ML20133H373
| ML20133H373 | |
| Person / Time | |
|---|---|
| Issue date: | 10/09/1985 |
| From: | Costanzi F NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Higginbotham L NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| Shared Package | |
| ML20133H376 | List: |
| References | |
| CON-FIN-A-3171 NUDOCS 8510170114 | |
| Download: ML20133H373 (9) | |
Text
- . - - - - .. . . - . - - - .--- --- . -_ __-_-_ -- - - - _ -
)
i j OCT e Iges
- l
! MEMORANDUM FOR: Leo'B. Higginbotham, Chief r
!' Low-Level Wastes and Uranium Recovery Projects Branch Division of Waste Management, NMSS 3
! FROM: Frank A. Costanzi, Chief i Waste Management Branch l Division of Radiation Programs '
and Earth Sciences, RES
SUBJECT:
ARCHIVAL TRENCHES FOR TESTING THE PERFORMANCE OF WASTE PACKAGES AT SPECIFIC LLW DISPOSAL SITES '
1 I have asked my staff to examine the materials you sent on the subject of
! archival trenches at LLW burial sites. In your cover letter, you posed three j separate questions. Specifically: "Is this type of project more suitable as a ,
i Research or Technical Assistance contract? Would data from archival trenches i benefit the NRC waste disposal program? If so, is the approach outlined by ;
j Brookhaven National Laboratory the correct protocol to answer the questions we
- are most interested in?"
l In the program plan you provided, the stated purposes of the archival trench project are first to verify the acceptability of current laboratory tests and $
to provide a site-specific data base for potential revision of these tests, and !
l second to assess long-term site performance in preparation for the application for site decommissioning. With respect to the first purpose, it is the role of l
i RES to assess the fundamental processes affecting waste disposal on a generic l j basis. Because the BNL archival trench program is site-specific, it is ;
' unlikely to provide such a fundamental, generic understanding of underlying processes. With respect to the second stated purpose, it is clearly not within the purview of RES to gather data necessary for a licensing decision. ;
Based upon the arguments above, the archival trench testing program at BNL is i most appropriately a Technical Assistance contract insofar as actual data 1
collection is concerned. On the other hand, the design of such a program, ;
including the sampling statistics and generic emplacement, retrieval, and l
- testing protocols, is an appropriate RES activity. This ties in to part of the ,
first purpose stated above, e.g., developing a new source of data so that i results of field tests can be intercompared to current laboratory testing.
Therefore, if an archival trench program were to be approved RES would support i I
4
\ I
- w -
I ,#
3pq x -
OCT 0 ses
.p.
the development of such a program on a generic basis, but not its implementation on a site-specific basis.
Your second question concerns the benefits of performing an archival trench program, which is likely to be ext *emely expensive. In general, an archivel trench is an appealing concept because many of the uncertainties associated with the durctions and conditions of laboratory testing, extrapelation of la! oratory results to an actual site, and variability between sites can be advessed. A carefully designed and executed program could provide assurance to the NRC and to the public that unknown phenomena are not adversely affecting the safe disposal of LLW. However, such an endeavor needs clearly defined
" success criteria" by which a determination of success or failure can be made; if a failure, some regulatory thought must be given to remedial action. In addition, note that an archival trench is only one of several approachts the NRC may implement in order to assess whether a given LLW site is performing as predicted, and thus meets the standards for site decomnissioning. I reconmend that we systematically explore all the options available to us, which will include the archival trench as one possibility.
Your final question asks whether the approach suggested by BNL is appropriate.
On the positive side, it is significant that BNL is considering the experiences of the National Bureau of Standards and Department of Energy field testing programs. This could enable cost-effective program design and implementation.
However, BNL has apparently adopted NBS procedures wholesale; the ar;ount of analysis given to the specific needs of an archival trench program is inadequate. bNL's lack of critical analysis can be diviced into two categories, e.g., similarities between archival samples and actual wastes, and sampling procedures and statistics. As currently envisioned by BNL, the results from an archival trench will be qualitative rather than quantitative, ar.d will be a better negative indicator than positive indicator. The main benefit of such an archival trench program may be a qualitative "early warning" system for unacceptable performance in waste package raterials at that particular site.
With respect to similarities between archival samples and actual wastes, BNL suggests non-standard burial techniques to accelerate testing. How can thc results of such accelerated testing be interpreted, either to check the results of accelerated lab testing, or as an indicator of the perforraance of the actual waste packages? BNL would test small coupons of materials which were specially made for the archival trench. If a metal shows (or doesn't show) stress corrosion cracking, what exactly does that say about waste packages of a different size and shape that were adnufactured differently? Further, what assurance is there that an external radiation source adjacent to simulated wastes, as proposed by BNL, simulates actual wastes? In sum, any archival 0FC: WMB:dm :WMB:SL :WMB:BC : : : :
NAME:CHac arth :RGrill :FCostanzi : : : :
...................... 390..............................................................
DATE: 10/1/T6 : /s /P k : / / : : : :
OCT 9m trench program must first establish the correspondence of test samples and conditions to the actual . waste packages and conditions, or else the results are qualitative at best, an'd meaningless at worst. If actual waste packages were exhumed for this program, then the similarity would be obvious. If simulated samples are used, the ways in which the samples are representative and the ways in which they are not must be precisely known.
With regard to sampling, BNL has adopted the number of replicate samples and
' the retrieval schedules from the NB5 program without consideration of sampling statistics, confidence levels, and the change in risk levels over time. If 100% of the actual waste packages were exhumed and examined, then NRC would have 100% confidence that waste package performance is well. characterized.
But, outside of this null case, how many replicates should be examined to give what level of confidence? Furthermore, over time, the probability of waste package failure goes up as the hazard from failure goes down. What is the optirun retrieval schedule? Both of these questions need a statistical evaluation.
In sum, I would like to meet with you at your earliest convenience to discuss an overall examination of assurances that a LLW site is functioning properly prior to site decommissioning. An archival trench project is one possible approach to collecting the data that NRC will need for this licensing decision.
Frank A. Costanzi, Chief Waste Management Branch Division of Radiation Programs and Earth Sciences RES e ._ .
cc: R. Browning . . . , ,
L. + tt m "2. '8 .2.93. _
DISTRIBUTION: . . . _ . . . . . . q Subj. FCostanzi : "c ta .
Nw as hTWt :;o.
l Rdg. KGoller ,
...._.......4 Cir. Dross fli' h . ........I Chron. RMinogue G'M : - . I CHackbarth Isu M .. ). . . . . . .
.h ii r, :a . '
RGrill tu. .
w.m 3?l)PL.. .. .
..e . m r r - M ~
t s c.: . m, 6, OFC: WMB:dm : : :
WMB:$ g : y :BC :
kkEECha rth RGrilf l4 o I12 1 ! : : :
............................... err....................................................
DATE: /0/f /36 :Ic/t /b : yj /[ : : : :
/ '%, UNITED STATES E' NUCLEAR REGULATORY COMMISSION
{ f ,
waswiwotou, p. c, roess e
i
\...../ oct .
1 1 MEMORANDUM FOR: Leo'B. Higginbotham, Chief
, Low-level Wastes and Uranium Recovery Projects Branch Division of Waste Management, NMSS FROM: Frank A. Costanzi, Chief Waste Management Branch F Division of Radiation Programs and Earth Sciences, RES ,
SUBJECT:
ARCHIVAL TRENCHES FOR TESTING THE PERFORMANCE OF WASTE :
1 PACKAGES AT SPECIFIC LLW DISPOSAL SITES
~
I have asked my staff to examine the materials you sent on the subject of archival trenches at LLW burial sites. In your cover letter, you posed three
)
separate questions. Specifically: "Is this type of project more suitable as a Research or Technical Assistance contract? Would data from archival trenches benefit the ARC waste disposal program? If so, is the approach outitned by
! Ercokhaven National Laboratory the correct protocol.to answer the questions we l are most interested in?"
In the program plan you provided, the stated purposes of the archival trench project are first to verify the acceptability of current laberatory tests and to provide a site-specific data base for potential revision of these tests, and second to assess long-term site performance in preparation for the application for site decommissioning. With respect to the first purpose, it is the role of RES to assess the furdamental processes affecting waste disposal on a generic basis. Because the BNL archival trench program 15 site-specific, it is
- unlikely to provide such a fundamental, generic understanding of underlying processes. With respect to the second stated purpose, it is clearly not within t
the purview of RES to gather data necessary for a licensing decision.
Based upon the arguments above, the archival trench testing program at BNL is
- most appropriately a Technical Assistance contract insofar as actual data
- collection is concerned. On the other hand, the design of such a pro l including the sampling statistics and generic emplacement, retrieval, gram, and testing protocols, is an appropriate RES activity. This ties in to part of the first purpose stated above, e.g., developing a new source of data 50 that results of field tests can be intercompared to current laboratory testing.
Therefore, if an archival trench progrem were to be approved. RES would support 4
k
2 the development of such a program on a generic basis, but not its implementation on a site-specific basis.
Your second question concerns the benefits of perfonning an archival trench program, which is likely to be extremely expensive, in general, an archival trench is an appealing concept because many of the uncertainties associated with the durations and conditions of laboratory testing, extrapolation of laboratory results to an actual site, and variability between sites can be addressed. A carefully designed and executed program could provide assurance to the NRC and to the public that unknown phenomena are not adversely af fecting the safe dispesal of LLW. However, such an endeavor needs clearly defined
" success criteria" by which a determination of success or failure can be made; if a failure, some regulatory thought must be given to remedial action. in addition, note that an archival trench is only one of several approaches the NRC may implement in order to assess whether a given LLW site is performing as predicted, and thus meets the standards for site decomissioning. I recommend that we systematically explore all the options available to us, which will include the archival trench as one possibility.
Your final question asks whether the approacn suggested by Bhl is appropriate.
On the positive side, it is significant that BNL is considering the experiences of the National Bureau of Standards and Department of Energy field testing programs. This could enable cost-effective progran design and implementation.
However, BNL has apparently adopted NBS procedures wholesale; the amount of analysis given to the specific needs of an archival trench program is inadequate. Bht's lack of critical analysis can be divided into two categories, e.g., similarities between archival samples and actual wastes, and sampling procedures and statistics. As currently envisioned by Bhl, the results from an archival trench will be qualitative rather than quantitative, and will be a better negative indicator than positive indicator. The main benefit of such an archival trench program may be a qualitative "carly warning" system for uracceptable performance in waste package materials at that particular site.
With respect to similarities between archival samples and actual wastes, BNL suggests non-standard burial techniques to accelerate testing. How can the results of such accelerated testing be interpreted, either to check the results of accelerated lab testing, or as an indicator of the perfonrance of the actual waste packages? BNL would test small coupons of materials which were specially made for the archival trench. If a metal shows (or doesn't show) stress corrosion cracking, what exactly does that say about waste packages of a different size and shape that were manufactured differently? further, what assurance is there that an external radiation source adjacent to simulated wastes, as proposed by BNL, simulates actual wastes? In sum, any archtval
t 1
l -
i 3 1 trench program must first establish the correspondence of test samples and 4
conditions to the actual, waste packages and conditions, or else the results are i i qualitative at best, and meaningless at worst. If actual waste packages were exhumed for this program, then the similarity would be obvious. If simulated l samples are used, the ways in which the samples are representative and the ways ,
- in which they are not must be precisely known. !
~
i With regard to sampling, BNL has adopted the number of replicate samples and
! the retrieval schedules from the NBS program without consideration of sampling
, statistics, confidence levels, and the change in risk levels over time. If
! 100% of the actual waste packages were exhumed and examined, then NRC would ;
j have 1001 confidence that waste package performance is well-characterized.
)
But, outside of this null case, how many replicates should be examined to give
- what level of confidence? Furthermore, over time, the probability of waste i package failure goes up as the hazard from failure goes down. What is the j optimum retrieval schedule? Both of these questions need a statistical evaluation. -
In sum, I would like to neet with you at your earliest convenience to discuss :
an overall examination of assurances that a LLW site is functioning properly 3
- prior to site decommissioning. Anarchiv4.1trenchprojectfisonepossible
- approachtocollectingthedatathatNjC i' lljit, ed his licensing decision. i
(- N
~
<1 l
1 ,\ 6 o ,V G:
Frank A. Costanzi, ie '3q ,
Waste Management Bran l Division of Radiation Program and Earth Sciences, RES ,
i i i
i cc: R. Browning b,
l i
t l
l l
I i i
~
i r
J
4 i
OCT sm l l
l i
l t
i trench program must first establish the correspondence of test samples and f conditions to the actual waste packages and conditions, or else the results are i qualitative at best, and' meaningless at worst. If actual waste packages were 1 exhumed for this program, then the similarity would.be obvious. If simulated samples are used, the ways in which the samples are representative and the ways l i in which they are not must be precisely known. !
i With regard to sampling, BNL his adopted the number of replicate samoles and ,
- the retrieval schedules from the.NBS program without consideration of sampling i
- statistics, confidence levels, and the change in risk levels over time. If l i 100% of the actual waste packages were exhumed and examined, then NRC would !
have 100% confidence that waste package performance is well. characterized. L But, outside of this null case, how many replicates should be examined to give )
what level of confidence? Furthermore, over time, the probability of waste i package failure goes up as the hazard from failure goes down. What is the i optimum retrieval schedule? Both of these questions need a statistical ,
evaluation, j In sum, I would like to meet with you at your earliest convenience to discuss !
- an overall examination of assurances that a LLW site is functioning properly i prior to site decommissioning. An archival trench project is one possible ,
4
! approach to collecting the data that NRC will need for this licensing decision. i
(
1 f
Frank A. Costanzi, Chief r i
Waste Management Branch (
Division of Radiation Programs :
and Earth Sciences, RES l:
l i cc: R. Browning nca n I i s 3 t .m u. E.-3]tL_ !
DISTRIBUTION: .. .. .._. . j l
i Subj. FCostanzi :. c :; . _ . . _ .
] Rdg. KGoller :'. w.-: n R; rm.t : o . ..._.. . .. . l. l Cir. Dross N h. _...........t l I I
! Chron. RMinogue ... El' .
CHackbarth ::a r,.e :. . a . ___...4 !
I l RGrill : ; . s.n.n ta . !
DR.x.. . ! 1
[
]
..'.e. a r x:. d e i t s .~ L.: . '.' %,8 y I
. . , . , . . ~ . . . . . . . . . . . _ . . - . . . . . . . j I !
0FC: WMB:dm :WMB:S g :1N : : : : !
........ p .......... p ..... f.:BC .................................................... t NAME:CHactbdrth :RGrill : J. o anzi : : : : !
.................... ........... m .................................................... .
i DATE: /0/f /36 :Ic/v /W : gf /[ : : : :
{
l.
4 1
OCT 9 25 i
) MEMORANDUM FOR: Leo'd.Higginbotham, Chief i Low-Level Wastes and Uranium Recovery
- Projects Branch Division of Waste Management, NMSS
] .
FROM: Frank A. Costanzi, Chief
. Waste Management Branch Division of Radiation Programs and Earth Sciences, RES
SUBJECT:
ARCHIVAL TRENCHES FOR TESTING THE PERFORMANCE OF WASTE PACKAGES AT SPECIFIC LLW DISPOSAL SITES J
4 I have asked my staff to examine the materials you sent on the subject of archival trenches at LLW burial sites. In your cover letter, you posed three separate questions. Specifically: "Is'this type of project more suitable as a 4
Research or Technical Assistance contract? Would data from archival trenches j benefit the NRC waste disposal program? If 50, is the approach outlined by 4 Brookhaven National Laboratory the correct protocol to answer the questions we
] are most interested in?"
4
] In the progran plan you provided, the stated purposes of the archival trench j
project are first to verify the acceptability of current laboratory tests and 1
to provide a site-specific data base for potential revision of these tests, and I second to assess long-tern site performance in preparation for the application i for site decommissioning. With respect to the first purpose, it is the role of i
RES to assess the fundamental processes affecting waste disposal on a generic I . basis. Because the BNL archival trench program is site-specific, it is i unlikely to provide such a fundamental, generic understanding of underlying
- processes. With respect to the second stated purpose, it is clearly not within
! the purview of RES to gather data necessary for a licensing decision, t
! Based upon the arguments above, the archival trench testing progran at BNL is j nost appropriately a Technical Assistance contract insofar as actual data-collection is concerned. On the other hand, the design of such a program, i including the sampling statistics and generic emplacement, retrieval, and
! testing protocols, is an appropriate RES activity. This ties in to part of the first purpose stated above, e.g., developing a new source of data 50 that i
results of field tests can be intercompared to current laboratory testing.
Therefore, if an archival trench program were to be approved. RES would support i
OCT 9 as
.p.
the development of such a program on a generic basis, but not its implementation on a site, specific basis.
Your second questien concerns the benefits of performing an archival trench program, which is likely to be extremely expensive. In general, an archival trench is an appealing concept because many of the uncertainties associated w,ith the durations and conditions of laboratory testing, extrapolation of laboratory results to an actual site, and variability between sites can be addressed. A carefully designed and executed program could provide assurance to the NRC and to the public that unknown phenomena are not adversely affecting
- the safe disposal of LLW. However, such an endeavor needs clearly defined i " success criteria" by which a determination of success or failure can be made; if a failure, some regulatory thought must be given to remedial action. In addition, note that an archival trench is only one of several approaches the NRC may implement in order to assess whether a given LLW site is performing as predicted, and thus meets the standards for site decommissioning. I recommend that we systematically explore all the options available to us, which will include the archival trench as one possibility.
- Your final question asks whether the approach suggested by BNL is appropriate.
On the positive side, it is significant that BNL is considering the experiences of the National Bureau of Standards and Department of Energy field testing programs. This could enable cost-effective program design and implementation.
However, Bhl has apparently adopted NBS procedures wholesale; the amount of analysis given to the specific needs of an archival trench program is
! inadequate. BNL's lack of critical analysis can be divided into two 1 categories, e.g., similarities between archival samples and actual wastes, and t sampling procedures and statistics. As currently envisioned by BNL, the 4
results fron an archival trench will be qualitative rather than quantitative,
- ar.d will be a better negative indicator than positive indicator. The main benefit of such an archival trench program may be a qualitative "early warning" system for unacceptable performance in waste package materials at that particular site.
j With respect to similarities between archival samples and actual wastes, BNL suggests non-standard burial techniques to accelerate testing. How can the results of such accelerated testing be' interpreted, either to check the results of accelerated lab testing, or as an indicator of the performance of the actual waste packages? BNL would test small coupons of materials which were specially
~
made for the archival trench. If a metal shows (or doesn't show) stress ;
corrosion cracking, what exactly does that say about waste packages of a '
, different size and shape that were manufactured differently? Further, what i assurance is there that an external radiation source adjacent to simulated wastes, as proposed by BNL, simulates actual wastes? In sum, any archival OFC: WMB:dm :WMB:SL :WMB:BC : : : :
NAME:CHac arth :RGrill :FCostanzi : : : :
........................@0.
DATE: 10/1/f 5 : /s/r er : / / : : : : I l
.-- . __. - - - . , _ . . - - _ - _ - _ _ _ - -