Forensic Engineering: Informing the Future with Lessons from the Past

ISBN 978-0-7277-5822-4
ICE Publishing: All rights reserved
doi: 10.1680/feng.58224.225

IMPACT OF ADJACENT CONSTRUCTION ON

EXISTING PILE FOUNDATIONS

Yang, Hanwei Taiwan High Speed Rail Corporation, Taipei, Taiwan

Ho, Shu-Ken Sino Geotechnology, Inc, Taipei, Taiwan

Introduction
Ground settlement or lateral soil movements due to construction activities are well understood.
Ground movement associated with nearby construction activities may have adverse impacts
on existing piles, including geotechnical or structural pile failure due to movement of the soil
in relation to, as well as to additional shear or bending moments acting on, the pile structure.
O’Neil (1983) showed that pile-soil-pile interactions consist of “installation” effects and
“mechanical” effects. Installation effects, which refer to the alteration of soil stress states,
densities, and grain size distributions due to pile installation, can be different for groups of
piles and single pile. Mechanical effects refer to the alteration of soil strains and failure zones
due to simultaneous loading of closely spaced piles. The mutual interaction of installation and
mechanical effects makes rigorous analysis difficult. Huang (2001) summarized data from
field tests including the standard penetration test (SPT), cone penetration test (CPT), and
dilatometer test (DMT) on a full-scale lateral pile load test project in Taiwan to evaluate the
effects of construction on the behavior of laterally loaded pile groups. This study showed that
pile group installation, bored or driven, tended to disrupt the material near the ground surface
and lower the in situ lateral stress. Bored pile group construction appeared to loosen the soil
surrounding the piles, whereas driven pile construction apparently caused densification.

The Taiwan High Speed Rail (HSR) Project is a key component of transportation
development in Taiwan that involves the construction of an approximately 345-km-long
railway between terminal stations at Taipei in the north and Kaohsiung in the south. The
southern section from TK193 to TK345 is constructed on an alluvial plain and nearly all
alignments are essentially on viaducts and bridges. Typical foundation conditions are deep
alluvial or colluvial materials, often overlaid by soft clay of varying thickness that require
piles to provide sufficient capacity and minimize the possible settlement.

To provide access between HSR stations and nearby urban areas, highway authorities
constructed three sections of highway parallel to the HSR alignments. The nearest distance

225

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Forensic Engineering: Informing the Future with Lessons from the Past

between the edges of the HSR viaduct structure and highway structure is 5 m. In general, the
structural type of a highway is an embankment 2–3 m high. However, viaducts with a ramp
structure of up to 6 m high were adopted for sections crossing over a river. For these reasons,
many nearby construction activities such as piling, earthwork, and excavation of pile caps
from the highway construction project were undertaken between 2004 and 2005. The effects
of nearby construction activities on existing HSR pile structures have been monitored and
documented (Sino Geotechnology Inc., 2006, 2007). In this paper, we first present two cases
of pier settlement, namely of Piers P627 and P1224, caused by adjacent construction activities.
Subsequently, we investigate the underlying causes by monitoring data.

Observed pier settlement: P627

On 10 June 2004, a visual inspection of the track alignment around pier P627 north of the
Yuanshuixi River area was undertaken following a verbal report of misalignment of the tracks
in this immediate area. The inspection confirmed the presence of displacements for both
north- and south-bound tracks. Immediate steps were taken to carry out a detailed survey of
trackwork bollards. The results indicated that a vertical settlement of around 55 mm had taken
place at some time between April and June of 2004. The horizontal alignment was also
displaced by around 50 mm.

The affected HSR structure is a standard 30-m concrete box girder viaduct simply supported
on four bearings on top of the pier head. The pile cap is 11 m × 11 m × 2.8 m and supported
on four 2-m-ij piles 58 m long. The piles are designed as friction piles; they were built
entirely within the alluvial layers and therefore not founded in rock. This is normal for all
piles within the southern section of the railway alignment.

We noticed that construction work was underway for a new highway that straddles the
existing HSR viaduct alignment and includes a 385-m-long bridge across the Yuanshuixi
River. This bridge is composed of twenty-six 25- to 35-m-span box beams and supported by
26 piers and abutments. Within the Yuanshuixi River area, the highway contractor was
constructing large bridge foundations. At the time of discovery of the track misalignment, pile
construction work was underway adjacent to pier P627. Figure 1 show the layout of structure
for HSR and adjacent highway. The highway contractor has reportedly been encountering
problems with the piling work at this location.

Excessive loading, insufficient pile capacity, and impact from adjacent highway construction
are three potential causes of the unexpected settlement of HSR piers. The loading from the
superstructure and available pile capacity were thoroughly assessed by considering the site
conditions and construction activities. The current loading was only about 16% of the pile

226

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Yang and Ho

capacity. Hence, insufficient design capacity is excluded as the likely cause of the observed
settlements. At the time the settlements of the piers were recorded, construction work on the
foundation for the adjacent highway was underway; in addition, these piers are rather close to
those of the HSR infrastructure. Therefore, the possibility that these two events were
connected is reasonable.

Figure 1. Structural layout Figure 2. Concrete consumption record

Experience shows that during construction of the cast-in-place piles, the possibility of
over-excavation and/or cave-in or lateral sand flow in the cohesive soil layer can be identified
by monitoring the quantity of concrete used in the filling stage. Monitoring records indicated
that during construction of the foundation for the four piers of the adjacent highway (A1W,
P1E, P1W, and P2E), noticeably excessive amounts of concrete were pumped. Figure 1 shows
the structural layout of the adjacent highway. Most of the over-use occurred between the
depths of 35 to 50 m, where a cohesive soil layer was present. Figure 2 gives the typical curve
of concrete consumption along the length of the pile for pier A1W. The four above-mentioned
piers under construction are located near P626–P629, of which P627 is positioned in the
middle. The total amount of over-used concrete was around 450 m3 for these four piers.
Therefore, estimating that the total volume of over-excavation, cave-in, and lateral sand flow
will be equal or somewhat greater than 450 m3 is reasonable. We note that the period of
construction for these four piers was between April and June of 2004. The effects of these
over-excavations on the site soils are difficult to assess. The presence of the thick cohesive
soil layer located above the 35- to 50-m sand layer would most likely mask the effects on
short-term bases. This is probably why the supplementary site investigation borings located
near the pier of the HSR did not reveal any significant disturbances in the soil layers. If the

227

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Forensic Engineering: Informing the Future with Lessons from the Past

construction resulted in disturbances in the sand layer and caused loosening of the sand soils,
the overburdening pressure acting on the sands might even re-compress the sand so that
subsequent N-values of SPT in situ would not detect significant density-associated changes.
Nevertheless, the effects of each over-excavation were greatest at the pier location and could
be accumulative in nature; hence, the combined effects would be greatest and most obvious at
or near the center of these four newly constructed piers. Given that the disturbance,
re-compaction of the sand layer, and subsequent downward movement of the cohesive soil
above the sand layer would be progressive, they may not be noticeable immediately after the
over-excavations. This could have caused the track distortion initially observed on June 2,
2004.

The construction of pier P1E-1 started on June 3, 2004. Major difficulties were encountered
during construction of this pier. An incident that involved upward movement of the steel cage
(reinforcements) suggested that serious lateral sand flow had occurred. Reinsertion of the
casing was stopped near the depth of 41.1 m, for 40 hours; this may have resulted in
additional over-excavation and/or cave-in of the sand soils.

Construction records also indicated that on June 10, 2004, an impact hammer was used to free
the casing. These impact loadings may increase pore water pressure in sandy soil and cause
additional disturbances and subsequent re-compaction of the sands. Figure 3 is a photo that
shows significant cracks on the ground after this incident. This indicates a serious disturbance
of the surrounding soil.

Figure 3. Ground cracks after incident

Quantitatively, when subsidence of the upper soil layer causes negative skin friction acting on

228

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Yang and Ho

the pile, the safety factor for the pile capacity is reduced. This factor becomes one when the
sum of the dead load and the magnitude of negative skin friction are equal to the residual
friction resistance and tip resistance combined. A safety factor of one usually results in large
vertical displacements. According to the original design parameters, the dead load on each
pile of the four-pile group is 438 tons (i.e., ¼ of 1750 tons). The amount of negative skin
friction that could develop from subsidence of the upper 41 m is about 1272 tons for each pile.
The residual positive skin friction from the soil below 41 m would be 969 tons. According to
the above discussion, at least 840 tons of tip resistance (78% of the original design tip
resistance of 1072 tons) is needed to maintain a safety factor of one for the pile. Experience
with the construction of cast-in-place piles by using the reverse circulation suggests that there
is mud at the base of the piles constructed at the site; however, based on the observed
settlement, we may conclude that the actual tip resistance is 840 tons or less.

We note that the discussion above was based on examining available data and performing pile
capacity analyses to evaluate the likelihood of each possible cause. Therefore, the conclusions
made are qualitative in nature; any quantitative assessments presented are approximations
engineering practice.

According to the latest survey data, no additional settlement has occurred since June of 2005;
some minute upheavals have been recorded, although this is within the acceptable range of
survey tolerance. This indicates that the piles have reached an equilibrium state and that the
reduction of pile capacity due to negative skin friction should decrease with time and lead to
full recovery of the friction pile capacity.

Observed pier settlement: P1224

On Dec. 2, 2004, twisted rails were reported in the vicinity of pier P1224 in the Chiayi area.
The suspected cause of rail movement was settlement of the piers, which in turn may have
been caused by the adjacent highway construction work underneath the viaduct. There were
excavations adjacent to the HSR piers in order to construct pile caps and a foundation for a
retaining wall. The soil from excavation was piled up along the site up to 5 m high in some
areas. The settlement profile in the vicinity of pier P1224 since June of 2003 indicated that
settlement (a few centimetres) had occurred between May and December of 2004. The
records between June of 2003 and May of 2004 did not reveal any significant settlement of
piers. Consideration of the site conditions suggests that the settlement of piers supported by
piles is initiated by uncontrollable soil movement due to the placement of fills, excavations, or
pile installations by the construction of a nearby highway. However, further evidence is

229

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Forensic Engineering: Informing the Future with Lessons from the Past

required to support this hypothesis.

Field monitoring program and data interpretation

In January of 2005, the foundations for the elevated section of the highway project were
completed. The remaining ramp structures located on the north and south sides of two
waterways, namely the Xinpi Channel and the Puzhi Creek, were expected to cause further
ground settlement that was likely to make a significant impact on the existing HSR
infrastructure. For this reason, a comprehensive monitoring program was set up prior to
commencing construction of the remaining ramp structures.

The monitoring program consists of three parts. The first includes four
extensometer-monitoring wells and two inclinometers installed in the area between the
highway ramp structures and the HSR infrastructure. The second involves the survey
monuments on the retaining structure for the ramp highway. The third is comprised of the
survey monuments built on the pier of the HSR viaduct.

The construction activities for the ramp structure include earthwork and retaining wall seating
on piles. The height of fill for a ramp structure varies from 2 to 6 m. Two rows of piles with a
diameter of 1 m and a length of 24–32 m were constructed to support the retaining wall.
Construction began in January of 2005 and was completed in June of 2005.

Data from a leveling survey for HSR pier elevation showed HSR piers in four areas that
underwent significant settlement since December of 2004. These four areas north of the Xinpi
Channel and south of the Puzhi Creek are located coincidently near the ramp structure for the
highway. Figure 4 shows the pier settlement time history of the four most affected piers and
nearby construction activities from 12/19/2004 to 08/11/2006. During this time, two stages of
significant settlement occurred during the periods from mid-March of 2005 to mid-June of
2005 and from mid-December of 2005 to mid-April of 2006. A similar trend was noticed
from the data of piers located south of the Puzhi Creek. Pile installation for retaining wall and
fill work for ramp embankment were the only construction activities during the same time
frame.

The data from the extensometers in Figure 5 show similar displacement behaviors during the
same period. Data from SIS4 (Figure 6) indicate that a distinct lateral displacement of soil at
depths ranging from 20 to 30 m below the ground surface was recorded during the pile

230

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Yang and Ho

installation. The data from SIS4 and EXT4 both indicate that major soil deformation occurred
at soil layers 20–35 m below the ground surface during pile installation. This clearly
demonstrates a mechanism of pile settlement due to soil deformation adjacent to the pile
structure. In addition, data from EXT4 (Figure 7) indicate 30 mm of accumulated settlement
due to the 6-m-high fill.

Figure 4. Settlement histories for four affected piers on north side of Xinpi Channel

Figure 5. Data from extensometers: north side of Xinpi channel

231

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Forensic Engineering: Informing the Future with Lessons from the Past

Figure 6. Lateral soil displacement observed by SIS 4 and EXT 4 during pile
installation

Figure 7. Soil deformation observed by EXT 4 during fill work for ramp embankment

232

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Yang and Ho

Table 1. Summary of pile capacity re-evaluation

^ŝŶŐůĞWŝůĞĂƉĂĐŝƚLJ;ŬEͿ
WŝůĞ WŝůĞ WŝůĞ ŝƐƚƵƌďĞĚ
,^Z ^ŝŶŐůĞWŝůĞĞƐŝŐŶ ;EĞŐĂƚŝǀĞ&ƌŝĐƚŝŽŶŽŶĚŝƚŝŽŶͿ
dŽƉ ŽƚƚŽŵ >ĞŶŐƚŚ ^Žŝů>ĂLJĞƌ
WŝĞƌ ĂƉĂĐŝƚLJ;ŬEͿ
;>͘ŵͿ ;>͘ŵͿ ;ŵͿ ĞƉƚŚ;ŵͿ ^Ͳϭ ^ͲϮ

WĨŶс ϵϴϴϴ ϵϴϴϴ

YƐс ϮϱϱϴϮ YƉƐс ϭϯϯϵϮ ϭϯϯϵϮ

WϳͬϭϮϮϰ
ϭ͘Ϭ ͲϱϬ͘ϱ ϱϭ͘ϱ ϯϰΕϯϳ
;ϴWŝůĞƐͿ
Yďс ϴϰϵϮ Yďс ϴϰϵϮ Ϭ

YƵůƚсYƐнYďс ϯϰϬϳϯ YƵůƚсYƉƐнYďͲWĨŶс ϭϭϵϵϲ ϯϱϬϰ

WĨŶс ϵϬϴϯ ϵϬϴϯ

WϳͬϭϮϮϱ YƐс ϮϲϬϯϬ YƉƐс ϭϰϴϱϰ ϭϰϴϱϰ

;ϰWŝůĞƐͿ ϯ͘ϴ ͲϱϬ͘Ϭ ϱϯ͘ϴ ϯϭΕϯϲ
Yďс ϴϰϵϮ Yďс ϴϰϵϮ Ϭ

YƵůƚсYƐнYďс ϯϰϱϮϮ YƵůƚсYƉƐнYďͲWĨŶс ϭϰϮϲϯ ϱϳϳϭ

Re-evaluation of pile capacity

Piles for two piers, P1224 and P1225, suffered significant settlement due to adjacent
construction were selected in this paper for re-evaluation of pile capacity. Referring to the unit
skin friction mobilization curves from the pile load test conducted in this area, we find that
the displacement required to mobilize ultimate pile skin resistance is 20–25 mm. The
observed accumulated ground settlement is within the range of 2–4 cm which is sufficient to
mobilize ultimate pile unit skin friction. That negative skin friction can be generated on piles
and result in a large reduction of pile capacity and pier settlement may be reasonably expected.
Table 1 gives the dimensions and design ultimate pile capacity for the piles of each pier.

It is assumed that the neutral point (zero relative movement between soil and pile shaft) is
located at the disturbed soil layer. Pile negative skin friction (Pfn) could be mobilized by the
soil layers along the pile structure above the neutral point. The ultimate pile capacity (Qult)
with consideration of end bearing capacity (Qb) for Pier 1225 and 1224 are 11,996 kN and
14,263 kN respectively (CASE-1) which are less than the dead load from the
superstructure combined with the live load from the train (17,150 kN). If no end bearing
capacity can be mobilized (CASE-2), the deficient gap of pile capacity between demand and
supply is even higher. The results of pile capacity re-evaluation from both cases indicate
that the pile capacity may not be sufficient to resist the loadings from the current conditions.
It can therefore reasonably conclude that the significant settlement of piles during adjacent
construction was resulted from in-sufficient pile capacity due to pile negative skin friction
acting on pile structure. Details of the assessment results of pile capacity are summarized in
Table 1.

233

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.
 Forensic Engineering: Informing the Future with Lessons from the Past

Conclusions
Here we reviewed two cases of pier settlement. We examined and assessed the possible
causes, including excessive live/dead loads, insufficient pile capacity, and construction of an
adjacent highway. We found that piles sitting on an alluvial plain were easily affected by
nearby pile installation and fill work. The main adverse impact of nearby construction on
existing piles is the densification of loose and sandy material after disturbances.

Monitoring data provide direct evidence of significant lateral soil movement and settlement
caused by bored pile installations with casings. In addition, we observed the squeezing of
loose material underneath a ramp structure built with fill. During the same period, the
settlement of HSR infrastructure piers supported by piles occurred.

The lesson we learned from these two cases of pier settlement is that insufficient pile capacity
due to generation of negative skin friction, which resulted from the relative movement
between the soil and existing piles caused by adjacent construction activities such as piling
and fill work, is particularly critical for construction sites seated on the alluvium. Zones that
are free from the impact of piling/fill construction should be evaluated and defined prior to
construction. If construction outside of these zones is unavoidable, engineering schemes to
minimize the impact from adjacent construction should be adopted. Monitoring should be
conducted jointly by the contractor tasked with new construction and the owner of existing
structures in order to clarify the liabilities for unexpected settlement that occur on existing
structures during construction.

References
Huang, A.B., C.K. Hsueh, M.W. O’Neil, S. Chern, C.Chen, (2001) ‘Effects of Construction
on Laterally Loaded Pile Groups Foundation’, Journal of Geotechncial and Geoenvironmental
Engineering, ASCE, Vol. 127, No.5, pp.385-397.

O’Neill, M. W. (1983), ‘Group Action in Offshore Piles’, Proc. , Spec. Conf. on Geotech.
Engrg. In Offshore Praact., ASCE, New York, 25-63.

Sino Geotechnology Inc.,(2006), C576 - Settlement Assessment Report for P627 submitted to
Taiwan High Speed Rail Corporation, Taipei, Taiwan.

Sino Geotechnology Inc.,(2007), C575 - The Investigation and Evaluation of Adjacent

Construction Effects to the THSR Infrastructure in Yunlin, Chiayi, and Tainan, submitted to
Taiwan High Speed Rail Corporation, Taipei, Taiwan.

234

Downloaded by [ York University] on [13/12/18]. Copyright © ICE Publishing, all rights reserved.