This is a repository copy of Bridge distress caused by approach embankment settlement.

White Rose Research Online URL for this paper:

http://eprints.whiterose.ac.uk/4769/

Article:
Jones, C.A., Stewart, D.I. and Danilewicz, C.J. (2008) Bridge distress caused by approach
embankment settlement. Geotechnical Engineering, 161 (2). pp. 63-74. ISSN 1353-2618

https://doi.org/10.1680/geng.2008.161.2.63

Reuse
See Attached

Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing [email protected] including the URL of the record and the reason for the withdrawal request.

[email protected]
https://eprints.whiterose.ac.uk/
Proceedings of the Institution of
Civil Engineers
Geotechnical Engineering 161
April 2008 Issue GE2
Pages 63–74
doi: 10.1680/geng.2008.161.2.63

Paper 14958
Received 25/09/2006
Accepted 07/06/2007
Charlotte A. Jones Douglas I. Stewart Christopher J. Danilewicz
Keywords: bridges/embankments/ Geotechnical Engineer, Buro Senior Lecturer, School of Principal Geotechnical
piles & piling Happold, Bath (formerly MSc Civil Engineering, University Engineer, Halcrow Group
student, University of Leeds, of Leeds, UK Limited, UK
UK)

Bridge distress caused by approach embankment settlement

C. A. Jones, MSci, MSc, FGS, D. I. Stewart, MPhil, PhD and C. J. Danilewicz, MA, MSc, CGeol, FGS

Surtees Bridge, which carries the A66(T) over the River mv coefficient of volume compressibility
Tees near Thornaby-on-Tees in the UK, has been ˜y horizontal deflection of pile cap
showing signs of distress that predate its opening in ˜yq non-dimensional pile cap deflection (defined by
1981. Subsequent investigations have shown that the equation 2(b))
bridge distress is related to unexpectedly large ó9v vertical effective stress
settlement of the eastern approach embankment.
Recent ground investigations prompted by a proposed
widening of the river crossing have produced many new 1. INTRODUCTION
data on the alluvial deposits underlying the site, and The foundations for highway bridges must satisfy demanding
explain why embankment settlement was so much larger movement criteria if a bridge is to perform satisfactorily over
than originally anticipated. Comparison of the its full design life. Many highway bridge foundations, however,
geotechnical parameters obtained from the original and fail to meet these limits. A survey in the 1980s of around 300
more recent ground investigations suggests that the bridges in the United States found that a third had undergone
original investigation significantly underestimated the intolerable foundation movements. 1 Movement of bridge
thickness of an alluvial clay layer underlying the site, and supports can affect all aspects of bridge performance, from
that its coefficient of consolidation was overestimated. visual appearance to vehicle ride quality, and in extreme cases
Settlement analyses using geotechnical data from the can affect the structural integrity of the bridge.
original ground investigations predict moderate
embankment settlements occurring principally during Most common types of highway bridge can tolerate reasonable
construction. Settlement analyses based on all the magnitudes of total and differential vertical settlement of their
available data predict far larger embankment supports without serious distress. For example, a longitudinal
settlements occurring over extended time periods. The angular distortion (differential settlement/span length) of 0.004
latter analyses predict an embankment settlement is likely to be tolerable for a continuous bridge. Horizontal
similar to that observed and of sufficient magnitude to movements, however, are much more damaging, and it is
cause the observed lateral displacement of the bridge usually recommended that horizontal movements be limited to
due to lateral loading of its piled foundation. less than 38 mm.1,2 Limiting the horizontal movement of
bridge abutments founded on soft soil is a challenge to
NOTATION designers. The use of piled foundations is generally effective at
˜q embankment load limiting vertical movement, particularly when end-bearing
Cc compression index (slope of one-dimensional normal onto a firm stratum or rock. Unfortunately, piles constructed
compression line on graph of e against log ó v9 ) through soft soil may be subject to lateral loads and
Cr recompression index (slope of rebound line on graph of movements as a result of time-dependent deformation of the
e against log ó v9 ) soil underlying the approach embankments. 3,4
cu undrained shear strength
Cv coefficient of consolidation Design guidance tends to focus on movements of the bridge
d pile diameter foundations, with less attention being paid to foundation
e void ratio conditions beneath the approach embankments. Differential
e0 initial value of void ratio settlement between a bridge abutment and approach
Ep Young’s modulus of pile embankment can be damaging to the road pavement, although
Es representative stiffness of soft clay layer such damage is easier to remedy than damage to the bridge
hs thickness of soft clay layer superstructure. A survey of several hundred highway bridges,
Ip moment of inertia of a pile carried out in Kentucky in 1968, found that about 80% had
KR relative soil–pile stiffness (defined by equation 2(a)) required some form of maintenance action to remedy faults
Leq equivalent length of pile between points of fixity caused by differential settlement.5 Piling a bridge abutment to
mOD elevation measured in metres relative to Ordnance limit movement of the bridge superstructure would tend to
Datum accentuate this problem.

Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al. 63
 This paper reports on the geotechnical performance of a 2. SITE DESCRIPTION
highway bridge that carries four lanes of traffic across the Surtees Bridge carries the A66(T) over the River Tees,
River Tees. Site investigations associated with plans to widen approximately 1 km to the south-west of Thornaby railway
the river crossing to cope with increased traffic volumes have station, Thornaby-on-Tees, Cleveland, at National Grid
revealed that this 25-year-old bridge has suffered significant reference NZ 446 178. Surtees Bridge is located upstream and
settlement-related distress. This distress is caused primarily by south of two railway bridges, one of which carries the
settlement of one approach embankment and the resulting Darlington to Saltburn railway line across the River Tees (Fig.
movement of the associated bridge abutment. Based on limited 1). Historical Ordnance Survey maps show that Victorian
construction information it is understood that the western railway sidings, related to iron works, covered much of the
approach embankment was built 18–24 months prior to bridge land now occupied by the eastern approach embankment of
construction, and surcharged to increase the rate of settlement. Surtees Bridge. These sidings were constructed on ground
It consequently shows no sign of recent movement. The eastern raised to the level of the railway, apart from one spur that ran
embankment, however, was built contemporaneously with the down through the filled ground to a quay on the river (Fig. 2).
bridge construction, with no measures to increase the rate of Some land between the main railway line and the siding to the
settlement, and has subsequently undergone large settlements. landing stage was used for allotments.
This paper investigates the reasons why movements of the
eastern bridge abutment are causing distress to the bridge, and Construction of the bridge took place between 1980 and 1982,
why these movements were not anticipated during bridge except for the western approach embankment, which was built
design. under an advanced works contract in early 1978. This
444600

444800
700

518000 518000
Sta

Depot
El Sub

9·1 m
) m
Path (u

9·8 m

Tees Bridge a

10·4 m

° SL #5 CH4760
BM 9·68 m m
10·4 m
MP.25°
dy
Ward B
SM

SL
) m

A 66
Path (u

#2
i
517900 517900

#1 #4

#3 Outfall
g
Surtees Bridge

i
h
Pat Path (u
m)

D
444800

700
(c) Crown copyright/database right 2004 An Ordnance Survey/EDINA supplied service

0 10 20 30 40 50 60 70 80 90 100 m

Fig. 1. Site plan (drawn using Edina Digimap Carto).

# Crown copyright and Landmark Information Group Limited (2007). All rights reserved (2005)

64 Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al.
 bedrock is generally encountered at a consistent level of

22 mOD across the site. Made ground, including former
railway siding fill and blast furnace waste, is found beneath the
ry
Tees nda bridge approach embankments. Engineering descriptions of the
rd bou
Bridge Wa main soil horizons, together with their reduced levels at
chainage 4760 (whose position is shown in Figs 1 and 4), are
Allotment
presented in Table 2.
Landing gardens
stage
There is no record of erosion, other than reworking by the
river, at the site after the alluvium was deposited. General
filling to form the level of the railway sidings would, over
time, have increased the effective stresses in the underlying
deposits. This fill is very extensive, and any removal during
construction of the bridge and approach embankments is
expected to have been relatively minor. Groundwater level at
the site is dominated by the river, which was tidal until
completion of the Tees Barrage in 1995. Tidal variations in the
Fig. 2. Historical map showing the site in 1962 (excerpt from water table under the embankment are likely to have been
National Grid 1:1250). relatively small, and thus it is reasonable to assume that the
# Crown Copyright and Landmark Information Group alluvial deposits at the site of Surtees Bridge are lightly
Limited (2007). All rights reserved (1962)
overconsolidated.

The original ground investigation for Surtees Bridge comprised

embankment was raised in three lifts, surcharged, and allowed 10 boreholes and one Delft sample hole (three boreholes near
to settle for 18–24 months prior to implementation of the main the western abutment, five in the river channel, and two
construction contract. Information on the construction of the boreholes and the Delft sample hole near the eastern
eastern approach embankment is more limited, but the absence abutment). The 1976 Soil Mechanics report7 presents a
of an advanced works contract for the eastern approach longitudinal geological section along the line of the road based
embankment indicates that it was built contemporaneously on these data. The relevant portion is reproduced in Fig. 5.
with the bridge abutments, whose construction started in 1980. Prior to construction it was thought that a layer of alluvial
The railway access to the riverside quay was infilled, but it is clays approximately 7.5 m thick was present under the eastern
not known how much of the old siding fill was incorporated abutment. The alluvial clay was thought to be underlain by
into the eastern approach embankment. 9 m of alluvial silty sand resting on glacial deposits (such a
ground model appears to be supported by the limited pre-
3. GEOLOGICAL SITE CHARACTERISATION construction SPT data shown in Fig. 3(a), but not by the more
At least six separate ground investigations have been extensive post-construction SPT data shown in Fig. 3(b)). The
undertaken at the site of Surtees Bridge since 1973: two to pre-construction ground model of the western approach
facilitate bridge construction 6,7 and four more recently to embankment indicated that it was underlain by up to 14 m of
assess the reasons for continued settlement of the approach alluvial clays. This significant difference in the thickness of the
embankment and their implications for bridge widening 8–11 alluvial clay layer in the east and west bank models may
(see Table 1 for a full chronology). During these investigations explain why a decision was made to construct the western
a total of 49 boreholes have been advanced in the vicinity of approach embankment as advanced works prior to bridge
Surtees Bridge, together with non-sampling CPT investigative construction.
methods. Uncorrected SPT data acquired from borings
immediately around the eastern abutment, presented in Fig. 3, 4. DETAILS OF SURTEES BRIDGE
are a guide to the ground conditions beneath this abutment. Surtees Bridge comprises a continuous deck supported by four
Based on all the ground investigation data, published intermediate piers (see Fig. 4). The distance between abutments
geological maps 12 and geological studies of the region, 13 a is approximately 125 m. Road level is approximately 11 mOD
ground model has been developed for the site (see Fig. 4). across both approach embankments and the bridge deck. Paved
earth slopes have been constructed in front of the abutments,
The recent geology at the site comprises alluvial deposits and these slope down to the river banks. There is a wide river
(brown alluvium and grey alluvium 13 ), which infill a valley bank under the eastern end of the bridge at approximately
that was cut through Devensian glacial deposits by post-glacial 5.5 mOD, and a narrower bank under the western end of the
erosion. The more recent brown alluvium occurs principally bridge at approximately 3 mOD. River level is approximately
within the present river channel, with grey alluvium 2.7 mOD and is maintained at that level by the Tees Barrage
underlying the brown alluvium and extending beneath the downstream of the site.
approach embankments. The buried early post-glacial
topography affects the level of the top of the glacial deposits The four intermediate bridge piers (labelled B to E in Fig. 4)
across the site. Beneath the bridge and much of the eastern are each supported by four bored, cast in situ, 1.35 m
approach embankment the glacial deposits are encountered at diameter piles founded within the underlying bedrock. The
approximately
15 mOD, but they rise to near ground level to western abutment (labelled A in Fig. 4) is a bank seat set
the east and west. The underlying Triassic Sherwood sandstone upon the western approach embankment. Formation level of

Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al. 65
 Pre-construction
Preliminary ground investigation 6 1973 One Delft continuous sampler exploratory hole in the area now
occupied by the eastern abutment.
Main pre-construction ground 1975–1976 Four boreholes (two in the river) and one field vane test near to the
investigation 7 eastern abutment. Three other boreholes in the river. Three boreholes
and one field vane near the western abutment.

During construction
Western embankment constructed under Feb–May Embankment raised in three lifts to 4.3, 6.5 and 8.0 m above ground
an advanced works contract 1978 level.
Monitoring of western embankment Jan 1979 Extensometers indicate up to 400 mm of consolidation settlement.
Inclinometers show 95 mm and 140 mm of lateral movement extending
to 4 m below original ground level.
Start of main construction contract 1980
Observation during construction 18 Nov 1980 Between 70 mm and 115 mm of ground settlement in the vicinity of pier
E. Forward displacement of the pier is recorded (40 mm of forward
movement after piling was completed).
Bridge open to traffic 1981

Post-construction
Principal inspection report 1982 Highlighted settlement behind the east abutment and recommended that
this should be monitored and made up as necessary. Noted that bearing
deflector plates on pier E were deformed.
Principal inspection report 1986 Cracking noted in the crossheads of piers B, C and D (B and C having
been repaired). Bearings at abutment A show cracking of supporting
plinths and deformation of deflector plates. Bearings on piers B, C and D
show cracking of supporting plinths but no deformation of the deflector
plates. Bearings on pier E are as per abutment A with some rusting on
the soffits of the sliding surfaces.
Aerial photograph May 1991 Differences in pavement colour indicate that the road surface had been
renewed on both bridge approaches.*
Ground investigation 8 1992 One objective was to assess the reasons for continuing settlement of the
embankments leading up to the bridge and the implications for the
proposed widening.
Surtees Bridge Category III check 1993 Minutes from a 1992 meeting state that settlement of the eastern
assessment of existing structure embankment may have exceeded 490 mm. Deflection of pier C and D
crossheads inferred from positional survey.
Aerial photograph July 1995 Extensive resurfacing of western approach embankment and over 100 m
of eastern approach embankment.*
Walkover survey 2000 Pier E was reported to be visibly ‘out of plumb’.
Ground investigation 9 2000 One objective was to monitor for ground movements in the east and
west abutments.
Monitoring of inclinometers installed in Aug 2001 Deflections compatible with settlement of alluvial deposits were
2000 recorded.
Ground investigation 10 2002 Aims included investigating excessive and continuing settlements of the
eastern abutment.
Bridge expansion joint repaired 2003 Repair instructed following inspection. Cause and nature of damage not
known to authors.
Ground investigation 11 2004 Introduction states that monitoring indicates that ground movement was
still occurring.
Bridge expansion joint repaired 2005 Additional repairs required due to inadequate work during 2003 repair.

* The road has been resurfaced periodically as part of routine maintenance, when minor differences in road level will have been
corrected. However, there have been several changes in the authority responsible for maintenance, and the records have been lost.

Table 1. Chronology of Surtees Bridge

the western abutment is approximately 8 mOD. The abutment are believed to be between 30 and 33 m, with end levels
is supported on 29 precast, driven concrete Herkules type 800 corresponding to bedrock.
piles, 0.3 m in section, installed in two rows. The row nearest
the river contains nine vertical piles installed alternately The bridge deck is horizontally restrained at abutment F and
between ten piles raking forward at 1H:5V. The heel of the partially restrained at piers C and D (labelled R in Fig. 4). There
abutment is supported on ten piles that are raked backwards are sliding bearings (labelled S in Fig. 4) at abutment A and
at 1H:5V. The eastern abutment (labelled F in Fig. 4) has a piers B and E. Abutment F provides horizontal restraint of the
formation level of approximately 4.7 mOD. It is founded on bridge deck via the pile arrangement. Piers C and D are
51 Herkules type 800 piles installed in three rows. Of the 51 restrained to reduce their effective length in relation to
piles, 13 form a single row of vertical piles beneath the heel buckling. Piers B and E are half the length of piers C and D and
of the abutment, with the remaining 38 piles forming two do not need to be restrained by the deck. There is an expansion
rows of raking piles that are pitched towards the river with a joint between the deck and abutment A that is filled with a
rake of 1H:3V. The driven lengths of all the abutment piles flexible plastic inlay. This allows for thermal expansion of the

66 Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al.
 Eastern side of river Eastern abutment Ref. 8 Ref. 9 Refs 10, 11

10·00 10·00

5·00 5·00

2 m aOD

0 0

Elevation: m OD
Elevation: m OD

⫺5·00 ⫺5·00

⫺10·00 ⫺10·00

⫺14·5 m aOD
⫺15·00 ⫺15·00

⫺20·00 ⫺20·00

N ⫽ 10 N ⫽ 10
⫺25·00 ⫺25·00
0 50 0 50
N-value N-value
(a) (b)

Fig. 3. SPT N-values measured in borings from within about 30 m of centreline of eastern bridge abutment: (a) pre-construction
data (data from nearby borings in the river channel have been included to provide sufficient data for comparison);
(b) post-construction data. N-values are uncorrected for depth, tests using a standard sampler and cone are not separately
identified, and open symbols are used to show tests driven to refusal

Approx.
A B C D E F vertical 10 m
20 m scale
Ch.4760
Approx. horizontal scale
S S R R S R
Western Eastern
embankment embankment

Key
Made ground Brown alluvium Grey alluvium Glacial deposits Sherwood sandstone bedrock
R Restrained bearing S Sliding bearing

Fig. 4. Simplified longitudinal geological section, showing details of Surtees Bridge

bridge, while the plastic inlay prevents runoff and debris from values of relevant geotechnical parameters based on data
the road entering the expansion gap. from all six ground investigations conducted at the site. The
settlement parameters for the glacial deposits should be
5. GEOTECHNICAL PARAMETERS treated with caution, as the borehole logs indicate that the
Geotechnical parameters for each lithology described in Table two samples tested were recovered from locally finer soil;
2 are summarised in Tables 3 and 4. Table 3 presents median however, owing to the depth of the glacial deposits, the

Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al. 67
 Soil horizon Description Top of formation at Ch.
4760: mOD

Made ground Medium dense/firm grey brown, sandy gravelly clay with partings of silt. Gravels of 3.5
(Victorian fill) sandstone, limestone, brick, slag and clinker, with sand, occasionally of ash
Brown alluvium Soft to firm, thinly laminated, brown sandy silty CLAY/sandy clayey SILT with a 2.0
little gravel and occasional organic fragments
Grey alluvium Soft to firm, thinly laminated, grey very silty CLAY/very clayey SILT with 1.5
occasional gravels and parting of sand
Glacial deposits Medium dense to very dense brown silty SAND and GRAVEL with some cobbles
14.5
Sherwood sandstone Very weak to moderately strong, thinly bedded, red brown, fine to medium
22.0
grained SANDSTONE

Table 2. Engineering descriptions of soil horizons beneath site of Surtees Bridge

(62 not plotted)

60 39 61 64 63 65 66 67 41

Proposed level

mOD
10
MADE GROUND
Ash and slag
Silty fine to
medium SAND with
occasional gravel

Grey and brown organic

0
SILT and CLAY
Laminated grey brown
organic clayey SILT and
silty CLAY with occastional
sandy layers and partings
Grey silty fine SAND
becoming sandy SILT
towards the east
⫺10

SAND and GRAVEL

Occasional with occasional cobbles
clay bands

⫺20
Organic silty fine SAND
(variation of strata normal
to section)

Ch. 4600 Ch. 4700

Fig. 5. Extract from the pre-construction longitudinal geological section

settlement of the eastern approach embankment is insensitive and median values for the full dataset (51 consolidation
to these parameters. Compression of the grey alluvium will tests). Approximately equal values for mean and median
have dominated the settlement of the eastern approach indicate that the data distribution is close to symmetric, and
embankment, and its parameters are therefore presented in either value can be used to characterise the dataset. A
Table 4 in more detail. Table 4 reports separately data from significant difference between the mean and median indicates
prior to construction (two consolidation tests) and the mean a skewed dataset, and in such circumstances the median is

68 Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al.
 Victorian fill Brown alluvium Grey alluvium Glacial deposits

Soil properties

ªbulk : kN/m3 18.15 17.85 17.95 21.09

ªsat : kN/m3 18.16 17.89 18.10 20.88
D10 : ìm 20 1 2 230
e0 0.81 0.99 0.93 0.48
cu : kN/m2 98 27 38 140

Settlement parameters

mv : m2 /MN 0.11 0.24 See Table 4 0.07

Cv : m2 /year 1.83 5.6 See Table 4 1.83
Cc 0.29 0.26 See Table 4 0.11
Cr 0.06 0.03 See Table 4 0.03

Table 3. Geotechnical parameters (median values)

Pre-construction dataset Complete dataset

Mean* Mean Median

mv : m2 /MN 0.08 0.22 0.22

Cv : m2 /year 10.90 12.31 2.74
Cc 0.22 0.29 0.26
Cr 0.03 0.04 0.03

* There are only two pre-construction values: thus the mean is equal to the median.

Table 4. Settlement parameters for the grey alluvium

considered a better indicator for characterising the by the specific volume (1 + e0 ). Thus the relationship between
dataset. 14,15 mv and recompression index Cr (the slope of a graph of e
against the logarithm of ó v9 ) is defined by the equation 17
The initial void ratio (e0 ) and coefficient of volume
compressibility (mv ) of the grey alluvium exhibit considerable Cr
1 mv ¼ :
scatter at shallow depths. This is probably because the 2 3ó v9 ð1 þ e0 Þ
measured values have been greatly influenced by stress
increases generated by construction and filling across the site.
Thus any apparent depth trends will not necessarily be a good If it is assumed that the grey alluvium is overconsolidated
guide to the situation prior to construction of embankments throughout the consolidation test stress increment, then
and are not considered further. equation (1) yields values of mv very similar to those obtained
from the pre-construction investigations. This observation
The mean pre-construction value of mv for the grey alluvium is tends to confirm that the mv values measured in advance of
significantly lower than either median or mean for the whole construction reflected an overconsolidated soil response, and
dataset. This is probably an artefact of the stress increment any calculation based on these values would underestimate the
used to evaluate mv , which was 50 kPa in the pre-construction likely settlement induced by embankment construction.
investigations but >100 kPa in the post-construction
investigations. Re-evaluation of the pre-construction data The compression index (Cc ) and recompression index (Cr ) of the
using Casagrande’s construction 16 indicates that the grey grey alluvium as calculated from pre-construction data are
alluvium was lightly overconsolidated, with a pre- similar to the median values for the whole dataset. This
consolidation pressure about 30 kPa greater than the vertical consistency between the pre- and post-construction
effective stress prior to sampling. Thus a 50 kPa stress consolidation data permits Cc , Cr and an estimate of the pre-
increment would have given a predominantly overconsolidated consolidation pressure to be used to back-calculate
response. The higher stress increments of the recent testing embankment settlement.
would have been dominated by compression in the normally
consolidated section of the consolidation curve. A further feature of the settlement parameters for the grey
alluvium is a significant difference between the median and
The coefficient of volume compressibility (mv ) is defined as the mean values of the coefficient of consolidation (Cv ) for the
slope of the one-dimensional consolidation response (on a whole dataset. This may be taken to indicate that the dataset is
graph of void ratio e against vertical effective stress ó v9 divided highly skewed, and that the mean value is strongly affected by

Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al. 69
 extreme values (the mean is three times the upper quartile resurfaced several times since the bridge opened) and lateral
value of 3.8 m2 /year). The Cv values from the two pre- tension cracks located within the soil of the soft verge on the
construction consolidation tests are similar to the mean Cv south side of the eastern approach embankment (#4 in Fig. 1).
value for the whole dataset. This suggests that they were not
representative of the grey alluvium. Monitoring of inclinometers installed beneath the eastern
approach embankment in 2000 and 2002 indicated horizontal
6. EVIDENCE OF BRIDGE DISTRESS displacement within the grey alluvium. The displacement
There have been reports of excessive ground movements at pattern takes the form of a bulge, with no preferred direction to
Surtees Bridge since its first principal inspection (details are the movement. The pattern is interpreted as buckling of the
given in Table 1). By 1992 concern about these movements was inclinometer tubes within the soft alluvium as it is compressed
sufficient to make assessing the reasons for continuing by the overlying made ground. Deformation does not occur in
settlement of the eastern embankment a primary objective of the stiffer made ground above or sand and gravel deposits
the ground investigations performed for the proposed widening below. It is believed that the inclinometer tubes are ‘gripped’
of the bridge. 8 by the made ground, and so cannot slide to relieve the axial
stress that eventually leads to buckling. The inclinometer data
A bridge inspection undertaken in 2000 indicated that the indicate the eastern embankment is still settling.
bridge deck had moved westward relative to abutment A and
pier B. The displacement of the bridge deck relative to pier E is Other structures in the vicinity of Surtees Bridge show signs of
easterly, however. The articulation of the bridge is such that distress, suggesting that excessive settlement is an issue in this
only piers C and D and abutment F can transfer horizontal locality. For example, the nearby Tees Bridge (location #5 in
forces to the bridge deck. Piers C and D are unlikely to be the Fig. 1) has a speed restriction imposed, and its most easterly
cause of the bridge deck movement, because their foundations pier has recently been strengthened.
are remote from any source of lateral load. It was therefore
deduced that the bridge deck had moved westward due to
forward movement of abutment F. The movement of the bridge 7. SETTLEMENT OF THE EASTERN APPROACH
deck relative to pier E therefore indicates that there has been a EMBANKMENT
westward movement of the head of pier E greater than The amount and rate of primary consolidation settlement at
westward movement of the bridge deck. Westward chainage 4760 have been calculated using Terzaghi’s one-
displacement of pier E during construction of abutment F is dimensional method. 19 The initial stresses were calculated
noted in a report prepared by Bullen and Partners in 1993. 18 from the soil profile shown in Fig. 8, assuming that the
Movement of abutment A is not thought to be the primary Victorian embankments could be represented by equivalent
cause of closure of the expansion joint above that abutment, trapezoidal pressure distributions acting at a ground level
because there are no visible signs of settlement of the western (3.5 mOD). Osterberg’s method 20 was used to calculate the
approach embankment, and it cannot account for the vertical stress increase, which idealises the foundation soil as
movement of the bridge deck relative to piers B and E. a homogeneous, isotropic, elastic material. The error caused
by assuming isotropic elasticity is typically about 20% even
During a walkover survey conducted in May 2005 a number of when the soil is systematically non-homogeneous, anisotropic
signs of distress were observed at locations indicated in Fig. 1. and non-linear. 21 Embankment construction was simulated by
The most obvious sign of distress (#1 in Fig. 1) is a step in the removing the pressure representing the Victorian fill and
public footpath adjacent to the westbound carriageway on the replacing it with a trapezoidal pressure distribution
eastern approach embankment. Assuming that originally there representing the new embankment. Removal of the Victorian
was no step in footpath height between the bridge and the railway embankments is assumed to occur simultaneously
eastern embankment, and that the footpath has not been with placement of the new fill.
resurfaced since it was
constructed, then there has
been a vertical displacement
of about 0.2 m at this point
(Fig. 6). Also there has been
an estimated 100–150 mm of
horizontal movement over
the sliding bearing on
abutment A (#2 in Fig. 1), as
shown in Fig. 7, where only a
small gap now remains at the
expansion joint. Other signs
of distress include lateral
tension cracks in the road
surface (#3 in Fig. 1) adjacent
to the joint between the
Fig. 6. (a) View westwards across Surtees Bridge showing a repair made to the footpath to
bridge deck and eastern
compensate for settlement of the eastern embankment; (b) the change in height across the
approach embankment repair.
(despite the road being

70 Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al.
 are presented in Fig. 9 (the analysis ignores the settlement of
the embankment fill). The maximum predicted primary
consolidation settlement is 0.71 m, with the asymmetric
distribution reflecting the smaller stress change under the site
of the Victorian embankments.

To apply Terzaghi’s analysis of rate of settlement to the ground

conditions at chainage 4760 it is necessary to further simplify
the ground model by assuming that the 0.5 m thick layer of
brown alluvium has the same consolidation properties as the
16 m thick layer of grey alluvium. This composite layer is
assumed to be undergoing two-way vertical drainage into the
Victorian fill and glacial deposits on the basis of particle size
(Table 3 gives the median D10 of each soil horizon). The mean
Cv value of the grey alluvium is unrepresentative, because the
Fig. 7. A sliding bearing on abutment A data distribution is highly skewed and strongly affected by
outliers, so the median Cv value of 2.74 m2 /year is used in the
rate analysis. The analysis suggests that the time for 95%
primary consolidation is 28 years, and that about 93% of the
consolidation settlement has occurred to date (Table 5).
15
For comparative purposes, the amount and rate of
10
consolidation settlement at chainage 4760 have also been
5 evaluated for the ground model available pre-construction. At
0 that time it was believed that the layer of alluvial clays was
Elevation: mOD

⫺25 ⫺20 ⫺15 ⫺10 ⫺5 0 5 10 15 20 25

only about 7.5 m thick under the eastern abutment, and
⫺5 Distance from centreline: m
underlain by silty sand. A total of six consolidation tests were
⫺10
conducted as part of the SI for Surtees Bridge: four on brown
⫺15 alluvium from under the western abutment, and two on grey
⫺20
Made ground (1·5 m) alluvium from under the eastern abutment. 7 Mean Cc values
Brown alluvium (0·5 m)
Grey alluvium (16 m) were 0.33 and 0.22, and mean Cv values were 9.8 and
⫺25 Glacial deposits (7·5 m)
Sherwood sandstone bedrock
10.9 m2 /year for brown and grey alluvium respectively. The
⫺30 pre-construction ground model does not differentiate the
brown and grey alluvium, but the borehole logs indicate that
Fig. 8. Idealised cross-section at chainage 4760 only grey alluvium was found under the eastern embankment.
Consolidation properties were not derived for the Victorian
fill, alluvial sands or glacial deposits during design of the
For the purposes of the analysis it is assumed that the bridge, probably because they appeared to be coarse and
groundwater level at the time of embankment construction was therefore were believed to be relatively incompressible
about 1.5 m below ground level (2 mOD), which is slightly (subsequent results have shown that this was a significant
above mean river level before construction of the Tees Barrage. error for the fill).
Median values of Cc and Cr were used in the analysis (Tables 3
and 4), and it is assumed that the maximum consolidation The maximum predicted primary consolidation settlement
stress is about 30 kPa greater than the current vertical effective calculated using the pre-construction ground model is only
stress. The results of the settlement analysis for chainage 4760 0.28 m (see Fig. 9) and the time for 95% consolidation only
1.5 years (Table 5). Thus it seems that, pre-construction, only
moderate embankment settlements would have been
Footpath anticipated, and these were expected to occur principally
N S during construction, when they were unlikely to cause bridge
distress.
⫺20 ⫺10 0 10 20 30
0 8. DISPLACEMENT OF PILES UNDER PIER E AND
Distance from centreline: m
Settlement: m

0·2
ABUTMENT F
0·4
It is believed that westward movements of pier E and abutment
0·6
F were caused by deformation of the alluvial clays beneath the
0·8
Pre-construction ground profile
eastern approach embankment causing lateral loading of the
1·0
Post-construction ground profile piled abutment and pier foundations. To test this hypothesis,
Settlement profile: pre-construction ground model the movements of the pile caps at pier E and abutment F were
Settlement profile: post-construction ground model
evaluated using a very simple empirical method. 3 The method
involves a design chart, developed from centrifuge model data
Fig. 9. Predicted primary consolidation settlement at chainage
4760 and field observations, that relates the non-dimensional pile
cap deflection ˜yq to the relative soil–pile stiffness KR , where

Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al. 71
 Based on pre- Based on all
construction data data

Cv : m2 /year 10.9 2.74

Drainage path: m 3.75 8.25
Time for 95% consolidation settlement: years 1. 5 28.0
Current amount of consolidation for t ¼ 24 years: % .99.9 92.6

Table 5. Results of time–settlement rate analysis

0.10–0.15 m, and the westward movement of the top of pier E

˜ yEp I p
2a ˜ yq ¼ exceeds this amount). Given the very approximate nature of
˜qdLeq 4
the analysis, this prediction is surprisingly good. (There is some
uncertainty about the base level of the pile cap under abutment
F, and the prediction is extremely sensitive to effective length;
Ep I p
2b KR ¼ also, no allowance is made for pile restraint within the
Es h4s
embankment fill.) The prediction for pier E must be treated
with particular caution because two further, possibly opposing,
and ˜y is the horizontal deflection of the pile cap, ˜q is effects have been ignored. First, the distance of pier E from the
embankment load, d is pile diameter, Leq is the equivalent front slope of the approach embankment is similar to the
length of the pile between points of fixity, Ep is the Young’s thickness of the soft alluvial layer, so the foundations pier E
modulus of the pile, Ip is the moment of inertia of the pile, Es is would not feel the full effect of the approach embankment.
the representative stiffness of the soft clay layer, and hs is the Second, the prediction is for pile cap displacement, whereas it
thickness of the soft clay layer. The equivalent length, Leq , is is not clear whether the displacement has been observed at the
equal to either 1.3L (where L is the pile length above the base pile cap or cross-head of pier E. If the displacement of pier E
of the soft layer) when horizontal movement is not restrained was primarily rotational, then the cross-head displacement
(pier E) or L when rotation is prevented by a rigid cap would be about 20% greater than the horizontal displacement
(abutment F). 3 The method takes no specific account of factors of the pile cap.
such as pile spacing, group size, group configuration or
embankment shape, although a broad range of pile abutment Despite the shortcomings of the method, it has correctly
configurations are represented in the dataset used to develop predicted the order of magnitude of the deflections of pier E
the design chart. and abutment F. It is therefore concluded that lateral
movement of the piled eastern abutment as a result of
The design chart recognises that there is a marked increase in settlement of the approach embankment can explain the
pile cap deflection as the stress increase due to the observed bridge deck displacement.
embankment exceeds about three times the undrained strength
of the soft clay layer, owing to the onset of significant plastic 9. DISCUSSION
deformation in the soft stratum. The stress increase due to Surtees Bridge is showing signs of distress due to excessive
construction of the eastern approach embankment was settlement of the eastern approach embankment, and associated
typically just under three times the undrained shear strength of movement of the eastern bridge abutment. As a result there
the alluvium deposits. The limit was, however, exceeded under have been extensive post-construction ground investigations in
a relatively narrow section of the full-height embankment in the vicinity of that abutment—a process that has been given
an area not previously loaded by the Victorian railway sidings extra impetus by plans to widen the river crossing. This gave
embankment. The pile cap deflections were therefore estimated the authors the opportunity to reanalyse the performance of
using the upper limit for loading less than 3cu . this structure using data not available to the original bridge
designers, and speculate as to the cause of the ground
The parameters used to analyse Surtees Bridge are presented in movements.
Table 6. The Young’s modulus of the grey alluvium has been
estimated from the volume compressibility mv by assuming a A new ground model has been developed for the site of the
drained Poisson’s ratio of 0.2. 22 The design chart assumes that eastern approach embankment, based on all the available data.
the piles are vertical, but the authors recommend that a This identifies the importance of a layer of soft grey alluvial
reduction of 25% be made in the predicted pile cap deflection clay, 16 m thick, only 2 m below the ground surface. A
if some of the piles are raking. This correction has been applied conventional analysis using this ground model predicts that the
to the predicted displacements of abutment F. maximum consolidation settlement of the embankment will be
0.71 m. Under the footpath on the south side of the road the
The lateral pile cap deflections at pier E and abutment F are predicted settlement is about 0.19 m, which is comparable with
estimated to be about 0.34 m and 0.30 m respectively (Table 6). the observed settlement at this point of around 0.2 m. The
This prediction, which assumes that pile installation occurred analysis also identifies that time for 95% consolidation of the
before embankment construction was complete, is about twice grey alluvial clay is only just being approached. This agrees
the observed movements (abutment F has moved forward with the inclinometer data, which show that small

72 Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al.
 Parameter Pier E Abutment F

Pile diameter, d: m 1.35 0.305

Moment of inertia of pile, Ip : m4 0.163 0.425 3 10
3
Young’s modulus of pile, Ep :* MN/m2 26 000 34 500
Young’s modulus of alluvium, Es : MN/m2 5.1 5.1
Thickness of soft clay layer, hs : m 18.0 16.5
˜q: MN/m2 0.126 0.126
L: m 21.5 20.5
Leq : m 28.0 20.5
KR 7.9 3 10
3 39 3 10
6
˜yq 0.014 0.87 3 10
3y
˜y: m 0.34 0.30

* Determined from an assumed characteristic strength for in situ cast concrete of 30 MN/m2 and
the reported characteristic strength for Herkules pile concrete of 53 MN/m2 using the
relationship reported by Mindess et al. 23
y
Estimated by extrapolation of the trend line.

Table 6. Summary of pile head displacement calculations

deformations are still going on in the alluvial clay layers The movements of the eastern bridge abutment and approach
beneath the eastern embankment. A simple empirical analysis embankment were not anticipated prior to construction,
of the horizontal displacement of the piled bridge abutment has because the original ground model for the site developed from
been used to show that the observed bridge deck displacements data available before construction underestimated the thickness
can be explained by this mechanism. of a soft alluvial layer, and overestimated its coefficient of
consolidation. This highlights the difficulties in characterising
This raises the question of why large settlements of the eastern alluvial soils from the limited ground investigations conducted
embankment were not anticipated from the original design. The for many construction projects.
answer seems to lie in the pre-construction ground model,
which underestimated the thickness of the compressible ACKNOWLEDGEMENTS
alluvial clay layer under the eastern approach embankment by The authors would like to thank the Highways Agency for
a factor of two. The data available before construction suggest access to site data. They would also like to thank A-one
that the alluvial layer changes from silty clay to sandy silt with Integrated Highway Services and Halcrow for their support,
depth on the eastern side of the River Tees. It also appears that and M. D. Parker and T. W. Cousens, University of Leeds, for
the values of Cv measured for the grey alluvium before helpful discussions during the drafting of the paper. The
construction were about four times larger than the median opinions expressed within this paper are those of the authors,
value of the larger dataset now available. Thus it seems that and not necessarily those of the parties who have supported
relatively modest embankment settlement was anticipated and, our work.
owing to the higher Cv value and shorter drainage path, the
majority of that settlement was expected to occur during REFERENCES
construction. 1. MOULTON L. K., GANGA RAO H. V. S. and HALVORSEN G. T.
Tolerable Movement Criteria for Highway Bridges. Federal
Interestingly, the pre-construction ground model identifies that Highway Administration, Wahsington, DC, USA, 1985,
the alluvial clay layer under the western abutment was 12 m Report No. FHWA/RD-85/107.
thick. Thus a larger amount of settlement occurring over longer 2. BARKER R. M., DUNCAN J. M., ROJIANI K. B., OOI P. S. K., TAN
time periods would have been anticipated prior to construction. C. K. AND KIM S. G. Manuals for the Design of Bridge
Presumably this is why the western embankment was built in Foundations. Transportation Research Board, Washington,
advance of bridge construction—to allow time for the DC, 1991, NCHRP Report 343.
embankment to settle before construction of the western bridge 3. STEWART D. P., JEWELL R. J. and RANDOLPH M. F. Design of
abutment. piled bridge abutments on soft clay for loading from
lateral soil movements. Géotechnique, 1994, 44, No. 2,
10. CONCLUSIONS 277–296.
Excessive movement of Surtees Bridge has been a concern 4. ELLIS E. A. and SPRINGMAN S. M. Full-height piled bridge
since shortly after construction was complete. To date, the abutments constructed on soft clay. Géotechnique, 2001,
eastern abutment has moved forward by 100–150 mm, and the 51, No. 1, 3–14.
eastern approach embankment is estimated to have settled by 5. HOPKINS T. C. and DEEN R. C. The bump at the end of the
about 0.7 m. These movements are the result of compression of bridge. Highway Research Record, 1970, No. 302, 72–75.
an approximately 16 m thick layer of alluvial clay beneath the 6. SOIL MECHANICS. Delft Continuous Sampling Investigation
site, and have resulted in lateral loading of the piled for the A66 Stockton–Thornaby By-Pass, Co. Durham. Soil
foundations of the bridge abutment. Distortion of the bridge Mechanics, Bracknell, 1973, Report No. 6145/1.
structure as a result of these movements has been a significant 7. SOIL MECHANICS. Site Investigation T.R. A66 Stockton–
factor in the decision to replace the bridge when the crossing is Thornaby By-pass. Soil Mechanics, Bracknell, 1976, Report
widened in the near future. No. 6627.

Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al. 73
 8. EXPLORATION ASSOCIATES. A66 Surtees Bridge Proposed Conference on Soil Mechanics and Foundation
Widening, Vols 1 and 2. Exploration Associates, Wallsend, Engineering, Harvard University, Cambridge, MA, USA,
1992. 1936, Vol. 3, pp. 60–64.
9. ALLIED EXPLORATION AND GEOTECHNICS LIMITED. A66 Surtees 17. LAMBE T. W. and WHITMAN R. V. Soil Mechanics, SI version.
Bridge Ground Investigation. Allied Exploration and John Wiley and Sons, New York, 1979.
Geotechnics Limited, Chester-le-Street, 2000. 18. BULLEN AND PARTNERS. Thornaby Bypass Stage 2C A66
10. RITCHIES. A66(T) Surtees Bridge Investigation, Vols 1–4. Surtees Bridge Widening Geotechnical Assessment. Bullen
Ritchies, Wigan, 2002. and Partners, Durham, 1993.
11. RITCHIES. 0521 A66 Surtees Bridge, Vols 1–3. Ritchies, 19. TERZAGHI K. Theoretical Soil Mechanics. John Wiley &
Wigan, 2004. Sons, New York. 1943.
12. BRITISH GEOLOGICAL SURVEY (BGS). Stockton, England and 20. OSTERBERG J. O. Influence values for vertical stresses in a
Wales Sheet 33 Solid and Drift Geology.1:50,000. British semi-infinite mass due to embankment loading.
Geological Survey, Nottingham, 1987. Proceedings of the 4th International Conference on Soil
13. AGAR R. Glacial and post-glacial geology of Middlesbrough Mechanics and Foundation Engineering, London.
and the Tees Estuary. Proceedings of the Yorkshire Butterworths Scientific, London, 1957, vol. 1, 393–394.
Geological Society, 1954, 29, Part 3, No. 13, 237–253. 21. PADFIELD C. J. and SHARROCK M. J. Settlement of Structures
14. MCCLAVE J. T. and SINCICH T. 2000. Statistics, 8th edn, on Clay Soils. Construction Industry Research and
Prentice Hall, Upper Saddle River, NJ, USA, 2000. Information Association, London, 1983, CIRIA Special
15. MOORE D. S. Statistics: Concepts and Controversies, 4th edn. Publication 27.
W. H. Freeman and Co., New York, 1996. 22. BOLTON M. D. A Guide to Soil Mechanics. Macmillan,
16. CASAGRANDE A. The determination of the pre- London, 1979.
consolidation load and its practical significance. 23. MINDESS S., YOUNG J. F. and DARWIN D. Concrete, 2nd edn.
Discussion D-34, Proceedings of the International Prentice Hall, Upper Saddle River, NJ, USA, 2004.

What do you think?

To comment on this paper, please email up to 500 words to the editor at [email protected]
Proceedings journals rely entirely on contributions sent in by civil engineers and related professionals, academics and students. Papers
should be 2000–5000 words long, with adequate illustrations and references. Please visit www.thomastelford.com/journals for author
guidelines and further details.

74 Geotechnical Engineering 161 Issue GE2 Bridge distress caused by approach embankment settlement Jones et al.