Safety studies related to the spent fuel storage of the Romanian VVR-S reactor

Safety studies related to the spent fuelstorage of the Romanian VVR-S reactor

Daniela Ene*

National Institute of R&D for Physics and Nuclear Engineering —‘‘Horia Hulubei’’,

IFIN-HH, RO-79617, Magurele, sect.5, Bucharest, Romania

Received 24 March 2003; accepted 24 April 2003

Abstract

The paper presents the methods developed and calculations carried out, using theSCALE4.3 code system to analyse the possibility of safe storage of the Romanian VVR-Sreactor spent fuel load inside the CASTOR MTR2 cask. A method has been developed toderive specific problem-dependent nuclear libraries, defined for groups of fuel assemblies. The

SAS2H calculation module has been used to simulate different irradiation histories of allspent fuel assemblies and to process the resulting libraries for input to ORIGEN-S codedepletion-decay calculations. Using this procedure the characteristics of the spent fuel

assemblies for six cooling times have been calculated. Obtained results are presented and dis-cussed in the paper. The CSAS6 module with KENO-VI criticality code based on the44-group energy cross-sections library has been used to solve criticality aspects. Some relevant

problems to justify the calculation model have been studied and are presented in the paper.The sequence SAS4 with MORSE, a Monte Carlo code based on the (27n–18g)-coupledgroup energy cross-sections, has been used for solving the shielding problem. Gamma and

neutron dose rates estimated on the surface of the container, and respectively at 1 m, 2 m, and3 m in both radial and axial directions for normal loading, and for audit conditions are pre-sented and discussed.# 2003 Elsevier Ltd. All rights reserved.

1. Introduction

Romanian VVR-S reactor has been shut down since December 1997 after morethan 40 years of operation without any major modification from the original Russian

Annals of Nuclear Energy 30 (2003) 1623–1643

www.elsevier.com/locate/anucene

0306-4549/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.

doi:10.1016/S0306-4549(03)00128-2

* Fax: +40-2145-7440.

E-mail address: [email protected] (D. Ene).

http://www.sciencedirect.com



http://www.elsevier.com/locate/anucene/a4.3d

mailto:[email protected]

design. Over the next four years, the status of the reactor has been ‘‘under care andmaintenance’’ and studies and analyses have been performed in order to decidewhich is the best strategy to be followed: (i) up-grading for restarting, or (ii)decommissioning. In 2002 the Governmental decision to start reactor decom-missioning has been taken and necessary planning activities are now in progress.The cumulative amount of spent fuel arising during the entire reactor operationalperiod is stored on site, away from reactor ponds. Until now no decision has beentaken concerning the final disposal of reactor spent fuel assemblies. There are twopossible strategies:(i) returning to the vendor, or (ii) emplacement into a geologicalrepository.

For the first variant mentioned above a container licensed for transport throughEurope is necessary. The second one needs a more safe interim solution, becausespent fuel clad corrosion has already occurred in the ponds. The dual purposetransport/storage CASTOR MTR2 German cask can be a suitable tool for bothaccounted situations and therefore methods have been developed and calculationshave been carried out in order to analyse its compatibility to the specific case.

2. Problem description

The dry transport/storage container CASTOR MTR 2 (Schneider and Hofmann,1994) is a cast iron cylinder with two stainless steel lids (Fig. 1). The interior iscompletely occupied by a cylindrical aluminium basket with seven cylindrical load-ing channels, one of them is central and the other six are arranged equally spacedaround the central one (Fig. 2). Loading units carrying the fuel assemblies will beinserted into each of these loading channels. A loading unit is an aluminium cylinderwith one central boron rod and some internal holes tailored in accordance with thegeometrical shape of the fuel assemblies (Fig. 3). Most of the fuel assemblies of theRomanian VVR-S reactor have a square outer shape, disturbed by one, two, orthree bevelled corners, whilst the remainder have a right square shape. Accordinglythere are two types of loading units: Type B carrying four fuel assemblies with aright square shape, and Type C for six fuel assemblies with the disturbed squareshape (see Fig. 4). Independently of the geometrical outer shape, there are two typesof fuel assemblies differing by the rod type: EK-10 fuel assembly consists of 16 rodswith MgO and UO2 (10% enriched uranium), and S-36 fuel assembly consists of 15rods containing a U–Al alloy (36% enriched uranium). For both types of fuelassemblies the clad and the structure of the assembly (i.e. the shroud and the rack ofthe fuel pins) are made from aluminium. A total of 152 EK-10 type and 70 S-36 typefuel assemblies have been irradiated in the reactor. Only for 73 EK-10 fuel assem-blies are the primary data available, while for the rest of 79 EK-10 fuel assemblies noinformation has been kept until now.

For the evaluation of the possibility of storage of the Romanian VVR-S reactor spentfuel in the CASTOR MTR2 cask it is necessary to study if the safety criteria are observed.

Calculation parameters necessary to be evaluated and analysed in order todemonstrate the compliance with the safety requirements under both normal and

1624 D. Ene /Annals of Nuclear Energy 30 (2003) 1623–1643

accident conditions are connected with: (i) criticality status, (ii) shielding againstradiation, (iii) heat transfer. The investigations mentioned above need properlyqualified, internationally accepted methods, computer codes and nuclear data. Forthis reason, the SCALE 4.3 code system that provides several analytical sequences(modules that link codes and data) for use in analysis for the nuclear fuel packageshas been implemented and used for solving the problem. In addition, to be on thesafe side, conservative assumptions have been made. Thus the paper presents espe-cially the results related to the variant of the CASTOR MTR2 cask that carries theType C loading units.

3. Spent fuel characteristics

Determination of the isotopic composition of the materials present in the VVR-Sreactor spent fuel and subsequent derivation of the heat generation and radiationsource terms has been the first study performed.

Further analyses of the chosen storage option such the evaluation of the burn-upcredit, the shielding and the heat transfer studies have to be based on the derivedamounts mentioned above. Additionally, detailed characterization of the entire

Fig. 1. CASTOR vertical section (with basket and central loading unit), all dimensions in cm.

D. Ene /Annals of Nuclear Energy 30 (2003) 1623–1643 1625

radioactive inventory as a function of time is required for the knowledge of the timeevolution of the fuel status during the storage.

3.1. Derivation of the VVR-S reactor problem-dependent libraries

The SAS2H module (Hermann and Park, 1997) of the SCALE 4.3 system hasbeen applied to produce several time dependent libraries as a function of the specificVVR-S reactor design characteristics, operating parameters and material composi-tion for input to ORIGEN-S (Hermann and Westfall, 1997) cases.

Based on the standard SCALE library (27 groups) the SAS2H sequence processesthe resonance cross sections (BONAMI, NITAWL), computes the neutron spectrumin an infinite lattice approximation, via 1D discrete ordinate code (XSDRNPM) andperforms a depletion calculation (ORIGEN-S) using the obtained three groups col-lapsed cross sections (COUPLE). Applying repeatedly this procedure as many timesas requested during the simulation of the operating history of the fuel assembly,produces cross section libraries for the specified irradiation intervals. These createdlibraries are used in subsequent ORIGEN-S runs to perform the depletion anddecay calculations.

Fig. 2. Basket (horizontal section), all dimensions in cm.


The procedure developed to model the different distributions of irradiationhistories of all spent fuel assemblies of a specific type has the following steps:

� defining the groups of fuel assemblies that have common irradiation history;� processing of the specific burn-up dependent library (SAS2H) for each group

of fuel assemblies, by using data of a representative fuel assembly inside thegroup;

� performing the depletion-decay calculations (ORIGEN-S) for all fuelassemblies of the group by using as input the corresponding burnup-depen-dent library.

Fig. 3. Loading unit, all dimensions in cm.


The groups with common irradiation history have been established based on thefuel irradiation analysis. Average power has been considered the relevant parameterthat identifies the complicated irradiation history of each fuel assembly (see Figs. 5and 6).

The representative fuel assembly of each group used for the simulation withSAS2H has been defined by an averaged power calculated as the arithmetic mean ofthe real values of the fuel assemblies inside the group (see the histograms of theFigs. 7 and 8) and an irradiation time given by the maximum operation time of thegroup components. According to the code requirement the entire residence time at

Fig. 4. (a) EK-10 fuel assembly, cross-section; (b) S-36 fuel assembly, cross-section.


the representative power has been split into cycles of 1000 days. Based on thismethod, for EK-10 fuel assemblies with known irradiation history the associatedproblem-dependent libraries have been produced for four groups:

group I
containing five fuel assemblies with averaged power <0.01MW/assembly;
group II
containing 33 fuel assemblies with averaged power of (0.01–0.015)MW/assembly;
group III
Fig. 5. Averaged power distribution of the EK-10 fuel assemblies.

Fig. 6. Averaged power distribution of the S-36 fuel assemblies.


group IV
containing six fuel assemblies with averaged power >0.02MW/assembly.
Derived oriented libraries for the specific S-36 VVR-S spent fuel assemblies havebeen obtained for three groups defined as follows:

group I
containing 12 fuel assemblies with averaged power <0.015MW/assembly;
group II
Fig. 7. Averaged power distribution of the EK-10 representative fuel assemblies.

Fig. 8. Averaged power distribution of the S-36 representative fuel assemblies.


group III
containing 28 fuel assemblies with averaged power >0.02MW/assembly.
3.2. Calculations undertaken

Using the created burnup-dependent libraries ORIGEN-S calculations have beenperformed to compute the radionuclide inventory, thermal power, gamma sourcespectra and neutron source strength.

For EK-10 fuel assemblies irradiated during the first period (7–8 years) of thereactor operation for which no primary data (irradiation history, burn-up, etc.) wereavailable an average burn-up has been necessary to be estimated.

Based on this, an ORIGEN-S calculation has been performed using a dummy fuelassembly representing an averaged behaviour (residence in reactor and cooling time).

Thus, each fuel assembly of the set of the 79 fuel assemblies was assumed to beidentical with this dummy fuel assembly.

Calculations have been done for the reference time of 31 December 1999 and fivesupplementary decay times of 5, 10, 25, 50 and 100 years after the first referencetime.

3.3. Results and discussion

The resulting total activities calculated at 31 December 1999 are shown in Fig. 9 forEK-10 spent fuel assemblies that includes the result for the dummy assembly andrespectively Fig. 10 for S-36 fuel type. The time evolution of the total activity anddecay heat arising from all fuel assemblies irradiated in the VVR-S reactor is presentedin Table 1. The results shown in the Table 1 are approximate due to the uncertaintygiven by the average simulation of the 79 fuel assemblies without information.

As expected, the study of the radionuclide inventory has shown that the majorcontribution to the fuel assembly activity is due to the fission products. The activityof the actinides has low levels, while the contribution of activation products to thetotal activity is negligible. Low maximum thermal power values of 2W assembly forEK-10 fuel and 3.98W assembly for S-36 fuel have been obtained at the referencetime. From the analysis of the thermal power results the conclusion has been drawnthat the heat transfer analysis is not necessary.

The detailed quantities of the fissile materials in the spent fuel have been alsoderived for safeguards requirements. Ranges of fuel enrichment values between 3.6and 8.4% for EK-10 fuel type and respectively between 18 and 36% for S-36 fueltype have been found at 100 years cooling time. Gamma and neutron spectra andtotal strengths have been analysed and used as source terms for further shieldingcalculations.

The reliability of the developed method has been tested by the comparison withthe calculations based upon in principle a similar method performed by meansHELIOS and ORIGEN-JR codes in VKTA Rossendorf Germany (Ene and Franke,2001).


For comparison two fuel assemblies: r7 of EK-10 type and c11 of S-36 type withthe highest burn up (50 283 MWd/tU and respectively 176 047 MWd/tU) irradiatedin the reactor were selected. The main issues of the comparison undertaken aresummarized below.

The values of the total activity and thermal power were in a good agreement(less than 5%). The photon source strengths obtained with SAS2H/ORIGEN-Sprocedure were about 65% higher than the VKTA results. The main reason of thisdifference were the low energy photons which were not at all taken into account inthe ORIGEN-JR code because they do not play a role in the shielding calculations.Without these three low energy groups the photon sources agreed within a fewpercent.

Fig. 9. Radioactive inventory of the total activities of the EK-10 fuel assemblies calculated for the

reference time.


The neutron sources resulted from the SAS2H/ORIGEN-S calculations wereabout 20% higher for both assemblies. Two thirds of the source arise from sponta-neous fission and one-third from (�, n) reaction for EK-10 fuel type whereas theS-36 fuel type has only a spontaneous fission source. Differences in the neutronsources reflected deviations obtained in the concentrations of few actinides mainly244Cm. The neutron spectra analysis has shown that the SAS2H/ORIGEN-S spectrahave more high energy contributions.

Taking into account the different database of the compared computational pro-cedures, the deviations of the calculated spent fuel characteristics were consideredacceptable. This has contributed to check the correct application of the SCALE4.3code and to verify the developed calculation procedure.

Fig. 10. Radioactive inventory of the total activities of the S-36 fuel assemblies calculated for the

reference time.


4. Criticality safety analysis

4.1. Modelling, calculations, and results

Module CSAS6 (Hollenbach and Petrie, 1997) based on KENO-VI Monte Carlocode has been used to study the criticality status of the spent fuel inside the cask.

The geometry model shown in Figs. 1–4 and material composition in Table 2 havebeen used for the SCALE calculations.

In order to assure that the cask is sufficiently far from the criticality even for theworst case the calculation model has assumed:

Table 2

Materials used in the model for criticality calculations

No.a
Material Location Density
(t/m3)

Element
Composition
(weight fraction in%)

M1
GGG CASTOR bottom and lateral walls 7 Fe 92.3
C
3.5
Si
2.0
Zr
2.2
M2
TSTE 355 CASTOR the first and second lid 7.83 Fe 96.7
Zr
3.3
M3
GKAlMg3Si Basket 2.7 Al 94.5
Mg
3.1
Si
1.1
Zr
1.3
M4
G-AlSi7Mg Loading unit 2.65 Al 92.5
Si
7.0
Mg
0.5
M5
DISPAL M180 Boron rod 2.5 Bnat 20.0
Al
80.0
a As used in Figs. 1–4.

Table 1

Time evolution of the total activity and decay heat arising from the VVR-S reactor spent fuel

Yeara
Activity (TBq) Thermal power (W)
2000
4130 335.5
2005
3020 234.1
2010
2640 208.1
2025
1828 146.9
2050
1048 85.8
2100
478 41.4
a 1 January of the year is considered.


� the loading of the fresh fuel inside the cask;� the use of the reflective boundary condition to simulate an infinite lattice of

containers;� water flooding of the volumes that are void under normal conditions.

Criticality calculations have been carried out for both types of fuel assembliesbased on the SCALE 44-group energy cross-section library. A supplementary cal-culation has been performed by means 238-group energy library only for the refer-ence case of S-36 fuel type. The obtained results are presented in the Table 3.

4.2. Investigations to justify the calculation model

Sensitivity calculations have been carried out to study some criticality relevantaspects and justify the calculation model.

4.2.1. Material used above and below the active zoneAs is shown in Fig. 1 the calculation model above and below the active zone uses

material M4 (Table 2) based on aluminium. Calculations performed by replacingaluminium above and below the active zone with water, led for a loading of S-36 fuelassemblies to a reduced keff value of 0.8735�0.0018. The complicated mixture of Aland water of these zones has been modelled by using solely aluminium, an optionjustified by the higher keff value found in this case (see Table 3).

4.2.2. Water fillingIn order to study the keff dependence on the volume of water that fills all gaps and

voids inside the cask, a hypothetical case consisting in the artificially increasing ofthe water amount has been run. A 10% increase of the density of the water betweenthe rods of the fuel assembly leads to a significant effect (a keff value of0.9100�0.0019 for an S-36 fuel load). The conclusion drawn from this analysis isthat to be on the safe side, the calculation model has to be conservative and there-fore should not underestimate the void volume.

4.2.3. Water gaps shapeThe regular annular gaps, between the basket cavities and loading units (�=2.75

mm each), as well as between the cavity of the cask and the basket (�=2.75 mm)occur for a centred insertion of these items. Due to their free mobility inside the

Table 3

Resulting keff values

Fuel type
CSASVI (KENO VI)
44GROUPNDF5 SCALE4.3 library
238GROUPNDF5 SCALE4.3 library
S-36
0.8901�0.0019 0.8879�0.0022
EK-10
0.8753�0.0019 –

cavities, the regular annular layer shape might be modified up to the extreme sit-uation when the gap thickness is zero on one side and double on the opposite site.The different gap shapes produce a change of the keff value. In order to account forthis effect a fictitious increase of the total amount of water not only locally but sur-rounding the gaps has been assumed. The result of the case that considers doublesize annular gaps is shown in Fig. 11. For validation reasons a parallel study hasbeen carried out by keeping unchanged the annular water layers but modifying thewater density inside the gaps (see Fig. 12). The two figures mentioned above show:

Fig. 11. keff dependence on the size of the water annular gaps.

Fig. 12. keff dependence on the density of the water inside the regular annular gaps.


(i) keff has a general tendency to increase with the increasing of the water amountinside the gaps and (ii) in spite of this, the keff value remains below the limit of 0.95even for the most conservative assumptions. It has been concluded from this analy-sis that the loading units and the basket can be arbitrarily positioned inside theircavities.

4.2.4. Dimensioning of the absorbing rodsA boron rod introduced in the centre of each loading unit is a necessary additional

absorption in order to assure subcriticality of the cask. For the 3 cm diameter of theboron rod set by the cask design the appropriate boron density has had to bedetermined in a study to this effect. From the analysis of the dependence of keff ofthe boron density inside the boron rods presented in Fig. 13, it has been concluded:(i) variation of the boron density has a small effect on the answer, (ii) the borondensity of 0.5 g cm�3 is a suitable value to be used because a further increase of thedensity is not efficient.

4.2.5. Removal of the fuel rodsThe effect on keff of the absence of the rods inside the fuel assembly has been

studied by filling with water the gap remaining after the rod removal (see Fig. 14). Ithas been concluded from the analysis of this figure that the full complement (15) ofrods produces the maximum keff value.

4.2.6. Additional resultsOne more calculation has been necessary to take into account the loading unit B

case for the EK-10 VVR-S fuel assemblies with square shape. The obtained result iskeff of 0.700�0.002.

Fig. 13. keff dependence on the density of the boron inside the rod.


5. Shielding investigations

The sequence SAS4 with MORSE, a Monte Carlo code based on the (27n–18g)-coupled group energy cross-sections has been used to study the shielding aspects ofthe problem. Only the results on S-36 spent fuel type inside the CASTOR MTR2cask with Type C loading units are reported in the paper.

5.1. Gamma and neutron source calculation

One of the main findings of the radioactive inventory calculations has been thederivation of the gamma and neutron sources. Analysis of those results has allowedfor further discussion in choosing the appropriate values of the radiation sourceterms needed for the shielding investigations of the cask.

Because, up to now the details of the different CASTOR loading patterns arenot yet established it has been considered that the entire loading of a caskconsists of fuel assemblies having identical properties to a representative one.Such a representative fuel assembly used in calculations has been defined asfollows:

� for the normal situation, when upper limits of the expected dose rates aredesired to be calculated, the representative assembly is characterized by themaximum values of gamma and neutron sources;

� for loading and audit conditions, when the detailed loading scheme has to betaken into account, the conservative criterion is not useful and therefore anaverage value of both gamma and neutron sources is taken into account.

Fig. 14. keff dependence on the number of rods inside the S-36 fuel assembly.


The representative fuel assembly for the closed cask case has been chosen the fuelassembly c28 that has the maximum source values. In the case of the open caskconfiguration the fuel assembly c08 has been considered appropriate to model theaverage sources. The source energy spectra and nuclide densities of these fuelassemblies provided by the ORIGEN-S calculations have been used as input for thefurther Monte Carlo transport calculations.

Constant radial and axial distributions for active part of the sources have beenassumed.

The contribution of the end parts of the fuel assembly has been considered alsofor axial dose rate estimates over an open cask. The end of the fuel assemblyconsisting in the aluminium rack of the fuel pins has been modelled considering avolume given by the surface of the fuel assembly and a height of 0.4 cm filled with10% aluminium.

With a view to obtaining the gamma source arising from the end of the fuelassembly an ORIGEN-S calculation has been performed separately addressing thecontent of the volume of the end part defined above (5 g of aluminium). From thealuminum impurities taken into account in the activation calculation 60Co was themain contributor to the gamma source.

All values used as input data in Monte Carlo shielding calculations have beencalculated for the reference time of 31st December 1999.

5.2. Modelling and calculations

The very complex and non-standard geometry of the cask has not allowed theelaboration of the calculation model with the help of standard casks available in theSAS4 (Tang, 1997) sequence of the SCALE4.3 system and therefore the detailedgeometry has been built by using the MARS (West and Emmet, 1997) module.

Using this utility a very complicated geometry model has been created as input forthe Monte Carlo code MORSE (West et al., 1997). The calculation model usedassumes (see Fig. 15) that the materials of fuel assemblies and their sources havebeen smeared over the horizontal cross section of the fuel assembly.

Additionally, other approximations in the calculation model were made, causingthe overestimating of the dose rates:

� the aluminium shroud of the fuel assembly (1 mm thick) has been added tothe surrounding loading unit and thus the smearing of the fuel materials hasbeen done over the surface interior to the shroud;

� symmetric homogenous cylinders 5 cm high have been modelled for the endsof the assembly and the corresponding remained space in the end zones havebeen filled with air;

� vertical symmetry of the cask has been assumed, replacing the existing slightasymmetry of the cask.

A full radiological characterization of the spent fuel cask has been performed bythe following scheme:


� neutron and gamma doses have been computed in separate calculations basedon the SAS4 automated biasing procedure;

� different calculations have been performed for radial and axial detectors.

For both gamma and neutron dose rate estimates over the open cask, only axial cal-culations have been performed. A supplementary calculation, for this last configurationwas needed for the estimation of the gamma dose contributed by the end part material.

The surface detectors have been located radial or axially on the outer surface ofthe cask and 1, 2 and 3 m from the surface. The axial detectors are circular discs,while the radial detectors are side surfaces of cylinders having appropriate radii andheights given by the height of active part of the fuel assembly.

The ANSI standard flux-to-dose conversion factors have been used. Theseresponse factors correspond to the coupled library (27 neutrons–18 gamma) energygroups used for all the calculations performed.

5.3. Results and discussion

The obtained results for both closed and open cask configurations are presented inTables 4 and 5.

The statistical errors (all under 5%) are insignificant compared to the moreimportant errors that come from modelling limitations. The error due to thesmearing of the fuel materials is about 30% (Franke and Seifert, 1997). An error ofabout �15% in the calculated gamma source per assembly arising mainly from theuncertainty of the power of the fuel assembly has been found from the comparisonwith measured gamma sources for some fuel assemblies of the RFR reactor (Frankeand Seifert, 1997). This error has been assumed to occur for the Romanian reactoras well, based on the similarity of those two VVR-S reactors and the fact that inboth cases the powers of the fuel assemblies were derived using simple models. Butthe large number of fuel assemblies of the cask loading (42) allows the compensation

Fig. 15. CASTOR MTR2: geometry model for the shielding calculations.


of the errors induced by the power uncertainties due to the various positions of thefuel assemblies during irradiation.

The values of the dose rates exterior to the cask loaded with VVR-S spent fuel (seeTable 4) are below the limit of 2 mSv h�1, accepted by the National Regulations forNuclear Safety.

In comparison with the gamma dose rates arising from active part of the fuel assem-blies, those that arise from the materials of the end part are negligible (see Table 5).

6. Conclusions

The main achievements of this work are:

� A methodology has been developed to obtain and apply problem-dependentlibraries, specific to the VVR-S reactor;

Table 4

Maximum dose rate values outside the CASTOR MTR 2 cask

Total sources

(g/s)

Surface
Dose rates
+1 m
+2 m +3 m
mSv/h
FSDa mSv/h FSD mSv/h FSD mSv/h FSD
9.7142E+14
Radial 68.63 0.02 6.32 0.02 3.22 0.02 1.89 0.02
Axial
23.32 0.04 4.03 0.04 1.85 0.05 0.97 0.05
(n/s)

7.8246E+04
Radial 2.45E-01 0.01 2.19E-02 0.01 9.59E-03 0.01 5.21E-03 0.01
Axial
2.73E-01 0.01 2.51E-02 0.01 8.78E-03 0.01 4.33E-03 0.01
a FSD fractional standard deviation.

Table 5

Dose rate values over the open CASTOR MTR 2 cask

Total sources

(g/s)

Surface
Dose rates
+1 m
+2 m +3 m
Sv/h
FSDa Sv/h FSD Sv/h FSD Sv/h FSD
Active part

5.8997E+14
1.88 0.05 1.22 0.02 0.64 0.04 0.35 0.05
End parts

1.5107E+09
2.29E-05 0.02 1.28E-05 0.01 6.37E-06 0.02 3.73E-07 0.02
(n/s)

3.1391E+04
3.66E-07 0.01 9.74E-08 0.04 4.08E-08 0.01 2.20E-08 0.02
a FSD fractional standard deviation.


� Detailed evaluation of the radioactive inventory, thermal power, source termshas been done based on these nuclear data sets;

� A calculation model for criticality safety analysis has been established andcalculations have been carried out based on this model;

� Sensitivity calculations have been carried out to justify the chosen model andto study extreme potential conditions;

� A calculation model for shielding safety analysis has been established anddose rate evaluations for normal and loading conditions have been carriedout.

It is concluded from these studies that:

� In respect of criticality safety, even under the most restrictive conditions, theCASTOR MTR 2 cask assures for the VVR-S reactor spent fuel the com-pliance with the safety condition (keff<0.95);

� From the shielding safety point of view also, the CASTOR MTR 2 cask offersa safe containment of the Romanian VVR-S reactor spent fuel. Resultingdose rates exterior to the cask are less than the accepted limit criterion evenfor the most conservative conditions.

Analysing the relevant phenomena related to safety, these studies have demon-strated that the CASTOR MTR2 is a safe option for the VVR-S spent fuel storage.

Additionally the resulting dose rate values will be used to assist the developmentof the working procedures that have to be observed during the loading of the cask aswell as for the audit activities. The conservative approximations used have to betaken into account in order to have more precise basic input needed to establish theconditions of personnel operation in the working area, and to ensure the safety ofthe staff under ALARA principles.

Acknowledgements

The author is grateful to D. Cepraga for advice related to the nuclear libraryprocessing, to E. Franke and E. Seifert for the cask data and information supplied,the discussions and observations while performing this work.

References

Ene, D., Franke, E., 2001. Comparison of different computational systems for calculation of the VVR-S

IFIN-HH reactor spent fuel characteristics. AER Working Group Meeting, Vuje Trnava, Slovakia, 23–

25 April (unpublished).

Franke, E., Seifert, E., 1997, E. Berechnung der Dosisleistung uber den nach oben offenen CASTOR-

Behaltern mit Rossendorfer Brennenelementen. Arbeitsbericht KWK-1/97.

Hermann, O.W., Parks C.V., 1997. SAS2H: A Coupled One-Dimensional Depletion and Shielding

Analysis Module (NUREG/CR-0200), Vol. 2, Section S2, Revision 5. ORNL.


Hermann, O.W., Westfall, R.M., 1997. ORIGEN-S: SCALE System Module to Calculate Fuel Depletion,

Actinide, Transmutation, Fission Product, Decay and Radiation Source Term (NUREG/CR-0200).

Vol. 2, Section F7, Revision 5. ORNL.

Hollenbach, D.F, Petrie, L.M., 1997. CSAS6: Control Module for Enhanced Criticality Safety Analysis

with KENOVI. ORNL/NUREG/CSD-2/VI/R5.

Schneider, B., Hofmann, G., 1994. Konditionierung bestrahlter Brennenelementen aus dem Rossendorfer

Forschungsreaktor RFR in Transport-und Lagerbehalter CASTOR MTR 2, VKTA-17.

Tang, J.S., 1997. SAS4—A Monte Carlo Cask Shielding Analysis Module Using Automated Biasing

Procedure (NUREG/CR-0200), Vol. 2, Section S4. ORNL.

West, J.T., Emmet, M.B., 1997, MARS: A Multiple-Array System using Combinatorial Geometry

(NUREG/CR-0200), Vol. 2, Section M9. ORNL.

West, J.T., Hoffman, T.J., Emmet, M.B., 1997. MORSE-SGS for the SCALE System (NUREG/CR-

0200), Vol. 2, Section F9, Revision 5. ORNL.


Safety studies related to the spent fuel storage of the Romanian VVR-S reactor

Documents

Transcript of Safety studies related to the spent fuel storage of the Romanian VVR-S reactor