Proposed electricity generation developments: peat landslide hazard best practice guide

Detailed site assessment

4.1 Overview

The content and structure of a peat landslide hazard and risk assessment ( PLHRA) is ultimately the responsibility of the developer or their nominated consultant/contractor, but should have the following key elements:

• An assessment of the character of the peatland within the application boundary including thickness and extent of peat, and a demonstrable understanding of site hydrology and geomorphology;
• An assessment of evidence for past landslide activity and present-day instability e.g. pre-failure indicators;
• A qualitative or quantitative assessment of the potential for or likelihood of future peat landslide activity (or a landslide susceptibility or hazard assessment);
• Identification of receptors ( e.g. habitats, watercourses, infrastructure, human life) exposed to peat landslide hazards; and
• A site-wide qualitative or quantitative risk assessment that considers the potential consequences of peat landslides for the identified receptors.

The assessment would normally comprise desk study, site reconnaissance, site mapping and probing, hazard and risk assessment and reporting. These steps are considered further below.

4.2 Desk study (review of existing information)

All development sites should be subject to a front end desk study which collates all data relevant to peat instability for the site and provides context to site based survey, mapping and peat characterisation.

The extent of the peat landslide study area should be clearly identified on all maps and plans at the outset of the desk study. The study area should not be restricted to the footprint of the proposed infrastructure but should take in any areas of the landscape that the development might impact on, or that might impact upon the proposed scheme. Typically, the study area will be determined by catchments and topography, sometimes extending downslope and upslope of the application boundary.

Once the extent of the peat landslide study area has been identified, appropriate efforts should be made to collect any and all relevant information relating to the site. The time spent in data collection and review should reflect the nature and scale of the investigation and the volume of information available for the site.

4.2.1 Sources of site information

Sources of information to be considered may include:

(a) Previous site information including technical reports, feasibility reports, and previous ground investigation information;

(b) Geological information, specifically the regional field guide relevant to the site in question and maps of superficial (drift) and bedrock geology;

(c) Soil Survey of Scotland (Macaulay Institute) Soil Memoirs;

(d) Site-wide aerial photography, both contemporary and historical;

(e) Digital elevation data;

(f) Academic literature and publications about the site;

(g) Newspaper archives;

(h) Internet searches, and

(i) Local or site knowledge relayed by landowners, farmers and local residents.

(j) Scotland's Soils website http://soils.environment.gov.scot/

a) to e) provide fundamental site information of relevance not only to peat landslide hazard and risk assessments but later stages of geotechnical design if and when consent has been granted. These should be considered essential sources for the initial assessment. f) to i) provide additional information and are often particularly informative where a history of instability at a development site has already been recognised by landowners, researchers or has been recorded by the local press.

It should be noted that maps indicating peat cover should not be taken as definitive statements on its presence or absence. The depth and extent of peat deposits may vary sharply over short distances as a function of local underlying geology, past and ongoing geomorphological activity and management history. It is for this reason that the desk-study must be informed by a site reconnaissance survey, to ensure that existing information is sufficient and reliable.

4.2.2 Desk based mapping and site modelling

Following review of the information available for the site, a base map of relevant data should be compiled. The data acquired may allow the extent of the peat deposits to be defined and basic geomorphological interpretation to be undertaken. This map would usually be sufficient to provide a basis for site reconnaissance and ground truthing. In some cases, digital datasets can be transferred to desk-based mapping tools such as handheld GPS, tablet computers or portable mapping devices.

4.2.2.1 Aerial photography

Since the previous issue of this guidance in 2006, the availability of digital aerial photography in Scotland has significantly improved. Ortho-rectified digital aerial photography is now available for a majority of the UK from a number of suppliers, including data from a major acquisition for Scotland in 2008/09. Data is readily available for inspection in Google Earth ^TM , including in some areas for multiple time periods (or epochs). The quality of data continues to improve, but in general, 25cm ground resolution is becoming widely available, even in remote areas.

Aerial photography may be used to:

• Identify the presence of existing failure scars and the extent of debris runout;
• Identify pre-conditioning factors for failure (where visible at the resolution of the photography);
• Identify evidence of other pre-development ground conditions of relevance to ground works but not exclusively associated with landslides, including vegetation cover, drainage regime and dominant drainage pathways; and
• Identify evidence for land management practices with the potential to influence ground conditions ( e.g. burning, artificial drainage, peat cutting, forestry).

While contemporary aerial photography is of value in determining current site conditions, historical photography can be particularly useful where:

• There is a requirement to identify features reported on site ( e.g. landslides) that may be decades old; and
• Where the site has been subject to significant changes in land-use and where there is a requirement to understand the site conditions prior to modification ( e.g. where the site has been afforested, burnt or cut over).

Digital datasets can be easily viewed in standard geographical information system ( GIS) software, and compared with infrastructure layouts and other GIS based site-wide datasets such as geology and topography. When using digital aerial photography datasets, the ground resolution of the photography should be clearly stated and the date of capture of the data provided.

Mapping of peat geomorphology (including landslides and erosion), hydrology (natural drainage features), land-use (forestry, artificial drainage) and any other pertinent factors can be undertaken in GIS in order to inform subsequent site visits and the scope of detailed peat probing work.

4.2.2.2 Soil and geology maps

Geological datasets available from the British Geological Survey ( BGS) are typically available for solid (bedrock) and drift (superficial) geologies. The former describe the hard rock underlying the softer overlying materials, while the latter provide detail of the overlying materials (such as alluvium, till or peat). Since issue of the previous guidance, BGS now offer access to much of this information via Geofacets, providing georeferenced geological maps which can be built into GIS and compared with other site data.

The BGS also offer landslide information via the GeoSure ^TM database, which incorporates landslides reported within the National Landslide Database and assigns an indicative landslide potential according to site characteristics (such as geology and slope). Information on collapsible and compressible ground is also available.

As of 2011, Macaulay Institute soils datasets became available digitally via lease or as paper copies from the James Hutton Institute, an amalgamation of the Macaulay Institute and Scottish Crop Research Institute. Data concerning a number of characteristic soil attributes for each soil type ( e.g. organic matter content, bulk density) are also available.

From 2014, a consolidated carbon and peatland map for Scotland was made available by Scottish Natural Heritage. This map was modified following an extensive consultation exercise in 2016 (Scottish Natural Heritage, 2016).

When using digital geological or soil data, the reliability of the data should always be considered, as should the scale at which the map was intended to be viewed. For example, boundaries displayed at 1:250,000 scale would not generally be regarded as reliably located when viewed at 1:10,000 scale in GIS software.

4.2.2.3 Digital topographic datasets

Digital terrain models ( DTMs) generated from LiDAR aerial surveys can provide detailed information on site topography including elevation, slope angle and slope aspect. These data should be used as follows:

• To characterise overall site relief e.g. steep with pronounced convex slopes, or gentle and undulating, and identify topographic controls on drainage e.g. hillslope summits and footslopes, major catchments, sub-catchments and gullies;
• To classify the site into slope classes ( e.g. 0-5°, 5-10°) on the basis that certain slope ranges may be more or less susceptible to specific failure mechanisms (see Section 2.1); and
• To identify north and south-facing slopes on the basis that slopes with differing aspects may have differing hydrological characteristics in relation to sun exposure ( e.g. rates of snow melt).

As with other digital datasets, digital terrain models are now widely available from a variety of suppliers, or can be flown by commission if required. Datasets are normally geo-referenced and can be layered in a GIS with the mapping and aerial photography datasets described previously.

When using digital topographic datasets, the resolution of the data should be stated ( e.g. 5m) and any limitations of the data resolution for the analysis clearly stated. For example, 25m or 50m ground resolution may be insufficient to pick up small scale but critical variations in topography on sites of quite variable relief.

4.2.2.4 Other digital mapping datasets

A large number of digital mapping datasets have become available as digital data delivery has become more popular. Some of these may be relevant to the assessment of peat landslide hazards, including hydrological features, the position of existing infrastructure, the location of environmentally designated sites (such as Special Areas of Conservation or SSSIs) and forestry plans. In most cases, this data is collated as part of the wider EIA process and should be made available to the PLHRA team.

4.2.2.5 Remotely sensed imagery

Data collected by remote sensing includes aerial photographs, digital topographic datasets ( e.g. from NEXTMap) and multispectral datasets illustrating ground conditions ( e.g. moisture content). Until recently, earth observation applications to peat landslide investigations have relied upon the interpretation of aerial photographs, with satellite imagery lacking the spatial resolution required to provide detailed images at the scale of an individual landslide. However, increasingly satellite and airborne technology offer opportunities to investigate and map individual peat landslides, and susceptible terrain. The main satellite and aerial imagery sources which may be applicable to landslide investigations are summarised below and covered in more detail in Appendix A:

• RGB digital camera imagery and hyper-spectral imagery (Unmanned Aerial Vehicles): for high resolution photography of ground conditions, high resolution digital elevation models and assessment of water content and forest condition;
• Optical satellite imagery (Landsat thematic mapper): for identification of flow tracks, ground fissures and subtleties in peat morphology;
• Microwave (Synthetic Aperture Radar Interferometry, InSAR): for vegetation type, moisture content and collation of digital elevation models;
• Multispectral video: for mapping of groundwater systems; and
• Hyperspectral scanners: mapping of geological units in areas of poor exposure using soil moisture content as a proxy, estimation of soil thicknesses prone to landsliding.

Where pre-existing datasets are available, these can be of value in understanding site conditions. However, commissioning of such datasets for a single scheme would normally be considered cost prohibitive.

4.2.3 Listing of data sources

Given the variety of datasets available, it is important that the PLHRA provides a clear listing of all data sources referred to during preparation of the desk study, including where relevant the age of the dataset ( e.g. for aerial photography) or where appropriate, its resolution. Not only does this demonstrate knowledge of the validity of the data, but where schemes are revisited in later years ( e.g. for wind farm repowering) it provides a data benchmark with which subsequent data acquisition can be compared.

4.3 Site reconnaissance survey

Site reconnaissance should be undertaken early in the EIA process to verify the features identified during the desk study, and to enable an interpretation of the site in the context of the surrounding environment. Ideally, this should be undertaken subsequent to review of aerial photographs of the site and any associated mapping.

Reconnaissance level peat depth probing should be undertaken to verify the presence of peat on published data sources (such as BGS drift maps) and check whether peat is more widely distributed than desk study data suggests, as can often be the case.

The reconnaissance survey may also provide helpful information on ease of site access ( e.g. accessibility of forest stands for peat probing or the location of particularly wet or dangerous ground), information which could be of value in scoping and planning more intensive peat probing work.

On the basis of the reconnaissance survey and validation of the desk study, ground conditions assessment (or detailed site survey) can be planned and implemented.

4.4 Ground conditions assessment

Ground conditions assessment should provide sufficient information to enable peat landslide hazard and risk assessment to be undertaken. Following this work:

• Any site evidence for past landslides or incipient instability should have been accurately recorded (including details of the field evidence and the locations observed);
• Site hydrology ( e.g. presence of flushes, soil pipes) and peat geomorphology (un-eroded, eroded, type of erosion) should be well understood and preferably summarised on maps, with the hydrological baseline sufficiently well understood to enable water table levels and their impact on stability to be modelled;
• The impact of land-use on natural hydrology and peat thickness should have been ascertained ( e.g. maps of drain distribution or locations of forestry or cuttings);
• Sufficient peat depth data should have been collected to enable characterisation of peat depth across the site, and in further detail at infrastructure locations;
• The character of the peat ( i.e. geotechnical properties, humification, wetness, nature of contact with substrate and the nature of the substrate itself) should have been determined sufficiently to enable stability analysis to be undertaken (if this approach is taken in the hazard and risk assessment).

The key objective of the ground conditions assessment is to obtain sufficient and reliable information in support of the PLHRA. This section provides recommendations for site mapping and ground investigation where peat is known to be present at a proposed development site. It does not cover the standard geotechnical investigations that would normally be required for development planning as specified in current standards ( e.g. BS EN 1997-2:2007 (EC7-2) and associated National Annex ( BSI, 2007); BS 5930:1999+A2:2010 ( BSI, 2010); BS 1377 (Parts 1 to 9) ( BSI, 1990).

4.4.1 Site mapping

Peat landslides and the pre-failure indicators which may suggest their future occurrence are geomorphological features, and as such should be represented on a geomorphological map of the site. At its most basic, a geomorphological map ( e.g. Cooke and Doornkamp, 1990; Griffiths, 2001) should show:

• The position of major slope breaks ( e.g. convexities and concavities);
• The position and alignment of major natural drainage features ( e.g. peat gullies and streams);
• The location and extent of erosion complexes ( e.g. haggs and groughs, large areas of bare peat);
• Outlines of past peat landslides (including source areas and deposits), if visible; and
• The location, extent and orientation of cracks, fissures, ridges and other pre-failure indicators (see section 2.3).

In addition, non-geomorphological features can be mapped where they have relevance to peat instability:

• The position and alignment of artificial drains (or grips);
• Turbary (deep cuts in peat associated with harvesting for fuel); and
• Forest stands (and their condition).

The resulting geomorphological map, when compared with a slope angle map derived from a digital terrain model and with solid and drift geology information should provide a basis for zoning the site into areas of similar character, and where peat landslides or indicators are present, determining what site-specific factors control their distribution.

The most effective way of preparing a geomorphological map is to undertake mapping from high resolution aerial photographs in GIS software. Features mapped at the desk study stage can then be verified in the field, and any features not visible at the scale or resolution of the photography can be added to the map. Alternatively, if aerial photography is not available, or is of insufficient quality, GPS survey can be used to demarcate areas of similar ground condition ( e.g. Dykes, 2009).

Preparation of these maps relies upon the skill, knowledge and experience of the interpreter in accurately recording features specifically associated with peat instability, including peat landslides themselves. Good examples of the mapping of peat landslide morphology can be found in Higgitt and Warburton (1999) and Wilson and Hegarty (1993).

4.4.2 Ground investigation

This section provides guidance on the general principles behind planning a ground investigation in support of a PLHRA, and then summarises a range of intrusive and non-intrusive approaches to characterising ground conditions within a development area.

Ground investigations ( GI) in peatlands serve a number of purposes, particularly in relation to the planning and siting of wind farm infrastructure:

• Ground investigation data may be used to identify the site baseline conditions as a context to environmental impact assessment ( e.g. the depth and extent of peat), and this may include association of specific habitat types with specific peat thicknesses;
• Peat depth data may be used in calculation of a mass balance for the site, i.e. quantifying the amount of peat that may be excavated as compared with the amount that may be reused (such as in a Peat Management Plan, e.g. Scottish Renewables and SEPA, 2012); and
• Geotechnical data may be used to characterise the properties of peat (and the substrate) to inform stability analysis.

Guidance in relation to characterising the geotechnical properties of soils on proposed construction sites is specified in Eurocode 7 - 'Geotechnical design - Part 2: Ground investigation and testing' ( BS EN 1997:2-2007; BSI, 2007). Eurocode 7 defines the scope of ground investigations at three levels, corresponding to successive stages in a construction project: i) preliminary investigations for positioning and preliminary design of the structure, ii) design investigations, and iii) control and monitoring. The planning stage of a wind farm development corresponds to a preliminary investigation, the scope of which is specified in Section 2.3 of Eurocode 7. This document does not offer guidance on design investigations, which would normally occur post-consent.

A preliminary investigation should be sufficient to assess the general suitability and stability of the site, its suitability in comparison with alternative sites, suitable positioning of structures, the possible effects of works on their surroundings, identify borrow areas, consider possible foundation methods and ground improvements, and enable planning of design and control investigations. In addition, estimates of soil data including stratification, pore pressure, strength properties and presence of contaminated ground or groundwater should also be possible from the preliminary investigation.

GI sampling locations should be optimised using the findings of the site reconnaissance and geomorphological mapping and should reflect the nature and extent of the proposed construction works. The Scottish Government (Scottish Government et al., 2014) provide information on the level of detail expected for site investigations on peatlands, which suggests both a site-wide density of approximately one probe per 100m (or one probe per hectare) supplemented with significant additional probing at infrastructure and along tracks.

4.4.2.1 Scoping ground investigation

A competent person should be responsible for identifying and justifying the numbers, locations and types of sample collected, and this will depend upon the size and variability of the development site. Grids, transects, random or targeted sampling strategies may be appropriate. In relation to infrastructure, sampling or probing locations should be considered for:

• All proposed turbine locations, possibly using a gridded sampling approach within the extent of the micro-siting allowance;
• Access tracks, with sampling along the centrelines and more widely within the area covered by the micro-siting allowance.
• Major areas of hardstandings ( e.g. site compounds); and
• Borrow pits.

The following additional factors should beonsidered during scoping of the ground investigation programme:

• Topography and morphology: peat depths are likely to be shallower on steeper slopes, and therefore sufficient samples/probes should be taken to reflect the range of slope angles identified over the development site;
• Vegetation: the physical characteristics of peat will vary according to their hydrological setting, usually reflected in surface vegetation, samples should be taken to reflect the range of major vegetation types ( e.g. heathers, mosses, grasses);
• Hydrology: the hydrology of the site should be recorded and mapped where possible including any evidence of surface and subsurface drainage pathways and the depth of water strikes encountered during peat probing; and
• Land management: peat will also vary according to local land management practices, with peat that has been subject to burning, draining or cutting exhibiting differing characteristics to adjacent undisturbed peat.

In general terms, sufficient sampling locations should have been investigated to produce an outline map of variability in peat depth across the development site (to inform layout iterations).

4.4.2.2 Peat depth probing and coring

Peat depths may be determined by coring and by probing. Coring involves the retrieval of material in a chambered sample ( e.g. a hand auger, gouge or 'Russian' type corer, Aaby and Berglund, 1986) and enables both the depth of peat and the variation in its character with depth to be observed and logged.

Probing may be undertaken using steel (or other) rods and does not involve the retrieval of material. Typically an avalanche probe or survey poles may be used. The depth at which the peat ends and the substrate ( e.g. rock or clay) begins is usually determined by a change in resistance (or rate of refusal) of the probe, or by a change in sound e.g. a granular sound as mineral material is encountered. Inevitably this is less accurate than coring for determining the depth of peat, particularly where the substrate is soft ( e.g. clay) and the change in resistance is subtle. Furthermore, no insight is given into the characteristics of the peat ( e.g. its humification or wetness).

The balance of probing and coring undertaken during site work should be considered during scoping of the programme. Typically, probing may be undertaken more rapidly than coring. Where time or budget is a key driver, allocating a certain proportion of sample locations to coring whilst undertaking the majority of investigation with probing may be acceptable.

Dependent on the probing technique, all materials encountered and depths of changes in strata should be logged and recorded. Logging should be undertaken on-site or samples removed from the site and logged remotely using the techniques described in Section 4.4.2.3.

Dynamic probing can also be used to provide information on peat depth and variability in strength with depth through the stratigraphy.

4.4.2.3 Logging of peat stratigraphy

Peat deposits form over many thousands of years through the slow decay and accumulation of successive layers of organic debris. As these layers build up, the prevailing climatic conditions and their influence on moisture content lead to decomposition of peat fibres (or humification), and compaction. Many UK peat uplands are characterised by fibrous upper peat layers overlying more humified amorphous basal peat, underlain by either weathered substrates of tills and/or bedrock.

The degree of organic content and the high percentage moisture content of peat materials mean that standard logging using BS5930 ( BSI, 1999) is not suitable to interpret and record the detailed differences between peat layers. Instead, it is advised that logging of the peat is conducted under two logging systems:

• Troels-Smith, as outlined in Long et al. (1999); and,
• Modified Von Post classification, as outlined in Hobbs (1986).

Logging should be carried out using both systems wherever trial pits and augured or cored samples have been extracted and where peat sections are exposed by existing cuttings or on naturally exposed and free draining faces such as gully sidewalls. The advantage of a combined system is that Troels-Smith affords logging of mineral deposits (washed fines, till layers) where von Post does not, and hence the use of both systems enables effective characterisation of the peat, the substrate, and the peat-substrate interface, a critical layer with respect to peat instability.

Otherwise, materials identified on site that are not peat should be described in accordance with the most current and relevant British or Eurocode standard.

4.4.3 Other ground investigation techniques

4.4.3.1 Non-intrusive (geophysical) techniques

Geophysical survey techniques may provide an alternative non-intrusive method for investigating peat environments. These technologies measure the vertical and lateral variation of physical properties of the ground. They are particularly useful in areas where biodiversity and other environmental issues preclude the use of more invasive methods (such as trial pits and drilling). Geophysical methods are generally used to support more traditional intrusive techniques and should be correlated with intrusive investigation to improve confidence in results.

Variations in physical properties may be collected along horizontal profiles to identify vertical variability of a physical property, or on a grid basis to allow contour plots of data variation over the development site (and therefore identification of local anomalies or spatial continuity). Data collection techniques for geophysical survey are outlined in Appendix B.

4.4.3.2 Intrusive investigation - trial pits

Trial pits provide an opportunity to log continuous sections of peat stratigraphy and extract representative, undisturbed block samples for subsequent laboratory testing. Detailed methods for shallow and deep trial pitting are discussed in detail in BS5930 ( BSI, 1999). Where possible, trial pits should be carried out using a tractor mounted excavator and should be dug to the level of the underlying substrate (if bedrock) or slightly into softer substrate materials ( e.g. till) in order to identify the presence of impermeable horizons at the top of the substrate.

Frequently, issues of site access and generation of potential instability prevent the safe use of excavators. In these circumstances, trial pitting should be undertaken by hand or alternatively hand augering techniques should be used. Further details on the safety consideration for excavations are included in the AGS Safety Manual for Site Investigations (2002). Consideration should be given to the suitability of the location for infrastructure if it is not suitable for plant access. All trial pits should be logged, photographed and back filled using the appropriate methods described in BS5930 ( BSI, 1999). Block samples should be taken from the walls of the trial pit where it is safe to do so. In this instance, disturbed samples may be collected from the representative excavated material. Importantly, after back filling, the surface vegetation mat must be re-laid to promote recovery of habitat

Consideration should be given to practical access routes. Peat covered areas often contain materials of variable compressive strength and vehicles may easily become stranded. Prior to the ground investigation fieldwork, the proposed access route should be checked and agreed in order to minimise damage to the peatland. The site team should be made aware of any potentially hazardous ground conditions and alerted to health and safety constraints.

4.4.3.3 Site instrumentation and monitoring

Instrumentation at the development site both during and subsequent to ground investigation can be of value in monitoring groundwater levels, characterising hydrological responsiveness of peat layers and identifying precursory slope movements at tension crack locations. Ideally a minimum twelve month cycle of monitoring is required to identify seasonal variability in hydrological conditions, and in turn, rates of slope movement. The shorter the monitoring the period, the less representative the data will be of longer-term trends and extreme responses. However while twelve months may not be possible, monitoring can continue beyond the investigation and even during determination.

The need for monitoring will depend upon ground conditions identified at the site. For example, if a planned turbine site is situated downslope of a pipe network or within a flush or particularly wet zone, groundwater monitoring may be valid. If the turbine site is situated in an area of cracking, pegs may be required to monitor potential crack displacement. A minimum monitoring period of 8 weeks would be anticipated. In areas where ground movement is possible, the monitoring would be longer and comprise a baseline survey and permanent monitoring network such that if movement were to occur, it could be accurately determined from retrospective surveys. Where the period is shorter, justification should be provided.

Appendix C identifies variables relevant to slope instability and the instrumentation available to monitor them.

4.4.3.4 Representation of peat depth data

Once peat depth data has been collected, the data can be portrayed on site plans in a number of ways:

• As colour coded points, with colours corresponding to specific depth ranges ( e.g. 0.0 - 0.5m, 0.5, 1.0m, etc);
• As contour lines of peat depth (or isopachs) with each contour representing a specific peat depth ( e.g. 0.5m, 1.0m, etc); or
• As filled contours of peat depth, with each shaded area representing a peat depth range between contours ( e.g. 0.0 - 0.5m, 0.5 - 1.0m).

In order to derive contours or filled contours, application of an interpolation process in GIS is normally required. Interpolation uses the values from existing data points to predict values at unsampled locations. Interpolation is valuable as it provides information on likely peat depths between sampled points, and the resulting interpolated surface can be used in a GIS package alongside other data to estimate risk levels at areas of the site that have not been directly sampled for depth.

There are a number of interpolation methods available, including natural neighbour, inverse distance weighting and kriging. If these methods are applied, it is important that any user defined inputs (other than peat depth) that may influence the resulting peat model are stated, and any limitations of the model clearly outlined. The temptation to 'over-interpolate' across large distances between real data points, or to extrapolate beyond the area covered by real data should be avoided.

In order that the model be considered reliable, sufficient coverage of data points should be available in areas covered by the model, hence infrastructure targeted probing (with limited coverage elsewhere) may be inadequate to produce a site wide peat model.

4.4.4 Laboratory testing

Peat landslides generally exhibit a shear zone or shear plane within the lower layers of the peat mass, at the interface with the substrate ( e.g. till or bedrock), within the substrate, or in a combination of these layers. As such, sufficient samples must be collected to describe the geotechnical parameters of both the peat mass and the mineral substrate.

While laboratory testing for mineral soils is well understood, peat 'soils' are composed of vegetative (organic) matter in various states of decomposition, rather than mineral particles. As such, conventional geotechnical theory based on mineral soils does not apply well to peat ( e.g. Gosling and Keeton, 2008). For example, the characterisation of cohesion, c, and the angle of internal friction, ϕ, is notoriously difficult due to the inherent variability in peat arising from its vegetational composition and degree of humification. Both of these values are critical in slope stability analysis, however, the behaviour of well humified peat samples can approximate the behaviour of clays (Evans and Warburton, 2007). For more fibrous peats, characterisation of geotechnical parameters is more difficult.

Section 4.4.4.1 provides a list of appropriate tests which should be considered to aid the characterisation of both the peat and underlying substrate in order that appropriate parameters can be incorporated into the site ground model.

4.4.4.1 Physical properties of peat

The following physical properties may be of value in characterising peat. Tests (iv) to (viii) are most relevant to characterising material properties of the substrate:

(i) Moisture content;

(ii) Bulk density;

(iii) Organic content (Loss on Ignition);

(iv) Plastic and liquid limit;

(v) Specific gravity;

(vi) Particle size distribution;

(vii) Shear strength parameters (c and ϕ);

(viii) Soil pH and sulphate content - (if concrete design is a consideration);

(ix) Linear shrinkage; and

(x) Fibre content.

Tests should be carried out in accordance with BS1377 ( BSI, 1990a; 1990b) except where superseded by BS EN 1997-2:2007 ( BSI, 2007). Details of the relevant standards are provided in the UK National Annex to Eurocode 7 ( BSI). Note that some variations are required in certain test procedures to account for the highly organic nature of peat materials. Key variants on these tests are described below.

Drying for moisture content determinations should be undertaken at less than 50°C to avoid charring the samples. Bulk density and linear shrinkage determinations should be carried out on peat from block samples, to ensure the moisture content and volume of the peat are preserved. Organic content (loss on ignition) tests should be performed at a maximum temperature of 375°C to avoid any weight losses associated with loss of water from clay minerals (Ball, 1964).

Particle size distribution tests can give misleading results, particularly for fibrous peats, and should be treated cautiously if used in geotechnical models and calculations. The use of fibre content tests, to be carried out in accordance with ASTM D1997-13 (due to a lack of equivalent British Standard) provides an additional and useful measure of peat composition and an indicator of potential tensile strength.

All shear strength tests should be performed on undisturbed samples taken from intact block samples of peat, substrate and peat/substrate interface as considered appropriate. Soil water samples should be collected from trial pits to determine pH. Shear tests can then be conducted with water pH equivalent to the natural condition.

Hobbs (1986) provides further useful practical advice on the applicability of standard index tests to peat soils, however caution should be exercised in any interpretation of ground conditions based upon these tests.

4.4.4.2 Shear strength tests in peat

Deriving credible shear strength parameters from conventional laboratory and field testing methods has been identified as problematic (Evans and Warburton, 2007; Gosling and Keeton, 2008; Winter et al., 2005). Available testing results should be analysed with reference to previously published parameters ( e.g. Carling, 1986; Warburton et al., 2004; Dykes et al., 2008). Although shear vanes can be used to derive site specific in-situ results, there is some doubt as to the reliability of these tests given fibre/vane interactions during testing (see Long and Boylan, 2012).

Recent testing developments such as T-bar and ball penetrometer testing ( e.g. Long and Boylan, 2012) may provide more accurate shear strength characteristics for peat materials, although these methodologies are still in their infancy. Given the uncertainties inherent in both established laboratory investigation techniques and in more recent innovative approaches to testing organic soils, this guidance does not seek to be overly prescriptive as to the types of testing to be undertaken. Rather, it is the responsibility of the competent person(s) to clearly state:

• The reasoning behind the schedule of laboratory tests;
• The uncertainties associated with the results in relation to the quality of the sample(s), the composition of the sample(s) (particularly in relation to their fibre content), and in relation to the equipment used to undertake the tests;
• How these uncertainties are accommodated in the range of parameters used in slope stability analysis, where undertaken.

The publication of the first edition of this guidance has contributed to an increased interest in geotechnical research into characterisation of peat soils, particularly in relation to stability analysis, and it is expected that further useful advances will be made subsequent to this reissue.

4.4.4.3 Selection of appropriate site plant and safe working practice

In planning any site investigation, particular attention should be paid to the safety of personnel and the public at all stages of the investigation. The implications of specific methods and their associated risk to workers and the public should be considered and accounted for during the fieldwork layout design.

Any investigation should, in the first instance, consider the results of the desk study to minimise the impact of subsequent investigation techniques on any of the peat landslide hazards identified at the site. Where practicable, the investigation methods used should have a minimal impact on the site and any arisings should be returned to their point of origin, with surface vegetation cover re-laid to promote rapid recovery of the peat surface and minimise degradation.

Similarly the impacts of the works and their potential to trigger peat landslide hazards should be considered during the planning and design of any ground investigation. Cutting of a peat slope toe to enable machinery access should be avoided where practical and similarly the loading of peat slopes by plant or by temporary peat storage areas at the slope crest should be avoided to minimise the risk of destabilising the slope. In many cases, lightweight plant and vehicles ( e.g. quad-bikes and trailers, six-wheel buggies) should be employed during the ground investigation, to avoid the impacts of heavy plant at the site. Where loading or cutting is unavoidable, efforts should be made to site the plant equipment away from slope crests and to drain the free faces of slope cuts.

The hand coring and trial pit techniques described above are most suited to areas of shallow peat cover (less than 2m). Depths greater than 2m can be achieved using hand corers, but only at the discretion of the site worker(s) in relation to ease of retrieval of samples and with regard to Health and Safety issues. Where peat depths are confirmed to exceed 2m the need for deeper ground investigation should be considered alongside the suitability of the location. A suitable combination of trial pits, boreholes and auguring/coring should be sufficient to identify variability in peat material properties across the development site.