- Email: [email protected]
Defuelling of the UTR-300 research reactor
Nuclear Engineering and Design 178 (1997) 99 – 105
Defuelling of the UTR-300 research reactor R.D. Scott a,*, H.M. Banford a, B.W. East a, M.A. Ord b, A.P. Gaffka b a
Scottish Uni6ersities Research and Reactor Centre, East Kilbride, Glasgow G75 0QF, UK b AEA Technology, Harwell, Didcot, Oxfordshire OX11 0RA, UK Accepted 11 August 1997
Abstract A description is given of the movement of fuel elements from the core of the UTR-300 research reactor to the UNIFETCH flask, which is normally loaded under water, through a specially designed shielding arrangement which permits a dry transfer. The regulatory requirements and the safety case are summarised along with the predicted and measured doses to operators. The task was successfully completed to a tight time schedule with recorded doses which were well within the allocated dose budget. © 1997 Elsevier Science S.A.
1. Introduction The UTR-300 reactor at the Scottish Universities Research and Reactor Centre was based on the original Argonaut design with two aluminium core tanks set in a graphite reflector each containing six fuel elements cooled and moderated by water flowing up through the tanks in a closed primary circuit. The fuel plates in the original 13-plate elements were uranium oxide-aluminium with a 22 g loading of 90% 235U. After 7 years of operation at 100 kW (10 kW average from 3 runs of 6 h per week), the maximum power was increased to 300 kW (30 kW average) and, in order to maintain the operational excess reactivity, it was necessary to add another plate to each element progressively over the years until they all contained 14 plates. These extra plates were uranium metal-aluminium with 24.5 g of 90% 235U. * Corresponding author. Tel.: + 44 1355 223332; fax: + 44 1355 229898.
Calculations of critical mass as a function of the fuel plate spacing indicate that the critical mass is near its minimum value for a 13-plate element and that the critical mass increases sharply as the spacing is decreased by the addition of further plates (the overall dimensions of the element are fixed) and the core becomes more under-moderated. Addition of a 15th plate would have had a marginal effect on reactivity and, in order to appreciably extend the operating life of the reactor, it would have been necessary to replace some of the elements with fresh ones made from our stock of unirradiated plates. However, irradiated fuel could not be stored on site and shipment of this material to the owners, the UK Atomic Energy Authority (AEA), would have triggered immediate reprocessing charges, with more to come in the future. The financial implications, coupled with low research usage and little demand for its use in teaching, made shutdown inevitable and the decision was taken in March
0029-5493/97/$17.00 © 1997 Elsevier Science S.A. All rights reserved. PII S 0 0 2 9 - 5 4 9 3 ( 9 7 ) 0 0 2 0 9 - 4
100
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
1995 that the reactor should operate for the last time early in September 1995. The unloading of the fuel and shipment to Dounreay Nuclear Establishment for reprocessing are described in this paper and constitute the first stage of the decommissioning of the UTR-300 reactor.
2. Regulatory requirements The operation was carried out as a plant modification in compliance with an arrangement made against one of the site licence conditions and was assessed in terms of safety as a category 1 modification, having major significance. Any such modification requires a proposal containing a detailed safety case which, after consideration both by the SURRC Nuclear Safety Committee and by an independent consultant, is submitted to the Nuclear Installations Inspectorate (NII). Depending upon the response from NII, work may proceed immediately or must not proceed without an agreement from NII issued in the form of a licence instrument. In the latter case, ‘hold points’ are sometimes identified beyond which work may not proceed without the further agreement of NII. Preliminary investigations established that the container most suitable for the transfer of highly enriched irradiated fuel and which could be handled in the Dounreay pond was the UNIFETCH flask belonging to AEA. This flask could, however, take only six elements at a time and was normally loaded under water so two shipments were required and a shielding system compatible with dry loading had to be designed. It was also necessary to obtain a certificate of approval from the Department of Transport for the design of the package for carriage of radioactive materials based upon activity estimates and supported by a criticality calculation.
specially designed shielding arrangement. The complete assembly is shown in Fig. 1. It comprised a basket, fixed relative to the UNIFETCH inner wall, with six circumferentially disposed fuel pockets and a blank position, a ratchet ring, a slewing ring supporting a rotating hub to which was attached a primary shield with a single hole matching the size of the fuel pocket and a removable top shield mounted on the primary rotating shield, again with a single hole to match the fuel pocket. The idea was to lower one fuel element at a time from the SURRC fuel transfer flask, which holds a single element, into each of the six pockets by rotating the shield after each insertion, so ending up with the hole over the blank position and the six elements shielded. The removable top shield had a 25 mm deep ‘footprint’ of the transfer flask cut into its top surface so that, as well as ensuring accurate positioning and providing extra shielding during the loading, it largely eliminated the g flash as the element was lowered into the basket. Fig. 2 shows more detail of arrangement. The ratchet and slewing rings were fixed relative to the basket and the hub could rotate on the slewing ring bearing (in one direction only) with the primary shield attached. Orientation of the various components was achieved by means of dowel pins before they were bolted together, although the removable shield was not bolted to the primary shield since it was held in place by the weight of
3. Design of the UNIFETCH basket and rotary shield Since there was no pond at SURRC to permit loading of the UNIFETCH flask under water, it was necessary to do a dry loading through a
Fig. 1. General view of the UNIFETCH flask, the fuel basket and shielding with the SURRC fuel flask in position.
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
Fig. 2. Detail of the rotating shield and removable top shield.
the SURRC flask. The position of a dowel pin on the primary shield relative to the wall of the UNIFETCH was marked with adhesive tape to indicate each of the six positions and precise alignment was provided by a spring plunger which dropped to engage the ratchet ring when the holes through which the fuel was loaded were aligned. The locking bar shown in Fig. 2 also engaged the ratchet ring when fuel transfer was taking place. There was no access to the spring plunger when the removable shield was in place and great care was therefore necessary when the system was rotated since, if the next empty position was passed, it was not possible to go back. Once the final blank position was reached, the spring plunger dropped into a slot in the ratchet ring and movement of the shield in either direction became impossible.
4. The safety case
101
pits in the floor of the reactor hall by means of the transfer flask referred to in Section 3 above and shown in Fig. 1, and it was decided that the loading into the UNIFETCH flask would be carried out in two stages, from the core to the storage pits and from there to the UNIFETCH by means of this flask. This meant that the fuel could be kept in the reactor from shut-down until the transfer with the core ventilation running continuously and the fuel tanks dry. The transfer from core to floor storage was carried out by SURRC staff according to documented procedures under the maintenance arrangements and lay outside the scope of the safety case, which was therefore required to cover only movement of fuel from the floor to the UNIFETCH together with procedures for handling the UNIFETCH and the rotating shield. The matter of training was considerably simplified by a division of labour between SURRC and AEA staff based upon past experience: all operations involving the SURRC transfer flask were undertaken by SURRC staff whereas the UNIFETCH was handled by AEA staff. It was possible to maintain this practice even when the two flasks were mated during the transfer of a fuel element. The case was produced by AEA and consisted of a combined preliminary safety report, pre-commissioning safety report and pre-commencement safety report containing a hazard identification (HAZOPS) study. It was backed by a quality plan and detailed operating instructions for the various tasks involved.
4.2. Calculation of fuel element acti6ity and dose rates
4.1. Requirements and scope The arrangement to comply with the Site Licence condition under which the operation was carried out requires a safety case based upon a comprehensive review and assessment of the hazards, particularly those which might result in increased radiation exposure to operators or to members of the public. During annual maintenance the fuel elements were routinely transferred into shielded storage
Radiation doses from both shielded and unshielded fuel elements constituted an important aspect of the safety case and the activity of the fuel was therefore calculated at AEA Winfrith by means of the FISPIN code (Webster, 1995). The calculation was performed with the neutron flux modified by the duty cycle of the reactor (18 h per week) and it is interesting to note that this gave an underestimate of the activity at shutdown of fission (or activation) products with half-lives
102
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
much less than the ‘on’ period of the duty cycle (6 h) since they saturate at a value corresponding to the true neutron flux. The effect is evidently a function of the half-life of the species and a detailed calculation shows that for our duty cycle it can be ignored for half-lives in excess of a few days. The code also gave the g-ray spectrum in 24 energy groups as a function of time from the end of operation and this spectrum was used for shielding calculations assuming a decay time of 1 month. FISPIN is a point code and account was taken of the distribution of activity in the fuel by assuming a cosine distribution of activity along the axis of the fuel elements with an extrapolation distance of 20 cm. Calculations in an accurately modelled geometry were made of the dose rate directly above a loaded fuel pocket in the UNIFETCH basket with and without the removable shield in place and of the dose rate directly above the last open empty pocket when the other five were filled, but shielded. The total predicted activity of 12 elements after 31 days was 4.5× 1014 Bq and the above rates were 390 mSv h − 1, 3.4 and 23 mSv h − 1, respectively. Measurements of dose rate performed during the fuel movement indicated the validity of the calculations. For instance, taking account of the fact that the actual cooling time of the fuel was 120 days and that this reduced the b activity by almost one half, but ignoring any change in the g-ray spectrum, the calculated dose rate in the last case mentioned above was 13 mSv h − 1 against the measured value of 12. This is very good agreement considering that the dose in this instance arose from g-rays scattered out of the empty pocket. The dose from an unshielded or partially shielded element was also obtained for use in the analysis of fault sequences in the HAZOPS study discussed below.
4.3. The HAZOPS study The HAZOPS study (Kletz, 1986) was a highly structured procedure which lasted 2 days and employed a committee of AEA and SURRC representatives covering design, operations and health physics aspects. The entire operation, start-
ing with the arrival on site of a low-loader carrying the UNIFETCH and finishing with the preparation of the flask for departure, was broken down into ten independent steps, each containing a number of identifiable tasks. The effect of deviations, defined by parameters (e.g. pressure) and guide words (e.g. higher), on each of these tasks was considered in terms of consequences, safeguards and recommendations. A total of 19 recommendations requiring either satisfactory responses or action emerged from the study and covered matters such as contamination, correct location of the shield relative to the basket and the need for a spring-loaded plunger to engage the ratchet ring. The analysis of possible accidents was based on a fault schedule produced from the HAZOPS worksheets in the form of a list of 132 faults, safeguards (engineered or management controls) and consequences, not all of which contributed significantly to risk. Low risk hazards were therefore screened out on the basis of a dose limit of 100 mSv for operators and 10 mSv for members of the public or a frequency of B 10 − 6 per year. A second screening using the product of dose and frequency was carried out and the remainder (14 in number) were treated in detail in an accident hazard assessment which produced a set of individual risks calculated from the product of frequency, dose and dose risk factor (taken to be 0.05 per Sv). The total individual risk to an operator arising from these 14 fault sequences was 4.4×10 − 6 per year with the dominant contribution (70%) arising from three accident sequences which could have led to exposure either to an unshielded or partially shielded fuel element. This slightly exceeded the NII basic safety objective of 10 − 6 per year (Health and Safety Executive, HSE, 1992). However, taking into account the fact that the sequences are dominated by operator errors, the probabilities of which were conservatively chosen, and the extra cooling time of the fuel, it was concluded that the objective was closely approached and unlikely to be exceeded. The NII basic safety limit of 10 − 4 per year was comfortably met.
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
4.4. Dose assessment during normal operation Each operation, starting with the arrival in the reactor hall of the low loader carrying the UNIFETCH flask and ending with the monitoring of the loaded, sealed flask was allotted a whole body dose based on the dose rate associated with the task and the estimated time for its completion. The estimates were pessimistic in that upper bounds were used both for calculated dose rates and times and resulted in a dose of 600 mSv summed over the SURRC operator working with the SURRC transfer flask, the AEA operator involved with the UNIFETCH, a supervisor and a health physics monitor. The maximum single dose was 239 mSv and a dose restraint objective of 250 mSv (for the transfer of six elements) was therefore set. The dose estimates could, of course, have been invalidated if there were any shine paths through the basket and lid assembly and a test was therefore carried out with a 1 GBq 60Co source in one of the fuel pockets. A scintillation probe was used to check round the outer rim of the removable shield and of the primary rotating shield. In the first instance only the background counting rate was observed and in the second, 500 cps were measured at a region on the joint between the lid and the flask body adjacent to the pocket containing the source, with a negligible number elsewhere around the rim. The sharp fall on moving the probe a few centimetres away from the position of maximum counting rate suggested that the counts were due to directly transmitted radiation, rather than to scatter, and this conclusion was supported by a shielding calculation.
5. Procedure and timetable Since there was some doubt as to the continued availability of reprocessing facilities at Dounreay, a deadline of January 1996 was set for the removal of fuel from the site. Discussions between SURRC and AEA began about a year in advance of shutdown and led to the preparation and acceptance of a formal quotation and initiation of the design and construction of the basket and
103
shield. The work proceeded in parallel with the preparation of the safety case which was submitted to the NII and to an external assessor at the end of October 1995. This was followed by an emergency exercise, inspected by NII, which had as the scenario a fire caused by an electrical short circuit, a radiation hazard from a partially shielded element and a personal injury. It was designed to test the various hazard and rescue procedures, emergency shielding arrangements and liaison with the local fire brigade (which was in attendance), and was judged satisfactory. Early in November 1995, the UNIFETCH flask, complete with basket, primary rotating shield and top shield was brought for testing to SURRC on a low loading trailer which was manoeuvred into position in the reactor hall and disconnected from the tractor. The UNIFETCH had to remain on the trailer because its weight far exceeded the capacity of the crane in the hall and scaffolding was assembled around it to allow the operator to gain access at the correct height. Two dummy fuel elements were then transferred from the storage pits in the floor of the reactor hall to the UNIFETCH with all work done according to the detailed operating instructions accompanying the safety case and controlled and recorded in a quality plan to be used with active elements. The transfer and all the mechanisms worked perfectly and the UNIFETCH was returned to AEA. The comments of the external assessor were received late in November and a meeting of the Nuclear Safety Committee took place early in December at which the safety case and the assessor’s report were discussed at length. The case was judged acceptable, subject to minor revisions which were incorporated in an up-dated version, and the minutes of the meeting were submitted to the NII. NII raised various issues in connection with the details of the safety case and, once these had been answered, a formal request was made to the NII for permission to proceed with the transfer of the first six fuel elements. A certificate of approval for carriage of the material was received from the Department of Transport on 20th December 1995 and a licence instrument agreeing to the first transfer was received from NII on 8 January 1996, on which day six fuel elements were
104
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
transferred from the reactor core to the storage pits in the floor. A note of the radiation doses received by SURRC personnel during this operation was sent to NII by fax the same day. The UNIFETCH arrived at SURRC on 9th January. The scaffolding was erected, tests of the rotating shield were performed and AEA personnel were given instruction in the SURRC safety procedures. The transfer from the floor storage pits to the UNIFETCH was successfully accomplished the following day and the flask was taken to Dounreay on 11th January to be unloaded in the pond there according to documented instructions for removing and reassembling the shield. This constituted a hold point in the operation: a full report, including dose measurements, was sent to NII with a request for permission to proceed with the transfer of the other six elements. Permission was granted by NII, the UNIFETCH returned from Dounreay, and the remaining six elements were transferred to the floor storage pits from the core, all on 15th January. When functional tests were performed the following day on the rotating shield it was found that the slewing ring bearing had seized after immersion in the fuel pond. This problem had been anticipated and an exact replica was available. With the agreement of NII the replacement was installed, tested and used for the second transfer which took place on 16th January. The flask was sent to Dounreay on 17th January and a report on dosimetry was submitted to NII.
order to record any dose arising from movement of the fuel across possible interface gaps. Local health physics rules were written in compliance with the Ionising Radiations Regulations. These rules established a controlled area and an additional area which was barriered off and had further entry requirements in respect of clothing. Respirators fitted with particulate filters were available but were not required during normal operations. The transfer of fuel from the core to the storage pits was not included in the dose budget derived from the safety case but the doses measured in this operation (for six elements) are shown in Table 1. Fuel is loaded into the SURRC flask by means of a grab on the end of a rod which runs down through the centre of the flask and the wrist dose to the operator arose from shine past this rod as the grab was being disconnected after a fuel element had been lifted into the flask. The ankle dose was a result of shine and scatter in the enclosed space on top of the core where the transfer took place. The doses recorded during transfer of the same six elements from the pits to the UNIFETCH are given in Table 2 together with those derived from the safety case. Part of the large discrepancy arose from the tendency to err on the side of caution when estimating doses. In particular, operations were probably broken down in too fine detail so that, when practically measurable doses were attributed to each step, these rapidly accumulated to give a substantial overestimate.
6. Health physics arrangements and radiation doses Area dose monitoring was provided by background g monitoring equipment in the reactor hall, neutron dose rate monitors, hand held gamma monitors and an air sampler. All operational staff were fitted with whole body thermoluminescent (TLD) and digital dosimeters and finger and head TLDs. The SURRC operator responsible for raising and lowering fuel elements and the AEA operator responsible for rotating the shield were issued also with ankle TLDs in
Table 1 Doses recorded during the transfer of six fuel elements from the reactor core to the storage pits in the floor of the reactor hall
SURRC SURRC SURRC SURRC
Op 1 Op 2 RPS RPA
Whole body dose (mSv)
Wrist
Ankle
24 9 11 18
110 – – –
58 – – –
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105 Table 2 Doses recorded during the transfer of six fuel elements from the storage pits in the floor of the reactor hall to the UNIFETCH flask
SURRC Op 1 SURRC Op 2 SURRC RPS SURRC RPA AEA Op 1 AEA Op 2 AEA project manager AEA HP AEA RPA
Whole body dose (mSv)
Calculated (mSv)
9 6 5 3 11 4 4
194 – 194 – 163 163 163
4 2
153 153
7. Conclusions The successful design, construction and use of a dry loading facility for a flask which is normally loaded under water, along with the preparation
105
and assessment of the safety case and the accompanying procedures, were completed in a little over 1 year. Doses were well within the restraint objective derived from the safety case and the difference between the measured and calculated values can be attributed to: a longer cooling time than was assumed in the safety case; overestimates of the time involved in the performance of the various tasks; too much detail in the assignment of doses and uncertainties in the shielding calculations.
References HSE, 1992. Safety Assessment Principles for Nuclear Plants, HMSO Publications. Kletz, T.A., 1986. HAZOP, HAZAN, Notes on the Identification and Assessment of Hazards, 2nd Ed. The Institute of Chemical Engineers. Webster, E.B. 1995. FISPIN — A Code for Nuclide Inventory Calculations, Introductory Guide, February. Available from AEA Winfrith.
.
Defuelling of the UTR-300 research reactor R.D. Scott a,*, H.M. Banford a, B.W. East a, M.A. Ord b, A.P. Gaffka b a
Scottish Uni6ersities Research and Reactor Centre, East Kilbride, Glasgow G75 0QF, UK b AEA Technology, Harwell, Didcot, Oxfordshire OX11 0RA, UK Accepted 11 August 1997
Abstract A description is given of the movement of fuel elements from the core of the UTR-300 research reactor to the UNIFETCH flask, which is normally loaded under water, through a specially designed shielding arrangement which permits a dry transfer. The regulatory requirements and the safety case are summarised along with the predicted and measured doses to operators. The task was successfully completed to a tight time schedule with recorded doses which were well within the allocated dose budget. © 1997 Elsevier Science S.A.
1. Introduction The UTR-300 reactor at the Scottish Universities Research and Reactor Centre was based on the original Argonaut design with two aluminium core tanks set in a graphite reflector each containing six fuel elements cooled and moderated by water flowing up through the tanks in a closed primary circuit. The fuel plates in the original 13-plate elements were uranium oxide-aluminium with a 22 g loading of 90% 235U. After 7 years of operation at 100 kW (10 kW average from 3 runs of 6 h per week), the maximum power was increased to 300 kW (30 kW average) and, in order to maintain the operational excess reactivity, it was necessary to add another plate to each element progressively over the years until they all contained 14 plates. These extra plates were uranium metal-aluminium with 24.5 g of 90% 235U. * Corresponding author. Tel.: + 44 1355 223332; fax: + 44 1355 229898.
Calculations of critical mass as a function of the fuel plate spacing indicate that the critical mass is near its minimum value for a 13-plate element and that the critical mass increases sharply as the spacing is decreased by the addition of further plates (the overall dimensions of the element are fixed) and the core becomes more under-moderated. Addition of a 15th plate would have had a marginal effect on reactivity and, in order to appreciably extend the operating life of the reactor, it would have been necessary to replace some of the elements with fresh ones made from our stock of unirradiated plates. However, irradiated fuel could not be stored on site and shipment of this material to the owners, the UK Atomic Energy Authority (AEA), would have triggered immediate reprocessing charges, with more to come in the future. The financial implications, coupled with low research usage and little demand for its use in teaching, made shutdown inevitable and the decision was taken in March
0029-5493/97/$17.00 © 1997 Elsevier Science S.A. All rights reserved. PII S 0 0 2 9 - 5 4 9 3 ( 9 7 ) 0 0 2 0 9 - 4
100
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
1995 that the reactor should operate for the last time early in September 1995. The unloading of the fuel and shipment to Dounreay Nuclear Establishment for reprocessing are described in this paper and constitute the first stage of the decommissioning of the UTR-300 reactor.
2. Regulatory requirements The operation was carried out as a plant modification in compliance with an arrangement made against one of the site licence conditions and was assessed in terms of safety as a category 1 modification, having major significance. Any such modification requires a proposal containing a detailed safety case which, after consideration both by the SURRC Nuclear Safety Committee and by an independent consultant, is submitted to the Nuclear Installations Inspectorate (NII). Depending upon the response from NII, work may proceed immediately or must not proceed without an agreement from NII issued in the form of a licence instrument. In the latter case, ‘hold points’ are sometimes identified beyond which work may not proceed without the further agreement of NII. Preliminary investigations established that the container most suitable for the transfer of highly enriched irradiated fuel and which could be handled in the Dounreay pond was the UNIFETCH flask belonging to AEA. This flask could, however, take only six elements at a time and was normally loaded under water so two shipments were required and a shielding system compatible with dry loading had to be designed. It was also necessary to obtain a certificate of approval from the Department of Transport for the design of the package for carriage of radioactive materials based upon activity estimates and supported by a criticality calculation.
specially designed shielding arrangement. The complete assembly is shown in Fig. 1. It comprised a basket, fixed relative to the UNIFETCH inner wall, with six circumferentially disposed fuel pockets and a blank position, a ratchet ring, a slewing ring supporting a rotating hub to which was attached a primary shield with a single hole matching the size of the fuel pocket and a removable top shield mounted on the primary rotating shield, again with a single hole to match the fuel pocket. The idea was to lower one fuel element at a time from the SURRC fuel transfer flask, which holds a single element, into each of the six pockets by rotating the shield after each insertion, so ending up with the hole over the blank position and the six elements shielded. The removable top shield had a 25 mm deep ‘footprint’ of the transfer flask cut into its top surface so that, as well as ensuring accurate positioning and providing extra shielding during the loading, it largely eliminated the g flash as the element was lowered into the basket. Fig. 2 shows more detail of arrangement. The ratchet and slewing rings were fixed relative to the basket and the hub could rotate on the slewing ring bearing (in one direction only) with the primary shield attached. Orientation of the various components was achieved by means of dowel pins before they were bolted together, although the removable shield was not bolted to the primary shield since it was held in place by the weight of
3. Design of the UNIFETCH basket and rotary shield Since there was no pond at SURRC to permit loading of the UNIFETCH flask under water, it was necessary to do a dry loading through a
Fig. 1. General view of the UNIFETCH flask, the fuel basket and shielding with the SURRC fuel flask in position.
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
Fig. 2. Detail of the rotating shield and removable top shield.
the SURRC flask. The position of a dowel pin on the primary shield relative to the wall of the UNIFETCH was marked with adhesive tape to indicate each of the six positions and precise alignment was provided by a spring plunger which dropped to engage the ratchet ring when the holes through which the fuel was loaded were aligned. The locking bar shown in Fig. 2 also engaged the ratchet ring when fuel transfer was taking place. There was no access to the spring plunger when the removable shield was in place and great care was therefore necessary when the system was rotated since, if the next empty position was passed, it was not possible to go back. Once the final blank position was reached, the spring plunger dropped into a slot in the ratchet ring and movement of the shield in either direction became impossible.
4. The safety case
101
pits in the floor of the reactor hall by means of the transfer flask referred to in Section 3 above and shown in Fig. 1, and it was decided that the loading into the UNIFETCH flask would be carried out in two stages, from the core to the storage pits and from there to the UNIFETCH by means of this flask. This meant that the fuel could be kept in the reactor from shut-down until the transfer with the core ventilation running continuously and the fuel tanks dry. The transfer from core to floor storage was carried out by SURRC staff according to documented procedures under the maintenance arrangements and lay outside the scope of the safety case, which was therefore required to cover only movement of fuel from the floor to the UNIFETCH together with procedures for handling the UNIFETCH and the rotating shield. The matter of training was considerably simplified by a division of labour between SURRC and AEA staff based upon past experience: all operations involving the SURRC transfer flask were undertaken by SURRC staff whereas the UNIFETCH was handled by AEA staff. It was possible to maintain this practice even when the two flasks were mated during the transfer of a fuel element. The case was produced by AEA and consisted of a combined preliminary safety report, pre-commissioning safety report and pre-commencement safety report containing a hazard identification (HAZOPS) study. It was backed by a quality plan and detailed operating instructions for the various tasks involved.
4.2. Calculation of fuel element acti6ity and dose rates
4.1. Requirements and scope The arrangement to comply with the Site Licence condition under which the operation was carried out requires a safety case based upon a comprehensive review and assessment of the hazards, particularly those which might result in increased radiation exposure to operators or to members of the public. During annual maintenance the fuel elements were routinely transferred into shielded storage
Radiation doses from both shielded and unshielded fuel elements constituted an important aspect of the safety case and the activity of the fuel was therefore calculated at AEA Winfrith by means of the FISPIN code (Webster, 1995). The calculation was performed with the neutron flux modified by the duty cycle of the reactor (18 h per week) and it is interesting to note that this gave an underestimate of the activity at shutdown of fission (or activation) products with half-lives
102
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
much less than the ‘on’ period of the duty cycle (6 h) since they saturate at a value corresponding to the true neutron flux. The effect is evidently a function of the half-life of the species and a detailed calculation shows that for our duty cycle it can be ignored for half-lives in excess of a few days. The code also gave the g-ray spectrum in 24 energy groups as a function of time from the end of operation and this spectrum was used for shielding calculations assuming a decay time of 1 month. FISPIN is a point code and account was taken of the distribution of activity in the fuel by assuming a cosine distribution of activity along the axis of the fuel elements with an extrapolation distance of 20 cm. Calculations in an accurately modelled geometry were made of the dose rate directly above a loaded fuel pocket in the UNIFETCH basket with and without the removable shield in place and of the dose rate directly above the last open empty pocket when the other five were filled, but shielded. The total predicted activity of 12 elements after 31 days was 4.5× 1014 Bq and the above rates were 390 mSv h − 1, 3.4 and 23 mSv h − 1, respectively. Measurements of dose rate performed during the fuel movement indicated the validity of the calculations. For instance, taking account of the fact that the actual cooling time of the fuel was 120 days and that this reduced the b activity by almost one half, but ignoring any change in the g-ray spectrum, the calculated dose rate in the last case mentioned above was 13 mSv h − 1 against the measured value of 12. This is very good agreement considering that the dose in this instance arose from g-rays scattered out of the empty pocket. The dose from an unshielded or partially shielded element was also obtained for use in the analysis of fault sequences in the HAZOPS study discussed below.
4.3. The HAZOPS study The HAZOPS study (Kletz, 1986) was a highly structured procedure which lasted 2 days and employed a committee of AEA and SURRC representatives covering design, operations and health physics aspects. The entire operation, start-
ing with the arrival on site of a low-loader carrying the UNIFETCH and finishing with the preparation of the flask for departure, was broken down into ten independent steps, each containing a number of identifiable tasks. The effect of deviations, defined by parameters (e.g. pressure) and guide words (e.g. higher), on each of these tasks was considered in terms of consequences, safeguards and recommendations. A total of 19 recommendations requiring either satisfactory responses or action emerged from the study and covered matters such as contamination, correct location of the shield relative to the basket and the need for a spring-loaded plunger to engage the ratchet ring. The analysis of possible accidents was based on a fault schedule produced from the HAZOPS worksheets in the form of a list of 132 faults, safeguards (engineered or management controls) and consequences, not all of which contributed significantly to risk. Low risk hazards were therefore screened out on the basis of a dose limit of 100 mSv for operators and 10 mSv for members of the public or a frequency of B 10 − 6 per year. A second screening using the product of dose and frequency was carried out and the remainder (14 in number) were treated in detail in an accident hazard assessment which produced a set of individual risks calculated from the product of frequency, dose and dose risk factor (taken to be 0.05 per Sv). The total individual risk to an operator arising from these 14 fault sequences was 4.4×10 − 6 per year with the dominant contribution (70%) arising from three accident sequences which could have led to exposure either to an unshielded or partially shielded fuel element. This slightly exceeded the NII basic safety objective of 10 − 6 per year (Health and Safety Executive, HSE, 1992). However, taking into account the fact that the sequences are dominated by operator errors, the probabilities of which were conservatively chosen, and the extra cooling time of the fuel, it was concluded that the objective was closely approached and unlikely to be exceeded. The NII basic safety limit of 10 − 4 per year was comfortably met.
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
4.4. Dose assessment during normal operation Each operation, starting with the arrival in the reactor hall of the low loader carrying the UNIFETCH flask and ending with the monitoring of the loaded, sealed flask was allotted a whole body dose based on the dose rate associated with the task and the estimated time for its completion. The estimates were pessimistic in that upper bounds were used both for calculated dose rates and times and resulted in a dose of 600 mSv summed over the SURRC operator working with the SURRC transfer flask, the AEA operator involved with the UNIFETCH, a supervisor and a health physics monitor. The maximum single dose was 239 mSv and a dose restraint objective of 250 mSv (for the transfer of six elements) was therefore set. The dose estimates could, of course, have been invalidated if there were any shine paths through the basket and lid assembly and a test was therefore carried out with a 1 GBq 60Co source in one of the fuel pockets. A scintillation probe was used to check round the outer rim of the removable shield and of the primary rotating shield. In the first instance only the background counting rate was observed and in the second, 500 cps were measured at a region on the joint between the lid and the flask body adjacent to the pocket containing the source, with a negligible number elsewhere around the rim. The sharp fall on moving the probe a few centimetres away from the position of maximum counting rate suggested that the counts were due to directly transmitted radiation, rather than to scatter, and this conclusion was supported by a shielding calculation.
5. Procedure and timetable Since there was some doubt as to the continued availability of reprocessing facilities at Dounreay, a deadline of January 1996 was set for the removal of fuel from the site. Discussions between SURRC and AEA began about a year in advance of shutdown and led to the preparation and acceptance of a formal quotation and initiation of the design and construction of the basket and
103
shield. The work proceeded in parallel with the preparation of the safety case which was submitted to the NII and to an external assessor at the end of October 1995. This was followed by an emergency exercise, inspected by NII, which had as the scenario a fire caused by an electrical short circuit, a radiation hazard from a partially shielded element and a personal injury. It was designed to test the various hazard and rescue procedures, emergency shielding arrangements and liaison with the local fire brigade (which was in attendance), and was judged satisfactory. Early in November 1995, the UNIFETCH flask, complete with basket, primary rotating shield and top shield was brought for testing to SURRC on a low loading trailer which was manoeuvred into position in the reactor hall and disconnected from the tractor. The UNIFETCH had to remain on the trailer because its weight far exceeded the capacity of the crane in the hall and scaffolding was assembled around it to allow the operator to gain access at the correct height. Two dummy fuel elements were then transferred from the storage pits in the floor of the reactor hall to the UNIFETCH with all work done according to the detailed operating instructions accompanying the safety case and controlled and recorded in a quality plan to be used with active elements. The transfer and all the mechanisms worked perfectly and the UNIFETCH was returned to AEA. The comments of the external assessor were received late in November and a meeting of the Nuclear Safety Committee took place early in December at which the safety case and the assessor’s report were discussed at length. The case was judged acceptable, subject to minor revisions which were incorporated in an up-dated version, and the minutes of the meeting were submitted to the NII. NII raised various issues in connection with the details of the safety case and, once these had been answered, a formal request was made to the NII for permission to proceed with the transfer of the first six fuel elements. A certificate of approval for carriage of the material was received from the Department of Transport on 20th December 1995 and a licence instrument agreeing to the first transfer was received from NII on 8 January 1996, on which day six fuel elements were
104
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105
transferred from the reactor core to the storage pits in the floor. A note of the radiation doses received by SURRC personnel during this operation was sent to NII by fax the same day. The UNIFETCH arrived at SURRC on 9th January. The scaffolding was erected, tests of the rotating shield were performed and AEA personnel were given instruction in the SURRC safety procedures. The transfer from the floor storage pits to the UNIFETCH was successfully accomplished the following day and the flask was taken to Dounreay on 11th January to be unloaded in the pond there according to documented instructions for removing and reassembling the shield. This constituted a hold point in the operation: a full report, including dose measurements, was sent to NII with a request for permission to proceed with the transfer of the other six elements. Permission was granted by NII, the UNIFETCH returned from Dounreay, and the remaining six elements were transferred to the floor storage pits from the core, all on 15th January. When functional tests were performed the following day on the rotating shield it was found that the slewing ring bearing had seized after immersion in the fuel pond. This problem had been anticipated and an exact replica was available. With the agreement of NII the replacement was installed, tested and used for the second transfer which took place on 16th January. The flask was sent to Dounreay on 17th January and a report on dosimetry was submitted to NII.
order to record any dose arising from movement of the fuel across possible interface gaps. Local health physics rules were written in compliance with the Ionising Radiations Regulations. These rules established a controlled area and an additional area which was barriered off and had further entry requirements in respect of clothing. Respirators fitted with particulate filters were available but were not required during normal operations. The transfer of fuel from the core to the storage pits was not included in the dose budget derived from the safety case but the doses measured in this operation (for six elements) are shown in Table 1. Fuel is loaded into the SURRC flask by means of a grab on the end of a rod which runs down through the centre of the flask and the wrist dose to the operator arose from shine past this rod as the grab was being disconnected after a fuel element had been lifted into the flask. The ankle dose was a result of shine and scatter in the enclosed space on top of the core where the transfer took place. The doses recorded during transfer of the same six elements from the pits to the UNIFETCH are given in Table 2 together with those derived from the safety case. Part of the large discrepancy arose from the tendency to err on the side of caution when estimating doses. In particular, operations were probably broken down in too fine detail so that, when practically measurable doses were attributed to each step, these rapidly accumulated to give a substantial overestimate.
6. Health physics arrangements and radiation doses Area dose monitoring was provided by background g monitoring equipment in the reactor hall, neutron dose rate monitors, hand held gamma monitors and an air sampler. All operational staff were fitted with whole body thermoluminescent (TLD) and digital dosimeters and finger and head TLDs. The SURRC operator responsible for raising and lowering fuel elements and the AEA operator responsible for rotating the shield were issued also with ankle TLDs in
Table 1 Doses recorded during the transfer of six fuel elements from the reactor core to the storage pits in the floor of the reactor hall
SURRC SURRC SURRC SURRC
Op 1 Op 2 RPS RPA
Whole body dose (mSv)
Wrist
Ankle
24 9 11 18
110 – – –
58 – – –
R.D. Scott et al. / Nuclear Engineering and Design 178 (1997) 99–105 Table 2 Doses recorded during the transfer of six fuel elements from the storage pits in the floor of the reactor hall to the UNIFETCH flask
SURRC Op 1 SURRC Op 2 SURRC RPS SURRC RPA AEA Op 1 AEA Op 2 AEA project manager AEA HP AEA RPA
Whole body dose (mSv)
Calculated (mSv)
9 6 5 3 11 4 4
194 – 194 – 163 163 163
4 2
153 153
7. Conclusions The successful design, construction and use of a dry loading facility for a flask which is normally loaded under water, along with the preparation
105
and assessment of the safety case and the accompanying procedures, were completed in a little over 1 year. Doses were well within the restraint objective derived from the safety case and the difference between the measured and calculated values can be attributed to: a longer cooling time than was assumed in the safety case; overestimates of the time involved in the performance of the various tasks; too much detail in the assignment of doses and uncertainties in the shielding calculations.
References HSE, 1992. Safety Assessment Principles for Nuclear Plants, HMSO Publications. Kletz, T.A., 1986. HAZOP, HAZAN, Notes on the Identification and Assessment of Hazards, 2nd Ed. The Institute of Chemical Engineers. Webster, E.B. 1995. FISPIN — A Code for Nuclide Inventory Calculations, Introductory Guide, February. Available from AEA Winfrith.
.