Debbie Williams, Andy Hurley and Stuart Bowe
BNFL

The Magnox fuel storage pond and decanning facility at BNFL, Sellafield, U.K., was built in the 1950’s for the purpose of receipt, storage and decanning of irradiated Magnox fuel. The main liquid effluent discharge routes were via purge and sludge pipe lines to settling ponds. These large diameter cast iron pipe lines are contained within an open trench which is approximately 125m long by 1m wide by 3m deep.

Major failure of the pipework will lead to loss of pond water and exposure of Magnox fuel and radioactive sludges leading to a serious site incident. High radiation levels of up to 7 Sv/hr, corrosion and the possibility of failure during pipe removal has led BNFL to a non intrusive method of reducing the risk to the site.

This paper describes how a solution was determined by applying project management, safety assessments and Value Engineering techniques.

A number of studies have been undertaken and encapsulation with a cementitious material is considered the best option. This technique can be implemented within the required timescale and will significantly reduce the risk of an uncontrolled leak of pond water by improving pipework integrity and eliminating impact damage. Connections into the pipes will be sealed in the longer term as sumps become accessable during clean out of the building.

Full scale inactive trials have been carried out to develop an acceptable cementitious material, which can be safely deployed, to ensure that no additional hazards are introduced and that emergency procedures are robust during placement.

This project demonstrates a method of risk management in a high dose nuclear environment as an intermediate stage prior to decommissioning.

British Nuclear Fuels plc (BNFL) owns and operates the nuclear reprocessing facility, Sellafield in North West England. The role of Sellafield as an integral part of the British Nuclear programme began in 1952 and since then several generations of reprocessing and support plants have been operated successfully.

The Magnox fuel storage pond and decanning facility at Sellafield was built in the 1950’s for the purpose of receipt, storage and decanning of irradiated Magnox fuel as the first stage of the reprocessing operation. Decanning operations were originally carried out under water in large concrete tanks (Wet Bays) prior to the provision of the dry decanning caves in the mid 1960’s. The facility consists of three reinforced concrete ponds, and a later extension, arranged in an east west direction with a fuel inlet building to the north side and the decanning complex to the south.

As part of the liquid effluent system the South Active Drain Trench (SADT) runs along the whole length of the south wall of the facility carrying purge and sludge pipe lines. These were connected to a settling tank, which was used for removing solids prior to further treatment and discharge to the sea.

Use of this system discontinued in 1986 when a new settling tank was commissioned.

The purge and sludge lines are generally a mixture of 8" diameter spun and cast iron flanged pipe using neoprene sealing rings. At the extreme west end of the trench the 4" down pipe includes approximately 300mm long mild steel spool piece, which has replaced the original spun section.

Several schemes have been considered to reduce the risk of failure of these pipes during the implementation of the programme to clean out the facility prior to full decommissioning. This paper describes how a solution was determined by applying Project Management, safety assessment and Value Engineering techniques.

The purge line was a means whereby the top layer of pond water was continually extracted to the settling tanks; the pond water level was kept constant by make up water. This procedure cleaned the top layer of the pond and worked well.

Corrosion of the fuel and cladding resulted in corrosion product sludge being deposited on the floor of the pond. The sludge line has connections into various sumps in the bottom of the pond and bays. It was provided to allow this heavy particulate matter to be transferred to the settling tanks.

An incident which occurred in 1986, in which a badly corroded 2" diameter mild steel feeder line was fractured resulting in a loss of contaminated water into the SADT, indicated the need for the SADT to be made safer i.e. removing the potential for major incidents arising from the trench and pipework.

The redundant pond purge and sludge lines which run in the SADT are essentially below ground level. Connected into these lines are many branches from the storage ponds, decanning and wet bays. (See figure 2). The SADT is 3m in depth, 1m wide and 125m in length. The base of the SADT and 30cm of the wall sides are covered with an impervious material. The trench is covered with lightweight thermally insulated metal covers along its length, apart from where the SADT runs under the road and certain areas where the original concrete covers are still in place.

Radiation dose rates vary along the length of the trench and are of the order of 6 mSv/hr rising to 7 Sv/hr on the trench top. Shielding of the SADT is provided by concrete blocks placed in front of the trench. Access is provided to the SADT by gates built into the concrete shield wall.

The SADT pipework is over forty years old and is showing signs of corrosion (loss of up to 50% of wall thickness) and loss of integrity.

Major failure of the pipework will lead to loss of pond water and exposure of Magnox fuel and radioactive sludges leading to a serious site incident. A comparison of risk has been carried out showing that a ‘do nothing’ option is not acceptable and that a strategy to reduce the risk of pipework failure is required.

Continuous dialogue with both the Regulators and the customers has been maintained throughout the project life.

Initial work was carried out to improve the emergency arrangements and reduce the consequences of a pipe failure. Pumps were installed to re circulate water back to pond. The risk of damage was reduced by removing ~ 250 reinforced concrete slabs (3’6" long by 12" wide by 4" deep and weighing 80kg each) from the trench top and replacing them with lightweight thermally insulated covers. The slabs were supported on corroded mild steel angle iron ledges. A few still remain where radiation levels are too high for man access.

The strategy was then developed to remove the sludge and purge pipe lines from the trench therefore taking away the potential for catastrophic failure of either line. Between 1989-91 various lines were re routed and sealed off to allow this work to proceed, and in 1992 work commenced on manually removing the purge line. It was intended that the purge line would be removed first:

• Because of its position in the trench (above the sludge line)
• So that experience would be gained in this sort of work prior to removing the sludge line (which was thought to have significantly higher dose rates associated with it compared to that of the purge line).

However, although the sludge line had been covered with water filled shield bags to allow workers to stand on them and remove sections of purge line, radiation levels of over 20 mSv were detected, meaning working time on the line would be less than one minute.

The strategy was reviewed and a remote cutting system was developed. Between May and December 1994, the pond purge feed lines and the 8" purge lines for three of the wet bays were removed, some 35m out of a total of 125m.

A planned major project review (as agreed with the Nuclear Installations Inspectorate (NII)) was carried out between January and May 1995 which concluded that progress made to date on removing the purge line was unacceptable. Progress was slow because the cutting equipment was less reliable than had been anticipated and the dose rates were much higher than predicted. If the project was to continue using the same methods, it would take until 2003 to remove the risk and would lead to much higher dose uptake than originally anticipated.

To identify the most appropriate way forward for the project a Value Engineering (VE) study was set up to meet risk reduction objectives. VE is a structured approach to problem, solution and decision making. The following four alternative methods were identified for improving the integrity of the sludge and purge lines.

• Fill or part fill the SADT with a cementitious material, encasing the purge and sludge lines, therefore greatly reducing the risk of catastrophic failure.
• Improve the integrity of the pipework (for example sealing flanges).
• Isolate the in-feeds to the pipes at the pond and bay ends.
• Improve protection around the trench (strengthen covers etc).

Following extensive further studies, the first option emerged as the preferred option on the basis of the following:-

• Encapsulation provided the best means of making the trench safer ie. minimising the potential for a major leakage.
• Encapsulation could be carried out in the limited timescale (as the condition of the pipework required that the trench be made safer as soon as practical).
• Encapsulation it is anticipated, will significantly reduce radiation dose rates emanating from a majority of areas along the trench.
• Encapsulation will protect the pipework from impact damage.

• Covering/sealing the flanges was rejected due to it being an intrusive method with high dose uptake. Also it was considered impractical to carry out this work in the timescale allocated to the project.
• Isolating the in feeds to the pipes from the pond and bay ends was rejected as the project would have difficulty accessing some of the sumps leading to the in feeds due to sludge and fuel skips in the pond and other debris in the wet bays. Also, it was considered impracticable to seal these in feeds in the timescales allocated to the project.
• If the trench was covered with plating to protect the trench and pipework from impact, the pipework in the trench could still fail and the leakage from the pipes would not be inhibited as with encapsulation.
• A ‘do nothing’ option was also considered, however, the risk associated with loss of containment of the Purge and Sludge lines was considered unacceptable and not in common with BNFL safety policy.

Kepner Tregoe (KT is a multi attribute assessment of options) studies were carried out in February and April 1996. These studies combined with the previous VE study enabled a strategy to emerge that would fulfil the broad requirements of removing the risk of an uncontrolled leak from the main pipe lines therefore making the SADT safe. All these studies involved team members who brought their own particular knowledge base, ensuring a broad approach to the issues. In conjunction with this, other work was undertaken to further strengthen the chosen strategy such as:

• Corrosion surveys
• Seismic calculations
• Safety Cases (to be amended to reflect exact current strategy)
• Design work undertaken
• Development Trials

Trials have been carried out on different cementitious materials both in laboratory and on full scale mock ups of the SADT. The trials have been designed to provide additional information on the proposed method of encapsulating the SADT. This information will give a better understanding of the performance of the cementitious material during and after casting and its affects on the pipework. The data and experience gained from the trials will be used to support the theoretical design solution allowing the detailed scheme for encapsulation to be finalised.

• To represent 25m length of the main SADT.
• Constructed from wooden shuttering supported on a concrete plinth.
• Overall dimensions 25m long by 3m high by 1m wide.
• Configurations to represent the range of pipes and layouts in the trench itself.

• To test the ability of the cementitious material to flow at least 14m, unaided.
• To predict the stresses on the pipes caused by encapsulation.
• To establish the maximum grout core temperature due to the heat of hydration effects.
• To measure any vertical displacement of the sludge line due to buoyancy effects of the line during casting.
• To establish the effectiveness of the grout in encapsulating pipework, obstructions and debris within the trench.
• To engineer a leak in a pipe flange with simulated pond pressure of 0.5 bar in the pipe to predict possible leak paths.
• To confirm that the design material strength can be achieved and that the quality is satisfactory.

• The cementitious material flowed unaided in excess of 14m.
• The resultant stresses on the pipes were considered insignificant (~5MPa)
  and would not themselves cause pipe damage.
• Due to the hydration effects of the cementitious material, temperatures in excess of 100⁰ C were recorded during the curing of the grout. This did not meet the design criteria of 56⁰ C.
• There was a vertical displacement of ~ 2mm during initial casting. The current wall thickness pipework (taking account of corrosion) was not available for use in the mock up. Calculations had predicted a movement of 2mm for the lighter pipework and this was seen during the trial. Calculations predict no buoyancy effects for the current wall thickness pipework.
• Sample cores and stripping of the shuttering have proven the effectiveness of the grout in encapsulating the pipework.
• The engineered flange leak and subsequent dye test initially indicated that the most likely leak path is via the joint between the two pours and any fracture plains within the grout. Also this material has the ability to hold significant quantities of water. Minor leaks are acceptable and can be managed.
• The designed material strength was achieved and the quality acceptable.

• The second trial was similar to the first and represented the under road section of the trench with overall dimensions 14m long by 3m high by 1m wide.

• To trial a different material strength against the criteria in Trial 1.
• To give an indication of the head of grout required to push the grout through a drain hole. The drain hole will act as an escape route for air and show when the trench is full.

• The results were positive and as for trial 1 except for the heat of hydration temperature which peaked at 53⁰C. This is within the design criteria and is due to the different constituent parts of the two mixes trialled.
• A 400mm head of grout is required to fill the drain hole.

It was believed that reducing the depth of pour of the first grout mix would reduce the maximum temperature within the grout during curing. Supplementary trials were therefore carried out, where grout was poured to varying depths in 1m² boxes.

However these trials have shown that the mix could not be adequately controlled to meet the required criteria and therefore the mix was not acceptable.

The second mix was acceptable on heat of hydration temperature but was slightly less resilient to the impact protection criteria. However by carrying out further calculations impact protection capabilities can be maintained by increasing the depth of the grout above the pipework. The seismic criteria has also been met by this mix.

• To represent 10m length of the main SADT
• Overall dimensions 10m long by 3m high x 1m wide.
• Constructed on a concrete plinth with reinforced concrete sides.
• 8" and 4" diameter pipe.

• To engineer leaks in 4" and 8" pipe flanges to predict possible leak paths and flow rates.
• To measure temperature and pressure changes to liquor contained within the 8" pipe during grout casting.
• To estimate shrinkage of the grout
• To confirm temperature rise in the grout.

• At the time of writing this paper results for the leak tests and grout shrinkage are not yet available.
• The temperatures associated with the grout and the water within the 8" pipe were well within the 56⁰C design criteria.

Significant progress has been made on this project and includes the following:

• A new emergency pumping system installed - capacity in excess of 1300 m³/hr.
• 35m of pipework removed.
• A costed strategy developed with a revised programme showing a significant time reduction compared to the initial strategy - end date pulled back from 2003 to 1998.
• Safety Case approved by the Regulators for debris removal and in principle for encapsulation.
• Debris removed from SADT in preparation for encapsulation to time and dose budget (40 mSv).
• Successful development trials culminating in an acceptable cementitious material demonstrating that it meets the design criteria.

This project demonstrates a method of risk management in a high dose nuclear environment as an intermediate stage prior to decommissioning. The strategy for safely dealing with this high risk pipework has been developed and implemented in a structured manner using project management, safety assessment, Value Engineering and Kepner Tregoe techniques.

Significant progress has been made and cost and time targets will be met. Major success factors include:

• Improved emergency arrangements
• Use of a team approach to problem solving and use of TQM techniques.
• Investment and off site development to support the theoretical design solution to allow the detailed scheme for encapsulation to be finalised.
• Continuous liaison with Regulators and customers.

BNFL has achieved significant progress in managing and reducing a serious risk to the Sellafield site.

BACK