SPE/IADC 108272

Borehole Shape Characterization With Azimuthal LWD Measurements: Evaluation,

Applications, and Limitations
P. Butt and A. Fareed, Schlumberger, and S.M. Ma and A.A. Al-Hajari, Saudi Aramco

Copyright 2007, SPE/IADC Middle East Drilling Technology Conference & Exhibition
calipers are that they are derived from LWD logs such as the
This paper was prepared for presentation at the SPE/IADC Middle East Drilling Technology density data. Therefore, they are valid only if the raw data
Conference & Exhibition held in Cairo, Egypt, 22–24 October 2007.
quality is good. In situations where borehole washout is larger
This paper was selected for presentation by an SPE/IADC Program Committee following
review of information contained in an abstract submitted by the author(s). Contents of the
than a few inches, the density data will be compromised.
paper, as presented, have not been reviewed by the Society of Petroleum Engineers or Consequently, the density derived calipers will not be
International Association of Drilling Contractors and are subject to correction by the author(s).
The material, as presented, does not necessarily reflect any position of the SPE, IADC, their accurate.
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers
WL formation imaging tools provide a pair of mechanical
and International Association of Drilling Contractors is prohibited. Permission to reproduce in calipers that are often used for washout and borehole shape
print is restricted to an abstract of not more than 300 words; illustrations may not be copied.
The abstract must contain conspicuous acknowledgment of where and by whom the paper analysis. However, these calipers are limited to two
was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836 U.S.A.,
fax 1.972.952.9435.
orthogonal directions.
This study will concentrate on ultrasonic and density
Abstract calipers and evaluate their applications and limitations. The
Caliper is a fundamental measurement of wellbores. It has focus will be placed on both drilling and geo-stress
significant impacts on decision making regarding well characterization since both will have significant effects on
completion and monitoring operations. Traditional Calipers borehole size/shape. Real-time borehole size/shape analyses
are measured mechanically, thus can not be performed while have the potential to allow a driller to optimize the drilling
drilling since they will not be able to survive the harsh practice to improve borehole quality and drilling efficiency.
conditions of drilling while the bottom hole assembly (BHA) Borehole information while drilling allows engineers to make
rotates. With logging while drilling (LWD) measurements, better decisions regarding well completion and cement volume
several methods have been developed over the years to derive computation.
borehole calipers or hole shape indicators from LWD logs. In The main objectives of this paper are to better understand
this paper, we will compare the LWD derived calipers with the causes of hole-size variations for horizontal wells and
that obtained from wireline (WL) and summarize the investigate how these variations may relate to drilling process
developments, utilizations, and limitations of LWD derived and/or rock mechanics. Attempts were also made to better
calipers. Factors affecting LWD calipers will also be understand effects of borehole shape on petrophysical
discussed to help operation petrophysicists for better log data measurements and methods to remove these environment
quality control. effects.
Examples will also be shown for using LWD calipers to Density Caliper Determination
identify borehole washout, elipticity, breakout, and spiral-hole LWD density-neutron tools (Fig. 1) provide compensated
conditions. Knowing these borehole conditions in real time density and neutron measurements for quantitative formation
can help to make action plans to improve drilling practices and evaluation while drilling. The apparent density (ρapp) from a
reduce drilling risks. After drilling, LWD calipers are used to single detector is a function of formation density (ρb), mud
correct LWD logs for borehole size effects and guide well cake density (ρmc), and the thickness of the mud cake (tmc), as
completions. illustrated in Fig. 2.
Introduction
Logging While Drilling (LWD) derived calipers (ultrasonic, k − tmc t
density, neutron, and resistivity) have been available for
ρapp = ρb + ρmc mc (1)
k k
several years (Maeso and Tribe, 2001), but their utilizations
are still limited to LWD log data quality control due to the
Where k, with a unit of length, is a function of both mud
belief that they are not as reliable as the mechanical calipers
density and mud photoelectric factor (Pe), and is determined
obtained from wireline (WL) logs. In fact, LWD calipers have
from the tool characterization data base. The value for k is
their advantages over the WL calipers. One example is that
typically ranging between 3 and 4.
they cover the whole circumference of the wellbore, thus
From Eq. 1, it is seen that ρapp = ρb when tmc = 0, i.e., the
provide a better characterization of borehole shape. A 3-D
detector density measures formation density if mud cake effect
caliper image can be obtained using 16 radii density
measurements around the wellbore. Major limitations of LWD
2 SPE/IADC 108272

is zero. The higher the tmc, the more effect of ρmc to ρapp. The Where So,D(b), So,D(u), So,D(r), So,D(l) are the standoffs in the
thickness of mud cake can be derived from Eq. 1 as, bottom, up, right, and left quadrants, respectively (Fig. 1B).

ρb − ρapp Factors Affecting Density Caliper

t mc = k (2) The azimuthal density and caliper can be acquired in both oil
ρb − ρmc
and water based mud systems. Comparison with WL calipers
indicates that the LWD derived azimuthal density caliper can
The distance between the detector and the formation is be reliable up to 3 inches over the bit size under optimal
often called tool standoff (So). In case of WL density logging, conditions.
the thickness of the mud cake is equal to the tool standoff (tmc Mud properties. In addition to the depth of investigation of
= So) since the pad tool is pushed against the borehole wall density measurements, the most critical factors limiting LWD
LWD density measurements are acquired using two derived caliper performance are the mud Pe and density. A
detectors. The bottom formation density (ρbot) is derived from mud with high Pe will affect k, thus the apparent density ρFar.
the far detector (ρFar), by adding an empirical correction term Also, as mud weight (ρNear) increases, the sensitivity of
Δρ that is a function of the discrepancy between the near and apparent density (ρFar) to stand-off (So,D) is reduced (Eqs. 1
far detector densities (Fig. 2). and 4). Experimental data show that LWD density derived
standoff is generally reliable for mud below 12 ppg (pound per
Δρ = f ( ρ Far − ρ Near ) gallon) and tends to be less accurate for mud heavier than 14
(3)
ρbot = ρ Far + Δρ ppg (Ortenzi et al, 2002).
In high Pe muds, these limits in mud weight are
In Eq. 3, Δρ can be either positive or negative depending significantly reduced. However, the current experimental
on mud properties. In high angle and horizontal boreholes, the database does not comprise of all possible combinations of
bottom quadrant density (ρbot) typically gives the best estimate mud weight and Pe. Therefore, the value of k is optimized for
of formation density. Assuming that the near detector density the most common range of mud properties.
(ρNear) measures the mud density (ρm), from Eq. 2, the density BHA rotation vs. sliding. As the azimuthal data are available
standoff (So,D) can be expressed as; only when the tool is rotating, the density caliper can not be
obtained when the drill string is sliding. The term “sliding”
refers to non-rotation of the bottomhole assembly (BHA), such
ρ bot − ρ Far
So,D = k (4) as drilling with a mud motor or tripping into or out of a well.
ρ bot − ρ Near The density calipers (horizontal and vertical) are available
in real time. Their accuracy depends upon the quality of the
One advantage of the new tool over the older one is its density measurements. The 16 radii array data can also be
enhanced azimuthal capabilities by computing density and Pe computed directly at the wellsite, if requested.
in sixteen individual sectors (Fig. 1B). That allows stand-off
from each specific sector to be calculated as: Ultrasonic Caliper Determination
The ultrasonic caliper is more commonly used for borehole
size characterization. Thus the theory of the ultrasonic caliper
ρ bot − ρ Far (i )
So,D ( i ) = k (5) is well known. The sonic sensor operates in a “pulse/echo”
ρ bot − ρ Near ( i ) mode to emit and receive an ultrasonic signal. An electrical
pulse is applied to a piezoelectric ceramic which generates an
These 16 radii measurements of the borehole can be ultrasonic wave. The sonic wave travels through the mud,
displayed as borehole shape images. A pseudo 8-arm caliper reflects off the mud/formation interface, and returns to the
can easily be derived from the 16-sector data by adding the ceramic and generates an electric signal (Fig. 3A). The sensor
tool size (Tsize) to the sum of opposite-sector standoffs: electronics measure the time elapsed between the transmission
and the arrival of the sonic wave to obtain its time of flight.
The standoff distance between the sensor and the formation is
CDi = Tsize + S o,D( i ) + S o, D( i +8) (6)
related to the time of flight by the speed of sound in mud (Fig
3B). For a given tool orientation, this gives a standoff in the
Where So,D(i) is the standoff of the ith section and So,D(i+8) the direction the tool is positioning. The new ultrasonic sensor has
standoff of the i+8 section. an accuracy of ± 0.1 inch.
These sector-data are subsequently grouped into four In tools such as azimuthal density-neutron tool,
quadrants (up, left, bottom, and right). The vertical (CDve), magnetometers are available to allow binning of the ultrasonic
horizontal (CDho) and average (CDav) calipers are then stand-off data into quadrants. Rapid sampling allows the
computed as following: construction of a stand-off histogram at each depth and for
each quadrant. Minimum, maximum and average quadrant
CDve = Tsize + S o,D(b ) + S o, D( u ) (7) stand-offs are obtained from a histogram statistical analysis
CDho = Tsize + S o, D(l ) + S o, D(r ) (8) for quality control purposes. Oriented (horizontal and vertical)
borehole diameters are computed by combining opposite
CDav = 0.5(CDve + CDho ) (9)
SPE/IADC 108272 3

standoffs with the tool diameter, similar to the derived density caliper obtained at the time of drilling shows the best borehole
caliper as discussed above. conditions. If decisions are to be made about hole volume or
completion design, then it is advisable to obtain a caliper as
Factors Affecting Ultrasonic Caliper Measurements close to the required completion time as possible. It is
A critical limiting factor for ultrasonic caliper measurement is recommended that for making well completion decisions, a
mud weight. Heavy mud attenuates the sonic pulse, thus repeat log should be made while coming out of the hole on the
reduces the amount of measurable standoff. Also, the strength final run. Comparing the repeat caliper pass with the original
of the reflected sonic wave decreases as the acoustic one gives confidence on the stability of the borehole.
impedance of the mud approaches that of the formation, which Decisions may be made regarding whether the borehole needs
often prevents accurate echo detection. to be cleaned or conditioned.
Other effects that compromise the accuracy of the
ultrasonic caliper are eccentering, cuttings and borehole wall
Stand-off Effect on LWD Data
rugosity (both scatter and reflect the sonic wave), and gas. If
LWD data is commonly used for placing difficult to drill wells
the tool is equipped with a single transducer, the drilling string
in target reservoirs. These LWD measurements are often
must be rotating in order to get meaningful results.
affected by borehole washout and rugosity. Understanding
Finally, the accuracy of this technique depends upon the
these effects on LWD data is vital for making real-time
value used for the speed of sound in mud, which is a function
geosteering decisions.
of several environmental parameters, such as mud type, mud
Well-C was drilled in thin limestone reservoir which was
density, mud salinity, pressure, temperature and gas content.
sandwiched between two anhydrites (Fig. 7). Density and
These parameters are normally derived from different sources
resistivity images were acquired in real-time for geosteering.
and they may not represent the real down-hole conditions.
Density derived caliper indicated up to 2 inches hole-
Where signal is good, the new ultrasonic sensor has an
enlargement. Density and shallow resistivity images were
accuracy of ±0.1 inch. In a 10 ppg mud, the ultrasonic caliper
affected by borehole rugosity and breakout at the top of the
will be accurate up to 3 inches from the tool face. In a 14 ppg
borehole and this affected the top anhydrite detection. Deep
mud, this is reduced to 1.5 inches. Vertical resolution is 6
resistivity image with depth of investigation of 5 inches was
inches when drilling at 90 feet per hour.
not affected by this borehole breakout and a geosteering
decision was made after detecting the approaching top
3D Visualisation of Borehole Shape anhydrite (Mudhhi, et al., 2005).
Conventionally, caliper information has been displayed as a In another example as shown in Fig. 7 (Farwaaz, et al,
1D log curve. Availability of the LWD 16-sector caliper 2006), a well with dual lateral boreholes was planned in
around the borehole makes it possible to generate a borehole Khuff-C gas reservoir in a structure with high formation dip.
shape in 2 or 3 dimensions, making it easier for visualization A high angle pilot hole was first drilled to identify the best
of the borehole shape thus aiding interpretation of the borehole porosity layer. Borehole shape analysis using ultrasonic
variations. The different ways of visualizing the borehole calipers indicated borehole ovalization along top and bottom
calipers are shown in Fig. 4. of the borehole. This indicated that density sensor is
Improved real-time data transmission makes it possible to contacting the formation along the left and right quadrants of
transmit and visualize multidimensional borehole shape at the the borehole. In this case, bottom quadrant density is not
time of drilling. It is even more powerful to display borehole representative of the formation density. Image derived density
images with lithological and borehole shape information. A (Radtke et al, 2003,) was more representive of true formation
further development is to display the calipers as textured 3D response, thus used for best porosity layer identification and
images. horizontal well placement (Fig. 7A). In this well, breakout
direction is vertical suggesting the horizontal stress σh is
LWD vs. WL Caliper greater than the vertical stress σv (Fig. 7B).
The mechanical WL single and dual calipers were acquired in The above conditions are commonly observed where the
a number of wells to validate the LWD derived calipers. hole is consistently enlarged and increased cuttings are
Comparison results are shown on Figs. 5 and 6. In well A, generated. Understanding and recognizing the causes of the
drilled in a clastic formation with an 8.5 inch bit. The WL variation in borehole shape will provide more efficient
caliper indicated a maximum borehole washout up to 2 inches. drilling. This is particularly important in differentiating
LWD density derived calipers were in good agreement with borehole shape changes due to drilling or due to formation
the WL calipers in this case. In sliding intervals around X700 stress re-distribution.
and X800, LWD calipers could not be computed as density Environment Corrections
azimuthal data was not available. Horizontal well placement in carbonate reservoirs is generally
In well B (Fig. 6), drilled in a carbonate formation with a targeted good porosity layers, in other words horizontal well
6.125 inch bit, LWD density derived caliper indicated an placement is a porosity geosteering. Presence of borehole
almost gauged borehole while drilling. The WL dual caliper enlargement adversely affects density and neutron porosity
suggested that the borehole was breaking out in intervals measurements. Borehole size information is critical for log
around x000 and X800. quality control and corrections are required for computation of
The shape of the borehole commonly changes with time, accurate porosity. Fig. 8 shows a well drilled in Arab-D
corresponding to redistribution of near-wellbore stresses. A carbonate, the neutron and bottom quadrant density were
4 SPE/IADC 108272

affected by borehole washout. Density measurement was unconsolidated formations. Hole-angle is built with minimal
reading mud response only and neutron porosity reads around dependence on the borehole for deflection. It is a fully
35 pu across the porous interval X350-X610. Porosity range rotating tool with no stationary external parts, which reduces
from nearby offset wells was about 20-22 pu. LWD derived the risk of sticking pipe. It looks and acts very much like a
calipers from density and ultrasonic were obtained. Ultrasonic conventional motor BHA although it has the ability to adjust
calipers showed maximum hole-washout up to 4 inches and azimuth and inclination of the bit during drilling. Fast drilling
density caliper showed that washout up to 2 inches. Analysis gives a gauged hole and consistent speedy-drilling gives a
indicated that density caliper was underestimating hole smoother hole. A RSS assembly would remove borehole size
washout due to measurement depth of investigation. variation between sliding and rotating with a mud motor.
Ultrasonic caliper was then used to perform environmental However, RSS is commonly utilized to achieve higher rate of
correction on neutron porosity. Corrected neutron readings are penetration in hard formations.
much inline with porosities from offset wells. To compare holes drilled with RSS with that drilled with a
conventional motor, Fig. 11 shows a 3D textural display of
Effect of Drilling Process on Borehole Shape
borehole shape for two sections of a horizontal well, drilled in
2D and 3D borehole shape analyses were performed in a
a deep carbonate reservoir. First section was drilled with mud
number of wells using ultrasonic caliper data. Results of the
motor that shows significant borehole wash outs (Fig. 11A).
study indicated that drilling process is one of the key factors in
The second section was drilled with RSS and a very smooth
controlling borehole shape variation.
borehole was drilled (Fig. 11B).
Borehole shape analysis results from a well drilled with
mud motor are presented in Fig. 9A. Upper section was Conclusion
drilled while rotating, and the middle section was drilled while A reliable borehole caliper measurement is required for
sliding first to build angle and then to ream down while drillers, petrophysicists, and geosteering engineers. LWD
rotating to acquire azimuthal density image. Bottom section derived density/ultrasonic calipers and borehole images
was drilled while rotating. The borehole for sections drilled provide encouraging results and can help to better understand
while rotating is smooth and in gauge and sections drilled with influence of drilling process on borehole shape, stress regime,
sliding/reaming-down showed borehole washout. Density and and, to some extent, the lithology of formation.
2D caliper image also indicated that the upper side of the hole Comparing to motor drilling, using rotary steerable system
was washed out. Example in Fig. 9B was from the same well. often produces very smooth boreholes.
Here the spiral borehole was associated with change in Spiral boreholes are often found to be associated with
lithology. change in lithology.
Availability of 2D and 3D borehole images enhances
Effect of Lithology on Borehole Shape
interpretation and understanding of borehole stability issues. It
In clastic reservoirs, one of the principle causes of drilling
can be utilized to optimize drilling techniques and provide
delays and, in some cases, even suspension of the well prior to
smoother borehole for expert completions.
reaching the target total depth is wellbore instability related to
shale. Chemical instability problems arise from the interaction
Acknowledgments
between the wellbore fluid and the formation. Shale, a
chemically active sedimentary rock, reacts with water. As a
The Authors would like to thank Saudi Aramco for their
result, mechanical properties change proportionally with the
permission to publish this work. We would also like to thank
volume of moisture absorbed. The property alteration is
D. Kersey from Saudi Aramco and R. Hussain and A. Alvi
nonlinear and is related to the moisture induced moduli change
from Schlumberger for their review and comments.
(Bol et al, 1994, Santarelli et al, 1992, and Wanget al, 1994).
Fig. 10 shows wellbore stability analysis results from a
well drilled in a clastic reservoir. The well was drilled with a References
new multi-functional LWD tool (Fawwaz, et al., 2007).
Detailed clay mineralogical analysis was carried out with 1. Maeso, C. and Tribe, I: “Hole Shape from Ultrasonic
spectroscopy measurements and 3D textural borehole shape Calipers and Density while Drilling,” SPE 71395, SPE
displays were obtained with measurements using dual Annual Technical Conference and Exhibition, New
ultrasonic calipers available with this new technology. The Orleans, Louisiana, 2001.
example in Fig. 10A showed an advanced formation 2. Ortenzi, L., M. Evans, and C.J. Maeso: “An Integrated
evaluation where detailed mineralogy is obtained from Caliper from Neutron, Density, and Ultrasonic Azimuthal
spectroscopy measurements. Fig. 10B showed borehole break- LWD Data” SPE 77479, SPE Annual Technical
outs in sand-shale stringers and Fig. 10C showed spiral Conference and Exhibition, San Antonio, Texas, 29
borehole associated with carbonate formation. September–2 October 2002.
3. Al-Mudhhi, M.A., S.M. Ma, A. Al-Hajari, K. Lewis, G.
Rotary Steerable System (RSS)
Berberian, P. Butt, and P. Richter: “Geosteering with
The new rotary steerable system (RSS) has a steering
Advanced LWD Technologies - Placement of Maximum
assembly that continuously orients the tilted bit shaft to
Reservoir Contact Wells in a Thinly Layered Carbonate
control the drilling direction and the dogleg severity of the
Reservoir”, IPTC-10077, International Petroleum
borehole. The tool does not push against the borehole to build
Technology Conference, Doha, Qatar, 21–23 November
angle, therefore is more effective for steering in
2005.
SPE/IADC 108272 5

4. Al-Fawwaz, A., N. Al-Mashrafi, P. Butt and A. Fareed:

“Formation Evaluation While Drilling of a Complex
Khuff-C Carbonate Reservoir in Ghawar Field, Saudi
Arabia”, SPE-105232, SPE Middle East Oil Show,
Bahrain, 17–20 March 2007.
5. Radtke, R. J., R. A. Adolph, H. Climent, L. Ortenzi, N.
Wijeyesekera: “Improved Formation Evaluation through
Image-Derived Density” SPWLA 44th Annual Logging
Symposium, June 22-25, 2003
6. Bol, G.M., S.W. Wong, C.J. Davidson, and D.C.
Woodland, “Borehole Stability in Shales”, SPE Drilling
& Completion, June 1994, pp. 87-94.
7. Santarelli, F.J., M.E. Chenevert, and S.O. Ossisanya,
“Stability of Shales and its Consequences in Terms of
Swelling and Wellbore Stability” IADC/SPE Drilling
Conference, New Orleans, Louisiana, Feb. 18-21, 1992.
8. Wang, C. L., C.H. Yew, and M.E. Chenevert, “The
Stability of a Deviated Wellbore in Shale Rocks”, Proc.
8th International Conference on Computer Methods and
Advances in Geomechanics. (Eds), 1994, H.J.
Siriwardane and M.M. Zaman, Balkema, Rotterdam, pp.
2195-2220.
6 SPE/IADC 108272

(A)

Far detector F
Near detector N
Source S

(B)

Up
Right
Near and
Left
Far
Down
Detectors

16 Sectors

Fig. 1 Density measurements are made with two detectors (Fig. 1A) which
provide azimuthal measurements, thus density image, while the LWD BHA
rotates (Fig. 1B). Complex horizontal wells can be geosteered with the help
of the azimuthal densities and/or density image.
SPE/IADC 108272 7

Far
Detector

ρ m ρa,ρ Far

ρ
Near
Detector
ρb
Near
tm

Source

Fig. 2 A LWD caliper can be derived from density measurement. Difference

between far detectors’ readings and formation density is due to stand-off. Knowing
mud density allows standoff to be computed.

(A) (B)

Fig. 3 A LWD caliper can also be derived from ultrasonic measurement.

High frequency sonic waves are transmitted by orthogonal transmitters and by
knowing the mud velocity and measuring the two way travel time for the signal
allow computation of standoff .
8 SPE/IADC 108272

(A) (B)
x445ft

x
x400

x450

3D View

Fig. 4 Well D: Track-1, caliper image from ultrasonic sensors, Borehole shape using
ultrasonic calipers (vertical and horizontal, track 2) indicates borehole ovalization in
top and bottom quadrants. Density (vertical and horizontal) calipers, track-3 reads
lower than ultrasonic calipers. Track-4, density measurements for top and bottom
quadrants are affected by breakouts. Image derived density is true representative of
formation response and heavier than both top and bottom densities.

x050

x100

x150

x200

Fig. 5 Well A: Comparing LWD density caliper (red in track 2) with wireline
mechanical calipers (light and dark blue in track 2) shows good match.
SPE/IADC 108272 9

x200

x600
FMI Caliper

x1000

Density Caliper
x1400

x1800

x2200

Fig. 6 Well B: Comparing LWD density caliper (black) with wireline

mechanical calipers (red) shows that they generally agree across the
bottom section. However, at the top, the wireline caliper is significantly
larger; probably due to borehole breakout after LWD logging.

1 ” RI 5 ” RI DI

Fig. 7 Well C: LWD density caliper (light blue curve in track 5) indicated
hole washout of about 2”. LWD resistivity image with 1” depth of
investigation was also affected by hole washout, while resistivity image
with 5” depth of investigation was not (after Mudhhi, et al., 2005).
10 SPE/IADC 108272

x200

x400

x600

Fig. 8 LWD density and neutron logs were effected by borehole washout (track 3).
Better neutron porosity was obtained after applying borehole size correction using
ultrasonic calipers. Density caliper was underestimating due to measurement depth
of investigation (track 1).
SPE/IADC 108272 11

(A)
(B)

Rotating zone
Good borehole
x500
x50 Sliding zone
higher borehole

Borehole
enlargement

Spiral
Rotating zone
x550
x100 Good borehole

Fig. 9 2D Borehole shape analysis using density calipers indicated that borehole
washout was often associated with BHA sliding. Spiral boreholes were related to lithology
change in this example.

(A) (B)

(C)

Fig. 10 Detailed mineralogical and 3D caliper analysis with multifunctional logging in this
example indicated that borehole breakout and spiral hole were related to mineralogical and
chemical composition of the rock.
12 SPE/IADC 108272

(A) Borehole was enlarged with high degree of rugosity in a section drilled with motor

(B) A generally gauged and smooth borehole was achieved with RSS

Fig. 11 Borehole shape can be directly related to the drilling process. A showed a
enlarged rugose borehole section drilled with mud motor, while B showed a generally
smooth and gauged borehole drilled with rotary steerable system.