ML20049H448
| ML20049H448 | |
| Person / Time | |
|---|---|
| Issue date: | 02/12/1982 |
| From: | Silling S NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | Knapp M NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| References | |
| REF-WM-6 NUDOCS 8203030073 | |
| Download: ML20049H448 (18) | |
Text
[Aj)7) - b PDva 3406.3.3/SAS/82/02/11/0 ggA
/o SM WMHL: 3406.3.3 d
~b (9 7'S J FEB 121987 t
i y
,sy n
MEMORANDUM FOR: fialcolm R. Knapp j
- 29 y~-
y %y.j 7,082s 9
FE;;
L High-Level Waste Licensing N
Management Branch Division of Waste Management MS ]y{%
FR0ft:
Stewart A. Silling h
N
.N High-Level Waste Licensing N D[T cT~
3 fianagement Branch Division of Waste Management
SUBJECT:
PROGRESS ON IN-HOUSE CALCULATIONS ON EPA STANDARD As you directed on January 7,1982, I have been performing DVM calculations in connection with the EPA standard and its relation to 10 CFR 60.
The first phase of this work was to set up models similar to Sandia's point value checks and base case sample runs. This first phase is essentially complete.
The attached report describes the highlights of this work.
ORIGI:7AL 37ggp m7 Stewart A. Silling High-Level Waste Licensing Management Branch Division of Waste flanagement
Enclosure:
As stated Distribution:
4fMHL file JBMartin J0 Bunting WMHL r/f REBrowning PDR WM r/f MJBell Nf1SS r/f SASilling & r/f DIST:
HJMiller TICKET N0:
OFC l2_ _ _ _ M: _ _ _ _WMH L NAME :SASilling:lmc l2____:____________:____________:____________:____________:____________:____________:___________
DATE :"2/-Q /82 L
r~
0203030073 820212 PDR WASTE WM-6 PDR
e CALCULATIONS ON EPA STANDARD:
PROGRESS THROUGH FEBRUARY 8, 1982 Point Value Calculations 4
As checks on my updates to the DVM code and on the reproducibility of i
Sandia's work, my first step was to rerun Sandia's point value calculation [1].
This calculation was a simulation of three nuclides in EPA's reference site in a routine re~1 ease scenario.
l The site being simulated is shown in Figure 1.
The medium is granite i
underlying a sandstone aquifer.
EPA describes the repository and site in reference [2].
Various numerical parameters for the site are listed in Table 1.
t My approach was to take the work as far as possible using only data i
provided in Sandia's and EPA's documentation.
Although I succeeded in setting up a DVM model based on this documentation, some of my results differed greatly from Sandia's.
After calling Margaret Chu and Richard Pepping, I found that there were two nonstandard aspects of Sandia's work that they did not document:
(1) Source model.
Sandia's version of DVM, unlike NRC's version, i
has a " mixing cell" source model which attempts to account for dilution of the dissolved contaminant within the repository void space.
I l
reconciled my approach with Sandia's by rederiving EPA's source model, which is very similar to Sandia's, and updating NRC's code version to use j
this source model as a user option.
s 2
Table 1.
Point value calculation parameters.
Variable Value Units Granite porosity 1E-4 Granite conductivity 1E-9 cm/s Aquifer porosity 0.15 Aquifer conductivity 1E-4 cm/s Upward gradient in granite 0.1 Horizontal gradient in aquifer 0.01 Depth of repository below aquifer 200 m
Distance in aquifer to discharge point 1600 m
Aquifer thickness 30 m
Repository height 5
m Repository length (perpendicular to flow) 4000 m
Repository length (parallel to flow) 2000 m
Extraction ratio 0.1 Dispersivity 0
Granite retardation factor 1
Aquifer retardation factors C14 1
Tc99 1
Sn126 10 Solubility limits C14 infinite Tc99 1E-9 g/g Sn126 1E-6 g/g Initial inventoriss C14 2.8E4 ci Tc99 1.4E6 ci
3 Sn126 5.6E4 ce Canister life 500 yr Leach time 1E4 yr l
l l
l 1
4 (2) Extraction ratio.
Sandia and EPA agreed on a fudge factor to account for the fact that not all of the 2km x 4km repository area is mined. This factor reduces the repository volume and the effective vertical flow area through the repository by a factor of 10.
I reconciled my model with Sandia's by changing the volume and areas accordingly.
Following these changes to my model, I got good agreement with Sandia's work.
Table 2 compares the results.
The analytic solutions shown in Table 2 come from my rederivati :n of EPA's source and transport models.
Figures 2 through 5 show as a function of time the discharge rates, integrated discharge, EPA release fractions, and health effects for the three nuclides.
The health effects values come from EPA estimates of the number of cancer deaths that occur from release of each nuclide to a river [4].
Calculations with a Statistical Sample I generated a Latin Hypercube sample of 90 vectors using the same distributions and ranges that Sandia used (see Table 3).
For initial inventories I used the spent fuel ORIGEN database M. J. Wise and I developed last summer.
In addition, I used the database to help model competition between isotopes of the same chemical element in going into solution [3].
My DVM model included all important nuclides, including all four actinide chains.
I used the EPA exponential leach source model in order to permit compatability with Sandia's work.
5 Table 2.
POINT VALUE CALCULATIONS EF.T REFERENCE REPOSITORY ROUTINE RELEASE i
4 INTEGRATED DISCHARGE OVER 10 YEARS (curies) t t
C14 analytic 375
[
Sandia 364 l
t NRC 365 L
Tc99 analytic 3.44 Sandia 3.47 NRC 3.40
+
Sn126 analytic 46.3 Sandia 46.5 NRC 47.7 L
My run modeled only 1000 MTHM of waste in order to make comparison with the EPA standard easier.
Canister life was fixed at 1000 years for all vectors, since it makes little difference in this scenario.
I ran the sample on DVM and obtained CCDF shown in Figure 6.
The CCDF shows that 27% of the vectors exceed the EPA release limits.
This agrees with the 26% figure that Sandia reported.
Health effects are plotted in Figure 7.
Note that these health effects are for 1000 MTHM,
6 Preliminary Work on Effect of 10 CFR 60 Criteria I generated separate CCDF curves for the 35 vectors in the sample which happened to comply with 10 CFR 60 in travel time and leach rate.
These curves are shown overlaid on the original curves in Figures 8 and 9.
Note that 6% of the vectors satisfying 10 CFR 60 (2 of 35) exceed the EPA standard.
This result agrees with that of Sandia, which found that 1 vector which satisfied 10 CFR 60 failed the EPA limit.
(I found 2).
However, Sandia failed to count how many vectors of the total sample were disqualified by 10 CFR 60.
(I found that 55 of the 90 vectors were so disqualified).
Consideration of the reduction in sample size by imposing 10 CFR 60 changes the statistical picture and makes 10 CFR 60 look slightly less effective.
Effect of 4n Actinide Chain One of Mike Bell's questions about Sandia's work was whether or not the 4n actinide chain, which Sandia did not model, makes a difference.
I tried running my LHC sample with and without this chain. I found that for a given release limit, consideration of this chain makes at most a 4%
difference in the probability of exceeding that limit.
Code Enhancements I have written many modifications to DVM which permit the work described above.
These modifications are contained in 21 pages of Fortran updates.
The modifications are made and the code is recompiled each time a run is made.
i
. + -
7 Table 3.
Ranges and distributions for for routine release LHS sample.
Range Distribution Variable Units 1
.01-500 Lognormal Kd for Am and Cm ml/g 2
.01-100 Lognormal Kd for Pu ml/g 3
.01-100 Lognormal Kd for U ml/g 4
.01-500 Lognormal Kd for Th ml/g 5
.01-50 Lognormal Kd for Np ml/g 6
.01-500 Lognormal Kd for Pa ml/g 7
.01-500 Lognormal Kd for Ac ml/g 8
.01-500 Lognormal Kd for Tc ml/g 9
.01-50 Lognormal Kd for Sn ml/g 10 1E-13 to 0.1 Lognormal Sol. limit for Pu g/g 11 2E-9 to 2E-3 Lognormal Sol. limit for U g/g 12 1.3E-9 to SE-6 Lognormal Sol. limit for Th g/g 13 4E-24 to 4E-6 Lognormal Sol. limit for Np g/g 14 1.6E-7 to 6E-4 Lognormal Sol. limit for Pa g/g 15 1E-8 to IE-4 Lognormal Sol. limit for Tc g/g 16 1E-8 to 1E-4 Lognormal Sol. Ifmit for Sn g/g 17 50-500 Uniform Dispersivity ft 18 1E3 to IE7 Loguniform Leach time yr 19 1.3E-3 to 50 Lognormal Conductivity in aquifer ft/ day 20 0.05 to 0.30 Normal Porosity in aquifer 21 1E-8 to 8E-4 Lognormal Conductivity in granite ft/ day i
22 1E-5 to 1E-2 Lognormal Porosity in granite
8 Non graphics changes accomplish the following:
Read LHS vectors from sample.
j Compute travel time from edge of repository.
Optionally use EPA source model.
Optionally use NRC square pulse source model.
Initialize inventories from ORIGEN database.
Set mass fractions for solubility calculation from database.
Compute EPA release fractions and health effects from f
integrated discharge.
Compute integrated discharge as a function of time.
Allow up to 120 nuclides to be modeled in a run (old limit was 10).
Sort vectors according to increasing EPA release fractions and health effects.
Graphics-oriented changes include the following:
Plot discharge rate, integrated discharge, EPA release fraction, health effects, and source rate as a function of time.
Plot CC0F curves for EPA release fraction and health effects for a sample.
9 References 1.
M. S. Y. Chu and R. E. Pepping, " Preliminary Comments on the EPA Draft Standard on HLW Isolation," Sandia National Laboratories Internal Report (1981).
2.
C. B. Smith, D.J. Egan, Jr., W. A. Williams, J. M. Gruhlke, C. Y.
Hung, and H. L. Serini, " Population Risks from Disposal of High-Level Radioactive Wastes in Geologic Repositories,"
Environmental Protection Agency draft report EPA 520/3-80-006 (1981).
3.
M. J. Wise and S. A. Silling, " Waste Composition Computer Databases," NRC letter report (August 14, 1981).
4.
J. M. Smith, T. W. Fowler, and A. S. Goldin, " Environmental Pathway Models for Estimating Population Health Effects from Disposal of High-Level Radioactive Waste in Geologic Repositories,"
Environmental Protection Agency draft report EPA 520/5-80-002, page 148 (1981).
i
I l
I discharge l
point i
(river)
I
/\\
200m surface deposits antrifor 9
/
r 7
y r,
30m 200m repository v
s 2000 m s
1600 m s
granite b
I' i
F.igure is EPA granite reference repository site, i.
4 I
l l
i l
I I
1 l
}
L l
f
r.
e NWFT/DVM ISOTOPE 02/09/82 o-C 14 ROUTINE RELERSE o - TC 99 EPR SOURCE MODEL
^ - SN126 o
~E 23
'o _
M f'
~
B C
o x
mm S'o r
~E o
H C
e u?
ca o a
=
,a a
e -E CC I
o en o~'o
-E:
?o 0.0 2.C
- 4. 0 6.0 8.0 10.O TIME (YR) x10' Figure 2.
Discharge rates for point value problem.
.m,*wm
+6e
m o
NWFT/DVM ISOTOPE 02/09/82 o-C 14 ROUTINE RELERSE o - TC 99 i'
EPR SOURCE MODEL
^ - SN126 A
"g I
E
_~
]
mo
_/
i M
O L
w
[.a I GE~o r
r." E c
i (D
)
e--*
O Bo
!J r
g -* {
O E
G tu H
Zr
-O_
i "E
(5
~
N'a i
i i
0.0 2.0 4.0 6.0 8.0 10.O l
TIME (YR) x10' l.
v Figure 3.
Integrated discharge for point value problem.
l I
l
i t
ISOTOPE NWPT/DVM o-C 14 02/09/82 o - TC 99 ROUTINE RELEASE
^ - SN126 EPR SOURCE MODEL
+ - ALL NUCLIDES t
~o-
-E:
i o
~s y
e o
l H.-. _
/
i
- o
/
c>
-E 11 CD 5
a e"o
/
~=
e a-c L1
?
o_
,s 4
~
'o O.0 2.C 4.O G.0 8.0 10.0
(
TIME (YR) x10 Figure 4.
Integrated discharge measured as ratio to EPA release limit for point value problem.
3 I
ISOTOPE NWFT/DVM a-C 14 02/11/82
- TC 99 ROUTINE RELERSE a - SN126 EPR SOURCE MODEL
+ - ALL NUCLIDES
.t
."o ?
l
-a
~5
/
i cn H
o
/
uO w o,
/
L_. ~ -
u I
-s 1
J cc 'o or~=
4 i
cc e
ct E
u
'o f
7 I
o.
i i
j J
~
0.0 2.0 4.0 6.0 10.0 l
TIM.
x10'
'l i i;
Figure 5.
Health effects-for 1000 MTHM in point value -problem, i
6
t I
(
NWFT/DVM 02/07/82 ROUTINE RELEASE RLL VECTORS i
RLL NUCLIDES i
a J-t t
m t
d
>c w d-
_a In_a C
1 w.
o, g
3 4
1-
\\
Nc-
}
i o
.a~,"
i i l iTT} '
i i iiIsig i i e ilist i i i i 11:13 i a ia e ilissi 10 10
- 10*
10*
10 10' EPR' RELEASE RATI0' l.
Figure 6.
CCDF for EPA release ratio.
l I
NWFT/DVM 02/07/62 ROUTINE RELERSE ALL VECTORS RLL NUCLIDES 9_
O
>9 u a-hs A
E D
_s-1.
1 5
o-s a
l %
0.
i i i a n til ' i i i i siTit i i a i iT1 i
i i i iTil i i i iiTTI 10' 10 10' 10 10 10'
~
(
EPR HERLTH EFFECTS l
Figure 7.
CCDF for health effects per 1000 MTHM.
T t
NWFT/DVM 02/07/82 VECTOR ROUTINE RELERSE a - RLL VECTORS RLL NUCLIGES o - VECTORS WITHIN 10CFR60 o
.a I
w o e s-
-a
.g-a 4
8.k y g--
T
\\
-L 0-c
,' s
/
o L3 i i, i i i i i j, o.
,,,,,,g i i a i s iiij '
3 I 3 3Illll i i i e 61ll 10 10 10' 10 10 30' q
EPR'RELERSE RATIO s
Figure 8.
CCDF for EPA release ratio for vectors satisfying 10CFR60 compared with original CCDF for entire sample.
M
y
=
S_
NWFT/DVM 02/07/82 VECTOR
/
ROUTINE RELERSE a - RLL VECTORS RLL NUCLIDES o - VECTORS WITHIN 10CFR60 o
/..
i e
O r{.
s o
~ s-hs t
g?
)
l s
1
,h, u-l t
N 5
9 L
5 o-s C
/
1 s
a i i i liilli i i i l iilli iI i i 1116) i i i sinii ia a i n till '
1 10" 10 10*
10' 10' 10
.:EPR HERLTH EFFECTS i
1 Figure 9.
CCDF for health effects per 1000 MTHM for vectors satisfying 10CFR60 compared with original CCDF for entire' sample.
ll w
m.
_ _ _ _. _ _ _ _ _ - - - - - -. -