A ponderous hazard

10 August 2006

Cooling ponds built for the UK’s Magnox programme now pose one of the most difficult cleanup challenges in the world. By Steve Topping and Steve Bruce

The First Generation Magnox Storage and De-canning Facility (MSDF) was constructed in the mid 1950s at Sellafield and performed a vital and integral role in the UK civil nuclear power programme.

It operated safely for nearly 30 years, storing irradiated Magnox fuel in a concrete open air pond before stripping the fuel of its cladding (decanning) prior to reprocessing at a separate Sellafield facility. The advent of commercial-scale nuclear power demanded recycling facilities on a large scale – able to accept 500 fuel elements per day – to run the 26 Magnox reactors across 12 sites in the UK.

During operational service a massive 27,000t of fuel was stored, decanned and then exported from the plant. It received its last batch of fuel in 1992 before entering its post-operational cleanout (POCO) phase.

However, its unique cleanup challenges were created in the mid 1970s due to a lengthy and unforeseen shutdown at the Magnox Reprocessing Plant and also a vastly increased throughput of fuel due to electricity shortages.

These factors caused the spent fuel to be stored in the pond for longer than the designed period, resulting in the corrosion of the fuel’s magnesium oxide cladding and degradation of the fuel itself. Ultimately this led to increased radiation levels and extremely poor underwater visibility in the pond. These difficulties slowed the rate of decanning, increased residence times for spent fuel and so created a perpetual delay to throughputs.

A number of steps were taken to counter these problems, including washing the fuel before decanning to remove sludge, and the use of zeolite ion exchange resins in skips to reduce local radiation levels in the pond water and in discharges of liquid effluent. However, these were only partially successful and MSDF continued to operate under difficult conditions until its replacement, the Fuel Handling Plant (FHP), was commissioned in 1986.

The pond water in MSDF was purged to assist in maintaining water quality by lowering the concentrations of radionuclides and improving visibility, with the purge water being passed through large settling tanks in an adjacent facility to remove particulate sludges prior to discharge.

In the mid 1980s, a replacement facility was built to cope with the greater throughputs of pond water. Both liquid wastes and sludges were transferred to the Site Ion Exchange Effluent Plant (SIXEP) which treated the purge water prior to discharge, reducing the environmental impact. The plant also provided some capacity for receipt and storage facilities for the sludges.

In the absence of waste treatment and disposal routes, legacy ponds, in particular MSDF, accumulated significant inventories of waste materials, including sludges from the corrosion of fuel cladding, fuel fragments and other debris during operating life. Although these practices were regarded as entirely acceptable at the time, their effects now pose a number of challenges to cleaning out and decommissioning the plant.

The facility currently contains large quantities of spent fuel, sludge and other materials that are classed as intermediate-level waste (ILW). Overall remediation and cleanup poses some of the most complex radiological challenges that exist in the world.

The safe and accelerated decommissioning of the facility is very high on the national agenda of the Nuclear Decommissioning Authority (NDA), which owns all civil nuclear assets in the UK, including Sellafield.

MSDF is one of several higher hazard facilities that exist onsite that the NDA has publicly stated are its number one cleanup priority. The cleanup of this facility will result in the sizeable step forward in hazard reduction that is necessary to make the site acceptably safer for future generations.

The estimated lifetime cost for this vitally important cleanup project is approximately £1 billion.


At the time of its conception and construction, MSDF was never intended to be in operation, nor to hold such an inventory in such a degraded condition for so long.

Considering that both these factors had been significantly exceeded, as the plant entered into its POCO phase in 1992, enormous attention had to be paid to assessing how the building’s infrastructure needed to be brought up to date so that decommissioning work could commence. This identified the following priorities for investment prior to full-scale decommissioning work being undertaken:

  • Risk reduction activities. Improving the integrity of the external drainage system; improving secondary containment; improving the integrity of the building fabric.
  • Contingency plans. Deployment of specialised equipment to prevent a loss of pond water and inventory.
  • Essential asset restoration and upgrade. Re-cladding the building; new lightning strike protection; surveys of power and control systems; upgrades of fire detection, monitoring equipment, ventilation to effectively tackle airborne contamination.
  • Preparation for retrieval activities. Identifying, preparing and trialling routes for sludge, fuel skip and fuel retrievals.
  • Early remediation. Where possible using existing facilities and technologies to accelerate the cleanup process.
Since 1992, work has been carried out to address the first three phases in order to prepare for retrieval activities. With the creation of the NDA and British Nuclear Group (BNG), not to mention an increasingly vigorous debate on the future of nuclear energy in the UK, the focus on accelerated decommissioning and cleanup has sharpened considerably.

In terms of the priorities for the overall cleanup of the pond, there are four elements, each of which are interdependent:

  • Sludge removal and processing into a safe, passive state for interim storage.
  • Fuel removal for reprocessing or encapsulation as waste and placement into storage.
  • Skip retrieval for decontamination, size reduction and encapsulation and interim storage.
  • Removal of miscellaneous waste for size reduction and encapsulation and interim storage.
Considering its potential capacity to become mobile, sludges represent the biggest radiological risk and are therefore the primary focus of cleanup work.

Regulators, the Nuclear Installations Inspectorate (NII), placed licensing specifications on BNG that required the sludges to be contained in modern stainless steel containers, thus removing the overall risk in the ponds.

To this end, the BNG project team recommended a buffer tank for interim storage with subsequent processing capabilities. The recommendation was to construct a local sludge containment facility in the short term, to enable the swift removal of the sludge into modern containment. This facility would be followed by the construction of a treatment facility that could encapsulate and prepare the sludges for safe storage as ILW.

In anticipation of the decision being made by the NDA to build these facilities, the demolition and clearance of redundant buildings adjacent to the MSDF was completed. This created a space which could house the sludge containment and treatment facilities.

To optimise productivity, multiple work streams were run simultaneously with sludge retrieval preparations. Strategies also have had to be implemented for addressing retrieval and encapsulation of fuel elements, fuel skips and miscellaneous wastes arising from past operations.

To this end, a focus has been on saving the taxpayer money by using existing facilities to process and store the contents of the MSDF more quickly than was originally planned. For example, rather than construct a purpose-built facility to take skips and miscellaneous wastes, it is proposed they be removed and decontaminated for encapsulation as ILW and stored in the modified Boxed Encapsulation Plant (BEP), which has already been constructed.

Additionally, pilot programmes have been initiated to demonstrate retrieval options. The best example of this type of approach has been the recent trial retrieval of fuel from MSDF by BNG engineers, which proved that the taxpayer can get better value for money and accelerate hazard reduction in nuclear cleanup by using technical innovation and making minor alterations to existing facilities. This trial provided an opportunity to further evaluate processing an additional 100t of fuel through existing facilities.


The NII considers MSDF to be the country’s highest priority in terms of hazard reduction, a view supported by the site licensing company and the NDA.

Remotely operated submersible vehicles (ROVs) were used by BNG to assess the condition of stored fuel in MSDF. The survey showed fuel condition to be better than expected and engineers were able to prototype retrieval on a selected skip of fuel.

The remotely operated vehicle could clarify the positions of the various skips

The remotely operated vehicle reveals the situation of the spent fuel in the MSDF fuel storage pond

The retrieval and subsequent export of the spent fuel to the FHP was executed in November 2005 and was completed successfully. Subsequently the fuel was to reprocessed using the Magnox Reprocessing Plant. This operation has demonstrated that BNG can:

  • Safely retrieve fuel from its legacy ponds.
  • Safely transport retrieved legacy fuel between facilities.
  • Eliminate the hazard presented by this legacy fuel by use of existing onsite reprocessing facilities.
This in turn would enable larger-scale fuel retrievals from MSDF, subject to regulatory approval, up to seven years ahead of the current plan, which assumes new plants would be available to handle all arisings from the legacy ponds in 2015.

Poor visibility and high radiation and contamination conditions exist in the pond creating unique difficulties for the inspection and monitoring of projected inventories, the building structure and radiological inventory by conventional means. BNG capitalised on available marine-based technology to confirm the inventory and assess the current condition of the fuel, Magnox sludge, storage skips and debris in the pond.

The completed survey delivered a phenomenal amount of new information on underwater conditions in the pond. This information on fuel, skips, sludge, debris and radiation profiles is currently being analysed by BNG to optimise the retrieval of materials and reduce the overall hazard and risks of the ponds. Additionally, the underwater surveys have provided Euratom with the necessary information to validate fuel storage records.

Most interestingly, since the fuel in the facility had been underwater for between 20 and 40 years, in the absence of any definitive information, it was assumed to be too badly corroded for reprocessing in existing facilities. The DVD footage showed clearly that some fuel was in a condition worth retrieving and reprocessing in such facilities.

The current lifecycle baseline plan for Sellafield assumes that the fuel, sludge, skips and debris in the ponds are all nuclear waste, and waste streams to deal with this waste are being developed. However, the plan would not deliver an end facility to receive and treat the waste streams until 2015.

BNG identified that a portion of the fuel in MSDF would be suitable for reprocessing, allowing it to be recovered and routed through existing Magnox reprocessing facilities, thereby reducing the radioactive inventory and pond hazards.


In March 2005 the BNG team initiated a pilot scheme to retrieve a single skip of fuel from MSDF and export it to FHP with a view to further assessment of condition and, possibly, transport it onwards to the Magnox Reprocessing Plant for final reprocessing. This pilot would prove that at least some fuel was retrievable; some of that fuel could be reprocessed; existing onsite assets could be used now; and hazard reduction could be initiated prior to 2015 when an interim treatment and storage facility would become available.

A suitable fuel skip was identified for retrieval, primarily chosen because it was recoverable into an existing inlet cell, Inlet Cell 3, but also because it contained fuel which the ROV footage showed to be in good condition.

The inlet cell was refurbished to enable it to receive the skip – all three inlet cells had been externally stripped out as part of a long-term refurbishment programme to convert the inlet cell building into a modern-standards export facility prior to pond remediation. Vital components were re-installed to provide for skip removal.

Therefore, vital components such as the skip hoist, grapple and flask lid lifter were inoperable because their drive mechanisms, controls and interlocks were no longer present.

Additionally, the cell refurbishments had to be completed during a very narrow window of opportunity – the retrieval period was limited to just eight months in line with the overall critical path to ultimate cleanup.

The first task therefore, was to determine specifically what had to be refurbished or modified in the Inlet Cell 3 to allow the skip removal. This was done by employing a one-off retrieval to determine fit-for-purpose systems. A minimalist approach was taken using simple but reliable equipment that could be outfitted quickly. This provided an opportunity for the efficient, safe and expedient removal of a skip of fuel.

Safety case and engineering

The approach to the safety case was also to ensure that it was fit-for-purpose, maximised the opportunity for supervisory control, and used simplistic and safe equipment, with less reliance on the type of engineered systems more suitable to routine operations over many years. However, the safety case was still subject to the normal peer review and approval process with oversight by the UK regulator.

New lifting devices for skip hoist and flask lid lift functions were designed, substantiated, fabricated and installed, as was a new manually operated hydraulic pump arrangement to operate the skip grapple fingers.

Extensive use was made of the existing building’s electric overhead travelling crane as a lifting beam from which to suspend the above devices. New high quality but readily available established technology, such as cameras and lights, were installed in the cell, giving better coverage than before of in-cell operations.

The existing flask bogie and shield doors were mechanically intact and were refurbished to an appropriate standard for the job. The ramp bogie, which carried the fuel skips up and down a ramp into the storage pond, was mechanically intact. An existing hand-wind feature through a reduction gearbox was identified and successfully tested – a decision not to re-energise the electric motor on this device was taken on the basis that it took away operator control on a sensitive operation.

By the time the safety case had been fully approved, Inlet Cell 3 was ready for inactive commissioning. A rigorous set of integrated plant and equipment tests were designed and incorporated into commissioning schedules – these tests used a clean flask and clean test skip loaded to simulate the fuel in the target skip. These tests were completed in a matter of days and an inactive commissioning report was produced.

This report identified that once initial plant and equipment positioning and alignment had been executed, the dynamic testing of the cell as an export facility proved the operation to be reliable and repeatable. The report was approved and the project was set to move into the retrieval phase.


The only part of the operation which could not be tested during authorised inactive commissioning was the grappling of the target skip in-cell. This was an active test which could only be done when the retrieval was actually made. In addition, new information from a more enhanced survey of the ramp using a smaller ROV identified debris on the ramp bogie rails.

A coil of redundant cable resident in the pond had become entangled in the ramp bogie mechanism five metres down in the pond. In addition some sturdy protective PVC bags, used to protect electrical junction boxes in-cell prior to high pressure water jet decontamination operations, had been blasted off and into the water in the cell. These were in the way of the ramp bogie and fuel skip and, if not moved prior to the retrieval, could:

  • End up in the skip and render it unacceptable to the receipt plant.
  • Interfere with the winch mechanism and threaten the operation.
  • Mask a lifting feature on the target skip and make it impossible to grapple.
A safe and conservative decision was taken to remove the debris before commencing the retrieval. The resolution of the debris issues required more safety documentation to be produced. This had to be submitted to safety committees, approved and formally issued before any work could start.

To remove the cable hazard, the ramp bogie was moved off the end stops by one metre and monitored by a camera on a long pole. A six-metre-long lightweight telescopic window brush shaft was used to manipulate the cable clear of the rails and bogie.

To remove the bag hazard, men on lanyards stood at the open cell door and used a similar device to relocate the bags into a safe area. The decision to take this conservative approach delayed the retrieval by two days.


The retrieval itself was executed flawlessly in the course of one work shift. There was some tension during the skip grappling operation as the grapple needed to be lowered 150mm more than it had been during inactive commissioning.

The designers were adamant that this was allowed for in the design, however until the skip was grappled this could not be physically proven. In the event the whole operation was very comfortably completed.

Camera coverage and control were excellent with clear views of each grapple finger finding full engagement prior to the lift. Radiation surveys were conducted throughout and no increase of any concern in the working areas was detected, although the in-cell gamma detectors recorded the significant radiation from the fuel as it arrived in the cell.

The skip was introduced to the flask and lidded in-cell; the radiation readings went to background and the retrieval was completed in a single workshift.


Export of the 50t flask was executed over the following two days. With the flask being contained in a PVC bag for the entire operation, extensive radiation and contamination surveys were successfully completed.

The flask was craned onto a low loader and transported by road to the FHP – this in itself requiring meticulous planning. The consignment was received by FHP and immediately removed from the flask and loaded into its storage pond without issue.

The next step involved an assessment of the condition of the fuel at FHP, which is a modern, indoor facility with all the necessary systems and equipment to manipulate and assess the consignment. The fuel was assessed as being suitable for reprocessing as had been expected, and it will now be loaded into a magazine and transported to the Magnox Reprocessing Plant.

At the reprocessing plant it will be rendered into reusable nuclear materials and returned to the UK reprocessing fuel cycle. Back at MSDF, work continues to assess how much of the inventory is exportable to FHP.

With this trial proving to be successful a potential opportunity now existed for approximately 100 skips of consolidated fuel (ie maximum payload) to be exported to FHP over two years.

The scale of the implications and opportunities resulting from this trial are very promising for reducing some of the pond hazards earlier than anticipated. MSDF may have a viable route out for a considerable percentage of its radioactive inventory, using existing onsite assets and technology. This route could be available to commence inventory reduction in the very short term, as much as eight years ahead of the current cleanup schedule.

Significant and early hazard reduction would result in what is deemed one of the higher hazard facilities by the UK government – making a massive impact on what is the largest cleanup challenge in the world – Sellafield.

Submersible remote operated vehicles
Remotely operated vehicles (ROVs) have been successfully deployed in recent years in assessment of undersea conditions, notable examples being the explorations of deep water wreckage of Titanic and Bismarck, not to mention rescue missions to the stricken submarine the Priz in August 2005. Technological advances have combined improved digital photography and vehicle control, with miniaturisation of components which provided highly effective ROVs ideal for applications where manoeuvrability in tight spaces is required. Fuel storage ponds are such an application and British Nuclear Group has successfully deployed ROV technology to survey the entire MSDF complex and inventory. This produced more than 5000 hours of detailed footage plus all the information required to develop the pond’s management plan for the required sequential activities to remove the waste products.

Fuel receipt
Flasks containing skips of Magnox fuel were lifted from a loading bay into one of three shielded inlet cells. The cell doors were closed, the flask lid was remotely removed and the skip placed on an inclined trolley which travelled on a ramp to the bottom of the pond using a winch mechanism. Flasks from the Calder Hall contained baskets of fuel instead of skips. In this case, the fuel elements were discharged down a chute where they were loaded into a fuel skip by remote handling tongs. The empty baskets were washed and decontaminated before returning to the Calder reactor to pick up more fuel. Empty skips, created following decanning of the fuel, were loaded back into the flask, which, after quality assurance checking, were returned to the power station.

Pond storage
Full skips at the bottom of the inlet cell ramps were removed and placed at a predetermined grid position in the pond using a skip handling machine. A crane with special grapples to pick up skips moved them across the pond and placed them down in predetermined locations. There were approximately 1000 grid positions stacked up three high. The original pond was extended when it became evident that more capacity was required to meet Magnox fuel imports and storage demand. These were all interlinked.

Skip washing
During extended storage periods in the ponds the fuel began to deteriorate. When the fuel corrosion problems became evident, a skip washing machine was installed in the pond. This machine was a lidded box into which the fuel skip was placed and the box was rotated slowly. The sludge was washed out of the skip and passed to a settling tank before being discharged to the Site Ion Exchange Effluent Plant (SIXEP).

The washed fuel skip was moved by the skip handling machine onto a trolley at the decanner skip transfer area. The trolley was moved hydraulically into an underwater bay. In this area, operators, using remote handling tongs removed each element from the skip and placed it on an elevator. The elevator raised the element above water into a shielded decanning cave. Inside the cave, each element was passed through a slitting head which peeled away the Magnox can from the uranium fuel rod. The pieces of can, or ‘swarf’, were swept up a chute by a sledge into a swarf bin. These bins of swarf are transferred to the intermediate-level waste silo, or the new Magnox encapsulation plant, in a swarf flask. The uranium rod was then fed into a magazine bundle. When 38 rods had been fed into the bundle, the magazine was winched up into a shielded, bottom opening flask on the roof of the cave. The flask was then transferred by crane to a road transporter, which took it to the reprocessing plant. Today, most of the fuel in the pond is corroded to such an extent that the Magnox can is absent.

FilesIllustration of MSDF

Full_length_of_pond Full_length_of_pond
ROV_skip ROV_skip
ROV_annotated ROV_annotated
B30Inlet_FlaskPrepArea_ B30Inlet_FlaskPrepArea_
B30DecanningCaves B30DecanningCaves
Illustration of MSDF Illustration of MSDF

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.