obsessive interest in
minimizing formation damage from drilling
fluids is a direct result of the increased
exposure of producing formations to drilling
fluids by horizontal drilling. This is as it should
be. Perforating or fracturing procedures in
vertical wells that penetrate feet or meters
deep into a formation often render irrelevant
the effects of invading drilling fluid within a
radius of inches in the wellbore, but the nearwellbore effects of fluid in openhole
completions can be highly significant. The
object of the mud engineers efforts should be
3

to design and run a drilling fluid for a


production well such that it is possible to
demonstrate with production data that, at least
with respect to the effects of drilling fluid, the
well produces to its maximum potential.
Laboratory studies
Data that unequivocally demonstrate this
are in short supply. Usually, the best we can
do is design a well simulation that helps
predict what the effects of various muds will
be, and apply that knowledge when drilling.
While not all of the mechanisms of formation
damage are mud-related, the majority are,

THE CHOICE OF DRILLING FLUIDS TO MINIMIZE RESERVOIR DAMAGE

Figure 3.7
kro

krw

Kro, Krw

0
0

(a)
Mud solids invasion
Filter cake plugging
Formation fines migration
Clay swelling
Polymer adsorption
precipitation
Scale formation
Wax formation (paraffin,
asphaltine)
Sludge formation
Stress-induced permeability change
Perforation plugging

A summary of the
mechanisms of formation
damage. The heavy type
indicates mud-induced
damage mechanisms.

Sw
(b)

Wettabililty change
Fluid saturation change
(blocking)
Emulsion formation
Water coning
Gas breakout
Condensate banking

and their prevention must be a major design


factor in most horizontal wells (Fig. 3.7).
However, the tools we use to predict
formation damage from drilling fluids are
flawed. Standardization of methodology
between laboratories is nonexistent. When
possible, cores from the actual formation in
question are used; more often standard core
material is used, e.g., Clashach and Birchover
Sandstone. There are at least six variables in
the experimental procedures for testing
return permeability that will affect the results:
1. A major variable is the decision to
compare permeabilities based on a given
time period or on a given volume
throughput. The results may vary
enormously depending on this choice of
parameter.
2. Core lengths and diameters vary greatly.
These decisions are largely based on a
decision on how much time should be
devoted to each experiment.
3. Core preparations vary greatly. In some
laboratories, the cores are cleaned with a
methanol-toluene or similar solvent
system to remove asphaltenes prior to
testing. Other labs use the cores as is.

4. The experimental conditions with


respect to pressure and temperature vary
greatly between labs.
5. Post-test procedures also vary. Some
laboratories simulate completion cleanup
steps with backflow pressures, which
also vary; others do not.
6. The drilling fluids themselves will perform differently depending on whether
or not the test mud formulations included drilled solids (real or simulated).
Given the range of possible mud
formulations (i.e., with the same name but
with and without drill solids), varying mud
weights with different weighting agents,
along with the range of possible test
conditions, laboratory data suggest that
relatively small changes in drilling parameters
or downhole conditions can radically affect
our efforts to choose the right fluid for a
particular formation.
As a result, the choice of drilling fluids is
often made by simple intuition; i.e., what
seems to make sense for a formation. Our
intuition leads us to believe that:
1. Saturated, sized salt drilling fluids are
formation-friendly, because it makes
sense to assume that salt particles that
may enter a formation will also be
dissolved away when exposed to
unsaturated fluid.
2. Polymer fluids with calcium carbonate as
a weighting/bridging agent might invade
a formation, but should easily be
removed with acid.
3. Any fluid that contains bentonite should
be avoided, because it seems likely that
fine bentonite particles will invade a
formation, hydrate and be difficult or
impossible to remove, thus impeding
hydrocarbon flow through the formation.
4. The use of barite should be avoided
because barite particles cannot be
removed from a formation by dissolution
in acid or water.

WELL PLANNING

Figures 3.8 & 3.9

100

Percent of initial permeability

Percent of initial permeability

100

50

Static

Dynamic

50

Static

Dynamic

Recovery in permeability for

Recovery in permeability for 13 ppg fluids,

10.5 ppg fluids Clashhach

Clashach Sandstone, 60 min filtration:

Sandstone, 60 min filtration,

MMH weighted with CaCO3, Polymer

MMH weighted with

weighted with CaCO3, Saturated sized salt

CaCO3, MMH weighted


with barite, Polymer
weighted with CaCO3,
Saturated sized salt,
KCl polymer weighted

with CaCO3

Recent data suggest that our intuition is


wrong. Not only does our intuition fail to
identify the most cost-effective solution to drilling a formation without damage, data suggest
that what we intuitively identify as the least
damaging, can be the most damaging fluid.
Why do observed data contradict what
seems like common sense? It may be that the
simple model to which our intuition leads us
is wrong. Our intuitive model fails to take into
account the true nature of the effects of a

Figures 3.10 & 3.11


100

Percent of initial permeability

Percent of initial permeability

100

50

Static

Dynamic

(a)

(b)

(c)

(d)

50

Different backflushing

Recovery in permeability for 10.5

Recovered permeability for Clashach Sand-

ppg fluids, Birchover Sandstone,

stone, 300 psi differential, 240 min

60 min filtration: MMH weighted

a) when filter cake left in situ, b) with 2 psi

with CaCO3, Polymer weighted

backflush pressure c) with 6 psi backflush

with CaCO3, Saturated sized salt

pressure d) with 10 psi backflush pressure:


MMH weighted with CaCO3, Polymer

weighted with CaCO3, Saturated sized salt

drilling fluid on and in a formation: filter cake


formation, particle invasion (added solids
and drilled solids), filtrate invasion (water, oil,
polymer), and the reverse of these: filter
cake, particle and filtrate removal.
Tests were conducted recently at
Schlumberger Cambridge Research (SCR)
comparing formation damage, as measured
by permeability after static and dynamic filtration, with three different types of fluids. The
results were somewhat surprising. Static filtration tests with 10.5 ppg fluids resulted in
similar (within experimental error) permeability recoveries from mixed-metal hydroxide
(MMH ) weighted with barite, MMH weighted
with calcium carbonate (CaCO3), polymer
mud weighted with CaCO3 and sized salt.
Only the results with potassium chloride (KCl)
polymer mud weighted with CaCO3 were
significantly lower. Dynamic filtration, on the
other hand, gave the best permeabilities with
MMH weighted with barite, and polymer
weighted with CaCO3, followed by the others
(Fig. 3.8). The same fluids in the same
formation, but weighted to 13 ppg, gave
different results: the best return permeability
was observed with the polymer fluid,
followed by saturated, sized salt and MMH
weighted with CaCO3 (MMH with barite was
not tested) (see Fig. 3.9).
The same fluids at 10.5 ppg were tested
in the far less permeable Birchover sandstone
(Clashach permeability averaged 644 md and
average pore size was 26.6 microns;
Birchover averaged 12.5 md permeability and
average pore size was 2.64 microns). All gave
complete (100%) return permeabilities (Fig.
3.10). The effect of applying simulated
reservoir pressure to remove mudcake
decreased return permeability using the sized
salt system more than with either MMH or the
polymer fluid (Fig. 3.11). When cores were
backflushed at a variety of pressures, the
sized salt system showed less recovery than
the others, especially at the lower pressures,
again suggesting the resistance to cake
removal with the sized salt system.

THE CHOICE OF DRILLING FLUIDS TO MINIMIZE RESERVOIR DAMAGE

Figure 3.12

A stable, effective, external filter cake (cake on top, formation underneath).


Figure 3.13

Partially internal and external filter cakes with sized salt mud.
Figure 3.14

Photographs of the cores used in the


above experiments help us make adjustments
to our intuitive model of how drilling fluids
affect formation damage. It is true that the
invasion of solids along with filtrate is a major
mechanism of formation damage. As any
mud engineer knows, the quality and
quantity of mud solids determine the quality
of the filter cake and quantity of fluid loss in
time. In the same way, the nature and amount
of solids in the fluids determines what does
and does not invade the formation, both solid
and liquid. Again, mud engineers know that
the ideal filter cake material is a mixture of
different size particles, including the flat clay
platelets that overlap each other to reduce the
permeability and porosity of the filter cake.
The MMH-bentonite complex contributes to
this feature by preventing invasion of the clay
particles. The result is a stable, effective
external filter cake (Fig. 3.12); i.e., a barrier
external to the formation that impedes
invasion of both solids and filtrate to the
formation. In some cases, the contrast with
the sized salt system is observable. The sized
salt system tends to form both external and
internal filter cakes because of the nature of
the bridging material that allows invasion by
salt crystals (Fig. 3.13). Often, the difference
is irrelevant because the salt may be easily
removed with undersaturated fluid. There is
reason to believe, however, that the polymercoated salt crystals are, in some cases, not so
easy to remove and persist in the formation,
impeding hydrocarbon flow (Fig. 3.14).
But do the solids that invade the
formation and stay there necessarily reduce
permeability significantly? Not at all.
The right size solids in relation to pore
throat size bridge the formation and result
in less permeability damage. In that case,
one would not want the salt or calcium
carbonate particles to be removed.
One would want them to continue their

Possible polymer-coated salt crystals within the formation.

3 10

WELL PLANNING

Figure 3.15

VISPLEX filter cake in the


process of being removed at
4 psi back pressure.

bridging function at least until the external


filter cake can be totally removed during a
completion or production phase (Figs. 3.15
and 3.16). However, on one point everyone
seems to agree: ultra-fine, low-density solids
are no good for either drilling fluid or
formations.
WBM versus OBM and
particle size distribution
For some operators and in some
situations the choice of fluid can be a moot
point. The difficulties and economics of
drilling high angle wells in deep, hot formations may dictate the use of Oil Based Mud.

Figure 3.16

Core surface after filter cake in Fig. 3.15 has parted.

3 11

The increased risk of stuck pipe and the


likelihood of exposing the formation for
longer periods of time to water based mud
(WBM) make the case for OBM a strong one.
Indeed, it is not uncommon for an operator to
have an intuitive preference for, for example,
salt saturated mud, when practicalities
encourage the use of OBM. It is not surprising
that the particle size distribution of the fluid,
an exceedingly difficult parameter to control,
and one which we are only recently starting to
measure with regularity and accuracy, is a
crucial factor in the results of return permeability tests. One would expect, for example,
that return permeabilities would change when
the average particle size is greater than ultra
fine, i.e. larger than 8-10 microns. It is worth
noting that API standard barite contains very
few particles smaller than that. The problem
is, of course, that if one tries to exclude fluids
with a high concentration of ultra-fines, then
one of the most important economic
advantages of OBM gets lost, i.e. the reservoir
drilling fluid can no longer be a used fluid
from previous holes.
One typical approach demonstrates the
difficulty of conducting a controlled experiment in order to determine the most appropriate fluid for a particular situation. In this
approach the first five of the six aforementioned variables are carefully defined and
maintained in the study: 1) Actual cores from
the formation may be used, 2) Suitable core
lengths and diameters can be chosen and
maintained, 3) Core preparations may include
a toluene wash and a KCl displacement, 4)
The temperatures and pressures of the
experiment are chosen to be consistent with
drilling conditions, 5) No completion simulation is planned. The sixth variable, i.e. mud
type, can be chosen both with respect to the
generic mud type used and the condition of
the mud, i.e., with respect to the size
distribution of the solids. Still, however, the
results of such work can be difficult to

THE CHOICE OF DRILLING FLUIDS TO MINIMIZE RESERVOIR DAMAGE

interpret. For example, the data may favor the


invert oil based mud (IOBM) as it may seem
to cause the least reduction in return permeability. But because the particle size distribution
of the muds in a field situation is so difficult
to control, it is often suggested that this parameter, rather than mud type, may cause the
observed differences in permeability reduction. Often the relative effect of overbalance
pressure is difficult to separate from the effects
of particle size distribution. So, the balance of
the data, added to practical considerations,
often results in a decision to use OBM, while
attempts are made to monitor and control
particle size distribution as well as to minimize
overbalance. Furthermore, we are only in the
early stages of investigating the impact of
completion fluids on permeability reduction.
So, it is far from clear what should be the
answer to the crucial question: what is the
relationship between these data, especially
the results of return permeability tests, and
actual production results?
The control of drilling solids quantity and
quality in OBM is difficult. With WBM the
rheological and fluid loss properties give clear
indications of the buildup of fine solids. This
is not so with OBMs, which are far more
tolerant to solids from a purely drilling point
of view. When a fluid performs well from a
drilling point of view, it is difficult to justify
replacing it based on retort measurements of
solids. Therefore, strict control of particle sizes
will require the availability of particle size
analysis on a frequent and rapid turnaround
basis, plus a policy of replacing fluids that fail
to meet a particle size specification. Improved
solids control equipment and practices may
well help to keep mud particles in an acceptable size range. As mentioned previously, to
dispose of contaminated OBM on this basis
is difficult in the absence of sufficient actual
well production data that demonstrate in the
field what is observed in the lab.
The only alternative to this practice may be to

drill under-balanced and therefore remove the


source of energy that forces drilling fluids into
the formation. However, other non-mud-related factors cannot be ruled out either, such as
wax formation as a component of the crude.
An integrated approach
First, it must be realized that contact with
any drilling fluid will cause some damage to
any formation. The question, then, is not how
to avoid formation damage, but how to avoid
formation damage that will affect hydrocarbon production. The ultimate goal is to
develop a tool for drilling fluid design that
can account for this as well as pure drilling
problems; in other words, one that combines
the skills and concerns of reservoir, petroleum, drilling and fluids engineers.
Millions may have already been spent
trying to avoid formation damage that really
doesnt matter. As stated before, when a
vertical well is fractured, the resulting induced
fractures may extend thousands of feet into
the reservoir, rendering irrelevant solids and
filtrate invasion within feet of the well bore.
From a formation damage point of view, the
fluid loss, fine solids content and the nature of
the weighting agent affected nothing. The
difficulty with horizontal wells that are to be
completed without cementing and perforating, is that more of the producing formation
is exposed to the over-balanced drilling fluid
and it is generally exposed for a longer time.
So, the drilling fluid must be chosen so as to
minimize the formation damage at initial contact, and reduce the time of contact through
better drilling rates. This choice must include
decisions on whether various washes and
breakers, designed to remove filter cake, are
likely to help or ultimately cause more
damage. An alternative is simply to skip
the washes and back-produce the drilling
fluid through the completion equipment.
The use of a fluid such as MMH, which

3 12

WELL PLANNING

Figure 3.17

Polymer
WBM

Mud types

Sized Salt
Brine (NaCl)

Formate
Brine

MMH

OBM

Advantages

Flexible
Cheap

Easy cake removal


NaCI compatible
with formation
fluids

Easy cleanup
Good density range
compatible with
most formation
fluids

Excellent hole
cleaning
Avoids erosion of
soft formations
Minimal loss to
fractures

Low fluid loss


Filtrate compatible
with most formation
fluids
Stable under
reservoir conditions

Disadvantages

Damage,
esp. from
PHPA?
Difficult
cleanup
because
of polymers

Density restrictions
(10.4 - 12.1; 13.3
with NaBr)
Possible problem
with polymer
cleanup
Difficult to yield
polymers in brine

Expensive
Difficult to yield
polymers in brine

Sensitive to
contamination by
many mud
polymers
Questions
regarding
bentonite in
reservoir fluid
Limited fluid loss
control compared
to polymer muds

Environmental
restrictions
Possible damage due
to surfactants,
especially in tight
formations
and gas reservoirs
Non-conductive
nature limits data
acquisition

The advantages and


disadvantages of various
reservoir drilling fluids.

produces a wall cake that is more easily


removed than that from salt-saturated or
OBM, may be the best solution for some
wells. But the question becomes still more
complicated when the produced fluid must
also pass through production screens. Large
particles that had no effect on the formation
may bridge off screens.
Conclusion
The central question, then, is how can
one predict when a particular fluid is likely to
affect the permeability of a formation or the
performance of production screens enough to
affect production? The answer will depend on
many factors that are particular to a well.
No one can responsibly suggest the use of a
particular fluid without considering the whole

picture, including consideration of reservoir


type and quality, completion design, depth of
damage versus length of perforations (if any),
and mechanism and efficiency of filter cake
removal (in openhole completions), in addition to the usual parameters directly relevant
to drilling. Figure 3.17 outlines the main
advantages and disadvantages of each of
various possible reservoir drilling fluids.
To fully answer crucial questions related
to formation damage caused by drilling fluid,
it is necessary to conduct tests with cores and
conditions as close as possible to actual reservoir conditions, and to confirm the results
with actual well production data from wells
drilled with a variety of drilling fluid types
and qualities.

PREDICTING WELLBORE STABILITY

The BLower reservoir of the Misoa


Formation is located in the north part of Block
III in Lake Maracaibo. The structural map
(Fig. 3.18) shows that the top of the reservoir
is on the upthrown side of a fault system that
acts as a seal for the hydrocarbon
accumulation to the north. An oil-water
3 13

contact, originally at 9741 ft, is the limit of the


accumulation to the south. The structural dip
is quite uniform with a magnitude of 2 to 4
in the south/southeast direction.
The cumulative production of the
BLower reservoir is 26 MMSTB and the remaining reserves are estimated at 36 MMSTB.

PREDICTING WELLBORE STABILITY

Figure 3.18
1116
1110

1050

1091

1184

994

1031

95

00

'

925

988

OOWC 9741

96

00

Producer

'

1009

1008

970

0'

500 m
Scale

Structural map of BLower reservoir.


Figure 3.19

Moved hydrocarbon
Water
Gas
Oil

Mudcake

Quartz

Washout
BS

Permeability to Oil

6
16
Caliper Permeability to Water
6 (in.) 16 Permeability to Gas
Intrinsic Permeability
Depth
0.1 100
1000
(md)
(ft)

Water

Bound water

Moved hydrocarbon

Montmorillonite

Gas

Kaolinite

Oil
Sw
(%)

Fluid analysis
0 0.5
(V / V)

Illite
Volumetric analysis
0 1
0
(V / V)

9600

9650

9700

The ELAN volumetric analysis over the B-6 reservoir shows two sand bodies
separated by a shale bed.

The current production of approximately 7000


BOPD comes from seven active wells.
The reservoir is characterized by sand
bodies varying from 40 to 120 ft thick and
separated by shaly sections 10 to 30 ft thick. It
is divided into three flow units, the most
prospective being the top unit, B-6. Apart
from being structurally higher, B-6 exhibits a
better sandstone development and better
continuity. Therefore, drilling the horizontal
well VLC-1184 in the B-6 unit was considered
the best option to more efficiently drain the
remaining hydrocarbon and avoid early water
production due to water-coning.
A vertical pilot hole was first drilled to
acquire essential information and perform the
analyses required to plan the horizontal section. This pilot hole was located as far updip
(i.e., as far north) as possible, while seeking
to minimize the risk of crossing the fault. The
horizontal distance needed to land the
horizontal hole was obviously an additional
criterion in the choice of this location.
Horizontal target selection
Extensive logging and coring were
performed in the 978-in. pilot hole to gather
information needed both to characterize the
BLower Formation and to evaluate wellbore
stability. The ELAN volumetric analysis obtained from the openhole logs over the B-6
reservoir (Fig. 3.19) shows two sand bodies
separated by a 20-ft shale bed. The lower sand
is 51 ft thick with an effective porosity
between 18 and 20 p.u. and an oil saturation
ranging from 70% to 90%. The upper sand,
approximately 40 ft thick, is slightly cleaner
and consequently the effective porosity is
larger by 1 to 2 p.u. This is in agreement with
the core analysis, which shows a generally
coarser-grained sandstone. The oil saturation
is similar and the permeability, being 500 md,
is marginally higher. The clay content,
however, increases significantly in the top 10
ft and reduces the effective porosity. The core
analysis reveals a very fine-grained sandstone
in this zone.
3 14

WELL PLANNING

Figure 3.20

LLHD
GR
Depth
(ft)

Presure
(psia)

LLHS

(gAPI)

150 0.20

(ohmm)

2000

2000

3000

B6

9600

9700

9800

The MDT pressure points show a single gradient over the entire B-6 reservoir.
Figure 3.21

Dev.1
Caliper 1-3
-20

(in.)

20

Azimuth

Caliper 2-4
-20
Depth
(ft)

(in.)

Differential caliper
20

Bit size
-20

(in.)

(in.)

10

Ovalization azimuth
0

20.
20

Rupture zones

(deg)

200

Pad 1 azimuth

Ovalization
0

(deg)

Pad 1
azimuth

N
W

400

9400

9500

9600

9700

9800

The Break-out Orientation Log is computed from both FMI calipers.

3 15

Hole
azimuth

0 (deg) 10

E
S

Hole
deviation
0 (deg) 10

The upper sand was selected for placing


the horizontal section. Being structurally higher, it will delay the production of water. Crucial
information was provided by the pressure
measurements obtained with the MDT* Modular Formation Dynamics Tester. The pressure
points are aligned over the entire B-6 reservoir
(Fig. 3.20) indicating that both sands are
hydraulically connected, most likely because
the shale bed separating the two sand bodies
is not continuous. Therefore, producing from
the upper sand will drain the bottom sand as
well.
The thickness of this sand should provide a reasonably safe margin for navigating
the horizontal hole. To reduce the risk of
crossing the fault, the horizontal section
should be drilled from east to west, the well
VLC-1050 providing an excellent control of
the fault in that direction (Fig. 3.18).
Borehole stability analysis
Using the information collected in the
pilot hole, a wellbore stability analysis was
performed to ensure that the horizontal hole
could be drilled within the targeted sandstone
in the planned direction without formation
failure, such as rock collapsing or fracturing.
After determining the direction of the minimum horizontal stress, which would be the
preferred hole direction for mechanical stability, wellbore stability simulation was performed to estimate the mud weight required
to support the borehole wall.
Determining the minimum
horizontal stress direction
The key data for determining the minimum horizontal stress direction were the twoaxis caliper log from the FMI* Fullbore Formation MicroImager, and the anisotropy log
computed from DSI* Dipole Shear Sonic
Imager data.
The Break-out Orientation Log (BOL)
computes the hole enlargement in the direction of the two FMI calipers. The difference
between the enlargements, or the differential

PREDICTING WELLBORE STABILITY

Figure 3.22
Fast shear T
Slow shear T
GR
0

(gAPI)

350
150

CALI
5
Depth
(ft)

(in.)

20

Hole azimuth
0

(deg)

360 0

(deg)

90

50
Processing window

Time-based anisotropy

Fast shear waveform

0
0

Fast shear azimuth

(ms / ft)

DT-based anisotropy

(%)
4
8

50
16 <

Slow shear waveform


1100

(ms)

6600

9500

9600

9700

The DSI anisotropy log


computes the direction of
the fast shear wave.

enlargement, indicates borehole ovalization


in the direction of the caliper showing the
greater enlargement. Break-outs are a form of
enlargement caused by unequal horizontal
stresses and are identified by abrupt
ovalization. Their azimuth indicates the
direction of the minimum horizontal stress.
The results of the BOL processing from
well VLC-1184 in the interval 9820 to 9330 ft
are displayed on Fig. 3.21. Track 1 shows
actual and nominal values of the caliper pair.
Track 2 shows the differential enlargement
with the yellow shading, which indicates
values greater than the ovalization threshold.
The flags on the right side highlight the
intervals with the most probable break-outs.
In track 3, the azimuth of the FMI pad 1 is
displayed together with the azimuth of the
long axis of the borehole. This azimuth is

computed only when the ovalization is above


the threshold. In track 4, the distribution of
break-outs is presented together with the
borehole orientation. There are no break-outs
in the B-6 reservoir; hence, the horizontal
stresses directions cannot be evaluated there.
However, the shale interval immediately
above exhibits some damage with break-outs
distributed around an azimuth of 118.
In an anisotropic medium, such as a rock
exhibiting differences in horizontal stresses,
sonic shear waves split into two waves with
orthogonal polarizations. One component, the
fast shear, is oriented in the direction of maximum horizontal stress and travels faster than
the second component, the slow shear, which
runs parallel to the minimum horizontal stress.
The anisotropy log computes the fast and
slow shear slownesses and the fast shear
polarization angle. It uses input from the DSI
run in Both Cross Dipole mode. This mode
consists of recording shear information at
each orthogonal dipole receiver of the DSI
tool for each firing of the orthogonal dipole
sources (DSI box 433). This acquisition mode
provides two in-line and two cross-line components. It is unlikely that any of these four
components are aligned with the fast or slow
shear waves. However, the 4-component data
contain all azimuthal information so that fast
shear azimuth, as well as slow and fast shear
slownesses, can be extracted using a component rotation technique (Alford, 1986). The
direction of the measured fast shear wave
provides the direction of the maximum horizontal stress. The DSI is run in combination
with an inclinometry tool so the computed
direction can be related to north.
Figure 3.22 presents the acoustic
anisotropy log obtained in well VLC-1184
over the interval from 9460 to 9770 ft. Track 3
displays both fast and slow shear slownesses
and the percent of anisotropy computed by
two methods. The two quantities, Dt-based
anisotropy and Time-based anisotropy, are
derived from the difference of slowness and
difference of arrival time between the two
3 16

WELL PLANNING

Figure 3.23
North
Depth (ft)
9450.0
9498.0
9548.5
9599.0
West

9649.5

East

9700.0

South

The DSI anisotropy results


can also be presented
on a crossplot.

shear waveforms, respectively. The azimuth of


the fast shear (the azimuth of the maximum
horizontal stress) is shown in track 2. Track 4
displays actual fast and slow shear waveforms
from receiver 7 together with the processing
window. At levels of higher anisotropy, the time
difference between the waveforms can be seen.
It is also convenient for trend analysis to
display the fast shear azimuth on a crossplot
scaled east-west and south-north on the
horizontal and vertical axis, respectively
(Fig. 3.23). The color of the dots indicates
depth range. The fast shear direction in well
VLC-1184 is consistent over the entire interval,
with an average azimuth of 15.
The core recovered from the interval 9540
to 9600 ft was oriented in order to perform stress
orientation measurements that would validate
the maximum stress direction derived from
wireline logs. Several methods were used, such
as Anelastic Strain Recovery (ASR), Shear Wave
Amplitude Anisotropy (SWAA) and Acoustic
Anisotropy Analysis (AAA).
Figure 3.24 compares the wireline-derived
horizontal maximum stress directions with the
directions obtained from core measurements.
The direction obtained by both wireline
methods are in good agreement with the
direction returned by the ASR core-based

Figure 3.24

N
AAA

SWAA

ASR DSI

BOL

S
Comparison of wireline- and

core-derived horizontal
maximum stress directions.

3 17

In the AAA method anisotropic velocities are used as indications of in-situ stress orientations. Velocities are measured in
multiple directions perpendicular to the axis of the core, and in the
axial direction. It is a two-dimentional analysis, with an assumption that one of the principal stress is parallel to the core axis.
The SWAA method is based on finding the maximum absorbtion of shear waves propagated vertically through a specimen.
Two shear wave sources are placed along the core axis and one of
the sources is rotated in small angular increments. The amplitude
of the shear wave is measured for each increment and with continued rotation the minimum shear wave amplitude is determined.
The direction of the core corresponding to the minimum amplitude
infers the direction of the maximum in-situ horizontal stress.
ASR relies on monitoring time dependant strain relaxation of
freshly-cut, oriented core. The strain recovery, due to the relaxation
of the sample detached from its parent rock mass (the core), is
measured and is related to the directions of magnitudes of the insitu stresses, using elastic or viscoelastic relationships. ASR is a
completely three-dimensional measurement and does not require
assumptions about the magnitudes or directions of any of the
principal stresses.

technique i.e NNE. The AAA and SWA methods


show NNW and do not agree with the previous
methods.
The three-dimension core measurements
also showed a maximum stress oriented at 45
from the vertical. The principal stresses are
therefore not in an horizontal plane and the
two-dimension measurements such as AAA
and SWA have to be taken with caution.
Borehole stability simulation and mud
weight prediction
The borehole stability simulation is done
with the interactive workstation program
IMPACT. IMPACT analyzes and predicts well
behavior by integrating information from logs,
well tests and core measurements. The computation is organized in individual modules that
compute elastic moduli, rock strength and
stresses (Fig. 3.25). Once the mechanical
properties evaluation is done, the user chooses
the application that corresponds to the problem
to be analyzed: hydraulic fracturing design,
wellbore stability evaluation, or sanding
evaluation.
The minimum inputs required by IMPACT
are compressional and shear slowness logs, a
density log, and a volumetric analysis of the
formation such as ELAN.
The first module computes the dynamic
elastic moduli using one of the following
models: Single Component, Biot-Gasmann or
Shear modulus. Statistical analysis tools are
available to relate log-derived elastic moduli
with static moduli measured from cores.
Dynamic elastic moduli in well VLC-1184 were
calculated using the single component model.
In this model, elastic moduli are a function of
the formation bulk density and the
compressional and shear velocities.
The second module calculates static
moduli and rock strength parameters from elastic moduli and formation volumes. Various
empirical correlations are available, and core
measurements can be input to user-defined
correlations. In well VLC-1184, UCS (Unconfined Compressive Strength) values computed

Figure 3.25

PREDICTING WELLBORE STABILITY

Figure 3.25

Log measurements

Sonic waveform
analysis (STC)
Dtc

Bulk volume
analysis (ELAN)

Dts

Minerals, fluids
volumes

rb

Impact
Static-dynamic
correlations

Elastic moduli

Lab measurements
on cores

Rock strength

Overburden,
pore pressure,
stress measurements

Stress computations
sx

Pressure steps,
perforations

Hydro-fracturing
height growth

IMPACT program flowchart.

Fracturing
models

sx sy

Failure
analysis

Borehole stability:
critical mud weight
and well deviation

Failure
criteria

Sanding:
critical drawdown

from several correlations were compared to


UCS obtained from laboratory measurements
made on cores from nearby wells. The Brie
correlation (Brie, 1996) gave the best match
and was therefore selected.
In the next module, the vertical stress is
evaluated from the integration of formation
density or from a geostatic gradient, the pore
pressure is calculated from a pore pressure
gradient, calibrated with MDT pressure points,
and the horizontal stresses are derived from the
vertical stress through a choice of models.
External information can be used; for instance,
in the IMPACT run for well VLC-1184, the
minimum horizontal stress was determined

from a minifrac performed in the well VLC-988,


close to the planned horizontal well (Fig. 3.18).
For wellbore stability applications, various
failure models are available from the failure
analysis module. For well VLC-1184, the MohrCoulomb model was used to perform wellbore
stability simulations for the two extreme cases of
a horizontal hole parallel to the minimum
horizontal stress direction and parallel to the
maximum horizontal stress.
Figure 3.26 displays the results of the
simulations for the horizontal holes (tracks 3
and 4) as well as the mechanical stability
analysis done for the vertical pilot hole (track
2). The safe mud weight window represents
the range of mud weight that will safely
support the borehole wall. If the actual mud
weight is less than the minimum safe mud
weight, breakouts, or hole collapse, can occur.
If the mud weight is too high, it may fracture
the formation.
In the vertical hole, the actual mud weight
(red line) is well within the safe mud weight
window over the entire B-6 reservoir, and effectively the caliper log in track 1 does not show
any borehole failure. In the shale above,
however, the caliper indicates formation
collapse when the mud weight is outside the
safe window.
When the hole is parallel to the minimum
horizontal stress, the safe mud weight window
is quite large, especially in the sandstones. This
window is actually larger than for the vertical
hole because the stress regime derived from
mini-frac led to the model:
Ty > Tz > Tx

where Tx, Ty, Tz are the minimum horizontal


stress, maximum horizontal stress, and vertical
stress, respectively.
When the horizontal hole is drilled perpendicular to the minimum horizontal stress,
the safe mud weight window shrinks significantly and the mechanical stability could
become marginal in the middle of the target
(upper) sandstone.

3 18

WELL PLANNING

Figure 3.26

Vertical hole

Simulation for horizontal hole

Hole parallel to
minimum
horizontal stress

Hole perpendicular to
minimum
horizontal stress

Caliper
Depth
(in.)
(ft) 15 0 15

Gas
Oil

Actual mud weight


(lbm / gal)
20
Stability

Stability

Stability

Quartz
Bound
water
Montmorillonite
Kaolinite

Instability

Instability

Instability

Illite

Mudcake
Washout

Moved
hydrocarbon
Water

9500

9600

9700

Wellbore stability simulation performed with


the IMPACT program.

Drilling of horizontal section


and production results.
After setting a cement plug in the vertical
pilot hole, a 978-in. deviated hole was kicked off
at 9128 ft and drilled with a radius of curvature
of 13 for 100 ft until the inclination reached
77.5 at the depth of 9737 ft MD (measured
depth). At that depth the hole azimuth was
313, which corresponds to an optimal trajectory both to avoid the fault and to minimize the
stress on the borehole wall. A 7-in. casing was
set and the hole size was reduced to 618 in. The
final angle of 90 was reached in the target at a
depth of 9579 ft TVD.
The horizontal section was drilled for 1237
ft with an azimuth ranging from within 311.5
and 315. Operational problems occurred
while running the completion and it was eventually decided to leave the horizontal section
open. The initial production of VLC-1184
reached 2000 BNP, and after seven months it
was 1700 BOPD with a water-cut of 0.5%.
These good results show that, as predicted by
logs and cores, the borehole was stable
enough to be left as an open-hole completion.

GEOSTEERING

Introduction
GeoSteering is the real-time steering of
horizontal and high-angle wells using whiledrilling formation evaluation data. It guides
wells to optimum geological destinations,
rather than directionally steering wells to
predetermined (possibly non-optimum) geometric locations.
Directional drilling has evolved to the
point where the geometric target can be hit
with high accuracy. Unfortunately, geologic
maps based on surface seismic data and offset
well data are not this accurate. Uncertainties
in the determination of the depth and lateral
position of the target are combined with the
inability to predict subtle features such as
small changes in formation dip, pinchouts and
small displacement faults.
3 19

In addition to a sensitive and fast-acting


directional drilling capability, a GeoSteering
system has three key components. First is the
ability to obtain azimuthal formation information from as close to the bit as possible.
The further a measurement is behind the bit,
the longer it will take to observe, for example,
that the well has left the reservoir, and the
longer it will take to correct the trajectory. The
azimuthal information indicates whether the
well is going out of the top or the bottom of
the reservoir. Second, in order to actually steer
the well, the directional driller needs to have
a good idea of what changes in the downhole
measurements mean in terms of formation
geometry, and to have some idea of what to
expect if he makes a course correction.

GEOSTEERING

Figure 3.27

Sensor looking up.


Low GR sees mostly sand

Sensor looking down.


High GR sees mostly shale

Roof and floor detection a)


by rotating the azimuthal
sensors to look up or
down. b) by comparing bit
resistivity, which reads the
average around the tool
near the bit, and button
resistivity which is further
behind but can be pointed
up or down.

This information is provided by pre-job planning and modeling of the target and surrounding formations, and needs to be displayed on
a wellsite screen along with the data recorded
while drilling. Since the exact geometry cannot be predicted with the accuracy required,
several different scenarios are prepared,
representing the response of the most likely
alternatives. Only in this way can the well be
steered geologically rather than geometrically.
Lastly, it is important to have a GeoSteering
coordinator at the wellsite to coordinate
between the geologist, the logging-whiledrilling (LWD) engineer, the directional driller
and anyone else involved in the operation.
The first capability is provided by the
GeoSteering Tool (GST), which provides an
azimuthal resistivity and gamma ray (GR)
measurement from as close as 2.5 ft from the
bit (Chapter 45). In oil-based or very fresh
mud, the only resistivity-based measurement
that works is the bit resistivity, and that only
qualitatively. This works because, unlike the
other electrodes, the bit is in contact with the
formation so that current can pass between it
and any other part of the drillstring in contact
with the formation, such as a stabilizer. In such
muds it is necessary to use an induction-based
measurement such as the ARC5* Array
Resistivity Compensated tool, even though the
distance behind the bit is much greater. The
planning and modeling capability is provided
by the PowerPlan package and INFORM
(Integrated Forward Modeling), and by the
GeoSteering screen at the wellsite.
In practice different variations are used
for different situations. One case is
Geostopping, meaning the use of LWD data to
set casing or coring points as close to the
target as possible. Clearly the closer the
measurement is to the bit, the more exactly
the casing point can be picked. In other
situations the GST measurements are
sufficient to indicate whether the borehole is
moving out of the current formation, and
whether it is doing so by the roof or the floor.
An example is shown below. In yet other

situations, pre-job modeling and geosteering


are used to keep the well within the target, as
shown in the last example.
Roof and floor detection
The principle of using the GST for roof
and floor detection is shown in Fig. 3.27. Single
point station readings are taken with the sensor
facing up or down. These readings can be
plotted on a polar plot with zero at the center
and the value increasing toward the perimeter.
In this case, when the GR sensor looks up the
readings are high, indicating shale, while when
it looks down it reads low, indicating
sandstone. The largest dot shows the most
recent reading. The azimuthal resistivity faces
in the opposite direction and is used in the
same way. Alternatively the azimuthal
resistivity can be compared with the bit
resistivity, which is an azimuthal average
nearer the bit. The difference between the two
indicates the trend.
Figure 3.28 shows the logs recorded
while drilling a horizontal well in Eastern
Venezuela. The target was a 3 ft. thick interval
with the highest resistivity (70 ohm-m) within
a sand about 12 ft. thick. The casing shoe was
set at 3872 ft with the well nearly horizontal.
Near 4000 ft, while sliding with the button
facing highside, the resistivity at the bit was
found to be higher than the resistivity above
the well. (76 ohm-m vs 60 ohm-m see Fig.
3.28. The tool face curve in the depth track
shows that it is facing up at this point).Given
that the well had previously moved through a
long shale section, it was felt that this was a
strong enough indication that the well was
moving in the right direction, and that drilling
need not be inconvenienced by making a roof
and floor test. The decision was made to
rotate and both ROP (rate of penetration) and
resistivity increased. For the next 250 ft. the
well dropped while the resistivity remained
high. However, at 4250 ft. the bit resistivity
started to fall. Since this could have been
either the roof of the sand moving down or
the floor being reached, it was decided to
3 20

WELL PLANNING

5000

4900

4800

4700

4600

4500

4400

4300

4200

4100

4000

(ohm-m)
0 180 0 180 4
(ft h)
500

(ohm-m)

Bit Resistivity
ROP

(ft)
3465

TVD

GR

(gAPI)

150

3445

Depth
(ft)
GTF
(deg)

Button Resistivity

400

400

Figure 3.28

Geosteering logs in a horizontal well. When the tool is sliding, the tool face readings are smooth and indicate the direction of the sensors.
Resistivity is up, GR down when Tool face = 0 (GTF in center of depth track); resistivity is down, GR up when toolface = 180 (GTF is at
either edge of depth track). Data gaps are caused by high ROP.

make an azimuthal check. The button


resistivity in Fig. 3.28 is seen to fluctuate as the
checks are made between 4250 ft and 4390 ft.
The button resistivity is clearly greater than
the bit resistivity when the button is highside
e.g. at 4265 ft. This showed that the well was
approaching the floor and needed to move
up. By 4330 ft. the well was being steered
upwards. Monitoring of the azimuthal
resistivity and GR continued.
By 4600 ft. MD, while sliding with the
bent sub turned lowside to reduce angle, the
bit resistivity started dropping and was lower
than the resistivity facing down. The GR has
increased while the lowside resistivity remains
high around 70 ohm-m. This, combined with
the drop in bit resistivity suggests that the well
has reached the roof of the sand. The
reduction in angle was continued and
resistivity remained high for another 330 ft.
After 4950 ft. MD the sudden fall in
resistivity induced another comparison of
both GR and resistivity above and below the
well. Several measurements taken around
4990 ft were plotted on a polar plot (Fig.
3 21

3.29) and show clearly that above the well the


GR is low and the resistivity high, while
below the well the opposite is true. The
better sand is therefore above the well.
Drilling continued sliding highside, with
some increase in resistivity, but it soon
became clear that the well had left the good
part of the sand. The well was sidetracked
and drilled down from 5000 ft., but with the
same result. It was concluded that the sand
had pinched out.
Overall the well was steered within an 8
ft. sand for more than 1000 ft., much of the
time in the high resistivity section. This was
achieved by regular comparisons of the bit
resistivities with the azimuthal measurements
taken while sliding to control angle. Thus
even though the sand pinched out 600 ft.
short of the expected horizontal reach, the
well produced 1400 bbl/day, 700 bbl/day
more than planned.

GEOSTEERING

Figure 3.29

On bottom

BITRA (TN) (ohm-m)


12.31

CRPM_N (RPM)
239.00

BUTRA (TN) = 5.11


BUTGTF (N) = 5.11

GRN = 16.00
GRGTF (N) = 16.00

BUTRA (TN) (square)


min max
interval
0.0
30.00 3.00

GR_N (dot)
min max
0.0
100.0

interval
10.0

Wellsite screen for azimuthal measurements near 4990 ft. The green GR points are
clearly higher below the well, while the red resistivity points are higher above. Both
indicate a better sand above the well.
Figure 3.30
400 Plan View
1 = 200'

MFA 188
MFA 204

< North

400

800
MFA 186
1200
MFA128

1600

GeoSteering with pre-job planning


The wellsite geologist or GeoSteering
coordinator needs an initial well plan that is
based on all available data. He also needs a
way to recognize when the geology changes
to a point where the initial well plan must be
modified to reach and stay within the targeted
pay zone. The procedure is described below
and illustrated with an example from a well in
the Arecuna Field.
First, logs from wells that bound the
planned well are plotted in map view and
vertical cross section with offset and true
vertical depth (TVD) referenced to the
planned well surface location. The vertical
section can be viewed from various azimuths.
Normally the azimuth of the planned well is
selected, but one might also want to see other
views, such as perpendicular to the geological
strike. Figure 3.30 shows the plan view for the
new well MFA 204 with the existing vertical
well MFA 128 and the horizontal wells MFA
188 and MFA 186.
A geological cross section with well-towell correlations (using tools such as StratLog
or WellPix) is useful to decide which wells
best represent the expected geology and
which marker beds are continuous. At this
stage, it should also be observed which
measurements best characterize the marker
beds in the offset wells in order to decide
which tools to run (i.e., if a gas/oil contact is
an important marker for well placement, then
density/neutron tools should be part of the
LWD BHA in order to identify this marker).
3-D seismic cross sections taken along
the azimuth of the planned well are useful in
determining the continuity of marker beds.
An idea of expected structural dip can also be
seen from the seismic section and used in
creating initial models.

2000

2400
2400

2000

1600

1200

800

400

400

800

East >

Plan view of the proposed well (MFA 204) with nearby existing wells.

3 22

WELL PLANNING

INFORM

Figure 3.31

Offset - logs

Squared - logs

Trejectory #2
Trejectory #1

Tool response

1100

G
1200

1300

1400

1500

r
1600

1700

Cross section

Layer column with


petrophysical properties

Relative angle tables

2000 Real-time logs


200
E
20
2
0.2

Rps

Rad Hres

2000 Modeled logs


200
D
20
2
0.2

Rps

Rad

Simulated logs
for different trajectories

MB1: Delete, MB2: Pick, Mb3: Correlate


MD (ft) : 3998.03
TVD (ft) : 3537.46
Rps (ohm-m) : 2000
Rad (ohm-m) : 2000
Hres (ohm-m) : ****
Grc (API) : 102.38
Trph (pu) : ****
A
Rhob (g cm3) : ****
Shift (ft) : 2.60

Actual trajectory
Trajectory (TVD vs. MD)

Hole depth (ft): 4113.0

Planned trajectory Deepest sensor (ft): 4052.0

Model under rig


Offset profile (Rt)

3000

3200

D
C

3400

3600
4100

Map

INFORM (Integrated Forward Modeling) is an


interactive workstation-based system that has been
developed to support horizontal well operations during the
pre-job planning and formation evaluation phases. At the
heart of the system is a comprehensive array of
Measurement While Drilling/Logging While Drilling
(MWD/LWD) and wireline logging tool forward models
linked to an interactive graphical display. Primarily a tool
for the petrophysicist, the system also provides tight
integration with other oilfield disciplines.
During pre-job planning, geological information,
including cross section data and marker positions, is used
to assist in the construction of a petrophysical model of the
prospect. This model is integrated with trajectory files from
the Drilling Planning Center to predict MWD/LWD tool
response. All these data are accessible from a database and
may be viewed and manipulated through interactive
graphical displays.

3 23

3900

3700

3500

3300

3100

2900

2700

200 20 2 0.2

Well site geosteering screen

The figure above breaks the INFORM process into


basic steps.
1. Geologic cross sections, seismics, isopach maps and log
data from nearby wells are examined along an azimuth
parallel to the planned wellbore trajectory.
2. Wells logs representative of the expected geology are
corrected and squared into common layers of virgin
formation properties.
3. A cross section of the Geological model is formed within
INFORM, represented by a series of one or more layer
columns of petrophysical properties.
4. A database of log responses to this Geologic model is
created, based on the tools that will be run in the
planned well (especially those tools used for geosteering
purposes), and using detailed knowledge of their
response to different geometries. This is called a Relative
Angle Table or RANGDB.
5. Planned well trajectories can be combined with the
database to simulate possible log responses in a
particular well before it is drilled. This simulation is also
known as Forward Modeling.
6. The RANGDB can be exported to the Anadrill
GeoSteering Screen to allow real-time investigation of
new geological scenarios as real-time data are compared
to the simulated logs.

GEOSTEERING

Figure 3.31

PSR (2,3)

0.2
0
2400 MFA 128

MFA186

ILD (1)
(ohm-m)
GR
(gAPI)

2000
150

MFA 204
MFA188 MFA 128 MFA 188 MFA 186
1
2
3

2500
2600
2700
2800

True vertical depth (ft)

2900
3000
3100
3200
3300
3400
3500
3600
3700
3800

Cross section with nearby well logs (right) in the vicinity of the well; an image of the
GR dipping at 1 between MFA 128 and MFA 188; and the GR predicted along the
trajectory of MFA 204.

Pre-planning with INFORM


The correlation wells chosen for modeling
are converted to layers of petrophysical properties. In this case, GR and resistivity were considered adequate for formation top identification. Computer models are created that represent the response of these logs to many of the
situations we expect to encounter. Currently,
major 2-D geological features such as layering,
dip and faulting, can be modeled (see INFORM
box). During the pre-well planning phase,
these models are used to create measured

depth logs for various geological scenarios.


These logs, normally delivered in hard copy,
can be used by the wellsite geologist (or
GeoSteering coordinator) to aid in picking
correlations during execution of the job.
Figure 3.31 shows one scenario where the
structural dip is 1 as correlated from well MFA
128 to MFA 188. The existing well data are
shown on the right. The image in the middle
shows the GR interpolated between the two
wells (the higher the GR, the darker the color).
The planned well is shown in dark blue with
the predicted GR along its trajectory. Other
scenarios might assume a horizontal section
followed by a section of increased dip, or
horizontal formations with a fault crossing at
some point along the planned trajectory.
For many reasons the targeted pay zone
may not be exactly as expected. These may
be geological, because of changes in
thickness, dip or the presence of minor faults;
or they may be more petrophysical, with
changes in fluids or shale content. Therefore,
an additional pre-well planning phase is to
predict what the logs will look like if the
target has changed. This is particularly a
problem with unorientated resistivity and GR
tools because we may recognize that we are
leaving the reservoir, but not be sure if we are
coming up into the roof or going down in to
the floor. Modeling can show what tool
response to expect for each scenario and aid
in making the appropriate correction to bring
the Bottom Hole Assembly (BHA) back onto
the correct path. Even with the GST,
the response of other measurements can
help determine what action to take in
complex situations.

3 24

Figures 3.32 and 3.33 show two


scenarios. In Fig. 3.32, the formation dips
down half way along the horizontal section,
and the well moves first into the shalier
section at the top and then out into overlying
shale. In Fig. 3.33, the formation dips up.

Figure 3.32

ATTN
(ohm-m)

Phase
(ohm-m)

GR
(gAPI)

Cross section N37E - GR formation property


150
120
90
60
30
0
1000

33.48
41.76
50.52
57.84
73.32
83.76
90.00
95.52
98.28
101.16
103.80
105.96
109.44
116.40
123.23

100
10

1
1000

Trajectory
Modeled
GR
Modeled
Phase
Modeled
Attn

100
10
1

TVD (ft)

3350

T-Sand
U1-Sand

3750
3000

800

Drift along the section (ft)

100

200

GR

Model and predicted logs if the trajectory leaves the target


by the roof of the U1 sandstone.

Figure 3.33

ATTN
(ohm-m)

Phase
(ohm-m)

GR
(gAPZ)

Cross section N37E - GR formation property


150
120
90
60
30
0
1000

33.48
41.76
50.52
57.84
73.32
83.76
90.00
95.52
98.28
101.16
103.80
105.96
109.44
116.40
123.23

100
10

1
1000

Trajectory
Modeled
GR
Modeled
Phase
Modeled
Attn

100
10
1

TVD (ft)

3450

3750
3000

800

Drift along the section (ft)

Model and predicted logs if the trajectory leaves the target


by the floor of the U1 sandstone.

3 25

0 50

150

GR

250

The GeoSteering operation


The INFORM models are imported to the
wellsite GeoSteering screen. During the
buildup phase of the well, the GeoSteering
coordinator, working with the geologist,
matches events seen on the real-time logs
with the synthetic curves calculated by the
model. The model is then updated, adjusting
formation top TVDs and inclination.
Figure 3.34 shows a printout of the realtime GeoSteering screen during the buildup
section. The display consists of five subpanels
labeled A, B, C, D, and E. The user makes
correlations between modeled logs in Panel
D and measured logs in Panel E. Digital
values of the most recent correlation (in this
case correlation 9) are displayed in Panel A.
In Panel B there is a log called the offset
profile. The screen can be used to geosteer
with either GR, resistivity, density or porosity
data. Note that these profile data are
displayed versus formation TVD projected to
a vertical line that lies directly under the rig
(this is an important factor to consider when
formation dip is significant). In Panel C, both
the planned (blue dash line) and the actual
measured (solid black line) wellbore
trajectories are shown. The measured
trajectory is updated in real time from the
measurement while drilling (MWD) survey
measurements and is also extrapolated ahead
by 150 ft using the last two surveys.
The real-time logs are shown in Panel E
at the top of the display screen. These logs
can come from a variety of downhole tools,
depending on the mode selected by the
user. Panel D represents the modeled or
expected tool response when drilling
through the lithology sequence, as defined
by the profile shown in Panel B, along the

Figure 3.35

GEOSTEERING

Figure 3.34

2000 Real-time logs


200
E
20
2
9
0.2
2000 Modeled logs
200
D
20
2
9
0.2

Rps

Rad Hres

MB1: Delete, MB2: Pick, Mb3: Correlate

76

Rps

Rad

76

Actual trajectory
Trajectory (TVD vs. MD)

MD (ft) : 3998.03
TVD (ft) : 3537.46
Rps (ohm-m) : 2000
Rad (ohm-m) : 2000
Hres (ohm-m) : ****
Grc (API) : 102.38
Trph (pu) : ****
A
Rhob (g cm3) : ****
Shift (ft) : 2.60

Hole depth (ft): 4113.0

Planned trajectory Deepest sensor (ft): 4052.0

Model under rig


Offset profile (Rt)

3000

D
C

3200

T-sand

3400

3600
4100

3900

3700

3500

The GeoSteering screen at

3300

3100

2900

2700

200 20 2 0.2

drilled trajectory (black curve in Panel C).


Note that both the modeled logs and realtime logs are shown versus measured depth.
Also, note that the modeled logs are
extrapolated 150 ft to provide the user with
look-ahead-of-the-bit capabilities.
To recompute the modeled logs, the
GeoSteering screen user clicks on events (bed
boundaries, horns, etc.) present on the real-

4113 ft during the buildup.

Figure 3.35
2000
200
20
2
0.2
2000
200
20
2
0.2

Rps

Real-time logs

Rad Hres

A
2
Rps

Modeled logs

0
2

Rad

MD (ft) : 5078.55
TVD (ft) : 3567.18
Rps (ohm-m) : ****
Rad (ohm-m) : ****
Hres (ohm-m) : ****
Grc (API) : ****
Trph (pu) : ****
Rhob (g cm3) : ****
Shift (ft) : 8.97

Trajectory (TVD vs. MD)


3400

3500

T-sand
U1 Objective

3600

3700
7500

7000

6500

6000

5500

5000

4500

4000

time logs (Panel E) and correlates them by


clicking on the same features present in the
modeled logs (Panel D). The vertical dashed
green lines show the picks made on these
logs, and the horizontal dashed green lines
show the boundary or event in question in
the profile data. If a good match is found, the
GeoSteering coordinator or field geologist
can identify exactly where the bit lies within
the lithology sequence shown in the profile
of Panel B.
Once an event has been selected on both
the model and real-time logs, the model logs
are recomputed so that the two events
correlate and line up at the same measured
depth. This is achieved automatically by either
applying a TVD shift to the offset profile or by
introducing a dip in the model.
Correlations can be removed by simply
clicking on them and the model logs revert
back to the state they were in before that
particular correlation was made. As a result, the
GeoSteering screen user can investigate many
scenarios in real time while drilling the well.
Each scenario calculates and displays a new set
of modeled logs. For example, in this well,
correlation 9 in Fig. 3.34 matches a polarization
horn on the resistivity, predicted by the model
and later seen in the real-time logs. A
polarization horn is a non-linear effect with
induction-type measurements created as the
tool crosses formations of large resistivity
contrast at a high angle. It is particularly strong
with 2-MHz tools. (Chapter 839)
The model shown in Fig. 3.34 was used
until the moment when the bit crossed an
upper shale into the top of the target sandstone U1 (see U1 top in Fig. 3.35). Casing was
run and the planned trajectory modified to
the actual TVD at which the sandstone was
encountered. Correlations from the GeoSteering screen also give a prediction of what
inclination to expect the target sandstone to
maintain as the well continues.

200 20 2 0.2

The GeoSteering screen at 6195 ft in the horizontal section.

3 26

WELL PLANNING

Figure 3.36
2000
Phase

(ohm-m)
0.2

A
Depth
(ft)

4200

4250

4300

4350

4400

4450

4500

4550

4600

4650

150
GR

(gAPI)
0

Real-time log display of GR and phase resistivity in the horizontal section.

Figure 3.37
MFA 204

MFA 128
3050

Vertical Section View


Section at 216.85

MFA186

3100
3150

True vertical depth (ft)

3200
3250
3300
MFA188

3350
3400
3450
3500

Original Plan

3550
3600
U1 Objective

3650
3700
3200

2600

2000

1400

800

200

Section departure (ft)

Comparison of the original (geometric) planned and actual (geosteered) trajectory.

3 27

Figure 3.35 shows continuation of the well


in a 618-in. hole running the SLIM1* and ARC5.
The GeoSteering screen has now switched to a
second model that better represents the U1
sandstone and can be used to try and maintain
the BHA within this sandstone.
Initially the well dropped low (difficult to
build an angle in the soft sandstone) and at
4400 ft (A) the resistivity began to fall. This is
shown more clearly on the real-time log display
in Fig. 3.36. Soon after, at B, the resistivity
started to increase again and so did the GR. The
question arose, are we at the bottom of the U1
sandstone or is it just another shale stringer as
seen at the top of U1? The pre-job modeling
gave the answer. The sharp rise at B,
accompanied by the rise in GR, was much
better explained by Fig. 3.33 than Fig. 3.32. The
conclusion was that the well was crossing the
floor of the U1 sandstone and that the sharp rise
in resistivity was due to a polarization horn. The
decision was made to come up sharply, so that
by C the GR had dropped and resistivity
returned to U1 consistent levels.

GEOSTEERING

At 5100 ft MD (-2200 ft drift), the


resistivity began to fall again, and at this point
the well was climbing at 92 (Fig 3.35). The
formation model (Fig. 3.35, correlation 2)
showed the well approaching the thin shaly
streaks at the top of the U1 sandstone. No
drilling parameters indicated any change.
However, the drop in resistivity and
confidence in the model were sufficient to
convince the wellsite coordinators to drop the
well slightly to maintain it within the target
zone and avoid the thin shaly layers.
Figure 3.37 compares the trajectory of the
original plan and the actual well. There is no
doubt that a purely geometrically drilled well
would have missed the objective entirely. In
practice, some adjustments would probably
have been made at the casing point. However,
without the pre-job modeling and planning,
and with the uncertainties in formation tops
and dips, it would have been difficult to know

whether the well was leaving the target by the


top or the bottom at 4400 ft, and it would
have been difficult to keep systematically
within the sandstone.
Conclusion
GeoSteering helps ensure that horizontal
and highly deviated wells reach and stay
within their targets. Measurements close to the
bit help decide the precise point at which to
stop drilling and set casing or start coring .
Azimuthal measurements near the bit determine if the well is leaving the target by the
roof or the floor. Pre-job modeling and the
wellsite GeoSteering screen allow comparisons of actual vs. predicted measurements as
a basis for steering the well. Not all effects are
modeled, but GeoSteering offers more choices
and instils more confidence in decisionmaking, thus minimizing the risk of drilling
horizontal wells.

AUTHORS AND CONTRIBUTORS

This chapter was written by D.Eubank, M.Silverstone, P.Laurent, and S.Low


with contributions from L.Gaetano (Corpoven), J.Fernandez (Lagoven), G.Gonzlez
(Maraven) and G.Farruggio and with the permission of Maraven, Lagoven and Corpoven to
publish data from their wells.

REFERENCES
Fraser, L., Reid, P., Williamson, D., and Enriquez Jr, F., 1995, Mechanistic
investigation of the formation damaging characteristics of mixed metal
hydroxide drill-in fluids and comparison with polymer-base fluids, SPE 30501,
Society of Petroleum Engineers Annual Technical Conference and Exhibition
held in Dallas, Texas, U.S.A., October 22-25.
Gonzlez, G., Coll, C., Gamero, H., Meza, E., Simon, C., Cespedes, A, and de
mena, J., 1997, Pozo horizontal VLC-1184 Reto Tecnologico, SVIP 084,
Sociedad Venezolana de Ingenieros de Petroleo, XI Jornadas Technicas de
Petroleo en Maturin, Monagas, Feb 19-22.
Esmersoy, C., Koster, K., Williams, M., Boyd, A., and Kane, M., 1994, Dipole
shear anisotropy logging, presented at the 64th Annual International meeting,
Society of Exploration Geophysicists, Los Angeles, California, October 23-28.
Brie, A., and Bratton, T., 1996, IMPACT, a geomechanical wellbore evaluation
system, ISRM news journal, vol2 no. 1.

Terratec, 1996, Static and dynamic properties and in-situ stress direction, well
VLC-1184.
Alford, R.M., 1986, Shear data in the presence of azimuthal anisotropy.
Schlumberger Oilfield Services 1995, Horizontal Well Planning, Evaluation
and Execution.
Prilliman, J. D., Allen, D. F., and Lehtonen, L. R., 1995, Horizontal well placement and petrophysical evaluation using LWD, SPE Paper 30549, Society of
Petroleum Engineers Annual Technical Conference and Exhibition held in
Dallas, Texas, U.S.A., October 22-25.
Wu, P., Tabanou, J. R., and Bonner, S. D., 1996, Petrophysical interpretation
of a multispacing 2-MHz MWD resistivity tool in vertical and horizontal wells,
SPE Paper 36547, Society of Petroleum Engineers Annual Technical
Conference and Exhibition held in Denver, Colorado, U.S.A., October 6-9.

3 28

REFERENCES

3 29

Vous aimerez peut-être aussi