MT Cross, MI Wareing, C Dixon and EH Perrott
Decommissioning & Waste Management (North) Group
AEA Technology plc
Windscale
Seascale
Cumbria, CA 20 1PF
UK

Decommissioning of the Windscale Advanced Gas-cooled Reactor (WAGR) is a major UK reactor decommissioning project co-funded by the UK Government, the European Commission and Magnox Electric. WAGR was a CO₂ cooled, graphite moderated reactor which served as a test bed for the development of Advanced Gas-cooled Reactor technology in the UK. It operated from 1963 until shutdown in 1981.

AEA Technology plc are currently the Managing Agents on behalf of UKAEA for the WAGR decommissioning project and are responsible for the co-ordination of the project up to the point when the contents of the reactor core and associated radioactive materials are removed and either disposed of or packaged for disposal at some time in the future.

Decommissioning has progressed to the point where the reactor has been dismantled down to the level of the hot gas collection manifold with the removal of the top biological shield, the refuelling standpipes and the top section of the reactor pressure vessel. The 4 heat exchangers have also been removed and committed to shallow land burial.

This paper describes the work carried out by AEA Technology under separate contracts to UKAEA in developing some of the equipment and deployment methods for the next phase of active operations required in preparation for the dismantling of the core structure. Most recent work has concentrated on the development of specialist tooling for removal of items of operational waste stored within the reactor core, equipment for cutting and removal of highly activated stainless steel ‘loop’ pressure tubes, diamond wire cutting equipment for sectioning large diameter pipework and equipment for dismantling the reactor neutron shield.

The paper emphasises the process of adaptation and extension of existing technologies for cost-effective application in the decommissioning environment, the need for adequate forward planning of decommissioning methodologies together with large-scale ‘mock-up’ testing of equipment to ensure confidence during the active work phase.

Decommissioning of the Windscale Advanced Gas-Cooled Reactor (WAGR) is a major UK reactor decommissioning project co-funded by the UK Government, the European Commission and Magnox Electric. WAGR was a CO₂ cooled, graphite moderated reactor which served as a test bed for the development of advanced gas-cooled reactor technology in the UK (Fig. 1.). It operated from 1963 until shutdown in 1981.

Following shutdown the decision was taken to dismantle the reactor. This work has progressed as follows:

• The turbine hall and ancillary equipment have been cleared.
• All fuel has been removed and sent for storage prior to reprocessing. Fuel-related components eg neutron shield plugs, have been conditioned for future disposal and stored in the reactor core. These components have been termed ‘operational waste’.
• The reactor pressure vessel and containment have been isolated from the rest of the primary circuit and a ventilation system installed into the reactor vault and pressure vessel.
• A waste packaging route for both Intermediate Level Waste (ILW) and Low Level Waste (LLW) has been constructed utilising access through one of the four heat exchanger bioshields. This route connects to a custom-built waste grouting facility.
• The top reactor bioshield, reactor refuelling machine, top of the pressure vessel and refuelling standpipes have been size reduced and disposed of as LLW to the LLW disposal site at Drigg.
• The reactor structure above the top of the hot gas manifold or ‘hotbox’ has been removed and size reduced.
• A Remote Dismantling Machine (RDM) has been installed over the reactor.
• The four heat exchangers have been prepared, lifted clear of WAGR and committed to shallow land burial at Drigg.

The next phase of operational work is to remove and dispose of the remaining reactor internals and the pressure vessel.

Initially this will involve a combination of semi remote and remote operations to prepare the reactor for fully remote operations using the RDM.

This paper describes the work carried out by AEA Technology under separate contracts to UKAEA in developing some of the equipment and deployment methods for the next phase of active operations required in preparation for the dismantling of the reactor core structure. Most recent work has concentrated on the development of specialist tooling for:

• Sectioning large diameter pipework using a diamond wire cutting tool.
• Equipment for the removal of items of operational waste stored in the reactor core.
• Customising standard equipment for size reducing and removing the hot gas manifold (the "Hotbox").
• Equipment for the removal and cutting of the highly activated stainless steel ‘loop’ pressure tubes.
• Equipment for dismantling and removal of the complex steel and graphite structures which form the reactor’s inner and outer neutron shields.

The following sections underline the need for adequate testing of equipment before deployment in the remote environment and, in particular, the use of large scale mock-ups to simulate in-reactor operations. This strategy has proven to be highly effective in increasing confidence in systems particularly in the development of dismantling methodologies to reduce operational timescales and operator dose uptake. Lessons learnt have been valuable in directing future work.

In order to simplify subsequent remote dismantling operations, using the RDM and dismantling manipulator, a series of operations using manual or semi remote tooling have been performed on WAGR. These operations are known as the ‘preliminary operations’.

The most significant of these operations was the severing of the four mild steel ducts which run from the hotbox to the heat exchangers. Each duct consists of a 685 mm diameter cylindrical section, with a 25 mm wall thickness, lined with a 25 mm depth of stainless steel foil insulation. As a result of burst fuel experiments, on completion of reactor operations, the insides of the ducts has also become heavily contaminated with Cs-137.

One method of cutting the ducts would have been to use a thermal technique such as oxy-fuel gas cutting or plasma arc, but this would have been difficult to accomplish in the confines of the reactor. It would also have generated significant secondary waste in the form of dross and aerosol and caused the caesium in the ducts to volatalise. To overcome this, equipment which uses a diamond impregnated wire was developed. The system is capable of semi-remote operation such that, for the duration of each cut, there is no requirement for the operator to be positioned in the radiation field around the reactor.

The equipment consists of: (Fig. 2.)

• A frame which is clamped to the outer coaxial duct attached to the pressure vessel wall.
• A hydraulic, motor-driven diamond cable to cut through the duct.
• A hydraulic ram to feed the diamond cable onto the duct.
• A liquid nitrogen cooling system to prevent the wire from overheating during cutting operations.

Cutting operations in the reactor using this equipment have now been successfully completed, with cutting of individual ducts taking approximately four hours with minimal wire wear. Higher cutting speeds were achieved but these resulted in damage to the diamond beads.

Operational waste is the term applied to all ancillary equipment sited in the fuel channels in WAGR, which could be handled using the original fuel handling equipment. During the initial stages of decommissioning, after Stage 1 defuelling, all the operational waste was removed, size reduced where necessary, and the ILW items returned to the reactor fuel channels for storage. Examples of operational waste are neutron shield plugs, auto control rods and arrestor mechanisms.

To recover the operational waste from the reactor a custom built grab has been designed. This grab has been based on the original refuelling machine handling equipment to minimise development costs.

The grab will be deployed from the RDM, 3 te slewing beam hoist via a device which prevents the hoist cables splaying out and keeps them within the confines of the fuel channel.

Extensive trials using this equipment have been carried out in AEA Technology’s test facilities to assess the reliability, safety features and to optimise waste recovery and handling procedures. This grab has also been deployed in the reactor to recover waste items for inspection purposes.

The hot gas manifold or ‘hotbox’ (see Fig. 2.) is a large flat cylindrical vessel situated near the top of the WAGR pressure vessel. It is approximately 5 m in diameter and 1 m high and has a wall thickness varying between 25 mm and 32 mm. It is penetrated by the 247 refuelling channels and 6 loop tube channels and is a complex structure of mild and stainless steel. It is also insulated internally with stainless steel foil known as ‘Refrasil’. Its purpose was to distribute the hot coolant gas emerging from the reactor fuel channels to the four heat exchangers.

The WAGR hotbox is an interesting item with regard to its radioactive inventory. It sits above the top core reflector and neutron shield, and, as such, has received little neutron activation. It is, however, the first permanent reactor component in which the coolant gas came into contact with after passing over the operating fuel pins. It is hence the first site for deposition of fission products which were from time to time released from failed fuel pins.

The initial strategy for size reducing the hotbox was to use oxy-propane powder injection cutting, but following trials it was assessed that the quantity of dross produced, and the release of contamination, would have an adverse effect on the subsequent dismantling of the neutron shield. Also the fume generated would spread contamination throughout the reactor which would result in an increased dose uptake to operators performing operational tasks and maintenance operations.

Therefore the proposed method for dismantling the hotbox involves a combination of remote and semi remote operations and aims to minimise the effects on subsequent dismantling operations.

The current strategy for dismantling the hotbox divides the structure into several areas:

• Severing the structure from the neutron shield.
• Top plate dismantling.
• Cutting and removal of the box internals.
• Side wall cutting and removal.
• Bottom plate dismantling.
• Removal of the burst cartridge detection pipework.

For these operations it is proposed to use a combination of a diamond tipped saw, grinders, a small hydraulic shear and a controlled plasma arc cutting system.

Some of this equipment will be deployed by personnel setting up the tools and then withdrawing from the area to operate it semi remotely. Other pieces of equipment will be deployed and operated using a combination of the RDM, manipulator and 3 te slewing beam hoist.

This proposed strategy of combining remote and semi-remote operations offers the most efficient and cost effective solution for dismantling the hotbox whilst minimising the impact on subsequent dismantling tasks.

Trials to optimise cutting parameters and equipment performance and to minimise operator dose uptake are currently in progress.

Six ‘loop’ pressure tubes were inserted into the core of WAGR to enable fuel performance experiments to be conducted at the full coolant pressure of the Civil Advanced Gas-Cooled Reactor (CAGR), 600 psi. The loops are constructed in stainless steel and are now highly activated accounting for around 25% of the reactor inventory. The complex design and necessity to avoid spreading highly activated secondary wastes within the reactor structure has necessitated the development of a ‘cold’, swarfless cutting method to be developed.

A number of different ways were originally considered for removing the loops. The initial methods were to use the reactor refuelling machine or a heavily shielded facility constructed on the reactor cap. The loops are so constructed that they could be drawn into such a facility and then size reduced ready for disposal. For various reasons, including the lack of a suitable repository for the cut tubes, a decision was taken to delay loop removal until after the RDM had been installed. As a consequence the loops must now be removed using remotely-deployed tooling mounted on the reactor internals.

A removal methodology has now been formulated, which is to raise each loop in its entirety into a purpose-built cutting system mounted on the reactor neutron shield and then to handle the cut sections using grabs attached to the RDM, 3 te hoist. To undertake this work the equipment below was identified:

• An external tube grab to transport the cut sections of loop from the reactor to the waste packaging route. The grab will be deployed from the RDM hoist.
• An internal tube grab to initially raise the loop section into a stand-alone lifting system mounted on the reactor internals.
• A lifting system which will lift the tubes out of the reactor in discrete lifts and cannot fail in such a way that a tube remains jammed in the device, or that a tube is accidentally released. This system releases the RDM hoist to allow transfer of the cut tube sections.
• A system for cutting the loop tubes, which is capable of cutting the tube while producing minimal secondary waste and reaction forces.

A custom-built cutting and handling system which meets the above requirement has the following key features.

The internal and external tube grabs are direct adaptations of the expanding mandrel and scissor-type, plate grabs used in previous decommissioning campaigns. The lifting equipment uses ‘lazy cams’ to grip the tubes since these provide both a degree of compliance in operation and are fail safe.

The cutting system has been the most difficult system to develop due to the requirement for reliability, minimal maintenance and control of secondary waste production. Extensive trials have been carried out on plasma cutting systems, swaging systems and systems for hydraulically shearing the tubes. Of the systems tried only plasma cutting and hydraulic shearing proved reliable at cutting the tube sections.

Plasma cutting was eventually rejected on the grounds that the amount of secondary waste produced as fume could recategorise other reactor components from LLW to ILW by deposition of particulates generated during thermal cutting of the highly activated loops. Hydraulic shearing could cut the tubes but generated large sideways forces and massive distortion of the tube end section, causing problems with grab placement. Experience suggested that filling of the tubes with grout prior to cutting could minimise end section deformation and reduce overall cutting forces. This was found to be the case and is now the adopted approach.

To reduce the radiation exposure of operators working on the reactor ‘pile cap’ and to minimise neutron activation of the reactor components above the reactor core, a neutron shield was installed in WAGR. The neutron shield was installed directly above the core reflector and is divided into two distinct regions known as the inner and outer neutron shield. Due to the operation of the reactor, both structures have been subjected to a neutron flux and hence trace elements within the materials of construction have been activated.

The inner neutron shield is a construction of graphite and steel components with refuelling standpipes running through it. Each standpipe has, at the top, the restrictor sleeve components which connect to the hotbox and, at the bottom, the stools which connect it to the reactor core. Between these two ends and arranged on the standpipe are three graphite bricks and two graphite spacers, brick layers 1 and 3 having intersecting boron steel plates. The top layer (3) also has a layer of 12 x ½" mild steel thermal shield and 1 x ¼" boron steel plates doweled into the top of each brick.

The outer neutron shield graphite is made up of seven layers of graphite bricks. Layer 7 is the top layer, layer 1 the bottom layer nearest the core. There are 72 bricks in each layer, giving a total of 504 bricks for the assembly. The bricks are located on the 25.4 mm (1") diameter tie bars which pass through holes within the bricks.

Layers 7 and 6 bricks are slightly large and overlap those of the inner neutron shield to prevent neutron streaming. This means that layers 7 and 6 of the outer neutron shield require removal prior to the removal of layers 2 and 1 of the inner neutron shield. The bricks, when assembled, have a minimum of 1.3 mm (0.05") clearance between adjacent bricks in the same layer.

The dismantling plan for the inner and outer neutron shield can be sub-divided into 31 tasks and will follow on from dismantling the hotbox. Dismantling of the inner and outer neutron shield will be a top down, layer-by-layer approach.

To dismantle the neutron shield graphite bricks a number of purpose-built grabs have been developed. These grabs consist of vacuum grabs which can pick up on a flat surface of a brick. Ball grabs which use the internal bore of a brick to lift from. There are also a series of grabs which have extended arms which can locate under the base of the bricks to achieve the required lift. To ease the removal of the inner neutron shield bricks, staged cutting of the standpipes will be carried out. Similarly, the tie bar s on the outer neutron shield will be cut at stages to ease the removal of graphite bricks in the outer neutron shield. To perform the standpipe cutting an internal plasma cutter has been developed which has an integral grab to remove the waste item after it has been cut. The tie bars will be cut using a grinder deployed by the RDM manipulator.

Using a combination of these grabs and cutting tools, in conjunction with purpose-built pallets for stacking the waste items, the neutron shield can be effectively dismantled.

The conclusions that can be drawn from this work are:

• Modification and adaption of existing technology, where possible, gives the most cost effective solution to some dismantling tasks.
• Detailed risk and safety assessments of the proposed dismantling strategies allow the high risk areas to be reduced or avoided resulting in greater confidence of the final solution.
• ‘As-built’ information should be obtained and surveys of the area should be carried out wherever possible. This will help minimise modification to equipment. Where accurate information cannot be obtained, a degree of flexibility should be incorporated into the design of dismantling equipment.
• Manual and semi-remote dismantling are quicker and generally cheaper than fully-remote operations and offer a high degree of control over the dismantling task. Where radiation fields permit, these options should be used in preference to remote ones providing ALARP criteria are met.
• The use of a large scale mock-ups for both equipment development and operator training is essential to the success of a decommissioning project. These mock-ups must closely mimic the actual working environment. Although detailed mock-ups are expensive to produce in the first instance, they generally will result in an overall cost saving and in increased confidence in the equipment and methodologies being proposed.

The work reported in this paper has been funded by UK Government, the European Commission and Magnox Electric.

BACK