ML20132D340

Forwards Description of Expert Sys for Evaluation of Retired Nuclear Matl Licenses

ML20132D340
Person / Time
Issue date:	12/16/1996
From:	Paul Goldberg NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:	Bellamy R, Cline B, Grobe J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I), NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II), NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
Shared Package
ML20132D318	List: IA-96-464, Forwards Description of Expert Sys for Evaluation of Retired Nuclear Matl Licenses ML20132D313 ML20132D332 ML20132D336 ML20132D340
References
FOIA-96-464 NUDOCS 9612190328
Download: ML20132D340 (6)

Text

_ . . . - . ~ . _ - _ _ _ . . . _- . - .

(l .'

i Note For: Ron Bellamy, RI Bill Cline, RII Jack Grobe, RIII Chuck Cain, RIV l Bob Pate, RV i

l From: Paul Goldberg, IMNS /l<'

Subject:

Oak Ridge System for Evaluating Terminated Licenses l I understand from George Pangburn that a number of people had questions at the recent Counterpart Meeting about ORNL's methodology for identifying licenses of concern. The evaluation is based on license files; in some' cases, the inspection files are with them, in some cases, they are not. This means, of course, that the information available to ORNL is limited and that the additional step of Regional and/or NMSS review is necessary to determine whether there is contamination at a site. We instructed ORNL to be i conservative in identifying sites, i.e. to flag all those which might have the l potential for serious contamination. The score assigned to a license does not indicate any absolute risk value; it establishes an estimate of the relative likelihood and magnitude of contamination among the licenses and permits l

ranking and comparison of the licenses.

As the attached description explains in more detail, ORNL uses an expert system to evaluate the likelihood and possible magnitude of contamination.

The system first looks to see if the license can be eliminated for administrative reasons, e.g. it was superseded by another license or it was transferred.to an Agreement State. There is provision to override this determination if the site had the potential for serious contamination. The system next assigns a score based on the nuclides and quantities the licensee was authorized to possess. This score considers the form of the material, and .

inhalation and ingestion dose factors and decrements for half-life. The score I is modified by information in the file on licensed activity, closecut and material dispostion information, and indications of possible contamination of buildings or the environment (releases, burials, incinceration, operational ,

incidents,etc.).

l The procedure for sealed sources not accounted for by documentation in the files, also described in the attachment, is simpler: the system calculates a score based on the isotopes and quantities authorized, decremented for half-life, cc: George Pangburn Fred Combs 9612190328 961216 PDR FOIA SAVAGE 96-464 PDR

,. i L' .l l ,

EXPERT SYSTEM FOR EVALUATION OF RETIRED l l NUCLEAR MATERIALS LICENSES I

Purpose and Function Written for l Office of Nuclear Materials Safety and Safeguards  !

Nuclear Regulatory Commission by Measurement Applications and Development Group Health and Safety Research Division, Oak Ridge National Laboratory j l

Purpose of the system i An expert system is broadly defined as any computer system which makes structured decisions l based on information provided to it. The expert system for nuclear materials license evaluation i makes a twofold decision aboout retired nuclear materials licenses:  ;

• decides whether a site or sites of use authorized under the license could presently be  ;

meaningfully contaminated, based on activities carried out UNDER THAT LICENSE, and not known to be directly authorized at that site by a subsequent license. See section I on computing the site score for a license.

• identifies sealed sources authorized under the license which were not accounted for at license  !

retirement, and which could represent a present public hazard. See section II on computing a  :

sealed sources score for a license. .

Method utilized in the evaluation i For a given physical license file, the initial entry into the system is an inventory record of relevant t l information present in the file. This inventory record is completed for the approximately 21,000 license files in the boxes in ORNL's possession. At the time the license number is te be evaluated, l , the expert system evaluator first looks at what types ofinformation are available in each physical license file available for that license number. About 25 per cent oflicense numbers have more than one physical file in different locations (boxes). About half of these situations are essentially duplicate files. For the other half, the files contain different information. For the expert system evaluation, the file containing the disposition and closeout information is used in the evaluation.

I. Computing the site score or rank of the license The following information pertains to lice: ses authorized to possess loose material, or materials handled loose. The information listed below is used in the expert system evaluation, in approximate chronological order of use. The system can loop back for additional information where needed. In each step, the goal of the system is to disqualify (assign a score of 0) to the license based on the information available at that point. For licenses not ultimately disqualinea trom consideration, the next goal is to compute a site score for the license, based on the information in the file. The score can be thought of as an approximate ' hazard ranking' for the license, and is based on both material hazard and other characteristics of the activity and site. The final license / site-specific score has meaning as a basis for comparison of relative hazard represented by

'he license, but has no independent numeric meaning.

l (1) DISQUALIFYING CONDITIONS l

• License superceded by a subsequent license used at the same site (s).

• License transferred to an agreement state (unless sites were closed out during the period of the license, and the state license does not cover those sites).

i '

i *No auth:riz: tion to poIsess material (distribution or transportation, for example), or no I actual possession of material under the license.

j' ' License to import or export only

• Fresh fuellicense

• License for a power reactor i

! For licenses which cannot be disqualified on the basis of such administrative information, i subsequent information on materials, licensee activity and closeout (surveys, inspections, i disposition) are used to compute the site score for the license.

! (2) MATERIALS l *For each material authorized in loose form, the amount of the material and the inhalation i and ingestion dose factors for that material are used to compute an initial ' score' for that i material 2. The estimated marimum amount of enaterial which could be remaining is first

reduced using the halflife, using the period of time from retirement of the license to 1990.

l The dose factors are then used to compute a hae.mti for the material / amount combination.

! For licenses with multiple materials, the individual scores are summed to get an initial

! estimate of the hazard from the loose materials possessed under that license. For licenses j with extremely small possession limits, or only low-hazard materials, the license may be j disqualified at this point, because of a low cumulative score. Licenses which authorized only i noble gases are not included for further consideration.

i l For broad byproduct materials licenses (those authorizing any byproduct material, or any i byproduct material with atomic no.1-83), the authorized use of the material determines the .

i hazard factors which are used.

1

(3) ACTIVITY j *The next questions which modify the score concern the activities carried out under the j license, and the user's estimate of the degree to which contamination could have been generated by the licensee's activites. This includes the manner in which materials were j handled, the frequency of turnover of the licensed amount, generation and handling of

waste, possible burials during the period of the license, number of sites, and moving of sites j

durmg the period of the license, and airborne or effluent releases. Accidents which could have resulted in serious contamination are also taken into account (e.g. explosions). l

i (4) CLOSEOUT AND DISPOSITION INFORMATION l 3

Any closeout survey for a license having a score above 20 is given to a person knowledgeable

. in health physics survey techniques, who then evaluates the survey information to ,

i determine to what degree the survey technique and results reduce the likelihood that the i site is contaminated, and to what degree it can be ascertained that present standards for unrestricted release were met. hnal inspections, disposition of licensed materials, generation and disposal of waste, degree ofintermediate and final decontamination are all used to modify the final site score received by the license. Following is a comparison of the degree to which the various factors can change the magnitude of the score. Both the questions asked at any point, and the change in score for a given answer depend in part on the answers to previous questions. Thus,it is difficult to give an exhaustive list of the exact

' For source materials and SNM, use of the dose factors would have resulted in enormous scores.

Nonlinear functions based on the dose factors are used for each material. Enrichment is taken into account in these functions.

i qurstions and values. The numbers below do give a range of the magnitude of the change in score from the important factors. The order given is an approximate order of magnitude of importance, and does not represent the order in which the system arrives at the questions.

Closcout Survey - The potential for change in score is highest for the closeout survey. When a closeout survey is present, andjudged to very thorough, with no significant findings, the score is reduced by a factor of 10 (e.g. a score of 250 would become a score of 25). Survey quality isjudged on the basis ofcoverage of the potentially contaminated areas, instrumentation and methods of analysis, reporting of results, and appropriateness of the i measurements. In particular, for licenses where alpha contamination is likely, a careful check is necessary to be certain that the survey included a thorough alpha survey, and that ,

the limits used matched the material possessed. '

For less thorough surveys, or surveys with some significant positive Sudings not ultimately resolved, the survey may ultimate result in a reduction to 30% of th previous value, up to .

90% of the previous score for a survey oflittle value. '

Questions about the closeout survey are at the end of the evaluation. If the score is below 20, no increase is score is given because oflack of a closeout survey. For a license with a score above 20 at that point in the evaluation, the score is increased by a factor of 1.8 in there is no closeout survey. For example, a license with a score of 20 is reassigned a score )

of 36 if no survey was done. On the other hand, a license with a score of 200 (fairly high) would be given a score of 360 if no survey was done Burial or dumping onsite under the license - Depending on the possible magnitude of the burial, the score can be increased by a factor of 2 (200%).

Degne of turnover of the licensed possession limit - The possession limit is an at any time' ,

limit. The actual amount of material procured and used depends on how often the licensee turns over the material. If very frequent turnover, the score can be increased by a factor of 1.4. If there were conclusive evidence that the licensee possessed material far below the possession limit, and did not turn the material over, the score can be reduced to 20% of the previous score.

I Degne to which building could have been contaminated, based on activities and usage - 50

% to 150 % of the previous score Degne to which outdoor contamination could have occurnd fmm routine operations or i nicases (separate fmm burial or dumping) - 85 % to 150% of the previous score.

i II. - Computing the sealed sources score For licenses with sealed sources, the procedure is somewhat different. For sealed sources, the initial questions deal with whether the disposition for the sealed sources at the time of license retirement was adequate. If so, then no further questions are asked. If not, then the materials and amounts are entered, and the expert system computes a hazard score for the sealed sources in the same way as previously described. For very low hazard materials and amounts, the system will then eliminate the license. For sealed sources, the system does not ask for further information, but assigns the hazard score based on the materials, amounts, and upper limit estimates of the remaining amount.

i t

Decision Structure for Evaluation  ;

l l

of a License Can license be given a score m User can override

! of zero for administrative Psystem decision reasons? \

v V Types and Forms of Materials l V

If all short-lived materials, license automatically receives both a sealed sources score and site score of zero.

V V Sealed material Loose material i

V V Amount of material Amount of Material v +

Sum of Materials Scores /5 Sum of Materials Scores Y +

Are sources accounted for? INITIAL SCORE:

If yes, sealed score to zero. IF BELOW 5, is there disposition l If not, score is assigned. or closeout information?

Is it adequate? If not, go to SUPPORTING information.

i l

V If score <.2- SITE SCORE TO O If score <5-ARE DISP./ CLOSEOUT SUFFICIENT TO MAKE SITE CONTAMINATION UNLIKELY?

If score >=5 GO TO:

SUPPORTING INFORMATION

_- y SUPPORTING QUESTIONS l -- --

MODIFYING FACTORS I EXISTENCE OF AN IDENTIFIABLE SITE ESTIMATE OF THE DEGREE OF CONTAMINATION THE ACTIVITIES UNDER THE LICENSE MAY HAVE PRODUCED WAS THERE DOCUMENTED DECONTAMINATION AT CLOSEOUT WAS THERE AN NRC VERIFICATION LETTER (RELEASE OF SITE) 1 WAS THERE EVIDENCE OF INCINERATION OR OTHER RELEASE ESTIMATE OF FREQUENCY OF TURNOVER OF THE MATERIALS j USE OF GLOVE BOXES, HOODS, OR HOT CELL GENERATION OF OTHER CONTAMINATED MATERIAL (PARTS, CLOTHS)

DID THE OPERATION GENERATE RESIDUES (SLAG, SLUDGE)

HOW WELL DOCUMENTED WAS DISPOSITION OF LICENSED MATLS WAS THERE ANY INDICATION OF DUMPING, BURIAL OR ABANDONMENT?

EXISTENCE AND QUALITY OF CLOSEOUT SURVEY (S)

EXISTENCE AND QUALITY OF PHYSICAL CLOSEOUT INSPECTION