HDD Collapse Case Study

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/276408268
Study of ground collapse induced by large diameter horizontal directional

drilling in sand layer using numerical modeling
Article in Canadian Geotechnical Journal · February 2015

DOI: 10.1139/cgj-2014-0388
CITATIONS READS
22 5,390
2 authors:
Biao Shu Baosong Ma

Central South University Sun Yat-Sen University
29 PUBLICATIONS 642 CITATIONS 85 PUBLICATIONS 864 CITATIONS
SEE PROFILE SEE PROFILE
All content following this page was uploaded by Biao Shu on 06 February 2020.
The user has requested enhancement of the downloaded file.

1562
ARTICLE
Study of ground collapse induced by large-diameter horizontal
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by DIRECTORATE OF COLDWATER FISHERIES RES on 10/01/15
directional drilling in a sand layer using numerical modeling

Biao Shu and Baosong Ma
Abstract: With large-diameter horizontal directional drilling (HDD) becoming the preferred method to construct oil and gas
pipelines and utility pipelines beneath rivers, the issue of potential ground collapse arises when drilling in loose geological
layers such as sand. Ground collapse is a result of borehole collapse and may cause significant damage to the topography and
nearby facilities. The present investigation considered the potential causes of ground collapse induced by an actual 1.219 m
diameter HDD river-crossing project, using the FLAC3D numerical modeling tool. The analysis showed that the failure zone first
developed at the crown of the borehole, resulting in subsequent borehole collapse due to instability of the sand above, and
eventually leading to ground collapse. Sequential reaming cycles have been simulated and the results indicate very little effect
in comparison with a single reaming cycle. The risk of borehole collapse, and consequent ground collapse, increases with
borehole diameter. Parametric numerical modelling has also been conducted to study the influence of soil parameters and
drilling mud pressure on the stability of the ground surface above the borehole. The results show that soil cohesion and friction
angle have a large influence on stability of the borehole and ground surface, while elastic modulus and Poisson’s ratio have
relatively little effect. It was also determined that mud pressure is a very important factor in maintaining stability of the
borehole, and therefore the ground surface as well.
Key words: ground collapse, horizontal directional drilling, numerical modeling, borehole stability, river crossing.
For personal use only.
Résumé : Le forage directionnel horizontal (FDH) de grand diamètre devenant la méthode privilégiée de construction des oléoducs,
gazoducs et canalisations passant sous les rivières, le problème du risque d’effondrement du sol se pose lorsqu’on perce de tels
conduits dans des couches géologiques meubles, telles que le sable. L’effondrement du sol résulte de l’effondrement de trous de forage
et peut gravement perturber la topographie et endommager les infrastructures avoisinantes. La présente étude s’est intéressée, au
moyen du logiciel de modélisation numérique FLAC3D, aux causes possibles d’un effondrement du sol provoqué par un projet existant
de FDH de 1,219 m de diamètre à travers une rivière. L’analyse effectuée a montré que la zone de rupture se développait d’abord sur
le pourtour du trou de forage, ce qui entraînait l’effondrement de ce dernier, en raison de l’instabilité de la couche de sable située au
dessus, et finalement l’effondrement du sol. Des cycles consécutifs d’alésage ont été simulés et les résultats ont montré que leur effet
était très limité en comparaison avec celui d’un cycle unique d’alésage. Le risque d’effondrement du trou de forage, et de
l’effondrement du sol qui en résulte, augmente avec le diamètre du trou de forage. Une modélisation numérique paramétrique
a été réalisée pour étudier l’influence des paramètres du sol et de la pression de la boue de forage sur la stabilité de la surface du
sol situé au-dessus du trou de forage. Les résultats révèlent que la cohésion du sol et l’angle de frottement influaient beaucoup
sur la stabilité du trou de forage et de la surface du sol, tandis que l’effet du module d’élasticité et du coefficient de Poisson était
relativement faible. On a également déterminé que la pression de la boue constitue un facteur important de maintien de la
stabilité du trou de forage et donc aussi de la surface du sol. [Traduit par la Rédaction]
Mots-clés : effondrement du sol, forage directionnel horizontal, modélisation numérique, stabilité du trou de forage, forage à
travers une rivière.
Introduction reaming, and pulling back the pipeline (ASTM 2011). In the first
stage, a small-diameter drill bit is used to create a pilot hole from
Horizontal directional drilling (HDD) is a technique derived
the starting point to the end point. In the second stage, the pilot
from the oil and gas drilling industry for the purpose of installing
hole is enlarged to the required diameter, usually 1.2–1.5 times
various utility lines, such as oil and gas pipelines, under rivers,
that of the pipe to be installed. In the final stage, the product pipe
streets, and mountains (Allouche et al. 2000). Initially it was used is connected behind a reamer and pulled back from the end point
primarily to install small-diameter pipelines under streets or to the starting point (Najafi 2004; Ma 2008). The use of HDD has
highways to avoid road excavation, therefore reducing disrup- greatly facilitated the construction of oil and natural gas pipe-
tions to traffic and businesses. As the technique developed, dril- lines, especially pipelines that must be installed beneath rivers
ling machine manufacturers built larger and more sophisticated and lakes (Iseley and Gokhale 1997; Willoughby 2005).
horizontal drilling rigs with the capability of installing very large- The growing deployment of maxi HDD has been supported by
diameter steel and high-density polyethylene (HDPE) pipes for an increasing amount of theoretical research. The studies have
long distances, including beneath wide rivers. The HDD proce- concentrated on the mud pressure (Baumert et al. 2005; Ariaratnam
dure typically includes three stages: drilling the pilot hole, back- et al. 2007), HDD pull loads (Cheng and Polak 2007; Slavin 2010;
Received 9 September 2014. Accepted 10 February 2015.

B. Shu. School of Geosciences and Info-Physics, Central South University, Changsha, Hunan 410083, P.R. China.
B. Ma. School of Engineering, China University of Geosciences, Wuhan, Hubei, 430074, P.R. China.
Corresponding author: Biao Shu (e-mail: [email protected]).
Can. Geotech. J. 52: 1562–1574 (2015) dx.doi.org/10.1139/cgj-2014-0388 Published at www.nrcresearchpress.com/cgj on 26 February 2015.
Shu and Ma 1563
Fig. 1. Location of the HDD river crossing (modified from Google Earth).
Slavin et al. 2011; Slavin and Najafi 2012), ground surface heave rivers or lakes, the borehole is often drilled in soft and loose layers
(Lueke and Ariaratnam 2006), reamer rotational resistance (Lan like clay, sand, and sometimes gravel, with the accompanying
et al. 2011), cuttings transport mechanism (Shu et al. 2014), and risk.
pipeline safety (Lan et al. 2014). There have also been investiga- The Kirsch (1898) solution provides the stress distribution around a
tions of borehole stability using finite element methods (Wang circular hole in an infinite elastic, homogeneous, and isotropic
and Sterling 2007; Zhou et al. 2012), as well as of borehole wall material. Unfortunately, Kirsch’s solution is based on the assump-
displacement using numerical modeling with FLAC3D (Sun et al. tion that the circular opening is in an infinite plate, and is there-
2011, 2012). Conroy et al. (2002) studied how an HDD river crossing fore more applicable to a deep tunnel or borehole (e.g., 300 m
affects the safety of levees, especially the effect of hydrafracture. deep). As HDD boreholes for river crossing projects are relatively
There has also been research concerning horizontal borehole shallow, at a depth of only 10⬃40 m, Kirsch’s solution may not be
stability in the mining and petroleum industries (Ewy 1993; reliably used for such applications, and therefore numerical mod-
Mohiuddin et al. 2007; Zare-Reisabadi et al. 2012; Hungerford and elling may be a better choice.
Ren 2014). However, these investigations of borehole stability fo- The main aim of the present study is to better understand the
cused primarily on small-diameter boreholes at shallow depth, reason, and mechanism, for ground collapse caused by large-
and did not consider the potentially large ground collapse caused diameter HDD river crossing projects in sandy geological condi-
by HDD. tions, using FLAC3D software. It is also desired to determine the
relative importance of various soil parameters, borehole depth,
Ground collapse induced by HDD is a result of borehole shrink-
mud pressure, and the effect of sequential reaming (i.e., pre-
age or collapse. The consequences and damage caused by bore-
reaming) cycles on the stability of the borehole and ground sur-
hole collapse and ground collapse can be summarized as follows:
face. To help achieve this objective, the detailed project information
1. Relatively minor borehole failure, such as a small degree of and geological conditions of a particular large-diameter HDD proj-
borehole collapse and shrinkage, can increase the frictional ect were introduced into the FLAC3D numerical model, and the
drag force between the borehole and the reamer and (or) prod- results analyzed to determine the reason for ground collapse. The
uct pipe, increasing the required pulling force. A major col- relative importance of individual parameters was determined by
lapse can prevent the passage of the reamer and (or) product changing the value of one parameter while keeping the others
pipe, which then become stuck in the borehole and can lead to fixed.
failure of the entire HDD crossing project.
2. Borehole collapse can also lead to collapse at the surface and
Project and geological conditions
consequent damage to other facilities on the ground. For ex- A steel natural gas pipe with a diameter of 1.219 m (48 in.) was
ample, when the borehole crosses beneath a river, government designed and installed under a river in Wuhan, China, using hor-
agencies are extremely concerned about potential damage to izontal directional drilling. The borehole diameter is usually de-
levees (Conroy et al. 2002). signed to be 1.2⬃1.5 times the pipeline diameter, corresponding to
a final borehole diameter of 1.575 m (62 in.) in this case. The
Borehole collapse and subsequent ground collapse can occur in location and route of the borehole are shown in Fig. 1, and Fig. 2
many different geological conditions, such as soft clay, loose sand, indicates the length, depth, entry angle, exit angle, etc. The bore-
loose gravel, and highly fractured rocks. The mechanism of borehole is approximately 25⬃29 m deep; the distance between the
hole collapse and ground collapse varies for these different geo- entry point in the northeast direction and the levee is 120 m; the
logical conditions. When large-diameter HDD is used to cross distance between the exit point in the southwest direction and
Published by NRC Research Press

1564 Can. Geotech. J. Vol. 52, 2015
Fig. 2. Cross section of the borehole path (note: length and angle are not to scale).
the levee on the opposite side is 225 m; the entry angle is 9.5°; and with the above discussion indicating there was no evidence of
the exit angle is 8°. The total borehole length is approximately deformation or collapse above the region in which the HDD dril-
945 m. ling was located in this layer.
Collapse description E
Three collapse pits (#1, #2, and #3) appeared approximately (1) K⫽
3(1 ⫺ 2␷)
50–60 m from the exit point during the HDD installation; see
Figs. 1 and 3. Collapse pit #1 appeared initially; after approxi- E
(2) G⫽
mately one-half month pit #2 appeared, followed by pit #3 an- 2(1 ⫹ ␷)
other one-half month later. At the time of the study, pit #3 had
only recently appeared and was not fully developed. It is antici-
where E is the elastic modulus and ␷ is the Poisson’s ratio.
pated that this pit has subsequently developed similar to that of
pits #1 or #2, as seen in Fig. 3. The diameters of collapse pits #1, #2, Numerical modelling of the collapse
and #3 were approximately 3, 5, and 1.5 m, respectively. The ap-
pearance of this ground collapse raised a serious concern about FLAC3D is a three-dimensional explicit finite difference pro-
the safety of the levee, and whether the collapse will expand gram for engineering mechanics computations, with the analysis
towards the levee. capability of simulating the behavior of three-dimensional struc-

As it is well recognized that a large-diameter borehole may tures built of soil, rock or other materials that may undergo plas-
cause borehole collapse and subsequent ground collapse in loose tic flow when their yield limits are reached (Itasca 2012).
sand layers, a casing along the entire sand layer in the vicinity of
Numerical model
the exit point was planned to be installed. The casing is a steel
tube that is slightly larger than the final borehole diameter, and is As shown in Figs. 5 and 6, the ground collapse occurred between
sometimes used in drilling projects to insulate the borehole from 50–60 m from the exit point, in the fine sand layer, just beyond
the surrounding weak or loose soils or rocks to prevent borehole the protection of the casing. The casing was successful in prevent-
collapse and drilling fluid loss. However, during the casing instal- ing ground collapse in the area above in the first 50 m away from
lation, high friction resistance was encountered when the casing the exit point. At the same time, once the borehole enters the
was hammered into the ground, allowing only 50 m to be in- sandstone layer, it also becomes very stable because of the high
stalled. Figure 4 shows the casing installed at the exit point. The strength of sandstone. The analysis therefore is focused on the
casing was not installed at the entry point because of relatively 50 m “weak section” as shown in Fig. 6, beyond the casing, where
stable geological conditions. Figure 5 shows the locations of the ground collapse did occur. The resulting FLAC3D model is shown
casing and the ground collapse, indicating the ground collapse in Fig. 8, based on the above analysis of the geological conditions,
occurred just beyond the protection of the casing, 50–60 m from and is 50 m long, 24 m wide, and 16 m high.
the exit point. Due to the stability of the sandstone layer below the fine sand,
the top of the sandstone layer was established as the bottom
Site geology boundary in the numerical model with respect to vertical dis-
The geology of the site is a typical alluvial plain geology of the placements. Horizontal displacements, however, in both direc-
Yangtze River, comprising loose silty clay, fine sand, and sand- tions, are allowed by using roller boundary conditions. For the
stone, from above to below, as shown in Fig. 6, as determined by four side walls of the model, roller boundary conditions are also
the preliminary geological subsurface investigation. The top silty assumed in both the vertical and lateral directions, while the
clay layer is very thin (0.5 m thick), over a 15.5 m thick fine sand displacements are restrained in the perpendicular directions. The
layer, below which is the sandstone layer. Figure 7 clearly shows top of the model is a free surface. The Mohr–Coulomb constitutive
the thin layer of silty clay above the fine sand layer. Due to the model is used for silty clay and fine sand and their properties used
relatively high strength of the sandstone, the borehole in this in the numerical modeling are consistent with those in Table 1.
layer was quite stable during the HDD construction, and there is
no evidence of ground collapse above this portion. Therefore, Numerical modeling and results
there is only a single short (50 m) “weak” section beyond the The initial state of the model under gravity, before the borehole
protective casing, as shown in Fig. 6, within which the ground was drilled, was obtained initially. The borehole was then created,
collapse occurred. and gradient mud pressure was applied on the internal wall of the
The values of the elastic modulus, Poisson’s ratio, cohesion, borehole. In drilling engineering, a type of clay (e.g., bentonite) is
friction angle, and density of silty clay, fine sand, and sandstone typically used to seal the borehole (Selvadurai 2002; Pusch 2011).
are provided in Table 1. The values of the bulk modulus (K) and Thus as the drilling mud is pumped into the borehole, a thin mud
shear modulus (G) may be calculated by eqs. (1) and (2), and are also (clay) cake gradually accumulates, becoming tightly attached to
included in Table 1. It can be seen from Table 1 that the elastic the borehole wall. It is therefore assumed that the borehole is full
modulus of sandstone, even though highly medium-weathered, is of drilling mud up to the ground surface and no mud leaks into
about 100 times that of the fine sand. This high value is consistent the surrounding soil, otherwise possibly affecting the physical

Shu and Ma 1565
Fig. 3. Ground collapse caused by HDD.

Fig. 4. Casing installed to prevent ground collapse at the exit point.

Fig. 5. Locations of casing and ground collapse.
characteristics of the sand. The drilling mud pressure is deter- where ␳m is the drilling mud density (typically 1050⬃1200 kg/m3),
mined by eq. (3); i.e., the deeper the borehole, the higher the mud assumed to be 1100 kg/m3 in this study; g is gravitational acceler-
pressure. ation, 9.8 N/kg; hm is the depth of the borehole centerline at the
location of interest to the drilling mud surface. For example, the
(3) p ⫽ ␳m ghm drilling mud pressure at point y = 50 m, where hm = 7.03 m, is

Fig. 6. Geological cross section.

Fig. 7. Outcrop of silty clay and fine sand on site.

Table 1. Physical and mechanical properties of the site soils and trating the development of the plastic zones at varying depths.
rocks. The results show that the size of the failure zone increases as the
Silty clay Fine sand Sandstone depth decreases. Thus, collapse is more likely to occur above the
shallow borehole sections, in excellent agreement with the on-
Elastic modulus, E (MPa) 8.8 20.7 2100 site observations, as indicated in Fig. 6.
Poisson’s ratio, ␷ 0.26 0.22 0.20
The failure zones at the bottom of the borehole shown in Fig. 10
Cohesion, C (kPa) 10 3 —
Friction angle, ␸ (°) 29 35 —
are, in general, due to a combination of the ground rebound and
Tensile strength (MPa) 0 0 — mud pressure after the soil in the borehole is removed. Small
Density, ␳ (kg/m3) 1750 1920 2360 failure zones observed at the bottom of the borehole at cross
Bulk modulus, K (MPa) 6.11 12.32 1166.67 sections y = 0, 10, and 20 m are due to the very thin soil layer under
Shear modulus, G (MPa) 3.49 8.48 875.00 the borehole at those locations, corresponding to a small rebound
effect. It shows that the failure zone above the borehole is small at
locations where the borehole is deep, which may be due to the
equal to 75 783.4 Pa; and at point y = 0 m, where hm = 14.05 m, is higher mud pressure and soil arching effect. There are large shear
equal to 151 459 Pa. failure zones, as well as tension failure zones on the ground sur-
Figure 9 shows the resulting failure (plastic) zone of the model, face, at cross sections y = 40 and 50 m, corresponding to the
as indicated by the different colors designating failure due to locations of the on-site ground collapse. It can also be seen that
shear or tension. In all of the failure zone plots in this paper (Web most of the soil above the borehole has not failed. However, once
version only), green indicates no failure, red indicates shear fail- the sand above the borehole fails and collapses, the sand above
ure zone, and blue indicates tension failure zone. Cross sections at will collapse continuously and eventually lead to ground collapse.
different locations along the borehole are shown in Fig. 10 illus- Also, following initial borehole collapse at y = 50 m, the area of

Shu and Ma 1567
Fig. 8. FLAC3D model.

Fig. 9. Failure zone.

collapse can expand along the borehole route to y = 48 m, and final diameter of D7 = 1.575 m (radius = 0.788 m), as indicated in
then y = 44 m, etc., consistent with the sequential appearance of Fig. 12, instead of the previously simulated single reaming opera-
ground collapse at points #1, #2, and #3, as discussed previously. tion.
As described in the previous analysis, the model was brought to
Sequential reaming initial equilibrium, and then the soil in the pilot hole removed
It is recognized that the final 1.575 m (62 in.) diameter of the
and the same mud pressure applied on the borehole wall. In a
borehole is achieved by sequentially using a series of reamers, of
similar manner, additional soil was sequentially removed, corre-
various diameters, to expand the initial small-diameter pilot hole.
sponding to reamer diameters ranging from D1 to D7. Figures 13
For example, the diameter of the pilot bit is 0.168 m (6.625 in.), and
the diameters of the sequence of reamers are 0.584 m (23 in.), and 14 show the failure zones after each step at y = 30 and 50 m.
0.762 m (30 in.), 0.914 m (36 in.), 1.067 m (42 in.), 1.270 m (50 in.), These figures confirm that when the borehole diameter is small,
1.422 m (56 in.), and 1.575 m (62 in.). Figure 11 shows some of there are only very limited failure zones, which would not be
different-sized reamers used in the actual HDD project to enlarge expected to cause borehole collapse and associated ground col-
the borehole. lapse. However, with increasing borehole diameter, the failure
Therefore, the sequential reaming process of how the borehole zone dramatically expands, leading to instability and ground col-
was drilled and enlarged was also simulated. The basic model in lapse. Furthermore, consistent with the numerical results based
Fig. 8 was again used, except that the borehole was gradually on the single boring operation, the failure zone at cross sections
enlarged from the initial borehole diameter D0 = 0.168 m to the y = 30 m is smaller than that of cross section y = 50 m; i.e., the more

Fig. 10. Failure zone at different cross sections.

Fig. 11. Different sizes of reamers for soft geological conditions.

Fig. 12. Model for sequential reaming.
shallow the borehole, the greater the instability and potential sequential reaming operations in Fig. 15. While there is essentially
collapse. In general, these results confirm that borehole and no difference in the more significant final failure zones at cross
ground collapse are more likely to occur due to a shallow, large- sections y = 40 and 50 m, those at cross sections y = 0 and 20 m are
diameter borehole. more extensive for the more realistic case of several sequential ream-
The numerical modeling results with the large borehole created ing operations. Nonetheless, for the purpose of studying the stability
in a single boring operation are compared with that based on the of borehole and ground collapse, in the vicinity of a relatively shal-

Shu and Ma 1569
Fig. 13. Failure zone under sequential reaming steps at y = 30 m.

Fig. 14. Shear failure zone under sequential reaming steps at y = 50 m.

Fig. 15. Comparison of final failure status of two different assumed boring procedures: (a) single drilling; (b) sequential reaming.
low borehole, it is sufficient to only simulate the borehole as if it gle, which are strength parameters. By changing only one param-
was created in a single drilling operation. eter at one time, while keeping the rest fixed, the influence of
each parameter on borehole and ground stability has been stud-
Further parametric study ied. As shown in Fig. 8, because the silty clay is only 0.5 m thick
The results presented above are based on the physical charac- and, in the area of interest, the borehole was drilled mainly in the
teristics provided in Table 1. To determine the sensitivity of the fine sand layer, only the parameters of the fine sand were varied.
results on the specific quantitative values, and to help understand Figure 16 shows that the failure zone was not significantly af-
the relative importance of the various parameters, a study has fected as the modulus of fine sand varied from 15 to 30 MPa,
been conducted using varying values for the soil characteristics including the initial value of 20.7 MPa. Figure 17 similarly shows
and drilling mud pressure. the relatively small effect of the Poisson’s ratio. In comparison,
cohesion is a very important factor, with higher values signifi-
Soil parameters cantly increasing the stability of the borehole and ground surface,
The soil parameters include elastic modulus and Poisson’s ratio, and vice versa, as indicated in Fig. 18. Similarly, a greater friction
which are deformation properties; and cohesion and friction an- angle has a stabilizing effect similar to that of cohesion; see Fig. 19.

Fig. 16. Failure zone under different elastic moduli.

Fig. 17. Failure zone under different Poisson’s ratios.
Mud pressure fracturing soil and rock; operation of mud motor (if used); cooling
The functions of drilling mud include suspension and removal down hole tooling and electronics; and, as directly relevant to the
of cuttings from the borehole; lubrication for drill bit, drill rod, present investigation, maintaining borehole stability, especially
and product pipe; hydraulic jetting for assisting in cutting and for loose and soft soils. Ideally, the mud will eventually solidify in

Shu and Ma 1571
Fig. 18. Failure zone under different cohesion values.

the annular space between the pipe and borehole, providing per- investigation included a parametric study to determine the effect
manent support. High mud pressure may cause ground heave in of various soil parameters and mud pressure on the borehole and
soft soils (Lueke and Ariaratnam 2006). However, when drilling a ground stability. The simulation considered the effect of creating
large-diameter horizontal borehole in sand, the greater concern is the borehole in a single step in comparison with the more realistic
providing sufficient mud pressure to maintain borehole stability. procedure utilizing several intermediate sizes of reamers. The
Different mud pressures were therefore applied on the borehole results indicate that the latter procedure, in which the borehole is
wall to determine the effect on borehole and ground stability. gradually enlarged by sequential reaming, results in somewhat
It is assumed that part of the borehole should drain its fluid, less stability, but this small effect may be ignored at shallow
such that there is an absence of mud pressure on the borehole depths corresponding to the area of interest, including the loca-
wall or that a portion will collapse immediately, leading to even- tion of ground collapse.
tual collapse of the ground surface above. Figure 20 shows the The parametric study clearly showed that larger boreholes re-
failure zone of a locally drained borehole and Fig. 21 shows the sult in more extensive failure zones and associated borehole col-
results for a totally drained borehole (no mud pressure applied on lapse and ground collapse. The numerical modeling and analysis
the borehole wall). Therefore, it is always important to keep the confirms that soils with greater cohesion and friction angle char-
borehole full of drilling mud. The loss of drilling fluid may arise acteristics significantly improve the stability of the borehole and
due to use of inappropriate drilling fluid for the soil (sandy layer), ground above. In comparison, the stability is relatively insensitive
not providing the required sealing function at the borehole wall. to reasonable variations in the elastic modulus or Poisson’s ratio.
However, a more difficult situation to prevent is that of a nonlevel The mud pressure helps support the borehole, thereby also pre-
grade, with entry and exit points at different elevations, leading to venting ground collapse. The absence of mud pressure, such as
a natural tendency for the fluid to drain to the lower of these may be caused by inadvertent (or, in some cases, possibly unavoid-
openings (Slavin et al. 2011). In such situations, it may be advisable able) drainage of the drilling fluid, can lead to collapse of an
to fill the voids with a fast-setting grout, as soon as practical. inherently unstable soil, such as fine sand.
According to the study, the potential reasons that caused the
Discussion and conclusions collapses should be the low cohesion of fine sand, borehole depth,
The cause of ground collapse followed by a large-diameter hor- and possible lack of enough mud pressure support. Because of the
izontal directional drilling crossing beneath a river in Wuhan, low cohesion of the sand layer that was studied in the geotechni-
China, in fine sand was investigated by numerical modeling. The cal investigation, it was planned to install a casing tube to isolate

Fig. 19. Failure zone under different friction angles.

Fig. 20. Failure zone for a locally drained borehole.

Shu and Ma 1573
Fig. 21. Failure zone of a totally drained borehole.

the loose sand layer to prevent ground collapse before the HDD Iseley, D.T., and Gokhale, S.B. 1997. Synthesis of highway practice 242: trench-
construction. However, such casing installation was not fully less installation of conduits beneath roadways. Transportation Research
Board, National Academy Press, Washington, D.C.
completed because of the high frictional resistance. The numeri- Itasca. 2012. FLAC3D: Fast Lagrangian Analysis of Continua in 3 Dimensions.
cal modeling results show that the large failure zone can occur in User’s guide. Itasca Consulting Group Inc., Minneapolis, Minn.
the shallow borehole part of the weak zone while keeping stable Kirsch, G. 1898. Die theorie der Elastizitat und die Bedurfnisse der Festigkeit-
in the deep borehole parts, which is consistent with the real situ- slehre. Zeitschrift des Vereines Deutscher Ingenieure, 42(28): 797–807.
ation. Lan, H., Ma, B., Shu, B., and Wu, Z. 2011. Prediction of rotational torque and
design of reaming program using horizontal directional drilling in rock
The results of this study are primarily applicable to a borehole strata. Tunnelling and Underground Space Technology, 26(2): 415–421. doi:
drilled in sand, using HDD. Other geological materials, such as 10.1016/j.tust.2010.11.014.
silty clay or clay, as well as gravels or highly fractured rocks, have Lan, H., Ma, B., Zhang, Y., and Shu, B. 2014. Buried pipe affected by river erosion
significantly different physical properties and require additional when crossing the Yangtze River. Journal of Pipeline Systems Engineering
investigation. and Practice. doi:10.1061/(ASCE)PS.1949-1204.0000128.
Lueke, J.S., and Ariaratnam, S.T. 2006. Numerical characterization of surface
heave associated with horizontal directional drilling. Tunnelling and Under-
Acknowledgements ground Space Technology, 21(1): 106–117. doi:10.1016/j.tust.2005.06.001.
The authors would like to thank China Petroleum Pipeline Bu- Ma, B. 2008. The science of trenchless engineering. China Communications
reau for providing the opportunity to conduct this research on Press, Beijing.
site. The authors also wish to thank Larry Slavin for his detailed Mohiuddin, M.A., Khan, K., Abdulraheem, A., Al-Majed, A., and Awal, M.R. 2007.
Analysis of wellbore instability in vertical, directional, and horizontal wells
review of this paper and many valuable suggestions.
using field data. Journal of Petroleum Science and Engineering, 55(1–2): 83–
92. doi:10.1016/j.petrol.2006.04.021.
References Najafi, M. 2004. Trenchless technology-pipeline and utility construction and
Allouche, E.N., Ariaratnam, S.T., and Lueke, J.S. 2000. Horizontal directional renewal. 1st ed. McGraw-Hill Company Inc., New York.
drilling: profile of an emerging industry. Journal of Construction Engineer- Pusch, R. 2011. A technique to delay hydration and maturation of borehole seals
ing and Management, 126(1): 68–76. doi:10.1061/(ASCE)0733-9364(2000)126: of expansive clay. Engineering Geology, 121(1–2): 1–6. doi:10.1016/j.enggeo.
1(68). 2011.04.017.
Ariaratnam, S.T., Harbin, B.C., and Stauber, R.L. 2007. Modeling of annular fluid Selvadurai, A.P.S. 2002. Influence of pressurized water influx on the hydro-
pressures in horizontal boring. Tunnelling and Underground Space Technol- thermal behavior of an engineered clay barrier in a waste emplacement
ogy, 22: 610–619. doi:10.1016/j.tust.2007.01.001. borehole. Engineering Geology, 64(2–3): 157–178. doi:10.1016/S0013-7952(01)
ASTM. 2011. ASTM F 1962-11 standard guide for use of maxi-horizontal directional 00105-3.
drilling for placement of polyethylene pipe or conduit under obstacles, in- Shu, B., Ma, B., and Lan, H. 2014. Cuttings transport mechanism in a large-
cluding river crossings. West Conshohocken, Pa. doi:10.1520/F1962-11. diameter HDD borehole. Journal of Pipeline Systems Engineering and Prac-
Baumert, M.E., Allouche, E.N., and Moore, I.D. 2005. Drilling fluid considerations
tice. doi:10.1061/(ASCE)PS.1949-1204.0000190.
in design of engineered horizontal directional drilling installations. Interna-
Slavin, L.M. 2010. Parametric dependency and trends of HDD pull loads. Journal
tional Journal of Geomechanics, 5(4): 339–349. doi:10.1061/(ASCE)1532-3641
of Pipeline Systems Engineering and Practice, 1(2): 69–76. doi:10.1061/(ASCE)
(2005)5:4(339).
PS.1949-1204.0000045.
Cheng, E., and Polak, M.A. 2007. Theoretical model for calculating pulling loads
for pipes in horizontal directional drilling. Tunnelling and Underground Slavin, L.M., and Najafi, M. 2012. Effect of pipe stiffness on maxi-HDD pull loads.
Space Technology, 22(5–6): 633–643. doi:10.1016/j.tust.2007.05.009. Journal of Pipeline Systems Engineering and Practice, 3(1): 22–31. doi:10.1061/
Conroy, P.J., Latorre, C.A., and Wakeley, L.D. 2002. Installation of fiber-optic (ASCE)PS.1949-1204.0000092.
cables under flood-protection structures using horizontal directional drilling Slavin, L.M., Najafi, M., and Skonberg, E.R. 2011. Maxi-HDD pull loads for non-
techniques. US Army Corps of Engineers. level grade for polyethylene pipe. Journal of Pipeline Systems Engineering
Ewy, R.T. 1993. Yield and closure of directional and horizontal wells. Interna- and Practice, 2(2): 64–69. doi:10.1061/(ASCE)PS.1949-1204.0000075.
tional Journal of Rock Mechanics and Mining Sciences & Geomechanics Ab- Sun, P., An, C., and Cao, H. 2011. Horizontal borehole displacement analysis
stracts, 30(7): 1061–1067. doi:10.1016/0148-9062(93)90072-L. based on drilling fluid pressure. Chinese Journal of Underground Space and
Hungerford, F., and Ren, T. 2014. Directional drilling in unstable environments. Engineering, 2011(7): 1168–1173.
International Journal of Mining Science and Technology, 24(3): 397–402. Sun, P., Cao, H., Zuo, L., and Wang, Z. 2012. Ground surface movement simula-
doi:10.1016/j.ijmst.2014.03.019. tion due to HDD based on soil parameters. In Proceedings of the International

Conference on Pipelines and Trenchless Technology (ICPTT), 19-22 October Zare-Reisabadi, M.R., Kaffash, A., and Shadizadeh, S.R. 2012. Determination of
2012, Wuhan, China. pp. 1977–1988. doi:10.1061/9780784412619.198. optimal well trajectory during drilling and production based on borehole
Wang, X., and Sterling, R.L. 2007. Stability analysis of a borehole wall during stability. International Journal of Rock Mechanics and Mining Sciences, 56:
horizontal directional drilling. Tunnelling and Underground Space Technol- 77–87. doi:10.1016/j.ijrmms.2012.07.018.
ogy, 22: 620–632. doi:10.1016/j.tust.2007.01.002. Zhou, P., Jiao, R., Ma, B., Liu, H., and Han, Y. 2012. Analysis of borehole wall
Willoughby, A.D. 2005. Horizontal directional drilling utility and pipeline appli- stability in horizontal directional drilling. In Proceedings of the ICPTT 2012.
cations. McGraw-Hill, New York. pp. 1816–1825. doi:10.1061/9780784412619.183..
View publication stats

HDD Collapse Case Study

Uploaded by

Copyright:

Available Formats

HDD Collapse Case Study

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

HDD Collapse Case Study

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Study of ground collapse induced by large diameter horizontal directional

Article in Canadian Geotechnical Journal · February 2015

Biao Shu Baosong Ma

SEE PROFILE SEE PROFILE

The user has requested enhancement of the downloaded file.

directional drilling in a sand layer using numerical modeling

Received 9 September 2014. Accepted 10 February 2015.

Published by NRC Research Press

towards the levee. capability of simulating the behavior of three-dimensional struc-

Published by NRC Research Press

Fig. 3. Ground collapse caused by HDD.

Fig. 4. Casing installed to prevent ground collapse at the exit point.

Fig. 5. Locations of casing and ground collapse.

Published by NRC Research Press

Fig. 6. Geological cross section.

Fig. 7. Outcrop of silty clay and ﬁne sand on site.

Published by NRC Research Press

Fig. 8. FLAC3D model.

Fig. 9. Failure zone.

Published by NRC Research Press

Fig. 10. Failure zone at different cross sections.

Fig. 11. Different sizes of reamers for soft geological conditions.

Fig. 12. Model for sequential reaming.

Published by NRC Research Press

Fig. 13. Failure zone under sequential reaming steps at y = 30 m.

Fig. 14. Shear failure zone under sequential reaming steps at y = 50 m.

Published by NRC Research Press

Fig. 16. Failure zone under different elastic moduli.

Fig. 17. Failure zone under different Poisson’s ratios.

Published by NRC Research Press

Fig. 18. Failure zone under different cohesion values.

Published by NRC Research Press

Fig. 19. Failure zone under different friction angles.

Fig. 20. Failure zone for a locally drained borehole.

Published by NRC Research Press

Fig. 21. Failure zone of a totally drained borehole.

Published by NRC Research Press

Published by NRC Research Press

View publication stats

You might also like