You are on page 1of 9

Tunnelling and Underground Space Technology 16 Suppl.

1 (2002) S47–S55

Mechanical modelling for pipes in horizontal directional drilling


Maria Anna Polak1,*, Afdal Lasheen2
Department of Civil Engineering, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1

Abstract

This paper presents a new technique to predict forces and deformations of pipes installed underground by horizontal directional
drilling (HDD). The proposed technique consists of several models each describing different aspects of mechanics of HDD. It
allows prediction of the shape of the pipe in the borepath. Based on the pipe shape, shape of the borepath and pipe stiffness, the
contact forces between the pipe and the borepath are calculated. The contact forces cause friction and thus axial pulling force is
induced to equilibrate the frictional forces and the weight of the pipe. The analytical model is used to study the stresses and
strains on two pipes that were tested in the field. The theoretical results are compared with the strains measured in the field.
䊚 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Horizontal directional drilling; Predictions; Pulling forces; Strains; Modelling

1. Introduction borepath, material properties of the pipe and the drilling


fluids.
Horizontal directional drilling (HDD) is a popular
method for installation of both steel and plastic under- 2. Mechanics of HDD
ground pipelines (Ghokale et al., 1999; Khan et al.,
1994; Kirby and Kramer, 1996). However, relatively The process of pulling a pipe in HDD installation
little theoretical research has been done in the area of involves interaction between the soil, the drilling fluid
development of rational analysis procedures for the and the pipe. The axial pulling force on the pipe is
behaviour of soil, pipe and drilling fluids during HDD primarily due to friction. This friction is composed of
installations. Such procedures are needed for the devel- friction between the pipe and the surrounding soil, and
fluidic drag of the drilling fluid flowing around the pipe.
opment of rational design guidelines for HDD, which is
The mechanisms that produce friction are flexural stiff-
necessary for growth of the industry. Some analytical
ness of the pipe as it negotiates the curves of the
procedures exist (Huey et al., 1996; Driscopipe䉸, 1993;
borepath, weight of the pipe, and the effect of pulling
Drillpath娃, 1996) and they were reviewed in the recent
the pipe around the curves.
publication by Baumert and Allouche (2000). However,
Flexural stiffness, weight of the pipe and the direc-
to a large extent, current design methodology relies on
tional changes in the borepath create contact forces
the experience and judgement of contractors, manufac-
between the pipe and the soil and these forces produce
turers and engineers. frictional forces. The velocities of the pipe and the
This paper presents a new analytical method to drilling fluid and the viscosity of the drilling fluid are
calculate forces and deformations in the pipes installed the reason that drag forces exist. The magnitude of the
by HDD. The method is simple but it accounts for all contact forces depends on the geometry of the borepath,
major mechanical factors influencing the pulling force the clearance between the pipe and the soil, the bending
and the bending deformations of the pipe. It considers stiffness of the pipe, the axial stiffness of the soil, and
the size of the pipe and the borehole, shape of the the weight of the pipe submerged in the drilling fluid.
All forces must be in equilibrium. The effect of
*Corresponding author. Tel.: q1-519-888-4567; fax: q1-519-888- actions of these forces is that the pipe is subjected to
6197.
E-mail address: polak@uwaterloo.ca (M.A. Polak). axial and bending deformations. Fig. 1 shows the forces
1
Associate Professor. that act on a segment of the pipe. N1, N2 and N3 are the
2
Doctoral candidate. normal reaction forces between the pipe and the soil, W

0886-7798/02/$ - see front matter 䊚 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 8 8 6 - 7 7 9 8 Ž 0 2 . 0 0 0 2 0 - 2
S48 M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55

Fig. 1. Forces acting on a pipe during HDD.

is the submerged weight of the pipe, Fd is the force due


to fluidic drag, T1, T2 are the tension forces on both
Fig. 2. Model for the geometry of the borepath.
ends of the pipe and m is the friction coefficient.

3. The proposed method 3.1. Pipe shape at the borepath corner

The described analytical method consists of models The model for pipe at the corner utilises the theory
for pipe shapes negotiating curves in the borepath, pipe of large deflections of flexible bars to derive the contact
friction due to weight, fluidic drag and changes of forces between the pipe and the soil. From the geometry
direction. The following assumptions are adopted.
of the pipe and the borepath, the following relationships
1. The borepath consists of straight sections with defined are derived:
angles of inclination for each section. This is a ymaxqc9sltanc (1)
conservative assumption because it creates sharper
deviations, i.e. greater than in reality angles for the where:
pipe.
B dqc E
2. The angles between the horizontal line and the seg-
c9sC ydF (2)
ments are a; positive anticlockwise. ai is the angle D cosc G
for the segment between points ‘i’ and ‘iq1’, Li is
the length for this segment and Li9 is a horizontal Therefore:
projection of the length (Fig. 2). c9ycosc
3. Soil is assumed to be a stiff support for the pipe. It ls (3)
does not deflect or collapse due to the pipe forces tancyymax yl
and the borepath retains its shape during the whole c is the clearance between the pipe and the borepath,
process. c is half the angle between the two line segments of
the borepath, and l is half of the distance between
The developed formulation allows evaluation of the
supports B1 and B2.
bending deformations that result from the pipe being
forced to negotiate angles in the borepath. Based on
these deformations, the frictional forces and the required
pulling forces are calculated.
In order to model the shape of the pipe in the
borepath, a borepath profile is approximated by inter-
secting straight lines with defined angles of inclination
as shown in Fig. 2 (Lasheen and Polak, 2001a,b) When
a pipe passes through a borepath, it must negotiate the
corners. To do so the pipe must deflect and contact the
soil. Normal forces acting on the pipe can be approxi-
mated using three contact points (B1, B2 and B), as
shown in Fig. 3. In the proposed model, the pipe exerts
forces on the soil, which applies equivalent reaction
forces on the pipe at points of contact. Within each
borepath section, the pipe profile can be approximated
as a simple beam supported on points B1 and B2 with a Fig. 3. Model for calculating flexural forces acting on a pipe negoti-
concentrated load at point B. ating an angle in the borepath.
M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55 S49

Fig. 4. Model for the fluid flow in the borepath.

In Eq. (3), ymax yl is determined using the following Where rp and ri are the pipe external and internal
equation (Frisch-Fay, 1962): radii, gp, gw are self-weights of pipe material and the
drilling fluid, respectively. The weights, w and wp are
w z
xy positive if acting downwards.
ymax y 2sinccosf0ycoscFŽp,f0.~
|

sw z
(4)
l xy
y 2cosccosf0qsincFŽp,f0.~ 3.3. Fluidic drag
|

Having calculated l, the normal force P is calculated During HDD, the pipe is pulled with the velocity, vp
from the following equation: while the drilling fluid is flowing in the opposite
EI w z2
direction. The average velocity of the drilling fluid can
Ps 2
cosc0xyy2cosc0cosf0qsinc0FŽp,f.|~ (5) be characterised by the total discharge Q wvolumeytimex.
l
This section explains the modelling method used for
where: computing drag forces exerted on a pipe during the
installation process due to the flow of the drilling fluid
1 and the velocity of the pipe (Schlichting and Kestin,
ps , f0swsinc0x1y2 and
y2 1968; Streeter and Wyle, 1975). The method assumes
FŽp,f0.s0.8472qFŽp,f0.y2EŽp,f0.. laminar flow of fluid in the borepath and the concentric
location of the pipe in the borehole. This last assumption
The functions FŽp,f0. and EŽp,f0. are elliptic inte- results in the symmetrical distribution of shear and
grals and can be evaluated using mathematical hand- velocity with respect to the centreline of the borepath
books (Abramowitz and Stequn, 1965). (Fig. 4).
Pipe maximum curvature can be estimated from: Let the pipe axis be selected as z-axis in the cylindri-
M 1 cal co-ordinates, and let r denote the radial co-ordinate
ks s wPlqPtanc0ymaxx (6) measured from the axis outward, as in Fig. 4. The
EI EI
velocity components of the flow in the radial and
tangential directions are zero. The velocity component
3.2. Weight of the pipe in the z-axis direction, denoted by v, depends on r alone.
The pressure is constant in every cross-section. The
One of the causes of frictional forces between the Navier–Stokes equation for the cylindrical co-ordinates,
pipe and the soil is the weight of the pipe. In addition for the z direction is as follows:
to the frictional forces, the pulling force must also
B d 2v 1 dv E dp
overcome the weight of the pipe itself. For the pipe
mC q Fs (9)
resting on the ground outside the borepath, the gravita- D dr
2
r dr G dz
tional weight is used:
dp
wpspŽr2pyr2i .gp (7) Where m is the viscosity of the fluid and is the
dz
Inside the borepath, the pipe is submerged in the energy gradient through the length of the pipe. The pipe
drilling fluid. The submerged weight of the pipe per moves with the velocity nsnp and the drilling fluid
unit length (when pulled empty) is: moves in the opposite direction. The total discharge of
the drilling fluid is Q wm3 ysx. Both vp and Q can be
wspŽr2pyr2i .gpypr2pgw (8) measured in the field. The boundary conditions are ns
S50 M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55

0 at rsrh and nsnp at rsrp, where rp and rh are the


radii of the pipe and the borepath, respectively. By
applying the boundary conditions the following equation
is obtained for the velocity of the fluid:
1 dp
vŽr.s
4m dz
B r2pyr2h E lnryrh
=Cr2yrh2q lnrh yrFqvp (10)
D lnrpyrh G lnrp yrh
The shear stress is given by the following equation

dv 1 dp B r2yr2 1 E mvp 1 Fig. 5. Forces acting on the pipe at the corner of the borepath.
tŽr.sm s C2ry p h Fq (11)
dr 4 dz D lnrp yrh r G lnrp yrh r
the pipe (Eq. 4). Considering that RsPycosc (Fig. 3),
The shear stress at the wall of the pipe is:
the following equation can be derived for DTsT2yT1:
1 dp B r2yr2 1 E mvp 1
t ps C2rpy p h Fq (12) B coscqm sinc E
DTsT1C y1F
b
4 dz D lnrp yrh rp G lnrp yrh rp
D coscymbsinc G
The total discharge of the fluid through the pipes is B 1 E
Q and can be calculated as: q4PmbC F (17a)
D coscymbsinc G

p dp B 4 4 Žrpyrh.
rh 2 2 2E

Qs | 2prvŽr.drs Crpyrhy F or
rp 8m dz D lnrp yrh G
DTsT1=C1Žc.qP=C2Žc. (17b)
p vp
q
2 lnrp yrh 4. Calculation of the pulling force
=Žrp2yrh2y2rp2lnrp yrh. (13)
dp The formulas for calculating pulling forces for the
From Eq. (13), we can calculate and substitute pipe, at different points within the borepath, are pre-
dz
back into Eq. (12) to find the shear stress at the wall sented in this section. In the descriptions the following
of the pipe. The force per unit length on the pipe, , is definitions are used:
given by: ● Ti-pulling force. The force required to pull the whole
fdsK=2prtp (14) pipe when the head (point of pulling) of the pipe is
in location ‘i’. From equilibrium, this force is equal
K is a parameter, which accounts for approximations to the axial force at the head of the pipe.
of the fluid model. The choice of the value for K is ● Tji-axial force in the pipe at location ‘j’, when the
explained later in the paper. head is in location ‘i’.

3.4. Forces resulting from the change of direction Let us consider a simple profile shown in Fig. 2. The
calculations are made for pulling forces at points 1 to
4. The total pulling force consists of the components
Fig. 5 shows a segment of a pipe ABC that changes
described below.
its direction from A to C. The pulling forces at end C
and at end A do not act on a straight line. An equili-
brating force exists and acts on the soil at the contact 4.1. Pulling forces due to weight of the pipe outside the
corner point B. The coefficient of friction between the borepath
pipe and the borepath is mb. The equation of equilibrium
in the direction of AC is: The frictional force resulting from pipe weight outside
T2coscsŽT1q2Rmb.coscq2PmbqNTmb (15) the borepath and the weight of the pipe create a
component Tig of the pulling force. At point ‘i’ it is
Equilibrium in the direction normal to AC: equal to:
NTsŽT1qT2.sinc (16) B iy1 E
TigsŽwpmgcosa0qwpsina0.CLy 8 LkF (18a)
Where PR are the forces due to flexural stiffness of D ks1 G
M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55 S51

or These components are equal to the axial forces . For


point 1 only change of direction is included:
B iy1 E
TigsC3=CLy 8 LkF (18b) T13fsT13=C1Žc1. (22)
D ks1 G
where
Where wp is the weight of the pipe per unit length,
mg is the friction coefficient between the pipe and the T13sŽLyL1yL2.C3
ground, L is the total length of the pipe, Lk are the For point 2 both change of direction and flexural
lengths of segments of the borepath and a0 is the angle stiffness of the pipe are included. Therefore:
between the horizontal and the ground surface in front
of the entry point. T23fsT23=C1Žc2.qP2=C2Žc2. (23)
where T23sT3gqT2sqT2dqT13f and P2 is the normal
4.2. Pulling forces due to weight of the pipe inside the force resulting from bending stiffness of the pipe at
borepath point 2 wEq. (4)x.
For the pulling head at point 4 and further points,
The component of the pulling force Tis results from forces T14f , T24f and T34f are calculated using the
friction due to the submerged weight of the pipe. procedure analogous to the procedure for point 3.
Included in Tis is also a component of the pulling force
due to the submerged weight. At point ‘i’ it is: 4.5. Total pulling force at point ‘i’
iy1
Tiss 8 ŽZLkwmbcosakZqLkwsinak. (19) The total pulling force at point i is equal to the sum
ks1 of the components:
iy1
where w is the submerged weight of the pipe per unit
length, mb is the friction coefficient between the pipe TisTigqTisqTidq8Tjif (24)
js1
and the borepath, ak is the angle between the horizontal
and the segment k and Lk is the length of this segment.
5. Analysis of HDD installations
Note that segment in front of k is denoted as ky1.
The derived theoretical formulations were used to
4.3. Pulling forces due to drag of the drilling fluid analyse forces and deformations related to horizontal
directional drilling installations. The comparison with
The fluidic drag results in the following increase in experimental data was done based on two field tests
the pulling forces Tid, at point ‘i’: described by Gelinas et al. (2000), Gelinas (1998). The
iy1 tests involved 203 mm (8 inches) in diameter, standard
Tidsfd 8 Lk (20) dimension ratio (SDR) 17, HDPE pipes pulled along 55
ks1 and 90 m (180 and 295 feet) borepaths. In test 1 (pull
Where f d is the force per unit length on the pipe due 1) a 340 mm (13.5 inches) backreamer was used while
to drag of the drilling fluid wEq. (14)x. in test 2 (pull 2) the 305 mm (12 inches) backreamer
was used.
4.4. Pulling forces from change of direction and flexural 5.1. Analysis of the drag force
stiffness of the pipe
The derived equations wEqs. (9)–(14)x were used to
The pulling forces due to bending stiffness and change calculate shear stress and drag force per metre at the
of direction are denoted as Tif. At the entry point T1f s pipe wall for different pulling conditions. The parame-
0. When the pipe reaches point 2, the component of the ters for pull 1 and pull 2 were calculated and are
pulling force T2f is due to the change of direction at presented in Table 1. Also included in the Table 1 are
point 1 (bending stiffness is ignored at point 1): sample results of the analyses for pull 2, with several
T2fsT12=C1Žc1. (21) parameters changed, namely pipe speed, fluid viscosity,
pipe radius and borehole radius.
where Fig. 6 shows the relationship between the ratio of
T12sŽLyL1.C3 borehole and pipe diameters and the drag force for the
pipe (for 203 mm, i.e. 8 inches, diameter pipe). It can
When the head of the pipe reaches point 3, the pulling be seen that as the ratio decreases (which means that
force contains components resulting from friction due the clearance between the pipe and the soil also decreas-
to bending and change of direction at points 1 and 2. es), the drag force rapidly increases. Fig. 7 shows the
S52 M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55

Table 1
Sample analyses of the fluidic drag forces for the pipe in the borehole

Test Hole Pipe Ratio: Viscosity Pipe Discharge Pressure Shear Fluidic
radius radius holey (Nys m2) speed (m3ys) gradient (Nym2) drag
(m) (m) pipe radius (mys) (Nym2 m) wkNymx
Ks1
Pull 1 0.17145 0.1016 1.7 0.02 0.01183 y1.73Ey03 1.65135 y0.0687 y0.000044
Pull 2 0.1524 0.1016 1.5 0.02 0.02633 y1.73Ey03 y5.00882 y0.15055 y0.000096
Pull 2, influence of different parameters on shear stress
influence of pipe speed
0.1524 0.1016 1.5 0.02 0.10533 y1.73Ey03 y8.18255 y0.2762 y0.000176
0.1524 0.1016 1.5 0.02 0.26333 y1.73Ey03 y14.5300 y0.5275 y0.000337
influence of viscosity
0.1524 0.1016 1.5 0.04 0.02633 y1.73Ey03 y10.0176 y0.3011 y0.000192
influence of pipe radius
0.1524 0.0508 3 0.02 0.02633 y1.73Ey03 y0.80559 y0.06348 y0.000020
influence of borehole radius
0.1270 0.1016 1.25 0.02 0.02633 y1.73Ey03 y39.7169 y0.5486 y0.000350
0.1097 0.1016 1.08 0.02 0.02633 y1.73Ey03 y1208.99 y5.0462 y0.003220
0.1068 0.1016 1.05 0.02 0.02633 y1.73Ey03 y4665.98 y12.2956 y0.007849

relationships between pipe diameter and the logarithm


of drag force, for three different ratios of pipe radius to
borehole radius. The smaller the pipe radius the larger
the drag for the same boreypipe radii ratio. The reason
for that is the fact that for smaller pipe and the same
ratio, the clearance between the pipe and the soil is
smaller. The smaller clearance is in this case, the cause
for drag force increase. Similarly, for all pipe diameters,
the smaller the ratio the larger the drag force. It should
be noted, however, that the results are calculated using
the same fluidic discharge rate of 0.00173 m3 ys (taken
from pull 1 and 2, Table 1).
The most important conclusion that can be derived
Fig. 7. Influence of pipe ratio on the fluidic drag force, for different
from these analyses is that, in most typical cases, the borehole to pipe radius ratios.
fluidic drag force does not significantly increase the
pulling force. It is worth noting that the equations used
for calculations presented in Table 1 were derived
assuming concentric location of the pipe in the borehole. Table 2
Also, real drilling fluids exhibit non-Newtonian type of Input data for the analyses in Tables 3–5
behaviour. To reflect these approximations adopted in
Input variable Value
Bore-pipe friction 0.3
coefficient
Ground-pipe friction 0.3
coefficient
Pipe internal diameter wmx 0.179
Pipe external diameter wmx 0.203
Bore diameter wmx Pull 1 0.343
Pull 2 0.305
Clearance wmx 0.102
Cross-sectional area wm2x 0.00718
Young’s modulus E wkPax 700000
Moment of inertia I wm4x 3.2955E-05
EI wkNm2x 23.0687
Drag force wkNymx Pull 1 0.00044
Pull 2 0.00096
Pipe weight wkNymx 0.0662
Pipe submerged weight y0.2519
Fig. 6. Influence of the borehole to pipe radius ratio on the fluidic wkNymx
drag force.
M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55 S53

Table 3
Theoretical predictions for Pull 1 and Pull 2

Point x y Bending Pulling Stress Axial


strain force wkPax strain
w%x wkNx w%x
Model Test Model Test
Pull 1 1 0 0 0.46 64 0.01 0.00
2 11.46 y1.91 0.35 0.50 1.71 239 0.03 0.00
5 44.26 y1.95 0.47 0.50 4.22 587 0.08 0.05
6 54.36 0 4.96 691 0.10 0.15
Pull 2 1 0 0 0.76 106 0.02 0.00
2 11.52 y1.94 0.49 0.75 2.03 282 0.04 0.14
5 72.00 y1.95 0.20 0.40 6.61 921 0.13 0.51
6 90.68 0 7.70 1073 0.15 0.55
Comparison with test results.

the model, a simplified approach was taken by introduc- relatively small component of the pulling force. For
ing the coefficient K)1 in Eq. (14). In the analyses of example, for a 203-mm (8 inches) pipe pulled through
the pulling force presented in this paper K was taken as a hole 1.5 times larger (pull 2), the fluidic drag is
10, however, further research is needed to determine the 0.000096 kNym. For a 90-m pull this means a total
frictional forces resulting from fluidic drag. pulling force from fluidic friction of 0.0086 kN for Ks
It is worth noting that the value of K has little 1 and 0.086 kN for Ks10. These constitute 0.11% and
practical meaning for the calculations presented in this 1.1% of the total pulling force, respectively (the total
work. All calculated values of shear stress result in a pulling force is approx. equal to 8 kN, Table 3).
Another interesting observation regarding the fluidic
Table 4 drag is that increasing the pipe velocity by a factor of
The effect of changing borepath profile for pull 2 predictions 10, results in the increase in the drag force by a factor
Point x y Bending Pulling Stress Axial
of 3.4 (y0.00034yy0.0001s3.4, Table 1). Also,
strain force wkPax strain increasing viscosity by a factor increases the drag force
w%x wkNx w%x by the same factor (Table 1).
Profile 1 1 0 0 0.76 106 0.02
(pull 2) 2 11.52 y1.94 0.49 2.03 283 0.04 6. Predictions using the proposed model
3 25.00 y1.945 0.00 3.45 480 0.07
4 65.00 y1.945 0.00 6.16 858 0.12 The presented mechanical formulations were com-
5 72.00 y1.95 0.20 6.64 924 0.13 bined into a predictive model. The model was used for
6 90.68 0 7.73 1076 0.15
Profile 2 1 0 0 0.76 106 0.02 the analyses of the pipes tested in the field (Gelinas et
2 11.52 y1.94 0.92 2.03 283 0.04 al., 2000). During the tests, the only measured variables
3 25.00 y1 0.25 3.86 537 0.08 were longitudinal strains at four locations around the
4 65.00 y3 0.68 7.34 1022 0.15 perimeter of the pipe (at 0, 90, 180 and 2708 from the
5 72.00 y1.95 0.04 8.64 1203 0.17 crown). From these strains, the strains due to axial
6 90.68 0 9.52 1326 0.19
Profile 3 1 0 0 0.76 106 0.02 pulling force and strains due to bending were calculated
2 11.52 y1.94 0.69 2.03 284 0.04 (using the input data given in Table 2, see Table 3).
3 40.00 y1 2.72 4.51 628 0.09 The comparison is made between field measured and
4 45.00 y3 2.79 11.10 1545 0.22 calculated strains. The predicted flexural deformations
5 72.00 y1.95 0.08 19.46 2710 0.39 compare well with the field data. The predicted axial
6 90.68 0 20.59 2868 0.41
Profile 4 1 0 0 0.76 106 0.02 strains are smaller than the experimental ones. This can
2 11.52 y1.94 0.69 2.03 283 0.04 be attributed to several factors the most important one
3 25.00 y1.5 0.06 3.62 504 0.07 being that the presented analyses are simplified because
4 65.00 y2.5 0.20 6.65 926 0.13 only three straight lines approximate the drill path and
5 72.00 y1.95 0.02 7.30 1016 0.15 thus not all forces due to bending deformations are
6 90.68 0 8.12 1131 0.16
Profile 5 1 0 0 0.76 106 0.02 included in the calculations (the borepath profile for
2 5.00 y1.94 2.18 1.64 228 0.03 pull 2 is shown as profile 1 in Fig. 8). The value for
3 25.00 y1.945 0.00 6.83 951 0.14 Young modulus was taken as 700 MPa, which represents
4 65.00 y1.945 0.00 9.50 1322 0.19 a short-term modulus for high-density polyethylene. In
5 85.00 y1.95 1.78 10.83 1508 0.22 the tests, the effective modulus was likely to be less
6 90.68 0 14.57 2029 0.29
than this adopted value. Also, a simplifying assumption
S54 M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55

Fig. 8. Borepath profiles used in the studies presented in Table 4.

Table 5
The effects of different parameters for pull 2, profile 1 predictions

Parameters Points x y Bending Pulling Stress Axial


changed strain force wkPax strain
w%x wkNx w%x
Young’s 1 0 0 0.76 106 0.04
moduluss300 2 11.52 y1.94 0.49 2.03 283 0.09
MPa 3 25.00 y1.94 0.00 3.22 449 0.15
4 65.00 y1.945 0.00 5.94 827 0.28
5 72.00 y1.95 0.20 6.41 893 0.30
6 90.68 0 7.43 1035 0.35
Young’s 1 0 0 0.76 106 0.00
moduluss 2 11.52 y1.94 0.49 2.03 283 0.00
200 000 MPa 3 25.00 y1.94 0.00 116.62 16239 0.01
(value for 4 65.00 y1.945 0.00 119.46 16634 0.01
steel) 5 72.00 y1.95 0.20 120.09 16723 0.01
6 90.68 0 155.75 21688 0.01
Borehole 1 0 0 0.76 106 0.02
diameters0.254 m 2 11.52 y1.94 0.97 2.06 287 0.04
3 25.00 y1.94 0.00 4.69 653 0.09
Drag 4 65.00 y1.95 0.00 7.50 1045 0.15
forces0.0035 k 5 72.00 y1.95 0.41 8.00 1114 0.16
Nym 6 90.68 0 9.50 13234 0.19
Borehole 1 0 0 0.76 106 0.02
diameters0.228 m 2 11.52 y1.94 1.97 2.19 305 0.04
3 25 y1.94 0.00 9.79 1363 0.19
Drag 4 65 y1.945 0.00 13.03 1815 0.26
forces0.0139 5 72 y1.95 0.83 13.61 1895 0.27
kNym 6 90.68 0 16.86 2347 0.34
Borehole 1 0 0 0.76 106 0.02
diameters0.216 m 2 11.52 y1.94 3.74 2.64 367 0.05
3 25 y1.94 0.00 27.53 3833 0.55
Drag 4 65 y1.945 0.01 32.33 4502 0.64
forces0.0525 5 72 y1.95 1.60 33.20 4623 0.66
kNym 6 90.68 0 42.69 5945 0.85

of linear elastic behaviour of polyethylene was used in pipe is forced to negotiate smaller angles in the borepath.
the presented calculations. Also, an increase in the number of angles increases the
The developed method was used for a sensitivity pulling force. Smaller angles also result in larger strains
study of the pulling force, stress and strain predictions due to bending deformations.
to different parameters influencing mechanism of HDD. Table 5 shows the influence of Young’s modulus and
The conditions for pull 2 were used as a base and the borepath vs. pipe size. When the calculations were
different parameters were varied, one at a time. Table 4 done using a smaller Young’s modulus (300 MPa), there
presents calculations for different borepath profiles. The was an increase in the predicted axial strains. However,
analysed profiles are shown in Fig. 8. The model the total pulling force decreased slightly (from 8.04 to
predicted the increase in the pulling forces when the 7.72 kN). This shows that rational modelling for poly-
M.A. Polak, A. Lasheen / Tunnelling and Underground Space Technology 16 Suppl. 1 (2002) S47–S55 S55

ethylene pipe behaviour should include viscous behav- Acknowledgments


iour—a factor that was not included in these
calculations. An interesting result was also obtained Research presented in this paper was supported by
when the Young’s modulus was given a value of ESTAC (Environmental Science and Technology Alli-
200 000.0 MPa equivalent to steel stiffness (the dimen- ance Canada) and NSERC (Natural Sciences and Engi-
sions of the pipe were however kept the same as for PE neering Research Council of Canada). The authors wish
pipe). The pulling force increased from 8 kN (for PE to express their gratitude for the received support.
pipe) to 164 kN. Bending strains did not change with
the change of the Young’s modulus because they are References
dependent on the borehole geometry and not the pipe
stiffness. Abramowitz, M., Stequn, I.A., 1965. Handbook of Mathematical
Also shown in Table 5 are the results for different Functions With Formulas, Graphs and Mathematical Tables. Dover
borehole dimensions. The smaller the borehole with Publications, New York, pp. 589–618.
respect to the pipe that is pulled through it, the larger Baumert, M.E., Allouche, E.N., 2000. Installation Loads on Pipelines
Installed Using HDD. CSCE Conference, London, Ontario, pp.
the pulling force and the corresponding axial strains. 173–180.
Bending strains also increase for the smaller borehole Drillpath娃, 1996. Theory and User’s Manual. Intrasoft, L.L.C,
size. Houston, Texas, USA.
Driscopipe䉸, 1993. Technical Expertise Application of Driscopipe䉸
in Directional Drilling and River Crossings, Technical Note 噛41.
7. Conclusions
Frisch-Fay, R., 1962. Flexible Bars. Butterworths, London, UK, pp.
73–80.
A method was presented that allows for rational Gelinas, M.M., 1998. Field Monitoring and Analysis of High Density
calculations of pulling forces and deformations in the Polyethylene Pipe Installed Using Horizontal Directional Drilling.
pipes installed using HDD. The method considers all MASc. Thesis. University of Waterloo, Waterloo, Canada.
Gelinas, M., Polak, M.A., McKim, R., 2000. Field tests on high
components of the mechanics of HDD installations. density polyethylene pipes installed using horizontal directional
The proposed method was used for the analysis of drilling. ASCE J. Infrastruct. Syst. 6, 130–137.
two field tested pipes and for the analysis of HDD Ghokale, S., Hamm, R., Sterling, R., 1999. A comprehensive survey
installations where different parameters were varied. on the state of horizontal directional drilling in North America
Predicted axial strains were smaller than the field meas- provides an inside look at this increasingly growing industry. Dir.
Drilling 7, 20–23.
ured strains. Some of the reasons for this are simplified
Huey, D.P., Hair, J.D., McLeod, K.B., 1996. Installation Loading and
assumptions regarding borepath profile and material Stress Analysis Involved with Pipelines Installed in Horizontal
behaviour of polyethylene. Changing the borepath pro- Directional Drilling. No-Dig, New Orleans, USA, pp. 37–60.
file and including more corner points increases substan- Khan, S., Bennett, D., McCrary, S., Iseley, T., 1994. Mini-Horizontal
tially the predicted strains. Similarly, decreasing pipe Directional Drilling: State-of-the-Art Review. Trenchless Technol-
ogy Center at Louisiana Tech University, Ruston, Louisiana.
stiffness increases the predicted axial strains. It should
Kirby, M.J., Kramer, S.R., 1996. Design guidelines and procedures
also be noted that the analysed sample (two tests) is for guided horizontal drilling, Part II. No-Dig Eng. 3, 13–15.
small and more test comparisons should be done in Lasheen, A., Polak, M.A., 2001. HDD Loads Imposed On A Pipe.
future. However, the comparison between theoretical No-Dig 2001, Nashville.
and experimental results shows reasonable agreement Lasheen, A., Polak, M.A., 2001. Predicting the Behaviour of HDPE
and good prospects for further research on improving Pipes in Horizontal Directional Drilling. URI Conference, Kitche-
ner, Ontario.
the models that form the presented predictive formula-
Schlichting, H., Kestin, J., 1968. Boundary Layer Theory. McGraw
tion. The proposed method is rational, based on the Hill, NY.
principles of mechanics, and at the same time simple, Streeter, V.L, Wyle, B.E., 1975. Fluid Mechanics. McGraw Hill, , pp.
robust and efficient in predicting pipe behaviour. 197–201.

You might also like