RS06RA115

XLV KONFERENCIJA ZA ETRAN, BUKOVlCKA BANJA, 4 - 7. / W, ^001.

GAMMA-RAY DOSE RATE ESTIMATION AT SURFACE OF STORAGE CASK FILLED WITH RADIOACTIVE SLUDGE Milan Peiic, Ilija Plecas, Radojko Pavlovic, Vw£a Institute, Belgrade, Yugoslavia Marina* SokCic-Kostic, Forschungszentrum ZZK/HZY/BTE, Karlsruhe, Germany

Abstract - - contaminated sludge by l,7Cs from 137Cs nuclide. Activities of wet sludge ('ws') and dry and Co from the spent fuel storage pool of the RA research sludge ('ds') samples, measured by coaxial Ge gamma-ray reactor is conditioned and stored in specially designed casks. spectrometers in the Vinca Institue [4, 6, 131 and in the IAEA Purpose of this paper is to describe an attempt to estimate a laboratories (51, are shown in Table I gamma-ray ambient dose equivalent rate from the cask with the conditioned sludge by the MCNP code and compare the Table 1. Measured gamma-ray activity in the sludge samples result to the measuring data Sludge Activity ± uncertainty (la) 1. INTRODUCTION sample I37Cs "Co

Almost all spent fuel elements, used during ws #1 [4]: 1.80±0.20 kBq/mL 15.0± 1.5 kBq/L operation of the RA reactor [1] from 1959 to 1984, are stored ws #1 [51: 1.83 ±0.27 kBq/mL in the temporary spent fuel storage pool. Initially, spent fuel ws #2 [51: 0.94 ±0.14 kBq/mL elements were stored in 300 original (made in Russia) ds#l (6J: 13.4 ±2.3 kBq/g 122.7 ±21.2 Bq/g stainless steel channel-type containers, filled with de-

From the beginning of sixties until 1984, about 5000 Recent inspections of various water samples, taken oldest LEU spent fuel elements were repacked from stainless from spent fuel aluminium storage barrels and stainless steel steel containers into 30 sealed aluminium barrels. About 1600 containers, confirmed that 137Cs activity in the pool water LEU spent fuel elements and 1000 HEU spent fuel element* originates from leakage of fission products. They passed remained in stainless steel containers in the storage pool, up through the aluminium cladding (1 mm thick) of the spent to now. It is estimated [21 that the total activity of all spent fuel elements and aluminium walls of storage barrels, LEU and HEU fuel in the pool is approximately 2500 ± 250 damaged in corrosion processes. TBq by end of 1998. Most of the activity (about 99%) 137 w originates from Cs and Sr nuclides, while activity 3 SLUDGE CONDITIONING AND STORAGE attributed to wKr is about 1%. Total volume of the sludge in the pool, based on 2. INSPECTION OF THE POOL average sludge height on the bottom of the pool and area of the pool surfaces, is estimated to be about 3 m3. According to Basins of the storage pool were filled by tap water results of activity measurement of sludge samples it was from 1960 to 1995. Unfortunately, there was none monitoring concluded that the sludge could be treated as the LLW. of chemical properties or activities of the pool water during that period, i.e. there were no personnel within the Based on the previous experience, a technology [7] RA reactor department or in the Institute responsible for the was developed for sludge immobilisaticn and conditioning in pool regular inspection. Written regulations rules regarding a cement matrix, inside casks produced using the standard work, maintenance and monitoring within the RA spent fuel 200-litre metal barrels. Thickness of the concrete walls is storage pool area were not existed as well. between 7 cm and 8 cm. Entire inner side of the cylindrical concrete wall is covered by a plastic tube (wall thick 1 cm) Inspection of the storage pool in 1994, discovered that serves as a first barrier in preventing radionuclides thick deposits at basins walls, aluminium barrels, stainless leaching from radioactive sludge, immobilised in a cement steel containers and sludge at the bottom of the pool. The matrix. The bottom cask concrete wall is also 7-8 cm thick. situation was examined thoroughly in 1995-1996. Samples of In order to prevent or reduce radionuclide leaching, inner the pool water and sludge were taken out from different wall has been covered with thin layer of epoxy resin. Volume location and measured to determine chemical parameters and available for waste storage in these casks is 80 ± 5 L. activity. Analyses [3] have shown that the water is high corrosive to aluminium alloys (pH - 8.5, conductivity in About 60 - 65 L of sludge was poured at a time range 500 - 800 |iS/cm). Pool water samples showed also from the sedimentation vessel (Figure 1) into a previously specific gamma-ray activity of 80 - 90 kBq/L, originated prepared cask. As soon as a cask is filled up, it is

80 hermetically covered with a lid and transported to the The final stage of conditioning of the radioactive laboratory for sludge amdilioning where additional settling sludge is the covering of radioactive sludge, immobilised in a of sludge is allowed. Separated water is pumped inlo a plastic cement matrix, with the concrete cork. After concrete cork can and returned back lo ihc RA reactor spent fuel storage hardening, the cask is covered by a lid supplied with Screws pool. Through the second stage of the sludge settling, volume of the sludge in the cask has been reduced to about 40 L. 4. CALCULATION MODEL

Three-dimensional (3D) model of the sludge storage mclal cask, filled with matrix of conditioned sludgc-ccmcnt mixture, based on geometrical dimensions and known or assumed composition of materials, is developed. Vertical and horizontal cross-sections of 3D model of ihc sludge storage cask, used in calculation, is given in Figures 3

Slu

■ ■ ■ _

;oncr»t» 1 ■ x£t"l'"? "v,''i 1,:,; I F»ban«l | [ Ai..r. (::: :: U—' — 1 '• D«f'•! L__J

Fig. 3. Cress sections of the 3D model of the cask

"Standard metal 200 L barrel" has outer diameter 57.0 cm. total height 87.5 cm and wall thickness of 0.1 cm. Volume of the barrel is about 220 L. Material composition of Fig. 1. Settler for sludge in the spent fuel storage pool area Ihc barrel walls is not known and for the calculation purposes, a pure iron with theoretical density of 7.874 g/cm3, The existing pilot cement mixer was reconstructed i.e., with atom concentration of 8.491- l(T~ cm , is assumed. to enable simple placing a barrel containing the sludge on its • platform without a risk of spilling. Rooms for conditioning Inner sides and bottom of the barrel arc filled by the sludge in a cement matrix, supplied with independent concrete (thick 7.5 cm), which exact composition is not ventilation system, and for storing the casks during the period known. Density is measured as 2.35 g/cm3. "KENO code needed for cement hardening, have been arranged. Regular Concrete Standard Mix" with 2.30 g/cm' density and nuclide composition (9J is used in the model. Homogenous When the cask with settled sludge is placed 00 the distribution of die concrete material wilh uniform thickness is platform of the mixer, an amount of about 90 kg of ccmcnl assumed. Space above matrix in the cask is completely (PC-45 MPa) was added into the cask. Mechanical manipulator covered by concrete (thickness 13.0 cm in the mncr ring. (Figure 2) was used lo mix this mixture until a homogeneous above the matrix, and 10.0 cm above outer concrete ring). No substance ('matrix') was obtained [8]. This technology for air gaps arc assumed to exist within cask in Ihe model. sludge conditioning eliminates all the risks related to pouring the sludge into the concrete mixer and pouring the cement- Central axial /one of Ihc cask is separated from the sludge mixture into the metal barrel. The barrel with Ihc surrounding annular concrete ring by plastic cylindrical tube. homogenised mixture is removed from the mixer platform and Height of the tube is 67.0 cm, inner diameter is 40.0 cm. and placed in a separate room to harden mixture for about 2 days. wall thickness is 1.0 cm. Exact composition of the tube material used is not known. Pure PVC (C2H3CI), with density of 1.65 g/cm3 and nuclide composition given in [9), is chosen in the 3D model description of the cask.

Sludge composition is determined by radiochemical analysis of few sludge samples in the IAEA laboratory [5] and had showed that main component of Ihc sludge is Fe^Oj (average 83.60% by weight). The remaining components of the sludge arc (wcighl pcrccnts arc given): 5.19% A^Oj, 1.86% SiOz, 1.86% G2(>,. 1.11% MnO. 1.78% PbO and 4.60% CaO (including other 'minor impurities'). Exact density of 'dry sludge* is not known and depends of water Fig. 2. Pilot mixer with concrete container in the cask bounded in "dry sludge*. Measured density values were in 81 range from 0.64 g/cm3 to 1.04 g/cm3. Estimated maximum Table 3. Atom composition of materials used in 3D model density value for "100% dry sludge" is 1.96 g/cm3. Atom concentration [10^ cm'J Atom According to the sludge conditioning procedure, Air PVC Concrete Matrix dried (but not 100% dry!) sludge is homogeneously mixed H | 0.04771 0.01374 0.030971 with cement in average mass ratio 1:2.1. Procedure proposes C 0.03180 the best mass ratio as 1:1.8. About 90 kg of cement (in N 4.336 10° average) is added to about 40 L of sludge/water mush ("dried 1.019 10° | 0.04606 sludge1'), well mixed, and left to harden for few days. O 0.030600 Volume of this mixture, prepared inside central space of the Na 0.00175 storage cask, is about 85 L, based on an assumption that there Mg 0.000322 was no contraction of mass and volume of materials used for Al 0.00175 0.001046 mixing. Measured density of such conditioned sludge/cement Si 0.01662 1 0.002496 CI 0.01590 for one sample shaped as a cube (10 cm edge length) is 1.80 ± Ar 1.653 10'' 0.05 g/cm3. Density of one random selected 'dry sludge' Ca 0.00152 0.007318 sample (used in mixing) was 1.04 ± 0.05 g/cm3. Cr 0.000006 Exact composition of the cement used for Mn 10.000004 conditioning is not known either, except that it is Portland Fe | | 0.00035 0.000597 type cement. For the calculation purposes, the Portland Pb 1 || 0.000002 cement with density of 2.1 g/cm3 and composition given in [10] is chosen (weight percent are given): 2.0% MgO. 23.0% Table 4. Intensity of gamma-ray source (Iy) in the cask SiOa, 8.0% AIjQj, 4.0% FejO* and 63.0% CaO. Water (with density of 1.0 g/cm3) in the sludge is assumed to have the Gamma-ray 1Y [gamma-ray per second] same volume fraction as it was measured in one sample of the Source Rcf. [13] Ref. [5] Ref. [6] sludge (95.17%), while volume fraction of 100% dry sludge Nuclide is 4.83%. From these data, given above, density of (MeV] Avg sludge/cement mixture ('matrix') is calculated as 1.585 g/cm3 "'a 0.662 7.675 10" 6.273-107 4.144-10'" and atom composition of the sludge/cement matrix is 1.1732 5.944-If/ 4.089- 1(T 4.431 I05" *°Co determined for use in calculations (Table 2). 1.3325 5.95110* 4.095-10s 4.436-10* Total intensity: 7.794-101 6.355-107 4.233-10*" Table 2. Composition data for sludge/cement matrix 5. MCNP CODE RESULTS Component Density Mass Volume I g/cm 1 [kg] ty_ The well-known MCNP code [12] is used for 100% Dry Sludge 1.86 3.59 1.94 calculations with the MCPL1B continuous energy photon Water 1.00 39.01 39.01 dam library. The gamma-ray flux al two points in air near the Cement 2.10 90.00 43.24 cask surfaces (top and side) is calculated as well as a gamma- Total 1.59 132.62 84.19 ray spectrum in 13 energy groups up to 1.5 MeV (Figure 4). It is converted to the gamma-ray ambient dose equivalent In calculation model, it is assumed that sludge rates using ICRP-21 conversion factors [12]. The code is run storage cask is surrounded by an idealised large air sphere for 5 million gamma-ray histories to obtain satisfactory low (Figure 3 and 4). Air with density of 1.20 kg/m and nuclide statistical errors in group spectrum (< 10%) and ambient dose composition given in [111 is used. In such way, influence of equivalent rate (< 1%). Initial calculations are carried out for gamma-ray reflection from ground and surrounding objects at the sample cube (10 cm edge) of the sludge/cement mixture, gamma-ray spectrum and corresponding dose rate in the and values of calculated gamma-ray ambient dose equivalent calculating points is neglected. Atom concentrations of rales arc compared to the measured values at the same spots materials used in 3D model of the cask are given in Table 3. of the cube. Further calculations are carried out for the sludge storage cask. Gamma-ray source in the sludge storage cask is

assumed to originate only from Cs nuclide (Ey = 0.662 SldaollhcoMk McV, yield - 0.851) and "Co nuclide (E, = 1.1732 MeV, Top of Ih« CM! yield at 0.99857 and Ey = 1.3325 MeV, yield = 0.99983 Homogeneously distributed source of gamma-rays within volume of the sludge/cement matrix in the cask is modelled. Intensity of particular gamma-ray line in the sludge/cement mixture is determined according to the measured activity data obtained for the different sludge samples (Table 1) and given in Table 4. 0.76 1 00 1AO '■y mrm'Vt [M»V]

Fig. 4. Gamma-ray spectrum at cask surfaces 82 Table 5. Ambient dose equivalent gamma-ray rates [31 M.V.Matausck, Z.Vukadin, R Pavlovic, N. Marinkovic, "Current Activities on Improving Storage Source ADERv[nSv/h] Conditions of the Research Reactor RA Spent Fuel". Case Activity Top surface Side surface Transactions of the 1st Int. Topical Meeting on Ref. Calc. Meas. Calc. Meas. Research Reactor Fuel Management, pp. 115-119, Bruges, Belgium (1997) [5] 7.3 [4] M. V. MatauSck, Z. Vukadin, S. Pavlovic, T. Maksin, Z. Cube 113] Avg 83.4 8-10 Idjakovic. N. Marinkovic, "Detection of Fission [6] 4.9 Products Release in the Research Reactor RA Spent [5] 2.7 11.6 Fuel Storage Pool", Proceedings of the Annual Meeting Cask [131 Av* 36.7 50-60 140.7 80 on , pp. 415-418. Aachen, Germany [6] 1.9 7.7 (1997) [5] *** Appendix II to IAEA Mission Report to Vinda Calculation results, compared to measured ones, Institute in February 1997, IAEA. Vienna (April 1997) show wide dispersion due to large discrepancies in measured |6] M. P. Pe5ic;, "Measurement of Activity of Mud from activities and unknown exact material composition of the Spent Fuel Storage Pool", Proceedings of the IRPA sludge and sludge/cement matrix. Best agreement is achieved Regional Symposium on in for the 'cube* and activity determined in Ref. [5]. For better Neiglibouring Countries of Central Europe 1997 (ed.: E. agreement, more reliable composition of the '100% dry Sobol), pp. 439-442, Prague, Czech Republic sludge' - water mixture and sludge/cement matrix should be (September 8-12,1997) determined. [7] I.Plccas, S. Pavlovic '^Development of Concrete Composition in Radioactive Waste Management". Proceedings of o** International Conference 6. CONCLUSION Radioactive Waste Management and Environmental Remediation ICEM'97, pp. 565-566, Singapore Values of calculated gamma-ray ambient dose (October 12-16 1997) equivalent rates are compared to the measured values at the [8] R. Pavlovic S. Pavlovic L Plecas, "Safety Aspects of same spots of the cask with conditioned sludge from the RA the Cleaning and Conditioning of Radioactive Sludge spent fuel pool. Partially acceptable agreements were found, from Spent Fuel Storage Pool of "RA" Research in spite of wide spread of (input) data for the source activity Reactor in the Vinca Institute", BGNS Transactions, and relatively large uncertainties in compositions and Vol. 4, No. 1, pp. 92-95 (1999) fractions of materials used in the calculation carried out by [9] CD. Harmon, II. RJD. Busch, J.F. Briesmeister, R.A. applying developed 3D model of the waste storage cask. Foster, "Criticality Calculation with MCNP™ - A Primer", Appendix C, LA-12827-M. LANL, Los REFERENCES Alamos, NM, USA (1994) [10] CR. Tipton Jr. (ed.), "Reactor Handbook, Vol I - Materials"t II edition, revised and enlarged, pp. 1075. [1] M. PeSic, S. Cupac, Z. Vukadin, "Management of 1974, Interscience Publ. Inc., New York/London (1960) Ageing Research Reactors in the 'Vinca* Institute", [11] DJ. Whalen, D.E Hollowell, J.S. Hendricks. "MCNP: IAEA International Symposium on Research Reactor Photon Benchmark Problems", LA-121%, LANL, Los Utilisation, Safety and Management, Lisbon, Portugal Alamos, NM, USA (1991) (September. 6-10, 1999), paper IAEA-SM-360/042P, [12] J.F. Briesmeister (ed.). "MCNP™ - A General Monte CD ROM CSP-4/C, pp- 042P.1-042P.9, IAEA ISSN Carlo N-Particle Transport Code", Version 4B, LA- 1562-4153, Vienna, Austria (June 2000) 12625-M. LANL, Los Alamos. NM, USA (1997) [2] M. Milosevic, Z. Vukadin, "An Analysis of Spent Fuel [13] R. Pavlovic S. Pavlovic "Measurement of Dry Sludge Characteristics for Reactor RA Spent Fuel Elements", Activity", Internal Report, Vinca Institute, (December paper presented at The 3rd International Conference of Yugoslav Nuclear Society - YUNSC 2000, Belgrade. Yugoslavia, paper A4.7, Book of Abstracts, p.32 2000) ' (October 2-5, 2000). to be published

83