A. Demeulemeester, V. Massaut, M. Klein
SCK· CEN
Boeretang 200
B-2400 Mol, Belgium
Tel. (+32-14) 33 26 07
Fax (+32-14) 31 19 93

A. Lefèbvre
Belgatom
Avenue Ariane 7
B-1200 Brussels, Belgium
Tel. (+32-14) 33 26 24
Fax (+32-14) 31 19 93

The BR3 reactor was the first Pressurized Water Reactor installed in Europe. With a quite low net power output, it was started in 1962 and definitely shut-down in 1987. In 1989, the BR3 was selected as one of the four pilot decommissioning projects by the European Commission in the framework of its five year programme on the decommissioning of nuclear installations.

Although the dismantling project is still in progress, the paper will highlight already the most important results gained during the execution of the project till now.

The full-system chemical decontamination of the primary loop showed that this operation was really effective in saving doses and cost. Moreover, the waste decategorization of some parts of the reactor internals was also a positive result of the operation. The use of the plant equipment allowed to carry out this operation easily and with a minimum cost. Nevertheless it requires that the plant equipment such as pumps and valves are still in full operational conditions.

Many different remote techniques were applied to dismantle the highly active internals: plasma arc torch cutting, electric discharge machining, circular sawing, band sawing, hydraulic cutter, ... This allowed the BR3 to compare on actual pieces and in an actual environment the different cutting methods, their advantages and drawbacks, in terms of waste generation, dose uptake and cost. Moreover, for the underwater cutting strategy applied, it appeared that the cutting time itself covers only a small fraction of the total operational time. Therefore improvements can best be made on the auxiliary operations instead of only on the cutting speed.

Another result concerns the comparison of the immediate dismantling with the 30 years deferred dismantling of highly active pieces. Indeed, BR3 disposed of two sets of irradiated internals, one of which having been unloaded 30 years ago, and being stored in the plant during this period. The actual dismantling of both sets of internals showed that no real advantages in terms of dose and waste can be expected from such a deferring period.

In the paper there is also a description of a basket system for the high level waste evacuation. Before transportation to the waste conditioner, the BR3 high level waste was prepared in accordance to the different conditions required by the producer himself, the legal Safety Authorities and the National RadWaste Authority (NIRAS/ONDRAF). To fulfil these conditions, the dismantling team developed, in collaboration with the waste conditioner, a complete system of dedicated racks and baskets.

The paper focuses also on the waste characterisation. The characterisation is based on swarfs analysis at different altitudes of the reactor internals. After measuring the specific activity of ⁶⁰Co of the swarfs, these values gave the specific activity distribution along the longitudinal axis of the internal as well as the dose rate, important for the waste handling and conditioning. Other critical isotopes were measured by radiochemistry and allowed to define an isotope vector (finger print) that could be used for all pieces generated by the dismantling of the internals.

In parallel with the pilot project, The BR3 team started with the dismantling of loops and very low level contaminated equipment. This creates a complex material flow management. The paper describes the different requirements to which this flow management has to answer. Because the aim is to free release and recycle as much material as possible, this paragraph describes also the different decontamination techniques used in BR3

Finally the experience gained during the project allows to better define the different costs factors for the estimation of decommissioning costs. A data base on unit cost estimate has then been set up and is used to forecast or assess the cost of decommissioning projects and operations.

Although the project is not completely finished, it has already proven that probably the most delicate operations of the decommissioning which is the dismantling of highly activated material, can be carried out safely, economically and with today's technology.

The Belgian Nuclear Research Centre (the SCK· CEN) has three main nuclear reactors of which BR3 (Belgian Reactor n° 3) was the only one to produce electricity. The BR3 reactor was the first Pressurized Water Reactor (PWR) installed in Europe with a net power output of 10 MWe . Its main purposes were the training of personnel for the Belgian commercial nuclear power plants and the testing of advanced nuclear fuels (such as MOX-fuels). The reactor reached criticality for the first time in 1962. After 25 years of successful operation, the reactor was definitely shut down in June 1987. In 1989, the European Union selected the BR3 as one of its four pilot decommissioning projects in the framework of its five year Research & Technological Development programme on the decommissioning of nuclear installations.

For the first two years, a 6.7 tons stainless steel set of internals, the Westinghouse internals, were in the vessel. These internals contained the nuclear fuel and the measuring and controlling instrumentation. After those two years a new set of internals, the so-called Vulcain internals, replaced the original ones for experimental purposes. The first set of internals was stored during more than 30 years in a lead shielded corner of the reactor pool. In annex, figure 1 gives an image of the two sets of internals.

Up to now the pilot project has produced a lot of results and experiences. This paper will highlight the most important ones through describing the major parts of the project. These parts are the full system chemical decontamination of the primary loop, the comparison of cutting techniques, the comparison between the cutting of the two sets of internals, the waste management of the highly activated internals and the waste characterisation and management. Finally, the paper describes two other topics of the project, the very low level waste management and the development of a tool for the cost estimation of future decommissioning projects.

Twenty-five years of running the BR3 resulted in a radiating primary circuit and primary equipment, more particularly the steam generator and the primary pumps, due to internal contamination. A preliminary analysis showed that a net gain in the radiation exposure of the personnel during the dismantling operations could be achieved by performing a chemical decontamination of the primary circuit.

After a literature study and the execution of tests on original samples taken from the primary loop, the BR3 dismantling team selected the Siemens KWU CORD process (Chemical Oxidation and Reduction Decontamination) for the decontamination of the primary loop. This chemical process attacks the corrosion layer formed on the inner surface of the loop and so dissolves the fixed contamination. The dissolved corrosion products and activity are fixed on ion exchangers (demineralizers). The selected decontamination process has the following main advantages : low chemicals concentration, no special additional equipment, a reasonable Decontamination Factor (DF) and a low volume of secondary waste. The cold tests showed that a DF-factor between 9 and 30 could be achieved.

The decontamination operation comprised several steps : the inspection of the primary circuit, installation of the equipment, the Hot Run Test, the decontamination itself and the waste evacuation.

Before the decontamination could start after a 4 years shut down period of the reactor, the BR3 team carried out an intensive inspection and maintenance of the primary loop and its auxiliary circuits. Several modifications were also made (operation at 20 bars, use of the steam generator as liquid-liquid heat exchanger, installation of additional ion exchange columns). The AMDA equipment (preparation and injection of chemicals) was then connected to the primary system and a hot run test was performed before the operation.

This test implied that the primary loop and purification system was operated two days at temperature and pressure with the objectives of checking if all the systems were in good running order, as well as measuring leak rates and various other operational data. Then, the decontamination took place in three cycles. At the end, the process removed 33 kg of oxides (Fe, Cr and Ni) for a total activity of 2.05 TBq (55 Ci).

As far as the primary loop is concerned, not considering the measurements done in the vicinity of the reactor upper flange where the gamma dose rate is mostly due to the radiations emitted by the activated reactor internals, a mean DF close to 10 has been obtained, with a broad spreading of individual values ranging from 0.1 (redeposition of activity in a horizontal pipe) to 31 (steam generator) according to the measurement location.

As expected, in the vicinity of the reactor upper flange, the impact of the decon process was less important; nevertheless, the radiation field at the bottom of the reactor pool has been reduced by a factor about 2. Along the purification system, where the operating temperature during the decon process was kept lower (40-80° C) than in the primary loop, (80-100° C), a mean DF close to 6 has been obtained.

In the plant container, the decontamination operation has substantially modified the picture as far as working conditions are concerned : the ambient dose rate has been reduced by a factor about 10 and amounts now to about 0.08 mSv/h. In the purification circuit, the ambient dose rate is now around 0.06 mSv/h.

Besides a low water volume collected as low level waste, the decontamination generated only exhausted anionic and cationic resins for a volume of 1.37 m³.

The full system decontamination of the BR3 with the CORD process demonstrated that:

• it is possible to perform such a task with only minor modifications to the plant;
• the 3 cycle application of CORD led to significant dose reductions;
• the application and continuous control was quite easy to perform;
• low waste volumes were generated.

The BR3 team studied and used a variety of cutting techniques for cutting the highly radioactive reactor internals. This paragraph will focus on the main techniques used and their comparison. Three techniques were used for cutting the thermal shield and then compared.

These techniques were plasma arc cutting, Electro Discharge Machining (EDM) and one mechanical technique, milling cutter. Figure 2 in annex shows where and how the different techniques were used on the thermal shield and table I summarises the most interesting results.

The table shows that the role of the different equipment manipulations is important and that the mechanical cutting technique creates the less volume of secondary waste. The plasma cutting technique points out the influence of the auxiliary manipulations. Although its instantaneous cutting speed is 50 times faster than that of the mechanical, the average effective cutting speed of the plasma cutting is only 1.6 times higher than the one of the mechanical cutting.

The column of the secondary waste shows that the plasma cutting and EDM produced five times more secondary waste volume than the mechanical cutting technique. This is mostly due to the fact that EDM and plasma arc torch produce fine particles (< 1 m m) to be filtered on fine-mesh filters with a low filling capacity.

Based on these results, the mechanical cutting technique was selected as main technique for the dismantling of the remaining sets of internals. These internals mostly consisted of large cylindrical shaped pieces. The cutting strategy was first to cut the cylindrical pieces into rings with a circular saw, followed by the segmenting of the rings with a bandsaw.

Additional techniques used were hydraulic shears to cut small tubes such as thimbles and drainpipes, manual, pneumatic and hydraulic unbolters disassemble some sub-assemblies. EDM, using hollow electrodes, was used to cut some top-down bolts.

Table II compares the obtained technical data of the circular saw and bandsaw.

What does this table show us ? (for the same cut cross section)

• Almost all technical parameters are quite similar.
• It is obvious that the circular saw technique produces more secondary waste (metal swarfs) than the band saw due to a greater kerf width. The removed volume of metal is three times higher than with the band saw.
• The number of used blades is two times higher for the band saw technique. Nevertheless, the volume of low level waste as a result of the blades conditioning is negligible.
• The purchase of the circular saw technique is 80% more expensive. Also the consumable cost (mainly the blade cost) is 13 times higher than with the band saw technique, although the consumption of blades with the band saw technique is higher.

Based on these results, the conclusions are :

• Both techniques are complementary, but
• When the two techniques are applicable, the bandsaw should be preferred for economical reasons, the investment and consumable costs being lower, for the lower volume of secondary waste and for the higher flexibility.

As already explained in the introduction, the activity of the Westinghouse internals decayed during more than 30 years in a lead shielded space of the Refuelling Pool. Because of the very similar geometry of both sets of internals, the unique opportunity existed to cut a set of reactor internals with a deferring time of 30 years and to compare it to direct dismantling. The BR3 team looked at the influence on the total dose uptake by the workers and on the production of the different waste types (low, medium and high level waste).

Concerning the dose, there is no significant decrease of the received dose by the workers. This can be explained as follows. During the whole cutting campaign, the core mid plane was under a minimum of 4 m water (and even more). Considering that the "half thickness" of water is about 12 cm for ⁶⁰Co, the gamma reduction factor is about 10.000.000.000. The activity reduction factor due to the 30 years cooling down, which is about 50, is negligible compared to this last factor. No significant influence of the cooling down period could then be expected for the direct radiation influence. But the remaining radiation level of the cooled down internals was still high enough to require remote shielded operation.

Regarding the waste production, no significant advantage of the cooling down period could be observed. If we compare the immediate dismantling of the Vulcain internals and the dismantling of the Westinghouse internals (after 30 years), the waste category transition levels were almost the same. The only difference concerned one cut ring. So, the gain made on the Westinghouse internals was the category change of 283 kg (4.3% of the total weight) of material from medium to low level waste.

The main conclusion is thus that no significative advantage can be drawn from waiting for 30 years before dismantling highly radioactive pieces. This seems to be valid for the radioprotection point of view as well as for the waste production. Regarding the similar activity levels of both sets of internals at the time of definitive unloading, this conclusion is based on sound comparison results.

This paragraph describes the export evacuation system for the high and medium activated metal waste and the volume of secondary waste, produced by the cutting of the internals. The system fulfilled the Belgian waste conditioning requirements and the specific internal infrastructure requirements.

It is the National RadWaste Authority (ONDRAF/NIRAS) who sets up the different acceptance criteria on waste types and waste packages. Concerning the solid waste (big pieces), there are three important groups of waste and the distinction between these groups is based on the contact dose rate. These three groups are : Low Level solid Waste (LLW) with a contact dose rate < 2 mSv/h, Medium Level solid Waste (MLW) with a contact dose rate between 2 mSv/h and 0.2 Sv/h and High Level solid Waste (HLW) with a contact dose rate > 0.2 Sv/h.

Up to now, 12.7 tons were evacuated as HLW or MLW to BELGOPROCESS, the Belgian waste conditioner and interim storage facility. The remaining part was LLW. Table III gives the weight for the different types of waste for the three main internals.

For the conditioning and transport, the waste has to fulfil different requirements from NIRAS/ONDRAF, the legal Safety authority and the waste producer himself.

• The waste must be put in a 400 l-drum.
• The weight of each unconditioned drum may not exceed 800 kg.
• There must be a free space of three centimetres between the waste pieces and the inner side of the drum and 5 cm between the waste pieces and the top/bottom of the drum.

For the Safety Authority :

• The transported pieces may not move during the transport.
• The national transport limits may not be exceeded (the contact dose rate of the container must be less than 2 mSv/h and the dose rate at 1 metre of the transport vehicle must be less than 0.1 mSv/h.
• The worker's dose uptake has to be minimised (ALARA principle).

• The waste manipulations must be carried out under shielding (in this case water) for protecting the workers against the high radiation.
• The filling must be optimized in order to produce as less number of drums as possibile (economic optimization).
• The cut pieces had to fit in a small internal transfer container. Indeed, the only location to fill the transport container was in the spent fuel pool. The cut pieces had to be transported from the reactor pool, where the dismantling actually took place, to the spent fuel pool. A 13 tons lead shielded transfer container, which had to pass the small personnel air lock, had been designed and fabricated to transport the waste to the spent fuel pool. Due to this transfer container (inner side dimensions : 800 mm x 500 mm x 500 mm), the dimensions of the cut pieces were limited.

To fulfil all these requirements, the export system consisted of a carrier structure and two baskets which fit over the carrier structure. Two baskets and a carrier structure formed a waste package. A special lifting device was designed to manipulate the waste package for loading and unloading the transport container at respectively BR3 and the waste conditioner. The whole system was of course designed for working under water, but also in hot cells (see fig. 3 in annex).

The characterisation of the radioactive waste implies the determination of physical parameters (weight of the primary waste, form, material composition), chemical parameters (eventually toxicity, reaction with waste matrix or host site) and radiochemical parameters. For this radioactive waste, the total activity is the sum of the radioisotopes formed by activation in the mass and by contamination on the surface.

For the contamination, gamma spectrometry was used to determine the different isotopes in the deposited oxide layer. With this value and an estimated surface of the waste in each drum, the activity from the contamination could be estimated. The analysis was done on representative pieces of the internals. A difference was made between pieces that had undergone the full system decontamination in 1991 and pieces that had not.

Table IV shows the contamination values for the different isotopes for the Vulcain internals. (Some pieces were stored in the Storage Well and were therefore not decontaminated).

For the activation, the activation level of the waste was estimated by taking metal swarfs during the execution of the different horizontal cuts on the internals. For the determination of the isotopes, three different techniques were used, namely gamma spectrometry, modelling and specialized radiochemical techniques.

The gamma spectrometry gave the specific activation, expressed in Bq/gr, for the isotopes ⁶⁰Co and ¹²⁵Sb. The isotopes ⁶⁰Co and ¹²⁵Sb were measured immediately at BR3 with a NaI-detector. A first estimation of the transported activity was based on these isotopes. The values of ⁶⁰Co and ¹²⁵Sb were plotted to estimate the specific activity in function of the height of each internal. Figure 4 in annex shows such a curve for ⁶⁰Co. With the specific activity and a calculation code, the contact dose rate of each piece of waste could be calculated. These data were then also used for an estimation of the specific activity (and the corresponding contact dose rate) on smaller pieces of the internals such as plates, tubes, top-hats, ...) for which no samples were taken. The "difficult to measure" J-emitters were determined in the radiochemistry department of SCK· CEN. This allowed to determine some important long lived radionuclides such as ⁶³Ni, ⁵⁹Ni and ¹⁴C.

For other isotopes such as ⁹³Zr, ⁵⁵Fe and ^121mSn, the specific activity was calculated using the ratio between ⁹³Zr/¹²⁵Sb, ⁵⁵Fe/¹²⁵Sb, ^121mSn/¹²⁵Sb and ⁵⁵Fe/⁶³Ni, based on the base alloy composition, the radiation history and the isotopes properties.

The combination of these techniques allowed to determine an "isotope vector" (finger print) which can be used to quantify the whole spectrum of the important radionuclides not only in the short term (short lived K-emitters) but also in the mid term (long lived J-emitters) to the long term (I-emitters).

The circuits and the infrastructure represent a large volume and mass, and a great variety of different materials and often low contaminated material. Different export routes are possible for these materials : to be evacuated as radwaste, to be recycled (possibly after melting) or to be free released (possibly after melting). The goal of the dismantling is to free release as much material as possible if this represents the most economic way of evacuation.

To reach the limits for free release, different handling and decontamination techniques are used :

• manual decontamination techniques;
• wet abrasive decontamination process;
• chemical process based on the use of Cerium⁴.

The variety on materials, the different export routes and the different decontamination techniques result in a quite complex material flow. This paragraph will describe first the different requirements from the authorities and waste conditioner, and then very shortly the installed follow-up system and the QA-system (being implemented) to better manage this material flow.

The asked requirements on the material flow can be divided into three groups, namely traceability, destination and measurements requirements.

For the radwaste agency, NIRAS/ONDRAF, the producer has to indicate the quantity, characterization and destination of the waste coming from the different dismantling activities. Also, the free-release procedure of the SCK· CEN Health Physics Department demands the knowledge of the history of the waste to be free released. This means that the producer must be able to give the origin of the material (from which circuit or location) and the different treatments undergone. The producer also wants to know at every moment the situation of the material flow.

As mentioned, BR3 has the choice between three destinations for its material; radwaste, melting and free-release. Each destination has its own requirements on the material.

When the destination is radwaste, the materials have to follow the different waste requirements, formulated by NIRAS/ONDRAF. For the Low Level Waste, the main relevant waste categories are supercompactible metallic, non-supercompactible metallic, and burnable waste. The number of waste types and the different requirements on each type imply an attentive breakdown of the dismantled materials.

Up to now, for the recycling within the nuclear industry, the SCK· CEN used only melting, and sent the metal scrap to SEG-Westinghouse (USA). This implies that the materials fulfil the corresponding requirements. A great advantage of melting is that whole structures can go to the melting factory which is, from a dismantling point of view, very interesting (less cutting operations).

The last but not least important destination is the free-release of material. In order to decontaminate the pieces up to free-release level, the SCK· CEN uses different techniques such as cleaning (manually with a detergent or with a high pressure water jet), wet-sand blasting or hard chemical decontamination process. For the sand blasting, SCK· CEN constructed a special treatment cabin with an enveloped volume of 27 m³. For the chemical process, based on the Ce⁺⁴ process, an installation is under construction. It consists of a batch process with a chemical reactor of about 1.5 m³. Each process demands other requirements on the materials to be treated. The kind of material and the good accessibility to the internal surface contamination determine the choice of the treatment process. This means that the pieces with a poor internal accessibility will mostly undergo the chemical treatment (or will be melted or treated as radwaste if the weight/surface ratio is too low).

The materials which will be unconditionally free released have to be carefully measured for controlling if they follow completely the complete requirements for free release, as compiled in a procedure of the SCK· CEN Health Physics Department.

This procedure foresees two ways for releasing materials unconditionally : on one hand, based on surface contamination and implying the 100% surface measurements and, on the other hand, based on mass contamination and requiring the specific activity determination. This procedure asks also for a historical description of the items considered.

For being able to give all the necessary information to the different organisations managing the material flow, a batch management system based on a computerised database, which contains all the relevant information on the batches for fulfilling the different requirements, was set up. A batch is a certain quantity of material destinated for a selected evacuation way. This implies that all the items in the batch meet the requirements of the chosen evacuation way and the attached measurements methods. For the traceability, the batch gets a unique identification number which is physically attached to it and will travel with it through the material flow.

To ensure always the good "quality" of the materials (quality means in our case that the materials fulfil the requirements on traceability, destination and measurement) a QA system is implemented at BR3, for the complete management of the contaminated or suspected materials.

The feasibility study of dismantling high activated reactor internals with existing but modified technology was one of the goals of the pilot project. Another goal of the project was to get data on cost of different actual dismantling operations. Therefore, the SCK· CEN designed a model computing the decommissioning costs. This model uses an interactive database covering all aspects of the physical and radiological inventory, the unit costs deduced from own experience such as decommissioning the BR3 reactor, old laboratories and external projects. The results of SCK· CEN's evaluations performed on different nuclear facilities and following various strategies were successfully compared with those obtained by other institutions.

The dismantling project has proven that it is possible to dismantle high radioactive reactor internals with present techniques. Optimization concerning the waste, dose and cost, was carried out throughout the entire project. Based on the gained experience, the BR3 team prepares the dismantling of the last high activated reactor part, the reactor vessel.

Parallel with the reactor dismantling, the circuits will also be further dismantled. With these dismantling, BR3 will collect a lot of data in the management of the waste, in the use of different decontamination techniques and in the free-releasing of different types of materials.

All the present and future results of the BR3 dismantling will be collected and will be used for a better and finer estimation of the dismantling costs.

Part of the BR3 pilot dismantling project was carried out under contract with the European Commission as one of the four pilot projects in the framework of its research programme on the decommissioning of nuclear installations. Industrial partners like Framatome, Siemens, Rolls-Royce & Associates and Belgatom were also involved in the success of the project.

1. V. MASSAUT, M. KLEIN, A. LEFEBVRE, H. WILLE, H. OPERSCHALL, P. SAUMON, M. DUBOURG, P. ROBERTS, "The BR3 Pressurized Water Reactor Pilot Dismantling Project", Draft Final Report CEC contract F12D-CT89-0003, 1989-1993, Ed. CEC, EUR16...EN, 1996 (Final Report to be published)

2. M. KLEIN, et al., "Radiological Characterization of Shutdown Nuclear Reactors for Decommissioning Purposes", IAEA Technical Report 5th draft, 1996 (Final Report to be published)

3. SCK· CEN, "Officiële Richtlijnen voor het Beheer van Radio-Actief Afval Deel 1 en 2", Ed. SCK· CEN, 1993

4. P. DEBOODT, "Vrijgave van materialen afkomstig uit gecontroleerde/bewaakte zones van nucleaire installaties", Ed. SCK· CEN, PO.SB.002/N, 1996

5. L. NOYNAERT, V. VAN ALSENOY, R. CORNELISSEN, S. HARNIE, "Decommissioning plan of a nuclear research centre : lessons learned by the SCK· CEN", WM'97, Tuckson, 1997

6. L. NOYNAERT, V. MASSAUT, R. CORNELISSEN, M. KLEIN, "Decommissioning costs of the BR3 reactor", Annual Meeting on Nuclear Technology '96, Mannheim, Germany, 21-23 May 1996, ISSN0720-9207, p543-546

BACK