ML19256E719
| ML19256E719 | |
| Person / Time | |
|---|---|
| Site: | Perkins, South Texas  |
| Issue date: | 10/29/1979 |
| From: | Martin D NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | Lowenberg H NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| References | |
| NUDOCS 7911150059 | |
| Download: ML19256E719 (8) | |
Text
...
8[km bCg30,,
UNITED STATES g
NUCLEAR REGULATORY COMMISSION 3
E WASHINGTON, D. C. 20555 o
o g
OCT 2
- 373 MEMORANDUM FOR:
Homer Lowenberg, Assistant Director for Operations and Technology Division of Fuel Cycle and Material Safety THRU:
Hubert J. Miller, Section Uranium Recovery Licensing Branch FROM:
Daniel E. Martin Uranium Recovery Licensing Branch Division *f Waste Management
SUBJECT:
REVIEW 0F MAGNO ESTIMATES OF RADON RELEASES PER AFR FROM URANIUM MILLING In response to your memorandum dated August 22, 1979, the Magno radon release estimates incorporated by affidavit in the Perkins record have been reviewed in light of new information.
Principal source documents reviewed in this effort include the Draft GEIS on Uranium liilling (NUREG-0511), a draft report by NUS Corporation entitled " Interim Report, Source Tems for Radon Releases from Uranium Mining and Milling" (dated March 12, 1979), and a PNL research report, "A Field and Modelt' g Study of Windblown Particles from a Uranium Mill Tailings Pile" (NUREG/CR-0629). Other available sources have been reviewed as required, and where hard data were lacking, staff judgement has been exercised.
As you may be aware, the 11agno estimates were based on a set of assumed values for the various parameters which determine the radon releases per AFR from uranium milling.
Changes to any of these parameters result in revisions to the associated radon release rates.
Therefore, the procedure adopted for this review was to identify the important parameters, review the available infomation to detemine appropriate parameter values, recalculate radon releases per AFR using this improved data base, and then compare the revised estimates with those obtained previously by Magno.
Three complete data bases have been tabulated and reviewed.
These data bases have been extracted from the Magno affidavit, the GEIS, and the NUS study. The Magno and NU(efforts were specifically geared to the detemination of radon releases per AFR, while the GEIS had the purpose of conservatively estimating and analyzing the environmental impacts of uranium milling as a function of the effectiveness and cost of various kinds and levels of effluent and regulatory controls.
Unfortunately, and inappropriately, certain characteristics of the GEIS base-case (no controls) model mill were carelessly incorporated into the NUS analysis.
i336 352 O A 7911150
Homer Lowenberg 2-The parameter values extracted from the three data bases reviewed, and the parameter values selected as being most appropriate, are provided in attached Table 1.
The variable identifier letters assigned to the parameters listed in Table 1 are used in attached Table 2 which provides the equations by which radon releases from tailings impoundments per AFR are estimated. The Table 2 equations illustrate the mathematical relationships involved and can be used to determine the relative impacts of specific parameter value changes.
Radon releases from sources other than the tailings impoundment are estimated using other data and equations in addition to those presented in Tables 1 and 2.
Attached Table 3 provides a comparison of the original Magno estimates with those considered to be "most appropriate" as a result of this review.
As analyzed here, Magno's total Ci/AFR estimate is within about 10% of what would be estimated on phg bgis of present information. Also, Magno's expressed estimate of Ci/yr-AFR for ongoing persistent releases fully encompasses the revised estimate, though the margin of conservation is somewhat reduced.
Further details of the basis for the revised estimates are provided by enclosure to this memorandum.
Aff Daniel E. Martin Uranium Recovery Licensing Branch Division of Waste Management
Enclosures:
As stated cc: D. Swanson, ELD, w/ encl B. Bordenick, ELD, w/enci W. Thompson, NMSS/FCMS, w/ encl R. Wilde, NMSS/FCMS, w/ encl R. Gotchy, NRR/DSE, w/ encl F. Congel, NRR/DSE, w/enci s
1336 353
Table 1 COMPARIS0ft 0F DATA BASES FOR MAGNO, GEIS, AND NUS ESTIMATES OF RAD 0tl RELEASES PER AFR FOR OPERATIONAL, DRYING, AND RECLAIMED TAILINGS PILES Variable Values Utilized Variable Magno Draft NUS Draft Values identifier Variable Definition and Units Af fidavi t GEIS S-3 Report Selected _
a A
Ore processing rate, MT/d 1500 1800 1800 1800 a
i B
Days /yr. operational 341 310 350 310 t
C Average ore grade, % U 0 0.10 0.16 0.10 0.15 D
Operating 11 fetime, yrs.8 26 16 16 20 3
E Required drying time, yrs.
5 5
5 5
F MT U,0 / AFR (in yellowcake) 245 196 185 185 g
G Uranfum Recovery, %
90 93 93 92 0
5b 5b ll MT tails /AFR 2.72x10
- 1. 317x10 1.987x10 1.355xl I
Total tailings area, ha 56.63 80 80 50 20 20 17 J
pond area, ha
-- d K
wet bealh area, ha 42.88 10 10 8
L dry beach area, ha 13.75 50 50 25 M
Tailings depth, m 11.58 (ayg) 8 (max)6 8 (max)7C 12 (avg)
N Ultimate tailings inventory, MT
.33x10/
8.93x10 1.0lx10 1.12x10 Specificradonexhalationrate, pCi/m -sec per pCi/g Ra-226 in tails:
0 0
0 0
for water-covered tailings
-2d,e P
for wet beach areas 2.908g10 0
0.3 0.3 Q
for dry beach areas 1.068 1.0 0.6 1.0 R
Post-reclamation radon flux, pCi/m -sec 2.5 2.0 2.0 7.0 a) Estimated from Magno values for variables D and N.
b) Derived from values given for variables C, F, and G.
c) Derived from values given for A, B, and D.
d) Value given is for combined pond and wet beach areas.
e) Value given is derived from Magno calculations.
U Os u
LTl 4
1
Table 2 EQUATIONS USED TO ESTIMATE RADON RELEASES FROM URANIUM TAILINGS PILES I.
Operational Releases:*
Ci/AFR = F[KP + LQ] (8.837 x 10 )
6 ABG II.
Releases During Pile Drying:
+ ( +L 0]
6 Ci/AFR =
(4.418 x 10 )
ABDG III.
Post-Reclamation Releases:
FIR 3
Ci/yr-AFR =
(3.156 x 10 )
ABCDG
- Variable identifier letters as assigned in the far left-hand column of Table 1 are used in the equations provided here.
These equations are demonstrative of the relationships of the various input parameter; with the estimated radon releases (e.g., operational releases do not depend on the anticipated mill lifetime).
s -
1336 355 1,
Table 3 COMPARISON OF MAGNO ESTIMATES WITH REVISED ESTIMATES RESULTING FROM REVIEW 0F AVAILABLE DATA Calculated Raden Releases Magno Estimates Revised Radon Source Cateaory Original Recalculated Estimates I.
Ore storage, handling, and milling, Ci/AFR....
30 30.5 55.8 II.
From tailings impoundment A.
During operation, Ci/AFR...
750 749 873 B.
During pile drying, Ci/AFR...
350 346 308 C.
Post-Reclamati on, Ci/yr-AFR...
1 to 10 0.92 0.38 III.
From dispersed particulate 4 to 0*
Ra-226, Ci /yr-AFR...
IV. All uranium milling sources A.
Total Ci/AFR 1130 1130 1240 B.
Total Ci/yr-AFR (ongoing) 1 to 10 0.92 4.4 to 0.4
- Estimated to be about 2 Ci/yr-AFR at time of reclamation and to gradually decrease the reafter.
Expressed as 4 to 0 Ci/yr-AFR to allow for uncertainty and indicate gradual decrease.
s -
.e w
Enclosure
SUMMARY
OF TECHNICAL BASIS FOR REVISED ESTIMATES OF RAD 0N RELEASES PER AFR FROM URANIUM MILLING I.
Ore Storage, Handling and Milling The previous Magno estimate was based on a one-time loss of 40% of the radon inventory in one AFR's worth of cre; all emanated radon was assumed to be released and the umanation fraction was conservatively assumed to be 0.4 to implicitly account for ore milling releases not specifically addressed.
The revised estimate accounts for a three-time loss of all emanated radon in one AFR's worth of ore. The emanation fraction is assumed to be 0.2, based on the average of U.S. Bureau of Mines measurements of over 1000 uranium ore samples.
Emanated radon is assumed to be released when the ore is dumped to the ore pile, when the ore is dumped to the grizzly, and when the ore is crushed and ground. Surface radon flux from the ore storage pile is also includet b sed on an assumed average pile height gf 6 meters, an ore density of 2 MT/m, and a specific radon flux of 1 pCi/m'-sec per pCi/g of Ra-226.
II.
Tailings Impoundment Radon Releases The revised estimates are based on the selection of values of parameters listed in Table 1 as being most appropriate.
These are discussed individually below, in alphabetical order.
A.
As stated in the NUS study an 1800 MT/ day capacity is representative of tne industry average after the anticipated retirement of three large and old mills in New Mexico.
B.
An average operating factor of 85% is chosen as representative based on industry-average figures of 83% and 87% for 1975 and 1976, as stated in the GEIS, page 3-14.
This is equivalent to operation at full capacity for 310 days per year.
C.
An average ore grade of 0.15% U 038 is selected as representative of current industry average conditions given that the average ore grade is decreasing gradually and that it was 0.16% in 1977 (GEIS, p. 3-4).
D.
This selection is somewhat arbitrary but represents a compromise between the Magno and GEIS figures.
E.
A drying time requirement of 5 years is thought to be conservative; present data availability does not allow for a more precise estimate.
F.
The NUS estimated uranium requirement is selected as most appropriate as it is based on the current enrichment tails assay (0.20% U-235) and a marginally conservative reactor plant factor of 75%, slightly above current levels.
G.
Uranium recovery is estimated at 92%, based on the industry average for 1977 (GEIS, p. 3-4).
1336 357
~ ~.
. H.
This parameter value is calculated from values of variables C, F, and G.
I.
The total tailings impoundment area of 50 he'ctares is sized to hold 20 years of tailings, ct the stated production rates, with an ultimate average depth of 12 meters. The GEIS estimates average depth of existing tailings to be 12 to 13 meters (p. S-2).
J, K, L.
In the absence of hard data it has been estimated, based on staff judgement and experience, that about 50% of tailings areas existing now are dry beach and about 33% are under standing water.
M.
An ultimate average depth of 12 meters is selected based on the GEIS estimate of 12 to 13 meters as the average depth of existing piles (p.S-2).
N.
This value is calculated from values selected for variables A, B, and D.
O.
A zero value is selected as reasonable in view of the fact that radon releases through water covered tailings are insignificant.
P.
The NUS value is selected as it is the most conservative and is not unreasonable.
Q.
The value selected is chosen because available data from many sources indicate that it is reasonable but slightly conservative. The NUS value is believed to be the most realistic but there is insufficient data available at this time to support its use.
R.
The value selected is that formally proposed as a regulation (44 FR 50012).
III.
Radon Released from Dispersed Radium During the operational and drying periods Ra-226 may be dispersed from the tailings impoundment in particulate form by wind dusting. After reclamation and site decommissioning, this radium will contiribe to emit radon and constitutes an on-going source not specifically addressed in the Magno affidavit.
On the basis of Ra-226 release rates presented in the GEIS for the base-case model mill, and assuming a conservative emanation fraction over the short term of 0.5, the immediate post-reclamation release rate is estimated to be about 2 Ci/yr-AFR.
This rata would decrease with time to a value approaching zero as the dispersed radium is gradually mixed deeper into the ground.
This value could be reduced by the assumption of anti-dusting actions, as required by the staff of all licensees, or increased by the assumption of greater dusting rates.
1336 358
3-NUS Corporation has estimated a value of about 10 Ci/yr-AFR (rounded up from 8.5) based on the field measurements and analyses presented in a PNL research report (NUREG/CR-0629).
Several considerations provide solid argument for a lower release.
These include the lack of any dust mitigation procedures or actions at the site where the field work took place, and the fact that the site considered is otherwise atypical with respect to age, tailings area, and environmental factors.
Review of the subject PNL report indicates that almost one half of the source strength would be removed by site decommissioning cleanup in accordance with the staff's current branch position or proposed EPA regulations.
Almost the entire remaining half of the source strength is calculated to arise from measured radium concentrations only about 1 pCi/g above an assumed background concentration of 1 pCi/g.
Thus, if the actual background concentration is 2 pCi/g the estimated source strength is too high by a factor of two.
Given that this particular source of radan is quite difficult to precisely quantify, the calculated release rate of 2 Ci/yr-AFR is expressed as 4 to 0 Ci/AFR to allow for uncertainty and to indicate the gradual decrease with time.
~.
1336 359
_m<
a r
New 9'