Optimisation and Development Plan

Transcription

1 The 2011 Environmental Safety Case Optimisation and Development Plan LLWR/ESC/R(11)10025 May 2011

2 This page is left blank intentionally.

3

4 This page is left blank intentionally

5 Preface The Low Level Waste Repository (LLWR) is the United Kingdom s principal facility for the disposal of solid low-level radioactive waste. The LLWR is owned by the Nuclear Decommissioning Authority (NDA) and operated on behalf of the NDA by a Site Licence Company (SLC) LLW Repository Ltd. We, LLW Repository Ltd, are committed to operating the LLWR as a safe and efficient facility that provides a continuing option for the disposal of low-level radioactive waste in the UK. This will be achieved consistent with good practice for the near-surface disposal of radioactive waste, in accordance with environmental and health and safety regulation and guidance, and in compliance with the terms of our Nuclear Site Licence and Permit to dispose of radioactive waste. This report is one of a series of reports that present the evidence underpinning the 2011 Environmental Safety Case for the LLWR the 2011 ESC. The report has been prepared by the Environmental Safety Case Project and is issued under the authority of the Managing Director of LLW Repository Ltd. ESC objectives Under the terms of our Permit granted by the Environment Agency, we are required to submit an Environmental Safety Case (ESC) for the LLWR no later than 1st May 2011 and at intervals thereafter as requested by the Agency. The ESC: presents the arguments and evidence concerning the environmental safety of disposals of solid radioactive waste at the LLWR, at present and in the future, consistent with the Agency s Guidance on Requirements for Authorisation; provides a basis for the environmentally safe management of the site by the SLC, and regulation of the site by the Agency, including setting of conditions on its future management and acceptance of waste. The ESC is addressed primarily to the Agency and is intended to inform and enable their regulation of the LLWR. It also provides a plan for the future management of the LLWR and a baseline against which proposed changes in the plan for the development of the facility can be tested. As such, it will be of interest to our other stakeholders, both local and national. ESC document plan The ESC consists of documents at two levels: A single Level 1 report outlines the plan for the development of the LLWR and the main arguments concerning environmental safety and how this is achieved. A series of Level 2 reports present the evidence that underpins our safety arguments, including descriptions of our management framework, system understanding, design and management choices, and assessments. This is a Level 2 report. The ESC Level 1 and 2 reports are listed in the table at the end of this Preface, which also shows for the Level 2 reports the set of arguments for which each report mainly provides evidence. The ESC is supported by a large LLWR/ESC/(R11)/10025 Page 1 of 119

6 number of technical and scientific reports and references that we refer to as Level 3 documents. The ESC documentation concept Scope and audiences The 2011 ESC is based on an optimised Site Development Plan developed under our Environmental Safety Strategy. The Plan sets out our proposals and assumptions on operations, remedial activities, vault design, capacity and future waste disposal practice, closure design and management up to the end of management and regulatory control. It provides a basis for our quantitative assessments. The Plan is flexible, however, and will be amended as necessary in the light of UK radioactive waste management needs, operating experience, results of monitoring, future iterations of the ESC, regulatory and planning guidance and decisions, and stakeholder views. The safety arguments set out in the Level 1 report comprise arguments concerning the development and safety of the Site Development Plan. The Level 1 report focuses on the arguments in principle, referring to the more detailed and quantitative evidence that is presented in the Level 2 reports. The main features and findings of the supporting reports are presented, demonstrating that the Site Development Plan is optimised, and that the assessed safety is consistent with the regulatory guidance over the lifetime of the facility, including after closure. The Level 1 report is intended to be complete enough to inform managers from the Environment Agency, Government ministries and local government representatives and officials on the environmental safety of disposal of radioactive waste at the facility. It is also intended to be an entry point to the safety case for the Agency s technical staff and assessors. The Level 2 reports present the evidence that underpins our safety arguments, including descriptions of our management framework, system understanding, optimisation, assessments and proposed conditions for acceptance of waste. The Level 2 reports are primarily addressed to the Agency s Nuclear Regulator for the site and technical staff, and may be of interest to experts in specific technical fields. To fully satisfy themselves, however, for example, to find supporting information and details of the model formulations and data used, technical specialists and reviewers in specific topic areas may need to refer to Level 3 documents. LLWR/ESC/(R11)/10025 Page 2 of 119

7 We have also produced a Non-technical Summary of the ESC, to help a wider group of stakeholders understand its nature, conclusions and implications. Level 1 The 2011 Environmental Safety Case Main Report [1] Level 2 Management and dialogue System characterisation and understanding Optimisation and Site Development Plan Management and Dialogue [2] Site History and Description [3] Inventory [4] Engineering Design [5] Near Field [6] Hydrogeology [7] Site Evolution [8] Monitoring [9] [10] Assessments Environmental Safety During the Period of Authorisation [11] Assessment of Long-term Radiological Impacts [12] Assessment of Non-radiological Impacts [13] Assessment of Impacts on Non-human Biota [14] Waste Acceptance [15] Assessment of an Extended Disposal Area [16] Audit Addressing the GRA [17] LLWR/ESC/(R11)/10025 Page 3 of 119

8 This page is left blank intentionally. LLWR/ESC/(R11)/10025 Page 4 of 119

9 Executive Summary Decisions taken now and in the future regarding the development and operation of the LLWR are framed by past actions and UK strategy for the management and disposal of LLW. Making the best use of existing LLW management assets is central to the current UK Strategy. The UK Strategy assumes the continued availability of the LLWR for the disposal of the majority of UK solid LLW requiring the protection of vault disposal (subject to an acceptable environmental safety case being made). Development of the Environmental Safety Case (ESC) for the LLWR is an iterative process, ongoing through the life of the facility. It involves progressive development with focused improvement of data, understanding, design options and assessments. Decisions taken regarding the development and operation of the LLWR to meet environmental safety objectives are captured in the Site Development Plan (SDP). Key considerations that need to be addressed in establishing the LLWR SDP include the legacy of past operations, the nature of the wastes likely to be consigned in future, and the expected evolution of the site. In particular, the potential hazards associated with long-lived radionuclides (in both past and future disposals) and the implications of the expected eventual disruption of the facility by coastal erosion are central to decisions taken by LLW Repository Ltd in making the best use of available capacity at the facility. Optimisation is the process by which our preferred approach to design, construction, operation, closure and post-closure arrangements for the LLWR is identified and justified. It draws on over-riding environmental safety management principles for the control of near-surface waste disposal and examines the alternative ways in which they can be addressed, consistent with the environmental setting, history and anticipated future role of the facility. Options are analysed and compared, based on the principle of ensuring that radiological risks to members of the public are as low as reasonably achievable (ALARA), both during the period of authorisation (PoA) and afterwards. The aim in presenting the optimisation process is to make visible the key underpinning evidence and associated logic that has led us to put forward a preferred set of controls for future management of the LLWR. The logic begins by consideration of the type of controls that are available over the environmental hazards presented by near-surface radioactive waste disposal, including: Controls over waste inventory What, if anything, should be done about the existing disposals? What controls are appropriate over the acceptance of wastes for disposal in future? What conditioning is appropriate for wastes consigned for disposal? Controls over design and operation What control functions are required of the different components of pre- and post-closure engineering, and how are those controls most effectively implemented in terms of: design specification; timing of construction/implementation? LLWR/ESC/(R11)/10025 Page 5 of 119

10 What controls are required in order to ensure that radiological impacts are ALARA with respect to: waste emplacement; operational discharges? What active controls will be needed and for how long during closure of the facility? All these factors are framed by the underlying strategic role assigned to the facility, which, over the foreseeable future, is to provide the necessary national capacity for the disposal of LLW that is not amenable to being managed at higher levels in the waste hierarchy. This document brings together the outcomes from a wide range of studies relating to the types of control listed above, which together have informed decision making relating to future development and operation of the LLWR. Aspects of the SDP defined through this process have, in turn, become incorporated into the more detailed expression of the Environmental Safety Strategy for the facility, which is a core element of the overall ESC. Thus, for example, the broad strategic objective of ensuring isolation and containment of the source is translated in the specific design and operational principles that we will follow in future development of the facility to provide isolation and containment on different timescales through an appropriate combination of active and passive controls. The relevant option assessment studies and their outcomes are grouped together in four main areas. Management controls and interventions relating to past disposals. Management and engineering controls over future waste disposals to the LLWR, including waste acceptance, treatment and packaging, and methods for waste emplacement. Passive engineering controls over the environmental safety performance of the LLWR during the PoA and beyond, taking account of the functional role of engineering features in overall safety strategy, as well as their design and timing of implementation. Active management controls over environmental safety performance, including implications for discharges during the PoA as well as post-closure arrangements for the LLWR site. Inevitably there are certain aspects of control where flexibility is appropriate, in order to enable options to be kept open for the future. There are examples of past decisions at LLWR (such as the potential adaptability of the containment strategy for Vault 9 to reflect a changing focus from long-term storage to disposal) where flexibility has been a relevant consideration in choosing between alternative courses of action. The SDP recognises this, and identifies the monitoring and related activities that will need to be done in order to enable future decisions to be suitably informed. The main outcomes of the analysis in each area are highlighted in turn below. LLWR/ESC/(R11)/10025 Page 6 of 119

11 Management of past disposals We have considered a wide variety of actions that have been identified as having the potential to achieve reduction in the radiological risk associated with past disposals to the LLWR. These options have been assessed from the perspective of the potential magnitude of reductions in dose and risk they may be capable of delivering, set against the wider implications that would be associated with their implementation. Consistent with the underlying objective of continuing to provide capacity for the disposal of LLW at the facility, the emphasis in the options analysis has been on the appropriateness of actions that would target the retrieval of wastes from, or the implementation of other types of remedial action on, specific areas of the trenches. These are where localised high concentrations of key radionuclides are present that play a significant role in determining overall impacts from the facility. Our options assessments demonstrate that the achievable scale of risk reduction (below what is already a low risk by comparison with the regulatory guidance level) is small compared with the costs and disruption that would necessarily be associated with targeted retrieval or remediation. We do not therefore propose to adopt any such actions within our SDP, and this has formed the basis for assumptions adopted in the ESC. Management of future disposals We have explored the implications of potential changes to waste conditioning, consistent with the role of the LLW Repository Ltd to support implementation of more sustainable waste management routes across our consignors and the Nuclear Decommissioning Authority s strategic goal of extending the operational lifetime of the LLWR. Our conclusion is that the use of incineration to optimise the use of available disposal volume is compatible with the objective of demonstrating ALARA in relation to overall control of radiological risks. Moreover, efforts to optimise the packing efficiency in waste containers (and thereby reducing use of grout) are also consistent with the objective of optimising overall safety performance. Our SDP therefore encompasses both these measures, although the strategy for grout volume control is being kept under review alongside the development of alternative packaging solutions. We are continuing to review our strategy for waste packaging, seeking to reduce the amount of radiologically clean material that is disposed. The design of a disposable liner, for use in re-usable containers for waste transport is being developed, and options are being examined as part of the overall development programme for packaging innovations. However, such innovations will not be introduced unless and until they have been subject to rigorous demonstration of their contribution to an optimised safety case. For the purposes of the current ESC, therefore, it is assumed that the standard disposal package continues to be the half-height ISO container. We have examined a wide range of potential enhancements to our approach to waste emplacement within future disposal vaults. Our conclusions are that the preferred options are those that can be implemented without excessive cost and complexity while providing clear benefits in terms of control over radiological doses and risks. Specifically, for certain waste streams and consignments, we have identified opportunities to minimise environmental impacts and potential exposures through controls on where individual higher activity waste packages are placed within the vaults. We have also identified situations where controls on the emplacement of waste types containing absorbed liquids, or with higher potential void space, will LLWR/ESC/(R11)/10025 Page 7 of 119

12 provide benefits in terms of mitigating the potential for release of liquids under load and confidence in the performance of the cap. Our operational controls will therefore be enhanced to ensure that those waste consignments identified as requiring additional control during emplacement are appropriately managed. Some containers will be emplaced such that there is no more than one in a single stack of containers and in any immediately adjacent stacks. This is intended to ensure risks are ALARA for persons who might contact the wastes during coastal erosion of the site. In addition, certain consignments will, so far as is practicable, be emplaced at lower positions within the waste stack, thereby contributing to the optimisation of controls with respect to doses that might arise in the event of inadvertent intrusion or exposure to radon released into buildings constructed on the final cap. The load on specific packages and voidage in stacks will also be controlled where required. The number of consignments that will require emplacement restriction on the basis of the nature of the wastes or their activity content is such that it can be readily achieved within the scope of current practices. Pre- and post-closure engineering design We have reviewed the basis on which the design for Vault 9 was originally determined, noting that it was specifically intended to underpin the provision of longterm temporary storage. With the perspective of the current ESC firmly on the role of the LLWR as a disposal facility, we have examined a range of aspects of the engineering design that would be employed in future development of the site. There are uncertainties associated with providing definitive estimates of long-term engineering performance at the system level, so the main emphasis in comparing engineering options for the purpose of design optimisation was whether there is a preference from the perspective of establishing confidence in demonstrating environmental safety. Having made such a comparison, we then assessed whether that preference would be materially affected by wider considerations. Our conclusions, embedded in our SDP for the LLWR are: Future vault design should be based on the principle of providing comprehensive capture and control of leachate during operations and until such time as active control over leachate discharge is ceased. However, no containment function will be assigned to the vault walls, other than to protect against uncontrolled overflow during extreme rainfall events prior to capping. This will help to preserve unsaturated conditions over the majority of the waste column, minimising interaction between infiltrating water and contaminants within the wastes and lending greater confidence to the demonstration of favourable environmental safety performance. Passive discharge of leachate beyond the PoA will be provided by horizontally extensive drainage layers beneath the vault bases. There is no requirement for an extensive vertical drain to provide contingency against the possibility of near-surface releases, because the design intent is to maintain a very low level of saturation within the wastes beneath the final cap. However, an encircling cut-off wall (COW) will be constructed around the whole facility, to a depth of approximately 2m below the underside of the vault bases, primarily with the aim of minimising the encroachment of saturated conditions within wastes closer to the perimeter of the facility. This will also play an added contingency role in providing reassurance against the possibility of near-surface release of leachate close to the facility. LLWR/ESC/(R11)/10025 Page 8 of 119

13 The COW will be keyed into the perimeter of the final cap, which will have a single dome profile. This will help to mitigate the risk of localised cap failure, which was coupled to the installation of a vertical drain in previous closure designs. The components of the engineered cap have been optimised for performance as a hydraulic barrier, consistent with established best practice and experience from landfill disposal design. The design of the cap and its thickness also provide effective protection to mitigate the likelihood of intrusion by humans, deep rooting plants and burrowing animals. During operation, leachate from the trenches and rainwater run-off from the open vaults will continue to be managed as now, by the collection, monitoring and controlled discharge to sea via the Marine Pipeline, subject to the terms of our Permit. We will seek to construct the final cap progressively in strips to reduce infiltration to the trenches and to ensure that leachate management is adequate even when there is a need to manage storm waters from extreme weather events as new vault slabs are constructed. However, flexibility is incorporated in our SDP to allow for the possibility that the current interim cap may not continue to remain the optimum barrier to infiltration over the medium term, necessitating replacement or earlier installation of the final cap over the whole trench area. Should this be the case, or if there are concerns over the amount of settlement that remains to be expressed within the trench waste column, our SDP allows for the option of accelerating construction while maintaining confidence in the longterm performance of the cap by undertaking its installation in two phases. A passive gas venting arrangement is currently incorporated in the design of the final cap, to provide confidence that pressure differentials, for example caused by the generation of bulk gas within the facility, will not threaten the performance of the cap as a barrier to infiltration. The decision will be made later on the final design and whether the vent will be closed prior to the release of the site from active management control. Operational and post-closure management controls We have conducted a comprehensive review of discharge and disposal practices for wastes generated on the site. Although potential alternative approaches to both aerial and liquid discharges exist, we consider that current arrangements represent an optimised approach. Nevertheless, we remain open to the possibility of change in future (for example, recycling leachate in the preparation of waste conditioning grout) to ensure best available techniques (BAT) are used in the development of possible new facilities that may needed on the site. It is not appropriate to define detailed plans for monitoring and management during the closure phase at this stage, for a site that we expect to remain in continuing operation for many decades and long-term direct management control thereafter. However, assessment calculations demonstrate that intrusion hazards associated with shorter-lived radionuclides are expected to fall substantially over the first 100 years following repository closure (and much less rapidly thereafter), and we consider such a period to be an appropriate baseline figure for the assumed length of active control over the site in our assessment calculations. We also recognise that some long-term control over use of the site may be required to ensure that potential exposures to releases of C-14 labelled gas do not lead to unacceptable risks from use of the cap for smallholder farming. LLWR/ESC/(R11)/10025 Page 9 of 119

14 We consider that the most effective way of ensuring some control over use of the site persists for as long as practicable is to ensure that relevant information is identified and managed in an appropriate way such that knowledge is retained regarding the nature of hazards associated with the disposals that have taken place. We believe long-term information management would be assisted by close involvement of the local community in making decisions on beneficial future use of the site, as well as by legal covenants to restrict the use of the land. We do not anticipate the installation of engineered controls at the site to protect the LLWR in the long-term from the expected effects of coastal erosion, and we have not taken credit for such actions as part of the ESC. From the perspective of achieving risks that are ALARA, we believe there is appropriate evidence from the results of our safety assessments to indicate that such measures (which would imply a long-term burden for maintenance and re-building) would not be proportionate to the risks involved. LLWR/ESC/(R11)/10025 Page 10 of 119

15 Contents 1 Introduction Objectives Scope Structure History and Basis for Optimisation Role of the LLWR Current Status of the Facility Site Trench Disposals Vault Vault Approach to Optimisation Control Measures Options Assessment Management of Past Disposals Initial Options Assessments Schedule 9 Studies Targeted Retrievals Waste Heterogeneity and Long-term Radiological Significance Waste Heterogeneity Implications for the Period of Authorisation Evaluation of Retrieval Options Other Remedial Actions Waste Heterogeneity and Radiological Significance Evaluation of Alternative Remedial Actions Status Summary Management of Future Waste Disposals Waste Acceptance Optimisation of Physical Capacity Radiological Capacity Waste Conditioning and Packaging Incineration Grout Formulation and Quantities Waste Package Waste Emplacement Pre- and Post-closure Engineering Design Vault 9 Planning Basis Design Optimisation for Disposal Scope of Design Optimisation Evidence and Approach Engineering Mitigation of Threats to Isolation and Containment Analysis of Design Features Timing of Cap Installation Management Controls on Discharges Operational Controls Aerial Discharges LLWR/ESC/(R11)/10025 Page 11 of 119

16 6.1.2 Liquid Discharges Solid Discharges Post-operational Controls Site Management Monitoring Land Use Control and Information Management Site Development Plan Overall Optimisation Strategy Ongoing Site Management Management of Past Disposals Controls on Discharges Future Operational Development Management of Future Disposals Pre- and Post-closure Engineering Design Post-closure Controls Site Management Active Controls on Releases Conclusions References Appendix 1: List of Acronyms LLWR/ESC/(R11)/10025 Page 12 of 119

17 1 Introduction 1.1 Objectives The principle of optimisation is a fundamental element of international approaches to protection against radiological hazards, including radioactive wastes. Within this broad context, optimisation is defined as determining what level of protection and safety makes exposures, and the probability and magnitude of potential exposures, as low as reasonably achievable, economic and societal factors being taken into account [18]. Optimisation has been identified as a governing principle in UK regulatory guidance on authorisation requirements for the near-surface disposal of radioactive waste (the GRA) [19]. In this respect, optimisation is described as a continuing, forward-looking and iterative process aimed at maximising the margin of benefit over harm with the aim of questioning whether everything reasonable has been done to reduce risks (GRA paragraph 4.4.3). In identifying the need for balance between the detriment associated with radiological risk and other benefits and detriments (economic, human, societal, political, etc.), it is further recognised that The result of optimisation provides a radiological risk at a suitably low level, but not necessarily the option with the lowest possible radiological risk (GRA paragraph 4.4.4). Demonstration of optimisation is also identified in the GRA as a specific radiological requirement for authorisation 1. Requirement R8: Optimisation The choice of waste acceptance criteria, how the selected site is used and the design, construction, operation, closure and post-closure management of the disposal facility should ensure that the radiological risks to members of the public, both during the period of authorisation and afterwards, are as low as reasonably achievable (ALARA), taking into account economic and social factors. In supplementary guidance to the GRA [20], related to implementation of the European Groundwater Directive, the Environment Agency underlines the importance of achieving doses and risks that are as low as reasonably achievable (ALARA) in relation to radiation exposure through giving proper consideration to the input of radioactive substances in groundwater, while at the same time meeting the wider provisions of Environmental Permitting Regulations 2010 (see footnote 1). The supplementary guidance specifically highlights the potential need to consider alternative design options and to establish an appropriate balance in preventing or limiting the input of pollutants to groundwater between the period of authorisation (PoA) and subsequently. 1 Under the Environmental Permitting (England and Wales) Regulations 2010, which postdate publication of the GRA, waste disposal arrangements falling under RSA90 are now authorised through an environmental Permit. LLWR/ESC/(R11)/10025 Page 13 of 119

18 Insufficient consideration of optimisation was one of the main criticisms by the Environment Agency following their review [21] of the 2002 Safety Cases [22,23] produced by the previous operator, BNFL. This report summarises the evidence and logic in support of arguments made in 2011 Environmental Safety Case (ESC) regarding the demonstration of optimisation in the management, design and operation of the LLWR. Such evidence also contributes to the underpinning of a range of reference assumptions for development of the facility that have been adopted in safety analyses undertaken for the ESC. 1.2 Scope The central role of an environmental safety strategy lies at the heart of expectations regarding the demonstration of consistency with regulatory principles and requirements (GRA, paragraph 7.2.2). A comprehensive account of the Environmental Safety Strategy (ESS) for the LLWR is therefore a core element of the overall ESC [1]. We consider that such a strategy encompasses both the way in which environmental safety is achieved and the means by which it is demonstrated. Hence, the ESS for the LLWR establishes first a set of over-riding safety objectives, consistent with the declared goal of providing a route for the environmentally safe disposal of LLW. It also defines a process by which those safety objectives are translated into specific actions (the development plan for future management and use of the facility), the management framework within which those actions are implemented, and the characterisation, monitoring and analysis undertaken to provide assurance that objectives are being met. The key strategic environmental safety objectives are identified as: Control of the source i.e. managing the total inventory and concentration of radionuclides being disposed, as well as the wasteforms in which they are disposed and the manner of their emplacement. Isolation of the source from disturbance i.e. providing optimised protection against the threats of disturbance by natural processes (erosion, water infiltration) and human activities. Containment of the source i.e. ensuring that the wasteform, packaging, emplacement arrangements and repository engineering act together to minimise the likelihood that contaminants will be released from the facility, for as long as practicable. Management of residual releases i.e. seeking to control any releases that may occur in such a way that their impacts are minimised. Consistent with overall safety management principles, discussed in the GRA [19], the emphasis during the PoA is on aiming for passive safety so far as is reasonably practicable, but with some necessary active engineered systems and human actions. After the end of the PoA, the ESC has to rely entirely on features of the system that do not depend on human intervention or active engineered measures. Within this system, optimisation is the process by which our approach to design, construction, operation, closure and post-closure arrangements for the LLWR is identified and justified. It takes the over-riding safety objectives and examines alternative ways in which they can be addressed, refining the broad strategic goals LLWR/ESC/(R11)/10025 Page 14 of 119

19 into a preferred set of control measures that address environmental safety management. These preferred approaches provide the foundation of our Site Development Plan (SDP) for the LLWR. Aspects of the SDP then, in turn, become incorporated into a more detailed expression of the Environmental Safety Strategy for the facility, taking into account its environmental setting, history and anticipated future role. Hence, for example, the broad strategic objective of optimising containment of the source is translated in the specific design and operational principles that will be followed in future development of the facility to provide containment on different timescales through an appropriate combination of active and passive controls. Optimisation in the design, construction, operation, closure and post-closure arrangements for the LLWR therefore covers overall safety strategy (i.e. the engineered components of the disposal system and their functions in assuring safety) as well as detailed aspects of implementation. In considering the principle of optimisation, the GRA acknowledges (paragraph 4.4.4) that Optimisation decisions are constrained by the circumstances prevailing at the time. This reflects recognition that demonstrating radiological doses and risks are ALARA is a continuing and iterative process through the lifecycle of a disposal facility. Because the LLWR is an operating facility with a substantial history, decisions regarding its future management, design and operation are framed, and to a certain extent also constrained, by past actions. The logic and evidence developed in support of decisions regarding the future of the facility therefore start from an understanding of decisions that were taken in the past and the considerations that related to those decisions when they were taken (GRA, paragraph ). Not all aspects of past decisions are considered in detail in this report, although the overall framing of future plans is an integral part of what is presented here. Where appropriate, cross-reference is therefore made to material presented in other ESC reports. In particular: the Site History and Description report [3] and the Waste Acceptance report [15] provide accounts of some of the background to the current status of the facility. All ESC reports are listed in the table in the Preface. Optimisation of the design and closure of the facility is closely linked to, and provides a basis for, the more detailed discussion of the reference engineering design [5]. The design of the facility, alongside consideration of the optimisation of future waste conditioning and emplacement as well as possible remedial actions for past disposals, in turn informs the development of a reference inventory to underpin the ESC [4] and understanding of how the wastes and engineered system are expected to evolve over time [6]. In addition, decisions regarding the management of waste conditioning and emplacement are closely tied to certain aspects of the development of controls on waste acceptance [15]. Optimisation of safety management during the PoA underpins assumptions relevant to the analysis of environmental safety on this timeframe [11]. Finally, the identification and justification of a preferred approach to post-closure arrangements is informed by, and also supports related assumptions made in, long-term assessments [12]. 1.3 Structure The structure of the report reflects the scope described above. It covers the framing of optimisation decisions and the range of management controls for which such decisions have been addressed within the ESC in defining the overall ESS and SDP for the LLWR. Hence, in this report: LLWR/ESC/(R11)/10025 Page 15 of 119

20 Section 2 sets out the basis for optimisation, including the current status and anticipated future role of the LLWR, the key areas for which decisions need to be made in forward planning, and the approach taken to develop an optimised SDP for future management of the facility; Section 3 discusses the strategy for management of past disposals to the LLWR, with a particular emphasis on the hazards associated with disposals to the trenches; Section 4 considers the optimisation of future disposals, including the definition and implementation of controls on waste acceptance, potential treatment and packaging innovations, and methods for waste emplacement; Section 5 examines the role of passive engineering controls in contributing to the environmental safety performance of the LLWR and their role (including the timing of their implementation) as part of an optimised SDP; Section 6 discusses the optimisation of active management controls on environmental safety, including post-closure arrangements for the LLWR site, and their implications for discharges during the PoA; Section 7 provides a summary of current state of optimisation in relation to the design, construction, operation, closure and post-closure arrangements for the LLWR, highlighting those areas where flexibility has been built into the SDP to enable options to be kept open in future. The overall conclusions of the work described in this report are summarised in Section 8. A list of acronyms used in the report is given in Appendix 1. A general glossary for the ESC is appended to the Main Report [1]. LLWR/ESC/(R11)/10025 Page 16 of 119

21 2 History and Basis for Optimisation Decisions taken now and in the future regarding the development and operation of the LLWR are framed, and to some extent may also be constrained, by past actions. Past and current developments at the facility, as well as the vision for its future, are in turn influenced by national policy and strategy relating to the management and disposal of LLW. In this section, we review the major considerations that establish the background against which the future development of the LLWR will be carried out. The control measures that need to be taken into account in determining an optimised safety strategy are identified, and the approach taken in informing decision making between options described. 2.1 Role of the LLWR The Nuclear Decommissioning Authority (NDA) has published a UK Strategy for the Management of Solid Low Level Radioactive Waste from the Nuclear Industry [24]. The strategy has been prepared by the NDA for the UK Government and devolved administrations in response to the Policy for the Long Term Management of Solid Low Level Radioactive Waste in the United Kingdom, published in 2007 [25]. The NDA has also developed a strategic partnership with the LLWR Site Licence Company (LLW Repository Ltd) in order to support the delivery and implementation of the UK Strategy. With respect to the role of the LLWR, the UK Strategy underlines that: systematic implementation of the waste hierarchy at waste producer sites is expected to reduce the volumes of LLW that need to be disposed. Where wastes cannot be prevented from arising, the UK Strategy seeks to optimise the segration of radioactive waste so that as much as practicable is handled as VLLW or exempt waste, and to increase opportunities for reuse and recycling of waste materials; disposal needs to be retained as an option for some wastes that are not amenable to being managed at higher levels in the waste hierarchy; capacity for receiving waste at the LLWR site, from both a volumetric and radiological perspective, is finite; consistent with the Policy requirement for a risk-informed approach to ensure safety and protection of the environment [25], waste consignors will be expected to make appropriate use of alternative waste management routes to support volume reduction as well as fit-for-purpose disposal facilities that reflect the lower hazard associated with very low-level waste (VLLW); a key NDA priority, driven by the requirement to use public money efficiently, is to make the best use of existing LLW management assets. Continued availability of the LLWR, where the majority of solid LLW is disposed of, is central to the strategy. The strategy seeks to extend the life of the LLWR by only disposing of wastes that cannot be managed via other routes, to ensure capacity for the long term. LLWR/ESC/(R11)/10025 Page 17 of 119

22 A Strategic Environmental Assessment (SEA), conducted to support development of the UK Strategy, concluded that best use of the LLWR is preferred over other options for the provision of LLW disposal capacity [26]. Best use of the LLWR was assumed to involve rigorous application of the waste hierarchy across NDA sites and disposing only those wastes that require the level of safety and security offered by engineered disposal 2. With respect to the provision of disposal capacity, the conclusions of the SEA (and hence the strategy that it informs) are nevertheless contingent on the identified key assumption that all alternative strategic options (i.e. variants based either on continued use or replacement of the LLWR) would be capable of meeting regulatory requirements and, specifically, that LLWR Ltd will be able to make an acceptable Environmental Safety Case for the LLWR [26]. Hence the identified candidate strategies were not differentiated in terms of their implications for health and safety, except with respect to the additional hazards that would be associated with retrieval, handling and transfer of wastes from the LLWR to a hypothetical replacement facility. Early replacement of the LLWR is not, therefore, anticipated in the UK Strategy [24], provided an environmental safety case for the continued use of the facility can be made and subject to any necessary regulatory and planning approvals [26]. The expectation is that an alternative national facility for the engineered disposal of LLW would need to be developed only if the additional capacity proves to be required. It is planned that the UK Strategy (including the role of the LLWR) will be reviewed periodically, at least in line with the overall NDA Strategy review cycle [24]. 2.2 Current Status of the Facility A detailed account of the past development of the LLWR is provided in the Site History and Description report [3]. In what follows, some relevant aspects of existing features of the facility and its history are discussed, to provide a context for the optimisation considerations addressed in the current ESC Site The site of the LLWR was first developed in 1939 as a Royal Ordnance Factory (ROF). Ownership subsequently passed to United Kingdom Atomic Energy Authority, which was granted planning consent in 1957 for the disposal of waste in the northern 40 ha of the site. The first Certificate of Authorisation for disposal of LLW was granted in 1958 under the terms of the Atomic Energy Act 1954, and disposal operations commenced in Ownership and responsibility for the site was transferred to British Nuclear Fuels Ltd (BNFL) when the company was formed in 1971, and the site became a part of the Nuclear Decommissioning Authority s estate when the organisation was established as a non-departmental Government body in Best use of the LLWR (in the context of its role as a component of the UK Strategy for LLW management) is not the same as optimisation with respect to the control of radiological risks. Nevertheless, operational considerations associated with managing the capacity of the facility are important in achieving the necessary balance between radiological detriment and other benefits and detriments (Section 1.1). Factors relevant to LLWR s role in implementing UK Strategy through the waste hierarchy are therefore highlighted in Section 4 as part of the wider discussion of issues relevant to the optimisation of risks associated with future waste disposal plans. LLWR/ESC/(R11)/10025 Page 18 of 119

23 Land ownership and the proximity of the LLWR site to Sellafield were undoubtedly important factors in its identification as a suitable site for LLW disposal. In addition, it was judged that the boulder clay, which exists at a shallow depth on the site, would provide an effective barrier to the downward infiltration of radioactivity into the underlying Triassic sandstone aquifer. Leachate could therefore be collected in the trench drains and subject to controlled discharge (originally to the Drigg Stream, but since 1991 via the Marine Pipeline). Proximity to Sellafield, which is the dominant source of waste consignments to LLWR, remains a relevant factor in the determination of UK Strategy for the disposal LLW [26]. The majority of waste disposed to the facility is transported by rail from Sellafield, the waste either originating at Sellafield or coming from other sites and consigned via the WAMAC compaction facility (see Subsection 2.2.3). It could be argued that the original approach of loose tipping and leachate collection (see below) placed the emphasis of safety arguments on controlled gradual release and dilution, rather than extended long-term containment. As such, siting considerations did not differ significantly from those associated with general practice in relation to the controlled landfill of waste. In any case, the long-term evolution of the site does not appear to have been a significant consideration in earlier decision making relating to disposals. Taking into account the increased emphasis on isolation and containment in international policy and practice for the disposal of solid radioactive wastes, and recognising the long-lived component of the radioactive inventory at LLWR, the implications of site evolution have played a greater role in more recent safety analyses. The potential implications of climate and landscape change were examined in some detail in studies supporting the 2002 Post-closure Safety Case (PCSC) for the facility [23] and in the subsequent performance assessment update [27]. Whereas prominence was originally given to the potential implications for the site associated with glaciation, several tens of thousands of years in the future, the focus now is emphatically on sea-level rise and coastal erosion, which have implications over shorter timescales [8]. Phenomena that were not relevant to the original selection of the LLWR site therefore play a central role in framing the ESC and related optimisation decisions. Following the Environment Agency s review of the 2002 PCSC, Requirement 2 of Schedule 9 of the resulting site Authorisation called for a review of risks arising from potential site termination events (including coastal erosion) and options for reducing them. In its subsequent assessment [28] of LLW Repository Ltd s submission in response to this Requirement, the Environment Agency noted that assessment of the impacts associated with coastal erosion of the facility should be considered within the expected ( normal ) evolution of the site, requiring a robust assessment of the exposures that could arise when that erosion takes place. Regulatory guidance also highlights that, whilst it is necessary to ensure that optimisation takes into account uncertainties in how the disposal system might evolve after closure of the facility, unlikely circumstances should not have undue influence on design, construction or operation (GRA, paragraph ). As far as the ESC is concerned, the anticipated destruction of the LLWR by coastal erosion (albeit with uncertainty as to when this would take place) [8] sets a natural limit to the timeframe over which detailed optimisation considerations are deemed to be relevant. LLWR/ESC/(R11)/10025 Page 19 of 119

24 2.2.2 Trench Disposals For the first three decades of operations, disposals were solely by tumble tipping drummed, bagged and loose wastes into excavated trenches. There are seven disposal trenches in total and disposals were completed to Trench 7 in Future actions relating to the control of hazards associated with authorised trench disposals are a relevant factor in optimisation of the overall disposal facility, to ensure that radiological risks are ALARA. Considerations include a combination of possible interventions and factors relating to the design of closure engineering for the site. As background to examination of the issues to be taken into account, it is relevant first to consider the basis for actions that took place in the past. The loose tipping of waste was a widespread and internationally accepted practice when the trenches were first commissioned. The approach was considered to be efficient in terms of space utilisation, radiation exposures of the workforce and the time taken for disposal operations [29]. It was also a flexible system that allowed trenches to be formed in a phased manner in response to the demand for disposal space. No records have been identified relating to the consideration of alternative approaches to disposal in the original development of the facility. The boulder clay on the site was expected to provide a low hydraulic conductivity base to limit the vertical migration of leachate through the trenches. In addition, a slight fall was incorporated in the base of each trench to cause leachate to be collected by an interceptor drain at the southern end of the disposal area. Experience in operations led to improvements in detailed design, construction and operation. While maintaining consistency with overall design principles, such improvements included: variation of plan area and depth to optimise the available disposal capacity; construction of firebreaks and covering waste with soil at the end of each day s disposal operations to minimise the risks of fire; installation of perforated drainage pipes along the trench bases to promote the flow of leachate to the interceptor drain; engineering the base of trenches by rotovation and mixing with sodium bentonite to a depth of 0.5m, at locations where the basal clay layer was absent. A review of the leachate management system for the site was undertaken as part of a general review of the disposal facility in the 1980s [30]. This determined that the monitored discharge of leachate through the Marine Pipeline to offshore diffusers, rather than to the Drigg Stream, was the preferred (ALARA) approach to the control of public exposures from site operations. Post-emplacement engineering has also been undertaken to mitigate the impacts of trench disposals. A COW was installed from the north-west corner of Trench 3 (the western-most trench) to the southern end of Trench 7 (the eastern-most trench) between late 1989 and 1995 [5]. This was undertaken as an operational measure in order to limit the potential for lateral migration of leachate from the trenches to the adjacent railway cutting. In combination with the interim cap (see below), the COW also supports the minimisation of leachate production by controlling the lateral migration of meteoric water into the trenches from the surrounding land. LLWR/ESC/(R11)/10025 Page 20 of 119

25 An interim cap, extending over the top of the COW, was completed over Trenches 1 to 6 in 1989 and was subsequently extended to cover Trench 7 following the completion of trench disposal operations in The purpose of the cap is: to isolate the waste from the near-surface environment, limiting the possibility of disturbance and protecting the waste during the initial phases of waste settlement; to control the release of gas from the trench disposal area; to limit the infiltration of meteoric water into the disposal area and hence the volume of leachate generated. In association with these post-emplacement actions, the decision was also made as part of the wider review to phase out the disposal of loose wastes in favour of the orderly emplacement of compacted, containerised and grouted wastes in engineered concrete vaults (see Subsection 2.2.3). Vault operations commenced in 1988, but the phasing out of trench disposals was not completed until Optimisation considerations associated with the future control of hazards from the trench wastes, during the PoA and beyond, include: the possibility of remedial actions to provide additional containment or to recover some or all of the trench waste inventory (Section 3); the implications of waste settlement and trench leachate management for the design and scheduling of future site engineering and closure measures (Section 5); the possible need to replace or refurbish existing engineered controls, including the interim cap (Section 6) Vault 8 The adverse visual impact created by the tumble tipping of wastes into trenches was highlighted, as part of a wider review of radioactive waste management in the UK, in a report of the House of Commons Environment Select Committee in 1986 [31]. As part of its review of site operations in the mid-1980s [30], BNFL sought to identify a disposal system to supersede the trenches that would improve management practices, enhance containment and mitigate the visual impact of operations. Options for facility design, waste packaging and site closure were considered in a series of qualitative and quantitative studies [29]. In addition to the development of the COW and interim cap for those parts of the facility where disposal operations had been completed, the decision was made to move to an engineered concrete vault disposal concept. The new disposal vault was identified as Vault 8 in order to maintain continuity of numbering from Trench 7. It was designed to fit into the available space in the northwest corner of site. The design incorporated a secant pile wall to provide structural stability as well as water retaining capability along the shared boundary with the adjacent Trench 3 while minimising wasted space that could otherwise be used for disposal capacity. Although the concrete base and walls provide a physical barrier to the movement of groundwater and leachate, the emphasis of the Vault 8 design was largely on LLWR/ESC/(R11)/10025 Page 21 of 119

26 operational aspects of waste emplacement and storage, including the use of waste containerisation while at the same time seeking to maximise the capacity (and hence operational lifetime) of the facility. The vault has surface water drains to collect rainwater from the surface of the base slab, while an under-slab drainage blanket and perimeter drains collect groundwater from beneath and around the vault. The ultimate barrier to downward migration of leachate is the same as for the trenches the low conductivity of the naturally-occurring boulder clay layer beneath the vault base. Where this could not be assured (e.g. where the top of the natural clay dipped below the vault formation level or where it was assessed to be less than 1 m thick), bentonite was mixed with the local soils to a depth of 300 mm to achieve an in situ permeability of m s -1 or less [5]. Containment of the disposed radionuclides within the vault was enhanced by changes to the wasteform to reduce the potential for interaction between the wastes and waters passing through the facility. In terms of long-term safety, the functional requirements assigned to the wasteform also included keeping the residual voidage low, in order to minimise the potential for settlement of the final cap. In addition, the aim was to distribute the load on the base slab as uniformly as practicable to minimise differential settlement [32]. Enhancement of the wasteform involved three main stages. High-force compaction is applied to the raw waste as far as is practicable, according to the type of waste material. This is mostly undertaken at the WAMAC facility on the Sellafield site, which was opened in WAMAC accepts compactable raw wastes packed into 1 m 3 boxes as well as 200 litre drums. A compaction service for 200 litre drums is also provided at the Winfrith site. The compacted pucks, together with other non-compactable wastes, are then emplaced within standard ISO steel containers. The wastes are finally grouted in place within the containers, prior to emplacement in the vault. The grouting of the compacted wastes within the final container takes place at the LLWR Grouting Facility, which also started operations in the mid-1990s. All wastes are subject to waste receipt monitoring, whether compacted or not. In determining the preferred wasteform, high-force compaction was judged to be consistent with good practice in other national programmes. It was recognised as being capable of providing the desired low product voidage as well as efficiency in the use of disposal capacity within the facility. Incineration and other thermal treatments were not favoured at that time as a standard practice owing to their anticipated limited overall impact on total volume reduction across all waste streams, the lack of suitable facilities for the complete range of LLW waste streams arising in the UK, and public concerns over environmental impact from incinerator operations [32]. Apart from low-force compaction, no records have been identified relating to the consideration of other approaches to primary waste treatment. The half-height 20 m 3 ISO waste container was selected as the standard container, being sufficiently large to minimise the number of waste handling operations and consistent with practical limits on weight during movements on the site. The container could also be readily demonstrated to be consistent with Transport Regulations during shipment from Sellafield and elsewhere to the LLWR. A metal LLWR/ESC/(R11)/10025 Page 22 of 119

27 container was also considered to be more cost-effective, as well as being more efficient, than concrete in making use of the available space for disposal. A low-viscosity grout formulation was determined with the aim of providing efficient flow properties during filling, to be self-levelling and to settle sufficiently quickly in order to limit the need for temporary buffer storage prior to final emplacement. The top layer of grout across the surface of the waste container contents helps to distribute load throughout the wasteform. It was also recognised that the grout could contribute to the long-term maintenance of a local environment that would be beneficial to containment of a number of radionuclides, even after the physical containment barriers had degraded [32]. As well as the standard half-height ISO container, a small number of additional designs were developed. These include a third-height 13 m 3 ISO container for use with very dense wastes (such as metal plates) enabling the weight handling limit to be met while maintaining efficient filling of the container. A nominal 10 m 3 ISO container, with reduced length and width, was also introduced for easier handling and use in smaller areas at waste producers sites. The majority of waste emplacements to Vault 8 prior to 1995 needed to be retrieved for upgrading (at the LLWR Grouting Facility) to meet the new standard. Not all of the re-worked containers completely met requirements for control of void space. In some cases (where full-height ISO containers has been used for non-compacted wastes) the contents were removed and processed (via WAMAC) prior to repackaging. Others were deemed unsuitable for processing of the contents and such containers were typically repositioned at the top of the container stacks, in order to minimise their load-bearing requirements. Vault 8 has also received small arisings of large items of waste (such as uranium hexafluoride cylinders and the heat exchangers from the Windscale Advanced Gas-cooled Reactor) that were judged at the time to be impracticable to size-reduce into standard containers. Such items have mostly been grouted in situ using mobile grouting facilities; however, not all large items disposed to Vault 8 in this way were grouted internally. Optimisation considerations associated with the future control of hazards from Vault 8 wastes, during the PoA and beyond, include: the consistency of the vault design with overall strategy for pre- and post-closure engineering of the LLWR, including the implications of waste settlement (Section 5); the scheduling of cap installation, including the management of vault drainage and the need to protect wastes from degradation (Sections 5 and 6) Vault 9 The 2002 PCSC [23] was originally expected to establish a basis for the continued authorisation of disposal operations at the LLWR. The safety case was supported by a Site Development Plan [32] that set out expectations for the development of future vaults and final closure of the facility within what was formerly understood to be the consented area for disposals (dating back to the original 1957 planning consent). It was anticipated that development would take place on a phased basis, following a broadly similar design philosophy to that adopted for Vault 8, but with an increased height of container stacking, achieved by increasing the depth of the vault base slabs. It was also expected that, subject to the detailed geology of the area, more LLWR/ESC/(R11)/10025 Page 23 of 119

28 use would need to be made of engineered clay as the foundation for future vaults because of their greater depth. These plans were not put into effect, for two main reasons. First, Cumbria County Council established that the concept of a consented area had no meaning in the context of current and future use of the LLWR site, and that future developments would require planning permission from the Planning Authority via the appropriate procedures. In addition, the Environment Agency, in its review of the 2002 Safety Cases and the subsequent decision on disposal authorisation, considered that the safety cases had failed to make an adequate or robust argument for continued disposals of LLW [21]. As a result, in its Decision Document [33], the Environment Agency determined that the continued disposal of LLW to Vault 8 would be authorised (subject to existing annual solid waste disposal limits) until its capacity was reached, but that any further consignments to the facility would be for the purpose of temporary storage only. Disposals to the proposed Vault 9 (and any future vaults) would not be authorised until the appropriate planning permission had been received, confidence had been established in estimates of the radiological capacity of the site, and the Environment Agency was satisfied that a sufficiently robust safety case had been made. In the light of decisions taken by the regulators and planning authorities, a detailed analysis and associated rationale was developed for a preferred Modular Vault design [34], to be adopted in the construction of Vault 9 and to serve as a baseline for any future vaults. Long-term environmental impacts were considered through qualitative judgements comparing different designs with the outputs of the 2002 PCSC. A range of information on cost and wider socio-economic impacts was also developed to aid this process, and the options assessment process was undertaken with the involvement of stakeholder representatives, including members of the West Cumbria Site Stakeholder Group. The implementation of a comprehensive doubleliner system for the vault base and walls was judged to be preferred because it: better satisfies the design principles, as defined in the applicable UK regulations and Best International Practice (i.e. containment) and will therefore be more acceptable to the regulators and other stakeholders. It was also the clear preference of the stakeholder elected representatives. [34] The options study recognised that extended leachate containment within the vault would lead to increased contact times with the waste and corresponding increased concentrations of radionuclides. In order to mitigate the risk of releasing this concentrated leachate to upper groundwaters (i.e. at the level of the top of the vault walls), the closure design also incorporated vertical drains and a substantial COW both of which would be constructed at a later stage in site operations. However, it was also highlighted that the preferred option also offered scope for flexibility, enabling the containment strategy to be adapted in future, if required. Temporary planning permission for Vault 9 was granted in January 2008, on the basis that it would be used for storage only. This did not preclude the possibility of subsequently achieving planning permission for the vault to be converted to use for disposal. However, such permission would necessarily be conditional on satisfying Environment Agency requirements for regulatory authorisation. The current ESC is a key supporting document in LLW Repository Ltd s plans to achieve both planning permission and regulatory authorisation for continued LLW disposal at the facility. LLWR/ESC/(R11)/10025 Page 24 of 119

29 Construction of Vault 9 according to the Modular Vault design commenced in September 2008 and was completed in July The detailed content of a further round of design optimisation studies (focused on the proposed use of Vault 9 for waste disposal) is discussed later in this report (Section 5); however, an important element was to revisit the evidence and rationale that underpinned the preferred safety strategy for long-term passive leachate management. Specifically, the apparent attractiveness to community stakeholders (and the assumed attractiveness to the Environment Agency) of the preferred Modular Vault design was embodied in its description as containment of leachate for as long as practicable [34]. On further examination, however, the confidence that could be attached to this description, and the extent to which it could be described as an optimised strategy for disposal, has been called into question (see Section 5). The implications of the latest passive leachate management studies for vault design have been incorporated in the reference engineering design [5]. No corresponding detailed review was undertaken of wasteform prior to the granting of temporary planning permission for Vault 9. However, in line with development of the NDA s UK Strategy for management of LLW [24], LLW Repository Ltd is working with waste consignors to implement waste management routes that increase opportunities for reuse and recycling, or otherwise support a reduction in volume or change in the nature of waste sent for disposal to the LLWR. Innovations in waste conditioning and emplacement have also been examined as part of the wider optimisation of facility operations (Section 4), and their implications incorporated where appropriate into the definition of waste acceptance controls [15]. 2.3 Approach to Optimisation Control Measures The essential nature of optimisation, whether in design, construction, operation, closure or post-closure management, or in the determination of waste acceptance controls, is to determine a preferred set of control measures consistent with the goal of achieving ALARA with respect to the management of radiological impacts. An effectively designed and executed optimisation process informs the development of plans and programmes for the lifetime management of the facility, by establishing the actions that need to be taken to ensure an appropriate level of both active and passive control over the hazards presented by the wastes. Optimisation is a continuing, forward-looking and iterative process aimed at maximising the margin of benefit over harm, which involves continually questioning whether everything reasonable has been done to reduce risks (GRA, paragraph 7.2.2). Given the strategic role assigned to the LLWR and the current status of the facility (i.e. understanding why the facility is the way it is), the requirement for the ESC is therefore to determine a vision for the controls that need to be put in place now and in the future. The optimised controls, both active and passive, that are established then effectively become the functional expression of the ESS for the facility, implemented in the SDP (Section 7). In broad terms, the controls that are available over the environmental hazards presented by radioactive disposal can be classified as: controls over waste inventory; what, if anything, should be done about the existing disposals? LLWR/ESC/(R11)/10025 Page 25 of 119

30 what controls are appropriate over the acceptance of wastes for disposal in future? what conditioning is appropriate for wastes consigned for disposal? controls over design and operation; what control functions are required of the different components of pre- and post-closure engineering, and how are those controls most effectively implemented in terms of; design specification; timing of construction/implementation? what controls are required in order to ensure that radiological impacts are ALARA with respect to; waste emplacement; operational discharges? what active controls will be needed and for how long during closure of the facility? All these factors are framed by the underlying strategic role assigned to the facility (Subsection 2.1), which is to provide, over the foreseeable future, the necessary national capacity for the disposal of LLW that is not amenable to being managed at higher levels in the waste hierarchy and requires the protection of vault disposal. Moreover, given that this strategy is itself contingent on being able to make an acceptable environmental safety case for disposal, the optimisation of control measures necessarily relates to the operation of the LLWR as a final disposal facility Options Assessment Regulatory guidance relating to the principle of optimisation highlights that optimisation provides a radiological risk at a suitably low level, but not necessarily the option with the lowest possible radiological risk (GRA, paragraph 4.4.4). It also acknowledges that measures taken today cannot guarantee a particular outcome in future, and that the question of whether sufficient measures have been taken is inherently a matter of judgement (GRA, paragraph 4.1.3). The need to demonstrate that an appropriate balance has been achieved between radiological risk and other possible detriments, or benefits, in the determination of management controls over waste disposal lies at the heart of the optimisation process for the LLWR. As a general rule, and particularly in relation to the definition of management strategy, this implies the need to examine what feasible options are available for a particular aspect of control. These options then need to be compared in terms of the balance they represent between radiological risk and other factors. This does not imply that a detailed, quantitative multi-criteria decision analysis will always inevitably be the most appropriate method of options assessment. With respect to long-term environmental management, in particular, significant uncertainties arise and judgements are required that cannot readily be captured through quantitative best estimate scoring and weighting exercises. A range of different approaches to options evaluation have been followed in the different aspects of optimisation reported here. Whatever the approach to analysis, however, the aim in what is presented is to make visible the key underpinning evidence and LLWR/ESC/(R11)/10025 Page 26 of 119

31 logic that has led us to put forward a preferred set of management controls for future management of the LLWR. With reference to broad scope of controls identified above (Subsection 2.3.1), the relevant option assessment studies and their outcomes are grouped together as follows: management controls and interventions relating to past disposals (Section 3); management and engineering controls over future waste disposals to the LLWR, including waste acceptance, treatment and packaging, and methods for waste emplacement (Section 4); passive engineering controls over the environmental safety performance of the LLWR during the PoA and beyond, taking account of the functional role of engineering features in overall safety strategy, as well as their design and timing of implementation (Section 5); active management controls over environmental safety performance, including implications for discharges during the PoA as well as post-closure arrangements for the LLWR site (Section 6). Inevitably there is a degree of overlap and dependency between the management actions and controls relating to different aspects of the facility. There are also certain aspects of control where flexibility is appropriate, in order to enable options to be kept open in future. These are highlighted as they arise in the discussion of individual factors. An overall synthesis of the outcomes, interactions and contingencies is provided in the SDP, summarised in Section 7. LLWR/ESC/(R11)/10025 Page 27 of 119

32 This page is left blank intentionally. LLWR/ESC/(R11)/10025 Page 28 of 119

33 3 Management of Past Disposals The baseline Site Development Plan that underpinned BNFL s 2002 PCSC [32] assumed that all past disposals of trench wastes would remain in situ, and that the risks from these wastes would be managed by post-closure engineering controls (cap, cut-off walls and drains). It was nevertheless acknowledged that estimated radiological impacts from wastes originally disposed to the trenches exceeded regulatory targets for a modern standards facility. The calculated conditional risks from the trench wastes were dominated by the disposed inventory of uranium isotopes for the groundwater pathway (beyond a few tens of thousands of years), and thorium in the case of hazards associated with natural disruption and human intrusion [23]. Potential risk mitigation measures, including a combination of engineering barriers and chemical conditioning of the trench wastes to enhance sorption control over uranium release, were recognised as having a potential role to play in improving containment performance [35], but no formal analysis of options was carried out. The Environment Agency s assessment of the 2002 PCSC [23] and its corresponding authorisation decision [33] highlighted concerns that, particularly in view of the magnitude if the risks calculated for the trench wastes, insufficient consideration had been given to optimisation and risk management. The Environment Agency interpreted the safety analysis results as suggesting that the peak risks were significantly affected by a small number of disposals of long-lived wastes that had been emplaced in the LLWR prior to Among other options, it was recommended that further work should be undertaken to consider the possibility of selectively removing those wastes in order to optimise the performance of the site as a whole. This was eventually incorporated as part of the Schedule 9 requirements for improvement and additional information, attached to the subsequent Authorisation under the Radioactive Substances Act (now Permit under the Environmental Permitting Regulations) [36]. Prior to the Environment Agency s assessment and authorisation decision, published in 2005 and 2006, BNFL commissioned an initial set of options studies with the aim of understanding better the contributions to the highest risk estimates presented in the 2002 PCSC and potential options for their control [37,38]. The stated aim of these studies was to identify whether any management strategies could be identified that offered significant potential for overall reduction of radiological impact from the facility, both for past disposals and potential future disposals. The conclusions drawn from these studies in relation to the management of past disposals are summarised in Subsection 3.1, below. However, it is relevant to note that these studies were reviewed by the Environment Agency too, and found to be inadequate [39]. LLW Repository Ltd subsequently submitted a set of responses to the Environment Agency s Schedule 9 requirements, accompanied by a range of supporting studies. These included consideration of both waste retrieval and in situ remediation as options for the management of hazards associated with past disposals to the trenches (see Subsection 3.2). In its overall review, the Environment Agency indicated that further work would be required in order to make a systematic analysis and comparison of options in support of a comprehensive ESC [28]. More detailed studies had already been commissioned, and these were extended and completed as part of the programme of work described in Subsections 3.3 and 3.4 below. LLWR/ESC/(R11)/10025 Page 29 of 119

34 3.1 Initial Options Assessments BNFL s 2004 management option studies [37,38] sought to provide a more systematic analysis than had been undertaken for the 2002 PCSC of the possible radiological performance benefits of alternative management options for the future management of the LLWR. The studies were centred on two expert multi-criteria decision analysis workshops and took into account revised (lower) estimates of baseline radiological impact that addressed the implications of what had been identified by BNFL as cautious assumptions in the 2002 PCSC. Judgments made on options that specifically relate to the control of hazards associated with past disposals are summarised here. Other options considered in these studies that related to management of the facility as a whole (e.g. post-closure management of the site or closure engineering options) are discussed elsewhere in this report. The provisional conclusion of the assessments was that the baseline option (i.e. as described in the 2002 Site Development Plan [32], but taking into account revised estimates of radiological impact) was an appropriate solution for the optimum management of the LLWR. Nevertheless, two management options were identified as showing significant potential for overall reduction of the radiological impact [38], and further work was recommended to improve understanding of a range of options. The two options considered to have the greatest potential for overall reduction of impact were: site closure and retrieval of all disposed wastes for disposal elsewhere. It was judged that the residual hazards following retrieval would be limited to those from contamination that had leached into the ground around the LLWR, prior to or during the excavation and retrieval process. Some environmental releases would be likely to occur during retrieval operations, but it was considered that these could be kept to a minimum by appropriate working practices on the site. It was also noted that impacts would inevitably be associated with disposal elsewhere, but these could not readily be quantified in the absence of a candidate location. Moreover, the exposure of site workers during retrieval operations, disturbance of local community associated with such operations and subsequent transport to an alternative location were identified as relevant negative considerations. The study recommended that possible retrieval strategies should be developed and analysed in further assessments; encouraging dilution and dispersion of trench wastes. Taking measures to increase the rate of infiltration through the wastes, in order to promote the earlier release of some radionuclides, was identified as having the potential to reduce significantly the impacts associated with site disruption from natural events and possible future human actions, as well as gas pathway exposures. However, impacts associated with the groundwater pathway in the shorter-term would increase, while corresponding longer-term risks were judged to be largely unaffected. More significantly, it was acknowledged that such action would be manifestly inconsistent with the principle of waste isolation and containment, embodied in regulatory policy and in the design of the LLWR vaults. Despite its potential as a risk reduction measure, the study concluded that it was not appropriate to pursue this management option further. A number of other management options were also examined in the 2004 studies. None was considered to have comparable potential for reduction of impact as those summarised above to contribute to reduction of detriment associated with the trench LLWR/ESC/(R11)/10025 Page 30 of 119

35 wastes, although it was acknowledged that further investigation might nevertheless be merited in some cases. Key observations from the study included: implementation of a chemically active barrier to control the transport of leachate released from trenches was rejected on the basis that its useful life would be far too short to contribute to a reduction in the release of more significant radionuclides to the wider environment; installation of an impermeable barrier beneath Trenches 1 to 3 (e.g. by jet grouting) was considered as an option to improve the control of leachate for an area of the site where the natural low-permeability clay was judged to be reduced in efficiency or absent altogether. Although recognised as having the potential to improve engineering performance, uncertainty was expressed regarding the impact on risks, since it was judged that the intervention might introduce a greater potential for near surface releases; in situ grouting of the trenches was judged to have the potential to reduce the permeability of the wastes, blocking leachate flow paths and helping to reduce the magnitude of waste settlement. Additional benefits might also be expected through alkaline buffering that would reduce rates of release of a number of significant radionuclides. However, it was also suggested that improved containment within the trenches was likely to increase slightly the risks associated with the gas pathway, disruption by erosion and future human actions. Overall, the potential for improvement over the baseline option in terms of radiological impacts from groundwater pathway release was not considered sufficient to justify the wider impacts associated with its implementation; in situ vitrification was identified as an alternative approach to grouting for the production of a monolithic, impermeable wasteform within the trenches. However, concerns were identified regarding the potential mobilisation and release of radionuclides during vitrification, as well as possible increases in the long-term mobility of uranium. Doubts were also expressed regarding the technical effectiveness of employing the technique at the LLWR. The overall judgment was that implementation would be highly complex, demanding on resources, with potential for significant local disturbance and cost, while achieving only very limited reduction in the site risk/impact; compaction of trench waste was considered as an option in view of the perceived need to improve confidence in cap performance, taking into account the implications of waste settlement. It was also judged that such an intervention might also help to block leachate flow pathways through the waste. Any improved containment performance (potentially reducing groundwater pathway risks) would, however, be at the expense of associated slight increases in risks associated with the gas pathway, termination events and future human actions. On balance, the conclusion was drawn that there was too much uncertainty regarding the balance in overall efficacy (in terms of radiological detriment) to justify selection as a preferred management option; a variant on the retrieval of all disposed wastes, considered in the assessment, was to excavate the trench wastes for packaging and disposal in new vaults on the LLWR site. This was judged to be capable of reducing significantly the long-term impacts associated with the groundwater release pathway; however, the assumed improvement in containment was also judged to give rise to a corresponding increase in long-term doses for the gas release pathway and site disruption. Given the disturbance, exposures of site workers, and potential for LLWR/ESC/(R11)/10025 Page 31 of 119

36 environmental release during retrieval operations, coupled with the limited improvement in assessed risks, the option was not considered an appropriate management solution. Hence, from the perspective of optimising the control of risks to the public, key radionuclides contributing to the long-lived component of the inventory in the trenches were identified as the effective determining factor in decision making, irrespective of potential improvements in their containment. This is because reductions in estimated risks associated with radionuclides contributing to the detriment from groundwater discharges in the medium-term and long-term (principally from Ra-226 and U-234) were judged to be accompanied by corresponding increases in risks associated with gas release (also from Ra-226 and U-234), disruption by erosion and future human actions (Th-232). The lack of a clear outcome in terms of improvement in overall radiological risk, coupled with other potential detriments associated with implementing such interventions (e.g. cost, operational safety, discharges, disturbance, etc), was a significant factor behind identification of the 2002 Site Development Plan [32] as the preferred option. The only management options identified in the 2004 studies as having clear potential to reduce risks associated with past disposals involved the removal of inventory from the site, either by enhancing the rate at which they are released by natural processes or by retrieval to an alternative location (see above). In this respect, and given the significance of a comparably small number of radionuclides to the determination of estimated radiological impacts in excess of the risk guidance level, it is notable that no targeted actions (i.e. more limited interventions focused on specific radionuclides or waste disposals) were identified among the candidate management options. 3.2 Schedule 9 Studies Following the Environment Agency s authorisation of continuing operations at the LLWR [36], LLW Repository Ltd undertook and published a series of studies in response to the associated Schedule 9 requirements. Among these was a wideranging, peer-reviewed assessment of options for reducing future impacts from the LLWR [40], largely focused on the control of hazards associated with disposals to the trenches. The assessment was implemented through a series of expert and stakeholder workshops, drawing on a new review of national and international best practice in controls relevant to repository design and management [41]. A key difference from the 2004 studies was that the appraisal of options sought to focus on the decision logic (i.e. the underlying reasons for determining preferences between options), rather than a quantitative multi-criteria comparison of options. A tool kit of candidate technologies was assembled, taking into account the feasibility of implementation and likelihood of contributing to a strategy that could achieve benefit in terms of a reduction in impacts. These were then identified with broad strategic management options judged to have the potential to mitigate radiological impacts. Strategic options included: repository cap optimisation; engineered barriers; coastal defences; vertical drains; LLWR/ESC/(R11)/10025 Page 32 of 119

37 bulk or local retrieval ; institutional control optimisation; in-situ remediation. Several of these factors (cap optimisation, engineering barriers, coastal defences and institutional control) are considered later in this document as part of the wider optimisation of management of the facility as a whole. Nevertheless, key outcomes are discussed here in so far as they relate to the control of impacts from trench wastes. In addition to the best practice review of technology and management options, the assessment and identification of a preferred management strategy was guided by inputs to the 2004 studies [37,38], as well as the most recent update to the long-term safety analysis for the LLWR [27]. Apart from radiological safety, a variety of other performance attributes (consistent with those used in previous stakeholder engagement on site end states [42]) were also taken into account in examining potential differentiating factors between options. The study concluded that active institutional control would be a necessary element in the management strategy for the site, but no specific recommendations were developed for how this should be implemented in practice (e.g. with respect to the nature of controls put in place, or their duration). Instead, it was noted that a formal analysis of options would be carried out in support of the current ESC (Subsection 6.2). An engineered final repository cap, optimised with respect to protection against the likelihood of human intrusion, limitation of infiltration and control of gas release, was also identified as an essential element in the management of trench inventory wastes. Finally, it was concluded that engineered barriers and drains should play a key role in future strategy for the facility. No attempt was made to develop a comprehensive optimisation of the engineering design, although it was acknowledged that a formal analysis to define the preferred strategy would be required to underpin the current ESC (see Section 5). Nevertheless, the only barriers considered appropriate for application at LLWR were those relating to the passive control of the site in general (drains, vault base liners, cut-off walls). None of the barrier technology options specifically related to the possible mitigation of impacts from the trenches (i.e. active barriers and insertion of impermeable barriers, which had previously been dismissed in the 2004 studies (Subsection 3.1)) was selected in the study. Other possible components of an optimised long-term management strategy for waste disposal at LLWR were largely dismissed in the conclusions drawn from the options assessment. With respect to the hazards associated with past disposals, the study concluded that intervention to retrieve most, or all, of the wastes was unjustified in terms of the balance between the estimated legacy risks from the wastes remaining in situ and the costs and risks of implementing the remedial actions. LLWR/ESC/(R11)/10025 Page 33 of 119

38 In particular, it was judged that the re-disposal of retrieved waste (whether on site or elsewhere) would need to offer substantially improved long-term safety performance for such a strategy to be worthwhile 3. Nevertheless, it was proposed that the option of selective retrieval of certain radionuclides from the trenches should be kept under review and would be considered further in development of the current ESC (see Subsection 3.3). Large-scale in situ remediation within the trenches was also largely dismissed as a potentially viable option. Grouting was judged to be technically difficult to implement on a large scale, requiring extensive disturbance and high cost for comparatively little performance improvement for the groundwater pathway, and no improvement at all for coastal erosion and human intrusion. Vitrification was assessed to be even more problematic in terms of technical viability, partly due to the low silica content of the wastes. Moreover, although vitrification would in principle offer significant potential to reduce contaminant releases, its application for the trenches as a whole would be at the expense of very high cost, energy use and potential environmental releases during implementation. Large-scale compaction of trench wastes was identified as a potentially viable component of the overall site management strategy, but would only be justified if there was otherwise concern regarding confidence in the long-term performance of the engineered cap. The possibility of localised implementation of remediation techniques was not comprehensively ruled out by the study; however, no formal analysis was made and no specific approach identified as a preferred option. More detailed examination of the potential for selective application of in situ remediation was therefore proposed as part of the current ESC (see Subsection 3.4). Finally, while recognising the threats to the site from sea-level rise and coastal erosion [8], it was judged that a safety case depending on the necessary maintenance and periodic rebuilding of coastal defences into the far future would be impossible to substantiate and would be inconsistent with the regulatory principle that unreasonable reliance on human actions to protect the public and the environment should be avoided [19]. The decision was therefore made to pursue the long-term safety assessment for the LLWR on the assumption that coastal defences would not be constructed. In its review of the options evaluation submitted by LLW Repository Ltd, the Environment Agency highlighted that it expected the review of risk reduction options to be taken to completion, by defining an optimised strategy for the site [43]. In particular, the regulatory review highlighted that the possibility of selective retrieval was a key question to explore further in support of the ESC. This aspect of optimisation and other components of the overall management strategy for the LLWR are addressed in the following sections of this document. 3.3 Targeted Retrievals Having ruled out the retrieval of all disposed wastes as a component of the management strategy, LLW Repository Ltd commissioned a detailed review of the potential for selective retrievals of key target wastes [44]. The summary that follows 3 A similar conclusion was reached in the NDA s more recent Strategic Environmental Assessment in support of UK Strategy for Management of LLW [26]. The NDA study argues that, so long as a safety case can be made for continued use of the LLWR (and subject to regulatory and planning approvals), any long-term benefits of waste retrieval in terms of hazard reduction would be outweighed by the detrimental impacts arising from the retrieval operations themselves (see Section 2.1). LLWR/ESC/(R11)/10025 Page 34 of 119

39 is based on detailed information provided in the full review report and other relevant studies, including data and results from the most recent assessment calculations Waste Heterogeneity and Long-term Radiological Significance The distributions of some radionuclides in past disposals to the LLWR trenches are notably heterogeneous [4,45,46]. Where such radionuclides also contribute significantly to the overall radiological hazard presented by those disposals, there is a potential opportunity to gain a benefit in terms of mitigation of detriment through retrieval or other remedial action implemented on a comparatively localised scale compared with that of the facility as a whole. For example, localised high concentrations of key radionuclides could in principle have a significant effect on the average concentration (and associated doses and risks) associated with gas and groundwater release pathways. Similarly, in the case of localised exposure cases (e.g. human intrusion, coastal erosion) a significantly higher than average concentration of a particular radionuclide at a specific location could give rise to estimated conditional doses that are much greater than the mean value for the facility as a whole. Potential opportunities for targeted retrieval were identified by first examining the results of preliminary assessments of long-term safety performance (expressed in terms of calculated contributions from key radionuclides and release pathways [27]). The calculated doses for the trench waste inventory only, shown in Table 3.1, are based on estimated average concentrations across the trenches. Table 3.1 Key radionuclides in trench wastes and their estimated contributions to maximum dose (µsv y -1 ) for different release pathways (from [27]) Disposed radionuclide Groundwater (well water) Gas Release Coastal Erosion Human Intrusion Intruder (a) Occupier Dwelling C Cl-36 7 (early) Tc-99 2 (early) I (early) U-isotopes 3 (late) Ra Th Pu (a) Calculated doses for the intruder pathway are expressed in µsv per event. The estimated radiological detriments derived in the preliminary assessment [27] differ in detail from those determined for the current ESC, but the general conclusions are broadly similar. In particular, for each release pathway, the same (few) radionuclides from the overall LLWR waste inventory tend to dominate the LLWR/ESC/(R11)/10025 Page 35 of 119

40 contributions to the estimated radiological impacts from disposals to the trenches, particularly in the base case calculations. In what follows, the emphasis is on the analysis undertaken to support an examination of selective retrieval options [44], which was based on earlier assessment results. Appropriate comparisons have been made, however, with the outcomes of the latest assessments [12]. Groundwater pathway Calculated peak doses for the groundwater pathway (occurring in the period prior to a few thousand years after disposal, identified as early in Table 3.1) relate to the assumed extensive use of well water by agricultural smallholders. By comparison, in both the earlier and more recent assessment calculations the natural discharge of contaminated groundwater at the coast (assuming no well water abstraction takes place) is estimated to give rise to much lower levels of potential exposure. Maximum doses for exposure to contaminated well water in the earlier calculations [27] correspond to an equivalent individual risk of approximately y -1. By contrast, the base case results for the same exposure pathway in the current assessment [12] indicate a lower peak individual risk (for the trenches and vaults combined) of approximately y -1 at 2250 AD; this takes account of the annual probability of a water abstraction well between the disposal area and the coast. Both sets of results, but more so those from the current ESC, indicate a measure of confidence that groundwater pathway risks from trench disposals (in the absence of remedial action) are no greater than the regulatory risk guidance level [19]. Such understanding diminishes the attractiveness of taking remedial action to control risks for this pathway, as part of an ALARA process, unless they are readily implementable at limited cost or other associated detriment. It is also notable that the estimated disposals of both Cl-36 and Tc-99 to Vault 8 (and their estimated radiological impacts [27]) are both significantly higher than those to the trenches, while the projected disposals of both radionuclides to Vault 9 and future vaults are greater still [4]. Disposal records indicate that the trench inventories of the principal radionuclides for this pathway are quite evenly distributed, although approximately 20% of the Tc-99 is localised in adjacent bays (39 to 40 and 66 to 70) of Trench 7 [45] (see Figure 3.1). Hence, although a few areas can be identified where substantially greater than average amounts of radiologically important radionuclides have been disposed within the trenches, the significance of these areas as contributors to overall risk for the groundwater pathway is not high. Efforts to undertake selective retrieval would therefore be likely to be grossly disproportionate to any benefits they achieved. LLWR/ESC/(R11)/10025 Page 36 of 119

41 Figure 3.1 Heterogeneity plot for Tc-99 disposal to the trenches (from [47]) Calculated peak doses for the groundwater pathway in the longer term (occurring in the period several thousand years or more after disposal, identified as late in Table 3.1) also relate to the assumed use of well water by agricultural smallholders. For this period, the preliminary calculations [27] showed that isotopes of uranium dominate the overall detriment from trench disposals, although the corresponding equivalent individual risk (approximately y -1 ) remains less than the regulatory risk guidance level. More recent calculations undertaken for the ESC [12] indicate that peak individual risks for this exposure route (summed over all radionuclides from both the trenches and vaults) could rise to about y -1 at about 5000 AD, but only if it is assumed there is no disruption of the facility by coastal erosion (which is LLWR/ESC/(R11)/10025 Page 37 of 119

42 considered highly unlikely according to our current understanding of future sea-level change and coastal processes [8]). Uranium disposals to the trenches are heterogeneous, with approximately 55% being contained within Trenches 4 and 5 [4,45]. The majority of this fraction of the inventory is distributed over more than half the total area of Trench 4 and between one third and a half of Trench 5. Estimated uranium disposals to the trenches are significantly higher than those to Vault 8, and are not exceeded by the projected disposals to future vaults [4]. In principle, therefore, selective retrieval of uranium from a comparatively small area of the trenches (approximately half of Trenches 4 and 5) would have an appreciable (approximately a factor of two) impact on long-term risks for the groundwater pathway. However, radionuclides in the uranium decay series contribute only a very small fraction of the peak risk from trench wastes for the groundwater pathway (well abstraction) in the short term and even in the longer term, should the site not be eroded, the assessed risk is less than regulatory risk guidance level [12]. Efforts to undertake selective retrieval of uranium are therefore judged to be inappropriate in view of the limited reduction in risk that is likely to be achieved, and that the assessed risk is in any case less than regulatory risk guidance level. Gas pathway Calculated doses for the gas release pathway are associated with C-14 and Rn-222 daughters (from the decay of Ra-226). If the vent system is closed before the end of active site management, then Rn-222 releases from the waste will be effectively zero; if the vent remains open or the engineered cap is damaged increasing its gas permeability, then assessment calculations have shown that the deep profiling over the trenches is sufficient to limit the amounts of Rn-222 that could reach the gas collection layer and thence be released to the atmosphere to very low levels [12]. Thus, the gaseous release of C-14 is of most concern. Assessed doses associated with C-14 release from the trenches were relatively low in the preliminary assessment (see Table 3.1) that informed the targeted retrieval options study [44], but have increased in assessments for the current ESC, mainly due to a much more cautious model of C-14 labelled gas release from the near field [12]. In the current ESC, calculated doses from C-14 labelled gas release from the trenches are about 15 µsv y -1 at 2180 AD (i.e. 100 years after closure engineering). These calculations assume, cautiously, that a self-sufficient agricultural smallholding is established on the area above the trenches. At 2380 AD (i.e. 300 years after closure engineering), the calculated dose from C-14 labelled gas release from the trenches has reduced to about 5 µsv y -1. These doses are, however, less than those estimated for the vaults, since the C-14 inventory of the trenches (0.1 TBq) is relatively small compared with that estimated for the vaults (5.5 TBq). There is substantial heterogeneity in disposals of Ra-226 (which contribute to Rn-222 release in the shorter-term), with a large proportion (about 95%) of the total trench inventory associated with a comparatively small fraction of Trenches 2 and 3 [4,45]. This inventory is associated with the disposal in the early 1970s of process residues with a high content of naturally-occurring radioactive materials. However, consideration of the gas pathway does not justify retrieval, because the estimated contribution from Rn-222 to overall detriment (within the expected evolution of the site) is very small [12]. LLWR/ESC/(R11)/10025 Page 38 of 119

43 Disposals of C-14 are widely distributed within the trenches, and hence there is no opportunity for targeted retrieval to have a significant mitigating effect on overall impacts. Moreover, the estimated dose from gaseous release associated with the inventory of C-14 in the trenches is much less than that associated with the vaults and consistent with the risk guidance level. Efforts to undertake retrieval of C-14 from the trenches would therefore be likely to be grossly disproportionate to any overall benefits they achieved. Coastal erosion The projected radiological impacts from disruption by coastal erosion are among the most significant aspects of the overall long-term safety analysis for the LLWR [12]. In the preliminary assessment calculations that informed the optimisation studies [27] attention was focused on an individual making use of the beach below the eroding facility for recreational purposes. In the assessment of coastal erosion in support of the current ESC, doses have been calculated for the local recreational beach user, plus to an occupational user of the coast between St Bees and Ravenglass, and to a consumer of marine foodstuffs harvested from coastal waters [12]. Here we focus on the calculated dose to the local recreational beach user because this gives the most immediate opportunity for exposure to eroding waste, including experiencing the effect of heterogeneous distribution of radionuclides in the eroding waste. In both the preliminary and more recent assessments, the dominant contribution to potential exposure to the recreational beach user is from direct external irradiation, with Th-228 and Ra-228 being the most important radionuclides. Both of these arise from the disposal of Th-232 (Table 3.1). Secondary contributions to external irradiation exposure are associated with the short-lived daughters of disposed Ra-226. Inhalation of airborne particulates also contributes to the total estimated radiological impact, but to a lesser degree. The primary contributors to estimated inhalation doses associated with coastal erosion are Pu-239, Th-232 and its daughters, and isotopes of uranium. Taken together, however, inhalation contributes only about 1% of the overall estimated doses from trench wastes associated with coastal erosion [12,27]. The calculated doses arising from coastal erosion in the preliminary assessment studies [27] (see also Table 3.1) were based on mean concentrations of the relevant radionuclides across the trenches as a whole, whereas those carried out for the current ESC take account of heterogeneity of radionuclide concentrations within the waste [12]. In the latter case, the calculated maximum dose of 19 µsv y -1 for the reference case (corresponding to a conditional risk of approximately 10-6 y -1 ) relates to exposures at the time that Trenches 4 and 5 (see discussion below) are being eroded. With the disposed inventory of Th-232 representing by far the most important contribution to calculated radiological impacts associated with coastal erosion of the trench wastes, the heterogeneity of its distribution within the LLWR is of particular interest. Trench disposals of Th-232 are significantly higher than those to Vault 8, and approximately 80% of the total LLWR inventory, even when projected disposals to future vaults are taken into account [4]. Moreover, disposal records indicate that the majority of the Th-232 activity in the trench inventory is located in Trenches 2, 4 and 5, associated with specific disposals of mineral sands and process residues [4, 45,48]. The greater part of disposals to Trench 2 (naturally-occurring monazite and thorite sands, representing approximately one third of the total inventory) were to a LLWR/ESC/(R11)/10025 Page 39 of 119

44 small number of bays (Figure 3.2). Areas of significantly higher than average activity concentration are also apparent in the records for disposal to Trenches 4 and 5 associated with process residues. Figure 3.2 Heterogeneity plot for Th-232 disposal to the trenches (from [47]) The most important contribution to the total estimated Ra-226 inventory within the trenches is from the disposal of mineral processing wastes to Trench 3 (Figure 3.3). Significant disposals of Ra-228 were also made to this area [45], but its comparatively short half-life means that it makes little contribution to the estimated doses at the time coastal erosion is assumed to take place. LLWR/ESC/(R11)/10025 Page 40 of 119

45 Figure 3.3 Heterogeneity plot for Ra-226 disposal to the trenches (from [47]) Disposal records indicate that some 70% of the total disposals of plutonium to the trenches were to two fairly small regions (bays and 64-70) of Trench 2 [48] (see Figure 3.4). These wastes consist of plutonium-contaminated materials mixed with non-contaminated soils within drums. As noted previously, the enhanced concentration of a specific radionuclide at a particular location means that doses (in this case via inhalation) could be higher at the time when these particular wastes are being eroded. Examination of the results from the current assessment shows, however, that because the exposure of a person using the beach depends on the average concentration of radionuclides across all wastes that are being eroded at given time, the inhalation dose from these plutonium-bearing wastes is very small compared with the external irradiation dose from Th-232-bearing wastes that are LLWR/ESC/(R11)/10025 Page 41 of 119

46 assumed to be being eroded at the same time [12]. The remainder of the disposed inventory of plutonium is generally homogeneously distributed, reflecting its presence as a minor contaminant in a range of operational wastes, although there is a less marked elevation above the average in Trench 4. Pu-239 contributes (via inhalation) only about 1% of the total calculated dose associated with the coastal erosion of trench wastes - a dose of about 0.2 µsv y -1 12], which corresponds to a conditional risk of 10-8 y -1. This indicates that efforts to retrieve of plutonium bearing wastes from the trenches would be likely to be grossly disproportionate to any benefit they achieved. Figure 3.4 Heterogeneity plot for total plutonium disposal to the trenches (from [47]) LLWR/ESC/(R11)/10025 Page 42 of 119

47 The disposals of uranium to the trenches are heterogeneous, with approximately 45% being contained within Trenches 4 and 5 [47]. However, even though uranium disposals to the trenches are significantly higher than those to Vault 8, and are not exceeded by the projected disposals to future vaults [4], U-234 and U-238 contribute only a small fraction the total inhalation dose in the case of coastal erosion, while total inhalation exposure contributes about 1% of the overall calculated dose. Targeted retrieval of uranium disposals would, therefore, not offer any significant reduction of radiological impact. Human intrusion Estimated doses associated with human intrusion cases are also a key consideration in the overall long-term safety analysis. Preliminary assessment calculations undertaken to inform the optimisation studies [27] indicated that Th-228 (from the disposal of Th-232) is the dominant contributor to external dose, which is the most important exposure pathway for the intruder and site occupier case (Table 3.1). These conclusions are supported by the more recent assessment calculations undertaken in support of the ESC [12], which also show the importance of Th-232 as a contributor to external doses. The calculated dose depends on assumptions regarding the scale of disruption arising from the intrusion, and the use of land contaminated by intrusion events. Whereas the dose estimates for the human intrusion cases undertaken in the preliminary assessment studies [27] were based on mean concentrations of the relevant radionuclides across the trenches as a whole those carried out for the current ESC [12] take account of heterogeneity of radionuclides within the disposal site, as well as the different levels of protection against intrusion afforded by the final cap and profiling material at different locations. The more recent calculations show that the doses to those that intrude into the waste are low, generally only a fraction of a msv, and substantially below a dose guidance level of 20 msv that would be appropriate for exposures of less than a year duration [19]. The key human intrusion case that gives rise to the highest calculated doses involves the assumption that a dwelling is constructed on excavated spoil (waste and cap materials) following excavation to waste depth. In this case, doses are calculated taking into account the release of radon into the building. The outcomes are sensitive to assumptions made in the assessment, but in this case doses are dominated by contributions from the disposal of Ra In the preliminary assessment (presented in Table 3.1) doses in the range 0.6 to 3.0 msv were calculated corresponding to the average concentration of radionuclides across the trenches. In the more recent assessments in support of the 2011 ESC [12], credit is taken for the depth of profiling over the trenches, which protects the trench waste from direct excavation except at certain locations around the cap margins. The heterogeneity of disposals of Th-232 and Ra-226 has been discussed previously (Figure 3.2 and Figure 3.3, respectively). More than 90% of the total trench inventory of Ra-226 is associated with a comparatively small fraction of Trenches 2 and 3 [4,45]. However, the thickness of the final cap and profiling material above these areas is expected to prevent excavations up to 5 m deep from intersecting the 4 In the preliminary assessments, a contribution from Rn-220 (thoron) from the decay of Th-228 (a daughter of Th-232) was also cautiously assumed. The evidence from levels of Rn-220 in dwellings in the United Kingdom show that this was a highly cautious assumption and that in practice ingress of Rn-220 to a dwelling from the ground on which it was built would be negligible. LLWR/ESC/(R11)/10025 Page 43 of 119

48 wastes in these areas. Indeed, there is no calculated dose to site occupants following construction activities above most of the trenches and the potential impact of uranium and radium disposals within these areas is not seen in the latest assessment calculations [12]. With the cap and profile material providing significant protection against the likelihood of incurring significant exposure as a result of intrusion, it is therefore questionable whether the targeted recovery of trench disposals of Ra-226 could be considered a proportionate contribution to optimisation. The role of heterogeneity in relation to decisions surrounding the potential retrieval of Th-232 is different to that for Ra-226. As noted earlier, trench disposals of Th-232 represent a significant proportion of the total LLWR inventory, even when projected disposals to future vaults are taken into account [4]. A significant fraction is localised to a limited number of bays in Trenches 2, 4 and 5. This includes the northern end of Trench 2, near the perimeter of the disposal facility, where the protection against disruption afforded by cap profiling material is at a minimum. Targeted retrieval of Th-232 to mitigate the potential radiological consequences of coastal erosion, an event that is considered to have a high likelihood of occurrence within a period of a few hundred to a few thousand years in the future, would therefore also serve to reduce the impacts of several (less likely) human intrusion cases. Summary Consideration of waste heterogeneity alongside the radiological significance of specific radionuclides leads us to the following conclusions in relation to opportunities for targeted retrieval. Whilst retrieval has the potential to influence releases of radionuclides from the LLWR via the groundwater pathway, the scope for targeted actions to achieve an effective mitigation of overall impacts is small. Although a few areas can be identified where substantially greater than average amounts of radiologically important radionuclides (particularly Tc-99) have been disposed within the trenches, the significance of these disposals as contributors to risk for the groundwater pathway is not high. Radionuclides in the uranium decay series contribute only a very small fraction of the peak dose from trench wastes for well water abstraction in the short term and do not represent a high risk by comparison with the regulatory risk guidance level. Efforts to undertake selective retrieval would therefore be likely to be grossly disproportionate to any benefits they achieved. In relation to estimated exposures associated with the groundwater pathway in the very long term, uranium disposals to the trenches are significantly higher than those to Vault 8, and are not exceeded by the projected disposals to future vaults. It might therefore be argued that the selective retrieval of uranium from a comparatively small area of the trenches (roughly half of Trenches 4 and 5) could have an appreciable (approximately a factor of two) impact on long-term risks for the groundwater pathway. However, given the long timescales on which such estimated exposures are calculated to take place, and the fact that calculated doses associated with the trench wastes do not in any case represent a high risk by comparison the regulatory risk guidance level, efforts to undertake selective retrieval of uranium are judged to be inappropriate in view of the limited reduction in detriment that is likely to be achieved. With respect to potential releases via the gas pathway, C-14 is the dominant contributor. However, disposals of C-14 are widely distributed within the trenches, and hence there is no opportunity for targeted retrieval to have a LLWR/ESC/(R11)/10025 Page 44 of 119

49 significant mitigating effect on overall impacts. Moreover, the estimated dose from gaseous release associated with the inventory of C-14 in the trenches is much less than that associated with the vaults and consistent with the risk guidance level. There is substantial heterogeneity in past disposals of Ra-226. However, the potential releases of Rn-222 are very small, because the engineered cap and profiling material will form an effective barrier to radon migration from the wastes. Consideration of the gas pathway does not therefore justify the selective retrieval Ra-226 disposals. The disposed inventory of Th-232 represents the most important contribution to calculated radiological impacts associated with coastal erosion of the trenches. Trench disposals of Th-232 are significantly higher than those to Vault 8, and approximately 80% of the total LLWR inventory, even when projected disposals to future vaults are taken into account. A significant fraction of the total inventory is localised to a limited number of bays in Trenches 2, 4 and 5. Although projected doses and conditional risks associated with exposure to eroded wastes are not particularly high by comparison with the regulatory risk target, the feasibility of targeted retrieval of these wastes is a relevant consideration in the potential mitigation of risk associated with one of the more significant environmental safety issues for the LLWR. This is therefore considered further in Subsection 3.5, below. Other radionuclides whose disposal contributes to exposures associated with coastal erosion (isotopes of plutonium, uranium, radium) are also distributed heterogeneously, to a greater or lesser extent, within the trenches. However, the contribution that targeted retrieval would make to overall radiological impacts associated with coastal erosion is comparatively small. Examining the potential for such action to contribute to optimisation in the control of radiological risk associated with coastal erosion is therefore of lower significance than for Th-232 disposals. With respect to the possible impacts from disruption of the site by human intrusion, the exposure case that gives rise to the highest estimates of conditional dose involves the assumption that a dwelling is constructed on excavated spoil (waste and cap materials). Estimated consequences are sensitive to assumptions made in the assessment, including the level of protection against intrusion afforded by the final cap and profiling material at different locations over the trenches, but are dominated by contributions from the disposal of Ra-226. In addition, a possible source of external exposure is the disposal of mineral sands at northern end of Trench 2, near the perimeter of the disposal facility, where the protection against disruption afforded by cap profiling material is at a minimum. Taking account of the protection offered by the cap and thick profiling over the trenches, selective retrieval of Ra-226 bearing waste would offer limited benefit. The possibility of Th-232 retrieval is identified as a relevant consideration in relation to coastal erosion, which is expected to occur within a period of a few hundred to a few thousand years in the future. Coastal erosion is the expected evolution of the facility; disruption by human intrusion over such a time period is less certain (although must be assessed on the basis that it occurs). The targeted retrieval of Th-232 to mitigate the potential radiological consequences of coastal erosion could, however, also serve to reduce the impacts associated with inadvertent human intrusion. LLWR/ESC/(R11)/10025 Page 45 of 119

50 3.3.2 Waste Heterogeneity Implications for the Period of Authorisation The emphasis throughout the above discussion has been on the underlying potential for targeted retrieval to play a significant role in the mitigation of long-term risks presented by past disposals. Another relevant consideration is the control of discharges during the PoA. The detection of tritium in groundwater was a relevant factor in the decision to install an interim cap and partial COW for the trenches (see Subsection 2.2.2). Whilst there is expected to be substantial decay of the disposed tritium inventory prior to end of the PoA (and its projected contribution to long-term risks is low), the possibility of taking additional remedial action is a relevant consideration as part of an optimised strategy to minimise releases in the shorter term. In relation to the possible targeted retrieval of wastes containing tritium, it is notable that the inventory within the trenches is dominated (approximately 90%) by the disposal of small volumes of luminised telephone dials in bays 15 to18 and bay 23 of Trench 6 [48]. Although the disposals are highly localised, they are not necessarily readily detectable and efforts to achieve targeted retrieval might therefore entail the disturbance of substantially greater volumes than are associated with the wastes themselves. Disturbance of the wastes could also potentially lead to uncontrolled releases of tritium during retrieval. The current impacts associated with tritium disposal to the trenches are very small (1 to a few µsv y -1 ) compared with source related dose constraint of 0.3 msv y -1 that applies to the PoA [11]. It is not known how much of the original inventory of tritium has already migrated from the original location at which it was disposed. However, monitoring of radioactive contaminants in groundwater [9] shows that the concentrations of LLWR-derived tritium in groundwater around the site have decreased markedly since the installation of the temporary trench cap and the COW. Monitoring is being continued to ensure that the performance of existing barriers is well understood, and to ensure that any increases in discharge can be detected (Subsection 6.1.2). There is no evidence that tritiated groundwater releases are causing any harm at the moment. Furthermore, the land between the disposal area and the coast is currently an SSSI, so that there is unlikely to be any development that would include a water abstraction well that might intersect tritium contaminated groundwater. Nevertheless, the possibility that tritium contaminated groundwater could be abstracted from an offsite borehole (assumed to be located between the site and the coast) is examined in assessments both for the PoA [11] and beyond [12]. These indicate that the potential dose from tritium via this hypothetical pathway (if abstraction were occurring now) would be a maximum of 3 µsv y -1, falling by more than an order of magnitude by Existing control measures (in particular the interim trench cap and COW) are therefore judged to be working effectively, consistent with ensuring radiological impacts are ALARA. The efficacy of such measures will remain under review during active management of the facility, while it remains under regulatory control (Subsection 6.1.2). Targeted retrieval of tritium sources from the trenches is in principle a remedial action that might be taken, but our best understanding is that such action and the associated detriments would be grossly disproportionate to any overall benefits achieved from the perspective of optimisation. LLWR/ESC/(R11)/10025 Page 46 of 119

51 3.3.3 Evaluation of Retrieval Options Based on the above analysis of waste heterogeneity and its potential significance, three illustrative retrieval options were formulated in order to examine the extent to which such a targeted strategy might reasonably contribute towards optimised management of the LLWR [44]. In each case it was assumed that the total volume of material excavated was dictated by removal of the relevant source one bay at a time, starting from the highest concentrations, until the residual average concentration in that trench was equivalent to, or less than, that in the next worst trench. It was further assumed that bulk excavation would take place and the materials sorted to enable the target wastes to be segregated, for disposal or long-term storage elsewhere, with the remainder being retained on the LLWR site. The identified targeted retrieval options were: option 1. Retrieval of thorium-bearing mineral sands (monazite and thorite) from two discrete locations in Trench 2. This would involve the excavation of seven bays in total (7000 m 3 waste), of which some 27% (1740 m 3 ) is target waste. The estimated implementation time is two to three years. Based on the preliminary assessment results available at the time [27], this was calculated to reduce the maximum individual dose to a recreational beach user during coastal erosion (i.e. when Trench 2 is exposed at the cliff face) by a factor of ten and the average risk (based on the mean concentration of Th-232 across all trenches) by approximately one third; option 2. Retrieval of radium-bearing process wastes from relevant regions of Trenches 2 and 3. This would involve the excavation of 30 bays in total (36,800 m 3 waste), of which around 2% (768 m 3 ) is target waste. The estimated implementation time is eight years. Based on preliminary assessment results [27], this was estimated to reduce the maximum individual dose for human intrusion (i.e. when a dwelling is assumed to be constructed directly above the relevant trench wastes) by approximately a factor of about ten and the average dose (across all trenches) by a factor of five 5 ; option 3. Retrieval of thorium-bearing wastes from more widely dispersed areas of Trenches 4 and 5. This would involve the excavation of 54 bays in total (82,200 m 3 waste), of which some 3.3% (2730 m 3 ) is target waste. The estimated implementation time is approximately 14 years. Combined with Option 1, the preliminary assessment calculations suggested that this could reduce the assessed average risk to a recreational beach user across all trenches by approximately two thirds. There are uncertainties regarding the precise location and amount of target wastes that would need to be removed. In practice, such uncertainties are likely to lead either to increased volumes of excavation (in order to ensure that the target wastes are effectively retrieved) or reduced benefits (if some of the target wastes are not recovered). The estimates of excavated waste volumes for each retrieval option also 5 Subsequent assessments suggest that the thickness of the final cap and profiling material above the trenches is sufficient to protect trench wastes from significant excavations [12], and thus the case of building on excavated spoil is not a high concern for the trenches (see Subsection 3.3.1). Assuming that the cap and profile material provide such protection against disturbance of the wastes, it is questionable whether the targeted recovery of trench disposals of Ra-226 can be considered a proportionate contribution to optimisation. LLWR/ESC/(R11)/10025 Page 47 of 119

52 exclude the removal of material associated with the interim cap. In each case, the total volume of capping materials to be removed would be larger than (but less than double) the total volume of excavated wastes. There are also uncertainties surrounding how each option would be planned and implemented in practice, including uncertainties in the sorting efficiency that determine the overall amount of waste requiring forward management as well as the waste routes themselves. Each of these factors would have implications for timescales and costs. The practicability and wider implications of implementing each option were assessed based on the assumption that current available technologies would be deployed [44]. Safety considerations include radiological and chemo-toxic exposures of the workforce, as well as conventional hazards associated with the operations themselves. Relevant factors include potential exposures to radon and thoron, leachate management and bulk gas control. Consideration was also given to environmental impacts and off-site nuisance (e.g. transport, noise and visual intrusion) as well as the cost of implementing each option, including the cost of forward management of both the target and background wastes. Summary comparisons showing the calculated implications of different options (and combinations of options) for estimated impacts associated with coastal erosion and human intrusion are shown in Figure 3.5 and Figure 3.6, respectively. The charts also provide notional estimates of the total programme cost and collective dose to site workers associated with each case. Figure 3.5 Trench by trench and overall mitigation of impacts from coastal erosion compared with costs and operator exposures (from [44]) LLWR/ESC/(R11)/10025 Page 48 of 119

53 Figure 3.6 Trench by trench and overall mitigation of impacts from human intrusion compared with costs and operator exposures (from [44]) It is important to note that the dose estimates in these results relate to the results of preliminary assessments of long-term safety performance [27], which were available at the time the study was undertaken, rather than the calculations undertaken for the current ESC [12]. The coastal erosion results can be considered to hold good, since results in the more recent assessment relate to the same exposure pathways and differences are relatively minor. We no longer place such importance on the case of human intrusion into the trenches because we take credit for the thick cover over the trenches. Nevertheless, the following broad conclusions can be drawn from these and other aspects of the review. The total cost of the smaller actions (Options 1 and 2) is estimated to be in the region of 150 million, with corresponding estimated collective worker doses in the region of 0.35 Man Sv. Option 3 is worth considering only in combination with Option 1, and (because it is intrinsically much less efficient in removal of the hazard at source) would increase the total cost and collective dose by around a factor of three to four. The scale of works associated with even the smallest of the targeted retrieval options (Option 1) would be larger and more complex than any comparable programme so far attempted in the UK (e.g. at the Harwell Southern Storage Area) and broadly on a par with the largest waste retrieval programmes undertaken elsewhere in Europe (France and Hungary) [49]. Nevertheless, retrieval operations on a similar scale have also been undertaken in the USA (e.g. at Hanford and Oak Ridge National Laboratory), which gives some confidence that the technical requirements are within current industrial capability and knowledge. LLWR/ESC/(R11)/10025 Page 49 of 119

54 The wider consequences of undertaking retrieval, particularly in terms of local nuisance, are considerable. The scale of operations incorporates not only the excavations themselves but also an extensive infrastructure in support of excavation, sorting, segregation and re-packaging, stabilisation and backfilling. Although such factors are not necessarily overwhelming and their impacts can be mitigated by appropriate management, retrievals from Trenches 2 and 3, in particular, might well have implications for ongoing operations in the adjacent vaults and would significantly delay final capping. Targeted retrieval of Th-232 could reduce considerably the maximum individual dose associated with coastal erosion of an individual trench. However, retrieval from Trench 2 alone (Option 1) would have only a modest effect on the average risk (about 30% reduction), not least because of the significant remaining contribution from Trenches 4 and 5, in which the Th-232 is more widely dispersed. Conditional doses associated with exposure to these sources are comparable with those from Trench 2 (Figure 3.5), so the maximum potential individual dose associated with the coastal erosion would not be significantly changed unless Option 3 was also implemented. Targeted retrieval of Ra-226 from Trenches 2 and 3 (Option 2) would reduce significantly (up to 95%) the trench inventory of this radionuclide. This would in turn reduce the estimated maximum doses associated with the dwelling occupation case, used to assess the potential implications of gross disruption by inadvertent human intrusion 6. Conditional doses averaged across all trenches would be reduced by approximately a factor of five. 3.4 Other Remedial Actions Previous studies (Subsections 3.1 and 3.2) examined a range of potential remedial actions for disposals to the trenches, reaching the broad conclusion that large-scale in situ remediation was unlikely to be consistent with the principle of achieving an appropriate balance between detriment reduction and cost. However, the possibility of localised implementation of remediation techniques to improve isolation and containment was not comprehensively ruled out and LLW Repository Ltd therefore commissioned a further detailed review of the potential options, with a specific focus on potential target areas for localised action [47]. The summary that follows is based on detailed information provided in the full review report and other relevant studies Waste Heterogeneity and Radiological Significance In examining the potential for achieving significant reductions in the radiological impacts associated with past disposals, the question of heterogeneity in the associated radiological hazard (Subsection 3.3.1) is equally relevant to the consideration of other remedial actions as it is to waste retrievals. However, the potential effectiveness of different approaches in mitigating the impacts associated with different release pathways is somewhat different from that associated with targeted retrieval. 6 But only if it is assumed that the final cap and profile above the trenches plays no role in mitigating exposure. The most recent assessment calculations [12] suggest that the cap and profile material provide sufficient protection against disturbance by inadvertent human intrusion, so that Trenches 2 and 3 do not contribute at all to potential exposures for such scenarios. LLWR/ESC/(R11)/10025 Page 50 of 119

55 The extent to which the most important radionuclides are evenly distributed or present in localised regions of significantly higher than average concentration is a critical factor in determining the potential effectiveness of targeted actions. Hence, for the groundwater pathway, where Cl-36, Tc-99 and I-129 are all fairly evenly distributed through the trench wastes, the scope for effective mitigation of risk in the shorter term by locally implemented actions, even if their effectiveness could be guaranteed, is very modest. The disposed inventory of U-238 is the dominant contributor to estimated impacts via this pathway in the very long term (i.e. assuming coastal erosion does not take place). Localised actions (focused on Trenches 4 and 5) have the potential to affect the containment of approximately half the total inventory of uranium within the trenches [4]. However, given the half-life of the radionuclides involved and the slow rate at which uranium is expected to be released in leachate [12], it is questionable whether measures other than retrieval would in practice be capable of significantly mitigating even the hypothetical impacts associated with this pathway. As noted previously, estimated impacts for the gas pathway (based on the assumption that the engineered cap continues to provide a functional barrier to radon migration) are dominated by the disposed inventory of C-14. Even so, the estimated dose from C-14 release from the trenches is consistent with the risk guidance level. Moreover, the fairly even distribution of C-14 in trench disposals means that the scope for effective mitigation of risk by targeted actions to improve containment, even if their effectiveness could be guaranteed, is again very modest. Only a large-scale approach to in situ remedial action would appear to have any chance of affecting projected radiological impacts. The radiological impacts associated with disruption of the LLWR by coastal erosion or by human intrusion are a consequence of assumed exposure to radionuclides that have been retained within the repository prior to the time when the event occurs. Whereas retrieval actions (targeted or otherwise) would effectively remove such hazards entirely, other types of remedial action inevitably tend to be targeted at inhibiting the release of key radionuclides from the trenches by enhancing isolation or containment. Whether or not the radionuclides are heterogeneously distributed in the disposed wastes, the scope for achieving significant reductions in the radiological impacts associated with these cases through such actions is therefore intrinsically small. Nevertheless, it might perhaps be argued (in relation to human intrusion exposures) that the rate of radon gas release could be reduced by modification of the wasteform, or that certain forms of in situ remediation might reduce the likelihood of a significant intrusion taking place Evaluation of Alternative Remedial Actions A range of possible remedial actions, aimed at improving the isolation and containment performance of the trenches, were examined in order to assess the extent to which they might reasonably contribute towards optimised management of the LLWR [47]. The intent was to provide a more systematic assessment of feasibility, potential benefits, adverse impacts and costs than had been undertaken in previous studies following the 2002 safety case (Subsections 3.1 and 3.2). Broadly speaking, other than waste retrieval, the options considered fell into three broad categories: modification of the chemical environment in the trenches. Such actions would be aimed at decreasing the concentration in leachate solution of key radionuclides LLWR/ESC/(R11)/10025 Page 51 of 119

56 relevant to the groundwater pathway (Cl-36, Tc-99 and I-129), either by altering their solubility or their tendency to be adsorbed to solid materials, and thereby reducing their mobility. This would be achieved through measures aimed at controlling the local ph or redox potential, or through the introduction of other agents. However, it is only the behaviour of Tc-99 that can be significantly affected by a change in chemical conditions, and the most effective course of action for this radionuclide would be to target redox conditions. It is already believed that Tc-99 exists in a reduced state within the trenches, but environmental modification could potentially increase the confidence with which such conditions can be expected to persist. Alternative technologies include the promotion of chemically reducing conditions through introduction of microbial activity or zero-valent iron; modification of the physio-chemical properties of the wastes. Rather than modifying the waste environment, such actions would be aimed at changing the characteristics of the wastes themselves to restrict groundwater access, inhibit gas transport and reduce radionuclide mobility. This would be achieved principally by converting the wastes into a monolithic form. The three main alternative technologies examined in the review were in situ vitrification, in situ solidification and stabilisation through cementitious grout injection, and waste retrieval followed by ex situ solidification and stabilisation; local protection. Such actions involve the installation of local barriers to improve the isolation and containment at identified key locations, either to prevent or reduce migration, or to reduce the likelihood/impact of human intrusion. Possible alternatives included increasing the thickness of the engineered cap, the installation of concrete slabs above targeted areas of the trenches, and the installation of localised cut-off walls. More details of the alternatives examined in the assessment process are provided in the detailed review report [47]. The first stage in the evaluation was to consider whether any of the associated technologies would be unsuitable for application to the LLWR trenches, and whether any of the approaches was suitable for selective, targeted application. Modification of the chemical environment would be implementable in principle, and there are some examples of its application for specific categories of hazard, including Tc-99 contaminated groundwater [50,51]. However, there are few significant fieldscale applications as a remedial technology in the context of radioactive waste management. One specific technology that has been applied at a number of sites in the USA (iron nanoparticles) relates to the treatment of chlorinated organics (rather than radionuclides) and is not yet permitted in the UK. Because Tc-99 is widely distributed through the trenches, any selected technique would require large-scale (e.g. whole trench) application in order to make a significant contribution to effective remediation [47]. Modification of waste properties is considered to be feasible in principle by cement encapsulation (either in situ or ex situ), but the application of in situ vitrification was screened out for similar reasons to those identified in previous studies (Subsections 3.1 and 3.2). In particular, the LLWR trench environment is expected to be at least partially wet and lacks the desired presence of significant quantities of silica to promote vitrification. Ex situ solidification and stabilisation necessarily involves the excavation and retrieval of the wastes (although not their segregation) and, like retrieval, is therefore considered suitable for application only as a targeted remedial action. By contrast, the in situ solidification and stabilisation by the injection LLWR/ESC/(R11)/10025 Page 52 of 119

57 of cementitious grout was considered suitable for practical application at either a smaller or larger spatial scale. Implementation of cut-off walls at a local scale would not offer scope for mitigation of impacts, because the disposals that dominate risks from the groundwater pathway are widely distributed throughout the trenches (see above). However, the local implementation of enhanced capping in specific locations, or the emplacement of a concrete slab above targeted areas, were both carried forward as potential mitigating actions for human intrusion cases. The identified potentially feasible options were then carried forward for assessment against a more extensive set of evaluation criteria, including their expected effectiveness in mitigating radiological impacts, practicability and operational safety considerations associated with their implementation, environmental impacts, secondary wastes and nuisance (including impact on ongoing LLWR operations) and implementation cost [47]. The following broad conclusions can be drawn from the evaluation process: Coastal Erosion None of the identified remedial actions (with the exception of waste retrieval considered previously Subsection 3.3) has the potential significantly to mitigate radiological impacts associated with the coastal erosion. Neither modification of the chemical environment nor the installation of local protection measures would be expected to reduce the extent of radiological exposure from the wastes once they had been uncovered by erosion. The formation of a monolithic wasteform could potentially increase exposures from external irradiation once erosion has taken place by reducing the rate at which the wastes are degraded and dispersed. Any minor effects from enhanced shielding or reduced dust generation (given that inhalation is in any case a minor contribution to total exposure in this case) are therefore unlikely to offer a significant net benefit. Protection against Groundwater and Gas Release The opportunity for remedial actions to mitigate radiological impacts associated with the gas and groundwater release pathways is modest. Of the identified options, only large-scale solidification and stabilisation offers any clear potential for risk reduction. Although modifying the waste chemical environment would be aimed primarily at reducing the mobility of radionuclides in groundwater, environmental changes targeted at some radionuclides (e.g. through control of oxidation potential) might equally well increase the mobility of others. For example, introduction of microbes with the aim of increasing confidence in the persistence of reducing conditions within the trenches might lead to an increase the gaseous release of C-14. Given such trade-offs, and the lack of any comparable experience or definitive evidence of significant beneficial effect in the context of the LLWR environment, there is little confidence in the potential for achieving effective impact mitigation. Local applied remedial actions, including the installation of additional barriers or solidification and stabilisation (both in situ and ex situ), will also have a minimal impact on both groundwater and gas release pathways because the key radionuclides are widely dispersed within the trenches. It LLWR/ESC/(R11)/10025 Page 53 of 119

58 is therefore not feasible that such actions could affect a significant proportion of the overall inventory. Large-scale in situ remediation based on the injection of a cementitious grouting agent would appear to offer some potential for reducing releases via the groundwater and gas pathways, by restricting groundwater access to the wastes and inhibiting radionuclide mobility (in both gaseous and solute form). The potential is likely to be greater for the groundwater pathway than for gas release. Human Intrusion Some remedial actions have the potential to mitigate radiological impacts associated with human intrusion. However, given that none (apart from waste retrieval) actually entail a decrease in the radiological hazard within the facility, the scope for any significant reduction is modest. Modifying the waste chemical environment would be targeted at inhibiting radionuclide release and will therefore, if anything, tend to increase the amount remaining in the trenches at the time intrusion is assumed to occur. No mitigation of impacts can therefore be expected. Local enhancement of capping in specific locations, or the emplacement of a concrete slab above targeted areas, might potentially deter some forms of intrusion (i.e. reducing the likelihood of exposure), or perhaps increase the effective dilution of excavated wastes by mixing with the overburden. Whilst there might be some impact on the likelihood of inadvertent intrusion, however, the impact on conditional exposures is expected to be low. The formation of a monolithic wasteform is expected to reduce the airborne concentrations of radionuclides in dust in the event of intrusion and the potential mobility of relevant radionuclides into the food chain. Both these mitigating factors will have greatest influence on conditional doses associated with the intruder and site occupier cases (for which Th-232 is more important). For the site dwelling case, the preliminary assessment results available at the time [27] suggested that the main contributor to conditional dose is from disposals of Ra-226. Because Ra-226 is heterogeneously distributed within the trenches (more than 90% associated with a small fraction of Trenches 2 and 3), either small-scale or large-scale stabilisation and solidification was identified as having the potential to mitigate impacts. Specifically, the formation of a monolithic wasteform may reduce the release rate of radon by inhibiting its migration to atmosphere. However, the effectiveness of such a measure will depend on the quality of the monolith, which may be difficult to assure. Moreover, if it is assumed that the final cap and profile material provide protection against disturbance by inadvertent human intrusion (as in most recent assessment calculations [11]) Ra-226 in the trenches makes only a limited contribution to potential exposures for such cases. Large-scale in situ solidification and stabilisation As had been noted in earlier analyses (e.g. Subsection 3.2), there is potential scope for inhibiting the mobility of radiologically significant radionuclides in groundwater by large-scale in situ grouting (via pressure injection or mixing) within the trenches. Large-scale application, targeted at the groundwater pathway, would also carry any associated benefits of LLWR/ESC/(R11)/10025 Page 54 of 119

59 solidification and stabilisation that might be attributed to mitigation of human intrusion exposures from more heterogeneously distributed radionuclides. The technical knowledge to implement such remedial action exists and has been demonstrated, but the effectiveness with which it could be implemented at the LLWR (e.g. in the context of highly compacted wastes, and tailoring the cementing or other chemical agents to the wastes) is uncertain [47]. Engineering work required to implement such an action (including the removal of the interim cap and adoption of a process to ensure effective mixing) would be extensive. Indicative costs for remediation of all trenches (to address hazards associated with the groundwater pathway) are estimated to be in the region of 120 million [47]. Wider impacts of large-scale in situ grouting include compromising development plans for final capping of the trenches and possible impacts on the final cap design. Remediation activities in Trenches 2 and 3 (including removal of the interim cap) could have implications for ongoing operations in the adjacent vaults and there would be off-site implications in terms of nuisance (from noise, visual impact, etc.) associated with operating machinery. Whilst some mitigation of impacts for the groundwater (at least in the shorter term) may be achieved, the overall effectiveness of the process remains to be demonstrated. Moreover, assessment calculations for the groundwater pathway [12] provide a measure of confidence that risks associated with trench disposals (in the absence of remedial action) are in any case less than the regulatory risk guidance level. Such an outcome tends to diminish the attractiveness of taking remedial actions to enhance containment and isolation, as part of an ALARA process, unless they are implementable with confidence at limited cost and with no other associated detriment. Smaller-scale in situ solidification and stabilisation Remedial action at a smaller scale would be geared towards targeting Th-232 and Ra-226 in order to mitigate the potential consequences of human intrusion. The estimated cost of applying the technique to Th-232 wastes (i.e. targeting selected areas of Trench 2 together with more extensive coverage of Trenches 4 and 5) is in the region of 13 million, while implementation for Ra-226 wastes (i.e. parts of Trenches 2 and 3) is estimated to cost in the region of 6 million [47]. However, similar uncertainties to those identified above regarding the potential effectiveness of the approach also apply at the smaller scale. More confidence in the performance of the final wasteform is likely to be gained via the implementation of an ex situ process. However, this would incur substantial additional cost, broadly equivalent to that associated with targeted waste retrieval (Subsection 3.3), albeit without the stage of waste segregation. Estimates for the ex situ remediation of those areas of the LLWR trenches where there are high concentrations of Th-232 and Ra-226 are 183 million and 83 million, respectively [47]. The wider impacts of smaller-scale in situ grouting are similar to those for large-scale operations, including the requirement to remove the interim cap; however, the overall remediation process (and the associated nuisance) would extend over a shorter time period. LLWR/ESC/(R11)/10025 Page 55 of 119

60 Whilst some mitigation of impacts may be achieved for a range of human intrusion cases (depending on whether Th-232 or Ra-226 were targeted), the actual extent of any reduction in dose is difficult to quantify. This is because of uncertainties in how the wasteform would perform in the event of intrusion, and not least because the hypothetical cases on which dose estimates are based are themselves necessarily highly stylised. Mitigation of conditional doses from Ra-226 for the site dwelling case might be more effective, if it could be demonstrated that the process proved was highly efficient in inhibiting radon release. However, under the assumption that the final cap and profile material provide effective protection against disturbance by inadvertent human intrusion (as in most recent assessment calculations [11]) Ra-226 contributes little to potential exposures for such cases, so the actual benefit in terms of risk reduction would be marginal. Local protection The primary aim of installing local protection measures (enhancement of the cap or the emplacement of a concrete slab above targeted areas) would be to deter some forms of inadvertent intrusion and thereby reduce the likelihood of exposure. As with the use of markers, and attempts to ensure the preservation of records, such action might be accepted as a valid in the context of reducing risk, but its quantitative impact is not readily evaluated and it can therefore be difficult to demonstrate that implementation would represent a proportionate risk-reduction measure (particularly given the comparatively low doses associated with intrusion cases for the trenches [12]). The estimated costs of installing such protection measures to target Th-232 and Ra-226 wastes within the trenches are 4.5 million and 2 million, respectively. The figures are based on the assumption that a 1m thick concrete slab would be installed above the relevant trench bays prior to construction of the final cap [47]. A further possible local measure might be to alter the profile of the cap such that the perimeter areas of the trenches (particularly those areas where significant Th-232 disposals took place) were provided with protection against intrusions to depths of greater then 4m. However, there are planning and engineering constraints on the extent to which a thicker cap could be developed, which we believe prevent this from currently being considered a practicable option (see Subsection 5.2.4). 3.5 Status Summary The optimisation of management controls and interventions relating to the disposed inventory from past operations of the LLWR is inevitably framed both by the way in which those disposals were originally implemented and the overall vision for the future role of the facility. As discussed above (Subsection 3.3.1), an important aspect of this is the distribution of radiologically significant radionuclides in past disposals, particularly in relation to consignments to the trenches. This enables consideration to be given to the potential for targeted actions to achieve significant overall benefit in terms of risk reduction without incurring the cost and wider impact of measures taken across the facility as a whole. LLWR/ESC/(R11)/10025 Page 56 of 119

61 The wider framing of optimisation measures also means that the viability and implications of possible remedial actions relating to past disposals cannot be considered in isolation from other aspects of the disposal facility. Because UK Strategy anticipates that disposal operations at LLWR will continue (based on optimised LLW management across consignor sites and conditional on achieving an acceptable environmental safety case) [24], the implications of any remedial actions for possible ongoing operations on the site are an important consideration in determining the overall balance between benefit and detriment. Moreover, the potential benefits achieved by actions implemented to manage risks associated with past disposals need to be considered alongside the implications of future waste acceptance and management controls over the future use of the site. Option studies undertaken shortly after submission of the 2002 PCSC (Subsection 3.1) reached the conclusion that the only management options with demonstrable potential to reduce risks associated with past disposals involved the effective removal of wastes from the site. In particular, it was evident that the radiological detriments relevant to decision making were those associated with longlived radionuclides, and that the magnitude of their inventory was the determining factor in relation to possible interventions, regardless of potential (but likely shorterterm) improvements in their containment. Enhanced natural attenuation (by increasing dilution) was rejected as inconsistent with the principles of isolation and containment, leaving waste retrieval as the only option offering significant potential for mitigation. Subsequent studies (Subsections 3.2 to 3.4) have done little to alter these broad conclusions, although they have added considerably to the underlying evidence base and analysis and have enabled a more refined examination of alternatives. In particular, they underline that targeted remedial actions based on promoting the containment within the trenches (i.e. inhibiting the release) of key radionuclides (Th-232 and Ra-226) would do nothing to mitigate the radiological impacts associated with coastal erosion, while their potential effectiveness in ameliorating exposures associated with human intrusion is tentative at best. In this context it is also relevant to bear in mind that, since coastal erosion of the LLWR is the expected outcome of site evolution, even though the timing is subject to some degree of uncertainty [8], it can be considered more likely to occur than disruption by human intrusion within a period of a thousand years or so after site closure. Leaving aside the wider impacts of remedial actions on ongoing site operations, as well as secondary environmental impacts and off-site nuisance, the financial cost of implementing actions to enhance the containment of Th-232 and Ra-226 within the facility, coupled with the lack of demonstrable effectiveness in significantly reducing risk, suggests that they would therefore be grossly disproportionate to any overall benefits they achieved. The more recent option studies [44,47] also show that, in the light of the expected long-term fate of the facility, the potential for remedial actions to provide effective mitigation of exposures associated with the groundwater and gas pathways is similarly modest. Aside from the intrinsic cost-effectiveness of the measures themselves, which would need to be undertaken across the whole facility because the relevant radionuclides are widely distributed, a relevant consideration is that the contribution from trench wastes to total inventory and potential exposures via these pathways (from Cl-36, Tc-99, I-129 and C-14) is small by comparison with that from Vault 8, where no comparable remedial actions (other than full-scale retrieval) are possible. Remedial actions implemented to address impacts from the trenches LLWR/ESC/(R11)/10025 Page 57 of 119

62 would therefore have a minor impact on overall site risks, even if no further disposals took place. The NDA s Strategic Environmental Assessment [26] identified that full-scale retrieval of wastes from former disposals (whether from the trenches alone, or both the trenches and Vault 8) would be very expensive to carry out, and would carry with it a range of other detrimental impacts associated with the retrieval operations themselves. That is not to say that it would be impossible; however, it would clearly call into question the overall viability of the LLWR as a disposal facility. In particular, it is not necessarily the lack of waste conditioning or simplicity of engineering associated with legacy disposals that is the main driving factor behind the scale of future impacts, but the fact that long-lived radionuclides contained within the facility are vulnerable to disturbance by natural processes or inadvertent human actions. Taken as a whole, and bearing in the mind the outcomes of the latest analyses of safety performance for the LLWR [12], the above discussion leads to the identification of the following set of alternative remedial action strategies for the control of past disposals, coupled with future management of the LLWR site: A: Full-scale retrieval of all past disposals into storage (or disposal to an alternative facility), with no further use of the site for disposals of long-lived wastes. B: Leave the majority of past disposals of waste in situ, but carry out selective retrievals (targeted on Th-232 and, possibly, Ra-226), and cease using the site for future disposals of long-lived wastes. C: Take no remedial actions in relation to past disposals (other than continuing leachate management and closure engineering) and operate the site as planned to capacity, subject to the definition and implementation of relevant controls on waste acceptance and emplacement [15]. D: Undertake selective retrieval of Ra-226 only, and carefully manage Ra-226 inventory in future disposals. E: Undertake selective retrieval of Th-232 and Ra-226 from the trenches, and carefully manage Ra-226 inventory in future disposals. The earlier conclusion for Option A still holds, that the costs and other impacts would be disproportionate to the benefits to be gained in terms of risk reduction. Option B is inconsistent with UK Strategy as currently defined [24] provided it is shown that continuing use of the LLWR is safe. The 2011 ESC demonstrates that it is safe to continue to use the LLWR for the disposal of LLW. Option C represents the effective default action, consistent with declared UK Strategy, while Option D considers the possibility of remedial action focused on a specific source and exposure case (Ra-226 and human intrusion). This is because the targeted retrieval of Ra-226 from localised areas of Trenches 2 and 3 would reduce significantly the trench inventory for this radionuclide and hence the associated hazard from specific intrusion cases. However, the extent to which such action might be considered a proportional contribution to optimisation is conditional on the role played by the final cap and profiling material in mitigating exposures for such cases. The most recent assessment calculations [12] indicate that the protection provided by engineered barriers against disturbance of the content of Trenches 2 and 3 is such that Ra-226 makes very little contribution to potential LLWR/ESC/(R11)/10025 Page 58 of 119

63 exposures associated with inadvertent human intrusion. In this context, retrieval would make only a very marginal impact on estimated impacts for the facility. Option E is effectively an extended variant of Option D. Although past disposals of Th-232 are relevant to some (lower impact) intrusion cases, this strategy would logically be directed towards mitigation of radiological risks arising from coastal erosion. In this case, however, even targeted action would need to cover a substantial area (i.e. not only Trench 2, but also Trenches 4 and 5), with associated implications for cost, in order to achieve a significant (i.e. better than factor of two) mitigation of conditional exposures. The operational requirements for implementing targeted remedial actions to retrieve wastes (Options D and E) include the development of an extensive infrastructure on the site in support of excavation, sorting, segregation and re-packaging, stabilisation and backfilling. Although such factors are not necessarily overwhelming and their environmental impacts can be mitigated by appropriate management, retrievals of Ra-226 (and possibly Th-232) from Trenches 2 and 3, in particular, would be likely to have a significant adverse impact on ongoing operations in the adjacent vaults. Th-232 retrieval from Trench 2 alone could reduce significantly the projected conditional doses associated with human intrusion into the trench wastes at that location. The estimated cost of 150 million is rather less than the 500 million associated with targeted retrieval of a majority of past Th-232 disposals (which would be necessary to mitigate significantly the impacts from coastal erosion). However, the conditional doses associated with intrusion into these wastes are consistent with the lower bound dose guidance level of 3 msv (for prolonged exposure) identified in the GRA [12]. Moreover, as noted previously, the latest assessments indicate potential doses associated with coastal erosion of the trenches that correspond to peak conditional individual risks in the region of 10-6 per year. This effectively leaves Options C and D as being the only future site use or remedial action strategies that reflect current understanding of the potential effectiveness of possibly remedial actions and are consistent with declared UK Strategy. The key question is then whether the cost (about 150 million [44]), disruption and potential risks associated with Ra-226 retrieval from Trenches 2 and 3 are proportionate to the benefit it would secure in terms of a reduction in the hypothetical consequences of radon release into a building that might be constructed on the LLWR cap at some time in the future. It is argued in the most recent assessment calculations [12] (see above) that Ra-226 in the trenches would in fact make very little contribution to potential exposures associated with inadvertent human intrusion, because of the protection against disruption of the wastes afforded by the trench cap and profiling material. Moreover, given that coastal erosion is expected to occur within a period of few hundred to few thousand years [8] and therefore a more certain concern than possible disruption by human intrusion, the overall magnitude of benefit achievable by Ra-226 retrieval is small. We therefore consider Option C to be the preferred course of action, provided it is accepted that the consequences of gross disruption of the site are tolerable. Ultimately, such decision making requires a wider perspective than simply demonstrating that calculated risks (with all the stylised assumptions and uncertainties they entail) are either above or below regulatory guidance levels for the period after authorisation. Nevertheless, were the exposure implications of coastal erosion or human intrusion (without waste retrieval from the trenches) to be judged unacceptable, the appropriate course of action would not simply involve remedial actions relating to past wastes but a change in overall strategy regarding the future LLWR/ESC/(R11)/10025 Page 59 of 119

64 operation of the LLWR. Since this would be incompatible with broader national objectives regarding the continuing role of the LLWR [24], and because we consider that leaving the wastes in situ does not give rise to unacceptable long-term exposure to radiological risk [12], Option C has therefore been adopted as the basis for the forward management plan and the ESC. Consideration of the options for trench remediation has nevertheless underlined the importance of questions regarding the acceptability of disposal of specific waste streams in future, particularly those with high concentrations of long-lived radionuclides. For example, even though the selective recovery of Th-232 and Ra-226 in mineral sands and process wastes from the trenches has been discounted as a potential intervention in current plans, there remains scope for optimising the acceptance, treatment and emplacement of similar waste streams within future disposals, in order to ensure that the potential consequences of site disruption (by human intrusion or coastal erosion) are ALARA. Further discussion is provided in Section 4 and more detailed analysis in the report on Waste Acceptance [15]. LLWR/ESC/(R11)/10025 Page 60 of 119

65 4 Management of Future Waste Disposals Optimising the operational management of the LLWR requires appropriate controls to be established regarding the acceptance of wastes for disposal, how those wastes are conditioned and packaged, and the way in which they are emplaced in the facility. Such optimisation considerations are framed by the design of the facility, now and in the future [5], as well as understanding of key issues influencing environmental safety performance in the shorter and longer term [11,12]. They are also informed by the underpinning UK Strategy [24], which anticipates best use of the facility 7 to maximise its operational lifetime, through rigorous application of the waste hierarchy across NDA sites to ensure that the LLWR receives only those wastes that are appropriate for the degree of safety and security offered by the facility. The baseline Site Development Plan at the time of BNFL s 2002 Safety Cases [22,23] anticipated that disposal arrangements for Vault 9 and future vaults would adopt essentially the same approach as that implemented for Vault 8. No additional risk mitigation measures relating to waste acceptance, treatment and emplacement (other than a review of radiological capacity) were examined as part of the 2002 PCSC [35]. High-force compaction of compactable raw waste, followed by grouting of the compacted pucks and other non-compactable wastes within ISO containers, was therefore assumed to be the standard method of waste treatment and packaging (see Subsection 2.2.3). Although the subsequent authorisation decision for Vault 9 restricted its use to temporary storage only, and a series of options studies were undertaken to evaluate overall design at that time [34], no corresponding detailed review was undertaken of wasteform prior to the granting of planning permission. The same methods for producing a final wasteform for emplacement have therefore continued to be used in Vault 9 operations. The Environment Agency s assessment of the 2002 PCSC [21] made no specific reference to waste conditioning and emplacement arrangements in raising concerns that insufficient attention had been given to optimisation and risk management. However, it was suggested that the safety analysis was deficient in so far as it had not evaluated the potential effects of grout superplasticiser on biogeochemistry and radionuclide migration (and hence had not justified that the grout formulation represented implementation of Best Practicable Means). More generally, the Environment Agency assessment recommended that, among other strategic management options, consideration should be given to the possibility of restricting future disposals to certain specific (e.g. short-lived) categories of waste. These observations were eventually subsumed as part of the Schedule 9 requirements for improvement and additional information (a comprehensive review of best practice for minimising the impacts from all waste disposals on the site ), attached to the subsequent Authorisation (now Permit) [36]. LLWR subsequently submitted a set of responses to the Environment Agency s Schedule 9 requirements, accompanied by a range of supporting studies. The overview report [52] set out a methodology for determining the radiological capacity of the site, to support the definition of waste acceptance criteria (WAC). In addition, 7 As noted previously, optimised use of the LLWR (in the context of its role as a component of the UK Strategy for LLW management) is not the same as optimisation with respect to the control of radiological risks. Nevertheless, operational considerations associated with managing the capacity of the facility are an important element of the wider discussion of benefits and detriments relevant to radiological optimisation. LLWR/ESC/(R11)/10025 Page 61 of 119

66 the analysis of near field performance [53] reported on the outcome of a comprehensive literature review considering the impact of superplasticisers [54]. The work reported below summarises the outcome of these and other studies commissioned by LLW Repository Ltd in support of the current ESC that have a bearing on how we plan to optimise the management of future disposals, including the use of controls on waste acceptance (Subsection 4.1), methods for conditioning and packaging (Subsection 4.2) and approaches to waste emplacement (Subsection 4.3). 4.1 Waste Acceptance Controls on waste acceptance to the LLWR are aimed at ensuring that both the physical and the radiological capacity of the site are used in an optimised way. On the one hand, this involves ensuring that waste volumes are minimised, consistent with the waste hierarchy and NDA s goal of extending the operational lifetime of the LLWR. On the other, waste acceptance controls are used to ensure that disposals to the facility respect radiological and non-radiological limits on hazardous material inventories, as well as engineering and waste performance requirements Optimisation of Physical Capacity Managing the physical capacity of the LLWR is an element of the wider framing that determines the inventory disposed to the facility. In line with development of the NDA s UK Strategy for management of LLW [24], LLW Repository Ltd is working with waste consignors to implement waste management routes that increase opportunities for reuse and recycling, or otherwise support a reduction in volume or change in the nature of waste sent for disposal to the LLWR. To this end, LLWR has published an operational strategy that is intended to support the optimal use of the facility [55]. One of the main objectives of the operational strategy is to provide innovative solutions for delivering a reduction in UK s LLW liability, which in turn will extend the operational lifetime of the LLWR. A new form of contract has been introduced for waste management consignors with the aim of better integrating waste management options and services. Where wastes cannot be prevented from arising, the strategy and services provided by LLW Repository Ltd seek to increase the amount of radioactive waste from consignor sites that is handled as VLLW or exempt waste, and to increase opportunities for reuse and recycling of waste materials. Key elements of this waste volume minimisation service [55] include: the application of robust characterisation at the point of waste arising to guide the utilisation of fit-for-purpose management and disposal routes, commensurate with the required levels of safety, security and environmental protection. The goal is to reduce the number of pessimistic declarations of low-activity waste streams by seeking opportunities to identify and segregate VLLW and exempt waste rather than requiring their disposal to the LLWR; the provision of packaging solutions that facilitate the segregation of waste (see also Subsection 4.2); promoting the use of waste treatment technologies and processes to reduce the volume of waste material requiring disposal by changing its physical form or by LLWR/ESC/(R11)/10025 Page 62 of 119

67 removing surface contamination. In addition to high-force compaction, treatment processes include the surface decontamination and melting of metals and the incineration of combustible wastes. Not all these services are yet fully implemented; however, it is envisaged that their implications for the control of waste consignments will ultimately be fully integrated into the acceptance procedures for the LLWR [15]. As part of the wider examination of alternative waste management strategies and their implications, LLW Repository Ltd commissioned review studies to examine the potential implications of waste volume minimisation innovations for development of the ESC [56]. Among other things, these studies observed that a significant proportion of the LLW that could be treated through incineration is expected to arise prior to The process will therefore need to be adopted extensively at a fairly early time if it is to contribute significantly to extending the operational lifetime of the LLWR. The studies have also informed the development of controls on waste acceptance, strategies for waste conditioning and packaging (Subsection 4.2) and the examination of alternative approaches to waste emplacement (Subsection 4.3). The Waste Inventory Disposition Route Assessment Model (WIDRAM) has been developed on behalf of the NDA and LLW Repository Ltd in order to make projections of the radionuclide and materials inventory of LLW consignments to LLWR, and to enable an analysis to be made of differing waste segregation, treatment and disposal scenarios. Assumptions regarding the optimisation of future disposal capacity in defining the baseline inventory for the current ESC are discussed in the Inventory report [4] Radiological Capacity There is an essential balance to be struck between the desire to extend the operational life of the LLWR and the need to ensure that the risks associated with radioactive inventory accepted for disposal are tolerable. Identification of the radiological capacity of the site in general, and the future disposal vaults in particular, is an important element in determining controls on waste acceptance. Moreover, as noted in regulatory guidance [19], the determination and application of such controls in management of the disposal facility is itself a component of overall optimisation (Subsection 1.1). Estimates of radiological capacity depend on the understanding we have developed of the isolation and containment performance of the disposal system with respect to the threats of groundwater and gas release, and disruption by natural processes or human intrusion. Starting from the baseline inventory [4], the first step is to examine the extent which both the projected volume and radiological content of the wastes can be accommodated within the facility. Optimisation of capacity (and the associated implications for waste acceptance and emplacement) then involves assessing the potential implications of varying the inventory against optimisation principles. A large proportion of the total projected baseline inventory of some key radionuclides (including H-3, isotopes of uranium and Th-232) has already been disposed to the trenches [4]. Imposing strict limits to future disposals of these radionuclides through conditions for acceptance is therefore unlikely to have a significant effect on longterm impacts for the site as a whole. For other radionuclides (including C-14, Cl-36, Tc-99 and Ra-226), the majority of the potential total site inventory is associated with LLWR/ESC/(R11)/10025 Page 63 of 119

68 disposals to future vaults, and there is more flexibility to reduce the total inventory for the site as a whole. Nevertheless, even though the selective recovery of wastes from the trenches has been discounted as a potential intervention, and regardless of the relative contribution from future waste arisings, there remains scope for excluding waste streams from future disposals (particularly those containing specific long-lived radionuclides) in order to ensure that radiological risks to members of the public are ALARA. This involves considering not only the scale of the risks involved, but also the source and magnitude of the waste streams involved and the potential consequences of taking alternative course of action (e.g. the availability of alternative management routes). Engagement with waste consignors (as well as regulators) is an important element of LLW Repository Ltd s approach to the optimisation of waste acceptance, taking into account the understanding we have developed through safety and performance analysis of the disposal facility. This helps to ensure that the strategic approach we take in managing the radiological capacity of the site takes account of key stakeholder perspectives regarding the implications of alternative options. In addition to assessing the contribution from different radionuclides and waste streams to radiological risk, the determination of controls on waste acceptance involves ensuring that disposals to the facility respect non-radiological limits on hazardous material inventories. It is also necessary to ensure that the physical and chemical properties of the materials disposed as LLW do not adversely affect the engineering performance of facility components (e.g. by causing excess settlement that might influence the functioning of the repository cap) or the containment functions of the vault design (e.g. through detrimental impacts on radionuclide mobility). A comprehensive summary of the development of WAC, taking these optimisation considerations into account, is provided in the Waste Acceptance report [15]. Key elements of waste control arrangements include: identification and management of the radiological capacity of the LLWR, consistent with the outcomes of the ESC and relevant Permit conditions; derivation and application of additional consignment limits to control activity concentrations; for those wastes that potentially have the greatest significance in environmental safety terms, additional emphasis is placed on ensuring that disposals at the LLWR represent the optimum strategy for the management of the wastes; the use of emplacement strategies for particular waste streams and waste consignments where specific arrangements are required to demonstrate optimisation with respect to environmental safety (see also Subsection 4.3); provision of a basis for revisions to WAC for the LLWR and the adoption of waste acceptance arrangements that incorporate the assumptions and findings of the ESC. LLWR/ESC/(R11)/10025 Page 64 of 119

69 4.2 Waste Conditioning and Packaging Incineration The possibility to further optimise the vault wasteform was identified as part of the management option studies undertaken by BNFL following the 2002 PCSC [37,38]. The studies anticipated that, alongside high force compaction and grouting, future vault disposals could potentially involve the incineration of organic materials, with the aim of inhibiting contaminant mobility as well as reducing the potential for waste settlement within the vaults. An incineration route for suitable wastes is now being developed as a component of the wider waste volume minimisation service provided to consignors by LLW Repository Ltd [55]. The impact of incineration on safety performance has also been considered in more recent assessments of the implications of introducing innovations to minimise waste volume [56]. Consistent with BNFL s earlier assessments, the latter study concluded that incineration was likely to enhance the long-term integrity of the disposed wasteform and to improve packing efficiencies (and hence increase the specific activity content) of disposal containers. Incineration of the organic components of wastes would mean there would probably be less reducing conditions within the vaults system in the period after final closure; there would also be a corresponding reduction in methanogenesis (potentially a route for C-14 release via the gas pathway). A reduction in degradation will also reduce the potential for generation of organic colloids within the vaults. An improved packing efficiency will tend to lead to a reduced quantity of grout within the vaults and hence a slightly lower overall ph, which could possibly mean less effective containment of uranium in the long term. Reduced grout volume would also mean a reduced potential for C-14 capture by carbonation. Overall, the estimated effect of incineration (coupled with other waste volume reduction measures) on the calculated peak risk, for both the groundwater and gas pathways, was not judged to be significant in reference [56]. Final analyses and assessment for the ESC [6,12] have assumed incineration of a large fraction of organic wastes [4]. We have therefore concluded that the use of incineration to optimise the use of available disposal volume (consistent with UK Strategy [24]) is compatible with the objective of optimising overall radiological risks Grout Formulation and Quantities In-container grouting was originally introduced to support the aim of keeping the residual voidage in the waste stack low and helping to distribute the load on the vault base slab. This in turn diminishes the potential for significant differential settlement of the final cap [32]. Although improved containment of disposed radionuclides within the wasteform was not a primary functional requirement, it is recognised that grouting also tends to reduce the potential for interaction between the wastes and any waters passing through the facility. Moreover, long-term safety analyses take account of the fact that grout has the potential to provide further inhibition of release for some radionuclides through controls on ph and the provision of sorption sites [6] and reduction in C-14 labelled gas release. The grout formulation selected for use in the LLWR Grouting Facility was optimised from an engineering perspective with the aim of providing efficient flow properties during filling, as well as ensuring that it is self-levelling and settles sufficiently quickly LLWR/ESC/(R11)/10025 Page 65 of 119

70 within the waste container to limit the need for buffer storage prior to emplacement in the vaults (Subsection 2.2.3). The primary components of the grout are pulverised fly ash and Portland cement, in a 3:1 (by weight) ratio; however, its flow properties have been improved by the addition of a vinyl co-polymer based superplasticiser. In its assessment of the 2002 PCSC, the Environment Agency suggested that the potential effects of grout superplasticiser on biogeochemistry and radionuclide migration should be examined, to ensure that its implications for optimisation and risk management were addressed [23]. In particular, concern was expressed that such materials (together with other vault-derived complexants and colloids) might enhance the mobility of uranium. A review undertaken for LLW Repository Ltd [43] has examined the evidence for the potential of grout superplasticisers in increasing mobility, with a particular emphasis on their diffusion within a cured cement matrix. Although the potential to increase mobility is noted, the conclusion was drawn that the overall effect was likely to be small. The implications of uncertainties associated with this and other colloid transport effects in the vault system are examined in the detailed near-field modelling that supports the overall safety analysis [6]. Our judgement, based on this work, is that the role of superplasticiser in ensuring effective distribution of grout within the waste package (and thereby reducing voidage) is more important in terms of contributing to confidence in safety performance (in particular for the cap) than its potential detrimental effect on uranium mobility. This is particularly important in view of the importance of sea-level rise and coastal erosion, which places the focus of the safety analysis on the expected evolution of the disposal system and its performance during the first 1000 years or so after closure [12]. Nevertheless, we intend to maintain a watching brief on developments in superplasticiser technology, to examine whether alternative additives might be capable of delivering the desired grout properties but with even lower effect on mobility, as part of the ongoing optimisation of waste conditioning arrangements. As far as the overall amount of grout used in waste packaging is concerned, efforts to optimise the use of available space will tend to improve the packing efficiency in waste containers. Less grout is therefore required in order to maintain a low void fraction within the waste package (and hence contribute to the stability of the waste stack), which will have implications for geochemical conditions within the vault [56]. As discussed previously (Subsection 4.2.1), the results of the latest assessments [6,12] support the conclusion that efforts to ensure efficient use of the available disposal volume (i.e. leading to reduced amounts of grout within waste containers) are consistent with the objective of optimising overall safety performance. Nevertheless, our strategy for grout volume control is maintained under review alongside the development of alternative packing solutions (see below) Waste Package Disposals to Vault 8, and emplacements to date within Vault 9, have largely been accomplished using half-height ISO containers. Exceptions have included the use of different sizes of ISO containers for specific purposes and in situ grouting of some larger items (Subsection 2.2.3). In line with the overall aim of improving the overall efficiency with which the available volume for disposal is used, LLW Repository Ltd has been examining the potential for introducing alternative packaging solutions [55]. The primary intention of this development work (alongside other efforts in waste volume minimisation) is to reduce LLWR/ESC/(R11)/10025 Page 66 of 119

71 the amount of radiologically clean material that is disposed. As well as improving the efficiency with which the available space is used, it is anticipated that packaging costs may also be significantly reduced. Work has focused initially on the adoption of re-usable containers for waste transport, with modular internal packaging solutions (including half-height ISO liners for noncompactable wastes) (Figure 4.1). It is anticipated that the use of modular packages will also serve to facilitate improved waste segregation and application of the waste hierarchy (for example, via the provision of waste treatment services). The design of a disposal liner, to be used as the final container for waste emplacement, is now also being developed. Figure 4.1 Conceptual approaches to modular waste packaging, based on the use of re-usable transport containers (from [55]) The potential for improving containment performance through design optimisation of a disposal liner system (including the effects on waste stack stability and cap performance, as well as near-field geochemistry and waste containment) is being examined as part of the overall development programme for packaging innovations. For the purposes of the current ESC, however, it has been assumed that the standard disposal package continues to be a half-height ISO container. 4.3 Waste Emplacement In examining opportunities for minimising the impacts from waste disposals on the LLWR site, consistent with the principle of optimisation, LLW Repository Ltd commissioned a review of the viability and effectiveness of alternative strategies for waste emplacement [57]. The goal of the study was to identify alternative approaches to the emplacement of key wastes and to assess their implications (in terms of impacts on site operations, operational safety, environmental impacts and costs). A range of alternative waste emplacement strategies, which could be implemented either individually or in combination, were identified. In each case the aim would be to reduce the overall impact of facility operation and closure, by focusing on the control of a particular hazard associated with specific categories of waste. The types of strategy covered in the review included: controlling where certain waste packages are located within the disposal vault, according to the types of material they contain, or their radioactivity content with respect to identified trigger levels for particular radionuclides (i.e. particular waste streams or consignments); LLWR/ESC/(R11)/10025 Page 67 of 119

72 use of engineered sub-cells within the disposal vaults, to improve containment, mitigate the likelihood of intrusion or prevent gas migration; use of grouting outside and between waste packages, with particular emphasis on C-14 bearing wastes (to enhance the likelihood of capture by carbonation) and uranium wastes (to influence mobility through geochemical conditioning); providing a locally-reducing environment for waste packages containing uranium and Tc-99, to reduce mobility and release rates in groundwater; ensuring that acidic ash is separated from wastes containing radionuclides whose mobility may be increased by a local reduction in ph. The analysis examined whether any of the above waste emplacement strategies offered the possibility a significant improvement in environmental safety performance (or improved confidence in performance) and, if so, whether any such improvements could be achieved without disproportionate cost or associated impacts. Most types of emplacement strategy were identified as offering only marginal benefit, at best, compared with drawbacks (in terms of added complexity and cost, or even operational hazard) that their implementation would entail [57]. In some cases, significant changes to operational procedures could potentially be required (for example, requiring that waste consignments were held in storage for longer periods at consignor sites, or that extended buffer storage was constructed at the LLWR), with little in the way of any demonstrable beneficial effect on radiological risks, either in the shorter or longer term. For nearly all of the above identified strategies, we have therefore concluded that the detriments of implementation would significantly outweigh any advantages they might offer over existing waste emplacement arrangements [57]. Moreover, some of the options could potentially decrease significantly the available physical capacity for waste disposal. The single strategy that offers scope for more substantial benefit at proportionate impact on operation of the facility is that of controlling where in the vaults the disposal packages associated with certain small volume consignments are emplaced. Options under this general heading have therefore been examined in more detail as part of the definition of arrangements for waste acceptance [15]. From the perspective of the radiological hazard presented by the waste packages, two main controls on emplacement have been identified as elements of such a strategy: consignments for which the intention is to avoid placing more than one in a single stack of containers and in any immediately adjacent stacks. The aim will be to spread out within the vault those packages that might lead to relatively high impacts during coastal erosion. Such consignments would only be accepted after a consideration of whether disposal at the LLWR is the BAT, the nature of wastes and the number such consignments being receive; consignments excluded from within 5m of the cap surface, on the basis of the potential implications of radon gas release, or disturbance associated with certain types of human intrusion. As part of the approach to optimisation of operational practices, all such identified consignments will, so far as is practicable, be placed LLWR/ESC/(R11)/10025 Page 68 of 119

73 at lower stack positions, to minimise the potential for smaller-scale intrusions to intercept the wastes. We have also identified situations where controls on the emplacement of wastes containing absorbed liquids, or with higher potential void space, will provide benefits in terms of reducing the potential for release of liquids under load and avoiding localised subsidence (and thereby increasing confidence in the long-term performance of the cap). We will also keep under review the potential benefits of preferentially emplacing wastes with radiation levels that are significantly higher than average in such a way as to maximise shielding from external irradiation while the facility is operational. The number of consignments that will require emplacement restriction on the basis of the nature of the wastes or their activity content is such that it can be readily achieved within the scope of current practices [15]. LLWR/ESC/(R11)/10025 Page 69 of 119

74 This page is left blank intentionally. LLWR/ESC/(R11)/10025 Page 70 of 119

75 5 Pre- and Post-closure Engineering Design Assumptions relating to engineering design play a significant role in the ESC. They are also key elements in supporting future plans, programmes and investment decisions for facility development, operation and closure. A review of the basis for previous plans, and analysis of options for future development, has therefore been an important element in our preparation of the SDP and the ESC. It is not the intention here to explore the engineering design in great detail, as it is covered in other relevant documents [5]. However, we do examine the optimisation of engineering controls and their role in providing confidence in environmental safety performance. The baseline Site Development Plan that underpinned BNFL s 2002 safety case [32] incorporated assumptions regarding the design of future vaults and arrangements for final capping. Although these assumptions were supported by engineering optimisation assessments (as discussed in [35]), the overall safety analysis and safety case nevertheless attracted criticism from the Environment Agency for offering insufficient consideration of optimisation and risk management, to demonstrate that impacts will be as low as reasonably achievable (ALARA) [21]. The requirements specified in Schedule 9 of the ensuing Authorisation specifically underlined the need to demonstrate optimisation in the 2011 ESC [36]. Although some aspects of engineering design were explored by BNFL in management options studies shortly after publication of the 2002 safety case [37,38], the analyses were fairly general in nature and were not focused specifically on the rationale for determining a preferred approach to the use of passive controls in maintaining the isolation and containment of disposed wastes. Given the history of the facility, up to and including the design of Vault 8 (Subsection 2.2.3), it is relevant to examine the arguments that underpin our current development plans by starting from the basis on which a baseline design was developed (and planning permission sought) for Vault 9 (Subsection 5.1). Variants from this baseline have subsequently been explored as part of a comprehensive re-evaluation of design options and their implications [58]. The process, analysis and outcomes from this updated study are discussed in Subsection Vault 9 Planning Basis The background to the development and current operations at Vault 9 is described in Subsection An important element of the process by which permissions were sought and granted for Vault 9 was the so-called Vault Single Option selection process [34]. This process sought to establish a rationale for the design, consistent with its permitted use for temporary waste storage. A fundamental element of the decision process supporting this outline disposal concept was the exclusion of radically different disposal strategies (e.g. mined cavern, or silo-type disposal systems) developed on the LLWR site. This was underpinned by a screening process examining viable technologies, disposal practices and design principles, undertaken on behalf of BNFL in 2003 [59,60]. Deeper facilities, in particular, were rejected because at this site they would need to be constructed in unsuitable rock formations below the regional groundwater table. LLWR/ESC/(R11)/10025 Page 71 of 119

76 Implementation would therefore incur excessive costs, with significant scheduling and environmental impacts that were inconsistent with the current and anticipated future role of the site. Nothing has altered these general conclusions since the screening studies were undertaken, and the options examined to support the Vault 9 planning submission, as well as subsequent studies (Subsection 5.2) were therefore restricted to vault-type disposal arrangements. Consistent with the previous Vault 8 design, the basic concept developed via this process incorporated the following engineering components: a concrete reinforced vault base, with suitable gradients to support active leachate management, incorporating a waterproof liner; a low permeability final engineered cap; passive leachate discharge control through installed vertical drains; a complete encircling, deep COW. Although the Vault 8 and Vault 9 designs involved similar engineering components, there were substantial differences in philosophy. For example, the Vault 8 design has a lower specification base (involving a single natural clay layer, enhanced where necessary by engineered clay), compared with the composite barrier (liner plus clay) adopted for Vault 9. Also, because the passive safety performance philosophy that underpinned the Vault 8 design did not anticipate an explicit leachate containment role post closure, it was anticipated that drainage of the site would gradually revert to a natural state, with the potential for discharge to the near-surface environment, particularly towards the southern part of the facility [32]. It was therefore envisaged that a large single vertical drain would be required immediately to the south of the facility, to optimise long-term performance by directing any infiltrating leachate to deeper systems. By contrast, the Vault Single Option process identified that the retention of leachate within the vault for as long as practicable should dictate the design basis [34], resulting in the assumption of a double composite geomembrane and bentoniteenriched soil (BES) layer within the base and walls to maximise water retention. In this concept, should the vaults eventually become fully saturated, overtopping leachate would then be controlled passively by directing waters to a series of vertical borehole drains implemented between the line of Vault 9 (and any future vaults) and the trenches. No specific assumptions were made regarding cap design in the Vault Single Option study; it was simply expected that any final cap installed as part of final closure would necessarily be the primary passive engineered barrier and would therefore need to be robust. Its components would include bio-intrusion layers, vegetative layers, barrier layers, conveyance layers and gas control system. Final profiling was considered to be appropriately deferred to a later optimisation decision [34]. The absence of detailed discussion of closure design was consistent with the objective of the design process, which (in the absence of authorisation to undertake final disposal) was to gain approval to develop a system capable of providing longterm temporary storage. This planning objective was reflected in the stakeholder engagement and options evaluation process, which emphasised the storage role of Vault 9. Nevertheless, it was recognised that flexibility was required to adapt for possible subsequent conversion to a disposal facility, and that (provided that LLWR/ESC/(R11)/10025 Page 72 of 119

77 appropriate regulatory authorisation could be obtained from the Environment Agency) the Site Licence Company expected in future to present a second planning application to convert the purpose of Vault 9 from storage to disposal. Hence the emphasis in determining the overall design philosophy was on maximising leachate retention for as long as practicable, based on the principle that such a system: better satisfies the design principles, as defined in the applicable UK regulations and Best International Practice (i.e. containment) and will therefore be more acceptable to the regulators and other stakeholders. It was also the clear preference of the stakeholder elected representatives. [34] Additional flexibility was provided by the fact that the vertical drains and deep cut-off wall would not need to be constructed until late in site operation, and there was potential (if desired) to adapt the leachate containment strategy prior to closure by deliberately degrading the hydraulic performance of the vault base and walls. The Vault Single Option selection process was itself based on an earlier design options evaluation process, conducted following the 2002 PCSC. The screening exercise that was undertaken as part of this process [59], and which led to a focus solely on vault-based disposal systems, has already been mentioned. However, the work was extended further to a comprehensive evaluation of options for a preferred disposal design [61], including aspects of cap design. That study concluded that a gull-wing cap profile was preferred, partly in order to minimise the requirement for profiling materials, but also to support progressive construction. Moreover, it was expected that a gull-wing profile would also be more practicable in terms of enabling the anticipated later punching through the cap to install vertical drains, just prior to site closure (see Figure 5.1). The gull-wing profile was incorporated into illustrative design concepts presented in the Vault Single Option selection report [34]. The difference in objective between design for long-term storage and disposal is a key distinction between the reasoning adopted in decisions surrounding the Vault Single Option process and that relevant to the current ESC. It is also the reason why we have revisited the overall engineering design philosophy in preparation for the current ESC (Subsection 5.2). LLWR/ESC/(R11)/10025 Page 73 of 119

78 Figure 5.1 Section view for the preferred vault conceptual design in the Vault Single Option selection process (from [34]) 5.2 Design Optimisation for Disposal Scope of Design Optimisation Consistent with UK Strategy [24], the focus of the ESC, and therefore on options for engineering design, is on the role of LLWR as a disposal facility (i.e. capable of final closure, with no intent to retrieve waste at a later time). Because there is no realistic scope at the LLWR site for developing a mined cavern or silo-type disposal operation [59], these options are restricted to vault-type concepts. The basis for the ESC is the assumed development of the LLWR within the Reference Disposal Area (RDA), extending vault disposals in a modular fashion contiguously to the south-east from Vault 9 until they reach a line roughly defined by the extended southern perimeter of the trenches [5]. The implications of a variant case designed with a larger physical capacity, in and Extended Disposal Area (EDA) beyond the trenches and Vault 14 are being explored in supplementary studies (also reported as part of the ESC). Nevertheless, it is relevant to consider the extent to LLWR/ESC/(R11)/10025 Page 74 of 119

79 which specific engineering options may offer (or constrain) flexibility regarding potential future use of the site. Within these constraints, alternative engineering design options have been identified taking into account the types of passive control that are relevant to the main threats to isolation and containment of the wastes. These threats are those associated with: natural disruption of the facility in particular as a result of coastal erosion or inundation; disruption of the facility by future human actions in particular those actions that have the potential to cause disturbance of the wastes, as well as direct or indirect exposures to contaminants within the wastes; release of contaminated gases, generated either in the form of gaseous radioactive elements or as radio-labelled gases produced within the wastes, or inflammable gases; generation and release of contaminated leachate by water entering the facility and contacting the wastes. An examination of the potential role of engineering controls in mitigating the impact of these threats is presented in Subsection Evidence and Approach The options evaluation study was conducted prior to the main assessment calculations undertaken for the current ESC. However, it was able to draw on understanding gained from existing studies, including the 2008 performance update [27], which had been developed as an interim measure in response to the regulatory Schedule 9 requirements in our Authorisation (now Permit) [36]. Other key sources of information to guide the assessment of engineering options included earlier design option studies and supporting analyses undertaken prior to gaining planning permission for Vault 9 [34,61]. In addition, an important contribution to the examination of barrier functions in relation to different design strategies was provided by preliminary analyses of the engineered system hydrological performance [62,63], undertaken in preparation for the latest round of safety assessments. Preparatory analysis culminated in a scoping workshop with key stakeholders in December 2009, including representatives from the Environment Agency, Nuclear Installations Inspectorate, Cumbria County Council, NDA and the ESC Peer Review Group [58]. Participants were invited to review the proposed assessment approach and the principles adopted in undertaking the options analysis, through a combination of presentation and feedback as well as interactive sessions. A key conclusion emerging from that process was that the main emphasis in comparing engineering options (for the purpose of design optimisation) should be on whether there is a preference from the perspective of establishing confidence in the environmental safety performance of the LLWR. Having made such a comparison, it was then important to examine whether that preference was materially affected by wider considerations. This overall approach was adopted as a key theme in the more detailed analysis that followed (see Subsection 5.2.4). The scoping phase also endorsed the principle of examining engineering options in terms of the safety functions associated with different passive controls in relation to LLWR/ESC/(R11)/10025 Page 75 of 119

80 identified threats to waste isolation and containment. In this respect, the potential roles of engineered barriers were identified as including: protection against disturbance so far as is practicable; minimising water flow through the system for as long as practicable; controlling the release of gas and leachate that may be produced within the facility; directing releases that may occur so as to minimise their impact. It was also recognised that would inevitably need to be a balance between qualitative and quantitative reasoning in the examination of engineering design options. With the focus on confidence in safety performance, the aim of the options analysis was to identify significant discriminators between options at relevant times. A final integration phase drew together the outcomes from the assessment process into a more formal proposal for the engineering design strategy. An important element of this was a further round of formal engagement with stakeholders (based on attendance at the original scoping workshop), as well as ensuring agreement of the conclusions by the Site Licence Company senior management. Outside the main stakeholder events connected with the work, engagement activities also took place throughout the optimisation and options assessment process. These included two presentations to the LLWR sub-committee of the West Cumbria Sites Stakeholder Group, discussions with Drigg and Carleton Parish Council, and presentations at monthly liaison meetings between the Environment Agency and ESC Project team (also attended by the NDA and Office for Nuclear Regulation). Feedback from all such activities helped to guide and inform judgments made in the identification and comparison of options Engineering Mitigation of Threats to Isolation and Containment The first step in conducting a systematic review and evaluation of engineering features and their roles is an examination of the potential role of engineering controls in mitigating the impact of threats to isolation and containment of the wastes. Natural disruption Engineering features, such as perimeter drains around the final cap, can be built into the design to minimise the likelihood of localised flooding as a result of storm events. However, it is much more difficult to protect against threats from encroaching coastline and rising sea levels. The precise timescale over which disruption by coastal erosion might occur is uncertain, but projections based on current best understanding of the implications of global climate change on sea level, coupled with analysis of coastal processes in the vicinity of LLWR, suggest that it is almost inevitable within a period of a few thousand years, and might happen considerably earlier [8]. The characteristics of the disposed waste inventory are such that, following the initial decay of shorterlived radionuclides within the waste, there will most likely be no significant decline in the residual hazard at the time disruption occurs. Features that might delay LLWR/ESC/(R11)/10025 Page 76 of 119

81 disruption by a few years, decades, or even hundreds of years are therefore unlikely to have a large effect on the radiological impact when disruption occurs. Because the primary mechanism of disruption is expected to be erosion by under-cutting at the base of the sea cliff (most likely below the base level of the vaults), engineered features of a near-surface vault disposal system are not expected to offer significant protection. It is therefore difficult to argue a case for engineered barriers associated with the facility itself playing any significant role in hazard reduction by delaying disruption. A monolithic vault design might conceivably be more resistant to erosion than current waste emplacement arrangements, but would in any case not protect significantly against erosion by under-cutting. Some form of resistant material could conceivably be embedded in land surrounding LLWR. However, given the contribution to overall radiological hazard from very long-lived radionuclides on the timescales associated with natural disruption, it would be very difficult to establish confidence in the effectiveness of such measures in providing significant mitigation of risk through delaying disruption. Moreover, given the mechanism by which erosion is likely to occur, such measures would have to be very deeply embedded to be worth considering at all. Shoreline and coastal defences to protect against threats from coastal erosion or inundation would need to be continuously maintained over a substantial period in order to be effective in risk mitigation. It would be difficult to justify claims for the value of such defences as part of the SDP for the ESC extending several thousands of years into the future. Nevertheless, there are likely to be wider plans for coastal management in Cumbria that may affect on coastline evolution in the vicinity of LLWR (Subsection 6.2). Given the difficulty of identifying engineering measures that can provide effective passive control against threats from coastal erosion and inundation, safety arguments in the ESC ultimately centre on the overall acceptability of disposal itself, given the potential radiological implications of disruption. Disruption by future human actions Measures to reduce the likelihood of disturbance by future human actions include the use of barriers and other controls to prevent access to the site, preservation of information relating to its presence and content in order to support planning and legal controls over use of the site, and the use of signage and markers. In general, these are all part of the site management arrangements (Subsection 6.2), rather than being passive engineering controls. Nevertheless, they are potentially significant in terms of minimising the possibility of disruption to the facility over timescales relevant to some of the shorter-lived radionuclides present in the waste. The primary engineering control to reduce the likelihood of disruption is the depth of waste beneath the ground surface. This is essentially related to the thickness of the final cap over the facility, coupled with the depth below surface at which disposal takes place. Markers, or marker layers, built into the cap design (or, indeed, the evident man-made origin of the engineered layers of the cap itself) may also be considered to provide a passive function in alerting inadvertent intruders to the hazards within the facility. The overall aim of such a barrier LLWR/ESC/(R11)/10025 Page 77 of 119

82 would be to minimise the likelihood that disturbance by reasonably foreseeable actions might give rise to radiologically significant impacts. Features of the cap design also play a role in limiting the possibility of intrusion caused by burrowing animals and deep-rooted plants. Release of gases During operations, gas release from the trenches takes place via controlled venting in the interim cap. Potential public exposures are limited by restricting access to the site. A central vent within the final cap enables controlled release and monitoring of gases to continue while public access is denied until the cessation of active controls over the site. The primary engineering controls on gaseous release pathways in the long term are either to create a long diffusion path length for shorter-lived radioactive gases (particularly Rn-222) or, so far as practicable, to absorb or disperse releases of longer-lived gaseous radionuclides (specifically C-14). As far as engineering controls are concerned, both these functions relate primarily to the design and properties of the repository cap. The wasteform also plays an important part in limiting the release of C-14. In so far as gas generation and release from the waste may be promoted by reactions with infiltrating water, engineering measures that are taken to reduce water flows through the wastes (see below) will also play a role in mitigating the threats associated with gas release. Even so, there may already be sufficient moisture within the wasteform to promote the generation of C-14 labelled or other bulk gases. Other (non-engineering) measures that have been considered in relation to controlling gas generation within the waste include those that affect the nature of the disposed wasteform (e.g. treatment of metals and organics) and the way in which particular waste streams are emplaced in the repository (see Subsections 4.2 and 4.3). Generation and release of leachate Engineering controls have the potential to reduce disturbance of the wastes by infiltrating water (i.e. minimising volumes of water entering and passing through the facility), to control the release of radionuclides from the facility (by ensuring that contaminated leachate production is minimised or contained as far as practicable), and to mitigate the consequences of leachate release (by managing the routes via which leachate is discharge to the wider environment). During operations, leachate from both the vaults and trenches is collected in sumps and drainage lines, and controlled discharges made via the Marine Holding Tanks (MHT) and the Marine Pipeline. These controls will cease at some stage, following the completion of waste emplacement, as the disposal system moves from active to passive control on leachate release. The interim cap currently provides an engineering control on the volume of leachate produced within the trenches; as the facility is extended and further vaults are added, additional engineering measures to control leachate production are required to ensure that the capacity of the leachate management system is not exceeded. The function fulfilled by engineered features in controlling the volume of water entering and passing through the wastes (both trenches and vaults) is essentially LLWR/ESC/(R11)/10025 Page 78 of 119

83 one of acting as a barrier to water entry. The most important component in this respect is the engineered cap. However, there is also a need to ensure that the significant volumes of water do not enter the vaults and trenches as a result of lateral inflows. In particular, measures are incorporated (e.g. COW and cap perimeter drains) to ensure that water shed by the cap does not enter the facility around the periphery. Although the cap and any cut-off walls should be designed to provide an optimised approach to minimising water ingress, it remains necessary to control any water that does enter the facility in order to ensure that the possibility for contaminant release is minimised. There are two basic strategies for this, both of which rely on engineering design to ensure appropriate passive control of leachate production (see Figure 5.2) 8. One is to encourage water that enters beneath the cap to have minimum contact with the wastes, by providing preferential flow pathways through the facility in an attempt to minimise waste saturation and opportunity for radioactive contamination. The other (broadly consistent with that adopted in the Vault Single Option study Subsection 5.1) is to attempt to capture the water within the vault and ensure that no leachate is released, for as long as possible. The leachate management philosophy that applies to these strategies is different; in both cases, however, the opportunity to optimise engineering controls for passive leachate management is largely restricted to the design of current and future vaults and not to components of the LLWR where authorised disposals have already taken place. Finally, given that circumstances may arise (e.g. when barriers to inflow have degraded) when it is no longer possible to avoid leachate release, it is possible to conceive of engineering measures that seek to direct the flow of leachate to those parts of the environment where the impact will be lowest. As a general rule, this means avoiding discharges near to surface in the vicinity of the facility, and instead seeking to direct leachate to depth where it can be diluted in larger volumes of groundwater before eventual release the marine environment. Potential passive controls on leachate release include vertical drains and downstream cut-off walls. 8 Note that the schematics in Figure 5.2 are intended for general illustration only and were drawn on the basis of an interpretation and extension of the conceptual design derived in the Vault Single Option selection process (Section 5.1). The detailed design adopted for the purposes of the ESC [5] differs (e.g. in relation to cap profile and cut-off wall depth) from both these schemes. LLWR/ESC/(R11)/10025 Page 79 of 119

84 Figure 5.2 Schematic illustration of alternative strategies for the passive control of vault leachate (from [58]) Analysis of Design Features Based on the above analysis of the potential role of engineering controls in relation to threats to waste isolation and containment, it is possible to identify a set of design features that embody the overall passive safety strategy for the LLWR. These are: The final engineered cap for the facility, including consideration of its role as a barrier to the inflow of water to the trenches and vaults, minimising the likelihood of disturbance by future human actions and bio-intrusion, and controlling gas release. Relevant factors also include the timing of installation and its implications for the balance between active and passive controls on leachate production and discharge; Future vault walls and bases, including their contribution to the control of leachate that may be produced within the facility. This demands consideration of the implications of alternative strategies for leachate control beneath the cap (Figure 5.2) under both normal (as built) and degraded cap performance conditions; Cut-off walls and their potential role in minimising the lateral inflow of water to the trenches and vaults, as well as minimising the possibility of leachate release from LLWR/ESC/(R11)/10025 Page 80 of 119

85 the facility the surrounding environment via near-surface pathways in the event of degraded cap performance; Passive leachate discharge controls, including consideration of the possible role of engineered drainage in routing the release of leachate from the facility, under both normal (as built) and degraded cap performance conditions. In the summary that follows, the options and analysis for each of these design features is described in turn. Cap The cap is a principal component of the overall engineering design for the LLWR. It plays key roles in relation to the passive control of leachate and the protection of disposed wastes from inadvertent disturbance. From a functional perspective, the design of the cap needs to be optimised to minimise infiltration to the extent that it is less than the drainage capacity of the underlying geology, thereby creating unsaturated zones beneath the vaults and trenches, for as long as reasonably practicable. Ideally, this should extend up to the time at which natural erosion processes are expected to disrupt the facility [8]. It also needs to provide effective protection against intrusion by humans, deep rooting plants and burrowing animals. The component layers of final cap design for the LLWR have been optimised in previous design cycles, based on principles consistent with best international practice [5,64]. A specific function has been identified for each layer, together with an appropriate approach to providing that function [64]. The baseline design includes a composite geomembrane on a BES (bentonite enriched soil) hydraulic barrier, overlain by an internal drainage layer, biointrusion barrier and upper layers designed to provide substrate for plant rooting, moisture retention and filtration (see Figure 5.3). The overall approach to cap design for longevity of performance as a hydraulic barrier is consistent with established best practice and experience from landfill disposal design and related facilities. Engineering judgement supporting the design is that long-term barrier performance is unlikely to be substantially improved upon by further design iterations [64]. Infiltration through the cap can therefore be expected to increase slowly over time, while sea-level rise is expected to cause a rise in the regional water table. Consequently, there will be an unavoidable change in levels of saturation, gradually bringing the saturated geology closer to the base of the vaults. LLWR/ESC/(R11)/10025 Page 81 of 119

86 Figure 5.3 Components of the final cap (from [5]) The optimised cap design illustrated in Figure 5.3 has provided the basis for expert elicitation of key parameters used as inputs to the ESC hydrogeological model calculations [63,65,7]. A key principle is that, when the rate at which infiltrating water enters the vault exceeds the rate of leakage through the base of the vault or the underlying natural drainage capacity, the wastes within the vault will tend to become saturated. In the calculations carried out specifically to support the design optimisation process [63], the assumed drainage capacity of the natural geology was 160 mm yr -1. This is shown as a dotted line in Figure 5.4. LLWR/ESC/(R11)/10025 Page 82 of 119

87 Flow (mm/y) Infiltration rate Initial leakage rate Maximum leakage rate 160 mm/y AD 2680AD 3180AD 7000AD Figure 5.4 Comparison of elicited infiltration rates (from [63]) Figure 5.4 also shows the elicited values (and range of uncertainty) for the performance of the cap as a barrier to infiltration as a function of time (i.e. diamond symbols, labelled infiltration rate ). Under these assumptions about cap performance, there is high confidence that the underlying drainage capacity beneath the facility will not be exceeded by infiltration for several hundreds of years after closure. In addition, Figure 5.4 shows the elicited values of the performance of the vault base as a barrier to vertical water flow (i.e. square symbols, labelled leakage rate ). The rate of leakage through the base varies with the head of water above, and therefore according to whether it is assumed that the pore space in the vault system is essentially unsaturated ( initial rate) or fully saturated ( final rate). The implications of such performance for system saturation over time were examined in hydrogeological model calculations supporting the design optimisation work [63] 9. The results (originally carried out assuming the vault and vertical drains conceptual design derived in the Vault Single Option selection process Subsection 5.1) illustrate that the repository system is expected to remain unsaturated for several hundred years after final closure. 9 Design optimisation and performance assessment is inevitably an iterative process and the hydrogeological calculations reproduced here (and in the remainder of this section) are those that informed and guided the design optimisation work [63], which were based on the conceptual design derived in the Vault Single Option selection process [34]. Subsequent work [65] to underpin the performance assessment for the ESC drew on an updated hydrogeological model as well as taking into account the outcome of decisions regarding a preferred engineering design. There are inevitable differences of detail between the results from different stages of modelling work, both in relation to design features and the parameterisation of some hydrogeological features (e.g. in relation to the significance of a changing regional water table). Nevertheless, the general understanding supporting design optimisation decisions remains the same. LLWR/ESC/(R11)/10025 Page 83 of 119

88 Figure 5.5 Calculated saturation profile from 100 to 1100 years after closure for a selected vertical section through the repository (from [63]) The spatially-averaged rate of drainage associated with the natural geology was subsequently revised to 100 mm yr -1 in refinement of the hydrogeological model prior to the main assessment calculations [65]. Nevertheless, the central value of the assumed infiltration rate through the cap in these later calculations remains less than that assigned to the geology for well over 500 years. The cap therefore plays an effective role as a hydraulic barrier, protecting the wastes from disturbance, for a substantial period of time. The overall thickness of the engineered cap in the baseline design (Figure 5.3) is 3m. Taking into account that there will be profiling material underneath the cap, this means that there will be a minimum distance of 4m between the cap surface and the underlying wastes. As such, we consider that the cap provides significant protection against inadvertent disruption of both trench and vault wastes as a result of commonplace human actions that might involve bulk excavations, such as house construction [66]. A thicker cap might potentially provide greater protection against disturbance associated with certain types of bulk excavation (e.g. foundations of a high-rise building). LLWR/ESC/(R11)/10025 Page 84 of 119