DC199640 DOI: 10.

2118/199640-PA Date: 15-July-20 Stage: Page: 1 Total Pages: 10

Directional Drilling Automation

Using a Laboratory-Scale Drilling Rig:
SPE University Competition
Emmanuel Akita, Forrest Dyer, Savanna Drummond*, Monica Elkins*, Payton Duggan,
and Ramadan Ahmed, University of Oklahoma, and Fred Florence, Rig Operations, LLC

Summary
The use of drilling automation is accelerating, mostly in the area of rate of penetration (ROP) enhancement. Autonomous directional
drilling is now a high focus area for automating drilling operations. The potential impact is immense because 93% of the active rigs in
the US are drilling directional or horizontal wells. The 2018–2019 Drilling Systems Automation Technical Section (DSATS)-led inter-
national DrillboticsVR Student Competition includes automated directional drilling. In this paper, we discuss the detailed design of the

winning team. We present the surface equipment, downhole tools, data and control systems, and lessons learned.
SPE DSATS organizes the annual Drillbotics competition for university teams to design and develop laboratory-scale drilling rigs.
The competition requires each team to create unique downhole sensors to allow automated navigation to drill a directional hole. Student
teams have developed new rig configurations to enable several steering methods that include a rotary steering system and small-scale
downhole motors with a bent-sub. The most significant challenge was creating a functional downhole motor to fit within a 1.25-in.
(3.18 cm) diameter wellbore. Besides technical issues, teams must demonstrate what they have learned about bit-rock interaction and
the physics of steering. In addition, they must deal with budgets and funding, procurement and delivery delays, and overall project man-
agement. This required an integrated multidisciplinary approach and a major redesign of the rig components.
The University of Oklahoma (OU) team made significant changes to its existing rig to drill directional holes. The design change was
introduced to optimize the performance of the bottomhole assembly (BHA) and allow directional drilling. The criteria for selecting the
BHA was hole size, BHA dynamics, a favorable condition for downhole sensors, precise control of drilling parameters, rig mobility,
safety, time constraints, and economic practicality. The result is an autonomous drilling rig that drills a deviated hole toward a defined
target through a 2  2  1-ft (60.96  60.96  30.48 cm) sandstone block (i.e., rock sample) without human intervention. The rig cur-
rently uses a combination of discrete and dynamic modeling from experimentally determined control parameters and closed-loop feed-
back for well-trajectory control.
The novelty of our winning design is in the use of a small-scale cable-driven downhole motor with a bent-sub and quick-connect-
type swivel system. This is intended to replicate the action of a mud motor within the limits of the borehole diameter. In this paper,
we present details of the rig components, their specifications, and the problems faced during the design, development, and testing. We
demonstrate how a laboratory-scale rig can be used to study drilling dysfunctions and challenges. Building a downhole tool to
withstand vibrations, water intrusion, magnetic interference, and electromagnetic noise are common difficulties faced by major
equipment manufacturers.

Introduction and Overview

An SPE DSATS subcommittee established guidelines for the annual international competition, which provides students from different
areas of study the rare opportunity of working together to design and build small-scale drilling rigs that drill autonomously through a
rock sample. This is achieved using sensors and control algorithms. Since the inception of this competition in 2015, interesting designs
have been proposed, and multiple-style mini rigs have been built to address each year’s challenge, as seen in the Drillbotics archives
(Drillbotics 2019).
According to the SPE DSATS guidelines statement, the official 2018–2019 challenge was to: “Design a rig and related equipment to
autonomously drill a well, using downhole sensors, that obtains as much horizontal displacement from the surface as possible along
the rock’s ‘north’ direction, as quickly as possible while maintaining borehole quality and integrity of the drilling rig and drillstring”
(Drillbotics 2018–19 Guidelines). Thus, all decisions made throughout the duration of this project were aimed at satisfying the
said challenge.
A 1.25-in. (3.18 cm) microdrill bit was provided halfway through the project year, and a rock sample with unconfined compressive
strength ranging from 2,200 to 6,200 psi (15.17 to 42.74 MPa) was provided on the day of the demonstration. Students had the option
of using this drill bit or developing their own for the competition. A round aluminum tube with dimensions 0.375-in. (0.95 cm) outer
diameter  0.277-in. (0.70 cm) inner diameter  36-in. (91.44 cm) length was chosen for the competition.
Eleven teams competed, and only two teams managed to successfully drill directional wells. The winning team was selected by the
DSATS subcommittee, as per the scoring metric in the guidelines. Safety, functionality, rig performance, ROP, and wellbore integrity
were among some of the metrics used in scoring. Some of the limitations put on teams were the total power consumption was limited to
25 hp (18.6 kW), and maximum expenditure should not exceed USD 10,000.

The OU Mini Rig

Several past OU teams spent a considerable amount of resources perfecting the mini rig since the inception of the Drillbotics competi-
tion in 2015. The rig (Fig. 1) was used in the 2019 competition because it is still mechanically robust.

* Now with Schlumberger Limited

Copyright V
C 2020 Society of Petroleum Engineers

This paper (SPE 199640) was accepted for presentation at the IADC/SPE International Drilling Conference and Exhibition, Galveston, Texas, USA, 3–5 March 2020, and revised
for publication. Original manuscript received for review 23 March 2020. Revised manuscript received for review 9 May 2020. Paper peer approved 14 May 2020.

2020 SPE Drilling & Completion 1

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 2 Total Pages: 10

Fig. 1—OU mini rig.

The rig structure is designed to support the linear motion of the traveling block. The derrick has hinges, which allow folding/
reclining on its back for transport. The foldability of the rig, coupled with the wheels at the base of each leg, allows it to fit through
small doorway spaces and to be mobile. All high-voltage components of the rig are in separate compartments for safety purposes.
When fully erected on the rig table, the rig stands at 109 in. (276.86 cm) in height, 82 in. (208.28 cm) in width, and 28 in. (71.12 cm) in
breadth. A 32-in. (81.28 cm) gap below the rig table allows rock samples to be placed for drilling. More details on the rig components
and design are provided in the design report submitted to DSATS in Phase I of the competition (Akita et al. 2018).

Directional Drilling Design

Well Path Design. The tried and tested technique of using bent motors while sliding the pipe have often been used for directional drill-
ing. Furthermore, current rotary steerable systems have multiple advantages in directional drilling, which include, but are not limited
to, effective hole cleaning and lower tortuosity of drilled wells (Weijermans et al. 2001). Thus, a logical product development phase
would have been to model and build a rotary steerable system on a laboratory scale for this project. However, based on the complex
nature of rotary steerable systems, time scale, and funding limitations for designing such a system, the team decided to build a bent
motor on a small scale.
The case of drilling path optimization is based on the expected bending radius of the pipe and the amount of total weight on bit
(WOB) required to hit the target located at the bottom edge of a rock sample with dimensions of 12  24  24 in. The drilling path com-
prises a maximum horizontal departure of 6 in. (15.24 cm) along the 12-in. (30.48 cm) side of the rock sample (Fig. 2). Using the Drill-
botics guidelines, considering a minimum 4-in. (10.16 cm) kick-off point (KOP), an inverse tangent function of the ratio 6 to 20 is
determined (Drillbotics 2018–19 Guidelines). Thus, the drilling plan was a path that subtended an angle of 16.7 and traversed a length
of 21 in. (53.34 cm). The well trajectory was modeled as an L-shaped (build and hold) well (Fig. 2a) to determine the maximum angle
needed to hit the target. In reality, however, it was a J-shaped directional well (Fig. 2b), with a curved path, building at a relatively con-
stant rate of 0.84 /in. (0.33 /cm) from the KOP. This curve arises due to the mechanics of the drillstring and bit-rock interaction,
coupled with the intricacies of the control algorithm.

BHA Design Options. This year, the team designed a new assembly for directional drilling. The engineering of the entire project was
aimed at designing a system with minimal energy losses, while simultaneously optimizing the well trajectory and maximizing the rig
and personnel safety. Rigorous testing was conducted to establish the limits of the mechanical design and to help engineer out safety
hazards. For instance, the rotating section between the base of the topdrive and the top of the metal plate of the top assembly (Fig. 3)
was shielded with fully transparent polycarbonate sheeting. Furthermore, drilling fluid (water) was effectively circulated out of the per-
sonnel path to avoid slippage. Also, control algorithms were written to only begin system rotation when the bit was below the rock sur-
face of a reference depth, and to stop rotation when it was above the reference depth. This prevented parts from flying off in case
something unexpectedly broke. Moreover, the team went through several iterations of the prototype. This process resulted in a robust
design that allowed autonomously hitting the target in record time.

2 2020 SPE Drilling & Completion

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 3 Total Pages: 10

12 in.

KOP KOP
24 in.

Target Target

(a) (b)

Fig. 2—Rock sample with (a) idealized and (b) J-shaped well trajectories.

Metal plate

Aluminum block

Cam lock system

Water hose
bypassing
the swivel

Rotary union

Gear/chain

Fig. 3—Top assembly.

To start this process, many options were considered. The initial concept was to build a laboratory-scale turbine or positive displace-
ment motor to allow for directional drilling. Data extrapolated from the resulting calculations indicated that a downscaled turbine motor
would require approximately 25 gal/min (1.58 L/s) to begin rotating. Because the maximum capacity of our pump is just over the mini-
mum turbine capacity, it did not leave much room for drilling-parameter optimization for efficiency. Also, the minimum length of a
downscaled turbine came out to be approximately 13 ft (396.24 cm), which is not practical for this small-scale application.
Furthermore, the extrapolated data for a positive displacement motor indicate a large 7.5-ft (228.6 cm) length, which is also not prac-
tical for laboratory scale. The highest power value was 0.12 hp (89.48 W), resulting from 800 rev/min and a 1:2 lobe ratio for our posi-
tive displacement motor calculations. This is much smaller than the expected power generated at our bit from the 1 hp (745.70 W)
electric motor used as a topdrive. Thus, the “cable-drilling” method was considered. More details on the downscaled turbine and posi-
tive displacement motor calculations are provided in the design report submitted to DSATS in Phase I of the competition (Akita et al.
2018). The cable-drilling technique efficiently achieves torque transmission to the drill bit, using a cable rotating within the drillpipe.
The drillpipe slides during the drilling process, and this allows for constant azimuth and easy upward and downward steering of the
drill bit because the BHA does not rotate.

2020 SPE Drilling & Completion 3

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 4 Total Pages: 10

Drillpipe Stress Analysis. To perform directional drilling, the theoretical stresses on the drillpipe had to be determined to get an idea
of the maximum design limits of our proposed system. A finite element method (FEM) analysis was performed using the Static Struc-
tural Analysis module in ANSYSÔ Workbench (Fig. 4). The FEM analysis (model) was simplified to analyze only the drillpipe stress
distributions, independent of bit-rock interactions, BHA angle considerations, and rock material strength. The KOP is set at a 4-in.
(10.16 cm) depth. The model ignores the changes in stresses from the surface WOB (i.e., WOB measured at the surface) because it con-
siders the resultant horizontal stress in isolation when the drillpipe hits the target.

Static structural
Equivalent stress
Type: equivalent (von Mises) stress
Unit: psi
Time: 1
27 December 2018 3:36 p.m.

1.1745×105 max
1.044×105
91,349
78,299
65,249
52,200
39,150
26,100
13,051
1.1457 min
x

0 10.000 20.000 (in.) 0 3.000 6.000 (in.)

y
5.000 15.000 z 1.500 4.500

(a) (b)

Fig. 4—von Mises stress analysis on a drillpipe: (a) full and (b) upper section (Akita et al. 2018).

The FEM model converges at a von Mises stress value of 117 ksi (806.69 MPa), for the approximate 16.7 bend. The von Mises
stress provides the yield failure criterion for the drillpipe. It is typically used in mechanical design because it allows for the representa-
tion of the 3D stress state as a single positive stress value. Also, it is used for predicting yielding in ductile material such as aluminum
(Fig. 5). Because aluminum 6061 has a yield strength of 35 ksi (241.32 MPa), it is abundantly evident that this is exceeded. Hence, the
pipe undergoes plastic deformation without failure (Fig. 5a). Aluminum 6061 exhibits a very high plastic strain before failure, indicat-
ing its high ductility. The FEM analysis does not account for plastic deformation; as a result, it overpredicts the stress when plastic
deformation occurs (Fig. 5b). Aluminum prevents a linear stress growth beyond its yield strength because of its high malleability. Thus,
for the loads applied with a limited angle of bend that correspond to the maximum strain (emax), the pipe just permanently deformed
and prevented the linear stress growth and subsequent pipe failure. For a complete pipe failure, the maximum strain must exceed
0.12 (Fig. 5a).

500 1200
Stress/strain curve
Plastic strain before failure FEM analysis
375
800
Stress (MPa)
Stress (MPa)

250

400
125
εmax
Stress/strain curve
0 0
0 0.050 0.100 0.150 0 0.005 0.010 0.015 0.020
Strain Strain
(a) (b)

Fig. 5—Stress/strain curve of aluminum 6061: (a) full curve (Ambriz et al. 2010), and (b) partial curve and FEM analysis prediction.

Due to the high ductility of aluminum, the team theorized that the failure could be due to erosional wear from the rotating cable,
which scratches the pipe interior. Several ideas to reduce erosional wear were considered, such as coating the metal rod with plastic to
take some of the wear or machining the square rod edges to rounded geometries to reduce wear. However, the results of tests conducted
at high rotational speed (1,150 rev/min) for more than 5 hours of drilling showed no pipe failure. Hence, the design proved to be robust
and adequate for the intended use, which is less than 15 minutes of drill time to total depth. More importantly, a 4-in. (10.16 cm) KOP
requires a force with a 42-lbf (186.83 N) horizontal component to generate 6 in. (15.24 cm) of horizontal displacement from the central
axis. This horizontal force is a function of surface WOB. Thus, well path optimization becomes a two-part problem: BHA bent-sub
angle to guide the drill bit to target and control algorithms to efficiently regulate surface WOB for desired results.

New BHA Concept

For cable drilling, efficient torque transmission downhole is paramount to ensure maximum possible power at the drill bit. This was
achieved by means of a metal rod. Stainless steel was the material of choice for the rod because it can withstand the maximum torsional
stresses and does not easily corrode. A square 1/8-in. cable was the geometry of choice because it eases assembling, allows sufficient

4 2020 SPE Drilling & Completion

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 5 Total Pages: 10

drilling fluid flow, and has enough torque transmission capacity for the intended purpose. Torque capacity tests conducted on the square
cable indicated a 90 lbf-in. (10.17 Nm) torsional limit. This translates to 207 ksi (1427 MPa) torsional stress. Torque capacity of the
rod is 80% higher than a torque limit of 50 lbf-in. (5.65 Nm) set in previous years, which corresponds to a 115 ksi (792.90 MPa) tor-
sional stress level. Thus, the square rod was adequate for the design and worked well up to a rotational speed of 1,150 rev/min. More
details on geometry optimization are presented elsewhere (Akita et al. 2018).

Top Assembly Modification. The top assembly comprises a top spider, Camlock system, drillstring, and a 1/ 8-in. cable. The major rig
modifications made this year started from the metal plate shown in Fig. 3. The cable rod is housed in the interior of the drillpipe and
inserts into a groove on the top spider seen in Fig. 6a. The top spider allows the flow of drilling fluid into the drillpipe and the transmis-
sion of torque from the top driveshaft to the metal rod, and subsequently to the drill bit. The 2-in.-thick aluminum block is bolted onto
the metal plate and allows for ease of future design changes. The Camlock allows for a quick disconnect of drillpipe connections while
holding a fixed position for sliding during the drilling process. The team bypassed the swivel this year and connected the water hose
directly to the base of the Camlock. This was a practical change because it reduced leakages from the current swivel. The rotary union
acts as a connection piece between the Camlock and the drillpipe. It allows the change of azimuth independent of the Camlock connec-
tion. The chain/sprocket (gear/chain) system couples the rotary union with a stepper motor. The drillpipe threads into the base of the
rotary union, and the chain/sprocket system allows azimuth control while drilling using the stepper motor.

(a) (b)

Fig. 6—Top spider: (a) initial and (b) final versions.

The top spider (Fig. 6b) is the final version used for the competition because the first version was difficult to put in place and reweld
broken components. This spider is threaded onto a shaft from the topdrive and is located at the center of the Camlock. It transmits the
torque downhole through the square cable and allows fluid flow from the water hose to the drillpipe.

The BHA. A steel rod with dimensions 1-in. (2.54 cm) outer diameter and 0.75-in. (1.91 cm) inner diameter serves as the stabilizer
housing for the BHA. Two metal pieces are welded together at a 12 angle for directional drilling, creating a bent-sub. Fig. 7 shows the
bent-sub built for this competition. The internal components on the left (Fig. 7a) comprises metal rods, threaded holes for set screws, a
metal bearing, a race, and the 1.25-in. (3.18 cm) Baker Hughes microbit. The microbit is 1.25-in. (3.18 cm) long, and it has 25 cone
cutters (one to two) and 20 shoulder cutters (PDC Microbit 2019). The BHA is approximately 6-in. (15.24 cm) long and slides during
the entire drilling process. The visible holes allow for water flow around the rotating cable.
The first trial value of the bent-sub angle was 7 . This was determined using the trigonometric relationship presented in Eq. 1. The
focus of the team was to keep the length of the bent arm of the BHA at less than 4.27 in. (10.85 cm) as shown by the red dashed line in
Fig. 8. The 4.27 in. (10.85 cm) is the maximum Euclidean distance for a 4-in. (10.16 cm) pilot hole with a 1.5-in. (3.81 cm) diameter.
The straight portion of the BHA was designed to be approximately 50% of 4.27 in. (10.85 cm). The team focused on the bent arm and
used 50% as an educated guess for the straight portion as the design. With a 10% safety factor, an “x” angle of 7 provided a hypotenuse
length of 3.69 in. (9.37 cm), which was the closest to the maximum of 4.27 in. (10.85 cm) required. That notwithstanding, multiple itera-
tions of the BHA with different angles were later built because further laboratory testing suggested a more aggressive value. Therefore,
the team decided to go with 12 because this provided the best directional results. More details on the angle optimization for the bent
motor and engineering drawings of the BHA prototype for the bent-sub are presented in the design report (Akita et al. 2018).
ða þ bÞ  0:9 0:5  0:9
H¼ ¼ ¼ 3:69 in:; H < 4:27 in:; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð1Þ
sinðxÞ sinð7Þ
where H is the hypotenuse of the right angle subtended by the BHA arm, a þ b is the total available lateral gap in a 1.5-in. (3.81 cm)
pilot hole after the BHA arm is seated, and x is the BHA angle from the vertical axis.

2020 SPE Drilling & Completion 5

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 6 Total Pages: 10

Holes

(a) (b)

Fig. 7—Bent-sub and bit assembly: (a) interior and (b) exterior.

a b

Fig. 8—Schematic for BHA angle determination.

Pipe Stiffness. The stiffness of the drillpipe used is critical for success. The setpoint for WOB in our controls was tied to how easy or
difficult it was to bend the pipe. Hence, the team stuck with the same pipe throughout the tests. The more tests that were run, the more
flexible the drillpipe became. The controls were written to factor in small changes in pipe stiffness resulting from multiple test runs.
Stiff pipes could benefit from heat treatment or multiple loading cycles to reduce stiffness.

Azimuth Control. The azimuth was controlled using a stepper motor and a 1:3 ratio chain/sprocket system (Fig. 9). The system is
attached to the top assembly (Fig. 3). The well plan was to come out in the N-S axis of the rock after drilling was completed. A stepper
motor was used to orient the pipe to the north direction (0 azimuth angle) at the start of drilling. The team attempted to use the down-
hole inclination sensor to autonomously adjust the azimuth. However, electromagnetic interference from the variable frequency drive
(VFD) corrupted the azimuth readings. Therefore, the N-S axis was determined from prior calibration using a compass. During drilling,
the torque generated downhole tends to turn the pipe counterclockwise to reduce the azimuth angle. Thus, the stepper motor was config-
ured with holding torque of 21 lbf-in. (2.37 Nm) to prevent unintended azimuth change. This was done by increasing the idle stepper
current to 90% of its total value, overheating the stepper. Because drilling time was approximately 11 minutes, the brief overheating
was not an issue. At a rotational speed of 1,150 rev/min, this holding torque was more than adequate to ensure a constant azimuth
throughout the drilling process.

Control Algorithms
The rig control is based on the analog signal (voltage) transmission. Input voltage is generated from electromechanical sensors and is
calibrated to specific field parameters. The sensors used this year are as follows: laser distance sensor, optical tachometer, torque trans-
ducer, WOB load cell, pneumatic air cylinder pressure sensor, water circulation flowmeter, and downhole inclination sensor (Fig. 10a).
The output values from these sensors are continuously monitored during the drilling process. The recorded signals go through a central
data hub or data acquisition (DAQ) system (Fig. 10b), which is connected to the rig computer through a USB port. The rig computer
used a control program written in National InstrumentsÔ LabVIEW systems engineering software. The program incorporates control
algorithms required for drilling process control and optimization.
The control algorithms generate and send appropriate output voltage responses to actuators (electropneumatic converters and motor
controllers or VFDs) upon receiving the input signals. These voltages are also calibrated to specific actuation parameters, such as the
bottom air pressure of the pneumatic cylinder for WOB control or frequency of the VFD for topdrive rotational speed control. The pro-
cess of continuous monitoring and actuation forms the basis of the rig automation during drilling (Fig. 11).

6 2020 SPE Drilling & Completion

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 7 Total Pages: 10

Fig. 9—Stepper motor with azimuth control mechanism.

(a) (b)

Fig. 10—(a) Downhole sensor unit, and (b) data acquisition system.

Drilling
process
parameters

Actuators Sensors

DAQ DAQ
system system

Rig
computer
LabVIEW

Fig. 11—Rig control architecture (after Akita et al. 2018).

2020 SPE Drilling & Completion 7

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 8 Total Pages: 10

Control Optimization and Rig Tests. The prior knowledge going into this project was that increasing WOB increases inclination
build (Yang et al. 2008). Also, from initial tests, it was seen that high rotations per minute (RPM) increased ROP and caused drifting of
the bit toward the right, away from the intended target described in the competition guidelines. This was theorized as being due to
downhole reaction torque on the bit. As previously discussed, the issue of pipe turning was solved by setting the stepper motor to its
maximum holding torque to counteract the turning action. However, the WOB is a limiting factor, and because this was the very first
time we drilled directional wells on this scale, we ran various tests to quantify the increase in inclination with WOB and established a
relationship between them to develop a control algorithm. A rotational speed of approximately 200 rev/min was used at the start of drill-
ing to get the BHA started on its trajectory. The RPM was then gradually increased to improve ROP. Because the holding torque of the
stepper motor was enough to maintain the pipe in place, pipe turning was not an issue. The graph in Fig. 12 shows data from the final
test that was used to categorize operating zones in our control algorithm.

30
–ROP @ 50 lbf (222.41 N) and 1,150 rev/min = 1.8 in. (4.57 cm)/min
25 –Average torque = 20 lbf-in. (2.26 N·m)
–11 minutes drill time
–25.7 in. (65.28 cm) total depth
Distance (in.)

20

15
Third zone, 37 to 50 lbf

10
Second zone, 25 to 30 lbf

5
First zone, 20 to 25 lbf

0
0 2 4 6 8 10 12
Time (minutes)

Fig. 12—Experimental test results for Table 1, Test 9.

As shown in Fig. 12, starting with 20 lbf (88.96 N), WOB allows the drill bit to get started on its trajectory. There was a successive
gradual increase to 37 lbf (164.58 N) and later to 50 lbf (222.41 N) WOB at specific intervals (first, second, and third zones) to
guide the bit to follow roughly the planned trajectory of the well. These predetermined setpoints were optimized by the feedback loop
[proportional integral (PI) loop] between the controls and the downhole sensor (Fig. 10a) that was installed to collect and send real-time
inclination data. When the sensor was operational, the WOB was fine-tuned to maintain a suitable dogleg severity to ensure inclination
change of 2 to 3 in the first zone, 3 to 5 in the second zone, and 6 to 7 in the third zone. During the competition, the sensor failed,
and drilling was conducted without setpoint optimization. Because a pneumatic power system is involved in the control process, the rig
reaction time (lag time) is factored into the algorithm using PI controllers. This continuous monitoring of the downhole sensor inclina-
tion data coupled with the discrete WOB values in the three zones and the PI controller guide the bit to the desired target.

Test L = Pilot Length in Inches (cm);

Number D = Pilot Diameter in Inches (cm) WOB in lbf (N) Target
1 L = 2 (5.08), D = 2.25 (5.72) 25 (111.21) to 28 (124.55) Yes
2 L = 2 (5.08), D = 2.25 (5.72) 23 (102.31) to 35 (155.69) 3 in. above bottom
3* L = 4 (10.16), D = 2.25 (5.72), 1.5 (3.81) 23 (102.31) to 30 (133.45) 3 in. above bottom
4 L = 5 (12.70), D = 1.5 (3.81) 16 (71.17) to 27 (120.10) Missed; bottomhole
5 L = 4 (10.16), D = 1.5 (3.81) 23 (102.31) to 28 (124.55) Stopped drilling
6 L = 4 (10.16), D = 1.5 (3.81) 23 (102.31) to 33 (146.79) Stopped; Allen screws and epoxy broke
7 L = 4 (10.16), D = 1.5 (3.81) 20 (88.97) to 33 (146.79) Yes, 14 minutes
8 L = 4 (10.16), D = 1.5 (3.81) 20 (88.97) to 33 (146.79) No, drill-bit inserts
9 L = 4 (10.16), D = 1.5 (3.81) 20 (88.97) to 50 (222.41) Yes, optimized
*Test 3: D = 2.25 in. (5.72 cm) for the first 1 in. (2.54 cm) of pilot and reduced to 1.5 in. (3.81 cm) for the
remaining 3 in. (7.62 cm) of pilot.

Table 1—Experimental results from nine zonal categorization tests.

Zonal Categorization Tests. Several zonal categorization tests were conducted to establish a relationship between inclination and
WOB. For each test (Table 1), a variable was changed to learn more about the directional steering mechanism and improve the control
algorithm. The team started with a 2-in. (5.08 cm) deep pilot hole at a 2.25-in. (5.72 cm) diameter and slowly worked toward a 4-in.
(10.16 cm) deep pilot at a 1.5-in. (3.81 cm) diameter because these were the competition stipulations for the pilot hole. From Test 1, at a
given WOB value, the larger the pilot, the more space is available for the BHA to follow its planned trajectory; hence, it hits the target.
Higher WOB results in a higher build rate (Tests 2 and 3), missing the target. The team attempted a deeper KOP in Test 4, with similar
WOB values, and missed the target. With the 1.5-in. (3.81 cm) maximum allowable diameter, the BHA had limited bend in the pilot
hole, thus missing the target, with comparable WOB values from Test 1. At that point, the team increased WOB to higher values to
increase inclination. The remaining tests (Tests 5 through 9) provided more insight into the categorization of WOB zones for the

8 2020 SPE Drilling & Completion

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 9 Total Pages: 10

controls. The team also tested the drill-bit inserts received from Baker Hughes but did not get successful results (Test 8); hence, the
inserts were not used during the competition.
Fig. 13a shows the bit exiting at the well target (Test 9), demonstrating successful drilling. Fig. 13b displays the side profile of a pre-
liminary well drilled on a sandstone rock of similar properties during the early testing phases.

Microbit

(a) (b)

Fig. 13—(a) Well on a target in Test 9, and (b) the well profile of the preliminary well.

Challenges and Lessons Learned

Working on this project provided deep insights into the automation of directional drilling on a small scale. The biggest challenge was
optimizing the bent-sub angle. Several subs with different bent angles (5 to 15 ) had to be built for the optimized 12 angle to be
achieved. Thus, the lesson learned is that the wellbore geometry and BHA dynamics are variables during the drilling process that
strongly affect directional trajectory. Also, the rotating cable kept on breaking during the first prototype testing. This was due to stress
concentration points within the BHA. Using a universal joint helped to solve the problem because it isolated the two welded portions
and allowed free bit rotation.
Another issue was the optimization of the drilling process itself. The challenge was how to develop a control algorithm to autono-
mously and efficiently drive the drill bit to reach the desired target. The lesson learned is to start from what was known about WOB and
RPM and run a series of well-thought-out experiments, making incremental improvements after every iteration. Coupled with electro-
mechanical sensors, this resulted in good optimization of the directional drilling process. Furthermore, developing a reliable downhole
measurement system was one of the major challenges. The system was needed for actual real-time optimization during the directional
drilling process. However, it has to withstand downhole vibrations to operate in the vicinity of the drill bit. Due to excessive vibrations,
the RPM had to be reduced every time an inclination reading was taken, which caused a slight increase in total drilling time.
A properly designed drill bit is critical for success in directional drilling. The team used an old drill bit from previous years, which
provided unsatisfactory results. Upon receiving the microdrilling bit from Baker Hughes, the team had a quantum leap in directional
drilling results. Moreover, it was observed that the pipe stiffness was critical for success on the laboratory scale. The team drilled nine
successful test wells. However, after purchasing a new drillpipe with the same specifications from the same vendor, the team experi-
enced difficulties in replicating the results. This was theorized as due to minor variations in the stiffness of the pipe. Thus, it is critical
to establish a relationship between pipe stiffness and variables in the control algorithm to successfully replicate results on this scale.

Conclusions and Recommendations

Conclusions.
1. The cable drilling was a novel technique in this project and proved effective in driving the bit to the target.
2. The modifications introduced to the rig and the designed bent-sub were very robust; as a result, no mechanical parts failed during the
drilling demonstration.
3. The zonal categorization of WOB for different ranges and dynamic modeling implemented in the control algorithm proved effective
in drilling the desired directional well.

Recommendations.
1. The downhole sensor was affected by bit vibrations and electromagnetic interference from the VFDs and could not provide accurate
inclination and azimuth readings. Using a Faraday cage is critical in isolating the VFD. Also, designing a BHA with a better down-
hole sensor would help in automating and improving azimuth and inclination control.
2. Further work should be done on quantifying pipe stiffness and establishing the relationship between pipe stiffness and angle of bend
in directional drilling. This will enhance directional results on a laboratory scale. Also, heat treatment of drillpipes and subjecting
pipes to multiple fatigue cycles could be studied to understand how this influences directional drilling.
3. More work can be done in optimizing the BHA for better results.
4. ROP optimization for directional drilling can be studied and improved from recorded data and published literature.

2020 SPE Drilling & Completion 9

ID: jaganm Time: 15:21 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024

 DC199640 DOI: 10.2118/199640-PA Date: 15-July-20 Stage: Page: 10 Total Pages: 10

Acknowledgments
We would like to express our gratitude to Catalin Teodoriu, Jeffery McCaskill, Aditya Sharma, and Jack Borer for their technical and
logistical support. We would also like to thank the Mewbourne School of Petroleum and Geological Engineering of the University of
Oklahoma for providing the support needed for the competition and for creating a stimulating learning environment for this
unique opportunity.
We extend our appreciation to all the members of the DSATS Drillbotics committee for helping to organize the competition, create
the guidelines, and supervise the tests during the competition demo. Furthermore, we are extremely grateful to all the sponsoring com-
panies of Drillbotics: Halliburton, Baker Hughes, Helmerich & Payne, ASME Petroleum Division, Evolution Engineering, Wintershall,
and Oxy. Additional thanks to Halliburton for hosting the competition in North America and for supplying the rock samples, and to
Baker Hughes for supplying the microbits.

References
Akita, E., Dyer, F., Duggan, P. et al. 2018. 2019 Phase I Final Design Report, DrillboticsÔ International University Competition, the Drilling Systems
Automation Technical Section (DSATS) of SPE, https://drillbotics.com/download/2019_design_submissions/2019-OU-Drillbotics-Design-Rpt.pdf
(accessed 29 October 2019).
Ambriz, R. R., Mesmacque, G., Ruiz, A. et al. 2010. Effect of the Welding Profile Generated by the Modified Indirect Electric Arc Technique on the
Fatigue Behavior of 6061-T6 Aluminum Alloy. Mater Sci Eng A 527 (7–8): 2057–2064. https://doi.org/10.1016/j.msea.2009.11.044.
Drillbotics. 2019. Archives and Previous Winners, https://drillbotics.com/archives/ (accessed 2 May 2020).
Drillbotics 2018–19 Guidelines, The Drilling Systems Automation Technical Section (DSATS) of SPE, https://drillbotics.com/download/guidelines/
guidelines_archive/Drillbotics-Guidelines-2019_4.pdf (accessed 29 October 2019).
PDC Micro-Bit for 2019 SPE Drillbotics Competition, the Drilling Systems Automation Technical Section (DSATS) of SPE, https://drillbotics.com/
download/2019-Drillbotics-Microbits-BHGE.pdf (29 October 2019).
Weijermans, P., Ruszka, J., Jamshidian, H. et al. 2001. Drilling with Rotary Steerable System Reduces Wellbore Tortuosity. Paper presented at the SPE/
IADC Drilling Conference, Amsterdam, The Netherlands, 27 February–1 March. SPE-67715-MS. https://doi.org/10.2118/67715-MS.
Yang, D., Rahman, M. K., and Chen, Y. 2008. Bottomhole Assembly Analysis by Finite Difference Differential Method. Int J Numer Methods Eng
74 (9): 1495–1517. https://doi.org/10.1002/nme.2221.

10 2020 SPE Drilling & Completion

ID: jaganm Time: 15:22 I Path: //chenas03.cadmus.com/Home$/jaganm$/SA-DC##200024