Matthew Keep
AWE

AWE uses coolants in the manufacturing of radio-active components in support of weapons production. Considerable quantities of the coolant have accumulated due to the loss of an agreed disposal route.

AWE has implemented a coolant management system (CMS) in one of the manufacturing areas on the Aldermaston site. The purpose of the CMS is to improve the control of fissile material within the area and allow the disposal of radio-active contaminated machine tool coolants.

Disposal of the coolant has been approached in two phases. The first phase involves the removal of radio-active gross particulates followed by the dewatering of the coolant by ultrafiltration. Water recovered from the process is sent to AWE’s effluent plant for treatment prior to discharge to the Thames. The second phase involves converting the remaining organic waste to a form whereby it can be readily disposed of. A number of novel techniques involving encapsulation in rubber matrices and bio-remediation have been investigated. Many of the techniques identified have revealed huge improvements in efficiency over those currently in use.

In the summer of 1993 an review in one of the manufacturing areas on the AWE's Aldermaston site revealed higher than expected concentrations of fissile material in the machine tool coolant. Concerns were raised regarding the management of the coolant in the manufacturing areas and a task force was set up to implement an integrated coolant management system (CMS).

One of the most pressing issues was that of coolant disposal. The task force insisted that all of the contaminated liquids held in the machine tools at the time of the incident was to be removed and disposed of. This generated several thousand litres of contaminated oils and coolant in addition to that already held in the facility. The previous disposal route for the coolant had ceased to be available due to concerns by the disposal contractor over the presence of solvents used in cleaning operations. In addition, interest from the regulator was putting the task force under increased pressure to come up with a solution.

The task force initiated the development of a number of processes for the management of coolant. It was felt that some of these could be used to deal with the coolant legacy. Minor modifications were made to the equipment and processes to allow processing of the legacy.

The work has been approached in two phases. The first, high priority, phase was to separate the machine tool liquid waste into their constituent parts. This was to allow the disposal of approximately 80% of the liquids to the AWE Liquid Effluent Plant. Phase 2, which is still in its development stage, is to allow the disposal of the remaining 20% of the waste.

The total quantity of machine tool liquid waste (MTLW) held within the facility after the incident was estimated at around 10 m³, of this it was further estimated that 95% was coolant. The remaining 5% was lubricating or 'tramp' oil washed from the machine tool slides during cutting operations. Most of the MTLW was held in either 205 litre steel drums or 45 litre HDPE containers or `acitainers'.

Equipment developed for the CMS included a Coolant Processing Unit (CPU) for removal of gross particulate and an ultra-filtration Unit (UFU) for the removal of water from the coolant emulsion.

The CPU is a mobile, self contained unit that allows coolant to be passed through a number of progressively finer filters to remove gross particulate. The coolant is then held in 100l tank prior to being pumped into a flat cistern. Attached to the cistern is a NaI assay package that allows the concentration of fissile material to be calculated provided that the concentration is below the agreed upper limited coolant is discharged into acitainers and stored awaiting further treatment.

The CPU was designed using favourable geometry principles. Coolant, of unknown fissile concentration, is held in criticality safe tanks in the manufacturing areas. The CPU guarantees criticality safety until the point, determined by the assay package, where the fissile concentration can be confirmed as within safe limits.

The requirement for favourable geometry placed a number of limitations on the design of the CPU. The unit is large for a device processing 100 litres of coolant at a time. Filter surface areas were small due to limits on the diameters of the filter housings. The capacity of the cistern was low requiring the contents at the main tank to be assayed in up to fourteen batches. Each batchtakes approximately twelve minutes to process adding significantly to the time taken to run the overall process.

*To date approximately 3 000 litres of coolant have been passed through the CPU. In addition about 1 000 litres of contaminated water from other processes have been used as flushings. Flushing is required on a regular due to concerns over the chloride concentrations in some of the MTLW.

Once the coolant has been discharged from the CPU it is passed onto the ultra filtration Unit (UFU). The UFU operates as a cross flow process pumping the coolant in a continuous loop. The coolant passes over an hydrophillic membrane which allows water to pass through leaving the oil behind. Approximately 80% of the coolant is recovered as water suitable for discharge to the AWE liquid effluent plant. The remaining 20% is concentrated coolant which is currently consolidated and held in storage.

The UFU is very simple, consisting of a pump, a coarse filter to trap any gross particulate and the cross-flow ultra-filter membrane. The design was taken from a conventional industrial model and enhanced for duty in a nuclear environment.

To date approximately 2 500 litres of coolant have been passed through the UFU.

2 000 litres of water have been recovered and 500 litres of concentrate have been sentenced for storage. This represents an operational efficiency of approximately 80 %.

Although it was originally intended that primary function of the CPU and UFU would be to manage and condition the coolant in use within the manufacturing areas their biggest impact to date has been to process and recover the legacy of coolant already held. The age and poor condition of the coolant meant that each container had to be individually inspected, sampled and characterised prior to each treatment. Visual inspection to determine the depths of each phase was carried out lowering an endoscope into a custom built glass tube. Samples at each phase were taken to determine the concentrations of fissile material. This was particularly important where it was impossible to subject the containers to radiometric assay.

It was appreciated early in the design stage that the viscous tramp oils would rapidly blind the filters in the CPU and cause a massive drop in performance. It was, therefore, decided that only coolant would be passed through the CPU and UFU. The tramp oils were either left behind in the container to be consolidated at a later date or removed for storage.

Despite this precaution during the early stages of commissioning problems were encountered with the filters blinding very rapidly. Investigation revealed that this was due to neither oil nor particulate accumulating on the filter media. Work being undertaken in parallel on non-contaminated coolant revealed the presence of biological growth at the interface between the coolant and the tramp oil. This growth was highly viscous and coated the surface of anything with which it came into contact. It was considered that the growth was causing the premature blinding of the filters. Efforts taken to compensate for the presence of the growth were adopted successfully and operational performance improved significantly.

Conventional philosophies for the disposal of organic materials involved the use of cement. The waste is adsorbed onto granules which were then encapsulated into a cement matrix. Trials on compression and leachibility revealed that, if the conditions of Acceptance for the licensed LLW disposal site at Drigg were to be met the organic matter would comprise only 1-2% of the final waste package volume. The cost of disposal of 1 litre of organic material is therefore estimated at between $100 and $200. Processing and administrative cost add significantly to these figures.

It was decided that a more efficient solution should be sought. The process was accelerated by concerns over the composition of the granules currently in use and their effects on the health of the workforce. After initial investigations two philosophies were identified that had the potential to be developed into processes. These were absorption and bio-remediation.

Absorption was regarded as a progression from cementation. A number of products are available that react chemically with the organic matter. The amount of organic matter that can be taken up by these materials is considerably greater than that of adsorbers (eg clay granules) which rely on a physical surface effect. Two materials were investigated in depth. The first was the Imtech 'Imbiber Bead, the second Zeon Chemicals 'Norsorex'.

Imtech have been actively marketing their `Imbiber Bead' product for a number of years and claim to have sold it to the Canadian Nuclear industry.

It consists of a fine line bead that is capable of absorbing up to sixteen times its own weight in organic material. Light hydrocarbon such as petrol and kerosene work best being rapidly absorbed by the material. Heavy organics take considerably longer and often benefit from being dissolved in a light hydrocarbon carrier. The beads swell and do not give up the liquid on compression or in contact with a solvent. They will, however, allow evaporation of volatile organic compounds from the surface of the bead. This effect is enhanced by the elevation in temperature. Imbiber beads do not take up water.

Norsorex was originally developed for dealing with marine oil spills. The concept was that the fine powder would be sprayed onto the surface of the oil spill, absorb the oil and be collected for disposal. In reality the powder and oil congealed and sank created an addition environmental impact.

Norsorex is also used in the manufacture of soft rubber components in the automatic industry. When mixed with oil the powder will absorb up to four times its own volume. The resulting compound is a sticky, viscous mass. When heated it turned to a dark brown semi translucent material. No information is available on the retention properties of Norsorex. Norsorex, like the Imbiber Bead, does not absorb water.

Whilst both materials looked promising in the fact that they absorbed organic material efficiently it was evident that an improved method of mixing was required. This was particularly the case when dealing with the more viscous organics. All the laboratory experiments had been undertaken using `beaker and spatula' stirring techniques. Considerable experience in batch mixes had been developed during previous projects concerning cement. All of the mixers involved holding a large quantity of material in the equipment. In the past this had led to difficulties if a problem was encountered with a batch. A solution that involved the minimum hold up was sought.

APV Baker's 'Industrial Extruder Division', based in Newcastle-under-Lyme, manufacture twin screw compounder extruders for a variety of commercial applications. Initial contact indicated that the use of extruders for mixing these types of materials was common practice. A series of trials was commissioned to test the capabilities of the machines with the AWE application.

The results of the trials were mixed. The Imbiber Bead did not mix well coming out of the extruder in an irregular form and often breaking up. The Norsorex product was, however, very encouraging. A solid string was produced. Compression and leaching trials undertaken at AWE with the extruded Norsorex compound suggested that this mixture would meet the acceptance criteria for disposal at Drigg. Further trials were commissioned at APV to optimise the process and obtain a wider range of samples for Drigg Acceptance tests.

Extrusion has the advantage of offering a high throughput with minimal hold up. The trials have indicated that up to three parts oil can be encapsulated with one part Norsorex. This gives a process efficiency of approximately 75%. An increase of up to 3750% in efficiency over that of cement is therefore achieved.

Bio-remediation is a technology that has yet to make a significant impact in the Nuclear Industry. Contact was made between AWE and the Department of Biological and Nutritional Science at the University of Newcastle-upon-Tyne in the summer of 1995. The department was visited in October 1995 and it was clearly demonstrated that the cutting oils could be destroyed by biological micro-organisms.

Due to the lack of experience in the bio-technology field AWE was introduced to the nuclear technology company, AEA(T)'s National Bio-technology Centre through the DTI's Biotechnology Means Business (BMB) initiative. Visits to a number of exhibitions allowed AWE to make additional contacts with companies with experience in the field of bio-remediation.

It became clear early on that AWE would have to rely totally on commercial experience to develop a process for removing the organic matter from the coolant. Three feasibility studies were commissioned from companies that each had a different approach to the project. The specification of the contract was to determine whether bio-remediation presented a practical solution to the disposal of the coolant and to propose plant and equipment that could be installed as a process.

All the companies involved reported that the coolant would support microbial activity and that the coolant was readily digested. This was not a surprise as work undertaken by Newcastle University had identified over 7 geni of bacteria and three geni of fungus. (It was the fungus that caused the stringy effect.) Each company's digestion time was slightly different and this was a reflection of the technology engaged.

Two principles were adopted in the feasibility studies. The first was the use of a bulk reactor where the coolant and bacteria were held in a large tank. Air was sparged through the tank and nutrients fed in as required. The second approach was to use a packed tower where the bacteria and coolant interacted on the surface of the packing.

With regard to operation in a nuclear facility the bulk reactor was an advantage of the packed tower because of its simplicity and ease of cleaning. However, in situations where criticality control concerns may be raised the packed tower system would be easier to design in favourable geometry. The tower, for example, could be packed with boronated glass to act as a neutron poison.

The final reports from the contracts have been delivered. After careful consideration of the details a process will be selected. It should be noted, however, that owing to the lack of experience of biotechnology within AWE, introduction of a plant may be met with an element of resistance.

AWE has developed a system to manage coolant of uncertain provenance from its manufacturing areas.

Phase I of the system has allowed the reduction of MTLW within the facility by approximately 80%.

MTLW is now fully characterised and held in approved containers.

Phase II of the system is approaching the final stages of development. Specification of the process is expected by the beginning of 1998.

New technologies have been investigated that have the potentially to improve the efficiency of the disposal routes for organic liquids enormously.

Further trials are in progress to develop process opportunities from the research and developement undertaken.

AWE has successfully attempted to utilise commercially available technology to accelerate the development of disposal techniques.

B. CROFT AND R. SWANNELL, ‘Used Cutting Fluids: Biotreatability Studies.’ AEAT-1598 Issue 2, AEA(T) plc (1997).

J. BENT AND L. FRANCIS, ‘Bioremediation Feasibility Trial of Solcut BR’, Reponse Environmental Services Ltd (1997).

A NEWEY AND N KNEE, ‘The Treatment of Oily Residues to Remove SNM and Produce Disposable Waste Forms’ NMD01/B/19/01/03/AWEN/071/96/19/11/96, AWE, November 1996.

N KNEE, ‘The Polymer Absorption and Solidifaction of Cutting Oil Emulsions’, NMD01/B/19/01/02/NDK/027/97, AWE, July 1997.

ã British Crown Copyright 1997/MOD

Published with the permission of the Controller of Her
Brittanic Majesty’s Stationary Office.

BACK