You are on page 1of 10

PO173103 DOI: 10.

2118/173103-PA Date: 15-March-16 Stage: Page: 1 Total Pages: 10

Analysis and Quantification of


Wellbore Tortuosity
Jon Bang, Onyemelem Jegbefume, Adrián Ledroz, and John Weston, Gyrodata Incorporated; and
Jay Thompson, SandRidge Energy

Summary many applications, we find it somewhat narrow, because it limits


Small-scale wellbore tortuosity—variations in attitude on a length the evaluation of tortuosity to a comparison between the outcomes
scale smaller than standard survey intervals of 30 m (100 ft)—is of two well-construction phases: planning and drilling (including
generally neglected because of its small effect on the final posi- surveying). It should be emphasized that tortuosity is purely a geo-
tion of the well and its unclear relation to traditional dogleg sever- metric property of the wellbore. Thus, the process that created the
ity (DLS). However, it is well-known that such tortuosity may tortuosity (drilling) is subordinate to this, and the reference should
have significant influence on the drilling process and on drilling be more basic than a planned trajectory. This leads us to propose
efficiency. Furthermore, it is a crucial factor for the design and in- the following definition, which we will use throughout this paper:
stallation of completions and production equipment, because a “Tortuosity is any deviation from a straight line.”
highly tortuous wellbore section, depending on borehole diameter This definition can be applied in two or three dimensions; fur-
and tortuosity amplitude and along-hole distribution, may exert thermore, it implies that tortuosity may exist on any length scale,
strong bending forces on such equipment, or high friction on mov- as illustrated in Fig. 1.
ing parts.
This paper describes a novel methodology for analyzing the Why Analyze Tortuosity? Small-scale tortuosity [i.e., tortuosity
tortuosity of openhole wellbores, casing, drillstring, and produc- occurring on intervals significantly shorter than a drillstring stand
tion tubing. Several related tortuosity parameters are described, (30 m)] has been difficult to quantify because, in general, high-re-
and examples of application to field data are included. The meth- solution data are not collected in traditional surveys. Nevertheless,
ods use high-resolution survey data [measured depth (MD), incli- such tortuosity is recognized as an important factor in the drilling
nation, and azimuth], which may, in principle, originate from any process (Chen et al. 2002; Samuel et al. 2005; Mason and Chen
surveying tool or service capable of providing such data. The 2006). High levels of tortuosity may imply significantly increased
methodology requires input data from a single survey only. On friction factors and increased torque-and-drag values (Skillingstad
the basis of a user-defined length as single external parameter, tor- 2000) and excessive casing wear (Samuel and Gao 2013). In
tuosity can be analyzed on any length scale greater than approxi- severe cases, this may lead to stuck pipe or limit the distance of
mately 10 times the input survey-data interval, and with a extended-reach-drilling operations (Stuart et al. 2003). To analyze
maximum resolution equaling twice the data interval. The proc- and optimize the drilling process, advanced computer models
essed parameters include relative elongation of a tortuous section have been developed, which take into account the interactions
compared with a straight-line section, transverse displacements between bottomhole assembly (BHA), drillstring dynamics, fluid
from the straight line, and maximum available diameter for a and formation properties, and wellbore geometry (Liu et al. 2004;
downhole device caused by the small-scale bendings of the sec- Menand et al. 2009; Marck et al. 2014). Such models may predict
tion. The results can be displayed as graphs vs. MD, or as 3D- the resulting small-scale tortuosity for a particular drilling sce-
rendered views of the actual wellbore or tubing shape. nario (Marck and Detournay 2015), or they use tortuosity as one
Results from various field cases are included, in which the tor- of many input parameters (see the subsection on DLS).
tuosity analysis was applied to high-resolution continuous gyro Tortuosity may also affect the completion and production
survey data collected in cased wellbores. In all the field cases, the phases of a well. The large-scale wellbore curvature is normally
novel methods revealed sections of considerable tortuosity that designed for problem-free transportation and installation of down-
were either unnoticed, or located with unacceptably low accuracy, hole devices (Reid et al. 2013). However, unforeseen and possibly
by conventional methods. These results led to re-evaluation of the undetected small-scale tortuosity may cause significant reductions
planned locations for completion and production equipment. in the effective inner diameter of the wellbore (cased or openhole).
The characterization of the wellbore in terms of tortuosity on This may lead to stuck equipment, or bending moments exceeding
various length scales may be of crucial importance for the func- the manufacturer’s specifications, with reduced functionality or
tionality and lifetime of permanently installed equipment. For lifetime as potential consequences. A concrete example is the use
example, identification of highly tortuous sections will aid the of artificial-lift methods for the extraction of oil with a rod-acti-
placement of rod-guide wear sleeves, increasing the rod and cas- vated pump. This operation relies on low friction between the rod
ing life and reducing the workover frequency. Another application and the tubing wall; however, severe wear may result from sharp
is the identification of low-tortuosity sections in which downhole bends and other directional changes in the wellbore profile (Mat-
pumps or other equipment will not be subject to excess bending. thews and Dunn 1993). The rod and tubing wear can be reduced
In addition, the tortuosity results may help evaluate the drilling significantly by positioning rod guides at the most effective loca-
equipment and the drilling process. tions along the wellbore (i.e., typically, at highly tortuous sec-
tions). It is therefore crucial to identify these sections and to
analyze the wellbore geometry of such regions in detail.
Introduction Because the tortuosity-related problems in the completion or
Definition of Tortuosity. Gaynor et al. (2001) defines wellbore production phases are caused by the actual geometry of the drilled
tortuosity as “the amount by which the wellbore deviates from the wellbore, they can be addressed properly only through case stud-
planned trajectory.” Although this definition may be useful in ies involving survey data from the actual well. To our knowledge,
no method for such analysis was demonstrated to date.
Copyright V
C 2016 Society of Petroleum Engineers

This paper (SPE 173103) was accepted for presentation at the SPE/IADC Drilling DLS. First, we need to warn against confusing the tortuosity phe-
Conference and Exhibition, London, 17–19 March 2015, and revised for publication. Original
manuscript received for review 21 November 2014. Revised manuscript received for review
nomenon itself, as defined previously or by similar definitions,
11 June 2015. Paper peer approved 6 September 2015. with parameters used for characterization of the phenomenon.

2016 SPE Production & Operations 1

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 2 Total Pages: 10

92
HRCG
MWD
91.5

Inclination (degrees)
91

Large scale 90.5


(wellbore) Medium scale Small scale
100–1000 m (survey interval) 1–10 m
10–100 m 90

Fig. 1—Wellbore tortuosity may occur on any length scale, here


exemplified by wellbore scale (left), survey scale (middle), and 89.5
1100 1150 1200 1250 1300 1350 1400 1450 1500 1550 1600
subsurvey scale (right). Large-scale tortuosity corresponds to
planned wellbore curvature. MD (m)
(a)

Tortuosity can be described theoretically as periodic or statisti- 15


cal fluctuations superimposed on a smooth wellbore trajectory HRCG
MWD
(Sugiura and Jones 2008; Samuel and Liu 2009). More important
in our context are tortuosity parameters derived from survey data.

DLS (degrees/30 m)
Commonly used parameters are the DLS and its inverse, the ra- 10
dius of curvature (Sawaryn and Thorogood 2005). DLS is defined
as the angular change in wellbore direction over a certain MD
interval, divided by the interval (the course length). It is usually
expressed in units of (degrees/100 ft) or (degrees/30 m).
Survey-derived DLS, and more sophisticated parameters that 5
are derived from DLS, are widely used in the drilling-optimiza-
tion computer models mentioned previously (Samuel and Liu
2009; Brands and Lowdon 2012; Mitchell et al. 2015). However,
it has been realized that dogleg parameters alone are difficult to
0
use as reliable tortuosity indicators (Matthews and Dunn 1993; 1100 1150 1200 1250 1300 1350 1400 1450 1500 1550 1600
Oag and Williams 2000; Brands and Lowdon 2012). In particular, MD (m)
Matthews and Dunn (1993) pointed out that “… dogleg severity (b)
values provide a measure of only absolute curvature …”, imply-
ing that important information on the curvature’s orientation in Fig. 2—(a) Inclination data from a conventional MWD survey
3D space is not conveyed by the dogleg parameter alone. with resolution of 10 to 30 m (red curve), and a high-resolution
Another fundamental problem is the noisy character of DLS continuous gyro (HRCG) survey with resolution 0.3 m (black
for short course lengths. Surveyed DLS is traditionally based on curve). (b) DLS on the basis of inclination and azimuth data
directional data at MD intervals corresponding to one drill stand from the same two surveys.
(30 m) or one individual section (10 m). However, attempts to
evaluate DLS at shorter intervals (down to 1 m or lower) show needed for calculation of the wellbore position. However, high-re-
that it becomes unstable as the interval is decreased. An example solution surveys with continuous MWD or continuous gyro instru-
is shown in Fig. 2. This effect was also observed by Sugiura and ments were run for various reasons, such as improved wellbore
Jones (2008) and by Brands and Lowdon (2012); the latter authors positioning (Stockhausen and Lesso 2003), formation evaluation
compare DLS curves over course lengths of 3 and 30 m and con- (Bordakov et al. 2009), drilling-performance evaluation (Sugiura
clude: “Although it is possible to obtain survey data in shorter et al. 2013), and evaluation of load-induced drillstring buckling
(1–3 m) intervals today with continuous MWD surveys, it is diffi- (Menand et al. 2006; Weltzin et al. 2009; Mitchell and Weltzin
cult for engineers to interpret the presented hole curvature (DLS). 2011). In the latter of these applications, the development of
The reported hole curvature can vary significantly on listings of buckling was evaluated by comparing different surveys, not by
the same borehole depending on the reported survey interval.” analyzing a single survey, which is central to the methods
The noise is an intrinsic consequence of the DLS algorithm: described here.
On the basis of the directions at the wellbore section’s endpoints The high-resolution data presented in the papers mentioned
only, the direction difference may change significantly when cal- previously and in similar papers are typically limited to inclina-
culated over a slightly different length, and this change is ampli- tion and azimuth data, and no attempts are made to process tortu-
fied when divided by a short course length. This contradicts the osity characteristics from these data. To the authors’ knowledge,
behavior that should be expected by tortuosity as a physical phe- no comprehensive method for tortuosity analysis on the basis of
nomenon—namely, a smooth variation with length scale. high-resolution survey data was ever published.
To overcome these limitations of DLS, this paper presents new
parameters for characterizing tortuosity—in particular, small-
scale tortuosity. Theory
Preprocessing. Raw Data Set. One basic implication of the defi-
High-Resolution Survey Data. The tortuosity analysis demon- nition of tortuosity used here is that tortuosity can, in principle,
strated in this paper was applied to real continuous gyro survey exist on any length scale. Small-scale tortuosity (on a length scale
data, sampled at intervals of typically 0.3 to 1 m (1 to 3 ft). How- smaller than traditional survey intervals of 10 to 30 m MD) will
ever, the methodology itself does not require any specific tool or be the primary subject of the analysis presented in this paper,
service. Any quality-checked survey data set with sufficiently although the methods can be applied on any length scale, as long
high resolution can be used as input, from measurement while as a sufficient number of data points is available.
drilling (MWD) gyro, inertial or other tools. High-resolution data It is assumed that the data are provided by a high-resolution
are not usually collected in standard surveys, because they are not wellbore survey [i.e., at intervals of typically 0.3 to 1 m (1 to 3

2 2016 SPE Production & Operations

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 3 Total Pages: 10

W Large-scale T Small-scale T

L S
Survey Fitted straight line
stations L

Fig. 3—Establishing a local reference line for the wellbore tra-


jectory: An analysis window W of constant MD length is moved Fig. 4—Tortuosity T can be defined by the ratio between actual
stepwise down the wellbore, and at each location of W, a (along wellbore) length S and straight-line length L. The dashed
straight line is fitted through the position coordinates of the vertical line approximately separates between a region with
survey stations covered by W. The illustration shows a wellbore large-scale tortuosity (left) and a region with small-scale tortu-
section with large-scale tortuosity on the left-hand side, and osity (right). The well path will, in general, be 3D, and S and L
small-scale tortuosity to the right. The well path will, in general, are calculated in 3D space.
be 3D, and the straight lines are fitted in 3D space.
Smoothed Data Set. On the basis of the straight reference
lines produced at each location of the window W, we may derive
ft)]. MD, inclination, and azimuth values are collected at these a new wellbore trajectory where the small-scale tortuosity is sup-
intervals along the wellbore section of interest. It is further pressed, hence, the term “smoothed data set”. The basic criterion
assumed that the data have passed standard quality-control proce- for one point to belong to this data set is that it lies close to the
dures. No correction is needed for the possible nonuniformity of centerpoint of the raw data subset used to establish the respective
the MD data. The survey data are converted into position incre- reference line. The obvious choice is the “center of gravity” point
ments dNr,j (north), dEr,j (east), dVr,j (vertical) at each measure- (N0 E0 V0) (Eq. 1) for each window location, although other defi-
ment station j, where the subscript r indicates that these are raw nitions may be used. The set of smoothed data points will trace
data (i.e., input data to the analysis). Conversion to position is out the overall (i.e., large-scale) shape of the wellbore, and this
performed by the tangential method (API D20 1985). This set will be almost insensitive to single raw data outliers. By their
method is not recommended for normal survey intervals (10 to very nature as “outliers,” such points should be considered as con-
30 m) because of poor accuracy; however, for the short-interval stituting the small-scale tortuosity.
data, initial comparison with other methods showed that the con- The length scale indicating the transition from large scale to
version method has no significant effect on the tortuosity results. small scale, and the degree of suppression of small-scale tortuos-
The positions Nr,j, Er,j, Vr,j obtained by cumulating the incre- ity, will be closely related to the length of the W window.
ments are termed “the raw data set.” It is assumed that these data Similar to the raw data set, the smoothed data set can be sub-
constitute a true representation of the wellbore trajectory. No ject to the processing of the tortuosity parameter T (below). This
other data, in particular, no knowledge of the drilling process, are is not pursued further in this paper. The possible benefits of proc-
needed in the analysis. essing both the raw and the smoothed data sets are outlined in the
Local Reference Line. An analysis window W is defined Discussion section.
along the wellbore (Fig. 3), with an MD length of typically LW 
10 to 30 m (30 to 100 ft). As the window is subsequently shifted
along the wellbore, its length is kept constant, except for negligi- Quantification of Tortuosity. Tortuosity Parameter. Tortuosity,
ble adjustments so that the window always begins and ends on a described as “a factor that accounts for effective elongation of fluid
survey station. The window typically covers 30 to 100 stations. A paths”, has long been an important parameter in the study of flow
straight line is fitted through the position coordinates of these sta- in porous media in petrophysics (Matyka and Koza 2012). In its
tions, by a least-squares fit in three dimensions, with equal free- most basic form, tortuosity T was initially described by the pioneers
dom in all directions. Mathematically, this is equivalent to a Kozeny and Carman (hence, the subscript KC) by the formula
principal-component analysis (Davis 1986). The straight line will
pass through the point given by the arithmetic means N0, E0, V0 of TKC ¼ Spore =Lslab ; . . . . . . . . . . . . . . . . . . . . . . . . . . ð3Þ
the coordinates:
where Spore is the distance along a flowline (following a pore)
1 XMW 1 XMW through a porous slab of thickness Lslab. More sophisticated for-
N0 ¼ N ;
j¼1 r; j
E0 ¼ E ;
j¼1 r; j mulas were later developed to account for the complexities of the
MW MW
1 XMW flow problem. However, Eq. 3 is very appealing in the context of
V0 ¼ V ;
j¼1 r; j                      ð1Þ this study because of the well-defined geometry of the wellbore,
MW and the type of data available.
where MW is the number of stations covered by the window W. It follows from Eq. 3 that TKC is always greater than or equal
The direction of the line is given by the eigenvector that corre- to unity, and attains the minimum value of unity for a perfectly
sponds to the largest eigenvalue of the matrix straight path. In the wellbore situation, the analyzed section will
normally be very close to straight, and initial testing showed that
P¼ TKC from Eq. 3 typically lies in the range from 1.000 to 1.001.
2 XMW XMW XMW 3 Thus, to make it easier to visualize the results, we modified the
ðNr; j  N0 Þ2 ðNr;j  N0 ÞðEr;j  E0 Þ ðNr;j  N0 ÞðVr;j  V0 Þ
6X
6 MW
j¼1 j¼1
XMW
j¼1
XMW 7
7 equation and define our tortuosity parameter T as
6 ðE  ÞðN  Þ ðEr;j  E0 Þ2 ðEr;j  E0 ÞðVr;j  V0 Þ 7
6 r;j E 0 r;j N0 7:
4 Xj¼1 XMW
j¼1 j¼1
XMW 5
MW
ðVr;j  V0 ÞðNr;j  N0 Þ ðV  V0 ÞðEr;j  E0 Þ ðV  V0 Þ 2 T ¼ ðS=LÞ  1
j¼1 j¼1 r;j j¼1 r;j

                   ð2Þ ¼ ðS  LÞ=L; . . . . . . . . . . . . . . . . . . . . . . . . . . ð4Þ

The straight line is used as a reference in the further where S is the along-hole length of the wellbore section being
processing. studied and L is the straight-line distance between the ends of the
The window W is shifted along the wellbore, and the previous section. T is thus the relative elongation of S with respect to L.
processing is performed at each location of the window. From one The situation is illustrated in Fig. 4. A nice feature of the modified
location to the next, W is shifted by the average survey interval. This definition is the favorable correspondence between the verbal
implies that, typically, one survey station is discarded at one end of expressions “no tortuosity” or “absence of tortuosity,” meaning a
the window, whereas a new station is included at the other end. perfectly straight section, and the numeric value T ¼ 0.

2016 SPE Production & Operations 3

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 4 Total Pages: 10

6. Calculate T according to Eq. 4.


2 7. Move the analysis window to another depth and repeat
Steps 2 through 6. This will produce T as a function of MD.
By changing the nominal analysis-window length S0, the tortuos-
1.5
ity parameter can be obtained for various length scales. This, and
High Side Displacement (in.)

the option to calculate T from either raw or smoothed data, allows


1 for the separation of T into large-scale and small-scale tortuosity,
as suggested by Gaynor et al. (2002). This is described further in
the Discussion section.
0.5
Transverse Displacements. The transverse displacement at a
wellbore location is its deviation from a straight reference line at
0 that location. The processing is carried out over a section of the
wellbore at one particular depth. The along-hole length of the sec-
tion and its location are determined by the user.
–0.5 The initial processing step is identical to the straight-line fit-
ting procedure described previously. An analysis window V is
–1 used to select a subset MV of raw data points at the desired depth.
In principle, V is independent of the analysis window W used for
–1.5 –1 –0.5 0 0.5 1 1.5 preprocessing of the raw data; however, V may also be chosen to
Lateral Displacement (in.) be the same processing window. A straight line is fitted through
(a) the MV points by the principal-component-analysis method. This
straight line will be the reference for the subsequent calculation of
transverse displacements. It may further be interpreted as the
“predominant direction” of the wellbore section, and as a “line-
of-sight” when considering the results.
The displacements for all MV raw data points are found by pro-
jecting each of these points perpendicularly onto the reference
6
line. It is a straight-forward calculation to express the projection
vectors in the 2D plane perpendicular to the line. Fig. 5a shows
4 an example of transverse displacements, calculated from real-
survey data.
The transverse displacements, and their distribution along the
High Side (in.)

2
analyzed wellbore section, directly describe the geometry of the
0 section. Thus, features such as the local amount of small-scale tor-
tuosity, or the possible spiraling of the wellbore, can easily be
–2 detected and quantified. Furthermore, the displacements can be
6380
applied to 3D renderings of the wellbore section, giving a direct
–4 6360
visualization of the geometric shape. Examples are shown in Fig.
6340 5b and in the Results section. In all 3D renderings, we have
–6 6320 assumed circular cross section of the wellbore/the casing, and
MD (ft)
4 2 0 6300 diameters corresponding to nominal casing inner diameter.
–2 –4
Lateral (in.) Effective Diameter. The tortuous buckling of a wellbore or
(b) casing section will reduce the effective inner diameter of that sec-
tion, similar to the area of free sight that can be seen through a
Fig. 5—Transverse-displacement vectors calculated at 1-ft curved tunnel. The effective diameter is defined as the maximum
intervals over a 100-ft wellbore section, from real survey data diameter of a straight cylinder that can be inserted into the section
(cased hole). (a) Endpoints of displacement vectors plotted in without distorting it (i.e., just barely touching the inner walls). To
the plane transverse to the wellbore [reference line 5 z-axis,
perpendicular to the paper plane through (0,0)]. The endpoints yield a meaningful definition, the cylinder must have a length equal
indicate a certain amount of wellbore spiraling. (b) The same to or greater than the length of the wellbore or casing section. If the
displacement vectors show the real shape of a wellbore or cas- cylinder represents a physical downhole device, the processing
ing when applied to a 3D model. The color indicates the effec- window V should be chosen equal to the length of the device.
tive diameter value (refer to Fig. 8). As a result of irregular borehole washouts or drillstring or
BHA wear on the formation, the openhole wellbore may not have
Both S and L are calculated in 3D space, and are easily deter- a circular cross section. A deviation from circular cross section
mined from wellbore-position coordinates [the raw data set (Nr,j, may also be found on tubings and casings for various other rea-
Er,j, Vr,j), or the smoothed data set]. The steps necessary to deter- sons. Such effects are neglected here, and it is assumed that the
mine T are as follows: cross sections are perfectly circular.
1. Select S0 as the nominal length scale at which the analysis The reduction in diameter does not depend on the tortuosity of
is wanted. S0 might be the length of one particular down- the wellbore or casing section only. The diameter reduction will,
hole device. in general, also change with the orientation of the inserted cylin-
2. Apply an analysis window of length S0 at a certain well der. Fortunately, the cases that would occur most frequently, and
depth. also would need the most precise analysis, are expected to be
3. Adjust S0 slightly so that the window starts and ends on two nearly straight wellbore sections. In such cases, it is reasonable to
survey stations. let the cylinder’s axis (the line of sight) coincide with the refer-
4. Calculate S as the sum of all intermediate survey interval ence line established for processing of transverse displacements.
lengths between each pair of adjacent survey stations. For This implies that one can use the displacements directly in the cal-
raw data, this sum is equivalent to the original MD interval culation of the diameter reduction.
between the end stations. For smoothed data, each interme- Heavily bent sections can more easily be identified and judged
diate interval must be calculated. as problematic, without extensive analysis. However, the process-
5. Calculate L as the straight-line distance between the start ing of a reference line and transverse displacements may also be
and end stations. applied to these cases. Thus, a single formula for diameter

4 2016 SPE Production & Operations

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 5 Total Pages: 10

60 80 where xi (xj) and yi (yj) are the displacement coordinates in the


transverse plot shown in Fig. 5a, and maxi,j(…) means the maxi-
mum of the square-root function encountered over all possible
55 75
combinations of i ¼ 1…MV, and j ¼ 1…MV. Dnom is the nominal
Azimuth (degrees) inner diameter of the wellbore or casing. If the true diameter is

Inclination (degrees)
50 70 measured, for example, by a caliper, these values might be used
for Dnom; however, a diameter varying with MD or a noncircular
cross section would require a more sophisticated equation for
45 65
Deff. We did not pursue this further in this study; thus, all proc-
essed examples assume a nominal and constant value for Dnom.
40 60 Deff is calculated with the length of the analysis window,
which represents the downhole device, as the only external pa-
rameter. By shifting the analysis window throughout the well and
35 55
calculating the effective diameter at each location, we obtain a
graph of Deff vs. MD, for the selected device length. This graph
30 50 may be plotted directly, or the results may be superimposed on a
2500 2600 2700 2800 2900 3000 3100 3200 3300 3D visualization of the wellbore or casing shape. Examples are
MD (m) given in the following section.
(a)
55
Results
In all field cases, the tortuosity was analyzed from a high-resolu-
tion continuous gyro (HRCG) survey, which was run in cased
50
Azimuth (degrees)

hole (post-drilling). The continuous gyro survey is sampled at


22 Hz and preprocessed to produce a survey point every 0.4 sec-
onds. The typical running speed in casing is 0.8 m/s, yielding
45 0.3 m (1 ft) between measurements. Smaller survey intervals are
possible by increasing the preprocessing output rate, or by
decreasing the running speed. The survey data are quality-
40 checked according to procedures that are based on published error
HRCG models (Ekseth et al. 2011).
MWD
35 Field Case 1. In a deepwater offshore well drilled in Latin America
2500 2600 2700 2800 2900 3000 3100 3200
MD (m) with a modern rotary-steerable system, a conventional MWD survey
showed a straight tangent section with small doglegs (Fig. 6), but
(b)
there were problems associated with running casing all the way to
25 total depth (TD) (3290 m MD).
HRCG
MWD There were indications of issues with the well on the trip out
20 of the hole with the 121/4-in. BHA. Back reaming was required
from TD to 2775 m MD. When running the 95/8-in. casing, the
DLS (degrees/30 m)

string hung up at 3264 m MD, and it could not go any further.


15 Many attempts were made to pull the string free, but it remained
stuck at 3261 m MD. A cleanout run of the 95/8-in. casing was
required and resulted in abandoning several meters of casing.
10 Even though there were several problems in this part of the
well, the low DLS suggested that the completion equipment could
5 be installed here. However, two production pumps failed prema-
turely, and it was decided to run an HRCG survey. Comparison
between the MWD and the HRCG showed no evidence of gross
0 error in the MWD data but indicated rapid azimuth fluctuations
2500 2600 2700 2800 2900 3000 3100 3200
not detected by the conventional MWD survey (Fig. 6b). The
MD (m)
DLS, based on the HRCG data, also indicated higher tortuosity
(c) below 3000 m MD but failed to provide a quantifiable measure
(Fig. 6c). The tortuosity parameters calculated from the HRCG
Fig. 6—Field Case 1. (a) Inclination and azimuth from conven- revealed much-more detail (Figs. 7 and 8) and explained very
tional MWD survey. (b) Comparison of azimuth from the conven-
tional MWD survey and from an HRCG survey. (c) DLS curves
well all the issues with the casing and the production equipment.
calculated from the conventional MWD survey (inclination and The evidence from the HRCG resulted in a decision to place
azimuth), and from the HRCG survey (inclination and azimuth). the production pumps 150 meters higher than initially suggested
to avoid future rework. The initial lack of information between
the conventional MWD survey stations (30-m separation) resulted
in several weeks of lost production and significant extra costs for
reductions can be developed, which is valid for both low- and the operator (estimated to be more than USD 7 million).
high-tortuosity regions. With the assumption of a circular cross
section and a sufficiently high number MV of displacements (the
number is discussed in the Discussion section), a good estimate Field Case 2. A land well at high-latitude north was drilled and
for the effective diameter Deff is obtained from: surveyed with conventional MWD systems. The well profile is
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi indicated in Fig. 9a. On the basis of the DLS plots (Fig. 9b), com-
Deff ¼ Dnom  maxi;j ðxi  xj Þ2 þ ðyi  yj Þ2 ; when pletion equipment was placed near 4,500 ft MD.
After only a few months of operation, the production pump
; failed, and an HRCG survey was requested to evaluate the geome-
Dnom  maxi;j ð…Þ
try and tortuosity of the wellbore. As in the previous example, no
Deff ¼ 0; when Dnom < maxi;j ð…Þ           ð5Þ gross error was identified in the conventional MWD survey, and

2016 SPE Production & Operations 5

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 6 Total Pages: 10

×10–4 the coarse resolution of the conventional MWD data (100 ft) led
2.5 to the initial conclusion that 4,400 to 4,500 ft MD was the region
with lowest tortuosity. The HRCG data (interval ¼ 1 ft) indicated
that the low-tortuosity region actually is found at 4,300 to 4,400 ft
2 MD. This is confirmed by the effective-diameter plot (Fig. 10b),
which further shows that, at 4,500 ft MD, there is no free line of
sight over a length of 100 ft, and therefore this location should be
Tortuosity (T )

1.5
avoided. A new pump was placed near 4,300 ft MD on the basis
of this information. Fig. 11 shows the actual shape of the cased
1 wellbore at the two locations.

0.5
Field Case 3. A land well in the southern USA was surveyed
with conventional MWD methods. The profile is shown in
Fig. 12a. The well is vertical down to 6,300 ft MD and then builds
0 rapidly to horizontal, reaching 90-degrees inclination at approxi-
2500 2600 2700 2800 2900 3000 3100 3200 mately 7,500 ft MD and remaining close to horizontal thereafter.
MD (m) To achieve maximum drawdown, it was required to position an
(a) electrical submersible pump (ESP) in the near horizontal-well sec-
tion at the maximum true vertical depth at which pump placement
was physically possible.
Maximum Outer Diameter of Device (in.)

8 To achieve the required pumping capacity, it was proposed to


7 deploy an ESP of length 124 ft and diameter 4 in. in the well. The
planned tangent section for pump placement was 7,626 to 7,934 ft
6 MD. On the basis of the conventional MWD survey and DLS
(Fig. 12b), the pump could have been placed at any depth between
5
7,750 and 7,950 ft MD.
4 An HRCG survey was requested to confirm the planned pump
location. The tortuosity analysis leads to a very different conclu-
3 sion (Fig. 12c). The effective diameter curve indicates that the
2 depths at which an ESP may be safely deployed in this well are
far more restricted than is evident from the DLS analysis. The
1 best locations at which a pump may be deployed are shown to be
0
at depths of 7,880 ft MD or possibly at 7,815 ft MD (depths refer
2500 2600 2700 2800 2900 3000 3100 3200 to center of pump), whereas depths less than 7,800 ft MD and
MD (m) greater than 7,890 ft MD should be avoided altogether. The Deff
(b) curve indicates that, if positioned at 7,880 ft MD, the ESP may
have a maximum diameter of 3.1 in. to avoid contact with the cas-
Fig. 7—Field Case 1. Tortuosity curve T (a) and effective diame- ing, unless its length is shortened. This information led to a recon-
ter curve Deff (b) calculated from HRCG survey raw data with sideration of the installment plans.
processing window lengths S0 5 30 m and V 5 30 m, respec-
tively. The nominal casing inner diameter is Dnom 5 8.83 in.
Discussion
when the final positioning of the well was evaluated, the differ- Processing and Results. Both the tortuosity parameter T and the
ence in coordinates between the two surveys was only 10 ft. transverse displacements represent geometric features of the well-
A good overall correspondence can be observed between the bore trajectory that can be intuitively understood—relative elon-
DLS curve (Fig. 9b) and the tortuosity plot (Fig. 10a). However, gation of the trajectory compared with the straight line, and

2500 8

2550
7
2600
2650
6
1900 2700
2750 5
Vertical (m)

2000 2800
2850 4
2100 2900
2950 3
2200
3000
700
3050 2
2300 800
3100
1000 900
1100 3150 1
1200 1000
1300 East (m)
1400 1100 0
North (m) 1500

Fig. 8—Field Case 1. Wellbore trajectory with superimposed transverse displacements over the whole HRCG survey interval 2500
to 3200 m MD. The analyzing window length is V 5 30 m. Transverse dimensions (wellbore diameter and displacements) were aug-
mented by a factor of 100 for clarity. The Deff curve (Fig. 7b) was added in terms of color coding, according to the color bar shown
on the right (blue 5 no diameter reduction from nominal 8.83 in.; red 5 diameter reduced to zero).

6 2016 SPE Production & Operations

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 7 Total Pages: 10

75 90
×10–4
5
60 75 4.5
4

Tortuosity (T )
Azimuth (degrees)

Inclination (degrees)
45 60 3.5
3
30 45 2.5
2
15 30 1.5
1
0 15 0.5
0
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000
–15 0 MD (ft)
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000
MD (ft) (a)

Maximum Outer Diameter of Device (in.)


(a) 9
6
8

5 7
DLS (degrees/100 ft)

6
4
5

3 4
3
2
2
1
1
0
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000
0
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 MD (ft)
MD (ft) (b)
(b)
Fig. 10—Field Case 2. Tortuosity curve T (a) and effective-diam-
Fig. 9—Field Case 2. Inclination and azimuth (a) and DLS (b) eter curve Deff (b) on the basis of HRCG data and a device
from conventional MWD survey (interval 100 ft). length of 100 ft. The nominal casing inner diameter is
Dnom 5 8.83 in.

sideways deviations from the straight line. The algorithms apply differences; however, this is outside the scope of this study. In all
data from a single survey and just one input parameter—length of field cases presented here, the HRCG survey was run in cased hole.
the analysis window, which effectively defines the length scale For cases such as Field Case 3, the feasibility of other device
of the tortuosity results. These features make the methodology dimensions can be evaluated easily by changing the length of the
simple, robust, and superior to established methods such as the Deff analysis window. However, this was not performed in the
DLS analysis. present case. Furthermore, in situations where the tortuosity leads
The accuracy of the results is considered to be good. Using all to bending of equipment, it would be possible to evaluate bending
information within the analysis window (from typically 30 sta- moments and final equipment shape by combining our detailed
tions), the parameters are virtually insensitive to small changes in tortuosity description with the geometry and the material (stiff-
the number of stations as the window is moved along the wellbore. ness) parameters of the device.
Such small changes may occur because of the possible nonuni- It should also be emphasized that the methodology proposed
formity of the survey intervals. In particular, the ratio S/L in the T in this paper relates only to the geometric shape of the wellbore,
parameter will be even less sensitive to these changes than S or L and factors such as reservoir pressure, tubing stiffness, or dynamic
alone. Furthermore, the use of all information within the window processes are not considered. Thus, the results should be inter-
implies that the parameters increase or decrease smoothly if the preted as recommendations that are based on wellbore shape only.
window length is gradually changed. In early tests for both T and Factors other than tortuosity may be considered more important,
Deff, the results showed this smooth behavior, indicating reliable and may therefore become the major factors in decisions regard-
algorithm outputs, as long as the number of stations was typically ing equipment installation and operation.
>10. For survey report intervals of 0.3 m, for example, tortuosity
can therefore be analyzed on length scales down to 3 m. The reso- Further Work. The smoothed data set proposed in the Theory
lution interval of the output, on the other hand, is much smaller section has not been examined in detail in this study. It is believed
than the window length. According to the Shannon-Nyquist theo- that it can be used to separate between large-scale and small-scale
rem (Ifeachor and Jervis 1993), the highest possible resolution is tortuosity, along the lines suggested by Gaynor et al. (2002). The
twice the sampling interval (theoretical limit, for band-limited sig- reason for this is that tortuosity details can be processed with a
nals); this amounts to 0.6 m for the previous example. maximum resolution equaling approximately twice the survey
The tortuosity analysis can be applied to data from openhole interval from the raw data set, whereas the smoothing procedure
surveys, or from surveys inside drillstring, casing, or production will efficiently suppress tortuosity on length scales shorter than
tubing. The results should be expected to differ among all these the analysis window W, which typically is 10 to 30 times the sur-
cases because of possible buckling and misalignment of drillstring vey interval.
or tubing inside the hole/casing, and changes to borehole-wall As mentioned in the Theory section, Dnom in Eq. 5 is assumed
roughness and undulations caused by insertion of the casing. One to be the nominal inner diameter of a wellbore or casing with a
application of the methodology may therefore be to analyze such constant and circular cross section. On the basis of the field

2016 SPE Production & Operations 7

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 8 Total Pages: 10

400 90

350 75

Azimuth (degrees)

Inclination (degrees)
15
300 60
10
High Side (in.)

250 45
5
200 30
0
4,550
–5 150 15

4,500
–10 100 0
10 MD (ft) 6,000 6,200 6,400 6,600 6,800 7,000 7,200 7,400 7,600 7,800 8,000
5
0 MD (ft)
–5 4,450
Lateral (in.) –10 (a)
(a)

DLS (degrees/100 ft)


8

6 4

4 3
High Side (in.)

2
2
0
1
–2

–4 7,550 7,600 7,650 7,700 7,750 7,800 7,850 7,900 7,950 8,000
4,350 MD (ft)
–6 4,300 (b)
6 4 2 0 –2 MD (ft)
–4 4,250
Lateral (in.)
Maximum Outer Diameter of Device (in.)

4
(b)
3.5
Fig. 11—Field Case 2. Wellbore shape computed from trans- 3
verse displacements over 100-ft sections at two depths: (a)
4,500 ft MD (high-tortuosity region with Deff 5 0 in.) and (b) 2.5
4,300 ft MD (low-tortuosity region with Deff  6.7 in.). The nomi-
nal casing inner diameter is Dnom 5 8.83 in., and the colors indi- 2
cate the calculated Deff values (refer to Fig. 8). 1.5

examples, we believe that this gives sufficient accuracy for the 1


present application. However, more accurate estimates for Deff 0.5
would be obtained when using real Dnom values, for example, from
0
caliper measurements. This also may require a more sophisticated
7,500 7,550 7,600 7,650 7,700 7,750 7,800 7,850 7,900 7,950 8,000
formula for Deff. Furthermore, the Deff analysis can be extended for MD (ft)
devices with noncircular or nonuniform (along the device) cross (c)
sections, and eventually allowing for some bending of the device,
within the manufacturer’s or operator’s specifications. Fig. 12—Field Case 3. (a) Inclination and azimuth from conven-
The T and Deff parameters may respond differently to various tional MWD survey. (b) DLS calculated from the conventional
tortuosity geometries, as illustrated in Table 1. Therefore, by com- MWD survey. (c) Effective diameter Deff curve on the basis of
bining the information from the two parameters, we may obtain HRCG data and a device length of 124 ft. The nominal casing
more detailed information on the nature of the tortuosity. For inner diameter is Dnom 5 6.28 in.

Wellbore Shape (local) Relative Elongation, T = (S–L)/L Span of Displacements Effective Diameter, Deff
0 0 Dnom (=max)

Small Small Near Dnom

Large Large Small

Large Small Near Dnom

Table 1—Response of tortuosity parameters T and Deff to various local shapes of the wellbore. The wellbore trajectories may be 3D.

8 2016 SPE Production & Operations

ID: jaganm Time: 14:55 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 9 Total Pages: 10

example, if T is large, as shown in the lower two lines of the table, W ¼ analysis window for establishing reference
the value of Deff may help to determine whether the elongation is line
caused by a single “long” curve (still within the analysis window), xi, yi, xj, yj ¼ coordinates of displacement vectors i and j in
or caused by several microbends. A “shape parameter” derived plane transverse to reference line, in.
from T and Deff, or from T and the span of transverse displacements,
could be presented as a graph vs. MD, which may be more conven-
ient than 3D renderings over large wellbore sections. Acknowledgments
The authors want to thank Gyrodata and SandRidge Energy for
permission to publish this work, and for providing field data. Fur-
Conclusions thermore, Rob Shoup, Gyrodata, is thanked for stimulating
This paper has presented a novel methodology to analyze well- discussions.
bore tortuosity, in particular, tortuosity on a length scale shorter
than conventional survey report intervals (30 m). The input data
are provided by a single high-resolution directional survey, which References
may use any surveying tool or procedure capable of delivering API D20. 1985. Bull. on Directional Drilling Survey Calculation Methods
such high-resolution data. By applying a user-defined data win- and Terminology. Washington, DC: API.
dow as a single external parameter, the method quantifies tortuos- Bordakov, G. A., Kostin, A. V., Rasmus, J. et al. 2009. Improving LWD
ity on any length scale greater than approximately 10 times the Image and Formation Evaluation by Using Dynamically Corrected
survey-data interval, and with a maximum resolution that equals Drilling-Derived LWD Depth and Continuous Inclination and Azi-
twice the data interval, in terms of muth Measurements. SPE Res Eval & Eng 12 (1): 137–148. SPE-
• Tortuosity parameter; the relative elongation of along-hole 109972-PA. http://dx.doi.org/10.2118/109972-PA.
length with respect to straight-line length. Brands, S. and Lowdon, R. 2012. Scaled Tortuosity Index: Quantification
• Transverse displacements of the wellbore trajectory from a of Borehole Undulations in Terms of Hole Curvature, Clearance, and
locally established reference line. Pipe Stiffness. Presented at the IADC/SPE Drilling Conference and
• Effective inner diameter of wellbore or casing as a result of Exhibition, San Diego, California, USA, 6–8 March. SPE-151274-MS.
small-scale bending. http://dx.doi.org/10.2118/151274-MS.
The outputs can be presented individually or in combination, Chen, D. C.-K., Gaynor, T., and Comeaux, B. 2002. Hole Quality: Why It
either as graphs vs. MD or in 3D renderings showing the actual Matters. Presented at the SPE International Petroleum Conference and
geometry of the wellbore. The results are easily visualized, inter- Exhibition, Villahermosa, Mexico, 10–12 February. SPE-74403-MS.
preted, and communicated to operating personnel, and represent a http://dx.doi.org/10.2118/74403-MS.
significant improvement from traditional tortuosity analysis meth- Davis, J. C. 1986. Statistics and Data Analysis in Geology, second edition,
ods on the basis of DLS. 527–531. New York, NY, USA: Wiley & Sons.
The methodology was applied to various field cases in which Ekseth, R., Weston, J., Ledroz, A. et al. 2011. Improving the Quality of
the results demonstrate the identification of highly tortuous Ellipse of Uncertainty Calculations in Gyro Surveys to Reduce the
regions that were unnoticed or poorly quantified by traditional Risk of Hazardous Events Like Blowouts or Missing Potential Pro-
methods. This information has direct relevance to decisions duction Through Incorrect Wellbore Placement. Presented at the
regarding the installation of equipment in the completion and pro- SPE/IADC Drilling Conference and Exhibition, Amsterdam, 1–3
duction phases. March. SPE-140192-MS. http://dx.doi.org/10.2118/140192-MS.
The tortuosity analysis may thus contribute to more efficient Gaynor, T. M., Chen, D. C.-K., Stuart, D. et al. 2001. Tortuosity versus
operations and improvements to equipment functionality, eventu- Micro-Tortuosity—Why Little Things Mean a Lot. Presented at the
ally leading to reduced workover frequency and considerable cost SPE/IADC Drilling Conference, Amsterdam, 27 February–1 March.
savings. The possible applications of the methodology range from SPE-67818-MS. http://dx.doi.org/10.2118/67818-MS.
evaluation of the drilling process to the installation of completion- Gaynor, T., Hamer, D., Chen, D. C.-K. et al. 2002. Quantifying Tortuosi-
and-production equipment. ties by Friction Factors in Torque and Drag Model. Presented at the
SPE Annual Technical Conference and Exhibition, San Antonio,
Texas, USA, 29 September–2 October. SPE-77617-MS. http://
Nomenclature dx.doi.org/10.2118/77617-MS.
dNr,j, dEr,j, dVr,j ¼ raw data position increments in N, E, V at sta- Ifeachor, E. C. and Jervis, B. W. 1993. Digital Signal Processing. A Prac-
tion j, ft or m tical Approach, reprinted 1999, Chap. 2. Harlow, England: Addison-
Dnom ¼ nominal inner diameter of casing or wellbore, Wesley.
in. Liu, X., Liu, R., and Sun, M. 2004. New Techniques Improve Well Plan-
Deff ¼ effective diameter; maximum-allowed outer ning and Survey Calculation for Rotary-Steerable Drilling. Presented
diameter of downhole device, in. at the IADC/SPE Asia Pacific Drilling Technology Conference and
L ¼ straight-line distance (between start and end Exhibition, Kuala Lumpur, 13–15 September. SPE-87976-MS. http://
stations of a section), ft or m dx.doi.org/10.2118/87976-MS.
MV ¼ number of survey stations covered by V window Marck, J., Detournay, E., Kuesters, A. et al. 2014. Analysis of Spiraled
MW ¼ number of survey stations covered by W Borehole Data Using a Novel Directional Drilling Model. Presented at
window the IADC/SPE Drilling Conference and Exhibition, Fort Worth, Texas,
N, E, V ¼ position coordinates: Northing, Easting, Verti- USA, 4–6 March. SPE-167992-MS. http://dx.doi.org/10.2118/167992-
cal depth, ft or m MS.
N0, E0, V0 ¼ average (arithmetic mean) of N, E, V over a Marck, J. and Detournay, E. 2015. Spiraled Boreholes: An Expression of
subset of survey stations, ft or m 3D Directional Instability of Drilling Systems. Presented at the SPE/
Nr,j, Er,j, Vr,j ¼ raw data position coordinates N, E, V of sta- IADC Drilling Conference and Exhibition, London, 17–19 March.
tion j, ft or m SPE-173156-MS. http://dx.doi.org/10.2118/173156-MS.
P ¼ matrix used in principal-component analysis Mason, C. J. and Chen, D. C.-K. 2006. The Wellbore Quality Scorecard
S ¼ along-hole distance; actual value of S0, ft or m (WQS). Presented at the IADC/SPE Drilling Conference, Miami, Flor-
S0 ¼ analysis window for tortuosity parameter; ida, USA, 21–23 February. SPE-98893-MS. http://dx.doi.org/10.2118/
length of this window, ft or m 98893-MS.
T ¼ tortuosity parameter (Eq. 4) Matthews, C. M. and Dunn, L. J. 1993. Drilling and Production Practices
TKC ¼ tortuosity parameter in petrophysics (Eq. 3) to Mitigate Sucker-Rod/Tubing-Wear-Related Failures in Directional
V ¼ analysis window for effective-diameter Wells. SPE Prod & Fac 8 (4): 251–259. SPE-22852-PA. http://
analysis dx.doi.org/10.2118/22852-PA.

2016 SPE Production & Operations 9

ID: jaganm Time: 14:56 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008


PO173103 DOI: 10.2118/173103-PA Date: 15-March-16 Stage: Page: 10 Total Pages: 10

Matyka, M. and Koza, Z. 2012. How to Calculate Tortuosity Easily? Sugiura, J. and Jones, S. 2008. The Use of the Industry’s First 3-D Mechan-
American Institute of Physics (AIP) Conference Proceedings 1453, ical Caliper Image While Drilling Leads to Optimized Rotary-Steerable
Potsdam, Germany, 17–22 June. http://dx.doi.org/10.1063/1.4711147. Assemblies in Push- and Point-the-Bit Configurations. Presented at the
[or arXiv:1203.5646v1 [physics.flu-dyn] 26 Mar 2012]. SPE Annual Technical Conference and Exhibition, Denver, 21–24 Sep-
Menand, S., Sellami, H., Tijani, M. et al. 2006. Buckling of Tubulars in tember. SPE-115395-MS. http://dx.doi.org/10.2118/115395-MS.
Actual Field Conditions. Presented at the SPE Annual Technical Con- Sugiura, J., Bowler, A., Hawkins, R. et al. 2013. Downhole Steering Auto-
ference and Exhibition, San Antonio, Texas, USA, 24–27 September. mation and New Survey Measurement Method Significantly Improve
SPE-102850-MS. http://dx.doi.org/10.2118/102850-MS. High-Dogleg Rotary Steerable System Performance. Presented at the
Menand, S., Sellami, H., Bouguecha, A. et al. 2009. Axial Force Transfer SPE Annual Technical Conference and Exhibition, New Orleans, 30
of Buckled Drill Pipe in Deviated Wells. Presented at the SPE/IADC September–2 October. SPE-166165-MS. http://dx.doi.org/10.2118/
Drilling Conference and Exhibition, Amsterdam, 17–19 March. SPE- 166165-MS.
119861-MS. http://dx.doi.org/10.2118/119861-MS. Weltzin, T., Aas, B., Andreassen, E. et al. 2009. Measuring Drillpipe
Mitchell, R. F. and Weltzin, T. 2011. Lateral Buckling—The Key to Buckling Using Continuous Gyro Challenges Existing Theories. SPE
Lockup. SPE Drill & Compl 26 (3): 436–452. SPE-139824-PA. http:// Drill & Compl 24 (4): 464–472. SPE-115930-PA. http://dx.doi.org/
dx.doi.org/10.2118/139824-PA. 10.2118/115930-PA.
Mitchell, R. F., Bjørset, A., and Grindhaug, G. 2015. Drillstring Analysis
With a Discrete Torque/Drag Model. SPE Drill & Compl 30 (1): 5–16. Jon Bang has been a development engineer at Gyrodata
SPE-163477-PA. http://dx.doi.org/10.2118/163477-PA. since 2013. Previously, he was with SINTEF Petroleum Research,
Oag, A. W. and Williams, M. 2000. The Directional Difficulty Index—A Norway, working on wellbore-position uncertainty analysis.
New Approach to Performance Benchmarking. Presented at the IADC/ Bang’s current research interests include error modeling, qual-
SPE Drilling Conference, New Orleans, 23–25 February. SPE-59196- ity control, and data processing related to wellbore surveying.
MS. http://dx.doi.org/10.2118/59196-MS. He holds an MSc degree in electronics and a PhD degree in
laser physics, both from the Norwegian University of Science
Reid, L. C., McDonald, C., Baker, T. et al. 2013. Pump-Friendly Wellbore
and Technology. Bang is a member of SPE, and an active
Design and Case Study From Mississippian Play of Oklahoma/Kansas. member of the Industry Steering Committee on Wellbore Sur-
Presented at the SPE Unconventional Resources Conference, Calgary, vey Accuracy (ISCWSA).
5–7 November. SPE-167156-MS. http://dx.doi.org/10.2118/167156-
MS. Onyemelem Jegbefume is a research and development engi-
neer at Gyrodata. Previously, he worked for 3 years as a systems
Samuel, G. R., Bharucha, K., and Luo, Y. 2005. Tortuosity Factors for
engineer for Rockwell Collins. Jegbefume’s current research
Highly Tortuous Wells: A Practical Approach. Presented at the SPE/ interests include algorithms for wellbore surveying, communi-
IADC Drilling Conference, Amsterdam, 23–25 February. SPE-92565- cation systems, and signal processing for oilfield applications.
MS. http://dx.doi.org/10.2118/92565-MS. He holds a PhD degree in electrical engineering from the Uni-
Samuel, R. and Liu, X. 2009. Wellbore Tortuosity, Torsion, Drilling Indi- versity of Texas at Dallas. Jegbefume is a member of SPE.
ces, and Energy: What Do They Have to Do With Well Path Design?
Adrián Ledroz is the Vice President, Survey Technologies, Tech-
Presented at the SPE Annual Technical Conference and Exhibition, nical Services for Gyrodata. He has worked with gyroscopes
New Orleans, 4–7 October. SPE-124710-MS. http://dx.doi.org/ for the last 15 years, and his research interests include inertial
10.2118/124710-MS. sensors, data processing, quality control, and error modeling
Samuel, R. and Gao, D. 2013. Horizontal Drilling Engineering–Theory, related to wellbore surveying. Ledroz holds a BSc degree in
Methods and Application, first edition, Chap. 9. Houston: biomedical engineering from Universidad Nacional de Entre
SigmaQuadrant. Rı́os, Argentina, and an MSc degree in electrical engineering
Sawaryn, S. J. and Thorogood, J. L. 2005. A Compendium of Directional from the University of Calgary, Canada. He is a member of
Calculations Based on the Minimum Curvature Method. SPE Drill & SPE, and an active member of ISCWSA.
Compl 20 (1): 24–36. SPE-84246-PA. http://dx.doi.org/10.2118/ John Weston is the Global Adviser Gyro Technology for Gyro-
84246-PA. data, and his research interests are in inertial systems and their
Skillingstad, T. 2000. At-Bit Inclination Measurements Improves Direc- application to wellbore surveying. He holds a BSc degree in
tional Drilling Efficiency and Control. Presented at the IADC/SPE electrical engineering from the University of Wales and an
Drilling Conference, New Orleans, 23–25 February. SPE-59194-MS. MSc degree in information and systems engineering from the
University of Birmingham, UK. Weston is coauthor of the text-
http://dx.doi.org/10.2118/59194-MS.
book Strapdown Inertial Navigation Technology. He is a mem-
Stockhausen, E. J. and Lesso, W. G. 2003. Continuous Direction and Incli- ber of the Institution of Engineering Technology and SPE, and
nation Measurements Lead to an Improvement in Wellbore Position- is an active member of the ISCWSA.
ing. Presented at the SPE/IADC Drilling Conference, Amsterdam,
19–21 February. SPE-79917-MS. http://dx.doi.org/10.2118/79917- John H. (Jay) Thompson is the Artificial Lift Specialist with San-
dRidge Energy in Oklahoma City, Oklahoma. He has been
MS.
with SandRidge for 3 years but has been in the pumping indus-
Stuart, D., Hamer, C. D., Henderson, C. et al. 2003. New Drilling Technol- try, both surface pumping and downhole, for more than 26
ogy Reduces Torque and Drag by Drilling a Smooth and Straight years. Thompson’s interests include the effects of wellbore
Wellbore. Presented at the SPE/IADC Drilling Conference, Amster- design and how it affects ESP pumps, root-cause failure analy-
dam, 19–21 February. SPE-79919-MS. http://dx.doi.org/10.2118/ sis, and how the longevity of downhole equipment can be
79919-MS. improved. He is a member of SPE.

10 2016 SPE Production & Operations

ID: jaganm Time: 14:56 I Path: S:/PO##/Vol00000/160008/Comp/APPFile/SA-PO##160008

You might also like