A Sound Approach to Drilling

Of the many decisions drilling engineers make, selecting an optimal mud

weight is one of the most challenging and far-reaching. Today, sonic logging-
while-drilling tools are instrumental in making these decisions.

Jeff Alford Generations of drilling engineers have struggled stability, interpret lithology changes, monitor
Roger B. Goobie to visualize the dark and formidable downhole fluid-flow effects in the borehole and pick
Colin M. Sayers drilling environment. Today, engineers and accurate casing-setting depths.2
Ed Tollefsen geoscientists rely on increasingly sophisticated More importantly, these data are available in
Houston, Texas, USA sensors to gather data from deep beneath real time to help engineers and geoscientists
the Earth’s surface, understand subsurface make critical decisions that affect drilling cost
Jay Cooke
lithology, identify geologic features, locate and efficiency (see “Acting in Time to Make the
Helis Oil & Gas
Houston, Texas hydrocarbons and make a host of drilling and Most of Hydrocarbon Resources,” page 4). In this
completion decisions. article, we describe how advanced sonic tools
Andy Hawthorn Even though our sense of sight is highly and interpretation techniques are helping to
John C. Rasmus developed, it has its limitations. So, early in the better define the safe mud-weight window, drill
Sugar Land, Texas 20th century, scientists began development of deeper and optimize casing-setting depths. Field
technologies that would allow visualization of examples from the Gulf of Mexico and offshore
Ron Thomas environments that could not otherwise be seen. Australia show how operators are using real-time
PPI Technology Services In 1906, Lewis Nixon invented the first sound acoustic data and wellsite-to-shore telemetry
Houston, Texas navigation and ranging, or sonar, device, as a way systems to limit risk and uncertainty while
of detecting icebergs.1 reducing well cost.
For help in preparation of this article, thanks to Ron
Blaisdell, New Orleans; Lennert den Boer, Calgary; Early sonar devices were passive; they could
Joaquin Armando Pinto Delgadillo and Egbonna Obi, only listen. However, between 1914 and 1918, A Pressing Need for Pressure Prediction
Youngsville, Louisiana; Nick Ellson and Dale Meek,
Sugar Land, Texas; and Ivor Gray, CJ Hattier and World War I accelerated interest in and develop- Key to the well construction process is an
Sheila Noeth, Houston. ment of active sonar tools for submarine detection. understanding of the subsurface pressure
APWD (Annular Pressure While Drilling), CDR (Compen-
sated Dual Resistivity), FPWD (Formation Pressure While The first active sonar technology transmitted environment.3 Changes in the normal pressure
Drilling), PERT (Pressure Evaluation in Real Time), a sound, or ping, through water. Multiple gradient affect drilling safety, casing design and
sonicVISION, StethoScope and TeleScope are marks of
Schlumberger. receivers called transponders detected the setting depths, and in particular, the mud-
1. For more on the development of sonar devices: returning sound echo, providing data on the weight window.
http://www.absoluteastronomy.com/reference/sonar relative positions of static and moving objects. Engineers restrict the mud-weight range to
(accessed February 6, 2006).
Today, advanced acoustic technologies have sustain borehole stability, control downhole
2. For more on sonic logging: Brie A, Endo T, Hoyle D,
Codazzi D, Esmersoy C, Hsu K, Denoo S, Mueller MC, many uses in areas such as medicine, military pressures and optimize casing-setting depth.
Plona T, Shenoy R and Sinha B: “New Directions in Sonic applications and oil and gas exploration and Most often, the mud weight is maintained above
Logging,” Oilfield Review 10, no. 1 (Spring 1998): 40–55.
3. Barriol Y, Glasser KS, Pop J, Bartman B, Corbiell R, production (E&P). the formation pressure—at a level required to
Eriksen KO, Laastad H, Laidlaw J, Manin Y, Morrison K, Acoustic-based logging-while-drilling (LWD) control formation stress and prevent kicks or
Sayers CM, Terrazas Romero M and Volokitin Y: “The
Pressures of Drilling and Production,” Oilfield Review 17, tools provide data that help reduce uncertainty influxes that can lead to costly well-control
no. 3 (Autumn 2005): 22–41. and allow engineers to make effective and timely problems—and below the fracture gradient to
4. For more on underbalanced drilling: Bigio D, Rike A, drilling decisions. Data from sonic LWD tools not prevent the formation from breaking down and
Christensen A, Collins J, Hardman D, Doremus D,
Tracy P, Glass G, Joergensen NB and Stephens D: only help establish pore-pressure gradients, but losing returns. Wells are also sometimes drilled
“Coiled Tubing Takes Center Stage,” Oilfield Review 6, also help define porosity and permeability, detect with the static mud weight below formation
no. 4 (October 1994): 9–23.
and type hydrocarbons, evaluate borehole pressure, or underbalanced.4

68 Oilfield Review
W
 Spontaneous Formation resistivityy depends on porosity, the compensated for, reliance on resistivityy data
Potential (SP) Depth Resistivity type off the fluid within the pore space and its alone forr pore-pressure prediction significantly
ionic strength. Under normal compaction increases drillingg risk.
N al trend
Normal t d conditions, an increase in shale resistivityy with Geoscientists can often identifyy abnormally
line
depth corresponds to a reduction in porosity pressured formations using seismic velocities.
R i tii it
Resistivity
deviation
i
(left). An anomalous change in formation For a given lithology, acoustic velocityy usually
pressure is usuallyy associated with a shift in the depends on porosity: the greater the porosity, the
normal compaction trend, indicated on an lower the acoustic velocity. In normallyy com-
electric log byy a reduction in resistivityy associ- pacted sediments, compaction increases with
ated with an increase in porosity. depth. Porosity, in turn, decreases with depth,
For the purpose off maintaining safe mud and so the velocityy off sonic and seismic waves
weight while drilling, information about traveling through the formation generally
abnormal pressure needs to be available while increases with depth (below). Deviations from
drilling. Although formation resistivityy is one of this trend can often be attributed to layers of
the most common LWD measurements, several sediments that have not compacted, signaling
factors can have a significant effect on the data, abnormallyy high pressure, called overpressure.
potentially masking changes in the normal However, uncertainties in seismic velocities
compaction trend and hindering the detection of commonlyy result in depth errors, making it
> Electric log analysis to predict pore pressure.
abnormal pressure.5 difficult to define exact distances to drilling
In normally compacted sediment, electrical Changingg temperature in the borehole with hazards and geological targets.
resistivity will increase with depth along a depth alters the resistivityy off formation water, Velocityy models created from seismic dataa can
normal trend line (red). A deviation in resistivity while the presence off hydrocarbons considerably be improved byy adding high-resolution informa-
from the normal trend may indicate abnormal increases resistivity. Large deposits off organic tion from sonic measurements obtained while
formation pressure.
matterr mayy also increase resistivity, obscuring drilling (next page). Today, geoscientists and
undercompaction indicators. Changes in the engineers combine while-drilling and wireline
condition off the borehole, such as an increase in
5. Aldred W, Bergt D, Rasmus J and Voisin B: “Real-Time
The optimal mud-weight range is frequently borehole diameter due to washout or caving, Overpressure Detection,” Oilfield Review 1, no. 3
narroww and difficult to define; this is especially furtherr increase resistivityy measurement error. (October 1989): 17–27.
true in tectonicallyy stressed regions and in Although manyy off these effects can be
deepwater environments. Within this narrow
Checkshot
mud-weight window, engineers balance several Tomography
factors, including the minimum flow w rate Interval velocity
0 calculated from
required for hole cleaning, downhole motor and stacking velocity
0.5
telemetryy operations and equivalent circulating
1.0
and static densities. Drilling fluids such as oil-
Distance, km 1.5
base and synthetic-base muds frequentlyy exhibit
Depth, km

0 9
thermomechanical and compressibilityy properties 0 2.0

that varyy with depth, making it difficult to 1

2.5

optimize drilling efficiencyy while maintaining 3.0

Depth, km

mud weight. 2 3.5

Operating within the mud-weight window 4.0
3 1.5 2.0 2.5 3.0
allows engineers to improve drilling efficiency Velocity, km/s
and set casing at the best possible depth. If 4
casing is set too shallow, well-construction cost
typicallyy increases, well depth is limited,
production rate mayy be compromised and, in > Predicting pore pressure in the Gulf of Mexico 16
some cases, the target mayy not be reachable. with seismic data. In this example, the initial 0.5 15
Pore pressure, lbm/galUS

Maintaining the mud weight within a specific velocity model based on conventional stacking- 1.0 14
windoww relies on accurate determination and velocity analysis (above leftt) predicts the 1.5 13
Depth, km

prediction off anomalous changes in formation presence of overpressure (black circle). 2.0 12
Although pore-pressure predictions based on 2.5
11
pressure. The analysis off shale resistivityy using this information are not sufficiently accurate for 3.0
10
wireline log data is one off the oldest methods for drilling, a higher degree of seismic-velocity 3.5
9
detecting abnormal pressure. resolution can be obtained by using tomographic 16
analysis and checkshott data to refine the velocity 14 8
12 16
model (above rightt). Further data processing 10 14
12
allows construction of a three-dimensional (3D) y, k 8
m 10 km
6 6 8 x,
pore-pressure cube (bottom rightt).

70 Oilfield Review
 Gamma Ray Attenuation Resistivity
0 gAPI 150 0.2 ohm.m 2.0
Rate of Penetration Depth, Phase Shift Resistivity Sonic LWD ∆t
Normal compaction
p 50 ft/h 0 ft 0.2 ohm.m 2.0 150 µs/ft 50
t d
trend
8,800

9,000

9,200
Divergence
9,400
Onset of
overpressure
9,600

9,800
O erpress
Overpressured
p redd
zone
10,000

10,200

150 50
∆t 10,400

MD, ft Mud Weight

Overburden Gradient
Fracture
0 psi 5,000
Maximum Stress
Losses Fracture
0 psi 6,000 Unconfined
Minimum Stress Compressive Strength
Kick Shear Failure: Breakout Overgauge
0 psi 5,000 0 psi 5,000
Poision’s Ratio Tensile Strength Caliper
Breakout Shear Failure: Echelon
0 () 0.5 0 psi 6,000 5 in. 25
Gamma Ray Depth, Vp/Vs Ratio Friction Angle Diameter
Shear Failure: Knockout
30 gAPI 150 ft 0 () 6 0 deg 50 5 lbm/galUS 25 5 in. 25
7,680

7,700
Safe mud-
7,720 weight window

7,740

7,760

7,780

7,800

7,820

7,840

7,860

> Defining mud-weight windows. Sonic velocity can be used to predict changes in the normal compaction trend that is often an
indicator of abnormal pressure (top leftt). Unlike resistivity measurements, sonic velocity is unaffected by changes in borehole
temperature and salinity. Real-time compressional-slowness measurements from sonic LWD tools are used to predict pore
pressure and help define kick and borehole breakout limits (top rightt). Adding sonic shear measurements (bottom), available in
fast formations, helps determine kick and mud-loss potential, fracture limits, and the safe mud-weight window shown in white
(Track 4). Various types of shear failure can also be defined (Track 5).

Winter 2005/2006 71
 Sonic Acoustic Synthetic
Density slowness impedance Reflectivity Wavelet seismogram

> Placing a bit on the seismic map using synthetic seismograms. Sonic LWD slowness data are inverted with the density
measurement to produce an acoustic-impedance (AI) measurement (process from left to rightt). The AI is converted to
reflectivity and convolved with a 35-Hz wavelet at each reflector to obtain the synthetic seismogram (rightt). Geophysical
analysis of the seismic data determines the wavelet frequency. With increasing depth, higher frequency seismic signals
are attenuated, so a lower frequency, generally 20 Hz instead of 35 Hz, is used to correlate the sonic LWD data to surface
seismic measurements. This helps engineers and geoscientists place the bit on the seismic map more accurately.

sonic data with checkshots to generate synthetic Sonic Measurements While Drilling As the bit approached the predicted
seismograms that are then correlated with Soon after the introduction of sonic LWD overpressured zone, acoustic velocity acquired
predrilling seismic measurements, providing the measurements in the late 1990s, an operator while drilling was used to continuously update
drilling team with a way to locate the drill bit experimented with using sonic LWD measure- the velocity models derived from existing surface
within the geophysical environment (above).6 ments to improve drilling efficiency in several seismic and VSP surveys. Simultaneously,
These real-time processes help engineers prepare major operating areas. On an exploration well in engineers at the wellsite used real-time CDR
for pressure changes before drilling into them. the Gulf of Mexico, USA, in an area known for Compensated Dual Resistivity data, sonic LWD,
Generating a synthetic seismogram from LWD abnormally pressured formations, sonic and weight-on-bit (WOB), rotary torque and rate-of-
data involves combining transit-time (∆t) data density LWD data were transmitted from the rig penetration (ROP) measurements, in conjunc-
with density measurements, to produce an to the operator’s office. There, geoscientists tion with the PERT Pressure Evaluation in Real
acoustic impedance (AI) model. This model is generated a synthetic seismogram, which was Time program, to monitor changes in pore
converted into a seismic reflectivity sequence, correlated to the surface seismic section imaging pressure a few meters behind the bit. This
and then convolved with a selected wavelet to the target zone and an overlying overpressured information was used to calibrate the pore-
produce a synthetic seismogram.7 A synthetic zone.8 The synthetic seismogram indicated that pressure predictions from the seismic and
seismogram is much more useful when it is the top of the overpressured zone was 60 ft VSP data.
depth-calibrated with either a wireline or while- [18 m] deeper than what the seismic section Using multiple techniques for pore-pressure
drilling checkshot or vertical seismic profile predicted. This information allowed engineers to prediction, the operator accurately predicted
(VSP). Although the synthetic seismogram can place the casing shoe significantly closer to the changes in formation pressure, identified mini-
be generated at the wellsite, more often, the real- overpressured zone, optimizing casing-setting mum mud-weight requirements and optimized
time data are transmitted to an engineering depth and improving the safety and drilling casing-setting depth to construct a successful
center for processing. efficiency of subsequent borehole sections. well in this hostile environment.
Correlating a synthetic seismogram with In another early example, BHP (now BHP
surface seismic traces helps geoscientists and Billiton) and Schlumberger demonstrated the Narrowing the Window of Uncertainty
engineers place the borehole trajectory on a use of sonic LWD measurements not only to Drilling in technically demanding areas is usually
seismic section. Calculation of the spatial position calibrate seismic reflections, but also to update associated with high cost and elevated levels of
of the borehole relative to seismic markers, or pore-pressure calculations ahead of the bit.9 risk and uncertainty.10 Sonic LWD data available
reflectors, allows the drilling team to look ahead Several exploration wells offshore Western in real time play a key role in reducing cost, risk
to abnormal changes in formation pressure. Australia had been abandoned prematurely due and uncertainty by updating models created
to wellbore-stability problems associated with before drilling. However, creating those models
overpressured formations.

72 Oilfield Review
in the first place can be a bottleneck. In 2000,
geoscientists began looking at opportunities to
increase the speed and accuracy of while-drilling
pore-pressure modeling and prediction.11
In the deepwater Gulf of Mexico (GOM),
overpressure causes major drilling hazards.
Overpressure is caused by Mississippi River
sedimentation that is rapidly buried compared
with the time it takes for sediments to expel
pore water. This prevents sediments from
compacting as they are buried and causes the
pore fluid to become overpressured. In under-
compacted sediments, sediment grain contacts
are weak, causing low rock strength and low
acoustic velocities.
Accurate determination of pore pressure is a
key requirement to making optimized drilling
decisions in these overpressured environments.
Before drilling, pore pressure can be predicted
using seismic velocities—assuming there is a
seismic survey available and processed—
together with a velocity-to-pore-pressure trans-
form calibrated to offset-well data. However, this
procedure takes considerable time. Synthetic
seismograms can be generated quickly, compared
with the time needed for analyzing seismic
velocities and creating a pore-pressure cube.
As engineers focus on ways to reduce risk and
uncertainty, the time required to process and
correlate seismic and sonic LWD data becomes
critical. To speed up this process for prospects in 10 11 12 13 14 15 16 17
the northern GOM, Schlumberger geoscientists Pore pressure, lbm/galUS
developed a pore-pressure cube for the entire
> Building a three-dimensional (3D) mechanical earth model in the Gulf of Mexico.
area using data released by the Minerals
Seismic, checkshot and sonic data released by the Minerals Management Service
Management Service (MMS) (right).12 (green dots) were gathered from wells in the Gulf of Mexico (top) where pore pressure
Checkshot data from the MMS in the Gulf of exceeded 10 lbm/galUS [1,198 kg/m3] and the predicted velocity error was less than
Mexico were inverted to obtain compressional ± 1,200 ft/s [± 366 m/s]. The data were then trend-kriged to predict pore pressure, and
then plotted in a 3D model (bottom).
velocity versus depth below the mudline. These
velocity functions were then combined with
upscaled sonic logs from deepwater wells and
kriged to populate a three-dimensional (3D) 6. A checkshot is a type of borehole seismic survey 10. Malinverno A, Sayers CM, Woodward MJ and
mechanical earth model (MEM) displaying both designed to measure the acoustic traveltime from the Bartman RC: “Integrating Diverse Measurements to
surface to a known depth. Formation velocity is Predict Pore Pressure with Uncertainties While Drilling,”
velocity and levels of expected uncertainty.13 measured directly by lowering a geophone to each depth paper SPE 90001, presented at the SPE Annual
By applying a threshold to the predicted of interest, emitting energy from a source on the surface Technical Conference and Exhibition, Houston,
and recording the resulting signal. A checkshot differs September 26–29, 2004.
kriging error, maps of undercompaction and from a vertical seismic profile in the number and density 11. Sayers CM, Johnson GM and Denyer G: “Predrill Pore
overpressure can be restricted to specific areas of receiver depths recorded; geophone positions may be Pressure Prediction Using Seismic Data,” paper
widely and irregularly located in the wellbore, whereas a IADC/SPE 59122, presented at the IADC/SPE Drilling
of interest. For commercial projects, a confi- vertical seismic profile usually has numerous geophones Conference, New Orleans, February 23–25, 2000.
dential client subcube may be extracted from the positioned at closely and regularly spaced intervals in
12. Sayers CM, den Boer LD, Nagy ZR, Hooyman PJ and
the wellbore.
full GOM pore-pressure cube. Any additional Ward V: “Regional Trends in Undercompaction and
7. A wavelet is a pulse representing a packet of energy Overpressure in the Gulf of Mexico,” Expanded
information provided by the operator and data from the seismic source. Abstractss, 75th SEG Annual Meeting, Houston
acquired during the drilling process with sonic 8. Hashem M, Ince D, Hodenfield K and Hsu K: “Seismic (November 6–11, 2005): 1219–1222.
Tie Using Sonic-While-Drilling Measurements,” 13. Kriging is a statistical technique used with two-point
LWD and real-time pore-pressure tools is used to Transactions of the SPWLA 40th Annual Logging statistical functions that describe the increasing
update the client model, increasing resolution Symposium m, Oslo, Norway, May 30–June 3, 1999, paper I. difference or decreasing correlation between sample
9. Tcherkashnev S, Rasmus J and Sanders M: “Joint values as separation between them increases, then to
Application of Surface Seismic, VSP and LWD Data for determine the value of a point in a heterogeneous grid
Overpressure Analysis to Optimize Casing Depth,” from known values nearby.
presented at the EAGE Workshop: Petrophysics Meets
Geophysics, Paris, November 6–8, 2000.

Winter 2005/2006 73
 1 2

500 500

1,000
1,000 1,000
1,000

1 500
1,500 1 500
1,500

h m
h m

Depthh,
Depthh,

De
De

2 000
2,000 2 000
2,000

Sonic
2 500
2,500 2 500
2,500

3 000
3,000 3 000
3,000

1 500
1,500 2,000
2 000 22,500
500 3 000
3,000 10 15 20 1 500
1,500 2,000
2 000 2,500
2 500 3 000
3,000 10 15 20
Vp, m/s
/ P pressure gradient
Pore-pressure
Pore gradient,
di t lbm/galUS
lb / lUS Vp, m/s
/ P pressure gradient
Pore-pressure
Pore gradient,
di t lbm/galUS
lb / lUS

3 4

500 500

Mud weights Mud weights

1,000
1,000 1,000
1,000

1 500
1,500 1 500
1,500
h m

h m
Depthh,

Depthh,
De

De

2 000
2,000 2 000
2,000
Pore-
pressure
Sonic Sonic data
2 500
2,500 2 500
2,500

3 000
3,000 3 000
3,000

1 500
1,500 2,000
2 000 22,500
500 3 000
3,000 10 15 20 1 500
1,500 2,000
2 000 2,500
2 500 3 000
3,000 10 15 20
Vp, m/s
/ P pressure gradient
Pore-pressure
Pore gradient,
di t lbm/galUS
lb / lUS Vp, m/s
/ P pressure gradient
Pore-pressure
Pore gradient,
di t lbm/galUS
lb / lUS

> Reducing uncertainty with pressure data from multiple sources. The degree of uncertainty in a pore-pressure gradient is exemplified by the width and
low resolution of the compressional-velocity (Vp) and pore-pressure gradient curves (1). Velocity data from sonic checkshots are added to the model,
somewhat reducing pore-pressure uncertainty (2). Adding mud weights from drilling reports (3) and physical pore-pressure measurements (4) refines
estimates and dramatically improves pore-pressure resolution.

and reducing pore-pressure uncertainty both in Until fairly recently, many believed that it tools were designed around what is referred to as
the immediate drilling environment and ahead of would be impossible to achieve sonic measure- the hoop-mode frequency range of the collars.
the bit (above). ments while drilling. Engineers thought that the This frequency depends on the collar thickness
Along with improved modeling, technological fast acoustic-signal arrival down the tool collar and diameter, but for most tools, falls in a narrow
advances in LWD tools and telemetry systems are from the transmitter to the receivers would band between 11 and 13 kHz.
yielding more accurate real-time measurements dominate all the arrivals, making it impossible to At the hoop-mode frequency, acoustic waves
and in greater quantities. The sonicVISION, new discriminate and record formation arrivals. attempt to expand the collar rather than travel
generation sonic-while-drilling LWD tool intro- With this in mind, the designers of first- down to the receiver, thereby attenuating the
duced in April 2004 has increased confidence generation sonic LWD tools focused on mitigating collar arrivals at the receivers. By designing the
in the accuracy of real-time compressional- direct collar arrivals. To accomplish this, the transmitters to fire within the narrow hoop-mode
wave velocities. frequency band and filtering received data to the

74 Oilfield Review
same range, engineers hoped to receive clean
and discernable formation arrivals, free from
distortion caused by collar arrivals.
This technique proved somewhat satisfactory
for fast formations where the excitation
frequency falls within the appropriate range.
However, for slower formations, larger hole sizes
and for lower frequency components of the wave
train, such as shear and Stoneley waveforms,
these first-generation tools did not excite the
formation at the optimum frequency and were
discarding data outside of the narrow band
around the hoop mode (right).
1,000 Stoneley energy
Narrow-band processing also promoted
spatial aliasing, a condition in which non-

Amplitude, mV
formation arrivals, or processing artifacts, 100
Collar attenuation
appear within the slowness time coherence
(STC) search-band window. Aliasing depends on 10 Compressional and
the frequency of the transmitted pulse, the shear energies
recorded waveform frequencies and the inter-
1
receiver spacing. With an almost monofrequency 0 5 10 15 20
system, aliasing was well-developed and led to Frequency,
Previous tool frequency kHz
incorrect picking of events that were not
range
formation arrivals.
Misinterpretation of signal arrivals can also sonicVISION frequency
limit the usefulness of acoustic data. Previous range
tools analyzed all acoustic arrivals within a time
window associated with a depth. So within this
dataset, there could be compressional, shear,
mud, collar and aliased arrivals. The tool’s
downhole processors then discriminated the
compressional arrival from other signals based
on the coherency of those events. With compres-
sional arrivals being one of the smallest events
discernible in the wave train, their coherency is
> Frequency range of the new tool design. The frequency ranges of previous
typically low when compared with other arrivals
tools were narrowly aligned within the collar attenuation frequency. Newer
(below right). Early tools often confused or tools have an expanded frequency range covering a broader spectrum of soft
misidentified the data, sending incorrect values and hard formations (yellow bar). Lower frequency arrivals such as Stoneley
to the surface. and leaky-P (not shown) are now captured.
To mitigate these problems, Schlumberger
engineers designed the sonicVISION tool to
transmit and receive wide-band acoustic signals
in a frequency range from 3 to 19 kHz, a range Total Shear Mud
more likely to generate a measurable response transit time arrivals arrivals
from most formations. Acoustic shear waves are
difficult to acquire with narrow-band tools
because they contain lower frequencies than
Amplitude

compressional waves. The sonicVISION tool

frequency is optimized to excite the formation Time, s
Transmitter
across a significantly wider frequency band than firing
that of previous tools. This allows both shear and
compressional measurements to be routinely
made while drilling in faster formations. Power- Compressional Rayleigh Stoneley
arrivals arrivals arrivals
output levels were also increased 10-fold to more
> Acoustic wave train. Once an acoustic signal is transmitted, it travels through the formation, annular
effectively couple the wide-band acoustic energy
fluid, and to some degree the tool, ultimately arriving at the receiver array. Low-amplitude compressional
to the formation. signals (red) arrive first, followed in harder rock by the shear arrival. Newer tools take advantage of
slower arrivals such as Rayleigh and Stoneley.

Winter 2005/2006 75
 To speed data to the surface, Schlumberger
Min Amplitude
p Max Slowness Time Coherence Peaks
recently released the TeleScope high-speed
0 Slowness Projection 1 Maximum and Minimum ∆t Compressional telemetry-while-drilling service. This new
Label Limits
Recorded Mode
Min Amplitude
p Max measurement-while-drilling (MWD) system is
40 µs/ft 240 Maximum and Minimum ∆t Shear Label Limits capable of providing enough power to run eight or
0 Slowness Projection 1
∆t Shear from Receiver Array Real Time more LWD tools while offering up to a fourfold
∆t Shear from Receiver Array
40 µs/ft 240 increase in data rate over comparable tools. Field
40 µs/ft 240 µs/ft 240
∆t Compressional from Depth, application of these new hardware technologies,
Receiver Array ft ∆t Compressional from Receiver Array ∆t Peak 1 Compressional Computed Uphole, Real Time
in combination with improved pore-pressure
40 µs/ft 240 40 µs/ft 240 40 µs/ft 240
modeling described earlier, promises to enhance
X,300
drilling efficiency and reduce geologic and well-
construction uncertainty.
Advancements in sonic LWD tool design and
telemetry systems have overcome many of the
inadequacies previously inherent in while-
drilling sonic measurements. New data proces-
sing techniques and improvements in telemetry
systems have minimized earlier limitations,
allowing real-time access to while-drilling sonic
compressional measurements in almost any
drilling environment.

Seismic, Sonic and Pressure Measurement—

Defining the Mud-Weight Window
In many GOM fields, pore pressure changes
rapidly with depth, and tight mud-weight
windows make drilling and completion difficult,
or even impossible. One example of an extremely
difficult environment is the Vermillion offshore
area. Here, mud weights often reach 18 lbm/galUS
X,400 [2,157 kg/m3], the risk of wellbore instability
> Compressional and shear peaks available in real time. Because of improvements in downhole tool and lost circulation is high, and six or more
and telemetry systems, slowness time coherence peaks can now be sent to the surface for casing strings are typically required to reach
evaluation and labeling while drilling (Track 2). Previously, the semblance projection was available target reservoirs.
only by processing tool memory after the tool was pulled from the borehole (Track 1). The semblance
projection based on the real-time peaks (Track 3) is consistent with the memory-mode data. Station
Today, operators use data retrieved during
measurements of compressional ∆t (white circles) acquired during quiet periods, such as when drilling from sonicVISION, StethoScope formation-
pumps were off during pipe connections, also confirm the accuracy of the real-time data. The pressure-while drilling service and other LWD tools
sonicVISION system has the unique capability to modify label limits (Track 2) at surface for better to help improve well-construction efficiency and
extraction of compressional data, and for the first time, real-time shear data. These improvements in
real-time quality control make the compressional input, used for pore-pressure calculation of reduce cost by accurately defining and managing
minimum mud weight, more robust. Combining the real-time compressional and shear data also the effective stress and mud-weight window.
enables geomechanical calculations of the maximum mud-weight window. Sonic LWD, real-time formation pressure and
other while-drilling tools were successfully used
The new design also transmits real-time ination of arrivals improves pore-pressure to reduce risk and operational uncertainty while
coherent events, called peaks. The sonicVISION measurement and allows geomechanical inter- drilling a well in Vermillion Block 338 during
tool can send up to four peak arrivals uphole at any pretation based on while-drilling compressional, 2005. In this well, which was owned by Helis Oil &
given time, enabling engineers at the surface to shear and density data. Gas LLC and operated by PPI Technology
accurately differentiate arrivals rather than relying Acoustic data can now also be further refined Services, engineers planned and executed an
on downhole processing. These peaks are then by acquiring data while the pumps are off. aggressive drilling program. This program
assembled to form an STC projection log that helps Background noise of the same frequency as sonic extended both the 95⁄ -in. intermediate casing and
improve data accuracy and provides a significant measurements, generated by drilling and 7-in. liner strings to sufficient depths to eliminate
step forward in data quality control (above). circulation, can be problematic for making a string of casing common to wells in the area, in
STC projection logs help engineers accurately accurate acoustic measurements. During a this case, a 5-in. casing string. These efforts not
differentiate compressional, shear and other drillpipe connection, the sonicVISION tool can only reduced well cost, but more importantly,
modes in real time. The novel design of the acquire real-time formation velocity measure- eliminated the difficulties associated with
sonicVISION tool now allows engineers to modify ments in a quiet environment, increasing slimhole drilling and the completion limitations
label limits at the surface thereby improving confidence in the STC projections and potentially inherent in small production casing.
extraction of the compressional ∆t and also allowing engineers to observe velocity changes
providing real-time shear data. Accurate discrim- caused by flow-induced stress variance.

76 Oilfield Review
 2 2 2 2
3 3 3 3
4 4 4 4
5 5 5 5
Depth, 1,000 ft

Depth, 1,000 ft

Depth, 1,000 ft

Depth, 1,000 ft
6 6 6 6
7 7 7 7
8 8 8 8
9 9 9 9
10 10 10 10
11 11 11 11
12 12 12 12
8 10 12 14 16 18 8 10 12 14 16 18 8 10 12 14 16 18 8 10 12 14 16 18
Pressure gradient, lbm/galUS Pressure gradient, lbm/galUS Pressure gradient, lbm/galUS Pressure gradient, lbm/galUS
A B C D

> Telemetry to engineering centers. The wellsite engineer collects drilling, mud and sonic LWD data,
then transmits this information to the engineering center where a team of experts analyzes and
processes the data. Once the results are returned to the wellsite, initial pore-pressure predictions
(A) are updated with pore-pressure estimations (B), ultimately reducing the cone of uncertainty
(C) and providing more accurate predictions of pore pressure ahead of the bit (D).

Accurate prediction of geologic target depth center where the wellbore hydrodynamics and Correlating data acquired from the
and pore pressure is essential to the success of geomechanical earth models were updated in sonicVISION and FPWD tools significantly
aggressive drilling plans. Helis and PPI engineers real time using data from the rig (above). To increased confidence in the real-time pore-
based their initial well design on mud weights account for variations in lithology and sediment- pressure prediction model. These measurements
from wells in the area. They next approached compaction rates, the nonlinear normal compac- allowed predrilling uncertainties associated with
Schlumberger to refine these predictions using tion transform established during predrilling the velocity-to-pore-pressure transform to be
the GOM 3D mechanical earth model, to be planning was validated and recalibrated while properly defined while drilling. The calibrated
further refined with while-drilling sonic data. drilling using sonicVISION data and direct transform was then applied to revise and update
While-drilling data were transmitted by pressure measurements from the FPWD
satellite to a remote operations and collaboration Formation Pressure While Drilling tool.

Winter 2005/2006 77
the predrilling pore-pressure model, both behind 4,000
and ahead of the bit (right).
The results, including the mud-weight
recommendations, were then conveyed to the rig, 5,000
and action was taken to ensure that the surface
mud weight, the equivalent circulating density
6,000
(ECD) and the equivalent static density (ESD)
were kept within the limits of the mud-
weight window. 7,000
The initial requirements for setting the 95⁄ -in.
casing were constrained by a 13-lbm/galUS
[1,558-kg/m3] fracture point derived from 8,000

Depth, ft
previous experiences in the field. However, the
calculated fracture gradient derived from while-
drilling formation velocity and density measure- 9,000
ments indicated that the rock strength was
substantially higher, and capable of accom-
modating a heavier drilling fluid. 10,000
Predrilling pore pressure
The mud weight was increased to Real-time ∆t (sonic) pore pressure
13 lbm/galUS based on the real-time pore- Real-time mud weight
11,000 Real-time equivalent circulating density
pressure analysis as drilling approached 6,800 ft
Real-time ∆t (sonic) fracture gradient
[2,072 m]. Using real-time sonic LWD data, Equivalent static density
while-drilling pressure measurements and 12,000 Formation integrity test
advanced data processing techniques, geo- Formation pressure while drilling data
scientists at the remote collaborative center Casing point
established a safe mud-weight range that allowed 13,000
7 9 11 13 15 17 19
the driller to reach a depth of 8,187 ft [2,495 m]
Pore pressure, lbm/galUS
before running 95⁄ -in. casing; at casing depth, the
ECD was within 0.1 lbm/galUS [11.98 kg/m3] of > Drilling in a narrow mud-weight window. The top of the pore-pressure ramp
the calculated fracture gradient. is confirmed at around 6,800 ft by the sonic (red) and formation pressure while-
drilling measurements (green diamonds). Between 7,000 and 8,000 ft
Once the 95⁄ -in. casing was set, drilling [2,134 and 2,438 m], a significant divergence between the predrilling model
resumed with an 81⁄ -in. bit. At 9,500 ft [2,896 m], (green curve) and actual pore pressure represents an example of the
the annular pressure exceeded the fracture importance of using real-time measurements to update the predrilling model.
gradient, and circulation was lost. Time-lapse As drilling progressed below 9,000 ft [2,743 m], accurate pore-pressure
prediction, pressure measurements and hydraulic modeling allowed the
resistivity analysis indicated two zones near the drilling team to maintain the mud weight (black curve), equivalent static (blue
previous casing shoe where the formation had diamond) and equivalent circulating (purple curve) densities within a narrow
probably been damaged. window just below the real-time fracture gradient (gold curve).
On further evaluation, engineers believed
that the cost of remedial squeeze operations
outweighed the risk of drilling ahead with tight Engineers significantly reduced the uncer- Today’s sonic LWD systems are providing
hydraulic control and a maximum mud weight of tainty associated with pressure-prediction accurate acoustic data that in turn, are being
17.5 lbm/galUS [2,097 kg/m3]. Carefully moni- models by updating the predrilling velocity-to- processed in real time to reliably determine pore
toring and maintaining the annular pressures pore-pressure transform using sonic LWD data pressure and the geophysical limits of formations
within an accurately calibrated hydraulic- and measuring true formation pressure. The being drilled. When combined with seismic and
pressure envelope allowed the operator to com- critical 95⁄ -in. casing depth was pushed 1,187 ft other real-time data, this information helps
plete the well at 12,507 ft [3,812 m] in the target [362 m] deeper than planned, eliminating an geoscientists see ahead of the bit to the next
reservoir without an additional string of casing. entire casing section and reducing well cost by geologic horizon and beyond. Defining the mud-
The combined efforts of Schlumberger, PPI more than US$ 1.7 million. weight window while drilling enables engineers
and Helis engineers eliminated the preplanned to deviate from predrilling casing designs,
5-in. casing string and avoided the difficulties A Sound Future for While-Drilling pushing casing seats to greater depths and
associated with slimhole drilling and completion. Acoustic Tools significantly reducing well cost.
Sonic LWD, while-drilling pressure measurements A new generation of sonic LWD tools is helping Much like the development of sonar early
and careful hydrodynamic monitoring using the drillers, engineers and geoscientists make many in the 20th century, advances in modeling
APWD Annular Pressure While Drilling tool decisions that facilitate safe and cost-effective software, acoustic tool design and decision-
succeeded in identifying pore-pressure changes well construction. By supplying timely infor- processing utilities are helping engineers
and fracture points, and allowed drilling to mation on formation velocity, while-drilling see the unseen and make sound drilling
proceed within the constraints of a narrow mud- acoustic tools have proved to be a valuable asset decisions, reducing cost and increasing well-
weight envelope. to the well-engineering team. construction efficiency. —DW

78 Oilfield Review