Dogleg
Dogleg
Dogleg
139-151
139
Abstract
The dogleg severity is one of the most important parameters in directional drilling. Improvement of
these indicators actually means choosing the best conditions for the directional drilling in order to reach
the target point. Selection of high levels of the dogleg severity actually means minimizing well
trajectory, but on the other hand, increases fatigue in drill string, increases torque and drag, particularly
in the rotation mode. Therefore the aim is to define the index in an optimal range which meets both
requirements. Particle swarm algorithm was used for optimization the dogleg severity. The final
measured depth and directional well pattern were considered as an objective function and Build & Hold,
respectively. Then the fatigue caused by the stresses exerted on the drill string, evaluated by modified
Goodman equation simultaneously. The relationship between path parameters and the obligation to
reach a target point in directional wells, converts the problem into a constrained optimization problem.
Comparing the proposed directional drilling path in a drilled well in the Ahwaz oilfield with the
responses obtained from the particle swarm algorithm indicated that the particle swarm algorithm is
converged in finding the shortest path, and on the other hand, it decreases the time of using directional
drilling equipment due to the selection of the proper dogleg severity. Note that it is likely to add other
constraints to the optimization process which indicates the particle swarm algorithm efficiency in
solving these problems.
Tel: +989166133411
140
Journal of Chemical and Petroleum Engineering, Vol. 48, No.2, Dec. 2014
Dogleg severity
Usually the dogleg severity indicates
deflection amount of oil or gas well per
hundred feet which is in two dimensions
and is one of the most important parameters
for directional drilling. In fact optimization
of this index means selecting the best
directional drilling conditions to achieve
the target point. Choosing large values for
the dogleg severity actually means
decreasing drilling path but on the other
side results in increased fatigue of the drill
string, increased torque and drag force,
especially in case of rotational drilling and
eventually increased the possibility stuck of
the drill string or creation key seat.
Therefore, the objective is that the
mentioned index can be defined in such an
optimal range which satisfies both
requirements.
The relation between the dogleg severity
and final measured depth through the well
path equations provides the possibility to
select the final measured depth as the
objective function to optimize it.
cos
(2)
(3)
(4)
(5)
Optimization of Dogleg ..
141
142
Journal of Chemical and Petroleum Engineering, Vol. 48, No.2, Dec. 2014
(6)
Optimization of Dogleg ..
143
(7)
1
(8)
(9)
3-
4-
5-
67-
Modeling
In mathematical, optimization means
minimizing or maximizing a function for
certain variables. In this case the variables
are called optimal variables. On the other
hand in search space the aim is to find
number of that the function such as
should be maximized or minimized,
although for finding the answer some
constraints should be considered. In this
paper, choosing the best dogleg severity is
the purpose, not necessarily the highest or
the lowest one. So the measured depth
144
Journal of Chemical and Petroleum Engineering, Vol. 48, No.2, Dec. 2014
(10)
According to the Build & Hold pattern,
there are three variables that are as follows:
KOP
Dogleg severity
Maximum well angle
But these variables are related to each
other according to the well path
relationships, so there are two degrees of
freedom in this pattern. Thus the number of
variables in this problem is two that are the
KOP and the dogleg severity. But the third
variable, maximum well angle can be
considered as a constraint.
In order to determine the initial
population of particles, the range of
variables must be specified. But due to the
range of start of KOP is different to dogleg
severity, also in order to create the initial
population, both variables are defined in
two separate vectors and then the initial
population will be formed with their
combination in a matrix. Therefore each
particle will have two properties that are
KOP and dogleg severity. The range of
KOP can be considered in two ways:
Optimization of Dogleg ..
145
Case study
According to the research subjects
Constraints
KOP,DLS,INC
Well Description
Field
Ahvaz
Well Number
359
Hole Size
12 in
Shoe 13 3/8
1893.5 m
Flow Rate
520 Gpm
Pressure
2800 Psi
ROP Average
1.3 m
Mud Properties
Mud Weight
145 PCF
M.F.Viscosity
55
Solid
45%
67
117
PV
49
YP
19
DEPTH
0
1909
2185.14
2683.54
ANGLE AZIMUTH
0
0
1.5
226
38.32
225.85
38.32
225.85
TVD
0
1908.78
2163.97
2555.00
N/S
0
-17.97
-81.73
-296.98
E/W
0
-17.97
-84.30
-306.04
Lower Limit
Upper Limit
KOP
DLS
1909
1
2555
7
SECTION DOGLEG
0
0
24.99
0.02
117.42
4
426.45
0
146
Journal of Chemical and Petroleum Engineering, Vol. 48, No.2, Dec. 2014
2696
2694
2692
2690
2688
2686
2684
2682
2680
2678
2676
Particle
20
40
60
160
180
200
Figure 5: Dispersion of solutions in search space. It shows the possible of reaching to target
base on population.
Table 5: Shortest path with respect to the path parameters
DEPTH
0
ANGLE AZIMUTH
0
0
TVD
0
N/S
0
E/W
0
SECTION DOGLEG
0
0
1909
1.5
226
1908.78
-17.97
-17.97
24.99
0.02
2052.17
34.91
225.85
2042.87
-48.07
-49.62
69.09
2676.65
34.91
225.85
2554.97
-297
-306.05
426.47
Optimization of Doglegg ..
147
Table 6: Shortest
S
path with consideering fatigue criteria
c
and mechanical
m
p
parameters.
Section Calculated
C
on
o 225.85
D
DEPTH
A
ANGLE
AZ
ZIMUTH
TVD
N/S
E
E/W
SECT
TION DOG
GLEG
1909
1.5
226
1908.78
-17.97
-
-177.97
244.99
0.02
0
2
2110.27
36.32
2
225.85
2
2096.27
-62.09
-
-644.06
899.22
5.19
5
2
2679.61
36.32
2
225.85
2
2555.00
-2
296.98
-3006.02
4266.43
Torqu
ue in String
D
Drag(Tension)
in
Figure 7:
7 Torque & Drag in drilll string, comp
parison betweeen proposed
d & optimized
d well path. Inn this case
differrence is negligiible so choosingg the proposed well
w path due to
o its lower dragg and torque is not considered
d as an
advantage.
do
ogleg severrity is thee highest possible
vaalue. Also in
i this casee, the safetty factor
du
ue to fatiggue regardding to maximum
m
applied stresss to the drilll string is below
b
1.
In the other words,
w
the ffatigue cond
dition is
148
Journal of Chemical and Petroleum Engineering, Vol. 48, No.2, Dec. 2014
Optimization of Doglegg ..
149
depth of
o the well iss shown in Figure 10 and
a
the sloppe of this chhart represents the doglleg
severityy.
In this
t
paper by choosing measuured
depth as an objecctive functioon, the suitaable
amount of the doogleg severity has beeen
obtainedd. In this prrocess well path and drill
d
string constraints applied simultaneou
s
usly
with fullly operationnal look.
In opptimized weell path thee final anglee is
2 degrees less thann proposed well
w path, this
t
can be better for drilling operational. The
T
measureed depth and curvee section in
optimum
m case was lesss. Thus by
considering the sppeed drillinng in drilling
well, thhe time of using
u
direcctional drilling
equ
uipment caan be decreeased by 58
8 hours.
Also about ussing particle swarm allgorithm
wee can impply to its convergen
nce for
ach
hieving the optimal sollution.
The impoortant pointt is that insstead of
cho
oosing opttimized paath based on the
fattigue of driill string, cchoosing op
ptimized
patth can bee done bassed on th
he other
con
nditions suuch as the smallest off torque
and
d drag. Thiis is especiaally very im
mportant
in wells extennded reach.
Accknowled
dgements
The preseent paper w
was preparred with
ment of
colllaboration with drilling departm
Naational Iraanian Soutth Oil Co
ompany.
2692
MeanMD
D
2690
MaxMD
D
2688
2686
2684
2682
2680
2678
2676
0
20
40
60
80
100
0
120
Iterattion
1
140
160
180
200
9 the highestt, the mean vaalue, and thee lowest measured depth at
a each step. C
Convergence of
o particle
Figure 9:
swarm
m algorithm which
w
results too find the shorrtest path is coompletely eviddence.
150
Nomen
nclature
,
Positive acceleration
a
c
constants
in
PSO
TVD to KOP
K
TVD to EOB
E
TVD to taarget
Safety Faactor
Random Number
N
(0,11) in PSO
Radius off Curvature
Endurancce Limit
Ultimate Tensile Strenngth
Velocity of Particle inn Search
Space
Inertia Weight
W
in PSO
O
TVD
T
Posiition of Partiicle
Incllination
Streess Amplitudde
Bennding Stress
Meaan Stress
Axial Stress
Vonn Mises Stresss
Shear Stress
Doggleg Severityy
Endd of Build
Incllination
Meaasured Depthh
MD
D at end of buuild
Truee Vertical Deepth
Refereences:
1- Samppaio, H.B.J. (2006). "Plaanning 3d well
w
trajectoriees using cubbic functionss." J. Energyy Resour.
Technnol., Vol. 128, No. 4, pp. 257-267.
2- Guo,Y
Y. and Feng, E. (2008). "N
Nonlinear dynnamical system
ms if trajectorry design for 3d horizontall well and
their optimal
o
controols." J. Compput. Appl. Matth., Vol. 212, No.
N 2, pp. 1799-186.
3- Zhangg, J.B., Xin Liu,
L
X. and Hong Lv, X.
X (2012). "M
Multi-objectivee optimizationn method forr borehole
trajectory design of directional well."
w
Appl. Mech.
M
Mater., Trans Tech Pub., Switzerrland,Vol.152
2, pp. 816819.
S
and Abdd Wally, A.A.W. (2004). ''O
Optimal 3-d directional & horizontal
4- Shokirr, E.M., Emera, M.K, Eid, S.M.
wells planning usinng genetic algoorithm.'' Oil Gas.
G
Sci. Technol., Vol. 599, No. 3, pp. 255-266.
"
efficient methodology of productionn strategy opttimization
5- Nogueeira, P.D.B. annd Schiozer, D.J. (2009). "An
basedd on genetic algorithms."
Optimization of Dogleg ..
151
7- Aadnoy, B.S. (2008). "Theory and application of new generalized model for torque and drag." IADC/SPE
Asia Pacific Drilling Technology Conference and Exhibition, Jakarta,Indonesia.
8- Keith, J.N. and Budynas, R.G.(2006). Shigley's Mechanical Engineering Design, 8th.Ed, McGraw-Hill, USA.
9- Engelbrecht, P. and Andries.A. (2007). Computational Intelligence, 2.Ed, John Wiley & Sons Ltd, South
Africa.
10- Lazinica, A. (2009). Particle Swarm Optimizaion, In.Tech, Croatia.