HRPP653

Scour prediction in non-uniform soils: undrained shear strength and erodibility
JM Harris and RJS Whitehouse
Scour prediction in non-uniform soils:

undrained shear strength and erodibility
JM Harris1 and RJS Whitehouse1
1
HR Wallingford, Howbery Park, Wallingford, Oxfordshire OX10 8BA, UK
Published in the Scour and Erosion, Proc. 7th Int. Conf. on Scour and Erosion, The University of Western
Australia, 2-4 December 2014 (2015)
Abstract
Scour development in non-uniform soils is still an area of great uncertainty and remains a challenge for
designing structurally efficient and effective foundations in the marine environment. Scour risk in cohesive
soils is made more uncertain by effects such as weathering and time-scale to scour. For large volume
installation of foundations such as those related to offshore wind farm developments there is a limit to the
amount of detailed geotechnical information that can be collected as part of the project. Therefore, reliance
in data such as undrained shear strength, derived from cone penetration tests, supplemented with borehole
data collected at a limited number of sites across the wind farm and laboratory analysis of soil samples
becomes the principal source of geotechnical information. Hence, the question arises as to whether the
undrained shear strength be used as a proxy for the erodibility of a soil as proposed in the approach of
Annandale (1995). This paper will present evidence from both field and laboratory measurements of
undrained shear strength and scour potential to test the hypothesis of undrained shear strength as a proxy
for scour.
1. Introduction
Scour is a hydrodynamic process related to the movement of the seabed sediment as a result of the flow of
water past and away from a structure. The soil conditions are described by geotechnical parameters,
therefore, scour is also of a geotechnical nature as it relates to the reduction in ground level around a
structure. Marine soils are often variable in their makeup, and consist of a range of sediments, including
sands, gravels, silts and clays. Van Ledden et al. (2004) defined six types of sediment bed based on
sediment particle size distribution and the remolded shear strength as a measure of cohesion.
The distribution of a given soil type within a sample will be site specific and will also vary with depth below
the seabed. Understanding of how the non-uniformity and layering of a soil and its variation with depth
impacts scour development around a structure is still limited and is not routinely accounted for in design.
The terms clay and mud are often incorrectly used interchangeably in sediment transport. Mud is typically
defined as a mixture of water, clay and silt, and also includes organic material and sometimes gas
(e.g. hydrogen sulphide resulting from organic decomposition). Mud is generically classified as particle sizes
of less than 63 µm, which includes silt. Whereas clays have a plate-like structure, and generally have a
diameter less than 2 µm. It should be noted that the United States Department of Agriculture classifies silt as
less than 50 µm, and other countries have their own particle size classifications.
HRPP653 1
Granular classification enables scientists and engineers to assess the nature of soil behaviour, i.e. cohesive,
granular or mixed, which has important implications for the mechanical and erosion properties of a sediment
(Whitehouse et al., 2000; Winterwerp and van Kesteren, 2004). The degree of cohesion increases with the
fraction of clay minerals within the sediment, and starts to become significant when the sediment contains
more than 5-10 % of clay by weight. The most important types of clay minerals are: kaolinite, chlorite,
montmorillonite, and illite.
The prediction of scour in cohesive or multi-modal soils is more complex. Typically the scour process is
much slower; as a result the effect of scour is very much dependent on the period of time that the structure
will remain at the site. Annandale (1995) proposed an approach to estimating the erosion potential of
complex soils through the use of the stream power parameter, P, and its relationship to the ability of the soil
to resist scour, defined through an Erodibility Index, K. The Erodibility Index provides a measure of the in-situ
strength of the material, whilst the stream power provides a measure of the rate of energy dissipation in the
near-bed region due to hydrodynamic forces. If P exceeds the erosion threshold then scouring will occur.
2. Erodibility
Soil mechanics testing provides workable definitions of the complete spectrum of soil types from pure
cohesionless sands to clays. Therefore a variety of soil parameter definitions have been prescribed and test
procedures defined to help the engineer assess the performance of foundations (e.g. Terzaghi et al., 1996).
These measures of soil properties have also been used in an attempt to produce correlations between the
hydraulic behaviour of the soil (i.e. threshold for erosion and erosion rate) and the properties of the bed such
as bulk density and shear strength. This is an important parameter in the assessment of scour hazard.
The erodibility of clay soils has been evaluated in terms of direct geotechnical measures of bed strength.
However, the analogy between the force required to strip the surface layers of a soil through the action of
flowing water and the force required to shear the bed using a shear vane (which measures the torque force
required to rotate a vane of a given shape and size) is not that clear. Individual studies have demonstrated
that a reasonable correlation exists between the two measures (e.g. Kamphuis and Hall, 1983). They found
that the value of shear stress required for erosion increased with unconfined compressive strength, vane
shear strength, plasticity index, and clay content of the soil, and with increasing pre-consolidation pressure.
The results of tests on a consolidated cohesive sediment (clay) showed the variation of threshold shear
stress for erosion with the first two parameters was linear.
Kamphuis and Hall found the ability of a cohesive soil to resist erosion increased with clay content and
plasticity index. One notable conclusion was, for the clay tested in their study, if the consolidation pressure
was greater than approximately 200 kPa then the soil was considered safe from erosion under “normally
expected, naturally occurring flow conditions”.
3. Scour in cohesive soils and clays

The principal body of work on scour in cohesive soils and clays is related to scouring around bridge piers.
Molinas et al. (1999) presented results from a study conducted using montmorillonitic cohesive soils. They
noted a distinction between pier scour occurring in unsaturated compacted soils compared with that
occurring in saturated soils. In general they found that the local scour depth decreased as the initial water
content decreases and increased as the Froude number (Fr) increases (Figure 1). They also state that the
HRPP653 2
side slopes of the local scour holes formed in the tests were much steeper than those encountered in non-
cohesive soils, although this may be a short-term effect.
Figure 1. Effect of initial water content on local pier scour for a saturated cohesive soil under different
approach flow conditions (after Molinas et al., 1999)
Briaud et al. (1999) proposed an approach to predicting the scour depth and rate of scour in cohesive soils
around a cylindrical bridge pier. The SRICOS method (Scour Rate In COhesive Soils) involved taking site
specific samples and testing them in an Erosion Function Apparatus to obtain the rate of scour against the
applied hydraulic shear stress, and combining this information with the maximum shear stress prior to scour
developing. Having undertaken these steps the maximum scour depth for the given pier structure could be
determined and then a hyperbolic function developed to describe the scour depth with time curve. The
predictions obtained using this method were compared against measurements from 42 flume experiments.
Briaud et al. proposed a simple relationship for the maximum scour depth, Smax, based on the pier Reynolds
number RD:
S max = 0.00018RD0.635 (1)
where,
VD
RD = (2)
ν
and V is the mean flow velocity, D is the pier diameter and ν is the kinematic viscosity. The formula is similar
to that proposed by Shen et al. (1969) for scour in non-cohesive soils.
S e = 0.000223RD0.619 (3)
where Se is the equilibrium scour depth. Comparing Eqns. (1) and (3) suggests the equilibrium scour in
cohesive soils and non-cohesive soils is similar.
Ting et al. (2001) reported on a series of experiments in the laboratory to study local scour at cylindrical piers
in clay. They compared their results with experimental data for local scour in sand and investigated the
effects of Reynolds number, Froude number and approach flow depth on the local scour depth. They
compared their results with predictions obtained using the approach outlined in Richardson and Davis
HRPP653 3
(1995). The results indicated that although the rates of scour in clay were much slower than in sand, the
equilibrium scour depth was similar to that obtained in sand (confirming the conclusion regarding Eqns. (1)
and (3) above). Ting et al. also noted that the shape of the scour hole correlated with the pier Reynolds
number, with the scour developing relatively uniformly around the pier at low pier Reynolds numbers. At
higher pier Reynolds numbers, the scour holes developed mainly behind the piers with much less scour
forming in front of the piers.
For the test conditions run, Ting et al. did not find any noticeable effect of Froude number and approach flow
depth on scour depth. The experiments were carried out using five different sediment types, three clays
(porcelain, armstone and bentonite) and two sands (fine, d50 = 0.14 mm; and medium, d50 = 0.60 mm).
Froude numbers were in the range 0.1 to 0.42 and the ratio of the approach flow depth to the pier diameter
varied between 1.43 and 16. Seven scour tests were carried out using the two sands for comparison
purposes.
Figure 2. Variation of equilibrium scour depth with pier Reynolds number for clay and sand (after Ting et al.,
2001)
The results as presented in Figure 2 show remarkable agreement with the results of Shen et al. (1969)
(compare with Eqn. (2) – the leading coefficient in Ting et al. relates to scour depth in mm). The results of
Shen et al. were valid for clearwater scour conditions only, as they found that in the non-cohesive soils that
they used under live-bed conditions the sediment transport into the scour hole resulted in a decrease in
scour depth. In cohesive soils where the eroded particles remain in suspension the scouring process
essentially remains within the clearwater regime and the equilibrium scour depth increases with pier
Reynolds number. Whether this approach is valid for cohesive marine soils is more difficult to determine as
stiff glacial tills once eroded may release both granular sediment, silts and clays and some bedload transport
may occur.
Ansari et al. (2002) investigated the influence of cohesive sediments on scour around bridge piers under
steady flows. Ansari et al. proposed a numerical model to calculate the temporal variation of scour depth in
cohesive sediments based on the characteristics of the horseshoe vortex. The shear stress under the
horseshoe vortex was determined through consideration of the initial magnitude of the principal vortex tube,
which Ansari et al. proposed at the sides of the pier was 0.316 times that formed at the nose. Their model
was compared against laboratory data. It is interesting to note that the results of the comparison are variable,
with some test results comparing very favourably with the model, whilst in other tests the model fails to
HRPP653 4
predict the initial onset of scouring well, but gives a reasonable fit towards the peak values. The authors
attribute the model’s poor fit at the start of the tests to dominance of scouring at the sides of the pier initially,
whilst later in the tests scour development at the nose of the pier is dominant. Therefore, the model would
appear to be better able to predict the maximum scour depth than the earlier stages of scour development.
Rambabu et al. (2003) presented results from a limited number of tests using three different model cylinders
with diameters of 50 mm, 90 mm and 110 mm embedded in a silty clay soil. The tests were of short duration
and a hyperbolic function was used to obtain the theoretical maximum scour depth.
Rambabu et al. concluded that the equilibrium scour depth decreased with increasing averaged undrained
shear strength and increased with increasing Reynolds number. Based on their experimental results they
0.5
proposed a general relationship for scour in terms of Froude number Fr, = Uc/(gh) pile Reynolds number,
RD, and soil strength parameter as presented in Eqn. 4 below.
−0.976
0.641 0.640  Cu 
S e = DF r R
D   (4)
 γh 
Where Cu is the undrained shear strength of the soil; γ is the unit weight of the soil; Uc is the current velocity;
g is the gravitational acceleration, and h is the flow depth.
Debnath and Chaudhuri (2010) undertook experiments using clay and clay-sand sediment mixtures and
investigated the effects of clay content, water content, Froude number and applied shear stress on the
equilibrium scour depth and associated scour hole geometry as well the scour development through time.
From their studies Debnath and Chaudhuri concluded that for the soils tested with a water content less than
24 % the equilibrium scour depth decreased with increasing clay content. For water content greater than 27
% and increasing clay content they found that the equilibrium scour depth decreased up to 50 % to 70 %
clay content and then increased with higher clay content beyond these limits (Figure 3). This point of
inflection was a function of the water content of the clay-sand soil mixture.
Figure 3. Variation in dimensionless scour depth against percentage clay content in the sample (after
Debnath and Chaudhuri, 2010)
HRPP653 5
Debnath and Chaudhuri (2010) also noted the shear strength obtained using a shear vane was an important
parameter for describing the equilibrium scour depth in the clay-sand mixtures tested. With respect to the
formation of the scour, it started at the sides of the cylinder and then propagated downstream in all cases.
Under lower shear stress conditions the maximum scour depth was located downstream of the cylinder,
whilst at intermediate conditions the scour formed simultaneously at both the upstream and downstream
sides of the cylinder. At higher values of shear stress the scour development upstream was greater and
faster than that at the downstream side of the cylinder. Debnath and Chaudhuri do not give corresponding
flow conditions and describe the results as relative to the different soil samples and corresponding
geotechnical characteristics of the samples.
Debnath and Chaudhuri (2012) extended their earlier studies investigating local scour around non-cylindrical
piers in clay-sand sediment mixtures. Round nosed, square and rectangular piers were chosen, with
identical widths to the diameter of the cylinder used in the previous tests (0.12 m). They examined the effect
of clay content, water content and shear strength of the clay-sand sediment mixtures on the scouring
process as well as the time development and scour hole geometry.
For the round nosed pier the equilibrium scour depths were typically lower than those obtained for the
cylindrical pier, whilst being greater for the square and rectangular shaped structures. They noted once
again that the vane shear strength is a key parameter in describing the equilibrium scour depth. Debnath and
Chaudhuri (2012) noted three different modes of erosion: Particle-by-particle; in flocs, and “chunk-by-chunk”.
A similar observation was made by Kothyari et al. (2014). For the particular tests conducted the chunks were
of the order of 2 to 5 mm.
Kothyari et al. (2014) investigated scour in the wake region of cylindrical piers using two different sediment
types: fine gravel mixed with clay in proportions ranging between 20 % and 60 % by weight; and fine gravel
and fine sand in equal proportion by weight mixed with clay in proportions ranging between 20 % and 60 %
by weight. They used a range of moisture contents, dry densities, clay content, void ratios and unconfined
compressive strengths in the soils tested.
Kothyari et al. noted in their tests that scour holes formed in their tests were significantly different than those
formed in non-cohesive sediments and, in a similar fashion to earlier studies (e.g. Ting et al., 2001),
significant scouring took place on the downstream side of the piers with, in some cases, negligible scouring
taking place at the upstream side. The maximum duration of the tests was about 15.5 hours, therefore, in the
longer-term (weeks, months, years) it is likely that the scour pattern will continue to evolve, based on
experience from field measurements, however, their results support previous studies which showed scour
development in cohesive soils to form at the downstream side of the pier with the maximum scour depth
often forming behind the pier.
At lower percentages of clay content (≤ 30 %) the scour took place on a particle by particle basis, whilst at
higher percentages of clay content the sediment scoured in the form of thick flakes. It was also noted that
with higher values of unconfined compressive strength and higher percentages of clay content (> 40 %) the
sediment scoured in the form of “chunks”. This “peeling away” of sediment has been observed by other
researchers (Whitehouse et al., 2000). Kothyari et al. also concluded that the clay content and unconfined
compressive strength of the soil were significant factors in affecting the scour depth in the wake region of the
pier. They do not comment on the effect of the granular material within the soil and as to whether its release
in the scouring process led to any weathering/ abrasion of the surface.
HRPP653 6
4. Erosion testing of marine soils

Lefebvre et al. (1986) undertook a series of tests on three eastern Canadian marine clays, investigating soil
specimens with an intact structure, specimens remolded and reconsolidated and specimens destructured by
consolidation. They found that the high resistance to erosion of the natural intact marine clay is primarily in
response to the relatively strong links between clay particles. Whilst the clay structure remains intact the pore
water - soil structure does not have a significant effect since the erodibility of the soil at the particle level is
controlled by the natural structure of the soil. However, remolding and reconsolidation of the samples led to
high erodibility of some samples, whilst other samples remained resistant to erosion. One of the controlling
factors was the salinity of the pore water - soil structure.
Consolidation of intact specimens beyond the pre-consolidation pressure was found, in some instances, to
lead to a significant increase in the erodibility of the soil.
Kamphuis et al. (1990) undertook erosion tests on four natural intact cohesive soils obtained from Ontario,
Canada, using a laboratory flume. The soils were extracted from the field using a rectangular cutting frame
(0.6 m long by 0.15 m wide and 0.1 m deep). The sample and cutting frame were removed together and
waxed. Prior to testing the sample was trimmed flush with the cutting frame and both the frame and sample
were placed in the flume.
In their tests Kamphuis et al. found that erosion of the samples was controlled by the discontinuities present
in the sample rather than the geotechnical properties with the inception of erosion beginning between a
shear stress of 0 and 6 Pa. Erosion of the sample by sand suspension appeared to start once the sand
particles became mobile at around a shear stress of 0 to 2 Pa. Erosion at 0 Pa suggests turbulent
fluctuations play a role at very low time-averaged stresses.
The sand suspension could increase erosion of the sample by up to five times beyond that for the clearwater
case and it was concluded that erosion of cohesive soils may be controlled by the size of the particles rather
than the properties of the cohesive soils in environments were the flows contain solid particles.
In fact, erosion testing of marine soils in the laboratory is generally considered to be the most reliable
approach to assessing field performance. There are number of suitable in situ erosion equipment that can be
used to determine the erosion threshold and erosion rate of the soil. These devices can be applied to barrel
core or box core samples obtained from site. A review of erosion devices has been made by Black and
Paterson (1997) and Annandale (2006). Devices of this kind include ISIS and SedErode (Williamson and
Ockenden, 1996; Mitchener et al., 1996), the Erosion Function Apparatus (EFA) (Briaud et al., 2001) and the
Simulator of Erosion Rate Function (SERF) device (Jiang et al., 2004). Such devices can take a section of
core or alternatively use a remoulded sample. The different approaches to erodibility testing and their
application has been analysed by Whitehouse and Harris (2014). Remolding effects that takes place during
foundation installation may be significant in determining the soil strength, and, therefore, the resistance to
erosion.
Measurements can be carried out on sediment samples from different levels in the seabed to determine the
difference in erodibility with depth. An estimate of the threshold for erosion and the rate of erosion can be
determined from the analysis of the time-series data as material is removed from the surface of the sample.
HRPP653 7
5. Abrasion of cohesive soils

Very few studies have been conducted to assess the effect of abrasion on the erosion of consolidated clay
soils. The role of granular material in erosion of a consolidated cohesive bed has been studied by Kamphuis
(1983). Two modes of erosion were identified, “sand blasting” caused by impact of individual sand particles
moving by saltation and “milling” as a result of sand particles moving as a bed layer. The impact induced
erosion resulted in more severe erosion. The study used a site specific soil and only a limited number of
tests (six) were conducted, but a number of general conclusions can still be drawn from the results. A fluid
containing sand particles will cause erosion of a consolidated cohesive bed at much lower fluid velocities (or
shear stresses) than if the fluid was clear. A maximum (local) erosion rate of 2.3 mm/hour was obtained in
one of the tests, but more typically the erosion rate was of the order of 1 mm/hour. Longer-term rates of bed
degradation of 0.2 mm/hour for consolidated clay (consolidated to 345 kPa) were obtained for velocities in
the range 0.6 to 2.2 m/s at 3 mm above the bed.
6. Evidence from field data

The evidence database on scour, created primarily from monitoring data collected at UK Offshore Wind
Farms (OWF), relates to monopile foundations in different sediment and hydrodynamic environments.
Despite the large number of wind farm developments the evidence base related to cohesive soils in the
marine environment requires further investigation.
Jiang et al. (2004) presented observations of scour development in firm clay adjacent to an oil-unloading
terminal in a tidal river. They noted a scour depth of 5 m, indicating that in some environments stiff clay may
need to be treated as a scour hazard.
Some of the early analysis and reviews of the monitoring data from Round 1 and Round 2 offshore wind farm
sites in the UK are presented in DECC (2008), COWRIE (2010), Whitehouse et al. (2010) and Harris et al.
(2011). Three of the wind farm sites where monitoring data was available for seabed sediments influenced or
underlain by clay are Barrow, Kentish Flats and North Hoyle. In all three cases the foundation consists of a
circular monopile.
The characteristics of the three sites are presented in Table 1 and the scour is described below.
6.1. Barrow OWF

Barrow OWF is a Round 1 wind farm consisting of 30, 4.75m diameter monopile foundations, situated about
8 km southwest of Walney Island in the Irish Sea. Scour was measured at thirteen of the foundations over a
number of years. The first scour survey was undertaken in 2005 and showed that in the glacial till to the
eastern part of the site, low scour depths (up to S/D = 0.04) were measured. There was some indication that
scour depths in the glacial till increased slowly with time following installation (DECC, 2008). Depressions
from the spudcan footings of the jack-up barge used for installation were also visible in the seabed.
HRPP653 8
Table 1. Site characteristics

MTL Tidal range
h Tidal Annual
D SP above Neap (n); Up
Site Description Seabed sediment (LAT) range Hs
LAT Spring (s)
(m) (m) (m) (m) (m) (m/s) (m)
Barrow OWF Open seabed, northwest Fine SAND to
coast of England: exposed muddy SAND,
to waves, moderate some GRAVELS 12 to 4.1 (n)
currents, sand and clay, overlying CLAY; 4.75 No 5.1 0.8 4.9 4.75
18 8.2 (s)
stable seabed environment, exposed CLAY
moderate water depth.
Kentish Flats OWF Open seabed Thames Fine SAND; infilled
Estuary, England: exposed paleo-channel with
to waves, moderate CLAYS and 2.9 (n) *
currents, sand and clay, SANDS; CLAY 5 No 3 to 5 2.74 0.9 3.3 5
4.7 (s)
stable seabed environment, near surface or
very shallow water. exposed
North Hoyle OWF Open seabed, North Wales: Gravelly medium
exposed to waves, SAND or sandy
moderate currents, GRAVEL overlying 4.1 (n)
gravel/sandy gravel, stable CLAY 4 No 6 to 12 5.15 1.17 4.9 4
6.1 (s)
seabed environment,
moderate water depth.
Where: D is the monopile diameter; SP is scour protection; h is water depth; MTL is Mean Tide Level; Up is peak current speed; Hs is significant
*
wave height, and is depth limited at low water levels
HRPP653 9
All thirty foundations were re-surveyed in September 2006. In those areas covered with a thin veneer of sand
the scour depths were limited by the thickness of that layer to scour depths of up to and around 0.5m or 0.1D
in clay sites. Figure 4 shows one of the clay sites with a shallow scour hole of up to 1.25 m in depth.
The key parameters which determine the amount of scour are the composition and thickness of the surficial
and sub-surface sediment layers as well as the prevailing hydrodynamic conditions. The monitoring data
from Barrow OWF demonstrated that for the clay dominated sites scour had been restricted by the thickness
of the surficial layer and the resistant properties of the underlying soils. The depth limited cases generally
lay to the east of the site where the seabed consisted of glacial till.
Figure 4. Example of measured scour hole in sand veneer overlying clay seabed at Barrow – scour contours
at 0.25 m intervals below ambient bed level specified as 0 m
6.2. Kentish Flats OWF

Kentish Flats OWF is located in the outer Thames Estuary, approximately 9 km off the north Kent coast. The
seabed is generally flat and subtly varied comprising mainly coarse sand, but with varying amounts of shell
gravel and small exposures of the underlying clay. Geotechnical surveys also showed the seabed to consist
of variable thickness of sand underlain by soft to firm clays overlying London Clay formation. From the last
available monitoring data at the wind farm site S/D values of up to 0.4 were obtained. In the initial monitoring
surveys, depressions were measured at four of the thirty 5 m diameter turbine foundations in January 2005,
some three months after completion. It is uncertain whether the initial “scour” depression around the turbines
is due to hydraulic scour processes, or whether it was caused by “drawdown” of the soil during foundation
installation or a combination of the two processes. The monitoring data also revealed depressions in the
seabed in response to where the jack-up barge legs had been present during installation, most probably
mainly due to penetration of the legs into the soil rather than through scour processes (see Figure 5).
HRPP653 10
Assuming the scouring is the principal cause of the depressions at the foundations, the maximum measured
depth was less than 0.28D in January 2005, increasing to 0.46D in November 2005 and decreasing again to
0.34D in April 2006. The picture of change was complex as the scour at one location increased with time
during the three surveys whereas the scour at the other three locations increased in the first two surveys and
then decreased in the last survey. Assuming consistency of the surveys, and the time variations were not an
artefact arising from survey error, this suggested that seabed sediment transport processes were able to
produce fluctuations in the depth of the scour around the foundations at this site. Figure 5 shows the scour
development at Turbine E2 as measured in January 2005. The scour depth is around 0.8 m giving an S/D
ratio of 0.16. It is interesting to note that the spudcan depressions to the southwest of the monopile location
have a maximum depth of about 1.4 m.
Figure 5. Scour development at turbine E2, Kentish Flats OWF survey, January 2005. Note jack-up spudcan
depressions to southwest of monopile. (Data collected by EMU Ltd)
6.3. North Hoyle OWF

North Hoyle OWF is located about 7.5 km from the north Wales coast offshore of Rhyl. Within the wind farm
site the seabed sediments generally consist of sandy gravel or gravelly sand with larger patches of gravel
found further offshore. Within these areas the gravels tend to exist as thin veneer overlying sand or boulder
clay. Results from various surveys shows the site as being strongly heterogeneous, having variability over
very short distances and composed of very poorly sorted sediments. The thirty turbine foundations were
installed over the period April to July 2003. Monitoring of the seabed post-installation was carried out over
the period August to October 2004.
HRPP653 11
Some limited scour (less than 0.125 D) was recorded in the 2004 survey at ten of the thirty foundations. In a
survey conducted in April-May 2005 no scour was recorded at any of the foundations. No scour protection
material was placed around the foundations although there was some redistribution of drill cuttings on the
seabed which had arisen during the drill-drive process used to install the foundations. Burial of the inter-array
cables was successful with target burial depths achieved in all but about 3 % of the total cable runs. Where
full burial was not achieved, rock protection was placed during 2004.
Figure 6. shows an example of the survey data for North Hoyle OWF. A mound of drill arisings can be
observed to the southeast of the monopile. There is little evidence of scour development at the foundation.
Figure 6. North Hoyle OWF measured bathymetry at turbine 10, survey 2005
6.4. Scour evolution through time

Scour development around offshore structures is primarily a function of the hydrodynamics, sedimentology
and geotechnical properties at a site. Under tidal flows the current reverses direction with the phase of the
tide and, therefore, scour development will take place, typically, in two directions. The time variation in scour
depth corresponding to the time of installation of the foundation structure and the monitoring survey(s) is
important as there will be a general increase in the scour depth to some equilibrium condition over a time-
scale that is site specific. In non-uniform soils it is possible that the equilibrium condition is not achieved over
the (design) life of the project and, hence, having a reliable estimate of the time-scale in these type of soils
is, arguably, of more importance than that in uniform non-cohesive soils. The time evolution of scour in
offshore sites with non-uniform soils is more difficult to ascertain from available data due to the limited
number of surveys available in time and the more gradual scour development. From the monitoring data for
Round 1 wind farm sites (Barrow and Kentish Flats) the evidence base, as mentioned earlier, suggests a
scour depth that is variable in time and spatially between different locations (Figure 7). The results show a
general growth in scour but also reductions in scour depth from one survey to the next. However, inferring a
general reduction in scour depth over time from this data should be cautioned against as this may just be a
function of the prevailing conditions at the time of the survey rather than some longer-term trend.
HRPP653 12
Figure 7. Variation of dimensionless scour depth with time at Barrow and Kentish Flats offshore wind farms
7. The challenge
The requirement to undertake a scour hazard assessment for offshore projects that involve large volume
installation of foundations, such as those related to offshore wind farm developments, will inevitably mean
that there is a limit to the amount of detailed geotechnical information that can be collected as part of the
project beyond the key requirements for foundation design and in cable corridors. Therefore, there is a need
to have a reliable scour predictor for a range of ground conditions. Harris et al. (2010) undertook an initial
assessment of the Erodibility Index approach for a number of case studies offshore. The approach of
Annandale (1995; 2006) was used to assess the scour potential at three contrasting offshore locations. The
principal attraction of the Erodibility Index method is that it allows for the physical properties of the soil to be
considered and although the method does not directly take into account the chemical properties of the
material, the mass strength number, MS, represents the relative influence of chemical bonding properties of
the soil through the unconfined compressive strength. In principal, the method represents an engineering
methodology that can be applied using information obtained during geotechnical site investigations.
However, to apply the approach to a given location requires a number of key considerations which include:
 The requirement for good information on the soil properties with depth through the seabed, including
grain size distributions, density, undrained shear strength, internal angle of friction, etc from the seabed
surface to the depth (at least) of Smax, the maximum anticipated scour depth.
 Knowledge of the metocean conditions for both typical and extreme events.
Furthermore, the method relies on previously calibrated formulae for the stream power at the seabed and its
variation with depth into the scour hole. There is still a requirement to determine the development of scour
through time in complex marine soils and this requires further research, especially for soils with multi-modal
grading distributions and with distinct layering. It is also important to determine any adjustment to soil
properties that might occur during foundation installation that could affect resistance to scouring and the
effects of abrasion by granular sediments is still not well understood (see Section 5).
HRPP653 13
The key assumption in the Erodibility Index approach is that undrained shear strength can be used as a
proxy for the erodibility of soil. What do we know already? Experimental evidence from the laboratory
suggests that in general:
 as water content increases the scour depth increases;
 scour depth decreases with compaction (density) of cohesive soil, although heavy overconsolidation may
reduce erosion resistance;
 scour depth increases with increasing Froude No.; and,
 scour depth decreases with increasing clay content.
Therefore, erosion is highly dependent on the degree of consolidation and moisture content and
hydrodynamic conditions.
There is also evidence to suggest that erosion of cohesive soils may be controlled by the size of the particles
rather than the properties of the cohesive soils in environments were the flows contain mobile non-cohesive
granular particles and shelly material.
Using available data from a range of offshore sites and laboratory test data where suitable geotechnical
properties are available, it is possible to assess whether the data supports the hypothesis that undrained
shear strength can be used as a proxy for the erodibility of soil. Figure 8 presents an initial review of the
various data shown as dimensionless scour depth (S/D) against undrained shear strength.
Figure 8. Field and laboratory evidence base of scour depth against undrained shear strength
The curve plotted in Figure 8 represents an envelope encompassing all the data only and is not intended to
represent a curve fit. There is a significant amount of scatter within the data. There is an inherent limitation
with the laboratory data as these data represent relatively low strength soils compared with those obtained
from the majority of the field sites. The exception is the bridge scour data presented by Straub and Over
(2010), which is in a similar range of soil strengths to those of the laboratory tests.
There are a number of difficulties with the field data including the representation of the layering effect of the
marine soils. By this we mean that in the majority of cases the depth at which the scour has currently
developed to has been achieved through eroding through a number of different soil strata with different
properties. A decision was made to take the properties of the soil strata to which scouring has currently
reached, but this may be somewhat arbitrary given the possible effects that the overlying layer(s) may have
HRPP653 14
on the erodibility of the underlying soil layer. The values of undrained shear strength represent a pre-
construction state and it is unknown what if any change in soil properties may have occurred during and
post-installation of the structure, for example due to piling. In addition, the results do not have any temporal
element and, therefore, may only be representative of a transient state, and the equilibrium scour depth has
yet to be achieved. The exception is for the two bridge sites in the tidal River Thames (Vauxhall and Putney
Bridges), which lie within the scatter of other data.
Comparing the results from the field with the Erodibility Index method as proposed by Annandale (2006)
suggests that the method is not always conservative (Figure 9). It should be noted that Figure 9 represents
only a subset of the data shown in Figure 8.
Figure 9. Comparison of predicted scour depth (Erodibility Index method) against measured scour depth
It should be noted, however, that in previous studies that have made use of the Erodibility Index method they
have had access to detailed soil analysis including erodibility testing. This is not a feasible option for large
volume installation projects were hundreds of foundations are being installed over an area of tens of
kilometres. Therefore, there may be a requirement to recalibrate the approach based on the evidence base
from the field. Where possible we would recommend some specific testing of soil erosion to increase
confidence in scour estimates in non-uniform soils (Whitehouse and Harris, 2014).
8. Conclusions
Based on analysis of the available field and laboratory data a number of conclusions have been made and
these are presented below:
 A range of soil parameters have been used in previous literature as a proxy for resistance to erosion and
hence scour potential. The majority of these have been based on laboratory data and are untested with
field observations. Evidence from field and laboratory measurements suggests undrained shear strength
can be used as a proxy for the erodibility of soil.
 The Erodibility Index method (e.g. Annandale, 2006) is still one of very few approaches available to
assess scour potential that combines both geotechnical properties and hydrodynamic forcing.
 The relationship may not follow that proposed by Annandale for circular piles.
HRPP653 15
 Use of the methodology as presented by Annandale (2006) may not give conservative answers in the
marine environment. However, it should be noted that previous use of the Erodibility Index method have
used results from detailed soil analysis including erodibility testing.
 The results require further careful analysis through consideration of both geotechnical and hydrodynamic
properties combined with knowledge of how the installation of the foundation may change soil properties.
 Better monitoring of in-situ structures and large-scale laboratory testing will improve the evidence base.
Where possible we recommend this is combined with erosion testing.
9. References
Annandale, G.W. (1995). Erodibility. Journal of Hydraulic Research, 33 (4), 471-494.
Annandale, G.W. (2006). Scour Technology. Mechanics and Engineering Practice. McGraw-Hill.
Ansari, S.A., Kothyari, U.C. and Ranga Raju, K.G. (2002). Influence of cohesion on scour around bridge
piers. J. Hydr. Res., Vol. 40, No. 6, pp. 717 – 729.
Briaud, J-L., Ting, C.K., Chen, H.C., Gudavalli, R., Perugu, S. and Wei, G. (1999). SRICOS: Prediction of
scour rate in cohesive soils at bridge piers. J. of Geotechnical and Geoenvironmental Engng., Vol. 125,
No. 4, April, ASCE, pp. 237 – 246.
Briaud, J.-L., Ting, F.C.K., Chen, H.C., Cao, Y., Han, S.W. and Kwak, K.W. (2001). Erosion function
apparatus for scour rate predictions. J.l of Geotechnical and Geoenvironmental Engineering, 127, pp.
105-113.
COWRIE (2010). A Further Review of Sediment Monitoring Data. Final Report prepared by ABPmer Ltd,
HR Wallingford Ltd and CEFAS for the Research Advisory Group, Project Ref. ScourSed-09, March, 115 p.
Debnath, K. and Chaudhuri, S. (2010). Laboratory experiments on local scour around cylinder for clay and
clay-sand mixed beds. Engng. Geology, 111, Elsevier, pp. 51 -61.
Debnath, K. and Chaudhuri, S. (2012). Local scour around non-circular piers in clay-sand mixed cohesive
sediment beds. Engng. Geology, 151, Technical Note, Elsevier, pp. 1 -14.
DECC (2008). Dynamics of scour pits and scour protection – Synthesis report and recommendations
(Milestones 2 and 3). Final Report prepared by HR Wallingford Ltd, ABP Marine Environmental Research Ltd
and Centre for Environment, Fisheries and Aquaculture Science for the Research Advisory Group,
Department of Energy and Climate Change (DECC) and Department for Environment, Food and Rural
Affairs (Defra).
Harris, J.M, Whitehouse, R.J.S. and Sutherland, J. (2010). Scour Assessment in Complex Marine Soils – An
Evaluation through Case Examples. Proceedings Fifth International Conference on Scour and Erosion,
Holiday Inn Golden Gateway, San Francisco, California, Nov. 7 -10.
Harris, J.M., Whitehouse, R.J.S. and Sutherland, J. (2011). Marine scour and offshore wind - lessons learnt
and future challenges. Proceedings of the ASME 2011 30th International Conference on Ocean, Offshore
and Arctic Engineering, OMAE2011, June 19-24, 2011, Rotterdam, The Netherlands, OMAE2011-50117.
Jiang, J., Ganju, N.K and Mehta, A.J. (2004). Estimation of contraction scour in riverbed using SERF.
Journal of Waterway, Port, Coastal, and Ocean Engineering, 130, pp. 215-218.
HRPP653 16
Mitchener, H.J., Whitehouse, R.J.S., Soulsby, R.L. and Lawford, V.A. (1996). Development of SedErode –
Instrument for In-Situ mud erosion measurements. Report TR17, HR Wallingford, 17pp (+tables and figures).
Kamphuis, J.W. (1983). On the erosion of consolidated clay material by a fluid containing sand. Can. J. Civ.
Eng., 10, pp. 223 – 231.
Kamphuis, J.W., Gaskin, P.N. and Hoogendoorn, E. (1990). Erosion tests on four intact Ontario clays. Can.
Geotech. J., 27, pp. 692 – 696.
Kamphuis, J.W. and Hall, K.R. (1983). Cohesive material erosion by unidirectional current. J. Hydr. Eng.,
Vol. 109, No. 1, January, ASCE, pp. 49 – 61.
Kothyari, U.C., Kumar, A. and Jain, R.K. (2014). Influence of cohesion on river bed scour in the wake region
of piers. J. Hydr. Eng., Vol. 140, No. 1, January, ASCE, pp. 1 – 13.
Lefebvre, G., Rohan, K. and Milette, J-P. (1986). Erosivity of intact clay: Influence of the natural structure.
Can. Geotech. J., 23, pp. 427 – 434.
Molinas, A., Jones, S. and Hosny, M. (1999). Effects of Cohesive Material Properties on Local Scour Around
Piers, J. of the Transportation Research Board, Transportation Research Record, No. 1690, National
Academy Press, pp. 164-175.
Rambabu, M., Narasimha Rao, S. and Sundar, V. (2003). Current-induced scour around a vertical pile in
cohesive soil. Ocean Engng., 30, Elsevier, pp. 893 – 920.
Shen, H.W., Schneider, V.R. and Karaki, S.S. (1969). Local scour around bridge piers. J of the Hydraulics
Div., Vol. 96, HY6, ASCE, pp. 1919 – 1940.
Straub, T.D. and Over, T.M. (2010). Pier and contraction scour prediction in cohesive soils at selected
bridges in Illinois. Research Report ICT-10-074, Illinois Department of Transportation, August, 119 pp.
rd
Terzaghi, K., Peck, R.B. and Mesri, G. (1996). Soil mechanics in engineering practice. 3 Edition, John
Wiley & Sons Inc., 592 pp.
Van Ledden, M., van Kesteren, W.G.M. and Winterwerp, J.C. (2004). A conceptual framework for the erosion
behaviour of sand-mud mixtures. Continental Shelf Res., 24, pp. 1 – 11.
Williamson, H.J. and Ockenden, M.C. (1996). ISIS – an instrument for measuring erosion stress in-situ.
Estuarine, Coastal and Shelf Science, Vol. 42, 1-1.
Whitehouse, R.J.S. and Harris, J.M. (2014). Scour prediction offshore and soil erosion testing. Proceedings
rd
of the ASME 2014 33 International Conference on Ocean, Offshore and Arctic Engineering, OMAE2014,
June 8-13, 2014, San Francisco, California, USA. Paper OMAE2014-24271.
Whitehouse, R.J.S., Harris, J.M., Sutherland, J. and Rees, J. (2010). The nature of scour development and
scour protection at offshore windfarm foundations. Mar. Pollut. Bull., doi:10.1016 /j.marpolbul.2010.09.007.
Whitehouse R.J.S., Soulsby R.L., Roberts W., and Mitchener H.J. (2000). Dynamics of estuarine muds.
Thomas Telford.
Winterwerp, J.C. and van Kesteren, W.G.M. (2004). Introduction to the physics of cohesive sediment in the
marine environment. Elsevier.
HRPP653 17

HRPP653

Uploaded by

Copyright:

Available Formats

HRPP653

Uploaded by

Document Information

Original Description:

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

HRPP653

Uploaded by

Copyright:

Available Formats

Scour prediction in non-uniform soils: undrained shear strength and erodibility

JM Harris and RJS Whitehouse

Scour prediction in non-uniform soils:

3. Scour in cohesive soils and clays

S max = 0.00018RD0.635 (1)

4. Erosion testing of marine soils

5. Abrasion of cohesive soils

6. Evidence from field data

6.1. Barrow OWF

Table 1. Site characteristics

6.2. Kentish Flats OWF

6.3. North Hoyle OWF

6.4. Scour evolution through time

You might also like