Medmerry Realignment Scheme:

Design and Construction of an Earth Embankment

on Soft Clay Foundation

Toru Higuchi † High-Point Rendel Ltd, London, UK

Tony Bruggemann † Atkins Ltd, Epsom, UK
Sunday Obeahon Jacobs UK Ltd, Croydon, UK
John Gosden Jacobs UK Ltd, Reading, UK
Alastair Elder Jacobs UK Ltd, Croydon, UK

Summary
The Medmerry Realignment Scheme will provide increased protection against flooding for the
communities near Selsey on the south coast of England by realigning sea defences up to 2km inland
whilst creating a new intertidal area for improving the region’s wildlife habitats. The new flood
defences comprise a 7km-long embankment constructed from layers of site-won cohesive fill with a
height of up to 5m. Since parts of the earth embankment were founded on alluvial soft clay, an
accurate prediction of pore water pressure dissipation in the foundation was the key to adoption of a
safer and more economical construction method.

Introduction
The Environment Agency (EA) is charged with maintaining the flood defence assets on the Medmerry
coastline, which is located on the south coast of England approximately 10km to the south of
Chichester, West Sussex (Figure 1). The Medmerry coastline extends from Bracklesham in the
northwest to Selsey in the southeast, being one of the stretches of coastline most at risk of flooding in
southern England, and being also threatened by sea level rise. The existing frontage comprises a
shingle flood defence ridge which is at risk of breaching annually.

Following a serious breach event in 2008, realignment of the sea defences inland was selected as the
best solution for this stretch of coast. The Scheme will not only provide a much higher standard of
protection for the local communities, but also allow once-lost wildlife habitats to be recreated in a new
intertidal area seaward of these new defences, which themselves will help absorb the force of waves.

One of the challenges in the design phase was the prediction of the excess pore water pressure
developed in soft clay foundation by the embankment loading since the rate of embankment fill
placement is limited by the rate of dissipation of the excess pore pressures. The length of the
embankment at Medmerry enabled the development of a staged construction approach, which utilised
the space available to keep earth fill operations progressing while allowing natural dissipation to take
place in most areas. This approach limited the need for costly foundation treatment to a short length of
embankment. In addition the observational method was utilised to validate the approach adopted.

Some of the key aspects of the Scheme from the project development through to design and
construction of the earth embankment will be presented in this paper.

Project Description
The frontage at Medmerry has a long history of management intervention. Between 1976 and 1980 a
recharge scheme was completed, placing 230,000m³ of shingle on the beach. In December 1989, the
shingle bank breached and approximately 70% of the imported beach material was lost.
Subsequently, regular and extensive beach management has been implemented by importing beach
material and re-profiling the shingle banks to ensure a suitably high (5m above Ordnance Datum
[AOD]) and wide (5m crest width) barrier each year at an annual cost of between £200k and £300k.
†
formerly Jacobs UK Ltd, Croydon, UK
Notwithstanding this level of maintenance, the banks have breached 14 times since 1994. The most
serious recent event occurred in March 2008, estimated to be a 1 in 20 year event, when a breach
occurred during a major storm, severing Selsey’s only road link (B2145) and isolating the town’s
10,000 inhabitants from the ‘mainland’. This caused over £5M of damage to local businesses and
required the evacuation of some areas due to the risk to life.

Following extensive consultation with local residents and interest groups who have actively contributed
to the design of the Scheme, the Pagham to East Head Coastal Defence Strategy, produced by the
EA, was approved by Chichester and Arun District Councils in October 2008. The preferred strategic
option selected for the Medmerry frontage was Managed Realignment of the existing shingle banks
and to hold this new realigned defence line in the medium to long term. The Scheme is initially to
provide a standard of protection in excess of a 1 in 1000 year event, falling to a 1 in 100 event over
the next 100 years, including an allowance for climate change. The new defences are to be built up to
2km inland from the coast and comprise a 7km-long earth embankment (Figure 1), managing the risk
of coastal flooding to 348 properties and also giving additional protection to key infrastructure in
Selsey including the road link, a wastewater treatment works and electricity substations.

Figure 1 Location and general arrangement of Medmerry Realignment Scheme

Figure 2 Geological section along the line shown on the plan in Figure 1
The Scheme will also enable the formation of 183 hectares of new intertidal habitats and 80 hectares
of new transitional grassland. This will offset losses of internationally designated intertidal habitats
through coastal squeeze, enabling other coastal flood and erosion risk management schemes to go
ahead across the Solent area. While the formation of the intertidal habitats (mudflat invertebrates and
saltmarsh plants) could typically colonise within the first two years, the development of the new
transitional grassland habitats will be ready after one season.

Scheme Objectives
The Scheme has the following key objectives, drawn up through consultation with stakeholders:
• to encourage community participation during development of the Scheme;
• to provide a sustainable flood and coastal erosion risk management scheme; and
• to maximise the extent of Biodiversity Action Plan habitat, making use of existing topography.

Project Inception
In January 2009 Jacobs was selected under the EA’s National Engineering and Environmental
Consultancy Agreement to undertake the project appraisal and, subsequently, was commissioned for
detailed design and site supervision roles in accordance with the framework procurement principles.

Team Van Oord was selected as the construction contractor through the EA’s National Contractors’
Framework and initially collaborated with the consultant under an ‘early contractor involvement’ role
during detailed design to ensure best value in terms of buildability.

Design and construction contracts with Jacobs and Team Van Oord were both procured as NEC
Option C (Target contract with activity schedule). The project value is approximately £20M.

Environmental and Land Ownership Constraints

Field surveys confirmed the presence of protected species, including water voles, great crested newts,
reptiles and badgers. A mitigation strategy developed for the protected species, in consultation with
Natural England to obtain the necessary licences, included fencing and trapping to move water voles
and reptiles away from the immediate construction areas, creating new setts for badgers and
discouraging nesting birds within the construction areas in partnership with farmers and the RSPB.
Within the intertidal area, parts of the ground were raised above the mean high water spring level to
form bird islands (Figure 1).

Approximately 500 hectares of agricultural land were purchased by the EA at a cost of £8.8M to
enable scheme construction including creation of intertidal habitat.

Ground Conditions
Ground Investigations
Four phases of ground investigations for the Scheme were carried out between 2003 and 2011,
including 34 trial pits, 26 boreholes and 58 cone penetration tests (CPT’s). Over 700 disturbed
samples were used for soil/water chemical analyses and geotechnical physical property testing while
98 soil samples were used for geotechnical mechanical property testing to ascertain types of soils and
assess the stability and the settlement characteristics of foundation soil layers.

During construction, soil characterisation tests (natural moisture content, sieve analysis, Atterberg limit
and compaction) were conducted on materials from borrow pits and perimeter drainage channels to
confirm the suitability for embankment fills in accordance with the project’s earthworks specification.

Ground Models
The geology on the Scheme site comprises Quaternary drift deposits (Alluvium) overlying the Middle
Eocene Bracklesham Group Formation, which in turn is underlain by the Lower Eocene London Clay
Formation. The London Clay was not encountered in any of the boreholes drilled to a maximum depth
of 15m during the ground investigations. Figure 2 illustrates an inferred geological section along the
line shown on the general arrangement in Figure 1.

The uppermost layer of the Alluvium, 1m to 2m in thickness, is a firm to stiff crust made of desiccated
clay/silt, which was formed mainly due to fluctuations in the groundwater level. The desiccated
Alluvium is underlain by moist alluvial soils, up to 9m in thickness, comprising a complex interbedded
sequence of very soft to firm clay/silt and very loose to medium dense sand with the occasional
presence of peat and organic clay.

The Bracklesham Group encountered on the Scheme is subdivided into four sub-units with its
stratigraphical sequence being the oldest at the north of the site and becoming younger towards the
south as shown in Figure 2. The material is over-consolidated due to the erosion of later overlying
deposits and generally classified as laminated/interbedded firm/stiff clay and dense sand.

Parts of the earth embankment, approximately 2km in total, are situated on low-lying areas at ground
levels typically 1.5m to 2.0m AOD. These areas are either in close proximity to the coastline or along
existing streams/creeks; their foundations consist of topsoil (< 0.3m-thick) and a 5m to 9m-thick layer
of Alluvium underlain by the Bracklesham Group with the groundwater strike recorded at a relatively
shallow depth (1.7m to 2.5m below ground level [BGL]).

The rest of the embankment, approximately 5km in total, is located on relatively higher ground,
typically, above 2.5m AOD as marked in Figure 1. These areas are frequently used as farmland and
their foundations generally consist of topsoil, directly underlain by the Bracklesham Group, with the
groundwater level typically deeper than 6m BGL. At the locations where existing streams intersect the
alignment of the embankment a 1m to 2m-thick layer of alluvium, comprising soft to firm clay/silt and
medium dense sand, was also present between the topsoil and the Bracklesham Group.

Cut-off Trench
A network of land drains running across farm fields was present at depths typically 0.5m below ground
level. These were intersected by means of 1m-wide cut-off trenches excavated, along the centre line
of the embankment, to 1m depth and backfilled with cohesive embankment fill materials as shown in
Figures 3 and 4.

Fill Material Assessment

The fill material for embankment construction, a total of 400,000m³, was sourced from within the
scheme boundaries. The borrow pit areas were designed to provide optimum fill material recovery
whilst maintaining adequate flow paths for drainage within the area of the finished scheme and
optimising the extent of intertidal habitat as shown in Figure 1.

During the ground investigations, undertaken at potential areas for the borrow pits, three types of
materials were evident: wet, very soft/soft sandy clay (Alluvium); clayey/gravelly sand (Alluvium); and
firm/stiff sandy clay (Bracklesham Group). It was considered that the wet, very soft/soft sandy clay
was not suitable as embankment fill material due to its low strength. Also, granular soils were less
favourable as fill material because of their higher permeability. Hence, the firm/stiff sandy clay was
selected as the main fill material.

Based on Proctor compaction tests carried out on 12 samples, the air void content at the optimum
moisture content for the majority of samples fell between 5 and 10% with a corresponding range of
unit weight between 19 and 20 kN/m³, regardless of the soil type. A set of 6 consolidated undrained
triaxial compression tests with pore water pressure measurement was undertaken to determine a
“cautious estimate” of the effective stress parameters for design of the embankment slopes.

In-situ Permeability of Alluvial Strata

For sections of the embankment founded on a thick layer of cohesive soils, an assessment of the in-
situ permeability of the foundation was required to estimate the rate of dissipation of the pore water
pressures induced by the embankment construction.

Results from 14 oedometer tests on the Alluvium clay samples gave a range of values for the “vertical”
coefficient of consolidation, cv, between 0.3 and 20 m²/year. However, it is known that the use of
small-scale laboratory tests for determining the permeability may give values from tens up to several
1 2 3
hundred times smaller than field test values [RRL (1971) ; TRRL (1973) ; TRRL (1974) ]. These
disparities between the field and laboratory values could result from the presence of natural drainage
paths which are produced by vegetation, varying climatic conditions and other phenomena which
occurred during the depositional history of the strata [TRRL (1973)].

Based on the CPT pore pressure dissipation test results, the “horizontal” coefficient of consolidation,
4
ch, was evaluated using the method proposed by Teh and Houlsby (1991) , giving a value of 10
m²/year for peat and a range of values for Alluvium clay between 45 and 500 m²/year. Taking a ratio of
ch/cv as 1.5 from published data [TRRL (1973)] and applying it to the lowest value of ch, the design cv
value for Alluvium clays was selected to be 30 m²/year. For peat and organic clays, cautious values of
cv and ch were taken as 3 and 5 m²/year, respectively.

Geotechnical Design Parameters

Based on results from the in-situ and the laboratory tests undertaken and study of published data,
cautious estimates of the geotechnical design parameters for design of the embankment were derived
as presented in Table 1.

Table 1 Geotechnical design parameters

Coef. of volume
compressibility,
shear strength,
Plasticity index

consolidation,
Compression
Unit weight,

cv [m²/year]
mv [m²/MN]
resistance,
Undrained

cu [kN/m²]

cohesion,
c’ [kN/m²]
γ [kN/m³]

Effective

shearing

φ’ [Deg.]
Angle of
Stratum

Coef. of
index,
Cc
Embankment Fill 19 40 50 3 30 - - -
Alluvium (desiccated) 17 40 40 0 25 0.15 0.1 60
Alluvium (clay) 17 35 10 0 25 0.3 0.4 30
Alluvium (granular) 17 - - 0 28 - 0.2 -
Alluvium (organic/peat) 12 90 2 2 18 2 2.5 3
Bracklesham Group (cohesive) 19 35 40 0 26 - 0.1 100
Bracklesham Group (granular) 20 - - 0 35 - 0.025 -

Embankment Design
The new flood defence comprised an embankment constructed from site-won cohesive fill as well as
some of the less cohesive material encountered in the borrow pits. The embankment had a height of
up to 5m with a crest width of 4m.

The crest level (5.2m AOD) of the embankment has been designed to withstand a 1 in 100 probability
event, including an allowance for climate change, based on the following current tide levels: MLWS =
-2.3m AOD; MHWS = 2.3m AOD; and HAT = 2.7m AOD. The inclination of the seaward slope of the
embankment has been defined to be 1V:10H so that it will not erode by wave run-up on the high tides
while that of the landward slope has been taken as 1V:3H to maintain its footprint to a minimum.

In this section the method of slope stability analysis and prediction of post-construction settlement as
well as the construction monitoring regime will be described.

Method of Analysis
Since the crest level was constant throughout the length of the embankment, the critical sections of
the embankment with the maximum height were inevitably located in the low-lying areas of weak
foundations with the groundwater table at or near to the ground surface. With the view that these weak
materials would consolidate through excess pore pressure dissipation after construction and gain in
5 6 7
strength [Gosden and Attewill (1985) ; Gosden (1986) ; Huat (1994) ], it was identified that the critical
stability conditions including deep-seated slips were likely to occur during construction. Accordingly,
the design analysis was focused on modelling of the short-term embankment behaviour with
surcharge loading from a working dozer on the fill in order to define safe and economical embankment
geometries and construction sequences. The rate of dissipation at each construction stage has been
8
modelled using the finite difference method developed by Gibson (1958) .

The analysis comprised the following steps:

• Selection of critical foundation sections for analysis
• Prediction of the excess pore water pressure generated in the foundation by the fill placement
• Prediction of the magnitude of the pore water pressure dissipation occurring during
construction due to consolidation of the foundations
• Slope stability analysis at the end of construction
• Settlement analysis during/post construction
The Ultimate Limit State (ULS) stability of the embankment was analysed using limit equilibrium
methods in conjunction with partial factors on soil strength and surcharge load in accordance with
9
Eurocode 7 [British Standards Institution (2004)] . The Serviceability Limit State (SLS) was
investigated as a checking exercise for embankment sections where the factor of safety calculated by
the Eurocode 7 method was close to unity. A SLS factor of safety of 1.3 at the end of construction was
10
taken as the acceptable factor of safety after U.S. Army Corps of Engineers (2000) .

Critical Foundation Sections

For design analysis of the 7km-long embankment thirteen critical sections were selected as being
representative of the foundations, based on their ground conditions and geometries. The length of
each section varied between 50m and 3060m.

Whilst the crest level, the crest width and the gradients of the slopes are constant, the height and the
base width of the embankment are dependent on the existing ground levels. For a range of the ground
levels 1.0m to 4.2m AOD the corresponding height and base width of the embankment had ranges of
4.2m to 1m and 59m to 17m, respectively.

Two of the critical sections are illustrated in Figures 3 and 4 and described below.
• Section at Chainage 6700 (Figure 3)
o Dimensions: Base width = 49.5m, 3.5m of fill to be placed in 3 stages
o Ground conditions: 5m-thick very soft cohesive alluvial soils (cv = 30 m²/year),
overlying interbedded sand/clay (Bracklesham Group)
• Section at Chainage 5600 (Figure 4)
o Dimensions: Base width = 59m, 4.0m of fill to be placed in 4 stages, incorporating a
3m-wide “set-back” terrace on the landward slope at the mid-height to relieve the
loading in the foundation.
o Ground conditions: 6m-thick very soft to soft cohesive alluvial soils including a 3.7m-
thick peat layer (cv = 3 m²/year; ch = 5 m²/year), overlying interbedded sand/clay
(Bracklesham Group)
o Prefabricated Vertical Drains (PVD’s) are installed in the foundation.

Figure 3 Embankment cross section at Chainage 6700

Figure 4 Embankment cross section at Chainage 5600

Prediction of Foundation Excess Pore Water Pressures
The construction of the embankment was modelled by the application of increments of loading
corresponding to a 1m to 4m-thick layer of earth fill per construction stage. Fill placement for each
stage was carried out by compacting multiple layers of 200mm-thick fill. These increments were
assumed to be added at the rate of 0.5m to 1m per week with an 8-week rest period between the
construction stages, giving a range of total construction periods from one week to 32 weeks.

The initial response of the foundation to placement of fill material was calculated in the form of two-
11
dimensional stress distributions using the method proposed by Das (1983) . It was assumed that an
increase in load in cohesive soils can result in an increase of the pore water pressure, and hence, the
amount of the excess pore pressure generated may initially equate to the degree of increased load.

Dissipation of Pore Water Pressure during Construction

The rate of dissipation of the excess pore water pressure is governed by two factors: the distance to a
drainage boundary and the coefficient of consolidation, cv.

The ground investigations indicated that granular soil layers are present at the project site. The
dissipation analysis assumed that these granular soil layers are free draining and will act as drainage
boundaries to the cohesive foundation soil layers. At sections where a granular layer was not present,
a 300mm-thick layer of granular fill was placed underneath the embankment core fill prior to fill
placement.

At a 100m-long section around Chainage 5600, where a 3.7m-thick layer of peat is present, a grid of
PVD’s with the maximum spacing of 1.2m was incorporated in the foundation design to accelerate the
rate of the excess pore pressure dissipation as illustrated in Figure 4. It was expected that a
considerable amount of drainage would occur horizontally to the PVD’s in this particular section.

A spreadsheet, based on the finite difference method incorporating the effect of the PVD’s, was
developed to model the dissipation of the excess pore pressure with time in the cohesive foundation
layers. Typical pore pressure dissipation prediction at Chainage 6700 is illustrated in Figure 5. In the
granular foundation soil layers the excess pore pressure was assumed to be zero.

Figure 5 Typical % pore water pressure dissipation predicted

[Example: two weeks after placing the second layer of fill at Chainage 6700]

Slope Stability Analysis

Slope stability analysis was carried out under both undrained and drained conditions based on the
Janbu method for circular slip surfaces using the Geosolve software SLOPE (Ver.12R.01). It was
considered that circular slip surfaces are the most likely mode of failure for cohesive earth fill
embankments founded on relatively thick deposits of soft ground and constructed.

Since the construction of the embankment was planned to utilise a multi-staged fill placing approach,
allowing the foundation to gain some strength in-between stages, the undrained slope stability
analysis tended to underestimate the stability of the embankment. Hence, analysis was carried out
using the effective stress soil parameters with the excess pore water pressures in the foundation,
based on predictions of the dissipation at each construction stage.

The pore water pressures in the fill were expressed in terms of the pore pressure ratio, ru, which is the
ratio of the pore pressure to the total overburden pressure, and the value assumed for ru in the fill was
0.2.
Design rates of fill placement were determined by calculating the maximum allowable increase in the
embankment height per stage, which maintained adequate slope stability. In general a design layer
thickness that can achieve >60% dissipation of the excess pore pressure over the eight-week rest
period was selected for each construction stage.

Settlement Analysis
Due to the mass of embankment fill that is to be constructed over thick deposits of soft clay, large
settlement is expected to occur. The analysis has rationalised the compressibility by using
consolidation coefficients, Cc, in preference to the modulus of volume compressibility, mv, as the latter
relates to a specific stress range.

Based on the analysis, the predicted maximum total settlement generally ranged from 250mm to
750mm. However, over the section where PVDs were required, the amount of settlement was
estimated to be in excess of 1,000mm. Most of this settlement will take place during construction with
the balance taking place after completion of construction. The post construction settlement was
estimated to range from 50mm to 200mm and can be accommodated by overfilling the embankment.

Geotechnical Instrumentation
Since there were limited data to assess the dissipation of the excess pore water pressures in the
foundation on which the stability analysis largely relied, the observational method was proposed to
control construction at key locations. Instrumentation was aimed primarily at measuring pore
pressures in the foundations to ensure that the predicted levels are not exceeded and to serve as an
early warning system and, secondarily, at allowing the designer to revise the rate of fill placement and
to adopt more economical construction method if measurements showed this were acceptable. The
settlement behaviour of the embankment was also monitored to calibrate the prediction of the post-
construction settlement.

The instruments installed comprised the following:

• Eighteen vibrating wire piezometers with data loggers at three depths per location at six
locations as illustrated in Figures 3 and 4.
• Five settlement profile tubes installed at 1m below the fill/foundation interface at five locations
as illustrated in Figures 3 and 4.
• Two sets of four settlement pins at two locations where a pre-loading method was used to
accelerate consolidation prior to construction of drainage outfall structures.

Details of Construction
In this section some of the challenges encountered during construction and their impacts on the
programme will be described. Also, the behaviour of the embankment observed will be discussed.

Programme
Construction commenced in September 2011 and was originally programmed for 22 months.
However, the weather conditions in 2012, being documented as the second wettest year on record in
the UK, had a significant impact on the progress of earthworks despite the efforts made by the project
team, including trying different construction techniques and equipment. The wet ground surface made
it very difficult for construction equipment to move around the site and saturated materials taken from
the borrow pits needed to be dried out before being placed as embankment fill. Also, additional works
on archaeology and ecological surveys have contributed to delays. As a consequence, the
construction programme has been extended and is due to complete by the end of 2013.

Completion of the majority of the earthworks will be marked by a breach of a 100m-long section of the
existing shingle bank (shown in Figure 1). The breach is to be created by excavation of material from
the existing shingle bank to a level of -1.1m AOD, which represents the MLWN tide level. It is likely
that the breach will now take place in September 2013.

Borrow Pit Design

During the outline design phase the borrow pit locations and shapes were determined primarily by
12
hydrodynamic modelling results undertaken by ABPmer (2010) so that they are to function as
dendritic channels allowing the in- and out-flow of the tide on each tidal cycle, creating further intertidal
habitat. A further review carried out during the detailed design/construction phase modified the borrow
pit areas, based on the geology and information taken from additional ground investigations.
Revised Rate of Fill Placement
Heavy continuous rainfall encountered on the Scheme site in 2012 forced to make a number of
changes to the rate of embankment fill placement. These included the following:
• In most of the low-lying areas removal of the topsoil was reduced to just comprise the near
surface layer containing substantial quantities of vegetation roots since the high groundwater
level prevented dozers from placing the first lifts of fill once the crust made of the topsoil and
the desiccated Alluvium was punctured;
• In most of the low-lying areas the rest period was reduced in exchange for reducing the
inclination of the embankment side slope from 1V:3H to 1V:5H and installing additional sets of
piezometers to more closely monitor the behaviour of the slopes;
• The core of the embankment was placed at a range of moisture contents subject to achieving
an average air voids ratio of 5% and a maximum air voids ratio ≤10% by compaction with a
minimum shear vane strength of 40 kN/m² controlling the fill strength; and
• Trials were conducted using various methods of compaction to establish the optimum
thickness of each compaction layer and resulted in an increase of the layer thickness from
200mm to 300mm, thus, improving the rate of filling and still maintaining the desired density.

Behaviour of the Embankment

Figures 6 and 7 show observed excess pore pressure and settlement against time at Chainage 170
and 6700, respectively. Piezometer readings at both locations, where the thickness of clay layer is in
excess of 5m, indicate that the pore pressures in the foundation rise immediately with fill placement.
While the excess pore pressure generated by the placement of the final layer at Chainage 170
dropped off to the hydrostatic
pressure level over a period of
two months, that at Chainage
6700 only dropped by 40% after
four months. This indicates that
some magnitude of settlement
is yet to take place as the
excess pore pressure continues
to dissipate at Chainage 6700.
Though a thick clay foundation
layer was encountered at both
the Chainage sections, the
presence of sand beds within
the clay layer at Chainage 170 Figure 6 Observed pore pressures and settlement at Chainage 170
accelerated its dissipation rate.

Settlement gauges have

revealed varying amounts and
rates of settlement due to
consolidation of the underlying
strata. For both the Chainage
sections a predicted value of
the total settlement was
600mm. The trend of the
settlement behaviour presented
in Figures 6 and 7 indicates the
actual total settlement is
unlikely to exceed the predicted
value. Figure 7 Observed pore pressures and settlement at Chainage 6700

Diversion of Existing Services

New drainage solution was designed to work with the existing drainage system of rifes as well as
making use of the local topography. Four drainage outfall structures were constructed along the length
of the new embankment.

The 11kV power cable, which runs across a large section of the site, has been diverted by re-routing
around the perimeter of the new embankment.
Archaeological Investigation
The fill materials for the embankment were dug from a number of large shallow borrow pits and
perimeter drainage channels within the Scheme. Prior to excavation, trial trenches (40m long, 3m wide
and 0.5m deep) were carried out for every 1000m² of area excavated. Upon discovery of signs of
archaeology from those trenches extensive archaeological fieldwork was undertaken and discovered a
variety of nationally and regionally important archaeological remains and artefacts, dated from the
Neolithic Age (4000 BC to 2500 BC) to World War II, which were submitted to a local museum
13
[Environment Agency (2012)] .

The archaeological investigation, however, forced changes to borrow pit design due to the
requirement for in-situ preservation of the remains and the amount of fieldwork carried out was far
greater than the level anticipated during the detailed design, which resulted in a significant impact on
the construction programme and, consequently, the cost.

Conclusions
This case study described a new concept for sustainable flood and coastal erosion risk management
undertaken at Medmerry, presenting challenges at both design and construction stages covering the
environment, archaeology, ground conditions and weather delays.

One of the challenges in the design analysis was the modelling of the embankment behaviour, i.e.,
prediction of the excess pore water pressure dissipation, using a simple finite difference method based
on a spreadsheet, which is cost-effective compared with finite element methods.

The instrumentation allowed the designer to revise the rate of fill placement and the duration of the
rest periods, ensuring the adoption of a safer and more economical construction method.

Under exceptionally adverse weather conditions encountered during works, the opportunities for
modifying work sequences and methods were severely limited for earth works on very soft soil
foundations. Use of specialised low ground pressure plant or hard core haul roads were considered to
be prohibitively expensive. At times much of the site was inundated and work was suspended.

Acknowledgements
The authors would like to acknowledge the co-operation and assistance of Mr Colin Maplesden
(Project Manager, Environment Agency) and Mr Howard Hibbard (Supervisor, Jacobs UK Ltd).

References
1
Road Research Laboratory (1971) Embankments constructed on soft foundations: settlement study at
Avonmouth, Report LR 419.
2
Transport and Road Research Laboratory (1973) Embankments constructed on soft foundations: settlement
studies near Oxford, Report LR 538.
3
Transport and Road Research Laboratory (1974) Embankments constructed on soft foundations: settlement
and stability study at Tickton in Yorkshire, Laboratory Report 643.
4
Teh, C.I. and Houlsby, G.T. (1991) An analytical study of the cone penetration test in clay, Géotechnique,
41(1), pp. 17-34.
5
Gosden, J.D. and Attewill, L.J.S. (1985) A comparison of the predicted and the actual performance of an
embankment dam constructed on a soft foundation, Proc. 15th ICOLD Congress, Lausanne, Switzerland, Vol.
1, pp. 1327.
6
Gosden, J.D. (1986) The design of Sungai Bekok dam, British Dam Society, Paper Competition.
7
Huat, B.B.K. (1994) Behaviour of soft clay foundation beneath an embankment, Pertanika Journal of Science &
Technology, 2(2), pp. 215-235.
8
Gibson, R.E. (1958) The progress of consolidation in a clay layer increasing in thickness with time,
Géotechnique, 8(4), pp. 171-182.
9
British Standards Institution (2004) Eurocode 7: Geotechnical design - General rules, BS EN 1997-1.
10
U.S. Army Corps of Engineers (2000) Design and Construction of Levees, EM 1110-2-1913.
11
Das, B.M. (1983) Advanced Soil Mechanics, McGraw-Hill, New York.
12
ABPmer (2010) Medmerry Managed Realignment coastal and hydrodynamic modelling results summary -
Phase 2.
13
Environment Agency (2012) Archaeological finds at the Medmerry site, http://www.environment-agency.gov.uk.