7.2.4 Selection of parameters General

NOTE 1 BS EN 1997-1:2004, 2.4.3 requires ground properties to be obtained from test results or from other relevant data. Such data might be, for example, back calculations, empirical or theoretical correlations, field measurements and observations or published data.

The assessment of geotechnical parameters from tests should take account of the difference between the properties obtained from the tests and those that govern the behaviour of the mass of the ground forming the embankment or cutting.

NOTE 2 Potential influencing factors: for slope design the potential for strain-softening behaviour or brittleness can be of particular concern for cohesive soils as a significant loss of resistance can occur if the peak strength is exceeded locally (see 7.2.7 and Table A.1 regarding progressive failure).

Characteristic values for the geotechnical parameters should be selected in accordance with BS EN 1997-1:2004, 2.4.5. The process of selecting the characteristic values from the ground properties may be divided into two stages. First, establish the values of the appropriate ground properties and second, select the characteristic value as a cautious estimate of the value affecting the limit state under consideration taking into account all relevant information. This process is illustrated in Figure 4.

Figure 4 General procedure for determining characteristic values from measured values
General procedure for determining characteristic values from measured values

Derived from BS EN 1997-1:2004, Section 2.

When selecting the characteristic values, due account should be taken of the items listed in BS EN 1997-1:2004,, the following items are of particular relevance to slope design:

  • geological, historical and other background data;
  • the amount of measured data relating to the parameter value under consideration;
  • the variability of the measured data and the degree of confidence in the data;
  • the extent of the zone of ground governing the limit state under consideration;
  • the ability of the ground to transfer load from weak to strong zones; and
  • the consequences of failure at the limit state under consideration.

NOTE 3 Statistical methods may be used to select characteristic values for the ground properties and procedures for the application of statistical methods for this purpose are described in detail by Frank et al [5]. The use of statistical methods implies that sufficient data are available to permit a meaningful evaluation to be undertaken and, where statistical methods are used, BS EN 1997-1:2004, recommends that the calculated probability of a worse value governing the occurrence of the limit state under consideration should be no greater than 5%.

Although the use of statistical methods is permitted by BS EN 1997-1:2004, the use of such methods should be adopted with caution unless a large population of data is available for the geotechnical parameter under consideration. Statistical methods should not be used as a substitute for reasoned judgement of the geotechnical parameter which takes account of all the relevant data related to the parameter.

The design values of geotechnical parameters should either:

  • be derived from characteristic values by the application of a partial factor in accordance with BS EN 1997-1:2004, 2.4.6; or
  • be assessed directly.

Partial factors are shown in NA to BS EN 1997-1:2004 and these factors indicate the minimum level of safety for conventional designs that should be used. An increased level of safety should be specified for unconventional designs or earthworks where the consequences of failure are especially onerous. If design values of geotechnical parameters are assessed directly, the partial factors given in the National Annex should be used as a guide to the required level of safety.

The selection of parameters for design of slopes should consider both the soil grading and, in fine grained soils, the nature of the fines content which is commonly defined by the soil plasticity. It should be remembered that, in UK practice, the definition of material described as "intermediate" differs when used for:

  • a) selection of material parameters for slope stability design (where a soil with more than 35% fines is normally defined as "fine grained"); as opposed to
  • b) classification of the material as a fill (where more than 15% fines is the change point within the SHW [1]), see Table 1c) and 7.6.2 for details.


Figure 5 illustrates two types of soil: one where the minimum conceivable value of soil strength is represented by the critical state parameters ϕ'cv, c'cv (where c'cv will normally be zero) and the second in which very low residual strengths ϕ'r, c'r (where c'r will also normally be zero) can develop at large displacements. These two types of soil may be categorized by plasticity index, Ip (see Figure 6). However, according to the data of Lupini [23], the distinction between turbulent shear and sliding shear for fine soils is not well-defined; there is a transitional zone. The distinction at Ip = 25% is an over-simplification but provides a useful rule-of-thumb. The parameter ϕ'cv will generally lie in the range 30° to 35° for granular fills and in the range 20° to 25° for clay fills.

Figure 5 Variations of ϕ' with displacement
Granular soils and cohesive soils for which Ip < 25%

a) Granular soils and cohesive soils for which Ip < 25%

Cohesive soil for which Ip ≥ 25%

b) Cohesive soil for which Ip ≥ 25%

Figure 6 Variation of ϕ', ϕ'r with Ip
Variation of ϕ', ϕ'r with Ip


  • ● Peak values for glacial tills
  • ○ Residual values for glacial tills and sedimentary clays (σ'n = 130 kPa to 180 kPa) Derived from Hight [24]. Coarse soils (and fine soils with Ip < 25%)

In the case of coarse (granular) soils and fine (cohesive) soils with Ip < 25%, shear box tests taken to large displacement or drained triaxial tests should be conducted until the post peak plateau is identified to obtain ϕ'cv, c'cv; the values of ϕ'cv from these tests are likely to represent conservative values for use in plane strain calculations. Alternatively, an estimate of the plane strain value of ϕ'cv may be based on the plane strain values of ϕ'pk and ψ measured in standard shear box tests, where ψ is the angle of dilation, using the relationship ϕ'cv = ϕ'pk – 0,8ψ (Bolton [25]). Or the plane strain value of ϕ'cv may be estimated from the angle of repose. Fine soils (with Ip ≥ 25%)

For fine soils where displacements are likely to be small, and no pre-existing relic shear surfaces have been detected then it is appropriate to use design values based on ϕ'pk, c'pk in conjunction with the partial factors given in BS EN 1997-1:2004 and its National Annex.

The use of the critical state parameters in conjunction with the BS EN 1997-1:2004 partial factors is likely to lead to over-conservative designs for all soil types where displacements are small; however, the use of critical state parameters (ϕ'cv, c'cv) should be considered where significant displacements are likely to occur over the design life of the slope.

In the case of fine/cohesive plastic soil with Ip ≥ 25%, consideration should be given to whether residual strengths are likely to develop during the design lifetime of the slope. If relic shear surfaces are known to exist, or if sufficient displacement is likely to develop (or has already developed) such that shearing resistance will reduce (or has already reduced) to residual values along any given surface then the design values for the soil shearing resistance should be taken as the residual values. In these cases, large displacement shear box tests (either ring shear tests or repeated standard shear box tests) should be undertaken.

If significant displacement is likely to occur or the soil is brittle, the possibility of progressive failure should be carefully considered.

NOTE BS EN 1997-1:2004, (Design values on actions) and (Design values of geotechnical parameters) provide the option of assessing the design value directly or by derivation from the representative value by the application of a partial factor defined in Annex A. BS EN 1997-1:2004 states "if design values of geotechnical actions (parameters) are assessed directly, the values of the partial factors recommended in annex A should be used as a guide to the required level of safety".

7.2.5 Pore water pressures

The approach selected for pore water pressure monitoring should reflect the quality of data available and accuracy of results required (see BS EN 1997-1:2004, 4.5 regarding the minimum pore water pressure monitoring requirements for different categories of project). That is, the approach selected should be one of the following:

  • ratio ru – experienced based approach, only adequate for general indication of performance;
  • defined groundwater table – overall generalized model determined from standpipe data, adequate for general design purposes; or
  • detailed grid of pore water pressure values – piezometer monitoring required (with data on response time to specific events if appropriate), enables detailed modelling of pore water response to external influences.

The designer should assess which approach and what accuracy of data is required for the project. Having selected the approach to be adopted, conservative pore pressure values should be used in design (see BS EN 1997-1:2004, 11.3, and BS EN 1997-2:2007, 2.1.4 and 3.6).

The effects of vegetation should be considered when selecting design pore water pressures. The designer should consider time of year and proximity of instruments to trees when assessing the monitoring data.

NOTE Figure 7 shows how pore water pressures change with time after excavation of a cutting and after construction of an embankment.

Figure 7 Short and long term stability of embankment and cutting slopes
Stability at cuttings

a) Stability at cuttings

Stability of embankments

b) Stability of embankments


7.2.6 Local and overall stability

When preparing designs for the alignment and slopes of a cutting, the possibility of local slips or falls occurring on the face of the slopes should be considered, in addition to the overall stability against the various forms of failure described in Annex A. Local slips or falls can occur owing to the presence of random pockets of weak, unstable, or water-bearing soils, or thin layers of weak or shattered rocks; in most cases local instability may be dealt with as and when it becomes evident by adopting one or more of the remedial approaches described in 11.5. An overall flattening of the slopes due to the occurrence of these local failures may rarely be justified.

7.2.7 Modes of failure of slopes

There are a number of potential modes of failure of slopes and the designer should ensure that all relevant failure modes are considered (see Annex A).

7.2.8 Influence of construction procedure on slope stability

The designer should be aware that the following construction-related factors can influence slope stability:

  • a) sequence and geometry of excavation, in particular temporary slopes and excavations should not be cut so steeply that ground movement is likely that would significantly reduce the stability of the permanent slope;
  • b) effect of explosives, vibrations from blasting should be considered within the design;
  • c) control of ground water, the potential for groundwater conditions during construction to have detrimental effect on the earthworks should be considered and where necessary appropriate measures shall be incorporated within the works; this might necessitate control of rate of excavation in a pervious water-bearing soil to achieve a gradual reduction in water table, or dewatering techniques to release porewater pressures trapped by low permeability strata;
  • d) control of surface water, shaping the works to prevent water flow or ponding conditions where these are likely to have a detrimental effect on the earthworks; and
  • e) construction of drain trenches at the base of the slope.