The Geology of the LLWR Site and
Transcription
The Geology of the LLWR Site and
This page is left blank intentionally. LLWR ESC The Geology of the LLWR Site and Surrounding Region Uisdean Michie John Hunter George Towler QRS-1443Y-R1 Version 2.0 (Final) October 2010 Document History Title: LLWR ESC Subtitle: The Geology of the LLWR Site and Surrounding Region Client: LLWR Document Number: QRS-1443Y-R1 Version Number: Version 0.1 Notes: Draft report submitted to peer review workshop involving Date: March 2010 LLWR, NNL, Halcrow, BGS and Serco Prepared by: Uisdean Michie, John Hunter, George Towler Reviewed by: Uisdean Michie and John Hunter are the principal authors. They cross-checked each others work, while George Towler provided an overall review. Version Number: Version 1.0 (draft) Notes: Draft final report submitted to LLWR for review. Prepared by: Uisdean Michie and John Hunter Reviewed by: George Towler Version Number: Version 1.0 (Final) Notes: Final report following workshop and LLWR review Prepared by: Uisdean Michie and John Hunter Reviewed by: George Towler Approved by: George Towler Version Number: Version 2.0 (DraftFinal) Notes: Incorporate updates to 3D Geological Model Quintessa Limited The Hub, 14 Station Road Henley-on-Thames Oxfordshire RG9 1AY United Kingdom Date: April 2010 Date: 23 October 2010 Date: 15 October 2010 Tel: +44 (0) 1491 636246 Fax: +44 (0) 1491 636247 [email protected] www.quintessa.org www.quintessa-online.com Prepared by: Uisdean Michie and John Hunter Reviewed by: George Towler Version Number: Version 2.0 (Final) Notes: Outstanding reference added. No other changes. Prepared by: George Towler Reviewed by: George Towler Approved by: George Towler Date: 5 October 2010 This report uses data extracted from British Geological Survey (BGS) maps and supporting documents under the LLWR license IPR/117-4DR QRS-1443Y-R1, Version 2.0 (Final) Summary This report is an updated and comprehensive review of the geology of the LLWR site and its wider region. It is one of a number of reports produced to underpin the 2011 Environmental Safety Case (ESC) that describe the understanding of the LLWR site and its wider environs. The geological understanding supports other aspects of the site understanding, particularly the hydrogeological understanding, but also the engineering design, design optimisation, and hence performance assessment. The geology of the LLWR site is described in the context of the wider regional Quaternary geology, both onshore and offshore under the East Irish Sea, and is related to recent syntheses of the geological understanding by the British Geological Survey (BGS) and others. The report also describes how understanding of the geology of the LLWR site has changed with time, and why the current lithofacies approach is considered to be the most appropriate approach to underpin the ESC. The LLWR site is located on a generally thick sequence of unconsolidated Late Quaternary sediments, which were deposited on an irregular surface of Triassic Sherwood Sandstone bedrock. The Quaternary sediments consist of a wide range of material types of significantly different properties, which have complex spatial relationships. They are the focus of the geological interpretation because the LLWR is developed within these sediments, and because the nature of the Quaternary sediments is an important control on the long-term safety of the site: in particular with regards to the site hydrogeology and aqueous contaminant transport; and the vulnerability of the site to coastal erosion. The complexity of the Quaternary sediments has been revealed as a consequence of the extensive investigations that have been undertaken at the LLWR, including drilling a very large number of site investigation boreholes. A much sparser dataset would hide the complexity. As a consequence of this complexity, despite the very large amount of data available for the LLWR site and wider region, uncertainties remain, particularly with regards to the chronostratigraphic (time) correlation of individual deposits. However, it is only uncertainty regarding the spatial distribution of materials that is important to the ESC. An important factor in the comprehensive geological understanding of the site is its interpretation within the well-defined framework of the regional onshore and offshore geology. There are still some spatial data limitations, for example only limited borehole data are available at depth and between the site and the coast, relative to the very large amount of shallow borehole data. However, it is proposed that collection of large i amounts of additional deep and offsite borehole data is unlikely to have significant consequences for the ESC. This is because sufficient understanding is available regarding the key parts of the system to identify any hydrogeologically significant features that might be present (e.g. incised channels), and to define length scales of heterogeneity and explore their impacts. More generally, it can be stated that due to the very large data set, the natural variability and complexity can be understood and bounded in safety assessment calculations. Therefore the geological interpretation is considered to be appropriate to underpin the ESC. A summary of the recent approach to the geological interpretation and developments subsequent to the Schedule 9 Requirement 2 site understanding update, provided to the Environment Agency in 2008, is given below. Prior to 2007, the geological understanding of the Quaternary glacigenic sediments underlying the LLWR site and its surrounding area had been described in a series of widely-read reports which had been prepared by a succession of several different authors. All of these reports utilised ‘layered sequence’ models to characterise the nature of the glacial deposits, i.e., the preserved sediments were defined as a sequential accumulation of layers of variable mixtures of clay, silt, sand and gravel material. Individual layers were assigned to particular glaciation processes and events that occurred during the Quaternary. These models made significant assumptions when assigning such processes and events to the resulting individual layers of sediment and also when identifying the boundaries between the layers and correlating these layers between borehole logs. Because of the limited number of surface outcrops of Quaternary deposits in the LLWR area and the adjacent inland district, both of the principal layered sequence models (i.e., the LLWR model for the site and the British Geological Survey (BGS) model for the surrounding district) were constructed from a limited series of two-dimensional cross-sections using a small number of selected boreholes – usually those boreholes that had been drilled through the complete Quaternary succession to intersect the underlying surface of the Triassic sandstone bedrock. The authors of these models found difficulty in equating the layered sequences of sediment identified within the LLWR (on-site) using one set of borehole logs, with those identified in the surrounding inland area (off-site, or regional area) from other borehole logs, and also with the preserved glacial sediments underlying the nearby seabed (offshore) identified from geophysical surveying and offshore boreholes. The correlation problems between the models were compounded by the spatial data gap (i.e., the lack of boreholes) between the LLWR boreholes east (inland) of the site and the area to the NE investigated in detail by the BGS as part of the Nirex investigations. This lack of simple correlation between different sets of layered ii QRS-1443Y-R1, Version 2.0 (Final) sequences resulted in the adoption of different stratigraphical names for the on-site, off-site and offshore sequences. Both of the principal layered sequence models also had difficulty in accommodating some of the variable features in the glacigenic deposits that had been discovered by different phases of borehole drilling within the LLWR. In 2007, the present authors prepared a report which reassessed the database of historical borehole lithology logs and which, utilising most of the available data rather than a limited selection of boreholes, proposed an alternative representation of the LLWR Quaternary geology – termed a ‘lithofacies’ model. In this model, more emphasis was placed upon descriptive analysis of the sedimentary materials and fewer assumptions and inferences were made regarding the processes and events that may have formed them. This approach was selected because the spatial distribution of material types is the key input from the geological understanding to the ESC (including the engineering design, hydrogeology, etc). The 2007 model subdivided the sedimentary deposits into lithofacies units composed of similar material types and the report outlined the procedure used to identify the likely boundaries between the units. Some of the lithofacies units were shown to occur as partial sequences of layers. However, this spatial arrangement is not a rigid requirement and the model easily accommodates the spatial variability of sediment types that is known from the borehole drilling records and also the apparent lateral discontinuities that separate them. In 2008 and 2009 two phases of additional borehole drilling were completed within the LLWR and in the immediate off-site area on the coastal side. Geophysical surveying and other scientific studies were also conducted. The lithology logs from the recent boreholes provide new information that in the opinion of the authors, supports the lithofacies model as a valid characterisation of the LLWR site. This report updates, clarifies and refines the lithofacies model, but there are no major changes to the framework established earlier. The 2007 report described the spatial distribution of materials at the regional scale as a number of lithofacies units (labelled A to D), while the spatial distribution of materials at the site scale was described as a separate number of lithofacies packages (labelled LP1 to LP7). Further work for this report has enabled the site and regional scale interpretations to be integrated, and the current interpretation now uses a unified set of lithofacies units, with the previous local LP nomenclature being considered unnecessary and therefore replaced by the A-D units. These units can be mapped to the BGS regional lithostratigraphy in a general sense. However, the BGS’ lithostratigraphy has not been adopted because it is less focused on the spatial distribution of material types, and because considerable uncertainties arise when trying to describe the complex sequence of deposits at the LLWR using the BGS lithostratigraphy, e.g. in a iii given LLWR borehole there may be more or fewer till units than described by the BGS’ lithostratigraphy. This report describes the spatial distribution of materials for the LLWR site and wider region using a significant number of maps and cross-sections, which utilise almost all the available data. These form the basis for a refined 3D model of the geology of the LLWR site and surrounding area and are being incorporated into an updated 3D computer model of the geology. The large amount of detailed geological data assessed for this study is presented in this report using a comprehensive set of (2D) crosssections and also some boundary surfaces presented as rendered 3D images. Some of the cross-sections for the regional area are diagrammatic, with a large vertical exaggeration, to illustrate the details of the information from borehole logs. The sections for the LLWR site provide a comprehensive network across the site. Since Version 1 of this report was issued in June 2010, two further reports have been prepared to accompany the updated 3D computer geological model. In the process of digitising geological cross-sections for that work, some minor inconsistencies between the exact positions of certain lithofacies boundaries on a few intersecting cross-sections were revealed. Adjustments have been made to the relevant cross-sections to eliminate these inconsistencies and to make the 2D cross-sections directly comparable to the 3D model. The amended cross-sections, appropriately identified, have been substituted into Version 2 of this report. There is currently ongoing work to investigate the palaeo coastal evolution in the area of the Drigg spit. The results from these studies will be available for inclusion into the next update of the geological interpretation. iv QRS-1443Y-R1, Version 2.0 (Final) Contents Glossary 1 1 Introduction 4 2 Geology of the LLWR area of West Cumbria 7 2.1 Bedrock geology 9 2.2 Quaternary sediments 2.2.1 Application of a lithofacies approach 3 13 16 2.3 Sources of data 18 2.4 Lithological description of superficial deposits 21 Regional scale lithofacies units 23 3.1 Regional lithostratigraphical units 23 3.1.1 Age sequence and correlation of the Quaternary lithostratigraphical units 29 3.2 Offshore seismic stratigraphy 31 3.2.1 Correlation of the offshore seismic stratigraphy and the onshore lithostratigraphy 4 36 3.3 Development of the regional lithofacies units 36 3.4 Lithofacies Units 43 Characterisation of the Quaternary geology at the LLWR site 51 4.1 The concepts of ‘Lithostratigraphical Formations’ and ‘Lithofacies’ 57 4.2 Additional review of data for this report 59 4.3 Updated contour map of sandstone bedrock surface 61 4.4 Updated LLWR lithofacies model: the east-west cross-section and southwestnortheast section 8 64 4.5 Updated LLWR lithofacies model: northwest-southeast cross-sections 1 & 2 71 4.6 Updated LLWR lithofacies model: northwest-southeast cross-section 5 80 4.7 Discussion of uncertainties 83 5 Integration of the regional scale and site scale lithofacies packages 86 6 Application of the lithofacies units to hydrogeological modelling 89 6.1 Heterogeneity 92 7 Conclusions 94 8 References 95 Appendix A – Development of the understanding of the Quaternary sequence for the area of the LLWR 99 Appendix B – LLWR site borehole cross-sections 117 v Appendix C – Summary of lithology code consolidation 137 Appendix D - Lithofacies colour palette 143 Appendix E - Adaptation of 2D cross-sections into a 3D model 144 vi QRS-1443Y-R1, Version 2.0 (Final) Glossary Term Description Deglaciation Time period at the end of a glaciation when ice sheets melted and retreated from their maximum extent. Climate reversals and ice sheet dynamics during the deglacial period produce readvances and oscillations in the position of the margin of the ice sheet. Diamict A non genetic term that describes unsorted and predominantly massive sediment consisting of variable admixtures of grain sizes, which may range from clay to boulders. If the diamict material can be shown to be of glacial origin, then the glacial diamict is classified as till. Glacial deposits Sediments deposited directly by a glacier or ice sheet. Glacigenic A deposit that is glacially related, but not necessarily deposited directly by ice. Glacioeustatic Global changes in sea level due to the transfer of water from the sea to ice sheets on land. Glaciomarine Sediments deposited marginal to ice sheets in marine environments. Glaciotectonic Deformation of previously deposited glacigenic sediments by vertical compression or lateral movement of later ice cover. Interfluve An area of land separating two rivers. In the floodplain of meandering and braided river systems, the interfluves often consist of areas of finer-grained (silty, clayey) sediments that have accumulated from repeated flooding events from the same rivers and which separate active river channels and their associated sand bars. However in the context of this report, the term is applied to the non-fluvial glacigenic material (generally clayey) that separates interpreted fluvial channel deposits (sands and gravels) that may have formed from either surface or subglacial meltwater drainage. Isostatic The depression of the underlying bedrock and adjoining 1 depression marginal area due to the weight of overlying ice. Bedrock at a greater distance from the ice sheet can be raised due to deeper sub-crustal movement of material from beneath the ice sheet. Late Devensian The British name for the sediments deposited on the present land area of the British Isles during the time period from about 30,000 to 10,000 years ago when large ice sheets developed and expanded over the northern parts of North America, Europe and Asia. (The name Devensian is derived from the Latin name for the River Dee in Cheshire, an area where there are thick deposits from the glaciation. The equivalent name Weichselian is used in Europe and for the deposits under the North Sea and other offshore areas.} Lithofacies The physical (and partly chemical) characteristics of a body of sediment that distinguish it from other sediments. It includes the internal features of the sediment unit and the external relationships with adjoining sediments. This approach allows for the discontinuous nature of layers of a particular sediment type, and lateral variation into different sediment types of the same age. Lithostratigraphy The formal vertical subdivision and age correlation of sequences of layers of sediments (beds) by means of sediment type. This approach relies on bed-for-bed correlation and can lead to assumptions of lateral continuity of units. Outwash Generally sandy and gravelly sediments deposited by meltwater outside an ice sheet margin. Palaeovalley A valley infilled by younger sediments. Pinning point An upstanding bedrock feature that forms a stable position over a period of time for the ice margin during ice advance or retreat. Subaqueous Sediments deposited into a body of water. Subglacial The complex range of environments beneath an ice sheet. Towards the margin of an ice sheet, subglacial meltwater can be very important in modifying the sediments deposited. Till Genetic term for the unsorted admixture of debris eroded and deposited by an ice sheet and not significantly reworked by 2 QRS-1443Y-R1, Version 2.0 (Final) meltwater. The thickest till deposits are formed at the margins of ice sheets. Where the subglacial deposits are generally structureless and contain boulder-sized clasts in a clay matrix, they are often termed boulder clay. Tills deposited directly beneath an ice sheet are termed lodgement till, which may be deformed by the movement and weight of the ice (deformation till). More clay-rich tills termed melt–out and flow tills are formed by the deposition of material from within or above the ice sheet during the later stages of the melting of stagnant ice. Transgression The rise in the sea over the land surface due to rise in the relative sea level. Sediments deposited during the transgression show lateral facies changes from coarse beach deposits adjacent to the land to deeper water silts and muds offshore. Tunnel valley A deep channel cut by a ‘river’ of subglacial meltwater into older sediments or bedrock. Because the water is confined by the overlying ice sheet, incision can develop to well below contemporaneous sea level. Infilling of the channel can be by a range of sediments from coarse gravel to marine silts and muds. Unconformity A break in deposition of a sequence of sediments. It is particularly noted where there was removal (erosion) and/or tilting of the older sediments, and subsequent deposition of the younger sediments across different layers of the older sediments (angular unconformity). 3 1 Introduction This report describes the geology of the LLWR site and the surrounding region. It updates the lithofacies (i.e., sediment material type) model of the site’s Quaternary geology, developed by Hunter et al. (2007a) in response to review of the 2002 Drigg PCSC. The report does not present a complete 3-dimensional (3D) reconstruction of the different lithofacies that comprise the Quaternary deposits beneath the LLWR site, and adjacent areas of interest (e.g. potential groundwater transport paths). However, this report approximates a 3D reconstruction by means of a considerable number of 2dimensional geological cross-sections orientated in different directions. A regional scale 3D geological model, with some localised refinement of the LLWR site, has previously been developed based on the first iteration of the lithofacies approach. The 3D model has been amended to incorporate the updated and detailed lithofacies interpretation at both the regional and the LLWR site-scale that is described in this report (Hunter 2010, and Smith 2010a). The key purpose of the development of a reliable geological understanding of the LLWR site is to provide an adequate representation of the site’s Quaternary geology for use in hydrogeological modelling, in studies of contaminant transport, in civil engineering (vault development) work, and in studies of coastal erosion. The report provides a clear explanation of why the lithofacies model is the most appropriate to meet this purpose. The report provides the information needed to support the development of the 3D geological model (Smith, 2009a, 2010a), and other geological information required for hydrogeological modelling, etc and production of an ESC. The LLWR site and its surrounding area has been subjected to intensive surface and sub-surface investigation for at least the previous three decades, yet despite this scientific effort, the detailed character of the late Quaternary and Holocene sediments underlying the LLWR site is still not completely understood, particularly with regard to the sequence of depositional events and the chronostratigraphic (time) relationships between different deposits. This is partly a consequence of the complexity of the local geology resulting from the location of the site where the Irish Sea and inland ice masses coalesced during the last glacial maximum, partly because most of the geology is concealed and can only be interpreted from borehole records and partly because of the high level of geological understanding expected by regulators for a licensed repository site. The complexity has also been revealed as a consequence of the very large number of boreholes that have been drilled at the LLWR site. A much sparser dataset would hide the complexity. The geology of the LLWR site is described in the context of the regional Quaternary geology surrounding the site and also of the nearby offshore seabed sediments. The 4 QRS-1443Y-R1, Version 2.0 (Final) lithofacies approach is applied at the site and regional scales. The report explains how the lithofacies approach provides a regional framework for understanding the complex nature of the thick deposits of Quaternary sediments preserved beneath the LLWR site on a concealed bedrock surface that has significant topographical relief. The advantages of this lithofacies approach in providing the basis for the production of 3D models of the spatial distribution of the material properties of the late Quaternary sediments in the region surrounding the LLWR site, and in detail for the LLWR site are discussed. Uncertainties and problems identified from incorporation of new data from the LLWR site into the 3D geological model prior to this report (Smith, 2009a, 2009b) have been addressed and new information resulting from ongoing investigations such as the geophysical studies (Halcrow, 2010) have been included where relevant. To achieve its objective, it is necessary for the report to include a summarised ‘audit trail’ of previous interpretations of the LLWR site showing how the geological understanding of the Quaternary deposits beneath the site has evolved and how the lithofacies model compares to the alternative models as a representative characterisation of the site. Section 2 describes the bedrock and Quaternary geology of the LLWR area of the West Cumbria (the regional geology). Section 3 then describes the development of lithofacies units at the regional scale, the associated material types and relationship with the BGS’ lithostratigraphy. Section 4 describes the Quaternary lithologies at the LLWR site and Section 5 describes the relationships between the site and regional lithofacies. Section 6 describes the likely hydrogeological properties of the lithofacies units as an input to hydrogeological modelling. Section 7 draws conclusions regarding the status of the geological interpretation for the ESC. Five Appendices are also presented. Appendix A provides further information summarising the development of the classification of the sediment units in the Quaternary sequence leading to the lithofacies framework. The successive classifications of the sediment sequence in selected cross sections of the LLWR site at key stages in the development of the geological understanding are presented and reviewed, and their relationship the lithofacies framework identified. Appendix B provides a complete suite of detailed cross-sections that illustrate the geology of the LLWR site and adjacent areas of interest. Appendix C outlines how the extensive database of lithological descriptions of borehole samples from the LLWR has been rationalised and condensed to enable the lithofacies cross-sections of the site to be produced. Appendix D gives details of the lithofacies colour palette so that users of this report can maintain a consistent colour scheme in future studies. Finally, Appendix E 5 summarises the process used to adapt the geological information contained in the 2D cross-sections for incorporation into the 3D computer model. 6 QRS-1443Y-R1, Version 2.0 (Final) 2 Geology of the LLWR area of West Cumbria This section describes the bedrock and Quaternary geology of the LLWR area of west Cumbria, i.e. the regional geology. It includes the geological setting of the LLWR site, and provides the background to development of the regional lithofacies units for the Quaternary deposits described in Section 3. It also describes the sources and the varying quality of the available data and defines the principal concepts that are used to explain the site-scale lithofacies units that are described in more detail in Section 4. Because the LLWR is a surface facility constructed within a generally thick sequence of Quaternary sediments, the nature and spatial relationships of the Quaternary sediments are the most important features of the geology for the ESC, engineering, etc. Therefore the emphasis in this description of the current understanding of the geology is on the Quaternary sediments. The nature of the underlying bedrock is only outlined where relevant to the hydrogeological modelling of the area around the LLWR. There is a broad division into the older rocks forming the higher area of the Lake District in the east, and the younger sedimentary rocks forming the bedrock to the coastal plain and the sediments under the East Irish Sea in the west. The older rocks are metamorphic (deformed and recrystallised at great depth), or igneous (originally molten), either formed at surface (volcanic and pyroclastic) or intruded into earlier rocks at depth (granite). These older rocks also occur at depth below the younger sedimentary rocks to the west and north. Figure 2-1 is a compilation of information on the thickness of the Quaternary sediments of the west Cumbria and offshore district from the Nirex studies of the Quaternary of the area (Nirex, 1997a). These unconsolidated sediments were deposited on bedrock by processes related to the advance of ice sheets over the area and subsequent post-glacial deposition. All the deposits exposed at surface are considered to be the result of the latest ice age and younger sedimentation. Onshore deposits more than 10 m thick are confined to the valleys of the main rivers crossing the area from the uplands of the Lake District to the Irish Sea. These valleys were deepened by river erosion at times of lower sea levels and by valley glaciers and are now infilled with both glacial and postglacial deposits. In the thickest onshore sequence, situated east and south of the LLWR, these deposits overlie a concealed older sequence of lacustrine and marine to estuarine sediments, situated beneath a level of 15 to 20 m below present sea level. (Note, the sea level during their deposition is uncertain. The marine sediments were deposited on the seabed below a rising/transgressing sea level that probably rose from below -40 m to 7 above -15 m. The marginal sandy facies adjacent to bedrock reaches to above -15 m and probably consists of a beach and onshore sequence.) The BGS Memoir on the geology of the west Cumbria district (Akhurst et al., 1997) provides a full description of the mapped subdivision and genetic interpretation of the Quaternary sediments both onshore and offshore. The interpretation of the Quaternary sequence by the BGS is focussed on the age relationships of the sediments to the advances and retreats of the eastern margin of the ice sheet that spread down the East Irish Sea from an ice centre in Scotland, as shown on the inset to Figure 2-1. Two of the advances of the margin of the ice sheet after a retreat into the Irish Sea during deglaciation are shown on Figure 2-1 as the Gosforth Oscillation and the later Scottish Readvance. A more detailed study of the age relationships and depositional history of the deposits in the Lower Wasdale area, which comprises the lower part of the valleys of the River Irt and River Mite and includes the area around the LLWR, is provided by an article by Merritt and Auton (2000). These studies are summarised and placed in their wider regional context in the BGS Regional Geology guide to Northern England and the Isle of Man (Stone et al., 2010). These geological studies are primarily concerned with understanding the origin of the deposits and their stratigraphical (time) relationships. However, this study is primarily concerned with understanding the nature of the deposits (i.e. material types) and their spatial relationships. Therefore for this study and earlier related studies (Michie et al., 2007; Hunter et al., 2007b), the nature of the deposits was used as the basis for their investigation and classification of units. However, the BGS studies of the genesis of the Quaternary sediments were used to support the interpretation of the spatial relationships of the units. 8 QRS-1443Y-R1, Version 2.0 (Final) Figure 2-1. Thickness of Quaternary sediments across the west Cumbria and adjacent offshore area 2.1 Bedrock geology Figure 2-2 shows the simplified bedrock geology of west Cumbria and the adjacent offshore area. In the Lake District, the metamorphic and igneous bedrock consists of the slates of the Skiddaw Group, dominantly in the north, the volcanic rocks of the 9 Borrowdale Volcanic Group, dominantly in the central area, and the granitic rocks of the Ennerdale intrusion in the north-east and the Eskdale intrusion in the south. These older rocks are all of Ordovician age, from about 500 to 460 million years old. They extend at depth under the younger sedimentary rocks to the west and north, and are collectively referred to as basement rocks. In the north of the west Cumbria area, the basement rocks are overlain by a sedimentary sequence of Carboniferous age (350 to 300 million years old) rocks, including Limestone, Coal Measures and Sandstone. In turn, these are overlain by a younger sedimentary sequence, the Cumbrian Coast Group of Permian age (290 to 260 million years old), which also oversteps across the Carboniferous rocks to rest directly on the basement rocks. Above the Permian age strata, there is a thick succession of sandstones of Triassic age, which forms the bedrock to the coastal zone and much of the adjacent offshore area. Figure 2-2. Simplified bedrock geology of West Cumbria and the adjacent offshore area Some of the younger sedimentary rocks forming the bedrock to the coastal area are well exposed in cliffs up to 100 m high at St Bees Head, which is formed of the St Bees Sandstone, the lowest unit of the Sherwood Sandstone Group of Triassic age, about 240 10 QRS-1443Y-R1, Version 2.0 (Final) million years old. However, the Quaternary sediments cover the sedimentary rocks over most of the coastal plain to the south, except for the valley of the River Calder, which cuts into the sandstone bedrock. This mostly consists of the Calder Sandstone, the middle unit of the Sherwood Sandstone Group. The transition to the Ormskirk Sandstone, the uppermost unit of the Sherwood Sandstone Group, is present near the coast, and this unit also forms low outcrops on the beach at Seascale. Figure 2-3 shows the bedrock geology of the LLWR and surrounding area, onshore and offshore. In the east of the area, the bedrock is formed of the igneous and metamorphic rocks of the Lake District. The oldest rocks are a small area of Skiddaw Group on Muncaster Fell east of Ravenglass. These have been brought up from depth by the intrusion of the Eskdale granite into the Borrowdale Volcanic Group. About 4 km to the east of the LLWR, a major fault (structural displacement) forms the boundary between the basement (igneous and metamorphic rocks) of the Lake District uplands and the younger sedimentary rocks of the coastal zone, which are downthrown to the west by several hundred metres. Therefore the complete succession of the sedimentary rocks above the older igneous and metamorphic rocks is not present at the bedrock surface to the east of the LLWR area. However, to the north of the fault zone that runs inland from the Seascale area, north of the LLWR, the sedimentary sequence above the igneous and metamorphic rocks is present at the bedrock surface. The basal unit is the Brockram, a unit formed of large clasts of the Borrowdale Volcanic Group in a muddy matrix. Above this unit, there is an upward succession of St Bees Sandstone, Calder Sandstone and then Ormskirk Sandstone , which forms the bedrock under the LLWR. Faulting associated with the SW-NE trending Seascale Fault Zone affects the Ormskirk Sandstone north of the LLWR site. Studies by Nirex of the likely effect of faulting on the hydrogeological properties of the sandstone concluded that there was unlikely to be any significant effect on the permeability of an already permeable sandstone. One of the Nirex deep boreholes (BH13) was drilled to the north of the LLWR site and provided a complete succession of the sedimentary strata above the Borrowdale Volcanic Group, which was reached at a depth of about 1636 m (1617 m below sea level). From the surface, there was about 34 m of Quaternary sediment and disaggregated sandstone (the informal ‘Snellings Sand’ unit) over the Ormskirk Sandstone bedrock. The Ormskirk Sandstone occurred to about 176 m depth, with surface-related weathering effects noted in the top 30 m. The sandstone is reddish brown, fine to coarse, moderately to well sorted, generally poorly cemented and friable. Two main facies were identified; a predominant aeolian (wind blown) dune facies and a damp interdune facies. 11 Below the Ormskirk Sandstone, the Calder Sandstone occurs to 646 m depth and is a generally coarser-grained sandstone. The Calder Sandstone contains both fluvial and aeolian units, with the top of the uppermost fluvial unit defining the top of the formation and the base of the lowest aeolian unit defining the base of the formation. Figure 2-3. Simplified bedrock geology of the LLWR and surrounding area Below this, the St Bees Sandstone occurs to 1270 m depth and is an entirely fluvial formation of fine- to medium-grained sandstone. The base of the St Bees Sandstone is transitional to the St Bees Shales, which extend to about 1442 m depth. The shales are transitional to a 30 m thickness of the St Bees Evaporites, which contain layers of anhydrite (calcium sulphate), a poorly-soluble evaporite mineral. These overlie and 12 QRS-1443Y-R1, Version 2.0 (Final) interbed with the Brockram, a coarse conglomerate, which extends to 1590 m depth. Between this and the Borrowdale Volcanic Group, a 46 m section of Carboniferous Limestone is preserved. The Ormskirk Sandstone forms the bedrock to the LLWR site and most of the coastal area around the site. The strata are tilted (dip) to the west and so younger strata are present to the west under the area offshore from the LLWR. About 2 km offshore, the overlying Mercia Mudstone Group forms the bedrock under the sea. This is a thick sequence of reddish brown mudstones, with layers of dolomitic mudstone and anhydrite. Thick layers of halite, natural rock salt, are preserved at depth below the zone of near-surface dissolution. 2.2 Quaternary sediments The glacial deposits covering the bedrock in the west Cumbria area were deposited from either the ice sheet that infilled the Irish Sea in the west, or the local Lake District ice cap in the east, depending upon the position of their margins at specific times. At the maximum of the Late Devensian glaciation from 27,000 to 22,000 years Before Present (BP), the ice sheet and ice cap coalesced into the British-Irish ice sheet, which also coalesced with the Scandinavian ice sheet. However, the Lake District ice cap maintained an independent existence, probably because the high-level ice cap was frozen to the underlying substrate. In contrast, the lower-level ice sheet flowing down the floor of the Irish Sea was highly mobile. The East Irish Sea between the coast of west Cumbria and the Isle of Man is shallow, generally less than 50 m deep and the bed is fairly flat, but this conceals the infilling of an incised bedrock surface with a variable thickness of Quaternary sediments. The Coulderton Channel shown on Figure 2-1 is the most easterly of a set of such incised and infilled channels running sub-parallel to the coast. These incisions may have formed during earlier, more extensive, glacial episodes of the 2 million year duration of the Quaternary, when there was deep erosion of the East Irish Sea basin. Within the channels, the thickness of Quaternary sediments can exceed 100 m. In addition, there are some transverse incisions extending out from the valleys of the major rivers. Offshore seismic surveys and earlier boreholes drilled through the Quaternary sediments by the BGS have revealed that there is a upward succession consisting of basal sands and gravels with diamictons, overlain by lacustrine, marine and glaciomarine sediments, a thin till layer, and then a cover of post-glacial marine sediments on the sea floor. The nature of the sediments provides evidence for major rises and falls in the relative sea level. 13 Between the Isle of Man and the Scottish coast, offshore boreholes penetrated at the base of the Quaternary sequence a similar succession of laminated glaciolacustrine clays, which pass upwards into marine silts and sands between 50 and 70 m below current sea level. These are capped by a layer of till, which is covered by post-glacial marine sediments. The age of the older lacustrine and marine sequences below the till is uncertain. In the north of the Isle of Man, boreholes have revealed a 250 m thickness of Quaternary sediments, of which the lowest 140 m is concealed below current sea level and includes marine deposits over 60 m below current sea level. The upper succession is of interbedded tills and sands and gravels, deposited during the advances, retreats and readvances of the late Devensian glaciation. In the south of the Isle of Man, a radiocarbon date of about 36,000 years Before Present was determined for peat layers buried beneath the Late Devensian glacigenic deposits. No glacial deposits underlie the peat layer, which indicates that there was no significant Early or Mid Devensian glaciation affecting the East Irish Sea basin. The study by Heathcote and Michie (2004) showed that Relative Sea Level in the Irish Sea area was more than below -50 m below present sea level from about 110,000 to 35,000 years Before Present, and during this time the East Irish Sea was not present. Instead the area would have been a tundra environment with lakes, rivers and local valley glaciers. However, after 35,000 years Before Present, the weight of the growing British and Irish Ice Sheet caused depression of the crust around the ice sheet and produced a rise in Relative Sea Level, sufficient to allow the sea to advance into the area and deposit marine and glaciomarine sediments to infill the incisions in the surface. The influx of a shallow sea in the area adjacent to the growing ice sheet in Scotland, and the presence of an infilled bed of sub-horizontal clay-rich sediments allowed rapid southward expansion of the Late Devensian ice sheet along the axis of the East Irish Sea. Such an interpretation of glaciomarine effects generating fast-flowing ice streams during the initial advance of the ice sheet is similar to the interpretation presented for the later stages of the ice sheet by Eyles and McCabe (1989, 1991). At its maximum development, the ice sheet completely covered the Isle of Man, with indications of an ice thickness of over 800 m. This thickness of ice would have remained grounded on the floor of the East Irish Sea basin even when the inferred Relative Sea Level in the area rose to over 25 m above current sea level at the maximum of the Late Devensian glaciation (Heathcote and Michie, 2004). However, the eastern margin of the ice sheet only gradually advanced over the coast of west Cumbria. Before the arrival of the main Irish Sea ice sheet, local valley glaciers from the Lake District ice cap flowed down the palaeovalley of the River Irt from the 14 QRS-1443Y-R1, Version 2.0 (Final) Wastwater incised basin and deeply eroded the valley, particularly over the sandstone bedrock in the lower part of Wasdale. The detailed studies by the BGS of the glacial deposits identified that the maximum inland extent of the ice from the Irish Sea ice sheet was probably only slightly inland of the limit reached by the ice sheet during deglaciation, shown on Figure 2-1 as the “Limit of Gosforth Oscillation of the Irish Sea ice sheet.” The inset to the figure also illustrates the general form of the ice sheet during deglaciation. A similar relationship applied during the early stage of arrival of the Irish Sea ice sheet, when the valley glacier retreated up Wasdale, probably due to rising relative sea level as the bedrock was depressed. This situation lead to the ponding of glacial meltwater into large glacial lakes against the barrier of the Irish Sea ice sheet, which was pinned to the bedrock high of Muncaster Fell, situated between the River Mite and the River Esk, and other areas of bedrock. Meltwater channels across Muncaster Fell and lake beaches on the hillsides indicate that the lakes reached elevations of over 100 m above present sea level. Laminated silts and muds (varves) were deposited in the lakes, as were outwash deposits of sand and gravel in deltas. The lakes also formed reservoirs of large bodies of water that could be released subglacially to form tunnel valleys if the Irish Sea ice sheet thinned and/or lifted off its base. Figure 2-1 also shows the position that the BGS identified for the inland “Limit of the Scottish Readvance of the Irish Sea ice sheet” at a later stage of the deglaciation, probably before 16,000 yrs BP. This advance of the ice sheet inland produced folding of the older Quaternary sediments along the coast south of St Bees Head and to the south of Seascale, both environments where the sediments could be compressed against bedrock ridges. Inland, the glacitectonic effects were more limited and lead to the production of lowangle thrusting at the margins of till and sand units, and the development of small normal faults due to flattening of the unconsolidated sediments. After the melting of the Irish Sea ice sheet and its final retreat from the west Cumbria area, silts and sands were deposited in local hollows and vegetation gradually colonised the tundra environment. This lead to the formation of peat, the oldest layers of which in the Hallsenna Moor area, NE of the LLWR site, were dated to 15,770 calendar years BP (Nirex, 1997b). During the post-glacial phase, relative sea level initially fell to tens of metres below the present sea level as the bedrock rose following removal of the weight of overlying ice. Then there was a rapid rise in sea level due to melting of the continental ice sheets (Heathcote and Michie, 2004). Relative sea level reached a maximum of several metres above present sea level in the mid-Holocene, around 7000 years before present. During that time, raised beach deposits formed to 8 m above present sea level along the coast 15 and marine silts were deposited up to similar elevation adjacent to the estuary of the rivers Irt and Mite, including the southern edge of the LLWR site. The post-glacial development of the estuary of the rivers Irt, Mite and Esk was studied by Kelly and Emptage (1992), based on interpretation of low-tide air photographs and field mapping. Work is currently in progress to refine their interpretation in the area of the LLWR using information from new geophysical surveys and shallow drilling. The units defined by the BGS from their mapping and related studies of the Quaternary sediments in the west Cumbria area are described in Section 3.1. However, these units cannot readily be applied to hydrogeological modelling across the area, because of the difficulty of identifying the properties of the units and the spatial extent of the multiplicity of thin units with only limited development. In the BGS regional guide for Northern England (Stone et al., 2010) it is noted that; “The sequence of events that occurred during the MLD [Main Late Devensian] glaciation is not fully understood since there is insufficient geochronological control, some phenomena result from more than one phase of glaciation, and the stratigraphical record is beset with difficulties of regional correlation.” 2.2.1 Application of a lithofacies approach One of the key findings of the review by Bond (2006) of the geological interpretation provided to support the 2002 Post-Closure Safety Case (PCSC) for the site (BNFL, 2002a) was that the geological understanding presented was not optimal to underpin the update of the 2002 PCSC hydrogeological interpretation. Instead of presenting a description of the spatial distribution of material properties for the Quaternary sediments necessary for modelling groundwater flow and radionuclide transport, the description focussed on describing the stratigraphy (time sequence) of geological events that had affected the sediments. In the report on the geological interpretation (BNFL, 2002a) that supported the 2002 PCSC, an event stratigraphy approach was adopted. Events are relatively short occurrences, in a relative geological timescale, of phenomena that affect the deposition or removal of sediments to produce features that can be correlated across an area. The BGS lithostratigraphical framework for the onshore Quaternary deposits of the west Cumbria area was also based on the application of an event stratigraphy, using features such as the interpreted advances and retreats of the margins of the Irish Sea ice sheet. For the BNFL (2002a) geological interpretation of the Drigg site, the BGS event stratigraphy was adopted as the framework, but a site-specific set of units were defined, which were different from the BGS lithostratigraphical units. 16 QRS-1443Y-R1, Version 2.0 (Final) The summary of the 2002 geological interpretation states that, “This allowed the development of an event stratigraphy for the geological sequence and provided the framework to allow the prediction of the lithological and structural distribution across the Drigg site.” and; “The event stratigraphy framework allowed a generalised conceptual understanding of the regional depositional events to be developed. This allowed the identification of glacial and post-glacial features that are likely to be present at the Drigg site based on knowledge of the Quaternary events and the resultant depositional and post-depositional geomorphological processes.” This event stratigraphy approach for the LLWR site produced units that could not be confidently correlated across the site, nor with the BGS lithostratigraphical units, and which did not have well-defined material properties. The requirement for an approach that identified and correlated common units with well-defined material properties across the region and at the LLWR site was considered in the 2007 review (Michie et al. 2007: Hunter et al. 2007a,b). That review identified that a lithofacies approach would provide these necessary features. The lithofacies approach was developed in Canada for the study of outcrop sections and borehole cores from the complex suite of sediments deposited by braided river deposits (Miall. 1978) and extended to the study of glacial diamict sequences by Eyles et al. (1983). The BGS applied these lithofacies classifications to the interpretive logging of the outcrops, pits and sediment cores from the boreholes drilled as part of the Nirex Quaternary characterisation programme (Nirex,1997a). The availability of regional lithofacies information from the Nirex/BGS investigations and comparable lithological information from the LLWR site investigations, which could be interpreted and assembled as lithofacies, allowed the application of the lithofacies approach to be undertaken during the 2007 update of the geological interpretation. Instead of using an event stratigraphy to provide a framework for interpretation of the possible distribution of units, the lithofacies approach works directly from the identification and correlation of well-defined units of similar material properties. Lithofacies units are not constrained to have formed in the same time interval, they may cut across other lithofacies units of different material properties, and they may also be discontinuous. Subsequent interpretation of the spatial distribution, nature and relationships of the lithofacies units may allow the determination of their stratigraphical relationships and refinement of the lithofacies units into individual units of stratigraphical significance. The lithofacies approach avoids the problems related to a requirement to determine the stratigraphy of the sediments as an initial and main purpose of the study. The following Section 3 explains how the lithofacies approach was developed. The lithofacies approach is further described in the context of the site geology in Section 4. 17 2.3 Sources of data The principal sources of data that can be used for characterising the Quaternary deposits beneath the LLWR site have been outlined by Hunter et al. (2007b), who also summarised the data-processing procedures that have been used to codify and condense the wide range of borehole lithology descriptions that comprise the bulk of the recorded information from the LLWR site. The variable quality of these descriptive lithology data and the limitations and caveats associated with their use are also described by the same authors. Despite some limitations, it is considered that the lithological data-set is a valuable source of information which can be interrogated to reveal gross patterns of distribution of different sediment types sufficient for characterising the Quaternary deposits at the LLWR. Additional sources of data have become available since the 2007 report by Hunter et al. most of which have been utilised in the preparation of this report. These additional sources have been appended to the list originally shown in the 2007 report to produce an updated, chronological list of principal data sources as follows: ‘Historical’: 79 boreholes drilled when the LLWR was a Royal Ordnance Factory. 1962: the first three ‘old series’ boreholes were drilled. 1962 – 1975: a further four ‘old series’ boreholes were drilled. 1977: the Institute of Geological Sciences (IGS) (now the BGS) investigated the site on behalf of BNFL. The study included examination of available geological exposures in trenches, stream banks, and coastal sections on the Drigg beach. It was followed by 28 DDS (Drigg Disposal Site) series boreholes drilled between November 1977 and January 1979. 1981: a further 14 hydrogeological boreholes (DDS31 to DDS35 and DDS41 to DDS49) were drilled. This additional information led to a revised understanding of the geology and hydrogeology of the site, which is described by Williams et al. (1985). 1981 – 1989: a phase of site investigations was instigated, which included more DDS series boreholes drilled at various times. 1989 – 1995: several individual boreholes were drilled as part of a number of separate site investigations. 18 QRS-1443Y-R1, Version 2.0 (Final) 1995 - 1996: as part of its Quaternary Characterisation Programme of the coastal area of west Cumbria, Nirex drilled 15 boreholes in the area that includes the LLWR site. The boreholes were sited following a programme of surface mapping, excavation of pits, and shallow ground geophysics to identify localities that would provide key information of the relationships of the Quaternary sediments and their hydrogeological characteristics. Most of the boreholes, drilled to a maximum depth of about 68 m, penetrated the Quaternary sequence into bedrock. Summary logs of the boreholes are shown as Figures 4 and 5 in the article by Merritt and Auton (2000), which also provides references to the Nirex reports in which the detailed logs and their interpretation are presented. 1995 – 2002: the Drigg Site Characterisation Programme (DSCP) involved drilling a network of 41 boreholes on-site and 8 boreholes off-site in the first phase of work, followed by clusters of between 2 and 4 closely-spaced boreholes drilled to different depths at 12 locations on-site (the C1 to C12 cluster borehole sets). Various geophysical surveys (resisitivity tomography and seismic reflection) were also undertaken. 2004 – 2005: the Drigg Vaults Phase II programme of site investigation involved drilling 34 boreholes both on-site and off-site. 2006 - 2007: the Drigg Vaults Phase III programme (Stages 1 to 5) of site investigation involved drilling 43 boreholes and excavating 17 trial pits (the 8400, 8500 and 8600 series boreholes). These boreholes were located both inside the LLWR and also off-site, on the western (seaward) side. Note: the descriptive lithology data generated by these intrusive investigations were not available in time for incorporation into Hunter et al (2007). 2009: the Drigg Vaults Phase III, Stage 6 programme of site investigation involved drilling 14 boreholes (8600 and 8700 series) to be used as groundwater monitoring wells. 2009: geological logging of the Vault 9 trench excavation (Smith, 2009a) 2009: digitisation of historical lithology descriptions from Royal Ordnance Factory (ROF) boreholes dating from the 1940s (Smith, 2009b). 2009: the offsite geophysical survey conducted by Halcrow-RSK (Halcrow, 2010). 2010: LLWR coastal investigation studies (not completed at the time this report was prepared). 19 [Note: copies of all of the site investigation records listed above and used in creating the 3D geological model described in this and in other reports are held by LLWR in digital AGS format]. The depths of recently-drilled Drigg Vaults Phase III (Stages 1-6) boreholes vary from 10 m to more than 60 m. Of the total number of these boreholes, 12 were drilled to a depth sufficient to penetrate through the entire thickness of Quaternary deposits and intersect the sandstone bedrock beneath. A further 8 boreholes were drilled to a depth sufficient to penetrate through almost the entire thickness of Quaternary deposits but did not reach the sandstone bedrock. These 20 deep boreholes greatly assist the interpretation of the Quaternary geology of the LLWR site at intermediate depths and also improve the definition of the concealed surface of the sandstone bedrock. The historical ROF (Royal Ordnance Factory) borehole data, comprising lithological descriptions from a total of 79 boreholes drilled principally within the LLWR site, were converted into a digital format suitable for appending to the main database of LLWR borehole lithology descriptions by Smith (2009b). The final depths of the majority of these boreholes are quite shallow – the average depth is 6.8 m and the deepest only reaches 10.9 m. Therefore their value to site understanding is limited to improving the definition of the near-surface deposits in areas of the site where no recent boreholes have been drilled. The quality of these ROF data is also not as good as the data from later borehole drilling programmes and in many cases the lithology descriptions consist of only a few words. Also, narrow borehole intervals of slightly different material, which by modern standards would be recorded separately on logs, have probably been combined together into single composite units in many cases. Another recent report prepared by Smith (2009a) describes in detail the exposure of shallow glacigenic sediments revealed temporarily during the excavation of a 200 m long trench for Vault 9 at the LLWR. The report provides a valuable record of the sedimentary material types, depositional fabrics and gentle deformation structures preserved just below the land surface (maximum depth of 6 m) in one area of the LLWR. The most recent study made available for the authors to consider during the preparation of this report is the compilation of the 2009 off-site geophysical survey conducted by Halcrow-RSK (2010). Because most of the borehole lithology data utilised in the preparation of this report is located inside the LLWR site, only partial lengths of some of the resistivity profiles from the geophysical survey which either coincide with, or are sufficiently close to the borehole cross-sections presented herein, could be used for comparative purposes. The new geophysical data have not been used to construct any of these borehole cross-sections, which are based entirely upon lithology descriptions. 20 QRS-1443Y-R1, Version 2.0 (Final) It should be noted that the geophysical investigations were focused on describing the near-surface Quaternary deposits in the context of their vulnerability to coastal erosion, and the coastal sediment balance. They were not optimised to investigate the deep geology, or undertaken to directly support the geological interpretation. Work to analyse the results of the geophysical investigations is ongoing at this time. The new borehole, outcrop and geophysical data acquired since the 2007 report can, in addition to expanding the LLWR geology database to characterise the site, offer a means of testing the robustness of the various geological models of the LLWR Quaternary deposits and, in particular, the provisional map of ‘channel’ and ‘interfluve clay’ units shown as Figure 21 in Hunter et al (2007b). 2.4 Lithological description of superficial deposits The bulk of the sample descriptions in the LLWR borehole lithology database are engineering-type descriptions of disturbed drilling samples recorded to the BS-5930 standard (British Standards Institution, 1999). As such, they often lack valuable information on sedimentary and glacitectonic fabrics and other useful micro-level observations that may assist with the identification of a material as being deposited by a particular genetic process. With borehole samples from glacial deposits, for example, this limitation may introduce uncertainty in the correct identification of different types of tills. To avoid onerous borehole logging work and voluminous records, materials with slightly varying sedimentological character but similar engineering characteristics are also often grouped together as a single unit in descriptive logs, thus losing the high resolution logging detail that is associated with scientific borehole logging and which may be critical in recognising potential event boundaries. The BNFL lithology coding system for Quaternary materials that has been applied to lithology descriptions in the LLWR borehole database is described by Hunter et al (2007a). It is comparable to and probably based upon the AGS coding scheme (AGS, 1999). More recently, Cooper at al. (2006) have proposed a revised coding scheme for unlithified sediments that can be applied to glacial deposits. However, for consistency with Hunter et al (2007), the additional lithology descriptions appended to the LLWR database for the Phase 3 and Stage 6 borehole samples during the preparation of this report have been coded in the same manner as the original BNFL scheme. That geotechnical scheme does not include categories for till and related glacigenic deposits. The document published by McMillan and Powell (1999) contains some useful definitions, as used by the BGS, of the various types of unconsolidated superficial sedimentary materials that comprise glacigenic deposits. For instance, paragraph 4.18 on page 15 of that document contains a definition of ‘till’, i.e., material “deposited 21 directly by and deformed underneath a glacier. It consists predominantly of diamicton, a mixed, unsorted sediment of sand and coarser grains set in a clay/silt matrix”. Tills can be subdivided and classified into a series of genetic categories depending upon the particular glacial environment in which they were deposited. Whereas such genetic tilltypes may be potentially recognisable from high-resolution logging of scientific boreholes and in outcrops, identification of such genetic units with any degree of confidence in BS-5930 engineering-type logs is probably not feasible and has not been attempted in this report. Consolidation of a large range of lithology codes into a limited set representing the dominant grain-size ranges of sediment, i.e., clays, silts, sands, mixed sands and gravels etc., is a commonly-used method of simplifying an otherwise complex data-set for presentation on report figures and accompanying geological interpretation. With glacigenic deposits, where sediments of varying grain-size are often mixed together in an unsorted mass (diamicton), reasonable representation of separate sedimentary units can be difficult to achieve and different outcomes may result when groups of lithology codes are consolidated in slightly different combinations. For the purpose of continuity, the same lithology code combinations adopted by Hunter et al. (2007b) have been used for the preparation of cross-sections in this report. This coding scheme appears to be satisfactory for distinguishing the relatively simple but distinctive lithofacies units of clay-dominant tills, laminated silts and channel-fill sands and gravels. The scheme is incapable of differentiating between the more subtle varieties of glacial till and it is likely that any attempt to achieve this will introduce too much uncertainty into the model. A summary of the code combinations is presented in Appendix C. 22 QRS-1443Y-R1, Version 2.0 (Final) 3 Regional scale lithofacies units This section describes the development of the regional lithofacies framework, and the relationships between the lithofacies units and the BGS lithostratigraphy. The methodology for reviewing the interpretation of the Quaternary sediments on a regional scale was straightforward. Information on the lithofacies and lithological descriptions of the materials examined in the Nirex and earlier studies of the west Cumbria area was assembled and examined to identify units of characteristic material properties relevant to hydrogeological modelling, etc. This focussed attention on the range of sediment types (lithologies) present in the sequence of sediments including clays, silts, sands and gravels, and diamictons, and their degree of consolidation. A limited range of dominant lithologies or combinations of lithologies were identified that formed assemblages within a small number of discrete units. It was recognised that the set of vertical cross-sections (transects) showing the regional Quaternary units that had been constructed during the Nirex investigations could be re-interpreted based on these lithofacies units. Some additional schematic sections were constructed as part of the process of determining the spatial distribution of the lithofacies units across the area and across the LLWR site. These transects formed the original basis for the 3D geological model (Smith, 2009a,b). Accordingly, it is useful to first describe the setting and development of the regional Quaternary units identified by the BGS studies as part of the Nirex investigations, and then show how these were re-interpreted. 3.1 Regional lithostratigraphical units Figure 3-1 is a map of the bedrock surface beneath the Quaternary sediments that was derived from information compiled from seismic surveys and regional borehole data (Nirex 1997a), supplemented by later information from the LLWR. This map shows that the LLWR site is located over the northern margin of an incised palaeovalley that underlies much of Lower Wasdale. The extent of the palaeovalley is illustrated by the sea level contour (0 m), which extends from below the LLWR site well inland towards Gosforth and up Lower Wasdale. The maximum depth of the incision into bedrock is over 60 m below present sea level. From the LLWR site north to Sellafield, the bedrock surface is above present day sea level and is generally covered by less than a 10 m thickness of Quaternary sediments. Sandstone bedrock is exposed on the beach at Seascale, in stream sections inland towards Gosforth, and in the River Calder at Sellafield. Under the Sellafield site, to the north of the present course of the River Calder, the thickness of the Quaternary sediments increases significantly over another 23 incised palaeovalley, which forms part of the major palaeovalley under the present River Ehen, which is incised to a maximum depth of about 50 m below present sea level. Figure 3-1. Contour map of the concealed bedrock surface beneath Quaternary sediments of the LLWR and surrounding area (dashed red line shows the boundary between igneous and sedimentary bedrock) The most comprehensive study of the Quaternary sediments in this area was by the British Geological Survey (BGS), who presented their results in a number of reports for 24 QRS-1443Y-R1, Version 2.0 (Final) Nirex (Nirex, 1997a), in the new memoir for the “Geology of the west Cumbria district” (Akhurst et al., 1997) and in a published article (Merritt and Auton, 2000). These studies established a highly detailed lithostratigraphy with 20 units (separate members, or formations where individual members were not identified) in the southern onshore area around the LLWR, which extends from Ravenglass north to Sellafield and inland to Gosforth and Lower Wasdale. However, it was noted in the article by Merritt and Auton (2000) that; “… many units are recognized primarily by their unconformable bounding surfaces, rather than by their lithology alone…”. This emphasis on stratigraphical features meant that some units (members) had variable lithologies (material properties). Individual members were treated as continuous layers in specific areas, but the same lithological layer could have different names in adjoining areas. Members were assembled into a complex set of overlapping formations, which were in turn aggregated into Groups. The post-glacial units were included in the Solway Drift Group of estuarine, alluvial (river channel or coastal plain environment), organic (peat), aeolian (wind-blown), and mass movement (scree etc,) sediments, previously classified as Flandrian or Holocene post-glacial deposits. The glacial/glacigenic units were subdivided into two Groups on the basis of the source of the deposits. In the BGS Geological Memoir (Akhurst et al., 1997) and in the article by the BGS geologists Merritt and Auton (2000), these were classified as the Central Cumbria Drift Group of units mainly derived from the Lake District; and the West Cumbria Drift Group of units principally derived from the Irish Sea basin and adjacent coastal areas. Younger members of the West Cumbria Drift Group overlie the members of the Central Cumbria Drift Group in the area below the LLWR site. Subsequently, the BGS (McMillan, Hamblin and Merritt, 2005) reclassified some of the members included in formations within the Solway Drift Group as belonging to the Cumbria-Lancashire Catchments Subgroup, part of the Britannia Catchments Group (for fluvial (river), mass movement, organic and cover sand deposits), and other members as belonging to the Great Britain Coastal Deposits Group (for marine, estuarine and coastal deposits, including beach and dune sands). They reclassified the Central Cumbria Drift Group as the Central Cumbria Glacigenic Subgroup, part of the Caledonia Glacigenic Group, and reclassified the West Cumbria Drift Group as the Irish Sea Coast Glacigenic Subgroup of the Caledonia Glacigenic Group. This reclassification of the Quaternary sediments was useful for stratigraphical studies related to surface geological mapping. The Quaternary sediments of the region around the LLWR site are very varied and were mostly deposited by processes related to the ice sheets that covered the area within the last 25,000 years, during the late Devensian period. These sediments include tills (glacial diamictons), coarse fluvioglacial sands and gravels, and silts of 25 glaciolacustrine (lake) origin. The late Devensian sediments are locally deformed by glacitectonic processes, generally related to local re-advances of the ice sheet during the later stages of the glacial period. Some of the land-based (terrestrial) glacial and glacigenic deposits extend beyond the present day shoreline due to deposition during periods of lower sea level before the crustal depression and related rise in relative sea level at the maximum of the late Devensian Glaciation (Heathcote and Michie, 2004). Additionally, some glacial deposits formed at the base of the grounded ice sheet on the floor of the offshore basin during the period of raised sea level at the maximum of the Late Devensian glaciation. These are overlain offshore by younger glaciomarine and marine sediments associated with marine transgression during the deglacial phase of the glaciation and also during the younger post-glacial rise in sea level. Significant falls in relative sea level during the deglacial and post-glacial phase produced breaks in sedimentation (unconformities) within the glaciomarine and marine sequences. The inter-relationships of the various members of the Quaternary deposits were presented in a series of transects across the area in Nirex Report no. SA/97/045 (Nirex, 1997b). These were later supplemented by the production of additional transects through the area around the LLWR (Hunter et al., 2007b; Smith, 2008) as part of the update to the understanding of the Quaternary geology of the LLWR area. These transects included some of the boreholes drilled as part of the investigations of the LLWR and allowed a comparison of the regional lithostratigraphical units of the BGS with some of the detailed lithology units identified in the LLWR boreholes. Figure 3-2 is a highly vertically exaggerated W-E section across the LLWR site and Lower Wasdale to the Lake District foothills. The schematic section was compiled to illustrate the correlation of the BGS Lithostratigraphical units identified in boreholes in the LLWR area in the west (Off-site borehole E, C8 and C11) with the BGS lithostratigraphical units defined in Nirex boreholes to the east (QBH 2A, QBH 16 and AF 2). There is a major facies change identified in the area of borehole QBH 2A. Thick units of lacustrine clays, silts and fine sands (the Holmside Clay Member and the Whinneyhill Coppice Clay Member) are present to the east in Lower Wasdale and are interpreted as the deposits of a deep lake dammed by the Irish Sea Ice Sheet across the mouth of Lower Wasdale against the basement ridge of Muncaster Fell. To the west of borehole QBH 2A, there is an alternating sequence of tills (glacial diamicton) and sandy units, which were linked to the sequence exposed on the low cliffs at Drigg beach. Below the sequence of alternating tills and sands, a sandy unit, the Barn Scar Sand and Silt Member, is interpreted to thicken towards the coast, partly at the expense of the underlying unit of thick till, the Holmrook Till Member. This unit was interpreted as a deposit from the local Lake District Ice Cap, probably from a 26 QRS-1443Y-R1, Version 2.0 (Final) valley glacier that flowed into Lower Wasdale. Below this local till unit, the marine to estuarine or lacustrine deposits of the Glannoventia Formation and the Carleton Silt formation are preserved in lower parts of the palaeovalley. Below these formations, a basal till, the Maudsyke Till contained clasts only from the Lake District and was interpreted as a local valley glacier from Wasdale. At the base of the sequence in a few of the boreholes, the BGS identified a ‘weathered’ till unit, the Drigg Till Formation, which was interpreted to contain in addition to Lake District clasts, a suite of “Scottish” granite clasts. 27 28 Figure 3-2. West-east regional cross-section through the LLWR from the coast to the Lake District showing the BGS lithostratigraphical units QRS-1443Y-R1, Version 2.0 (Final) 3.1.1 Age sequence and correlation of the Quaternary lithostratigraphical units The BGS interpreted the sequence of lithostratigraphical units identified within the Quaternary sediments of the onshore area in an age (time) sequence of environmental changes. Their interpretation, which is presented fully in Merritt and Auton (2000), is summarised and reviewed here from the base of the sequence upward, from oldest to youngest, based on the section shown in Figure 3-2. The weathered appearance of the Drigg Till and its sporadic occurrence at the base of the Quaternary sequence was taken to indicate an early to mid Quaternary age, older than the Devensian glacial deposits of the late Quaternary. Weathered diamictons/tills and palaeosols (fossil soils) of unknown, pre-Devensian age are preserved in valleys within the Lake District (Boardman, 1985), but are much more weathered than the Drigg Till. The separate identity, age and origin of the diamicton identified as the Drigg Till is regarded in this review as uncertain. The spatial relationships of its identified occurrence suggest that it may actually be a variant of the diamicton classified as the Maudsyke Till in the boreholes near the LLWR site, and to be an altered form of basal tills or other coarse basal deposits elsewhere. Fossil material within the Carleton Silt Formation and overlying Glannoventia Formation indicates the upward transformation of a lake or estuary (ria) into fully marine to glaciomarine conditions as a result of rising sea level. Fragments of marine shell from the uppermost (youngest) member of the Glannoventia Formation, the Kokoarrah Shelly Sand Member, were analysed for the degree of degradation of contained amino acids. The peak height ratio of (D)- alloisoleucine to (L)-isoleucine was interpreted to indicate an age of c. 60, 000 years (discussed in Merritt and Auton, 2000). However, review of the original analysis shows that the inferred age was not corrected for temperature or other environmental factors and the age is highly uncertain, but is more likely to be in the range 35,000 to 30,000 years Before Present. The Holmrook Till Member overlies the estuarine to marine sequence and is interpreted as a valley glacier that flowed down Wasdale from the Lake District ice cap, before the arrival of the Irish Sea lobe of the British- Irish ice sheet offshore. With the arrival of the major ice barrier offshore, meltwaters were ponded in the area of Lower Wasdale to form deep lakes, which lead to retreat of the valley glacier up into Wasdale. Sands, gravels and minor silts were deposited from the advancing Irish Sea ice sheet in the west to form the Barn Scar Sand and Silt Member. At the same time, laminated clays and silts were deposited in the deep lake environment of Lower Wasdale to form the Whinneyhill Coppice Clay Member. 29 At the maximum of the Late Devensian Glaciation, the Irish Sea ice sheet advanced over the Barn Scar Sand and Silt Member to deposit the Ravenglass Till. It continued inland over the lake sediments to become confluent with the Lake District ice cap and deposit the Green Croft Till Member. There was then a retreat of the margin of the Irish Sea ice sheet. The outwash deposits of the Kirkland Wood Sand and Gravel Member were deposited adjacent to the retreating ice sheet in the west. At the same time, the valley glacier of the Lake District ice cap had retreated up into Wasdale. Another suite of laminated silts and clays, the Holmside Clay Member, was deposited in the icedammed lake between the two ice masses. A readvance of the edge of the Irish Sea ice sheet, termed the Gosforth Oscillation, deposited the Drigg Beach Till, which only advanced a limited distance into Lower Wasdale. There was no significant readvance of the valley glacier down Wasdale and lake deposits continued to accumulate in Lower Wasdale. Another retreat of the margin of the Irish Sea ice sheet lead to the deposition of the outwash sands of the Drigg Holme Sand Member in the west and continued deposition of lake sediments in the east. This was succeeded by another readvance of the edge of the Irish Sea ice sheet, the Scottish Readvance, which deposited the Fishgarth Wood Till, but to a generally shorter distance inland into Lower Wasdale than the Drigg Beach Till. During the recession following the Scottish Readvance, an interbedded sequence of laminated silty clays and sands was deposited in the coastal area of the LLWR in shallow lakes and braided rivers. Within this Drigg Moorside Silt Member, a thin massive clay unit is interpreted as melt-out or flow till, formed by the decay of stagnant ice. To the east, fluvial and deltaic deposits of sand and gravel formed the Mainsgate Wood Sand and Gravel Member, which covered much of the area of lake deposits. An equivalent unit, the Peel Place Sand and Gravel Member is present to the north and north-east of the area of the LLWR. These sands and gravels were deposited as ice contact deposits, outwash and deltas. A thin clay unit on top of the Peel Place Sand and Gravel may be an equivalent flow till unit to the unit within the Drigg Moorside Silt. Post-glacial deposits partially overlie, and infill hollows and erosion channels in, the glacial and glacigenic deposits. Some of these deposits infill hollows (kettle holes) formed by the melting of large isolated masses of stagnant ice. As part of the Nirex investigations, the peat infilling the hollow at Hallsenna Moor to the NE of the LLWR site was dated by Accelerator Mass Spectrometry of carbon-14 (Nirex, 1997b). The basal peat layer produced an age of 15,770 calendar years Before Present and lies on top of a sequence of organic-poor sediments. This indicates an age for the last glacial advance over the area of more than 16,000 years Before Present. This outline of the revised sequence of development of the onshore lithostratigraphical units agrees with studies of similar thick sequences of Quaternary sediments adjacent 30 QRS-1443Y-R1, Version 2.0 (Final) to the Irish Sea basin in the Isle of Man, eastern Ireland, NW England and North Wales. The revised time sequence indicates that the majority of the sediments were deposited in a fairly continuous manner from the period shortly before the arrival of the Irish Sea ice lobe of the British-Irish ice sheet. Changes in the marine to non-marine nature of the sediments can be directly related to the timing of major changes in relative sea level accepted for the northern part of the Irish Sea over the period. 3.2 Offshore seismic stratigraphy A separate stratigraphy was developed for the offshore Quaternary deposits adjacent to the west Cumbria area before the development of the lithostratigraphical framework for the onshore Quaternary deposits of west Cumbria, (Nirex, 1993; Nirex, 1997c). This was based on the interpretation of detailed seismic surveys, with some calibration from earlier BGS offshore boreholes in the area. A layered, but overlapping and partly crosscutting succession of seven seismically distinctive sequences was identified, which were separated in part by unconformities. Figure 3-3 is a transect offshore from the LLWR to illustrate the relationship of the offshore seismic sequences along the SW-NE line shown on Figure 3-1. The offshore part of the transect is based on seismic line 89-211, which was interpreted by Eaton and Williams in Nirex Report 519 (1993). This interpreted seismic section is reproduced in colour in a recent report on re-examination of the offshore seismic interpretation (Smith, 2010b). That report also reviewed the correlation of the offshore and onshore sequences, but is superseded by this report. For this study, the vertical scale of the seismic section was adjusted from two-way time to physical depth based on estimates of the seismic velocities. Additionally, the horizontal scale was calibrated to distance and the position of the line was adjusted to match the mapped location. 31 32 Figure 3-3. SW-NE cross-section from offshore to the LLWR showing offshore seismic stratigraphical units and onshore BGS lithostratigraphical units QRS-1443Y-R1, Version 2.0 (Final) The offshore portion of Figure 3-3 shows a simplified cross section of the relationship of the seismic sequences. Extension of the section to the SW to cross the incised and infilled Coulderton Channel shows the overlapping nature of the sequences and their thickening offshore, especially into the channels incised into bedrock. The nature of the seismic sequences, calibrated by correlation with the sequences intersected in the BGS offshore boreholes, is described and interpreted in the following summary. This includes discussion of the general relationships in the other seismic sections, not just the section illustrated in Figure 3-3. Sequence 1: Diamict, sand and gravel This consists of an irregular assemblage of diamictons, sands and gravels deposited under terrestrial (land surface) conditions. It is generally less than 10 m thick over bedrock highs, but thicker sequences are preserved in channels and other bedrock incisions. The sequence is interpreted as the accumulated deposits of the tundra (cold) environment during the long period of low sea level in the early and mid Devensian before the late Devensian glaciation. The deposits may include remnant deposits from the previous glacial and interglacial cycle. Sequence 2: Lacustrine This sequence is generally only 2 to 5 m thick. It forms a draped deposit over Sequence 1 and is sometimes absent over topographic highs, but thickens in the channels and incisions. It is interpreted as passive, low-energy deposition in a lacustrine (lake) environment and as rain–out from floating ice in a glacio-lacustrine environment. These conditions are considered to result from increasing crustal downwarping towards the growing Scottish ice sheet that created large lakes, which covered the earlier land surface. Sequence 3: Marine This sequence is dominated by layered clays/muds and silts, which have kilometres of lateral continuity on seismic sections. It cuts across and overlaps the lower sequences, and infills the channel features, where it can be tens of metres thick, but thins over the topographic highs to approximately 5-10 m. It is interpreted as a transgressing marine sequence that replaced the lake environment when increased crustal depression allowed entry of the sea. Sea level continued to rise, producing clay/mud-rich marine sediments deposited in deep water well below wave base in the western part of the area. Rising sea level allowed the sequence to advance and rise towards the onshore area to the NE, where more nearshore sediments were deposited. 33 Sequence 3A: Till This is a locally preserved sequence of clay-rich till, generally no more than 10 m thick. It rests on a planar unconformity, called the “X” unconformity (Nirex, 1993, 1997c), which cuts across the underlying Sequence 3. Sequence 3A and the underlying sequences are over-consolidated, that is their degree of compaction is much greater than that due to the present thickness of overlying sediment. The degree of overconsolidation indicates a thickness of several hundred metres of ice applied for thousands of years. This sequence is interpreted as the product of a continuous cover of the Late Devensian Ice Sheet until the deglaciation phase. During deglaciation, the margin of the ice sheet generally retreated to the north, but there were oscillations in the position of the margin, with readvances and retreats. In the axis of the East Irish Sea, this process in the marine environment caused erosion, reworking and partial removal of the till. At a higher level on land, the similar oscillations in the position of the ice margin produced interbedding and channelling of till and sand and gravel units. Sequence 4: Glaciomarine This sequence is lithologically similar to Sequence 3, but is not over-consolidated. It is dominated by very fine-grained silts and clays with occasional pebbles and shells, which indicate deposition in cold-water conditions. The sequence thickens offshore from <5 m towards the present coast to over 20 m thick in the west. It is interpreted as glacio-marine deposits in relatively deep water during the later phase of the deglaciation of the Irish Sea Ice Sheet. Pebbles in the dominantly muddy sequence are interpreted as dropstones from floating ice. Possible iceberg, or ice-flow, ploughmarks were interpreted in the lower part of the sequence. The base of the sequence is an unconformity that has involved removal of the underlying Sequence 3A till in places, particularly towards the coast. Sequence 5: Marine regression This sequence is not present in the offshore part of the transect shown in Figure 3-6. It is only present from more than 7 km off the coast of west Cumbria and forms a lenticular wedge reaching about 6 m thick in the west. It dominantly consists of fine silty sand, but there is also a limited development of medium to coarse sand towards the coast. The lower boundary is an unconformity, the “Y” unconformity, which truncates the stratification of the underlying Sequence 4. The shelly fauna indicates deposition in brackish and shallow subtidal environments under temperate conditions. This sequence is interpreted as the initial post-glacial deposits during a period of much lower relative sea level. The coarser sands are interpreted as an offshore bar and 34 QRS-1443Y-R1, Version 2.0 (Final) mollusc shells from these sediments were dated to 8310 (±160) radiocarbon years BP (Nirex, 1997d). Sequence 6: Marine transgression This sequence thickens from an edge at about 12 m below present sea level close inshore to about 15 m thick at about 20 km offshore. It is dominated by layered silts with varying proportions of sand and clay and overlaps all lower sequences. In the SW-NE transect (Figure 3-3), sequence 6 rests directly on Sequence 4 with unconformity and oversteps across lower sequences to the NE to rest directly on bedrock highs. The deposits are interpreted as the passive infill of a rising sea level, the main Holocene transgression. The rise in relative sea level involved still stands at some levels, which produced some accumulations of coastal deposits. The post-glacial sediments are continuous upwards with the present sea-bed sediments. Table 1. Proposed correlation of offshore-onshore stratigraphies OFFSHORE SEQUENCE STRATIGRAPHY ONSHORE LITHOSTRATIGRAPHY Drigg Point Sand Formation; Ehen Sequence 6: marine transgression Alluvial Formation; upper part of Hall Carleton Formation Sequence 5: marine regression Sequence 4: glaciomarine Blelham Peat Formation Lower part of Hall Carleton Formation (estuarine and lower members) Holmrook Till and all equivalent and Sequence 3A: till higher till/glacigenic units and interbedded sand and gravel units, and lacustrine units. Sequence 3: marine Glannoventia Formation Sequence 2: lacustrine Carleton Silt Formation Sequence 1: diamict, sand and gravel Maudsyke Till, Drigg Till, and basal deposits 35 3.2.1 Correlation of the offshore seismic stratigraphy and the onshore lithostratigraphy Although there is a data gap in the nearshore zone between the offshore and onshore sequences, this review using the improved understanding of the nature of the sediments, their facies relationships and the unconformities present has identified their simple and direct stratigraphical (time) correlation. This is shown in Table 1. 3.3 Development of the regional lithofacies units More detailed review of the information from the numerous boreholes in the LLWR and Sellafield areas showed that many of the BGS lithostratigraphical units did not have the continuity indicated from the various transects and sections (Michie et al., 2007). In particular, the numerous till units in the upper part of the sequence could not confidently be correlated. A major purpose of identifying units within the Quaternary sequence was to allow classification into units with similar material properties, e.g. hydrogeological properties for modelling groundwater flow. This lead to the adoption of a lithofacies approach to classification of the Quaternary sediments to identify units of similar properties. These were directly related to the lithological/lithofacies logging of the sediments identified in the boreholes. On a regional scale suitable for general hydrogeological modelling, regional lithofacies units of average thickness about 5 m or more could be identified. On the local scale of the immediate LLWR area, individual units down to 1 m thickness or less could be considered. The development of site scale lithofacies units is considered in Section 4. Appendix A provides more detail of the lithofacies approach and the relationship of the lithofacies units to the earlier classifications of the Quaternary sediments in the area of the LLWR. Figure 3-4 shows the re-interpretation of the regional West-East section, (Figure 3-2), through the LLWR area to identify the regional lithofacies units. An additional borehole, C7 within the LLWR site is included to show the lack of continuity of the BGS lithostratigraphical units and the difficulty of correlation of the numerous clay/till units in the upper part of the sequence. In addition, a large part of the middle part of the sequence was reclassified as dominantly sand and not diamicton/till. 36 Figure 3-4. West-east cross-section through the LLWR from the coast to the Lake District foothills showing the regional lithofacies units 37 QRS-1443Y-R1, Version 2.0 (Final) The lithofacies units were assigned the letter codes A to D, in a general sequence from youngest to oldest, with the additional sub-division of the B lithofacies into B1 to B4 from west to east, corresponding to the transition from offshore marine to onshore inland sediments, and recognising lateral facies changes across the area associated with the palaeogeography and position of the edge of the ice sheets. The details of the nature and distribution of the regional lithofacies units are described in Section 3.4, but their characteristics and possible origin are summarised here: A. Onshore and coastal unconsolidated deposits; a heterogeneous collection of generally thin post-glacial deposits near the present land surface. B1. Silt and fine sand; glaciomarine and marine deposits from about 2 km offshore; interbedded diamictons/tills, and coarser sands and gravels are concealed at depth infilling incised depressions into bedrock. B2. Interbedded clayey diamicton (till) and sandy to gravelly layers; formed by deposition from the advances and retreats of the Irish Sea ice sheet. B3. Sand and gravel; glaciofluvial deposits formed marginal to and also under the Irish Sea ice sheet. B4. Laminated silt and clay; glaciolacustrine deposits formed in ice-dammed lakes in Lower Wasdale. C. Diamicton; generally sandy diamicton where derived from the Lake District but more clay-rich under the East Irish Sea. Usually thin units over the bedrock of the Lake District, but thicker accumulations deposited by the valley glacier within Lower Wasdale. D1. Dominantly coarse sands and gravels, with diamictons at the base and margins of the unit against the irregular bedrock surface. These coarse deposits pass laterally into medium-grained to finer sands and silts. The D1 lithofacies is similar to the B3 lithofacies, but is more heterogeneous and more consolidated (overconsolidated) D2. Laminated clays, silts and fine sands, often dark coloured. They indicate an upwards transition from lacustrine to estuarine and fully marine deposits. These sediments are similar to the sediments of the B4 lithofacies, but are more consolidated (overconsolidated). Figure 3-5 is another schematic transect to illustrate the complexity and apparent continuity of the BGS lithostratigraphical units identified in the boreholes along the North-South transect shown on Figure 3-1. Figure 3-6 shows the corresponding 38 QRS-1443Y-R1, Version 2.0 (Final) lithofacies interpretation of the same section as Figure 3-5, where the lithofacies units correspond directly to the dominant lithologies and lithofacies identified in the boreholes. The actual relationships of the lithofacies units are more complicated than indicated schematically in Figure 3-4 and Figure 3-6. The distribution of regional lithofacies units was initially determined by reassigning all the lithostratigraphical units in the BGS transects in Nirex Report SA/97/045 (Nirex, 1997b) and the new BGS transect in Michie et al. (2007) to the new lithofacies units. From these two-dimensional transects, a map of the surface distribution of the lithofacies was produced and a simplified version is shown as Figure 3-7. 39 40 Figure 3-5. North-south regional cross-section through the LLWR showing the BGS lithostratigraphical units Figure 3-6. North-south regional cross-section through the LLWR showing the regional lithofacies units 41 QRS-1443Y-R1, Version 2.0 (Final) Figure 3-7. Map of the LLWR and surrounding area showing the surface distribution of the regional lithofacies units 42 QRS-1443Y-R1, Version 2.0 (Final) This map also shows as a dashed line the concealed extent and limit of the C lithofacies unit. In detail, the limit of the unit is more complex and this is discussed for the LLWR site area in Section 4. A simple three-dimensional digital model of the Quaternary sediments across the area was created (Hunter et al., 2007a). This was subsequently updated in Smith (2008) and in later studies of new information from the LLWR site (Smith, 2009a). In the following description, the general nature of the lithofacies units and their relationships to the BGS lithostratigraphical units is summarised. Details of the nature of the BGS lithostratigraphical units and their occurrence are provided in the publication by Merritt and Auton (2000). 3.4 Lithofacies Units Lithofacies Unit A This unit comprises all the post-glacial (latest Devensian and Holocene/Recent) sediments across the area and their distribution is shown in detail on the published BGS Solid and Drift map of the area (British Geological Survey, 1999). The sediments include scree and head (rock debris and clayey hillwash) on hill slopes, peat, alluvium, alluvial fan and river terrace deposits, lacustrine deposits, blown sand, estuarine and beach deposits, and offshore sediments, dominantly marine sands and muds. BGS lithostratigraphical units onshore included in this lithofacies unit are the Hall Carleton Formation, the Ehen Alluvial Formation, the Blelham (Peat) Formation and the Drigg Point Sand Formation. This unit comprises a wide range of lithostratigraphical units with likely markedly different hydrogeological properties. However, these lithostratigraphical units are of limited spatial extent and / or thickness such that it is not practicable to distinguish and characterise them all at the regional scale. However, they can be subdivided into a limited number of lithofacies units in the local area around the LLWR site. Lithofacies Unit B1 This unit includes offshore glaciomarine and marine deposits, which are partly underlain by glacial, lacustrine and terrestrial deposits. The glaciomarine and marine deposits partially infill deeply-incised depressions (palaeovalleys) cut down into the pre-Quaternary bedrock. The offshore deposits are predominantly bedded muds, silts, sands and minor gravels. These formed in a dominantly marine environment, even those laid down during the Late Devensian glaciation, and therefore can be treated as a single lithofacies assemblage. For the 2007 study (Hunter et al., 2007b), the B1 lithofacies was also taken to include the underlying glacial, lacustrine and terrestrial 43 deposits. This review has identified the correlation of the older sequence with the regional lithofacies units. This subdivision of the offshore sequence is discussed below. Offshore Lithofacies In the 2007 studies, the detail of the offshore sequences was not addressed and the correlation with the onshore units was not determined. The B1 lithofacies unit was applied to encompass all of the offshore Quaternary sediments until the recognised deficiencies in their understanding, nature and correlation could be resolved. That was subsequently achieved and was summarised in Section 3.2 of this review. Because the upper part of the offshore sediments is dominantly formed of clays/muds, silts and fine-sands, their overall categorisation as B1 is a reasonable approximation. However, a more detailed correlation is now possible. Figure 3-3 shows the separate offshore seismic sequences and the onshore lithostratigraphy in the SW-NE, offshore to onshore transect. Figure 3-8 shows the integrated lithofacies interpretation of the transect. This shows that Sequence 1 is equivalent to lithofacies unit D1, that Sequences 2 and 3 are equivalent to lithofacies unit D2, that Sequence 3A is equivalent to lithofacies unit C, and that Sequences 4, (5) and 6 are equivalent to lithofacies unit B1. Lithofacies unit B1 forms the sea floor and covers all the underlying units. In the nearshore environment, the B1 unit forms only a thin layer above the coarse deposits of lithofacies unit D1, which are similar to the coarse sands and gravels of the B3 lithofacies unit. Lithofacies Unit B2 This unit includes a sequence of interbedded tills (glacial diamictons), and sands and gravels. All these deposits were principally derived from the Irish Sea and formed during the Main Devensian glaciation and subsequent readvance stages. From the BGS lithostratigraphical scheme, the lowest part of the unit is formed by the Ravenglass Till Member, and upwards there is an alternation consisting of the Kirkland Wood Sand and Gravel Member, the Drigg Beach Till Member, the Drigg Holme Sand Member, the Fishgarth Wood Till Member and the Drigg Moorside Silt (including an unnamed Till) member. All the till (glacial diamicton) units are noticeably clay-rich and the upper till units are almost entirely clay, formed as melt-out or flow tills. North of the LLWR site, the lithofacies cuts across older sediments to rest directly on the bedrock. Because this lithofacies is the host for the LLWR, a brief summary of its characteristics as observed at local coastal exposures is provided here. The type locality of the Ravenglass Till Member is in low sea cliffs south of Ravenglass, where 4 m of the till is exposed. It is a very stiff, moderate brown, massive, matrix supported, sandy, silty, clay diamicton, which contains moderately well dispersed, 44 QRS-1443Y-R1, Version 2.0 (Final) subangular to subrounded clasts up to boulder size. These clasts are of both Lake District and more northerly/ coastal provenance. This mixed origin may be due to reworking of clasts into the till from the underlying older deposits derived from the Lake District. The lower boundary of the till is not seen at the outcrop, but in boreholes the till is interpreted to rest unconformably on older sands and gravels. The upper boundary of the till is exposed at the coastal outcrops near Ravenglass and is generally erosional, being overlain by sands and gravels, and also glacitectonites associated with the emplacement of younger till. However, the equivalent of the younger Drigg Beach Till was not identified and the till overlying the sands and gravels (the Drigg Holme Sand Member) was interpreted as the Fishgarth Wood Till. The upper part of the overlying sequence of interbedded tills, and sand and gravels is well exposed on the coastal section near Drigg west of the LLWR. On the foreshore and base of the sea cliffs, 3 m of the Drigg Beach Till Member is exposed. This comprises a dusky red to weak red, stiff, plastic, clayey silt diamicton with well dispersed sand grains and subangular to well rounded pebbles up to 3 cm in diameter, and also some reworked marine shells. The till is interpreted to have formed from a readvance of the ice sheet from the Irish Sea. Overlying the Drigg Beach Till Member with an undulating erosion surface is the Drigg Holme Sand Member, which is about 2.5 m thickness of fine to medium-grained sand with stringers of granule gravel and channels containing clast-supported gravel. The member becomes increasingly affected by planar, subhorizontal shearing towards the base of the next overlying till. These features could indicate that the member was laid down as glaciofluvial outwash between two glacial readvances, but could be produced sub-glacially by lifting/floating of a thin ice sheet margin. The overlying Fishgarth Wood Till Member is only 0.8 m thick and is a hard, moderate reddish brown, calcareous, massive, matrix-supported, sandy silty clay diamicton with well-dispersed (rare), subangular to rounded clasts up to 15 cm diameter. Coal, black shale and shell fragments are prominent. Conformably overlying the thin till is a 2.1 m thick sequence of pebbly gravels and sands. These may be a thin lateral equivalent of the upper part of the thick, coarse sand and gravels that form the Peel Place Sand and Gravel Member inland that is part of the B3 lithofacies described next. There is a planar, tectonised layer developed below the uppermost glacigenic unit, the Drigg Moorside Silt (and Till) Member, which is 1 to 1.9 m thick. Later deformation features are seen to affect the interbedded till, and sand and gravel units at the Ravenglass and Drigg beach outcrops and elsewhere. These include folding of the Ravenglass and Drigg Beach Till Members and complex faulting of the higher units. Faulting includes low-angle (near horizontal) thrust structures, which can duplicate the strata, and high-angle (near vertical) fault structures, which can disrupt 45 the continuity of layers. These deformation features are related to readvances, and other marginal processes, of the ice sheet affecting the underlying deposits. Lithofacies Unit B3 This unit is dominantly composed of thick sand and gravel deposits and appears to have two main developments. A lower development occurs below the B2 unit in the area of the LLWR site and appears as a series of channel developments cutting into older deposits. Some of these channels also cut into and remove the lower tills within the B2 unit and indicate a series of phases of incision during the development of the B2 lithofacies. The incisions were probably related to periodic subglacial release of icedammed lake water from Lower Wasdale. This lithofacies unit includes the BGS lithostratigraphical unit, the Barn Scar Sand and Silt Member, but also includes parts of the sequence previously defined as the underlying Holmrook Till unit, and also parts of the overlying sequence where the till units are not present. Another development occurs inland marginal to the B2 lithofacies unit, and further inland similar sands and gravels overlie the following B4 lithofacies unit. The equivalent BGS lithostratigraphical units are the Peel Place Sand and Gravel Member and the Mainsgate Wood Sand and Gravel Member respectively. At the Peel Place Quarry, about 2 km NE of the LLWR, the workings are exploiting about 11.5 m thickness of the Peel Place Sand and Gravel Member. Excavations have revealed an underlying sequence of about 1 m of lacustrine finely laminated sands and silts. There is a gradational upwards transition into thickly interbedded, cross-stratified sand and upwards fining gravel at the base of the Peel Place Sand and Gravel Member. This mainly comprises well-bedded, cross-stratified gravel and sand that typically forms an upwards coarsening sequence, from thickly interbedded sand and cobble gravel, into gravel with low-angle planar cross-stratification and trough cross bedding. Interbeds of sandy silt with weakly developed lamination occur towards the top of the member. Above this, forming the top of the quarry face and underlying the land surface behind is about 1 m thickness of a contorted succession of interstratified hard, sandy, clayey diamicton (flow till) and silty fine-grained sand. This unit may be the lateral equivalent of the uppermost and youngest till unit, within the Drigg Moorside Silt (and Till) Member in the Drigg beach succession. In places, the diamicton is overlain by a thin (up to 70 cm) bed of medium to fine grained sand, possibly of aeolian origin. Lithofacies Unit B4 This unit, which occurs in Lower Wasdale, dominantly inland of the glacial readvance limit during the Gosforth Oscillation, consists of glaciolacustrine deposits of 46 QRS-1443Y-R1, Version 2.0 (Final) interlaminated silts and clays, but also includes in the middle a till unit that is equated with the Ravenglass Till Member. The BGS lithostratigraphical units are the Whinneyhill Coppice Clay Member, the Green Croft Till Member and the Holmeside Clay Member, all of which relatively clay–rich and of fairly uniform properties. Lithofacies Unit C This unit consists of till as well as sands, gravels, silts and clays, deposited across the area from ice moving westwards from the Lake District and extending beyond the area of the LLWR site at depth. The till is generally a sandy stony diamicton, in contrast to the clay-rich upper till units in unit B2, and contains clasts predominantly of the Borrowdale Volcanic Group, the Eskdale Granite and the Ennerdale Granophyre, all derived from the Lake District. The BGS suggest that the till was deposited before the build-up of the main Late Devensian ice sheet that travelled southwards along the Irish Sea and coastal plain (Merritt and Auton, 2000). The lithofacies unit is partially equivalent to the BGS lithostratigraphical unit, the Holmrook Till Member at depth beneath Lower Wasdale and the LLWR Site, and the equivalent Scale Beck Till Member at outcrop inland in the Lake District. However, it was identified that the lithofacies unit under the LLWR site area is generally thinner than the thickness of the Holmrook Till unit interpreted by the BGS, and is removed by channel incision of the overlying B3 unit in places. Over the bedrock high to the north of the LLWR site, the lithofacies unit is absent, probably removed by the Irish Sea ice sheet advancing from the north-north-west. In the offshore area, where only one till unit (Sequence 3A) is present, this is of the same lithofacies as the C lithofacies in the onshore area. As explained earlier in Section 3.2 and shown in Table 1, Sequence 3A is stratigraphically (time) equivalent to all the till and interbedded sand and gravel units of the onshore area. The fluctuations in the position of the eastern margin of the ice sheet over the onshore area did not extend into the area of the East Irish Sea. In that area, a continuous cover of grounded ice was maintained until the final deglaciation and therefore there was no development of interbedded tills, and sand and gravel deposits. Sequence 3A is equivalent to lithofacies C in the offshore area, and was partly removed by erosion during the marine regression and transgression in the post-glacial phase. Lithofacies Unit D In the onshore area, the D lithofacies consists of all the older sediments present at depth infilling the palaeovalley in Lower Wasdale below the C lithofacies (there partly equivalent to the BGS lithostratigraphical units of the Holmrook Till Member and Scale 47 Beck Till Member to the east), which oversteps the D lithofacies to the east to rest directly on the bedrock as shown on Figure 3-3. The basal part of the lithofacies may be significantly older than the overlying sediments and is weathered in places and of mixed character. It contains a wide range of sedimentary types ranging from diamictons of mixed provenance (containing clasts from the Lake District, and from the floor of the Irish Sea to Southern Scotland) to sands and silts. The upper part of the lithofacies consists dominantly of lacustrine to glaciomarine clayey siltstones and fine sandstones. There are little data with which to characterise these deeper sediments across the Lower Wasdale area, but they can be separated into different subfacies under the LLWR site and adjacent area where the density of data is greater, although still relatively limited compared to the borehole information on the upper lithofacies. In the offshore area, the D lithofacies is interpreted to be equivalent to offshore Sequences 1, 2 and 3. These were described in Section 3.2 and can be summarised as follows: Sequence 1 at the base consists of an irregular assemblage of diamictons, sands and gravels; the overlying Sequence 2 consists of a thin drape of lacustrine sediments; and Sequence 3 consists of layered marine sediments that overlap the lower sequences. These sequences lie beneath the single till of Sequence 3A and the late-glacial to postglacial marine deposits of Sequences 4 to 6. Sequences 1, 2 and 3 were identified as similar to the succession forming the D lithofacies in Lower Wasdale. The consolidation of the sediments forming Sequences 1 and 3 was measured by oedometer tests on undisturbed clay samples from offshore borehole B3 and reported in the BNFL/Nirex report on the interpretation of their offshore high-resolution seismic survey (Nirex, 1997e). The results showed overconsolidation due to loads of over 780 kPa, much greater than the present sediment overburden, and were interpreted to indicate the load from a grounded ice thickness of several hundred metres, the main ice sheet. For the 2007 study (Hunter et al., 2007b), the D lithofacies was treated as a single unit of generally silty character, but this study has confirmed that the lithofacies can be subdivided regionally as follows: The D1 lithofacies dominantly consists of coarse sands and gravels, with diamictons at the base and margins of the unit against the irregular bedrock surface. Some of the diamictons are probably not glacigenic, but may have formed as talus/scree accumulations of coarse debris adjacent to cliffs (suggested by Eaton et al., 1996/97), or as lacustrine to marine beach deposits, which may have reworked older sediments. 48 QRS-1443Y-R1, Version 2.0 (Final) These coarse deposits pass laterally into medium-grained to finer sands and silts, which may have formed in nearshore environments. The D1 lithofacies is similar to the B3 lithofacies, but is more heterogeneous and more consolidated (overconsolidated). In the onshore area, the D1 lithofacies is equivalent to the sandier parts of the BGS lithostratigraphical units, the Maudsyke Till Member and the Drigg Till Formation, but also to the sandier parts of the Carleton Silt Formation and the Glannoventia Formation. In the offshore area, the D1 lithofacies is equivalent to Sequence 1. The D2 lithofacies consists of laminated clays, silts and fine sands, often dark coloured. They indicate an upwards transition from lacustrine to estuarine and fully marine deposits. These sediments are similar to the sediments of the B4 lithofacies, but are more consolidated (overconsolidated). Equivalent BGS lithostratigraphical units in the onshore area include the Glannoventia Formation comprising the Kokoarrah Shelly Sand Member, the Stubble Green Silt Member and the Carleton Hall Clay Member, and the Carleton Silt Formation. The D2 lithofacies was also taken to include clay-rich basal sections classified as parts of the Maudsyke Till Member and the Drigg Till Formation. Equivalent units in the offshore area are Sequences 2 and 3. At the LLWR, the detailed cross-sections presented in Section 4 indicate that it is possible to sub-divide the D2 lithofacies into the lower fine-grained lacustrine and estuarine deposits (termed D2) and upper interbedded silt and sand marine units (termed D3). This detailed subdivision is not possible at the regional scale due to the limited data available with which to characterise the D lithofacies. Therefore, in this regional Section and Appendix A, the upper interbedded (brown) silt and sand marine units are included within the D1 lithofacies. Within the site and in the adjacent area, thick sands, either near-shore facies variants of the silty units or possibly incised fluvial channels, are identified within the D lithofacies. On the regional sections, these sands are also included in the D1 lithofacies. However, in the detailed sections, these are identified as separate sand units and they have not been assigned to one of the named sub-units of the D lithofacies for the present. The spatial distribution of these units as currently understood is shown in the cross-sections in Section 4, but there is uncertainty about their spatial distribution. This review has identified the correlation between the regional lithofacies discussed above and the lithofacies packages previously defined at the LLWR site. Therefore, the classification of the Quaternary sediments was unified and the regional lithofacies classification applied at the LLWR site, but with more detailed classification of the lithofacies and higher-resolution identification of their spatial distribution. This is discussed more fully in the following section. 49 50 Figure 3-8. SW-NE cross-section from offshore to the LLWR showing the regional lithofacies units QRS-1443Y-R1, Version 2.0 (Final) 4 Characterisation of the Quaternary geology at the LLWR site As discussed in the earlier sections of this report, the methods used by previous researchers to characterise the Quaternary sediments at the LLWR site (and also in the surrounding region of west Cumbria) can be summarised in terms of three separate, alternative models. These are: The ‘events-based stratigraphy’ developed by G. Eaton and subsequently applied by BNFL to characterise the LLWR site for the 2002 PCSC (BNFL, 2002a). The regional lithostratigraphical framework developed by the BGS during and subsequent to the Nirex deep repository investigations (Nirex, 1997a). The ‘lithofacies’ concept proposed by Hunter et al. (2007b) for the Reinterpretation of the LLWR Site Quaternary Geology (LLWR Lifetime Project) in 2007. Each of these models is based upon interpretations of essentially the same source data, namely, outcrops of Quaternary sediments, sample lithology descriptions from site investigation boreholes, geological descriptions from scientific investigation boreholes and pits (varying from low to high quality), and non-intrusive geophysical surveying. However, each model assessed the relevant importance of these separate sources of data for the purpose of characterising the Quaternary geology at the LLWR site in different ways. Although each of the three geological models is ultimately linked to current understanding of the glaciation history of north-western England during the period known as the Main Late Devensian (MLD) (described in Section 3), each model incorporates different assumptions for defining the processes involved in the events, the sedimentary material deposited (or eroded) by an event and also for recognising such events in the available data. Furthermore, the objective of each of the three models was also slightly different: the events-based stratigraphy model was developed to describe a general understanding of the geology beneath the LLWR site and to be able to predict lithological distribution across the site (BNFL, 2002a); the lithostratigraphical model is part of the standard approach to regional geological understanding of Quaternary deposits throughout Great Britain adopted by the BGS (McMillan et al., 2005); while the lithofacies model developed by Hunter et al. (2007) was devised in response to the need to provide a 51 materials focused description of the geology of the LLWR for hydrogeological modelling, etc. The first two of these models were derived principally from the construction of a few schematic geological cross-sections using selected groups of simplified borehole lithology logs - usually the limited set of deeper boreholes that penetrate through the entire Quaternary sequence. However, the variability of the Quaternary sedimentary deposits beneath the LLWR site is sufficiently complicated such that slight changes in the orientation of the cross sections and the particular groups of borehole logs selected (and omitted) for each cross-section can result in markedly different interpretations of the Quaternary geology and stratigraphy. The third (lithofacies) model was derived from a non-selective interrogation of all of the available borehole data. Such an approach was deemed to be necessary to investigate and characterise the rapid changes in lithology that can be observed between certain groups of relatively closely-spaced boreholes. This alternative approach was also deemed to be appropriate for a lithology dataset that exhibits no obvious layered sequence that can consistently, or confidently, be matched (unit for unit) to the published BGS regional Quaternary lithostratigraphy. In a general sense, some or many of the unit boundaries that define the different lithofacies of the third model are also event boundaries of the type proposed for the first (i.e., events-based stratigraphy) model. Similarly, the lithofacies units can be generally related to the BGS’ lithostratigraphy, as described in Section 3. However, analysis of the LLWR borehole lithology database in the context of lithofacies revealed evidence for the existence of channel-like bodies of coarse-grained sediments (sands, gravels and cobbles etc) which are interpreted as being deeply incised into underlying layers of finer-grained sediment. Extensive coalescing of these channels also seems to have resulted in the erosion of most of a layer of thick, clayey till that was probably deposited across the entire LLWR site, but now only exists as limited remnants. Evidence for the existence of the channel-like bodies of coarse-grained sediments can be found in the database of borehole lithological descriptions recorded from the LLWR site, where comparison of particular pairs of boreholes located in relatively close proximity to each other show significant contrasts in lithology at approximately equivalent depths. Two examples of such contrasting lithology in borehole pairs are shown in Table 2 and Table 3. The boreholes are also plotted on the cross-sections shown as Figure 4-9 (and Figure C-2) and Figure 4-12 (and Figure C-3). 52 QRS-1443Y-R1, Version 2.0 (Final) Table 2. Comparison of sample lithology for boreholes 8646 and 7525 below 0 m OD Borehole 7525 Borehole 8646 Top elev -0.91 -2.91 -3.11 Bottom elev -2.91 -3.11 -9.81 -9.81 -14.61 -14.61 -16.11 -16.11 -16.61 -16.61 -17.11 Lithology description Firm to stiff brown slightly sandy slightly gravelly CLAY. Gravel is subangular to rounded fine to coarse of basalt and granite. Grey/brown fine to coarse SAND and subangular to subrounded fine to coarse GRAVEL of mixed igneous and metamorphic lithologies. Stiff brown slightly gravelly slightly sandy CLAY with occasional cobbles. Gravel is subangular to rounded fine to coarse of basalt and granite. Reddish brown very clayey fine to coarse SAND and subangular to subrounded fine to coarse GRAVEL of mixed igneous and metamorphic lithologies. With some subangular to Red slightly gravelly fine to coarse SAND. Gravel is subangular fine to coarse of sandstone. (Highly Weathered Sandstone) Moderately weak to moderately strong red fine to medium grained SANDSTONE. Highly weathered. Recovered as sandy subangular fine to coarse gravel sized fragments. Moderately strong red fine to medium grained SANDSTONE. Moderately weathered. Recovered as subangular cobbles and gravel sized fragments. End of borehole Top elev Bottom elev 0.15 -3.15 -3.15 -8.35 Lithology description Very dense brown slightly gravelly very clayey SAND with some thin bands of brown clay. Gravel is subangular to subrounded fine to medium of mudstone and igneous lithologies. Very dense multicoloured sandy very clayey angular to subrounded fine to coarse GRAVEL of sandstone, mudstone, shale, limestone and igneous lithologies with frequent subangular to subrounded cobbles an… (description unfinished) End of borehole 53 Table 3. Comparison of sample lithology for boreholes C8 and 8649 below -10 m OD Borehole 8649 Borehole C8 Top elev Bottom elev -10.87 -11.69 -11.69 -15.08 -15.08 -17.02 -17.02 -18.63 -18.63 -18.71 -18.71 -19.57 -19.57 -26.23 -26.23 -26.49 -26.49 -27.32 -27.32 -33.88 -33.88 -34.14 -34.14 -36.71 Lithology description moderate brown very sandy CLAY with many fine sand bands fine to coarse GRAVEL with a silty sand matrix fine to coarse GRAVEL with a silty sand matrix (no recovery) Top elev -10.1 Bottom elev -16.1 -16.1 -18.6 -18.6 -20.1 -20.1 -31.6 olive grey silty CLAY dark grey very silty fine SAND olive grey laminated sandy SILT brownish grey silty CLAY with gravel brownish grey silty CLAY with gravel brownish grey silty CLAY with gravel light brown silty fine SAND with gravel SANDSTONE End of borehole -31.6 -33.6 Lithology description Brown gravelly fine to coarse SAND. Gravel is subrounded fine to medium of basalt, sandstone and granite. Dark brown fine to coarse SAND and subangular to subrounded fine to medium GRAVEL of basalt, sandstone and granite. Brown fine to medium SAND with frequent subrounded cobbles of granite and basalt. Multicoloured slightly sandy subrounded coarse GRAVELS and COBBLES of granite and basalt. Moderately weak red fine to medium grained SANDSTONE. Moderately weathered. End of borehole In these two examples, thick borehole intersections of mixed sand, gravel and cobbles occur at similar elevations and are laterally equivalent to comparable thicknesses of finer-grained lithologies that will have been deposited by different sedimentary processes. In the first example, at an intermediate depth, the stiff, slightly gravelly, slightly sandy clay intersected in borehole 8646 most probably represents a thick till and in this study it has been assigned to the ‘C’ lithofacies. In the second example, at a deeper level, several metres of olive-grey, laminated, sandy silt most probably represent widespread marine sediments of the Glannoventia Formation (BGS) and this lithology has been assigned to lithofacies ‘D2’ in this study. In both of these examples, the intersections of comparable thicknesses of mixed sand, gravel and cobbles in boreholes 7525 and 8649 – here both assigned to lithofacies B3 – indicate deposition by an entirely different process. While glacitectonic deformation is one possible mechanism to explain the juxtaposition of these contrasting lithologies, fluvial incision by sub-glacial outburst drainage is a simpler and more realistic option when the surrounding palaeogeography during the 54 QRS-1443Y-R1, Version 2.0 (Final) late Quaternary Period is taken into consideration. Periodic outbursts of glacial meltwater (and sediment) from one or more ice-dammed lakes in the Wasdale area are likely to have occurred while the water level in these lakes was significantly higher than the relative sea-level at the time (and both the relative sea level and the level of the water in the ice-dammed lakes was higher during the glacial phase than the relative sea level, ordnance datum, at present). The incised channel exposure at Nethertown (north of LLWR) is an excellent example of such a deposit (Figure 4-10). If the existence of incised, sand- and gravel-filled channels is accepted, then such features represent distinct ‘events’ whose boundaries may be deeply incised into underlying sediment packages. Thus the lithofacies model allows for the possibility that components of younger sedimentary units (i.e., events) are likely to occur within, or even below, older units. This differential vertical displacement of events is matched in the model by similar lateral displacements because sedimentary material of local (i.e., Lake District ice) provenance is likely to exist co-mingled with material of Irish Sea ice provenance. In the lithostratigraphical framework model, glacial deposits derived from these two, different source areas are classified as separate geological ‘formations’. In a simplistic sense, the lithofacies model of the Quaternary deposits at the LLWR site can almost be regarded as an inverse representation of the events-based stratigraphy model. This is because the latter model (illustrated by Figure 4-14) depicts the Quaternary deposits at LLWR as consisting of a thick ‘Main Diamict Formation’, composed principally of various types of till deposited by a MLD glacier, which is overlain by a sequence of mainly fluvial and lacustrine sediments, whereas in the lithofacies model there is no equivalent of the Main Diamict till assemblage and most of the preserved, recognisable tills exist at a shallower depth and are seen to overlie a thicker deposit of sand and gravel which is more likely to be of fluvial outwash origin. In many of the cross-sections of the LLWR Quaternary geology prepared for this report, the lithofacies units can give the appearance of being simply another layered sequence model. This is the consequence of each cross-section representing only a small, 2-dimensional slice of one part of the LLWR site. When these lithofacies units are considered in 3 dimensions on a wider, semi-regional scale, particularly the B3 facies, their localised spatial distribution (controlled in part by the bedrock topography and the Wasdale palaeovalley) and their cross-cutting and lateral-transitional relationships to adjacent glacial deposits become more evident. These important features of the lithofacies model distinguish it from the alternative, layered-sequence models at the LLWR site and some of the essential differences between the three models are summarised in Table 4. 55 Table 4. Summary of the features of the 3 basic geological models used to characterise the LLWR and the immediately adjacent areas of interest Geological Methodology Principal assumptions model The Gosforth Oscillation glaciation event compacted the underlying MLD sediments, thus creating a density-contrast surface between the two Used selected site events which can be detected by geophysical investigation boreholes from (seismic) surveying. Events-based within and in the immediate stratigraphy vicinity of the LLWR site; model of the ‘event units’ are identified from each event can be identified as a unit and the LLWR site partly on the basis of seismic boundaries separating these units can be identified reflectors and partly upon from borehole lithology descriptions (in lithology. combination with seismic reflections). Each glacial event deposited sedimentary material in an accumulating sequence and the materials Each event unit may contain a wide variety of material types, depending upon particular glacial environments (e.g., sub-glacial, marginal ice etc.). Used a limited number of Regional lithostratigraph -ical framework surrounding the LLWR site Individual, discrete lithostratigraphical units, scattered outcrops, pits and including very thin units, can be correlated as a scientific boreholes, mainly uniform, systematic, layered sequence across large outside of LLWR, logged in areas on a regional scale between scattered, great detail and correlated isolated boreholes, outcrops and other data over large distances. Units sources. identified using a combination of lithology, sedimentary Each lithostratigraphical unit may consist of a gradation of different material types. fabric and additional ancillary information. Lithofacies LLWR site scale model of both LLWR site and the surrounding region Correlation of similar material types between Used as many of the site boreholes in close proximity to each other as investigation boreholes from simple lithology units, with no inferred genesis, within and around the LLWR offers the least interpretive and the most site as possible, basing reproducible characterisation of the LLWR correlations entirely upon Quaternary deposits for hydrogeological simplified lithological modelling even if the resulting spatial distribution descriptions to identify gross of units is not a uniform sequence of layers. patterns of distribution of different material types. Nearby coastal exposures are useful analogues for the materials being described in samples from the LLWR site borehole database. 56 QRS-1443Y-R1, Version 2.0 (Final) Regional scale The complex spatial distribution of bulk lithology types seen within the LLWR site is supported by a large dataset and there is nothing to suggest that the site is radically different from the immediate surrounding area. Used the same limited Ice–dammed lakes in Upper Wasdale formed regional borehole dataset as reservoirs of water and outwash sand and gravel the lithostratigraphical that could be released subglacially to form tunnel framework model, but re- valleys if the Irish Sea ice sheet thinned and/or interpreted in the context of lifted off its base. These tunnel valleys probably simplified, bulk lithology and incised deeply into older glacial and marine with regard to the regional deposits, replacing tills and laminated silts with palaeogeographical setting. mixtures of sand and gravel. Several other palaeographical features of the Wasdale valley area (described in Sections 2 & 3) probably restricted ice movement and influenced local depositional environments and sediment facies. 4.1 The concepts of ‘Lithostratigraphical Formations’ and ‘Lithofacies’ One recurring question which is often associated with the events-based stratigraphical model is that the exact basis for determining the bounding discontinuities between the separate event formations in many of the borehole lithological logs, often within an interval of seemingly uniform or similar lithology, is not recorded and, despite significant searching, could not be found for incorporation into this report. Similarly, the distinction between inference (of deposition and provenance) and description is not always clear. McMillan et al. (2006, page 6, para 1.2) state that: “By their nature, many Quaternary units are strictly allostratigraphical, i.e., [they are] defined and identified on the basis of their bounding discontinuities”. Also, “The lithology of many Quaternary deposits is determined by the medium of transport, the medium of deposition and by provenance. Inferred genesis and provenance may aid lithostratigraphical classification of heterogeneous superficial deposits but always there should be a clear distinction made between inference and description”. The lithofacies model proposed by Hunter et al. (2007a) attempts to follow these important principles by avoiding both the attempted identification of stratigraphical bounding surfaces and also the direct association of particular stratigraphical units 57 with established glaciation events and their assumed depositional processes. (Although, the lithofacies can be related to the lithostratigraphy and hence such events: Section 3). However, the lithofacies model does ascribe a general genetic origin to the clayey, stony diamictons, which are assumed to be sub-glacial tills, and the thick, elongate zones of sand and gravel, which are assumed to be either sub-glacial or subaerial melt-water channels. Both the events-based stratigraphical model for the LLWR site and the regional lithostratigraphical model for surrounding region utilise the term ‘formation’ in naming the sedimentary units. McMillan et al. (2006, page 21), quote the definition of a formation as “the primary formal unit of lithostratigraphical classification used to map, describe and interpret the geology of a region”. Also, the formation “is generally defined as the smallest mappable unit and has lithological characteristics that distinguish it from adjacent formations”. The significance of regional dimensions for the definition of geological formations is emphasised by the statement that “In Britain, formations should be mappable and readily represented on a 1:50,000 scale map”. McMillan et al. (2006) discuss the issues associated with the application of these definitions to glacigenic deposits and note that the key tests (of a formation) include demonstrating type section(s) where the top and base of the unit can be observed (recognising that these boundaries may vary laterally) … and tracing lateral continuity (accepting that lateral and vertical variation is likely). A geological formation may or may not be divided into ‘members’, which are “lithologically distinctive units interbedded within more regionally significant glacigenic formations”. Even lower in stratigraphical rank than the ‘member’ is the ‘bed’, “commonly applied to distinctive units that may be thin and laterally inextensive”. This discussion of terminology is not of mere academic interest; the use of the term ‘formation’ in a geological model of the LLWR Quaternary deposits implies that the sedimentary units have recognisable bounding surfaces and a degree of lateral continuity. The use of the concept of lithofacies has been described by McMillan and Powell (1999). In paragraph 4.5 on page 8 of that report, it is stated that “Bodies of unconsolidated sediment with specified physical and sedimentary characteristics are commonly referred to either (as) a single lithofacies or a lithofacies association. BGS geologists have in recent years employed lithofacies codes to enable a first-stage, nongenetic description of field and borehole sequences.” Such a ‘first stage’ approach is ideally suited to characterising the Quaternary deposits beneath the LLWR site where the limited number of outcrops reveal only a small part of the stratigraphical sequence and where the bulk of the information is in the form of engineering-type lithology descriptions from borehole logs. A comparison of Figure 3-2 and Figure 3-4 shows how a sequence of complex glacigenic sediments can be re58 QRS-1443Y-R1, Version 2.0 (Final) interpreted using the lithofacies concept as a more realistic alternative to a layered sequence model. 4.2 Additional review of data for this report The borehole cross-sections presented in Hunter et al (2007b) were preliminary in nature and were intended to interrogate the entire borehole lithology database on a non-selective basis and without any pre-conceptions of stratigraphical equivalence or depositional direction being applied. The outcome of that process was the observation that the spatial distribution of Quaternary lithologies beneath the LLWR exhibits distinctive lateral discontinuities between separate zones that apparently consist of different lithologies (i.e., lithofacies). These lithofacies zones (which in 2007 were termed lithofacies ‘packages’), are probably composite bodies of sediment resulting from more than one single geological ‘event’, may be potentially mappable and could exert significant influence in a hydrogeological regime. As noted by Jefferies (2009), one obvious disadvantage with the cross-sections presented in the 2007 report was that it was not possible to make direct comparisons between those sections and the existing, more widely-used cross-sections from older reports. Therefore the additional cross-sections presented in this report are intended to redress the omission of such comparative sections in the 2007 report. The new crosssections also afford the opportunity of assessing how closely the lithofacies zones defined in the 2007 report predicted the lithologies that were recorded in the recent Phase 3 and Stage 6 boreholes, particularly at intermediate depths. The additional geological studies of the LLWR site undertaken for this report included the following: Updating of the contour map of the sandstone bedrock surface underlying the Quaternary deposits at LLWR. Preparation of a set of borehole cross-sections approximately equivalent to Sections 1, 2 and 3 of Eaton (1996), all of which are approximately northwest-southeast in orientation and generally parallel to the long axis of the LLWR site. Preparation of four additional borehole cross-sections also orientated parallel to the long axis of the LLWR site to fill the gap between the replicated Eaton sections. All of the longitudinal sections are orientated with a ‘sea to shore’ viewpoint to enable easier comparisons to be made with previously published work. Preparation of ten borehole cross-sections orientated generally perpendicular to the long axis of the LLWR site. Some of these sections were located and orientated to 59 investigate specific lithology zones identified on Figure 21 of Hunter et al. (2007b) – the ‘channels’ map, while others were located to allow comparison with regional transects and one of the Halcrow (2010) geophysical profiles. All of these crosssections have been extended to reach the beach so that the relative position of the Drigg Beach Till and other exposed formations can be compared to the borehole lithofacies units. The locations of these new borehole cross-sections are shown on an index map of the LLWR site as Figure 4-1. The cross-sections have been prepared in a similar manner to the systematic ‘thick sections’ (i.e., ‘wide’ sections) described in Hunter et al (2007b) using the software Surfer9 (©Golden Software). However, unlike the 2007 crosssections, the positions and orientations of these new sections have been chosen to investigate specific geological features at the LLWR. Another difference between the 2007 cross-sections and those prepared for this report is their reduced width, typically 100 m (i.e., boreholes from up to 50 m on either side of the cross-section lines have been ‘pulled’ into the sections) rather than the 200 m width used in 2007. A further difference is the abandonment of the site-scale ‘Lithofacies Package’, or ‘LP’ nomenclature for identifying the separate lithofacies units at the LLWR and the adoption of the ‘A – B – C – D’ nomenclature previously used only for the regional lithofacies (Figure 5-1). This merging of the two parallel naming systems into a single nomenclature is now possible because of the improved understanding, as part of this study, of the relationships between the site- and the regional-scale lithofacies units and also their combined relationship with the BGS regional lithostratigraphical framework. The same RGB colour palette has been used for lithofacies units shown on the regional and the LLWR maps and on the cross-sections to enable direct comparisons to be made. The colour palette for the lithofacies units is attached to this report as Appendix D. Also, wherever possible, the same colour scheme has been used to depict the principal lithology types (as intersected by individual boreholes) shown on the borehole 'sticks' on the various cross-sections. However, it should be noted that, for improved clarity and contrast with the background, the colours used to depict lithologies on the thinner borehole 'sticks' used for the LLWR site-scale cross-sections in this section are brighter, primary colours and therefore have different RGB values to the corresponding lithology colours used for the regional cross-sections in Section 3 60 QRS-1443Y-R1, Version 2.0 (Final) Figure 4-1. Index map of LLWR cross-sections prepared for this report 4.3 Updated contour map of sandstone bedrock surface The contour map of the bedrock (the Triassic sandstone surface) concealed beneath the Quaternary deposits at the LLWR site shown as Figure 15 in Hunter et al. (2007b) has been updated utilising the additional data points acquired from the Phase 3 and Stage 61 6 boreholes. Most of the new sandstone bedrock intersection data did not materially change the previous depiction of the buried sandstone bedrock landform, however the information from three particular boreholes (BH8772, BH8774 and BH8776) have resulted in the discovery of a new topographical feature in the landform. This takes the form of a narrow, incised valley, or channel, orientated in a northwestsoutheast direction, parallel to the prominent sandstone bedrock ridge that underlies the northern half of the LLWR site (and perpendicular to the axis of the Wasdale palaeovalley). This narrow channel-like feature suggests that the concealed bedrock landform beneath the LLWR site may not be a simple, glaciated (i.e., ‘U’-shaped) Wasdale palaeovalley feature, but is possibly a more complicated, dissected topography which has implications for reconstructing the history of glaciation events during the Main Late Devensian period. The revised model of the sandstone bedrock surface is shown in the form of a flat contour map as Figure 4-2 and in the form of a shaded 3D relief image as Figure 4-3. Slices taken through this interpolated surface have been used to depict the position of the sandstone bedrock on the borehole crosssections shown in subsequent sections of this report. 62 QRS-1443Y-R1, Version 2.0 (Final) Figure 4-2. Contour map of the concealed sandstone bedrock surface beneath LLWR 63 Figure 4-3. Shaded 3D relief image of the concealed sandstone bedrock surface beneath LLWR 4.4 Updated LLWR lithofacies model: the east-west cross-section and southwest-northeast section 8 This section and subsequent sections of the report describe the updated understanding of the nature of the Late Quaternary deposits preserved beneath the LLWR site, expressed in terms of lithofacies, that has been derived from a review of the additional available data listed in Section 2.3 above. One means of explaining this updated understanding is to begin with the lithology zones (i.e., the channels and interfluves) identified on Figure 21 in Hunter et al (2007b), which is shown here as Figure 4-4. On Figure 4-4 are also shown the locations of the East-West (‘thick’) Slice 3 - Figure 18 of Hunter et al. (2007b) reproduced below as Figure 4-5, and two new cross-sections which have been prepared for this report: Figure 4-6 and Figure 4-7. All of these borehole cross-sections intersect the larger clayey interfluve zone identified as zone ‘F’ on Figure 4-4. 64 QRS-1443Y-R1, Version 2.0 (Final) The East-West Slice 3 from 2007 (Figure 4-5) shows how the glacigenic sediments were re-interpreted as a set of LLWR site-scale lithofacies packages (LP1-LP7), namely; the post-glacial peat and dune sands (LP1), a zone of interbedded clays (tills), sands and gravels (LP2), a sand- and gravel-filled channel (LP3) (shown schematically), which is apparently incised into a thick clay unit (LP4) and an underlying unit of silty sands (LP5). The lowest part of the sequence is completed by an apparently deeper-water deposit, LP6 and then some older tills and gravels (LP7). The thick clay unit (LP4) is not identified specifically in this section, although its existence is implied by ‘zone F’ on the channels map and it is likely to have been intersected by boreholes C8 and DDS6. Figure 4-4. Map of LLWR showing the locations of selected borehole cross-sections superimposed on lithofacies zones proposed in 2007 65 66 Figure 4-5. The East-West borehole thick-section 3 from Hunter et al. (2007b) QRS-1443Y-R1, Version 2.0 (Final) The updated East-West cross-section presented here as Figure 4-6 can be compared directly with East-West Slice 3 from the 2007 report. It is also coincident with the regional-scale cross-section shown as Figure 3-4 in Section 3.2. On this figure, the different lithofacies units (A, B2, B3 etc) are represented by different colours and are separated by thin grey lines which represent their bounding surfaces. These lithofacies boundary lines have been ‘correlated ‘ between boreholes to connect intersections with apparently similar bulk lithological characteristics, as shown on the figure key and defined by the coding system described in Appendix D. Of necessity, because of the nature of that facies and paucity of data at intermediate depths, the stacked, coalesced channel deposits of the B3 lithofacies are depicted as a single unit in a semi-schematic manner and without any attempt to show internal detail. Similar to the correlated stratigraphical boundaries used in the layered sequence models, these lithofacies boundaries are also subject to some uncertainty. Therefore the particular arrangement and relationship of the separate lithofacies units shown on this cross-section is only one of a number of alternative variations that can potentially be made. Despite this uncertainty, the lines of correlation shown on this and on other figures are recommended as being appropriate to the generality of the lithofacies concept and to the lack of precision in the much of the source data. In addition, the subtle, variable discontinuities between individual layers of thin, interbedded clayey till, sand and gravel of the type described in the Vault 9 excavation by Smith (2009a) and in Trench 3, as shown on Figure 62 in the 2002 PCSC (BNFL, 2002a), cannot be identified with any confidence from the quality of borehole lithology descriptions and the spatial distribution of the boreholes. The method by which these correlations were derived and interpreted should be sufficiently clear to the reader from the information provided in this report. A longer discussion of the issues of confidence and uncertainty is presented in Section 4.7. 67 68 Figure 4-6. New East-West borehole cross-section QRS-1443Y-R1, Version 2.0 (Final) The new information available from several additional boreholes completed in the last two years supports the interpretation shown on E-W Slice 3 from the 2007 report by emphasising the significant discontinuities that exist between the large, thick, sandygravely channel facies on each side of the cross-section and a prominent zone of interbedded clays, sands and gravels located in the centre. The new information from the recent boreholes also provides further evidence for repeated channel-incision of earlier sedimentary deposits, most noticeably at the base of borehole 8649 and in the upper half of borehole 8772. The improved lateral continuity of the grey silty layer, assumed to be marine in origin, is evident, except where it has been eroded by later channel-incision. The new boreholes 8772 and 8774, in combination with ROF borehole 19, suggest that a relatively uniform till overlying the B3 lithofacies unit on the west side may correlate with the Drigg Beach Till. This potential correlation may be improved if the elevation of ROF borehole 18, as recorded in the database, can be confirmed to be erroneous and it is actually 3-4 m lower than shown. Unfortunately at the present time the transcribed elevation of ROF borehole 18 cannot be verified because the original borehole logs are not available to be consulted. The configuration of the highest clayey till(s) has the appearance of having been deposited across a ground surface that suffered variable compaction during the final ice advance. Thus the remnants of the slightly older tills which are preserved in the centre of the cross-section (and in the centre of the LLWR site) may have been compressed into a hollow ‘depression’ between two, more resistive, channel lithofacies zones on either side. Similar preservation of a B2 lithofacies zone of interbedded clayey tills, sands and gravels in this central ‘depression’ can be seen on several of the other cross-sections. Also identified in the new East-West borehole cross-section is a more detailed subdivision of the deeper silty facies into three separate units (D1, D2 & D3). The most distinctive of these units is the D2 layer, which is usually described as a laminated silt in varying shades of grey and can be easily distinguished from the more variable sands, gravels and clays beneath it (D1). Overlying D2 is a unit (possibly a stratigraphical layer) of sands with multiple thin layers of brown silt, distinguishable by the change in colour compared to the underlying grey silt. This can also be recognised in several boreholes in a few of the other cross-sections. It is now identified separately as Unit D3. 69 70 Figure 4-7. SW-NE borehole cross-section 8 QRS-1443Y-R1, Version 2.0 (Final) The new borehole cross-section SW-NE 8, shown above as Figure 4-7, was constructed specifically to display the best combination of borehole logs which intersect the ‘thick clay’ lithofacies unit from Hunter et al. (2007b). Previously identified as ‘LP4’ and now re-named as lithofacies unit ‘C’ for consistency with the regional units, this unit is presumably a zone of preserved till which was not removed by subsequent incisedchannel erosion. It has been intersected by one of the new boreholes (8647) close to the eastern LLWR site boundary, suggesting that it may extend beyond the site on the inland side. However, it is absent from another of the new boreholes (8772) located offsite on the western, coastal side, presumably where it was eroded by an incised channel and replaced by a thick accumulation of sands and gravel (B3). The greater thickness of clayey material recorded in the log for borehole DR4114 suggests that lithofacies unit C may be a compound till, comprised of deposits produced by more than one glaciation ‘event’. This cross-section also shows a similar pattern of interbedded brown silts and sands comprising the proposed D3 lithofacies unit in both the older offsite borehole D and the recent Stage 6 borehole 8772, although the distinctive grey, silty D2 unit is absent in borehole D. Borehole 8772 also intersected a clayey till at a shallow depth, similar to that intersected in offsite borehole D. As suggested above for the East-West crosssection, this till may possibly correlate with the Drigg Beach Till, or it may be a younger (and higher) till, i.e., the Drigg Beach Till could have been removed by glaciofluvial erosion and channel incision. 4.5 Updated LLWR lithofacies model: northwestsoutheast cross-sections 1 & 2 Figure 4-8 is a map of the LLWR showing the locations of three new borehole crosssections which have been constructed for the purpose of enabling comparisons to be made with existing cross-sections of the LLWR site showing the ‘events-based stratigraphy’ and the BGS regional lithostratigraphy interpretations of the glacigenic deposits (i.e., from Eaton, 1996; Nirex, 1997b; and BNFL, 2002a). The new cross-sections are not exactly comparable to the older sections because the latter are constructed from selected boreholes which have been ‘pulled into’ the sections from a wider spatial extent than the borehole sets used for the new cross-sections. For example, NW-SE section 1 does not include offsite borehole G nor the C10 cluster boreholes, which are beyond the approximate 50m-wide selection zone (on both sides of the section line) used for the new cross-sections. The inclusion of selected boreholes from a wider offset distance into the new crosssections is no longer necessary because of the greater choice of more recently-drilled deep boreholes which were not available for use by the previous authors. The inclusion 71 of wider-offset boreholes could also introduce distortions and unnecessary complexity into the geological interpretations. Figure 4-8. Map of LLWR showing the locations of new borehole cross-sections NWSE 1, NW-SE 2 and Sections 1 and 2 from Eaton (1996) The new NW-SE borehole cross-section 1, presented here as Figure 4-9, is equivalent to Section 1 from Eaton (1996), shown as Figure 4-11. The horizontal : vertical scale ratio of 20:1 used by Eaton has been reduced to 10:1 for the new cross-section to make the 72 QRS-1443Y-R1, Version 2.0 (Final) details easier to see. The same scale ratio of 10:1 has been used for the majority of the other new borehole cross-sections presented in this section and in Appendix B. The preserved remnant of the thick-clay Unit C is evident in the centre of the section, apparently truncated by thick accumulations (20-25 m) of incised, channel-fill sands and gravels. There was no equivalent of this particular cross-section in the 2007 report by Hunter et al., but if one had been included, the remnant of Unit C would have been defined by only a single borehole (DDS41) and the Unit B3 on the right (south-east) side would have been defined principally by three boreholes (C9, C11 and DDS134). The new boreholes from the Phase 3 and Stage 6 drilling programmes provide support for this lithofacies interpretation because two more boreholes (8646 and 8647) have also intersected Unit C in the vicinity of DDS41, while two boreholes (8584 and 8664) have been drilled into Unit B3. The thick accumulation of clayey till located around borehole cluster C4, which was shown on the channels-interfluve map in the 2007 report (Figure 21 therein) as being probably part of the same thick clay lithofacies intersected by borehole DDS41, is now considered to be too high in elevation to be directly correlated to Unit C and it is therefore more likely to be a compound till deposit derived from younger glaciation events. This re-assessment of the correlation of the Unit C lithofacies has allowed a revised update of the 2007 channels-interfluve map to be prepared (Figure 4-16). Lithofacies Unit B3 is depicted, schematically, on this and on other borehole cross-sections as almost a single sedimentary unit. In reality, these thick intersections of sand and gravel are very likely to be compound stratigraphical units formed as a result of multiple, separate fluvial incision events that occurred at different times. Individual, linear channels in-filled with sand and gravel that have been incised into older glaciofluvial sediments, similar to the example exposed in the sea cliffs at Nethertown (Figure 4-10), are unlikely to be distinguishable as separate stratigraphical units from the borehole log lithology descriptions. These multiple channel incision events have probably coalesced to create the large masses of slightly silty, slightly clayey, sand and gravel which are preserved today beneath the LLWR site. In the context of this lithofacies, the occasional, thin intersections of finer sand and silt shown on some of the boreholes within Unit B3 are interpreted as evidence of fining upwards cycles related to one or more of the these fluvial depositional events. These thin intersections of finer sediment may or may not exhibit lateral continuity and therefore no attempt has been made to correlate them at this stage of the lithofacies conceptual model. 73 74 Figure 4-9. NW-SE borehole cross-section 1 QRS-1443Y-R1, Version 2.0 (Final) Incised channel deposits Figure 4-10. Photograph of a gravel- and boulder-filled channel incised into coaly, braided, stacked fluvial sand deposits at Nethertown sea cliffs. Height of cliffs approximately 5 m. (J. Hunter, January 2007). Section 1 by Eaton (1996) (Figure 4-11) also shows the thick-clay lithofacies zone, which is identified in this report as Unit C – it is correctly correlated between boreholes DDS41 and cluster borehole C10. Its correlation with the thick clay in offsite borehole G is probably also correct. In this report, borehole C10 is included in NW-SE borehole cross-section 6, while offsite borehole G is included in SW-NE borehole cross-section 4, both of which are presented in Appendix B. These intersections through probable separate remnants of Unit C by boreholes offsite-G and C10 can be understood by reference to the revised map of the lateral distribution of that unit shown on Figure 4-16. However, the combination in Eaton’s Section 1 of this clayey lithofacies (Unit C) together with the laterally adjacent sands and gravels of lithofacies Unit B3 (intersected by other boreholes) into a single stratigraphical unit called the Main Diamict Formation (MDF) can no longer be supported by the presently available data. One reason for this is that no distinctive bounding surfaces (represented in cross-section by lines of correlation) for individual geological formations are suggested by the borehole lithology logs. Furthermore, the accumulation of sands and gravels that comprise Unit 75 B3 were probably deposited by separate, younger geological event(s) and by a different genetic process to the clayey diamict of Unit C. Even if the bounding surface of the MDF was initially defined from the results of a geophysical survey, it is a reasonable expectation that the ‘geophysical surface’ should have a reliably recognisable expression in the borehole lithology log data, particularly in boreholes located away from the geophysical survey lines. The same borehole cross-section shown as Section 1 in Eaton (1996) was adapted by BGS authors as part of the development of the regional lithostratigraphical framework and published originally in a report by Nirex (1997a). It is reproduced in this report as Figure 4-11. The relationship between these regional lithostratigraphical formations and the lithofacies units proposed in this report are discussed in Section 3. The new NW-SE borehole cross-section 2, presented here as Figure 4-12, is equivalent to Section 2 from Eaton (1996), shown as Figure 4-13. As before, the new cross-section 2 includes most, but not all of the boreholes that were used to construct Eaton’s Section 2 because they are located beyond the 50 m-wide selection zone along the section line. The remnant Unit C is only distinguishable in this new cross-section in cluster borehole C8 (see SW-NE cross-section 8 in Figure 4-7 for a view of Unit C perpendicular to this section). It may also be present forming part of a thick compound clayey till unit that has been intersected by boreholes 8426 and 7524, but such a possibility cannot be verified and at this higher-elevation, this zone thick clayey-till is currently being assigned to the Unit B2. A more likely candidate for an extension of Unit C at the north-west end of the section and of the LLWR site is an intersection of stiff, laminated, silty clay between the elevations of –6.25 m and – 8.46 m in offsite borehole F. The laminated, grey-silt D2 Unit is clearly correlatable between the older cluster boreholes C8, C7 and C12, however the removal of the grey silt layer by erosion from a narrow, deeply-incised, gravel-filled channel intersected by the Stage 6 borehole 8649 is quite evident. The remarks made previously with regard to the MDF and the Unit B3 sands and gravels in Eaton’s Section 1 also apply to Eaton’s Section 2. 76 Figure 4-11. Section 1 from Eaton (1996) 77 QRS-1443Y-R1, Version 2.0 (Final) 78 Figure 4-12. NW-SE borehole cross-section 2 Figure 4-13. Section 2 from Eaton (1996) 79 QRS-1443Y-R1, Version 2.0 (Final) 4.6 Updated LLWR lithofacies model: northwestsoutheast cross-section 5 Figure 4-14 shows a schematic cross-section of the LLWR site that was included with the 2002 PCSC (BNFL, 2002a). Similar to the Eaton cross-sections, this has been drawn using a horizontal : vertical scale ratio of 20:1 and it was produced without any indication of the boreholes that had been used for the interpretation. For comparative purposes, Figure 4-15 shows the closest of the new borehole cross-sections to the PCSC section, i.e., NW-SE section 5. This has been constructed using a H:V scale ratio of 10:1 and its location is shown on Figure 4-8. In the 2002 PCSC section, most of the lower half of the glacigenic deposits has been interpreted and described as an accumulation of tills assigned to a single geological formation (the MDF), shown by the dark red-brown colour. The new borehole crosssection shows a more detailed re-interpretation of the borehole lithology data which reveals a different configuration of sedimentary materials and suggests a different glacigenic history for the LLWR site. The most notable feature of these two crosssections is the contrasting interpretation of the mixed sand-and-gravel material, which constitutes the bulk of the glacigenic deposits beneath the LLWR site. The possibility of partial erosion of older glacigenic deposits by later fluvial channel incision events is suggested by the shapes of some of the differently-coloured units shown on the PCSC section and also by the removal of part of the lower laminated silt unit. The preservation of a series of discontinuous, clayey tills in the upper part of the section, shown on the PCSC section as the Pebbly Clay Formation (PCF), is also consistent with the Unit B2 lithofacies shown on Figure 4-15. However the depiction of the MDF as a single stratigraphical formation of mixed types of till which is laterally continuous across the entire site makes no attempt to differentiate between the separate spatial zones of clayey and sandy diamicton that are apparent from the borehole lithology logs. The new borehole cross-sections shown as individual figures in this section of the report are presented together with all of the other cross-sections as a compete set of new borehole cross-sections in Appendix B. The general discussion of the lithofacies units described in this section applies equally to the other sections, which have been provided to help visualise the lateral distribution of lithofacies in a pseudo-3dimensional context across the LLWR rather than to illustrate any particular new features within the glacigenic deposits. The preparation this new set of borehole crosssections has also enabled the map showing the distribution of remnants of the ‘thick clay’, i.e., lithofacies Unit C, originally shown as Figure 21 in Hunter et al (2007b), to be revised and updated. This is presented below as Figure 4-16. The reduced areal extent 80 QRS-1443Y-R1, Version 2.0 (Final) of Unit C shown in this figure is a consequence partly of the changed correlations of zones of clayey till at the northern and southern ends of the site and partly a consequence of improved borehole control in the centre of the site. Figure 4-14. NW-SE cross-section through the LLWR from the 2002 PCSC (BNFL, 2002a) 81 82 Figure 4-15. NW-SE borehole cross-section 5 QRS-1443Y-R1, Version 2.0 (Final) Figure 4-16. Map of LLWR showing the distribution of lithofacies Unit C revised using information from the Phase 3 and Stage 6 boreholes 4.7 Discussion of uncertainties The large amount of data available for the LLWR site that is presented, in condensed format, on the detailed cross-sections shown in this section and also in Appendix B of this report serves to emphasise the natural variability and complexity of these 83 glacigenic sediments. Such complexity is inevitably associated with a degree of uncertainty which is difficult to quantify. Nevertheless, it is proposed that the lithofacies model offers increased confidence in the understanding of the spatial distribution of sedimentary materials beneath and adjacent to the LLWR compared to alternative approaches because it maps the bulk distribution of the principal material types. The principal uncertainties associated with the lithofacies model are the correct identification of the bounding surfaces and the heterogeneity of the material within each lithofacies unit. However, even if modifications are made to the borehole lithology coding system, or to the consolidation of the codes into the reduced set of main lithology groups (clays, silts, sands, gravels etc.), it is unlikely that the bulk spatial distribution of these lithology groups will change significantly, even if the exact boundaries of some lithofacies are not always well-defined. There is confidence in the evidence for the existence of apparently abrupt, large-scale lateral discontinuities in bulk lithology. More examples of these discontinuities were revealed by the Phase 3 and Stage 6 boreholes. There is also confidence in the evidence for deep, downward-narrowing, erosive channel incision into underlying glacigenic deposits by younger glaciofluvial drainage, which is assumed to have been caused by periodic sub-glacial outbursts from ice-dammed lakes. Once again, more examples of these features were discovered by the recent borehole drilling, including the very deep, gravel-filled incision through the unit D silt layers intersected by borehole 8649 (Figure 4-12). It is difficult to imagine a feasible alternative explanation for this particular example. Finally, the initial lithofacies model developed in 2007 was able to 'predict' important intersections of units C & B3 (i.e., the thick stony, clayey till beneath the centre of the site and the thick sequences of sand and gravel) made by some of the Phase 3 and Stage 6 boreholes. Some other examples of uncertainties include: the detailed genetic processes that formed the principal lithofacies (including the exact origin of the tills); the significance of smaller-scale sedimentological changes that occur within bulk lithofacies units (e.g., do the thin silty layers in unit B3 represent the upper layers of fining-upward fluvial sequences?); and the correct correlation of separate, detached clayey lithofacies to other similar lithofacies units. Fortunately, such uncertainties are of only limited significance in the context of describing the bulk spatial distribution of material types. Furthermore, while there is significant scope for alternative arguments regarding the late Quaternary glacigenic history of the site and its surrounding region, and also for the chronostratigraphic correlation of the glacial deposits, it is only the uncertainty relating to the spatial distribution of materials that is of direct relevance to the ESC. A significant advantage, therefore, of the lithofacies model for the LLWR site compared 84 QRS-1443Y-R1, Version 2.0 (Final) to alternative models is that it is essentially independent of any future revisions of the regional lithostratigraphical framework and nomenclature. In addition to the data quality limitations, there are also some spatial limitations associated with the LLWR borehole lithology dataset. For example, compared to the very large amount of shallow borehole data, only limited borehole data are available from the deeper Quaternary deposits beneath the LLWR site and also from the area between the site and the coast. There is also a large spatial data gap between the beach and the nearest ends of the offshore seismic survey lines. As a result, the boundaries between different lithofacies units cannot be defined with confidence in these areas and they will also be more sensitive to changes in lithology coding of the borehole logs (see Appendix C) than in other areas of the site. However the lack of additional deep and offsite borehole data may not necessarily have significant consequences for the ESC and for hydrogeological modelling. 85 5 Integration of the regional scale and site scale lithofacies packages The re-interpretation of the site-scale glacigenic deposits beneath the LLWR and the regional Quaternary lithostratigraphy of the surrounding area in the context of a lithofacies model was first proposed by Hunter et al. (2007b) and Michie et al. (2007). Figure 6 in Hunter et al. (2007b) was an attempt to summarise in a single table the proposed inter-relationships between the lithofacies units identified in the glacigenic deposits beneath the LLWR site and the likely correlatives to these units in the surrounding region. This table, which has been reproduced in several subsequent reports, sometimes in a slightly modified form, is also included here for convenience as Figure 5-1. Figure 5-1. Summary of the possible relationship between site and regional lithofacies units proposed by Hunter et al. (2007b) In these two reports from 2007, a separate system of temporary names was deliberately adopted to identify the different lithofacies units (or ‘packages’ of glacial sediment) that had been interpreted within the LLWR site and to enable them to be distinguished from the regional-scale lithofacies units interpreted in the surrounding area. The sitescale lithofacies units/packages were named ‘LP1’ to ‘LP7’, from the youngest to the 86 QRS-1443Y-R1, Version 2.0 (Final) oldest unit, in contrast to the simple alphabetic system (‘A’, ‘B’, ‘C’ etc) applied to the regional lithofacies units. The purpose of this deliberate separation of nomenclature was to avoid any premature assumptions being made about correlations between one or more site-scale lithofacies unit with any particular regional lithofacies unit without additional consideration of all of the available data. The use of two separate nomenclatures was also intended to avoid, with the exception of the Holmrook Till Member, any premature correlations between the site-scale lithofacies units and the published BGS lithostratigraphical framework. The additional studies completed during the preparation of this report have improved the level of confidence in the interpretation of the Late Quaternary site- and regionalscale glacigenic deposits using the lithofacies concept to a sufficient extent to allow a single, integrated lithofacies framework to be proposed. In this integrated correlation, it is proposed to abandon the site-scale ‘LP’ nomenclature and adopt the regional alphabetic nomenclature of lithofacies at all scales. An updated version of Figure 5-1 incorporating these changes is therefore presented as Figure 5-2. For consistency, the single (alphabetic) lithofacies nomenclature of Unit A, Unit B, Unit C etc., has been adopted for use throughout this report. It has been used to describe the regional lithofacies concepts in Sections 2 and 3, as well as the discussion of the sitescale lithofacies analysis in Section 4. Sections 2 and 3 also provide information showing how this unified nomenclature may be correlated to the individually-named formations and members in the BGS lithostratigraphical framework. It should be noted in this context that for some of the BGS regional lithostratigraphical formations, which in the case of concealed formations may have only be defined by a small number of borehole intersections, more information may be available for these formations from the greater number borehole intercepts within the LLWR site than are available in the surrounding area. Adoption of the single lithofacies nomenclature means that the silt sands assigned to LP5 in 2007 are now assigned to unit B3, while the silt and older tills and gravels previously assigned to LP6 and LP7 respectively are now assigned to unit D. 87 Drigg Point Sand, Ehen Alluvium Fm, Blelham Peat, Hall Carleton Fm Peel Place Sand & Gravel Mem, Mainsgate Wood Sand & Gravel Mem LLWR Lithofacies A Lithofacies B3 Lithofacies B2 Drigg Moorside Silt Fishgarth Wood Till Drigg Holme Sand Drigg Beach Till Kirkland Wood S & G Ravenglass Till Lithofacies B1 Offshore sequences 4, 5 & 6 Lithofacies B3 Barn Scar S & S C Offshore sequence 3A Offshore sequences 2&3 Offshore sequence 1 Lithofacies B4 Holmside Clay Green Croft Till Whinneyhill Coppice Clay Lithofacies C Holmrook Till Scale Beck Till Lithofacies D (D1-D3) Glannoventia Fm Carleton Silt Fm Drigg Till? Maudsyke Till? Figure 5-2. Updated summary of proposed correlation between site and regional lithofacies units (schematic and not to scale) 88 QRS-1443Y-R1, Version 2.0 (Final) 6 Application of the lithofacies units to hydrogeological modelling As highlighted throughout this report, the lithofacies approach was chosen because it focuses on describing the spatial distribution of Quaternary geological materials. The spatial distribution of geological media, and their associated properties, is the key geological issue of interest to the ‘end users’ of the geological interpretation. This report describes the spatial distribution of materials at two scales: the detailed distribution of materials, principally from a very large number of boreholes, at the LLWR site and in its immediately adjacent area; and the more generalised distribution of materials across the wider region. These two scales of information are consistent with the available data, and also with the anticipated requirements for hydrogeological modelling. The regional interpretation provides sufficient detail to underpin a regional scale groundwater flow model that can be used to provide boundary conditions for a more detailed model of the site and the adjacent areas of interest (i.e. those areas through which contaminant transport might occur). The detailed description of the spatial distribution of materials at the site is also suitable to underpin assessment of the impacts of heterogeneity for groundwater flow, and in particular contaminant transport. In some respects the regional approach is comparable to the hydrogeologial domains approach applied in Nirex’s studies (McMillan et al., 2000). However, Nirex were predominantly interested in the impact of the Quaternary materials on deep groundwater recharge and discharge from depth. Therefore Nirex only discretised the Quaternary deposits in plan, as geographical domains, not in section, as particular layers. However, the LLWR is a near-surface disposal system such that it is the shallow groundwater flow system that is of primary interest. The regional lithofacies approach is more qualitative than Nirex’s quantitative approach, which was limited by its reliance on spatially limited regional hydro-test data for the purposes of upscaling (the programme of work to address these limitations was not completed when the Nirex programme ceased in 1997). The 3D regional geological model, which includes some refinement in the area of the LLWR, is being used as the basis for hydrogeological modelling at the regional and site scales. Additional studies are examining the impacts of smaller scale heterogeneities at the site scale. 89 The anticipated hydrogeological properties of each of the lithofacies units are described below as an input to the hydrogeological models. Heterogeneity is further considered in Section 6.1. Lithofacies Unit A Lithofacies unit A comprises a wide range of geological materials including scree and head (rock debris and clayey hillwash) on hill slopes, peat, alluvium, alluvial fan and river terrace deposits, lacustrine deposits, blown sand, estuarine and beach deposits, and offshore sediments, dominantly marine sands and muds. It also includes made ground. The individual deposits that have been grouped within lithofacies unit A are of limited spatial extent and / or thickness such that it is not practicable to distinguish and characterise them all at the regional scale. However, their differing properties may be highly significant for local groundwater flows, and these will have to be recognised within the site scale groundwater flow model, and potentially even in the regional groundwater flow model for key deposits such as river alluvium. In addition, lithofacies unit A includes the soil zone, where the hydrogeological properties are influenced by vegetation and rooting zones, biological activity, weathering and anthropogenic activity such as ploughing and drainage. The National Soil Resources Institute (NSRI) at Cranfield University describes the soils in the area of the LLWR as generally loamy and free draining, although this is most relevant to the agricultural soils, rather than surface peat and till outcrops. Lithofacies Unit B1 This unit includes offshore glaciomarine and marine deposits, which are partly underlain by glacial, lacustrine and terrestrial deposits. The glaciomarine and marine deposits partially infill deeply-incised depressions (palaeovalleys) cut down into the pre-Quaternary bedrock. The offshore deposits are predominantly bedded muds, silts, sands and minor gravels, and therefore can be treated as a single lithofacies assemblage of generally low permeability. Lithofacies Unit B2 This unit includes a sequence of interbedded tills (glacial diamictons), and sands and gravels. All the till (glacial diamicton) units are noticeably clay-rich and the upper till units are almost entirely clay, formed as melt-out or flow tills. The clay-rich tills are anticipated to be of low permeability, while the sands and gravels are likely to be permeable. The impacts of this contrast can be observed in the Drigg beach cliffs, 90 QRS-1443Y-R1, Version 2.0 (Final) where groundwater seeps are observed at the interface between tills and overlying sands. As demonstrated in Section 4, this unit is structurally complex, with clays sands and gravels all having variable lateral continuity. Where the unit is dominated by clays it is likely to be of low permeability, and where it is dominated by sands it is likely to be permeable. In some areas of the site the till and sand / gravel layers are relatively continuous forming a sequence of impermeable and permeable layers. Such areas might exhibit low vertical permeability but relatively high horizontal permeability. This complexity might be upscaled and represented in groundwater flow models by specification of an appropriate horizontal to vertical anisotropy. Lithofacies Unit B3 This unit is dominantly composed of permeable to very permeable thick sand and gravel deposits. The permeability of this unit will vary depending upon the proportion of silt and clay fines in the matrix component and also upon the presence of siltyclayey layers that may represent the upper layers of fining-upward fluvial sequences. Lithofacies Unit B4 This unit consists of glaciolacustrine deposits of interlaminated silts and clays, but also includes in the middle a till unit. It is dominated by impermeable materials. Lithofacies Unit C This unit consists of till as well as sands, gravels, silts and clays. Within the LLWR the till is generally a stony clay diamicton, in contrast to the clay-rich upper till units in unit B2. This unit is anticipated to have a similar permeability to the clay-rich tills in unit B2, but it is laterally discontinuous. It is likely to be much less permeable than the thick sands and gravels in unit B3. Lithofacies Unit D Unit D1 comprises coarse sands and gravels, and diamiction. It is anticipated that D1 will have moderate to high permeability. Unit D2 comprises laminated clays, silts and fine sands. It is anticipated that D2 will have moderate to low permeability. Unit D3 (where differentiated) comprises interbedded silt and sand. It is anticipated that D3 may have similar properties to B2. There are few data with which to characterise these deeper sediments with confidence. However, it is noted that unit D is generally restricted to the very deep parts of the Wasdale Palaeovalley, and it is only present in the most southerly parts of the LLWR 91 site. This unit not anticipated to be of significant interest to the ESC, such that the limited data is not considered to be a significant issue. Bedrock The upper parts of the Sherwood sandstone are weathered and in some places they have degraded to unconsolidated material termed “Snellings Sand”. These Snellings sands and the weathered bedrock are both anticipated to be permeable. The bedrock underlying the site is Ormskirk sandstone, which is a permeable dominantly cross-bedded aeolian sandstone which is likely to exhibit a significant fracture-porosity. The Ormskirk sandstone dips to the west, and approximately 2 km offshore it is overlain by the Triassic Mercia Mudstone Group, which comprises relatively impermeable siltstones, claystones and fine grained sandstones, although the permeability may be significant where there is significant fracturing. 6.1 Heterogeneity One of the limitations of the lithofacies approach is that the complex near-surface geology has been bulked into a two lithofacies units, i.e. A and B2. It is anticipated that the small scale near-surface heterogeneity may be important in the context of local groundwater flow paths and hence contaminant transport. This is because the magnitude and temporal variability of hydrological processes at the site is greatest in the soil zone, e.g. Henderson (2008), Black et al. (2009). This complexity is reflected in the complex spatial and temporal patterns of H-3 contamination in shallow groundwaters (Hartley et al. 2009) and the historical responses of H-3 concentrations in shallow groundwater to engineering activities, e.g. dewatering of the trenches (BNFL, 2002b). In particular these heterogeneities have the potential to: Affect the lateral extent of contaminant transport in shallow groundwaters, and hence the potential for contaminant transport to surface waters. Affect the locations where contaminants migrate from shallow local groundwaters into deeper regional groundwaters, for example focussing contaminant transport paths through spatial discontinuities in the clays. Heterogeneity at depth may also be important, particularly in terms of the potential impacts associated with a groundwater well: well location and sustainable abstraction rate; plume capture and dilution. However, the deeper deposits are apparently more homogeneous (although this may be an artefact of significantly less data) and the scale 92 QRS-1443Y-R1, Version 2.0 (Final) of heterogeneity is apparently larger such that it can be reflected in the definitions of the lithofacies units. Despite this limitation of the lithofacies units, the geological interpretation still provides the information necessary to assess the length scales and impacts of heterogeneity. There are a number of lines of evidence: Direct observations from the Drigg beach cliffs, and site excavations including the trench walls, Vault 9 excavation and trial pits (both onsite and on the beach). Evidence from analogue exposures along the coast, e.g. the large scale incised channel at Nethertown (Figure 4-10). Evidence for continuity of distinct material types from boreholes and sections such as those presented in this report. Evidence for the continuity of distinct material types from continuum geophysics data (Halcrow, 2010). Inferred length scales based on facies information, i.e. the types of depositional environments, the depositional mechanisms, the energy state of the depositional environment, the resultant geological features and their potential length scales. 93 7 Conclusions The present study has updated the understanding of the geology of the LLWR. It explains how a lithofacies approach to classification of the Quaternary sediments was developed and first applied in 2007, after it was identified that previous classifications of the sediments were not optimal to support the ESC, e.g. for hydrogeological modelling. The lithofacies units are directly related to the material properties of the sediments and their relationship to the earlier classifications is illustrated. The lithofacies framework is able to ‘accommodate’ laterally discontinuous units that reflect the laterally discontinuous deposits observed at the LLWR. However, the depositional history of the sediments is not ignored, and where facies information is available it is used to inform correlation of deposits of similar materials and inform the potential length scales of heterogeneity. As a result of the current update, a set of lithofacies units are identified that unify the previous separate set of units for the wider region and the LLWR site area. In addition, the update has extended the detailed correlation of these units to the offshore sequence under the East Irish Sea. The comprehensive understanding of the nature of the sediments across the area shows that they were deposited in a fairly continuous manner during the later part of the Quaternary. The situation of the LLWR site at a position marginal to the ice sheet that occupied the East Irish Sea basin produced spatial relationships of the lithofacies units that were not a sequence of continuous subhorizontal layers. Instead the lithofacies units at the LLWR site are laterally discontinuous, with some younger sandy units cutting down into older units to form thick bodies of sand and gravel. The resultant geological interpretation is considered appropriate to robustly underpin the ESC. Although uncertainties remain, the geological interpretation provides sufficient evidence to bound these uncertainties within the ESC with good confidence. The knowledge gained from the geological interpretation is enhanced within the ESC when it is combined with other sources of evidence regarding the features and length scales of interest, e.g. hydrogeological and hydrogeochemical data. 94 QRS-1443Y-R1, Version 2.0 (Final) 8 References Akhurst, M. C., et al. 1997. Geology of the west Cumbria District. Memoir for the 1:50,000 Geological Sheets 28 Whitehaven, 37 Gosforth and 46 Bootle (England and Wales). British Geological Survey. AGS (Association of Geotechnical and Geoenvironmental Specialists). 1999. Electronic Transfer of geotechnical and geoenvironmental data. Edition 3. Association of geoenvironmental and Geotechnical Specialists. Beckenham, Kent. British Geological Survey. 1999. Gosforth, Solid and Drift geology map. Geological Sheet E037. Black, J., Barker, J., and Towler, G. 2009. LLWR Environmental Safety Case. Review of LLWR Hydro-tests. QRS-1443B-R1 Version 1.0. BNFL. 2002a. Drigg Post-Closure Safety Case. Geological Interpretation. BNFL. 2002b. Drigg Post-Closure Safety Case: Hydrological Interpretation. Boardman, J. 1985. The Troutbeck Palaeosol, Cumbria, England. In: Boardman, J. (ed.). Soils and Quaternary Landscape Evolution. Wiley, Chichester, pp 231-260. Bond, A. E. 2006. LLWR Lifetime project: Review and Update of the 2002 Hydrogeological Conceptualisation. Report QRS-1354B-HCUv1.0. British Standards Institution. 1999. BS5930, Code of practice for site investigation. Cooper, A. H., Kessler, H. & Ford, J. 2006. A revised scheme for coding unlithified deposits. British Geological Survey Internal Report IR/05/123. Eaton, G. P. 1996. The Geology of the Drigg Site: 1996 Update. Geological Consultants Ltd Report No GEO/96/25. Environment Agency. 2009. Review of LLW Repository Ltd’s “Requirement 2” Submission, Vol.4. Eyles, N., Eyles, C. H. and Miall, A. D. 1983. Lithofacies types and vertical profile models: an alternative approach to the description and environmental interpretation of glacial diamict and diamictite sequences. Sedimentology, vol. 30, pp 393-410. Eyles, N. and McCabe, A.M. 1989. The late Devensian (<22,000 BP) Irish Sea Basin: the sedimentary record of a collapsed ice sheet margin. Quaternary Science Reviews, vol.8, pp307-351. 95 Eyles, N. and McCabe, A. M. 1991. Glaciomarine deposits of the Irish Sea Basin: The role of glacio-isostatic disequilibrium. In: Ehlers, J., Gibbard, P. L., and Rose, J. (editors), Glacial deposits in Great Britain and Ireland. Balkema, Rotterdam, pp. 311331. Halcrow Group Limited. 2010. Low Level Waste Repository Ltd: Drigg Coastal Erosion. Geophysical Report. Report 291038-001. Hartley, L., Applegate, D., Couch, M., Jackson, P., James, M., and Roberts, D. 2009. Hydrogeological Modelling for LLWR 2011 ESC Phase 1. SERCO/TAS/E003632/002. Heathcote, J. A. and Michie, U. McL. 2004. Estimating hydrogeological conditions over the last 120ka: an example from the Sellafield area, UK. Journal of the Geological Society, vol.161, pp 995-1008. Henderson, D.E. 2008. LLWR Lifetime Project: Estimation of LLWR Stream Baseflows and Implications for Water Balance and Recharge. NNL (08) 9768. Hunter, J. 2010. The Geology of the LLWR Site – adaptation of 2D cross-sections into a 3D model. Quintessa report QRS-1443ZD-R1 Version 1.0. Hunter, J., Michie, U. Mcl., Smith, N.T., and Towler, G. 2007a. LLWR Lifetime Project: Reinterpretation of the Quaternary geology of the LLWR Site and the surrounding region. Quintessa report QRS-1354F GeologyV1.0. Hunter, J., Smith N., Towler G. & Michie, U., 2007b. LLWR Lifetime Project: Reinterpretation of the LLWR Site Quaternary Geology. Nexia Solutions report (07)8345 – Quintessa report QRS-1354F LLWRGeologyV1.0. Jackson, C.P. 2010. A Preliminary Heterogeneous Model for Lithofacies Package 2. SERCO/TAS/003270/002 Issue 2. Jefferies, N. 2009. LLWR Environmental Safety Case: Alternative Hydrogeological Models for LLWR. Serco report SERCO/TAS/002985/001, Issue 1. Kelly, M. and Emptage, M. 1992. Distribution of radioactivity in the Esk estuary and its relationship to sedimentary proceses. DoE Report No. DoE/HMIP/RR/92/015. McMillan, A. A., Heathcote, J. A., Klinck, B. A., Shepley, M.G., Jackson C.P. and Degnan. P.J. 2000. Hydrogeological characterisation of the onshore Quaternary sediments of Sellafield using the concept of domains. Quarterly Journal of Engineering Geology and Hydrogeology, vol.33, pp301-323. 96 QRS-1443Y-R1, Version 2.0 (Final) McMillan, A. A., & Powell, J. H. 1999. BGS Rock Classification Scheme, Volume 4: Classification of artificial (man-made) ground and natural superficial deposits – applications to geological maps and datasets in the UK. British Geological Survey Research Report RR 99-04. McMillan, A. A., Hamblin, R. J. O. & Merritt, J. W. 2005. An overview of the lithostratigraphical framework for the Quaternary and Neogene deposits of Great Britain (onshore). British Geological Survey Research Report RR/04/04. Merrit, J. W. and Auton, C. A. 2000. An outline of the lithostratigraphy and depositional history of Quaternary deposits in the Sellafield district, west Cumbria. Proceedings of the Yorkshire Geological Society, Vol. 53, Part 2, pp 129-154. Miall, A. D. 1978. Lithofacies types and vertical profile models in braided river deposits: a summary. pp 597- 604 in Miall, A.D. (editor), Fluvial Sedimentology. Canadian Society of Petroleum Geologists, Memoir 5. Michie, U., Smith, N., Towler G. and Hunter, J. 2007. LLWR Lifetime Project: Reinterpretation of the LLWR regional Quaternary geology. Nexia Solutions report (07)8346 – Quintessa report QRS-1354F RegionalGeologyV1.0. Nirex. 1993. The Quaternary geology of the Sellafield area (Eaton, G.P. and Williams, G.D.) Nirex Report No. 519. Nirex. 1997a. The Quaternary of the Sellafield Area. Nirex Report no: S/97/002. Nirex. 1997b. The Quaternary lithostratigraphy of the Sellafield District. Report No. SA/97/045. Nirex. 1997c. Interpretation of the 1993 East Irish Sea High Resolution Seismic Survey (Eaton, G. P.) Nirex Science Report No. SA/97/042. Nirex. 1997d. A summary of the correlation and nature of onshore and offshore Quaternary sediments of the Sellafield area (Wingfield, R.T.R., Merritt, J.W. and Eaton, G.P.) Nirex Science Report SA/97/003. Nirex. 1997e. Interpretation of the 1993 East Irish Sea High Resolution Seismic Survey (Eaton, G.P.) Nirex Science Report SA/97/042. Smith, N. 2008. LLWR Lifetime Project: Phase II Geological Reinterpretation: Further 3D geological modelling of the Quaternary geology of the LLWR site and the surrounding region. Update – April 2008. Nexia Solutions Report (07) 8885, Issue 3. 97 Smith, N. 2009a. Interpretation of geology exposed in Vault 9 excavation and integration with other geological information. NNL Report (08)10176. Issue 0.1. Smith, N. 2009b. Incorporation of latest site investigation and previously unutilised Royal Ordnance Factory (ROF) data into the LLWR 3D geological model. NNL Report (09)100695. Issue 0.1. Smith, N. 2010a. 3D Geological Interpretation of Geophysical Profiles and Further 3D Geological Modelling at LLWR site and surrounding area. NNL Report No. (10) 11217. Smith, N. 2010b. A Re-examination of the Correlation of the Offshore Quaternary Seismic Reflection with the Onshore Lithostratigraphic Sequence in the LLWR area. Report NNL (10) 11013 Issue 0.1 April 2010. Stone, P., et al. 2010. British Regional Geology: Northern England (Fifth edition). British Geological Survey. 98 QRS-1443Y-R1, Version 2.0 (Final) Appendix A – Development of the understanding of the Quaternary sequence for the area of the LLWR The development of the understanding of the nature of the Quaternary sequence is illustrated by presenting the successive interpretations of 3 sections across and adjacent to the site. The locations of the sections are shown on Figure A-1. Section 1 along the NE side of the site provides 4 different classifications of the nature of the Quaternary sediments intersected in boreholes drilled through the sequence. A.1 Williams et al. (1985) The first systematic attempt to classify and correlate the Quaternary sediments of the area of the LLWR, then the Drigg Disposal Site, was by Williams et al. (1985) of the Environmental Protection Unit of the British Geological Survey. Their study of the geology and hydrogeology was based on only 35 boreholes, 7 of which were Old Series (OS) boreholes with only basic (and suspect) lithological logs. The boreholes logged by the BGS were of the initial phase of the DDS (Drigg Disposal Site) series and the BGS logs included the category ‘boulder clay’, a category that was not included in later geotechnical logging of boreholes on the site. In their interpretation of the glacial and glacigenic deposits intersected in the boreholes at the site, the BGS noted that they were unable to correlate lithological units with the two major phase of glaciation identified in the Geological Survey of the wider area (Trotter et al., 1937). The survey had recognised a complex and partially repeated sequence of boulder clays, and sands and gravels and suggested that after the deposits of the ‘Main Glaciation’ there was withdrawal of the ice sheet followed by the ‘Scottish Re-advance’, which contained periods of glacial oscillation. In the 1985 report on the Drigg site it was noted that: “The complicated series of events which produced the deposits at Drigg makes it extremely difficult to correlate lithological units to a particular phase and this has not been attempted.” Instead, the BGS study at LLWR set up an independent stratigraphy for the site based on a succession upwards from the bedrock of G1 to G8 glacigenic units and some post-glacial units. The odd-numbered G units were identified as clays and boulder clays, and the evennumbered G units were identified as sands and gravels. The post-glacial units included peat, alluvium (in stream courses), blown sand and made ground. Although the Quaternary sequence in the boreholes was interpreted in a series of cross-sections as a set of near continuous layers across the site, the study noted that: “In most cases these 99 cross-sections assume continuity of a given clay stratum between boreholes. This may not necessarily be the case especially where glacial wash-out channels cut through older sediments.” and also that: “ …variations especially with respect to the number of horizons and their thickness frequently occur.” Their interpretation of the Section 1 cross-section is shown as Figure A-2, which is redrawn to a uniform scale for this report from Figure 7(a) of their report, and the units identified. The identification of the bedrock as St Bees Sandstone is due to that name being applied at that time for all of the Triassic sandstone bedrock. Only later was the Triassic sandstone in the west Cumbria area identified as the Sherwood Sandstone Group and the name St Bees Sandstone restricted to the lowest formation of the group. Figure A-1. Map of LLWR showing locations of Sections 1, 2 and 3 (as discussed in text). 100 Figure A-2. Williams et al. (1985) Section 1 101 QRS-1443Y-R1, Version 2.0 (Final) A.2 Eaton et al. (1996/7) Subsequent drilling of a much greater number of boreholes across the site, including a large number into the deeper bedrock identified in the southern part of the site, revealed that the simple “layer cake” interpretation was not supported, particularly for the layers of clay/till, which were identified as discontinuous in places. In addition, an older sequence of sediments was identified below the sequence intersected in the shallower boreholes examined by the BGS. This lead to the adoption of a new classification of the Quaternary sediments based on an Event Stratigraphy approach (Eaton et al., 1996, 1997). An Event Stratigraphy approach determines the stratigraphy (time sequence) of sedimentary sequences based on the identification of events, which are broadly defined as comparatively short occurrences, in a relative geological sense, during the sedimentary deposition. Such an approach had been proposed for interpretation of the Quaternary sediments of the wider Irish Sea basin by Eyles and McCabe (1989). In addition, Eaton and Curtis (1995) had produced an interpretation of the Quaternary sediments under the adjacent area of the Irish Sea from the results of the offshore shallow seismic survey and had developed a seismic stratigraphy consisting of several individual sequences. They correlated the offshore sequences with the onshore stratigraphy being developed at that time by the BGS as part of the Nirex investigations of the Quaternary of the wider area. Figure A-3 is their interpretation of Section 1, which shows both their event stratigraphy units as well as a correlation along the right hand side with an early version of the BGS lithostratigraphy. The lowest unit shown in the section is the Lower Diamict Formation, which they interpreted to have been deposited from an initial mid to early Devensian glacial event. This was succeeded by the Marine Silt Formation, which was interpreted to have formed during a lacustrine to marine/intertidal event. Following this the thick Main Diamicton Formation was interpreted to have formed during the main phase of Devensian glaciation. After the diamicton emplacement, a period of erosion and reworking of the diamicton was envisaged, with deposition in glacial outwash channels and channelized fluvial deposits to form the Fluvial Outwash Formation. The Pebbly Clay Formation was considered to lie unconformably above the outwash deposits and represent late Devensian till deposition associated with the Gosforth Oscillation event. Above this the Lacustrine/Fluvial Formation was linked to deposition during a period of sea level rise, followed by glacitectonic deformation of this formation and the Pebbly Clay Formation by the Scottish Readvance of the Irish Sea ice sheet, which produced thin deformation tills. The Holocene/Recent Formation 102 QRS-1443Y-R1, Version 2.0 (Final) consisted of peats developed in kettle holes formed by the melting of blocks of stagnant ice, and the youngest deposits of wind-blown (aeolian) sand. The section showed that units such as the Pebbly Clay formation could be discontinuous and repeated, and in places only sandy units were present above bedrock. However, the section also shows that the lithologies intersected in the boreholes did not have a direct relationship to the interpreted formations. In particular, the Main Diamict Formation was interpreted in some boreholes to consist almost entirely of sand, or sand and gravel lithologies. A major conclusion of the Eaton et al. (1996/7) report was that;“The Main Diamict Formation deposited during the late Devensian was previously described as a boulder clay. However, recent investigations have shown this till to be highly variable but generally sand rather than clay-bound and can therefore not be classed as a boulder clay.” There were also some problems with interpretations of the logs and elevations of the older DDS series boreholes. For example, borehole DDS 33 is common to this section and the previous Williams et al. (1985) section. However, in the Williams et al. section, the top of the borehole is correctly shown at its measured elevation of 19.39 m, but in the Eaton et al. section it is shown several metres higher. In addition, the lower part of the borehole in the Eaton et al. section is shown without any lithological symbol, but is interpreted as equating to Fluvial Outwash Formation on top of Main Diamict Formation. The Williams et al. section correctly shows that the lower part of the borehole is formed of a ridge of sandstone bedrock. (Conversely, the Williams et al. section incorrectly interpreted that bedrock could be present in the base of Old Series borehole OS 6 in the middle of the section at about 12.5 m below sea level. The later drilling shown in the Eaton et al. section shows that bedrock in that area is at about 24 m below sea level.) The imposition of a genetic framework based on an interpreted event stratigraphy for the Quaternary sequence produced a mismatch for some of the units, particularly the Main Diamict, between their unit description and their actual sediment type. This produced difficulties in trying to assign hydrogeological properties to the units for modelling of groundwater flow. For the 2002 PCSC, Cooper (2002) re-interpreted this and other of the Eaton et al. sections to remove some of these problems, but his revision of this section was not able to be found for this review. 103 104 Figure A-3. Eaton et al. (1996/7) Section 1 QRS-1443Y-R1, Version 2.0 (Final) A.3 BGS, 1997 and 2000 In their report to Nirex (Nirex, 1997) on the lithostratigraphy of the west Cumbria area, the BGS included a re-interpretation of the previous section. They did not examine the borehole material or review the detailed logs, but based their re-interpretation on the description of the section and their correlation of the sequence with the sequences observed at outcrop or in Nirex boreholes in the area. The lithostratigraphical units are described in detail in the Nirex report and also in a published article (Merritt and Auton, 2000). Figure A-4 shows their interpretation of the section onto which the corresponding borehole logs have been added. Comparison with the interpretation by Eaton et al. shows that some of the mismatch between units and borehole lithology has been reduced by the identification of separate till and sand units within what had been classified as the Main Diamict. However, a major feature of the BGS interpretation is an attempt to produce continuous clay/till units across the section, but this does not agree with the lithologies intersected in the numerous boreholes on the LLWR site. 105 106 Figure A-4. BGS Section 1 QRS-1443Y-R1, Version 2.0 (Final) A.4 Drigg PCSC (2002) Cooper (2002) developed a number of sections in support of the 2002 PCSC using the Event stratigraphy approach of Eaton. These detailed sections were not presented in the PCSC geological report (BNFL, 2002). Instead two schematic cross-sections were presented, orientated parallel and perpendicular to the long-axis of the site. These schematic sections were used to explain the geological conceptual model. One of the schematic sections is shown as Figure 4-14 in the main text of this report. 107 A.5 Michie et al. (2007) As part of the 2007 review of the geology of the Quaternary sediments of the LLWR area, the sequence in BGS section 1 was re-interpreted using the Regional Lithofacies framework. The resulting interpretation was not included in the report (Michie et al., 2007) because it was regarded as an intermediate stage in developing a detailed lithofacies interpretation of the LLWR site. However, it is useful to include the regional interpretation in this review of the development of the understanding of the Quaternary sequence in the LLWR area. Figure A-5 shows the classification and correlation of the borehole sequences into the regional lithofacies units. The A unit at the surface consists of sands and peats, all of post-glacial age. Under this is the B2 unit of alternating clay/clay diamicton (till) layers and sand/gravel layers. These are partly removed by the B3 unit material of thick sands and gravels, which also cut down into lower units. The C unit is of stony diamicton, which may thicken or form compound layers with the overlying tills adjacent to the bedrock high. The underlying D unit dominantly consists of laminated muds and silts of estuarine to marine aspect below about 20 m below present sea level, but consists of sands and coarser material where preserved at higher level adjacent to the bedrock high. The basal conglomeratic material may be a beach facies formed during a marine transgression. Because of its method of construction based on actual borehole lithologies, this interpretation of the section matches the units to the dominant lithologies. This framework was applied in this report to the development of the detailed lithofacies interpretation of sections across the LLWR using all the available borehole information. 108 Figure A-5. Michie et al. (2007) Section 1 109 QRS-1443Y-R1, Version 2.0 (Final) A.6 Additional section comparisons In order to illustrate more fully how the regional lithofacies interpretation was applied to the development of understanding of the Quaternary sequence, some additional cross sections are provided. Section 2 is approximately along the SW side of the LLWR site. Figure A-6 shows the Event Stratigraphy interpretation of Eaton et al. (1996/7) and Figure A-7 shows the corresponding regional lithofacies interpretation. There was no corresponding BGS interpretation of the section. The major difference in the regional lithofacies interpretation of identifying a thick B3 development of sandy material instead of a thick Main Diamict of variable characteristics is apparent. Section 3 is a long section through a set of off-site boreholes, west and SW of the site and parallel to the coast. Figure A-8 shows the Event Stratigraphy interpretation by Eaton et al. (1996/7) and Figure A-9 shows the BGS interpretation. The BGS interpretation is similar to the Eaton et al. interpretation except that the BGS show more continuous clay/till units in the upper part of the section. The regional lithofacies interpretation of the section is shown on Figure A-10. There are more similarities of this interpretation to the other interpretations in identifying the development of thick sandy material, but only the regional lithofacies interpretation marks the major development of sand and gravel filled incisions that remove most of the C unit (lower diamicton) and part of the D unit (marine facies). This 2007 interpretation was confirmed by subsequent deep drilling in the area and the results are included in the detailed lithofacies interpretation. 110 Figure A-6. Eaton et al. (1996/7) Section 2 111 QRS-1443Y-R1, Version 2.0 (Final) 112 Figure A-7. Lithofacies units Section 2 Figure A-8. Eaton et al. (1996/7) Section 3 113 QRS-1443Y-R1, Version 2.0 (Final) 114 Figure A-9. BGS Section 3 Figure A-10. Lithofacies units Section 3 115 QRS-1443Y-R1, Version 2.0 (Final) A.7 References BNFL, 2002. Drigg Post-Closure Safety Case. Geological Interpretation. Cooper, S. 2002. Drigg Technical Programme. Geological cross-section lines. DTP 02/03 TN142 (DTP/185 for issue to The Environment Agency). Eaton, G. P., Cooper, S. and Speakman, P. 1996 (revised1997). Drigg Site Characterisation : Geological Interpretation (Interim Report), 258 pages. Geological Consultants Report No. GEO/96/25. Prepared for BNFL. Eaton, G. P. and Curtis, N. J. 1995. 2D transects through the Quaternary of West Cumbria. Geological consultants limited report No. GEO/95/10. Prepared for Nirex. Eyles, N. and McCabe, A. M. 1989. The Late Devensian (<22,000 BP) Irish Sea Basin: the sedimentary record of a collapsed ice sheet margin. Quaternary Science Reviews, vol.8, p 307-351. Merritt, J. W. and Auton, C. A. 2000. An outline of the lithostratigraphy and depositional history of Quaternary deposits in the Sellafield district, west Cumbria. Proceedings of the Yorkshire Geological Society, vol. 53, pp129-154. Michie, U., Smith, N., Towler G. & Hunter, J., 2007. LLWR Lifetime Project: Reinterpretation of the LLWR regional Quaternary geology. Nexia Solutions report (07)8346 – Quintessa report QRS-1354F RegionalGeologyV1.0. Nirex, 1997. The Quaternary lithostratigraphy of the Sellafield district. Nirex Report no. SA/97/045. Trotter, F. M., Hollingworth, S. E., Eastwood. T. and Rose, W. C. C., 1937. Gosforth District. Memoir of the Geological Survey of Great Britain, Sheet 37 (England and Wales), 136 pp. Williams, G. M., Stuart, A. and Holmes, D. C., 1985. Investigation of the geology of the low-level radioactive waste burial site at Drigg, Cumbria. British Geological Survey Report, Vol. 17, No. 3, 24 pp. 116 QRS-1443Y-R1, Version 2.0 (Final) Appendix B – LLWR site borehole crosssections This Appendix contains a complete set of the new borehole cross-sections constructed during the preparation of this report, together with an index map showing their locations across the LLWR site. For some of the borehole cross-sections, it has been possible to overlay partial sections of some of the Halcrow (2010) geophysical survey (resistivity) profiles where they are either coincident with or in close proximity to the borehole sections. For use in these borehole cross-sections, the resistivity profiles have been re-scaled and, in some cases, the images have been reversed. It should be noted that the analysis and interpretation of the spatial patterns of resistivity values has not been completed at the time of writing of this report and therefore any apparent lack of immediate and obvious correlation between the resistivity patterns and the distribution of lithofacies units shown on the borehole cross-sections should not be construed as an indication than either one or the other of these methods of investigation is invalid. The geophysical survey work was undertaken in support of the coastal erosion modelling and was optimised to investigate the post-glacial Holocene deposits along the spit, rather than the detail of the deeper, Late Devensian glacigenic deposits. 117 Figure B-1. Index map showing locations of borehole cross-sections 118 Figure B-2. NW-SE borehole cross-section 1 119 QRS-1443Y-R1, Version 2.0 (Final) 120 Figure B-3. NW-SE borehole cross-section 2 Figure B-4a. NW-SE borehole cross-section 3 121 QRS-1443Y-R1, Version 2.0 (Final) 122 Figure B-4b. NW-SE borehole cross-section 3 with resistivity data sumperimposed Figure B-5a. NW-SE borehole cross-section 4 123 QRS-1443Y-R1, Version 2.0 (Final) 124 Figure B-5b. NW-SE borehole cross-section 4 with resistivity data sumperimposed Figure B-6. NW-SE borehole cross-section 5 125 QRS-1443Y-R1, Version 2.0 (Final) 126 Figure B-7. NW-SE borehole cross-section 6 Figure B-8a. SW-NE borehole cross-section 1 127 QRS-1443Y-R1, Version 2.0 (Final) 128 Figure B-8b. SW-NE borehole cross-section 1 with resistivity data superimposed Figure B-9. SW-NE borehole cross-section 2 129 QRS-1443Y-R1, Version 2.0 (Final) 130 Figure B-10. SW-NE borehole cross-section 3 Figure B-11. SW-NE borehole cross-section 4 131 QRS-1443Y-R1, Version 2.0 (Final) 132 Figure B-12. SW-NE borehole cross-section 5 Figure B-13. SW-NE borehole cross-section 6 133 QRS-1443Y-R1, Version 2.0 (Final) 134 Figure B-14. SW-NE borehole cross-section 7 Figure B-15. SW-NE borehole cross-section 8 135 QRS-1443Y-R1, Version 2.0 (Final) 136 Figure B-16. E-W borehole cross-section QRS-1443Y-R1, Version 2.0 (Final) Appendix C – Summary of lithology code consolidation C.1 The Drigg Site Characterisation Database Similar to the study produced by Hunter et al. in 2007, this updated interpretation of the Late Quaternary glacigenic deposits preserved beneath the LLWR site is based upon an interrogation of the accumulated database of lithological descriptions of Quaternary material recovered from boreholes (i.e., arisings). These descriptions are derived from numerous programmes of site investigation that have been conducted at the LLWR at various times since planning consent for the facility was originally granted in 1957. The history of these different phases of site investigation has been summarised in Section 2.3 of this report. These various programmes of historical site investigation at the LLWR were undertaken by a number of different operators for a variety of engineering projects. They will have used slightly different methodologies and different quality assurance and reporting standards. As a result, although a large quantity of descriptive geological data is available for the LLWR site, the quality of these data is variable. Some lithology descriptions are lengthy and detailed, while others are very brief. Some descriptive lithology records relate to very thin borehole intersections (<0.5 m), while others may represent thicknesses of several metres. The written lithological descriptions of borehole arisings from all of the pre-2007 site investigation drilling work described in Section 2.3 have been transcribed by previous investigators into a digital database (called the Drigg Site Characterisation Project database) and quality control checked. The same investigators have also assigned a one- or two-character alphabetic code to each lithology record to facilitate the computer-plotting of appropriate standard rock patterns on borehole log crosssections. These lithology codes are the means by which the large database of descriptive text can be condensed into manageable data suitable for plotting on crosssections and then interpreted into geological conceptual models. However, the reliability of the lithology codes for this purpose is no greater than the variable quality of the original lithological descriptions in the database records. The reliability also depends upon the judgement used in assigning the most appropriate codes to the descriptions. Furthermore, in common with most large databases of this type, the DSCP database will inevitably contain an unknown, but small percentage of transcription errors which can result in occasional incorrect lithology codes. 137 The DSCP database also contains all of the available coordinates and elevations of the surface locations of the boreholes. The small number of borehole records that lack spatial coordinates have not been used in this study, however, where possible, for the few boreholes that lacked only ground elevation values, these missing data were estimated from the digital ground surface elevation model. As part of the preparation of this report, the digitised lithological descriptions of borehole arisings derived from recent site investigation work – data that were not available for use in the report by Hunter et al. (2007) – have been appended to a copy of the DSCP database by one of the present authors. These additional records have been assigned the same type of lithology codes so that they are consistent with the existing data. C.2 Interpretation of the lithology data The sample lithology descriptions stored in the Drigg Site Characterisation Project database constitute the basic ‘raw’ data that have been used (selectively, and sometimes in combination with geophysical surveys) for all of the previous studies of the Quaternary deposits beneath the LLWR site. Some of these data (i.e., selected individual boreholes) have also been used by the BGS to assist with constructing the regional lithostratigraphical framework, while the entire database has been utilised by Hunter et al. (2007) and by the authors of this report. However, despite making use of the same raw data, the various studies of the LLWR Quaternary geology that now exist, including this report, differ in their interpretations of the data and accordingly can be classified into three separate models, as described in Section 4 of this report. A basic requirement for constructing the borehole cross-sections presented by Hunter et al. (2007) and also for the enhanced cross-sections included in this report, is a simplification of the larger set of lithology codes into a condensed set of groups which are most likely to represent the dominant sedimentary material types. The condensed grouping is shown in the following table, each group being distinguished by a colour which matches the lithofacies colours plotted on the borehole ‘sticks’ used in the crosssections. The principle groups are: clays, silts, sands, mixed sands and gravels, cobbles, peat and sandstone. 138 QRS-1443Y-R1, Version 2.0 (Final) Code Lithology A SAND AS Silty SAND AC Clayey SAND A1 Silty SAND with gravel A2 Clayey SAND with gravel AG SAND with gravel GA Gravel with sand Q1 Clayey sand and gravel Q3 SAND and GRAVEL G GRAVEL GC Clayey GRAVEL CO COBBLES BO BOULDERS C Clay C3 Sandy CLAY with gravel CA Sandy CLAY CG Gravelly CLAY CS Silty CLAY P PEAT PA Sandy PEAT PT Clayey PEAT S SILT SA Sandy SILT SC Clayey SILT SS SANDSTONE Condensed lithology Sands Sand and gravel Cobbles and boulders Clays Peat Silts Sandstone An example of the interpretation of the lithology descriptions and codes by the three types of interpretive geological models for the LLWR site and the surrounding region is shown in the following table using data from the log for borehole DDS134. This borehole log has been used in cross-sections for all three models and it appears in several borehole cross-sections in this report, including Figures 4-10, A-3, A-4 and A-5. The borehole was drilled in 1989 and the log was transcribed and coded in the DSCP database prior to 2007. 139 Descriptive lithology log for borehole DDS134 Top elev (mOD) 11.4 Bottom elev (mOD) 8.5 8.5 7.2 7.2 5.5 5.5 4.7 4.7 -1.3 -1.3 -5.3 -5.3 -9.3 -9.3 -10.3 -10.3 -13.3 -13.3 -14.3 140 Lithological description Red brown slightly silty fine to medium SAND with a trace of fine to medium subangular assorted gravel and occasional pockets of very silty CLAY and occasional cobbles. Red brown fine to coarse assorted subangular to subrounded sandy GRAVEL with some subrounded cobbles and occasional boulders. Red brown slightly sandy silty CLAY, firm, with some subrounded fine to coarse assorted gravel and occasional cobbles. Grey fine to coarse angular to subrounded GRAVEL and COBBLES with occasional pockets of very silty clay and occasional boulders. Grey fine to coarse subangular to subrounded sandy GRAVEL and COBBLES with occasional boulders. Predominantly grey fine to coarse subangular to angular COBBLES. Assorted clasts of granite, granodioritic and basaltic rocks. Predominantly medium-dark grey green medium-coarse subangular to subround subspherical sandy GRAVEL. Clasts of granitic, granodioritic and basaltic rock. Predominantly grey green fine to occasionally medium subangular to subrounded silty sandy GRAVEL. Predominantly grey green fine to occasionally medium subangular to subelongate slightly sandy GRAVEL. Predominantly grey green medium to occasionally coarse subangular to occasionally subrounded slightly sandy GRAVEL with occasional COBBLES. Lith code Events-based stratigraphy BGS Regional Lithostratigraphy Condensed Lithology Litho-facies A Lacustrine Fluvial Formation FWT, DHS, DBT & KWS members, (undifferentiated) (Gosforth Glacigenic Formation) Sand & gravel GA CS Pebbly Clay Formation G Fluvial Outwash Formation Clay Ravenglass Till Member (Seascale Glacigenic Formation) GA CO Main Diamict Formation Barn Scar Sand and Silt Member (Seascale Glacigenic Formation) GA GA G GA B2 (interbedded clayey tills, sands and gravels) Holmrook Till Member (Blengdale Glacigenic Formation) Sand and gravel (with a layer of cobbles) B3 (undifferentiated, coalesced, multichannel sand and gravels) QRS-1443Y-R1, Version 2.0 (Final) -14.3 -17.3 -17.3 -19.3 -19.3 -23.05 -23.05 -25.8 -25.8 -33.3 -33.3 -36.8 -38.32 -40.3 -45.28 -45.9 -45.9 -49.05 Red brown medium to occasionally fine SAND with some dark grey green fine angular to subangular GRAVEL, with a trace of red brown silty sandy CLAY at the base. Medium to dark red brown medium to occasionally coarse SAND, with increasing red brown silty sandy CLAY towards the base, firm to stiff. Predominantly grey green medium to coarse subangular to subrounded GRAVEL with boulder fragments. Red brown silty sandy CLAY, firm, with occasional gravel lenses. Medium red brown to occasionally orange red fine to medium SAND with occasional gravel lenses. Orange red fine to medium SAND with increasing dark green fine to medium GRAVEL. NO CORE RECOVERY. Cuttings description: Orange red fine to medium grained completely disintegrated SANDSTONE, with some dark green fine grained gravel. Medium red brown medium grained moderately to weakly cemented SANDSTONE soft to weak, locally crumbly and friable, with closely spaced cross bedding. NO CORE RECOVERY. Cuttings description: Medium red brown fine to medium grained completely disintegrated slightly micaceous SANDSTONE, with some dark mineral plates of Mn02. Note: the log continues to a greater depth within the sandstone bedrock AG Glannoventia Formation (includes an unnamed upper sequence of silts, fine sands & gravels which are possibly glacitectonised deposits of the Glannoventia Fm.) A G CA Marine Silt Formation Clay Lower Diamict Formation Sand A AG Drigg Till Formation D1 (interbedded, brown, laminated silts) D2 & D3 (sands with some gravel – laterally equivalent to grey laminated silts) SS SS Ormskirk Sandstone Ormskirk Sandstone Sandstone Sandstone SS This tabulated borehole log conveniently illustrates the ‘averaging’ of lithology information that occurs during the conversion of moderately detailed engineering 141 descriptions into the one- and two-character codes. It also illustrates the differences between the subsequent interpretation of both the descriptions and the codes by each of the three geological models that have been developed to characterise the LLWR Quaternary geology. The different stratigraphical interpretation of most of the borehole lithology intervals by the events-based stratigraphy and the BGS regional lithostratigraphy models is immediately evident. Also evident is that the identification of certain borehole intervals as tills by both of these models involves assumptions and inferences about genetic processes which are not unequivocal, based as they are upon the descriptions shown in the third column from the left. The last two columns on the right hand side of the table show the simplified bulk lithologies produced by condensing the original lithology codes and the lithofacies units that accommodate them. Although the condensation of detailed descriptions into generalised, bulk lithologies and lithofacies may appear to be an oversimplification, the derived lithofacies units form coherent spatial zones in three dimensions beneath the LLWR. Also, because the majority of the borehole logs in the DSCP database have been utilised in the mapping of the lithofacies, the derivation of the lithofacies units is evident from the data shown on the site cross-sections, as is the associated uncertainty when some lithologies in individual boreholes do not appear to ‘fit’. C.3 References Hunter, J., Smith N., Towler G. & Michie, U., 2007. LLWR Lifetime Project: Reinterpretation of the LLWR Site Quaternary Geology. Nexia Solutions report (07)8345 – Quintessa report QRS-1354F LLWRGeologyV1.0. 142 QRS-1443Y-R1, Version 2.0 (Final) Appendix D - Lithofacies colour palette 143 Appendix E - Adaptation of 2D cross-sections into a 3D model This appendix contains a brief summary of the contents of the report prepared by Hunter (2010), which describes the procedure adopted to convert the 2D lithofacies cross-sections of the LLWR site presented in this report into a format suitable for direct importation into the 3D computer model of the Quaternary geology of LLWR and its wider surrounding region. The Hunter (2010) report was prepared after Version 1 of this report was completed. Some of the figures from Hunter (2010) are included in this appendix (Appendix E) in Version 2 of this report, because they provide an additional means of visualising the Quaternary lithofacies and therefore they complement the cross-sections shown in the Section 4 and Appendix B. Hunter (2010) gives a detailed description of the process of digitising the lithofacies boundaries shown on the 2D vertical cross-sections across the LLWR site and the recombination of these data into a set of x-y-z coordinates representing lithofacies boundary surfaces in a more horizontal orientation. By this means the 3D spatial configuration of the different Quaternary lithofacies present beneath the LLWR site is represented as a series of irregularly-shaped surfaces which are stacked sequentially above each other. In the 3D model these spatial coordinates are combined with additional data derived from sources external to the LLWR site to produce a single, comprehensive regional model. Paradoxically, these multiple, spatially-continuous surfaces, which resemble a lithostratigraphical succession, are being used to represent a set of lithofacies boundaries that are often spatially-discontinuous and are not part of an established lithostratigraphy. This procedure is necessary because the 3D computer model is a layered model and it requires the data in each layer to be continuous across the full extent of the model, i.e., it does not accommodate lateral discontinuities within individual layers. Furthermore, in order for the spatial complexity and interrelationship of certain lithofacies units to be reproduced as accurately as possible in 3D, it was necessary to subdivide these particular units into multiple sub-units to enable abrupt lateral discontinuities to be characterised in a digital form. Internal surfaces that subdivide individual lithofacies units are entirely artificial in concept and have no geological significance. When the series of stacked surfaces is combined together in the model, the internal subdivisions within a lithofacies unit are no longer apparent because the properties of each of the subdivisions are the same. 144 QRS-1443Y-R1, Version 2.0 (Final) An additional level of spatial control was imposed upon the lithofacies boundary surfaces by digitising data from imaginary, or ‘virtual’, boreholes. These additional xy-z coordinates were included to try to ensure that the interpolation algorithm used by the 3D model produced reasonable and realistic surfaces in areas of the model where actual borehole control is lacking, i.e., data-gaps. In some parts of the model there are lateral spatial data gaps, requiring complete virtual boreholes to be inserted through the entire thickness of the Quaternary deposits. Elsewhere, there is adequate geological control from existing boreholes at a shallow depth level, but some of these boreholes were required to be projected to the sandstone bedrock by means of ‘virtual depth extensions’. A limited selection of the figures presented in Hunter (2010) is shown below. Each of these figures illustrates a digitised lithofacies boundary surface projected as an apparently-solid, colour-rendered 3D view for easier visualisation. These surfaces are usually composite in nature, i.e., different parts of the surface may separate different lithofacies in different areas of the model and therefore they are not necessarily of the same age across their spatial extents. They generally do not represent geological events marking particular time intervals in the established Quaternary stratigraphy. The effect of combining all of the digitised lithofacies boundary surfaces into a single stacked sequence to represent the Quaternary lithofacies beneath the LLWR site only becomes apparent when they are imported into the 3D computer model itself (Smith, 2010). 145 Figure E-1. Contour map representing the base of lithofacies unit A 146 QRS-1443Y-R1, Version 2.0 (Final) Figure E-2. 3D surface representing the base of lithofacies unit A Figure E-1 is a 2D contour map and Figure E-2 is a 3D surface representing the digitised base of lithofacies unit A, which is comprised of all post-glacial deposits together with the made ground inside the LLWR site. Therefore these two figures should depict the likely topography of the final glacial land surface if the post-glacial sediments are removed. Both of these figures reveal the shallow, stream valley feature that intersects the southeastern half of the LLWR site, known as the Drigg Stream. The green line added to Figure E-1 represents the spatial limit of the post-glacial peat layer which is intersected by many of the boreholes within the LLWR site. The peat layer post-dates and is draped across this shallow valley landform, including the higher-elevation ground to the north-west, but it is absent from the slightly elevated area in the south-east corner. This surface can be considered, in part, as a geological event boundary. 147 Figure E-3. Contour map of base of lithofacies unit A beneath regional land surface Figure E-3 shows the Figure E-1 contour map of the base of lithofacies unit A visible through a ‘window’ in a contour map of ground-surface LIDAR data for the surrounding area, the latter shown as a greyscale to enhance the contrast. On this figure the apparent continuity of the Drigg Stream valley (shown as a dashed-blue line) from outside to within the LLWR site is evident, with the 15m contour line appearing 148 QRS-1443Y-R1, Version 2.0 (Final) to define the stream bank. Some of the ‘dimples’ and irregularities in this lithofacies surface, particularly at the north-western end, are the result of various engineering excavations that have been made within the LLWR site. These excavations are an integral part of the LIDAR topographical surface, i.e., the upper surface of the model, and where they are of sufficient depth, they extend through lithofacies unit A into unit B2 beneath. Figure E-4. 3D surface representing an intermediate surface within the B3 lithofacies. Figure E-4 is a 3D image of the surface created to subdivide the B3 lithofacies unit within the LLW site into upper and lower sub-units. This procedure was necessary to enable the discontinuous remnants of the C lithofacies unit to be digitised in their correct, interpreted spatial relationship to B3. Although the surface is described in Hunter (2010) as an artificial concept, it nevertheless enables the remnants of the C lithofacies to be visualised as a single entity in a manner which is not so easy to achieve using a combination of cross-sections and plan-view figures. Also visible on this image is the onlap of the B3 lithofacies unit onto the sandstone bedrock hill, which protrudes through it. 149 Figure E-5. 3D surface representing the base of the B3 lithofacies. Figure E-5 shows a surface representing the base of the B3 lithofacies unit, which is a composite unit interpreted to be the consequence of multiple, coalesced glaciofluvial events, some of which formed incised, infilled channels eroded into the D lithofacies units beneath. The shapes of two of these incised channel features are revealed on this image. Only a small proportion of the ~650 boreholes drilled within and around the LLWR site penetrate below the B3 lithofacies. These deep boreholes provide valuable clues about the nature of the deepest glacial deposits in this area, however much of the geological interpretation between the boreholes is speculative, including the shapes and orientation of these incised channels. A complete set of similar contoured maps and rendered 3D images for all of the digitised surfaces are presented in Hunter (2010). E.1 References Hunter, J. 2010. The Geology of the LLWR Site – adaptation of 2D cross-sections into a 3D model. Quintessa report QRS-1443ZD-R1 Version 1.0. Smith, N. 2010. 3D Geological Interpretation of Geophysical Profiles and Further 3D Geological Modelling at LLWR site and surrounding area. NNL Report No. (10) 11217. 150