Geotechnical Eurocodes

OBJECTIVES
■■ To describe the Eurocodes relevant to geotechnical design.
■■ To emphasise the responsibilities placed on the geotechnical engineer and the need for close
liaison with other disciplines.
■■ The designer has a mandatory obligation to prepare a Geotechnical Design Report. This includes
the Ground Investigation Report. Input into the geotechnical aspects affecting the construction
phase are also required.
■■ To outline the general approach to risks in geotechnical engineering.
■■ To introduce the concept of limit states and methods of their verification.
■■ To describe the different types of actions and effects of actions.
■■ To explain the derivation of the design resistances, the design approach and the partial factors.
■■ To describe techniques for the determination of characteristic values of geotechnical
parameters.

Relevant Eurocodes consideration of the actions must be made simultane-

ously with consideration of the geotechnical resist-
The relevant Eurocodes for geotechnical design are
ances. This will require adequate (generally more)
described in Table 1. Other geotechnical Eurocodes
communication between the disciplines.
are introduced elsewhere in the relevant chapters of
Equally, engineers who are not specialised in geo-
Soil Mechanics: Principles and Practice. For the
technical engineering should be wary of planning and
latest information, please see http://eurocodes.jrc.
conducting geotechnical investigations and designs
ec.europa.eu/ © European Union, 1995–2017.
given the requirements of appropriate qualifications
and experience and the range of geotechnical design
Personnel situations to consider.

Persons carrying out the design and choosing the

Geotechnical Design Report
relevant structural components such as type of pile,
anchorage and geotextile are required to be appro- This is a mandatory requirement formalising the
priately qualified and experienced in geotechnical assumptions made, data used, methods of calculation
engineering. and the results of the verification of the limit states.
Traditionally, a structural engineer has provided It should describe the ground model and the effects
the geotechnical engineer with the actions (loads) on and of the proposed construction. It should justify
applied by a structure to the ground such as for the the assumptions and design calculations, make recom-
design of a foundation. With the limit state approach, mendations and identify risks.

1
2   Soil Mechanics: Principles and Practice

Table 1   Relevant Eurocodes for geotechnical design

Standard and title Description

BS EN 1990:2002 These describe the terms and definitions, the principles of limit state design,
Eurocode – Basis of the various actions and the verification of the limit state by the partial factor
structural design method.
NA to BS EN 1990:2002

BS EN 1991 In the UK, the structural British Standards, BS 5400 and BS 6399, are
Eurocode 1 – Actions on superseded by this Eurocode. The geotechnical engineer is required to
structures understand how the actions are derived, usually by a structural engineer, so
that the appropriate limit states and partial factors are applied.

BS EN 1997-1:2004 This describes the basis of geotechnical design and the derivation of
Eurocode 7: Geotechnical geotechnical parameters. The supervision of construction, the monitoring of
design – Part 1: General rules the performance of the structure and its subsequent maintenance are now
requirements not always previously addressed by the designer.
There are sections on various aspects of geotechnical design including
spread and pile foundations, anchorages, retaining structures, embankments,
hydraulic failure and overall stability.
All aspects of the geotechnical design are to be incorporated into a
Geotechnical Design Report.

BS EN 1997-2:2007 This describes the planning of ground investigations, soil and rock
Eurocode 7: Geotechnical sampling, groundwater measurements, field and laboratory tests in soil and
design – Part 2: rock, and the requirements of the Ground Investigation Report.
Geotechnical investigation BS 5930:1999, the UK code of practice for site investigations is to be a
and testing withdrawn/revised to remove conflicts.
For laboratory testing, in the UK, BS 1377:1990 will remain the preferred
standard.

NA to BS EN 1997-1:2004 Each country has its own National Annex for each Eurocode but they may
UK National Annex to only contain information on those parameters which are left open in the
Eurocode 7: Geotechnical Eurocodes for national choice, known as Nationally Determined
design – Part 1: General rules Parameters. In this book, the UK National Annex (NA) is referred to.

A Ground Investigation Report must be included Geotechnical risk

as part of the Geotechnical Design Report. This must
include the entire factual field and laboratory investi- It is generally recognised that most of the risks on con-
gations and a full geotechnical evaluation of the pro- struction projects lie within the ground conditions. These
ject, see Chapter 14 in Soil Mechanics: Principles and risks must be identified and managed, not ignored.
Practice for details. EC7 requires that structures are classified into Geo-
The designer is no longer allowed to be divorced technical Categories 1, 2 or 3, as detailed in Table 2. For
from the construction phase. A plan of supervision, each of these categories, a plan of supervision and a mon-
monitoring and checking during construction must be itoring programme is required in the Geotechnical Design
prepared and provided to the owner/client. The checks, Report. For categories 2 and 3 the report may state the
once completed, must be recorded in an addendum to sequence of construction envisaged in the design or it
the Geotechnical Design Report. may allow this sequence to be decided by the contractor.
 Geotechnical Eurocodes   3

Table 2   Geotechnical categories

Source: based on information in http://eurocodes.jrc.ec.europa.eu/ © European Union, 1995–2017

Geotechnical
Description
Category
Small and relatively simple structures for which it is possible to ensure that the fundamental
requirements will be satisfied on the basis of experience and qualitative geotechnical
1 investigations with negligible risk
For example, straightforward ground conditions, local experience, no excavation below the
water table.
Conventional types of structure and foundation
No difficult soil or loading conditions
Quantitative geotechnical data and analyses required
2 Routine procedures for field and laboratory testing
No exceptional risk
For example, spread, raft and pile foundations, retaining walls, bridge piers and abutments,
embankments, ground anchors, tunnels and excavations.

Those structures not in categories 1 or 2

Very large or unusual structures
Difficult ground or loading conditions
3
Abnormal risks
Highly seismic areas
Areas of ground instability. For example, mining, solution, collapsible soils, frost action.

Different regimes for each category are adopted for 22/02 Managing Geotechnical Risk (DMRB, 2002),
inspection of the construction works, quality control, now superseded by HD 22/08, available as a pdf at
checking the exposed ground and groundwater condi- http://www.standardsforhighways.co.uk/ha/standards/
tions and performance of the works in relation to the dmrb/vol4/section1/hd2208.pdf. The main require-
design assumptions. ments of this standard are the provision of Geotechnical
For Category 1 structures, the minimum require- Certification and risk reports prepared by the designer.
ments for ground investigations, design procedures, The risks are assessed at all stages of a project from
construction control and performance evaluation may initial inception to post-construction feedback.
be satisfied by experience, simple inspections and To formalise the process, a Geotechnical Risk Reg-
qualitative geotechnical investigations, provided this ister is produced at project inception and reports are
is agreed with the client. prepared at different stages. These documents remain
For categories 2 and 3, measurements of the ground live throughout the design and construction processes.
properties should be conducted and additional ground In the Risk Register, hazards/risks are identified,
investigations may be needed. Monitoring of perfor- their site-specific causes are recorded and the conse-
mance in relation to the sequence of construction is quences to the project if they occur are assessed in
required with measurements of displacements and terms of health and safety, cost, the environment and
appropriate ongoing analyses. the construction programme.
A risk rating can be applied based on the probability
of occurrence and the impact on the project. Risks with
Geotechnical Risk Register
low ratings may be accepted or monitored during con-
In response to cost overruns on highway projects in the struction while risks with high ratings should be avoided
UK, the Highways Agency introduced a standard, HD by modified designs and/or construction procedures.
4   Soil Mechanics: Principles and Practice

Durability works and the depths of excavations (over-excavation).

These should be included in the design.
Deterioration of materials buried in the ground can
affect their long-term properties. These may include It should be appreciated that the accuracy of the
acid or sulfate attack of concrete, corrosion of steel, geotechnical geometrical data used in a design is
fungal, bacterial or microbial action on timber and only as good as the quality and adequacy of the
degradation of geosynthetic fabrics. ground investigations.
This may result in reduction of the effective geo-
metrical dimensions; for example steel sheet piling in
contact with sea water is designed with a greater thick- Limit states – Ultimate
ness to allow for a sacrificial layer to corrosion. It is necessary to distinguish between ultimate limit
Reduction of the material strength will affect the resist- states and serviceability limit states:
ing capacity of the materials. For example, the strength
of geogrids used to reinforce soil walls and embank- Ultimate limit states are defined as states asso-
ments is reduced to allow for long-term deterioration. ciated with collapse or with other similar forms
of structural failure and concern the safety of
people and the structure.
Geometrical data
These are defined as states beyond which the struc-
A foundation may be designed as 2.5m square, say, ture no longer satisfies the design performance
supporting a central column, but it may not be con- requirements. This could represent a range of limits
structed to these dimensions exactly due to the from:
inaccuracies of construction. In EC7 the effects of
imperfections in the geometrical dimensions of a ■■ Collapse of the whole or part of a structure due to
structural member are considered, and are related to excessive movement such as subsidence, settlement,
the tolerances in construction, for example, the size, heave, seepage force, wall deflection.
shape and location of a spread foundation. ■■ Collapse of the whole or part of a structure due to
ground failure by exceeding the bearing resistance,
Design values of geometrical data, ad, may be repre- sliding resistance or overall instability, such as in a
sented by nominal values anom: slope, embankment or retaining wall.
ad 5 anom (1) EC7 recognises five ultimate limit states:
Where the effects of deviations in geometrical data ■■ EQU – loss of equilibrium of the structure or the
may be significant the design values are defined as ground, considered as a rigid body, in which the
strengths of structural materials and the ground are
ad 5 anom 6 Da(2)
insignificant in providing resistance. This limit state
Da takes account of unfavourable imperfections and is mostly relevant to structural design.
the cumulative effect of simultaneous intolerances. ■■ STR – internal failure or excessive deformation of
For example, a spread foundation is made smaller the structure or structural elements, including, for
than designed and set out in a position such that the example, spread foundations, piles or basement
supported column is no longer central but produces walls, in which the strength of structural materials
eccentric loading. Does the contractor remove the is significant in providing resistance.
offending foundation or can the designer verify that ■■ GEO – failure or excessive deformation of the
the limit state is still not exceeded? ground, in which the strength of soil or rock is sig-
The sensitivity of the design should also be checked nificant in providing resistance.
for the variability of geometrical data that may be ■■ UPL – loss of equilibrium of the structure or the
anticipated in the ground model, such as the levels of ground due to uplift by water pressure (buoyancy)
the interfaces between strata and groundwater levels. or other vertical actions.
In addition, there may be changes to the ground sur- ■■ HYD – hydraulic heave, internal erosion and piping
face level during construction, such as for temporary in the ground caused by hydraulic gradients.
 Geotechnical Eurocodes   5

Ultimate limit state design is then carried out by setting Ed # Cd(4)

up models for the structure with load cases in various
design situations, setting up ground models with design Rd is the design ground resistance available to prevent the
values of geometrical data and geotechnical parame- occurrence of the ultimate limit state and Cd is the limit-
ters, calculation models to determine the design ground ing design value of the relevant serviceability criterion.
resistances and verifying that the effect of the actions Different design actions are determined for these
or load cases does not exceed the ground resistance. two limit states. The effect of the design actions
attempting to exceed the ultimate limit state would
include force, stress, strain, moment and for servicea-
Limit states – Serviceability bility limit states this could include settlement, deflec-
tion, tilt and rotation.
Serviceability limit states correspond to condi- Limit states should be verified by one or a combina-
tions beyond which specified service require- tion of:
ments for a structure or structural element
are no longer met and concern the comfort of ■■ Adoption of prescriptive measures.
people, the functioning of the structure and its ■■ Experimental models and load tests.
appearance. ■■ An observational method.
These are defined as states beyond which the structure ■■ Use of calculation methods.
no longer satisfies the design performance require-
ments. This could represent a range of limits from: Design by prescriptive measures
■■ Cracking which gives an unsightly appearance. These measures involve conventional and gener-
■■ Unacceptable vibrations. ally conservative design rules used where calcula-
■■ Distortion or deflection that leads to loss of weather- tion methods are not available or are not necessary.
tightness or impaired durability or loss of function. Although not stated they would apply to Geotech-
Serviceability limit states can be a little more flex- nical Category 1 structures. The UK National Annex
ible in that they depend on subjective views regard- requires that the use of prescriptive measures should
ing people’s perceptions of function, appearance and be agreed with the client.
comfort. Avoiding these limit states in all events may
lead to poor economy of design so although perfor- Design by experimental models and load
mance criteria must be deemed unacceptable possi- tests
bilities, they should not be unnecessarily severe.
These could include loading tests on shallow spread
The Eurocodes give guidance on the choice of ser-
foundations, trial embankments, piles and anchorages.
viceability criteria but these should be specified for
From the load-deformation or load-pore pressure rela-
each project and agreed with the client.
tionships geotechnical parameters can be back-analysed
and the monitored performance used to justify the full-
Verification of limit states scale design.
The advantage is that a large mass of soil in its in
The approach in limit state design is to verify that: situ stress state undergoes testing. For small-scale
model tests scale effects must be considered and where
■■ For ultimate limit states, the effects of the design the test duration is shorter than the long-term behav-
actions (Ed) do not exceed the design resistance (Rd) iour of the structure, time effects such as pore pressure
of the structure or ground. dissipation and creep should be allowed for.
Ed # Rd (3)
Observational method
■■ For serviceability limit states, the effects of the
design actions (Ed) do not exceed the performance This does not mean design by observation but where
criteria (Cd) of the structure. reliable prediction of the geotechnical behaviour
6   Soil Mechanics: Principles and Practice

would be difficult or uncertain, the behaviour of Models and model factors

the structure is monitored during construction and
A calculation model may consist of an analytical
thereafter.
model, a semi-empirical model or a numerical model
The results from various surveying and geotechnical
and may include simplifications, but it must be either
instruments would be recorded and analysed in order
accurate or err on the side of safety. For this reason, the
to review and adjust, if necessary, the design of the
Eurocode introduces the concept of the ‘model factor’.
structure.
gR;d is the model factor associated with the design
The ‘design’ must include the determination of
resistance, Rd, and gS;d is the model factor associated
the acceptable limits of likely behaviour, for exam-
with the design value of the effect of actions, Ed.
ple the range of deflections of a propped embedded
Model factors are applied to certain aspects of the
retaining wall. This would be checked against the
design of pile foundations. They may be applied when
actual behaviour obtained from a planned monitor-
an innovative method of analysis of a building or
ing programme.
bridge is adopted or the reliability of the calculation
Contingencies for unacceptable behaviour must
method is uncertain.
be predetermined. EC7 makes it clear that during
For buildings designed with conventional calcu-
construction, the monitoring shall be carried out as
lation methods, it is assumed that the partial factors
planned; it shall be assessed and appropriate decisions
quoted in the Eurocodes include the model factors.
made. This may seem obvious but complacency can
lead to dire consequences.
The method is particularly useful where ground–
Actions
structure interaction cannot readily be determined,
such as for excavations for basements. Here the differ- These are all of the factors that may attempt to exceed
ent stiffnesses of the structural elements, the variable the limit state and include, with the appropriate
stiffness of the ground and the state of stress existing symbol:
in the ground will affect the magnitude and distribu-
tion of earth pressures and hence the internal structural ■■ Direct action (F) – a set of forces (loads) applied
forces, bending moments and deflections. to the structure, such as column loads, struts, self-
weight, earth pressures, water pressures, seepage
forces, removal of load, excavation.
Design by calculation ■■ Indirect action (F) – a set of imposed deformations
or accelerations such as moisture variation (such as
This is the process that most designers would expect to
swell/shrinkage of expansive soils, effects of tree
undertake in their working lives. It is the reason why
roots, vegetation), temperature changes (such as
students study the analytical parts of the subjects in
frost action, heated soils), uneven settlement (such
order to become designers. However, in geotechnical
as mining, tunnelling, slope creep, tilt, angular dis-
engineering, information about the ground has to be
tortion), changes in soil composition (such as ero-
sought out by ground investigations and there will
sion, liquefaction, dispersion, solution, degradation,
always be uncertainties remaining.
decomposition, self-weight compression).
The reliability of a design is only as good as the
■■ Permanent action (G) – action that does not vary
quality and quantity of information obtained for each
with time and persists throughout the design life of
site. To reinforce this plea for quality investigations
the structure; for example self-weight of structures,
the following is quoted from EC7:
anchorages, water.
2.4.1(2) ‘It should be considered that knowledge ■■ Variable action (Q) – action that varies with time,
of the ground conditions depends on the extent for example, imposed loads, storage contents, traf-
and quality of the geotechnical investigations. fic, snow, wind, water. Also, combined variable
Such knowledge and the control of workman- actions should be considered.
ship are usually more significant to fulfilling the ■■ Effect of action (E) – the effect on the structural
requirements than is precision in the calculation member and/or the ground from the applied actions,
models and partial factors.’ for example internal force, moment, stress, strain,
 Geotechnical Eurocodes   7

or on the whole structure, for example deflection, Variable actions (Qk) are specified as either:
rotation.
■■ An upper (Qk,sup) or lower (Qk,inf) value depending on
■■ Accidental action (A) – this is an action of short
whether the conditions are unfavourable or favour-
duration but significant magnitude. It is unlikely
able, or
to occur during the design life of the structure but
■■ As a nominal value, usually specified for a particu-
would cause severe consequences unless appropri-
lar use, based on experience.
ate measures were undertaken, for example impact
of a vehicle on a bridge pier, explosions. Variable actions may also be represented (Qrep) as a:
■■ Seismic action (AE) – this is an action that arises due
to earthquake ground motion. ■■ Combination value, given by coQk
■■ Geotechnical action – this is an action transmit- ■■ Frequent value, c1Qk, or as
■■ Quasi-permanent value, c2Qk
ted to the structure by the ground, fill material or
groundwater. An example is earth pressure acting where values of the factors c are #1.
on a retaining wall.
■■ Fixed action – the position, magnitude and direction
of the action are determined unambiguously. Design values of actions
■■ Free action – action that may have various spatial The design value of an action is given by the general
distributions, for example, traffic loading, impact expression
loading.
Fd 5 gF. Frep(6)
where gF is the partial factor on actions.
Design situations and values of actions
Consideration of the ‘design situation’ should be
Effects of actions Ed
made. These are mainly either persistent design situa-
tions, which refer to conditions of normal use, that is, The effects of actions could be in terms of internal
long-term conditions, and transient design situations, force, stress, strain and moment for structural mem-
which refer to temporary conditions such as during bers and deflection and rotation for overall structural
construction or repair, that is, short-term conditions. performance.
There are also accidental design situations and seis- The design value of the effects of the actions, Ed,
mic design situations. is the outcome of a verification procedure, such as
that resulting from a calculation method, using the
design geometrical data, ad, and all of the actions
Characteristic and representative values (generally termed F rep, both permanent and vari-
of actions able) multiplied by their respective partial factors,
The characteristic value of an action is its principal gF, combined with the design values of the geotech-
representative value. In general, therefore, the repre- nical parameters (Xd).
sentative value of an action is given by When the partial factors are applied to the actions
themselves, Ed can be expressed as
Frep 5 Fk(5) Design effect of actions 5 Effect of {factored rep-
resentative actions; factored geotechnical parameters;
For permanent actions (Gk) this value is specified as:
geometrical data}, i.e.
■■ A mean value if the variability is small.
Ed 5 E{gF,i Frep,i; Xk/gM; ad}(7)
■■ An upper (Gk,sup) or lower (Gk,inf) value. This would
refer to a statistical distribution if the variability is where i is the number of actions, which may be
significant. greater than or equal to 1.
■■ A nominal value. This does not refer to a statisti- In some situations it is more realistic to apply the
cal distribution but would be determined from partial factors to the effects of the actions such as
experience. when earth or water pressures are determined, as earth
8   Soil Mechanics: Principles and Practice

pressures determined from factored geotechnical This has then been reduced by an overall factor
parameters may lead to unreasonable design values. to provide for safety and stability. This ‘factor of
Then safety’ has, without any in-depth consideration, also
Design effect of actions 5 partial factor on effect been deemed sufficient to allow for mobilisation of
of actions x Effect of {representative actions; factored strength values, to provide for acceptable deforma-
geotechnical parameters; geometrical data}, i.e. tions and possibly even to cater for durability and
deterioration.
Ed 5 gE E{Frep,i; Xk/gM; ad}(8)
Partial factors were introduced into Danish geotech-
nical practice by Brinch Hansen in 1953 and now form
Design resistances the basis for limit state design in EC7. Together with
characteristic values they draw attention to the sepa-
These can be resistances determined from factored rate consideration of load conditions, material proper-
representative values of actions with partial factors ties and design situations and provide a more robust
applied to: approach compared to the global ‘factor of safety’
■■ The ground properties: Rd 5 R{gF Frep; Xk/gM; ad}(9) method.
A statistical approach to their application is illus-
■■ Or resistances: Rd 5 R{gF Frep; Xk; ad}/gR(10) trated in Figure 1, showing the relationship between
■■ Or both: Rd 5 R{gF Frep; Xk/gM; ad}/gR(11) design loads and design resistances for Combinations
1 and 2 of Design Approach 1.
Partial factors are chosen to ensure that the risk of
Design approach and partial factors failure of the foundation, and consequently the struc-
ture, is minimal so a combination of structural factors
Prior to the Eurocodes, the concept of limit state and geotechnical factors must be considered. These
design and partial factors was common in structural include:
design but not in geotechnical design. The traditional
approach in geotechnical design has been to follow an ■■ Uncertainty of loading
analytical method incorporating reasonable estimates With non-routine buildings and live loading, these
of the load and material parameters to obtain a derived effects are difficult to quantify, for example, wind,
ultimate value, such as bearing capacity. water forces, moving loads, dynamic forces.

Design Approach 1
Combination 2 Combination 1
design load or action Vd

design load or action Vd

design resistance Rd

design resistance Rd
Frequency

load or action

MATERIAL
characteristic

characteristic

PROPERTIES
resistance

LOADS
OR OR
ACTIONS RESISTANCES
∴ the limit state is not exceeded

Figure 1   Verification of the ultimate limit state

 Geotechnical Eurocodes   9

Table 3   Combinations of partial factors for Design Approach 1

Source: based on information in http://eurocodes.jrc.ec.europa.eu/ © European Union, 1995–2017

The limit state of rupture or excessive All design situations except

Design situations for axially
deformation will not occur with either axially loaded piles and
loaded piles and anchors
of these combinations anchors
A1 A1
combined with combined with
Combination 1 M1 M1
combined with combined with
R1 R1

A2 A2
combined with combined with
Combination 2 M2 M1 or M2
combined with combined with
R1 R4

■■ Likelihood of maximum design load The UK National Annex requires that Design
For non-routine structures, it is likely that unfa- Approach 1 is adopted. This provides the partial fac-
vourable variations of loading will occur, whereas tors on the actions, material properties and resistances
routine buildings are often designed on nominal for two combinations, 1 and 2. These combinations are
loading which is unlikely to occur. detailed in Table 3. The values of the partial factors for
■■ Consequences of failure Design Approach 1, combinations 1 and 2, are given in
The public will expect less risk to be taken with the UK National Annex.
structures where failure could result in catastrophic
consequences. More risk is taken with temporary
works than permanent works. Characteristic values of geotechnical
■■ Uncertainty of soil model parameters
Geological variations, inaccuracy of strength values, The assessment of characteristic values of geotechni-
water table fluctuations, mode of failure and limita- cal parameters causes the geotechnical engineer the
tions of the analytical method all provide uncertainty. biggest headache. It is not easy to provide a definition
■■ Extent of investigation of this value. EC7 requires that:
Sufficient depth of ground must be investigated to
assess the layering of deposits, the uniformity of 2.4.5.2(2)P ‘The characteristic value of a geotech-
the ground conditions and a sufficient number of nical parameter shall be selected as a cautious
tests should be carried out to enable a reasonable estimate of the value affecting the occurrence of
choice of parameters. the limit state.’
The more extensive the site investigation, the more Other terms that appear in the literature include:
confidence there will be in the choice of the soil model
■■ Moderately conservative value.
and geotechnical parameters.
■■ Best estimate of the field value.
For the verification of the serviceability limit state,
■■ More adverse than the most likely.
the partial factor applied to the permanent and variable
■■ Pessimistic estimate.
actions and to the ground properties is unity.
■■ Worst credible.
EC7 permits the adoption of three Design
■■ Midway between expected and worst credible.
Approaches, each one determined by different consid-
erations of the actions (A), material properties (M) and You will by now have realised that selecting soil
resistances (R). parameters for design purposes is the most difficult
10   Soil Mechanics: Principles and Practice

yet most important task of the geotechnical engineer. Consideration should also be given at the design
Unfortunately, there are no clear guidelines that can be stage to changes that can be caused by construction
offered to address this problem. A useful discussion activities including soil disturbance, swelling, shrink-
on this subject is given in Simpson and Driscoll, 1998. age, weather deterioration and poor workmanship.
Depending on its effect in the calculation, a charac- For example, the appropriate modulus value for
teristic value may be an upper or a lower value. These a soil retained by a flexible wall will depend on the
are described as higher or lower than the most proba- amount of deflection permitted during construction. At
ble value, respectively. EC7 requires that for each cal- the design stage, it may be assumed as a small strain
culation, the most unfavourable combination of lower value but if large deflections are permitted, a lower soil
and upper values of independent parameters is used. modulus will be available.
For example, lower values of density would be The accuracy adopted in calculation methods should
used where the soil acts favourably in support of the be tempered in the light of the above and output val-
structure, such as passive pressure in front of a retain- ues obtained for, say, settlements or bearing capacities
ing wall. Upper values of density are used where the should be rounded up or down and reported in approxi-
soil applies load unfavourably to the structure, such as mate terms.
active pressure behind a wall.
We should encourage engineers from other dis-
The properties of soils can be determined from:
ciplines to be more sympathetic to the difficul-
■■ Test results, field and/or laboratory. ties faced by the geotechnical engineer so they
■■ Direct determinations such as weight density, g, could accept some of the uncertainties inherent
from triaxial specimens. in the ground. Unfortunately all too often we are
■■ Indirect determinations from correlations, either expected to provide accurate answers to prob-
theoretical or empirical, such as su from SPT N. lems where the ground conditions have been
However, these properties can vary spatially (across poorly investigated.
a site), vertically (with depth) and even for a homoge-
neous soil statistical variation arises due to the vagar- Determination of characteristic values –
ies of obtaining values of the soil properties from field in situ
or laboratory tests on selected samples.
Selection of the characteristic value from the soil In EN 1990:2002 the choice of a characteristic value
properties must take account of: for both loads and soil parameters is directed towards
a statistical approach and EC7 allows for these meth-
■■ Geological and other background information.
ods to be considered for geotechnical parameters. The
■■ Data from previous projects.
characteristic value may be derived based on the frac-
■■ The extent of the ground investigation, number of
tiles of the statistical distribution, Table 4.
boreholes, samples, in situ and laboratory tests.
This approach is appropriate with structural materi-
■■ The variability of the soil property values in situ,
als such as steel and concrete since their properties can
both with depth and spatially, and statistically.
be specified and their manufacture controlled before
■■ The extent of the zone of ground affected by the
incorporation into the works to ensure compliance.
structure.
■■ The presence of weak zones and the ability for
stress transfer to stronger zones either by local soil
Table 4   Statistical approach to characteristic values
yielding or by structural interaction. Source: based on information in http://eurocodes.jrc.ec.europa.eu/ ©
European Union, 1995–2017
There may be differences between the soil property
values and the parameter chosen due to different time Design situation is Use for the
effects, scale effects, mass fabric features and strain unfavourable with a characteristic value
compatibility effects. The results of a ground investi-
gation should not be considered in isolation, valuable low value of the material property 5% fractile
input may be obtained from relevant published data or high value of the material property 95% fractile
local experience.
 Geotechnical Eurocodes 11

Take, for example, one cubic metre of concrete. and tested relatively infrequently so a limited amount
Nearly 300 standard cube specimens could be pre- of information is obtained.
pared and tested to give a representative series of data
Thus statistical approaches to characteristic val-
for statistical analysis and to give confidence in the
ues of soil properties for a whole site are fraught
value of the characteristic strength derived. Providing
with difficulties.
the ingredients and their proportions are maintained
and the placement and construction controlled, the
structural engineer will have confidence in the prop-
Determination of characteristic values –
erties of the concrete incorporated into the structure.
engineered fill
Then consider one cubic metre of soil in the ground.
Up to 500 undisturbed triaxial specimens could be pre- For a man-made structure using an engineered fill,
pared and tested to give a representative series of data such as in an earth dam, where the material properties
for statistical analysis and give confidence in the char- can be specified, then providing the soil material type
acteristic value of this cubic metre of soil. is constant, its remoulded properties are known and
However, this would be a laboratory derived result monitored throughout construction, and the placement
probably requiring adjustment for in situ mass effects and construction are controlled, the geotechnical engi-
such as fissures, differences in the field and laboratory neer will have more confidence in the properties of the
behaviour such as plane strain and triaxial conditions, soils incorporated into the structure.
differences in stress applications and stress effects pro- Each soil will have a characteristic property value,
duced by the curvature of the failure envelope, scale say strength or compressibility, and, by definition, the
effects, time effects and temperature effects. This is geotechnical engineer will accept that no more than
just for one cubic metre of soil. Note also that methods 5% of the soils incorporated may not provide the char-
for such adjustments are not readily available. acteristic value.
In the geological context the ground model can For earth structures it is not common for the designer to
vary dramatically due to stratification, weathering and specify particular materials, rather the materials sourced
stress history. This will produce significant variations by the contractor (for economic reasons) must meet the
with depth and across a site so that in effect many minimum soil property values adopted by the designer.
cubic metres of soil would have to be sampled and The designer must be prepared to reassess the design if
tested. In ground investigations, the ground is sampled the contractor proposes an economic alternative material.

SUMMARY
The Eurocodes relevant to geotechnical design are outlined.
The requirements of the personnel conducting geotechnical design are formalised together with the
requirement of a Geotechnical Design Report. This report must include and justify the assumptions, data,
calculations and limit state verifications.
The report must also include a plan of supervision, monitoring and checking during construction.
The risks associated with the ground conditions need to be recognised in relation to the Geotechnical
Category of the project with appropriate performance monitoring.
The preparation of a Geotechnical Risk Register is introduced as a tool to minimise the consequences to a
project, including health and safety, the environment, cost and programme overruns.
Aspects of durability and construction tolerances are to be considered at the design stage.
The ultimate limit state and the serviceability limit state are described with the means of verification,
including design by prescriptive measures, experimental models and load tests and calculation methods
and, where appropriate, the adoption of the observational method.
The various actions, effects of actions, resistances and partial factors are outlined. The design approach
using characteristic values of geotechnical parameters and partial factors is described.
Characteristic values of geotechnical parameters are discussed.