Presented at the COMSOL Conference 2010 Boston The Dssoluton and Transport of adonucldes from Used Nuclear Fuel n an Underground epostory Y A N N I C K B E A U E G A D 1 M A K G O B I E N 2 F A N K G A I S T O 2 C O M S O L C O N F E E N C E, B O S T O N, M A O C T O B E 7 TH 2 0 1 0 1 Unversty of Western Ontaro, London, ON 2 Nuclear Waste Management Organzaton, Toronto, ON
Concept for the long term management of Canadan spent nuclear fuel 2 Used fuel bundles are placed n durable contaners Contaners are emplaced wthn vaults excavated n a stable geologcal formaton Contaners are surrounded by self-sealng clay materal Garsto, F., T. Kempe and P. Gerszewsk. Techncal Summary of the Safety Aspects of the Deep Geologcal epostory Concept for Used Nuclear Fuel. Nuclear Waste Management Organzaton, report: NWMO T-2009-12, (2009)
COMSOL Model Goal: Develop a model to calculate the release of radonucldes from a defectve contaner and ther subsequent transport through the vault Key aspects: Accurate representaton of the vault geometry Vertcal contaner emplacement Pn-hole defect n the contaner Tme dependent radonuclde source term (functon of the dose rate and spent fuel dssoluton) 3 Non-adsorbed (I-129), moderately adsorbed (Ca-41), strongly adsorbed (Cs-135) radonucldes
Model Geometry Contaner Empty regon representng the contaner walls Pnhole r = 8.25 x 10-4 m Flux measurement boundary Buffer compacted bentonte Backfll bentonte, clay, grante Inner EDZ Outer EDZ ock 4
Governng Equatons 5 S c t b d c t SDoc L P SC θ s porosty κ d sorpton coeffcent ρ b bulk densty τ tortuosty D o free water dffusvty L, P lqud and sold reacton terms (radoactve decay) S c Source term
adonuclde release 6 Instant release fracton adonucldes present at the fuel claddng gap and n the gran boundares eleased mmedately upon contact wth groundwater Intal concentraton of radonucldes n the contaner Congruent release ~95% of radonucldes are present wthn the fuel grans elease s dependent on the dssoluton of the fuel matrx Shoesmth, D.W., evew Fuel corroson processes under waste dsposal condtons, Journal of Nuclear Materals, 282, 1-31(2000)
Congruent release - Fuel Dssoluton 7 uo2 dss A A A fuel fuel fuel A fuel G G G D D D α UChem ( t) ( t) ( t) dss Alpha Beta A fuel - Fuel surface area G α, β, γ Emprcal fuel dssoluton rate constant D α, β, γ (t) Tme dependent dose rates Gamma Garsto, F., D.H. Barber, E. Chen, A. Inglot and C.A. Morrson. Alpha, Beta and Gamma Dose ates n Water n Contact wth Used CANDU Fuel. Nuclear Waste Management Organzaton, report: NWMO T-2009-27, (2009)
Congruent elease (cont d) ( t) I S UO2 c ( t) V (1 ( t) I f o, UO2 contaner r I ) I o, UO2 UO2 ( ln(2) / t f I- Instant release fracton I UO2(t) Inventory of radonuclde at tme t I 0,UO2 Intal nventory of UO 2 I 0,UO2 Intal nventory of radonuclde m u Mass of uranum n the contaner e 8 ( t) 1 / 2 uo2 t ) ( t) m u
Boundary Condtons 9 Boundary condton No Flux Contnuous C,1 =C,2 Boundary name Inner contaner walls Outer contaner walls Hole walls All nternal boundares Constant concentraton Outer boundares
Geometry: Smulatons Contaner and pnhole Contaner, pnhole and buffer Complete vault adonuclde source Constant concentraton n the contaner Constant fractonal dssoluton rate (1 x 10-7 a -1 ) Dose dependent dssoluton rate 10 Compared COMSOL results to analytcal calculatons Used COMSOL vault model to verfy SYVAC-CC4
esults Dose dependent source term 11 Analytcal COMSOL Analytcal COMSOL Geometry: contaner, pnhole, buffer Peak contaner release rate occurs ~10 5 a Overall strong agreement
esults eleases to the geosphere 12 I-129 no sorpton Ca-41 moderate sorpton Cs-135 hgh sorpton Contaner flux Vault Flux Contaner flux Vault Flux Contaner flux Vault Flux Geometry: Complete vault Cs-135 source term hghest due to hghest nventory and hgher IF than Ca-41 Sorpton causes a tme delay n peak flux to the geosphere and a reducton n ts magntude compared to the source (κ d I = 0, κ d Ca > κ d Cs )
Verfcaton of SYVAC-CC4 Near-Feld Model Engneered barrer system represented by a seres of concentrc cylnders 13 Developed for horzontal n-room contaner emplacement The vault porton of the COMSOL model was used to calbrate SYVAC-CC4 for vertcal contaner emplacement
esults Verfcaton of SYVAC-CC4 Near Feld Model 14 I-129 no sorpton Ca-41 moderate sorpton Cs-135 hgh sorpton Source SYVAC-CC4 COMSOL Source SYVAC-CC4 COMSOL Source SYVAC-CC4 COMSOL The buffer, backfll and EDZ layer thckness were selected so that the agreement between COMSOL and SYVAC-CC4 s strong for low and non-sorbng elements (I-129), whch are the hghest dose contrbutors Preferental pathway for lower sorbng elements s up through the buffer and nto the tunnel. A large buffer thckness s requred n SYVAC-CC4 Preferental pathway for hgher sorbng elements s thourgh the sdes of the borehole and nto the rock due to the hgher transport resstance n the buffer. Therefore SYVAC-CC4 underpredcts Ca-41 and Cs-135 releases from the vault Dfferences n peak fluxes of approxmately 3%, 40% and 60% for I-129, Ca-41 and Cs-135 respectvely
Model Conclusons and future work Developed a COMSOL model to account for a dose dependent radonuclde source term, radonuclde release from a pnhole defect n a vertcally emplaced contaner and transport through the buffer, backfll and EDZ Model was bult n a seres of ncreasngly complex steps Vault porton of the model used to calbrate SYVAC-CC4 Future work can nclude examnng expandng pn-hole sze, multple defectve contaners, advectve flow and geosphere transport 15
Acknowledgements Nuclear Waste Management Organzaton 16 MITACS Accelerate Dave Shoesmth Group, Unversty of Western Ontaro
Assumptons Water enters the contaner after the buffer saturates wth water, whch corresponds to a model tme of zero (fuel age = 130 a.) The groundwater s reducng and neutral Transport s dffuson domnated All materals are fully saturated Steel canster nsert and fuel claddng are not consdered transport barrers 17
Intal Condtons Instant release fracton 18 C 0, f I I V UO2 ( vod t f ) V vod Internal vod volume f I-129 I=0.04 f Cs-135 I=0.04 f Ca-41 I=0 All other subdomans, ntal radonuclde concentraton s zero
esults constant concentraton n contaner 19 I-129 flux [mol/a] No buffer Wth buffer COMSOL 1.05 x 10-6 1.95 x 10-7 Analytcal 1.13 x 10-6 1.85 x 10-7 Case wthout buffer: COMSOL flux s lower due to local concentraton depresson at the entrance of the pnhole A smulaton wthout the contaner yelded a flux that s exactly as predcted analytcally Case wth buffer: COMSOL flux s hgher, possbly due to the fact that the analytcal soluton s applcable to a sem nfnte geometry whereas COMSOL uses a C=0 boundary condton, whch would result n larger concentraton gradents and hgher fluxes
esults Constant fuel dssoluton rate 20 Intal concentratons and fnal fluxes are the same Dfferences n ntal flux and fnal concentraton due to dfferences n resstance COMSOL soluton s senstve to solver tolerance
Dfferences n fluxes calculated at the buffer-hole boundary and flux-measurement boundary 21 ed: I-129 flux calculated at the buffer-hole boundary Blue: I-129 flux calculated at the 1 mm below the buffer hole boundary (flux measurement boundary) Green: I-129 flux calculated at the outer buffer boundary
Analytcal Soluton 22 elease ate buffer pnhole 4 r 1 buffer 1 2 r D L s o C D o cont pnhole
adoactve Decay 23 L P ln(2) s t 1/ 2 c ln(2) b d t 1/ 2 c t 1/2 radonuclde half-lfe t I-129 1/2 : 1.57 x 10 7 a t Ca-41 1/2 : 1.02 x 10 5 a t Cs-135 1/2 : 2.30 x 10 6 a