ML20116B052

Acceptance Criteria for the LOW Enriched Uranium Reform Amendments

ML20116B052
Person / Time
Issue date:	04/30/1985
From:	Emeigh C, Gundersen G, Withee C NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:
References
NUREG-1065, NUREG-1065-R01, NUREG-1065-R1, NUDOCS 8504240693
Download: ML20116B052 (53)

Text

.,

i-NUREG-1065 Revision 1 Acceptance Criteria for the Low, Enriched Uranium Reform Amendments 1

l U.S. Nuclear Regulatory

! Commission

- Offico of Nuclear Material Safety and Safeguards C. W.~ Emeigh, G. E. Gundersen, C. J. Withee:

p " "'%<

p l

{-

l l

l l

1 l

eg 42 g g 850430 106 5 R PDR

e x

1 2G M- _.0;;

g,

.c

,g, c.

y'

_n n

b G

' NOTICE S,

c 7

' Availability of Reference Materials Cited in NRC Publications i

4 W

[Most documents cited in N RC publications _will be available from one of the following sources:

61kThe NRC Public Document Room,1717 H Street, N.W.

y,

1. -M iWashington, DC 20555 l2J:The N RC/GPO Sales Program, U.S.. Nuclear' Regulatory Comrmssion, 1

1 Washington, DC 20555 '

l 3ciThe National Technical Information Service, Springfield, VA 22161 q

~

y Although$e listing that follows represents the majority of docurr.ents cited in NRC publications,.

~

/

t it is not intended to be exhaustive.

1Ref'erenced documents available for. inspection and copying for a fee from the NRC Public Docu-

~

~-ment Room include NRC correspondence and internal NRC memoranda; NRC Office of Inspection and Enforcement bulletins,: circulars, information notices, inspection and investigation notices;.-

Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and t licensee documents and correspondence.

The following' documents in the NUREG ' series are available for purch$se from the N RC/GPO Sales Program:-formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and -

a

' NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code o/c

? Federal Regulations, and Nuclear Regulatory Comminion issuances

~ Documents available from the Nationa! Technical information Service' include NUREG series

reports and technical reports prepared by other federal _ agencies and reports prepared by the Atomic'-

Energy Commission, forerunner agency'to the Nuclear Regulatory Commission.

iDocuments available from public and special technical libraries l'ncl0de all open' literature items, such as books, joumal and periodical articles, and transactions. Federal Register notices, federal and l

(state legislation, and congressional reports can usually be obtained from these libraries.

Documents such as theses, dissertations, foreign reports and translations, and non-NRC conference

, proceedings are available for purchase from the organization sponsoring the publication cited.

1 Single copies of NRC draft reports are available free, to the extent of supply,' upon written request--

to the Division'of Technical Information and Document Control, U.S. Nuclear Regulatory Com-mission, Washington, DC 20555.

iCopies of industry codes and standards used in a substantive manner in the NRC regulatory process.

are maintained at the NRC Library, 7920. Norfolk Avenue, Bethesda, Maryland, and are available 1there for reference use by the public.= Codes and standards are usually copyrighted and may be '

purchased from the originating organization or, if they are American National Standards, from the -

^

Amer,ican N'ational Standards Institute,1430 Broadway, New York, NY 10018.'

x..,

r GPO Printed copy price: $4.50

n

'I

NUREG-1065 Revision 1 Acceptance Criteria for the Low Enriched Uranium Reform Amendments M:nuscript Completed: Febrnry 1985 Data Published: April 1985 C. W. Emeigh, G. E. Gundersen, C. J. Withee Division of Safeguards Office of Nuclear Material Safety and Safeguards U.S. Nuclear Regulatory Commission Wrhington, D.C. 20666

,$* **%e,,

(g) l l

l i

r ABSTRACT Revisions have been made to the material control and accounting require-ments for NRC licensees authorized to possess and use more than one effective kilogram of special nuclear material of low strategic significance to have MC&A systems able to (1) confirm the presence of special nuclear material, (2) resolve indications of missing material, and (3) aid in the investigation and recovery of missing material. This document presents criteria that can be used to aid in judging the acceptability of licensee plans that would be sub-mitted to the NRC for implementing these capabilities. General performance objectives, system capabilities, and recordkeeping are addressed.

iii

PREFACE This document provides criteria for acceptance of licensee fuel cycle facility material control and accounting (MC&A) plans for possession and use of greater than one effective kilogram of SNM of low strategic significance.

MC&A plans would be submitted by licensees in compliance with 10 CFR Part 74,

" Nuclear Material Control and Accounting for Special Nuclear Material of Low Strategic Significance." The acceptance criteria are to be used to enable NRC license reviewers to assess the licensees' plans.

It is intended that the criteria allow licensees the flexibility to achieve the required performance using equipment and procedures of their own choice, rather than prescribe a fixed, standard method for all licensees.

The MC&A rule for special nuclear material of low strategic significance is relatively brief and performance oriented. The acceptance criteria are intended to help the licensees to understand what performance is required to meet the requirements of the rule as well as help the NRC staff in reviewing license applications.

Changes in the acceptance criteria may be appropriate as additional insights and information are obtained. Constructive thoughts that will lead to improvements in the acceptance criteria should be addressed to the Chief, Fuel Facility Safeguards Licensing Branch, Division of Safeguards, U.S. Nuclear Regulatory Commission, Washington, D.C.

20555.

(

(

/

v

TABLE OF CONTENTS Pan ABSTRACT.................................................................

iii PREFACE..................................................................

v ACKNOWLEDGMENT...........................................................

ix INTRODUCTION.............................................................

x1 1.0 CONTENT OF THE FNMC P LAN AND ANNEX..................................

1-1 2.0 ACCEPTANCE CRITERIA FOR CONFIRMING THE PRESENCE OF SNM, RESOLVING INDICATIONS OF MISSING MATERIAL AND AIDING THE INVESTIGATION AND RECOVERY OF MISSING MATERIAL.................

2-1 3.0 ACCEPTANCE CRITERIA FOR MANAGEMENT STRUCTURE.......................

3-1 4.0 ACCEPTANCE CRITERIA FOR CONTROLLING MC&A ERRORS....................

4-1 5.0 ACCEPTANCE CRITERIA FOR PHYSICAL INVENTORIES.......................

5-1 6.0 ACCEPTANCE CRITERIA FOR ITEM CONTROL...............................

6-1 7.0 ACCEPTANCE CRITERIA FOR RESOLVING SHIPPER / RECEIVER DIFFERENCES...... 7-1 8.0 ACCEPTANCE CRITERIA FOR PERIODIC ASSESSMENT OF THE MC&A SYSTEM..... 8-1 9.0 ACCEPTANCE CRITERIA FOR RECORDKEEPING.............................. 9-1 REFERENCES...............................................................

R-1 vii

ACKNOWLEDGMENT The NRC wishes to acknowledge the valuable technical contributions to this

_ document from the Pacific Northwest Laboratory, Richland, Washington, under contract to NRC.

Principal contributors from PNL were L. C. Davenport,

. R. J. Brouns, and F. P. Roberts.

In addition, the contributions of D. R. Joy to Revision 1 of this document are acknowledged.

ix

INTRODUCTION The proposed reform of the Material Control and Accounting (MC&A) regula-tions for licensees authorized to possess and use low enriched uranium neces-sitates the development of objective criteria for evaluating the acceptability of MC&A plans that licensees are required to submit. The revised Fundamental Nuclear Material Control (FNMC) Pla.s would demonstrate how the requirements of the LEU Reform Amendments would be satisfied.

(As appropriate, examples of how certain commitments are to implemented would be included in an Annex which would not be incorporated as a condition of the license.) After acceptance by the NRC of a Plan as a condition of the license, the NRC will judge the adequacy of MC&A performance of a licensee by inspecting for compliance with commitments made in the Plan.

The recommendations presented in this report are subject to revisions and modifications as additional insights and information are developed.

Each section in this report except the first is divided according to the following format:

Rule Intent Affirmations: the licensee must make the stated affirmations for the Plan to be acceptable Information to be included: the Plan and Annex should contain the stated information to demonstrate adequacy of the licensee's MC&A system Acceptance criteria: the ground rules used by NRC license reviewers to determine if the Plan is acceptable; and Questions and answers:

these help to resolve uncertainties about the acceptability of specific approaches by explaining the intent of the regulation and including some examples of acceptable ways to comply with the regulation.

xi

1.0 CONTENT OF THE FNMC PLAN AND ANNEX The licensee's submittal to the NRC will consist of twc documents:

1) a Fundamental Nuclear Material Control (FNMC) Plan, referred to herein as the Plan, that will contain the licensee's affirmations and will demonstrate how the licensee will comply with the requirements of the LEU Reform Amendments.

The Plan will be inspectable for compliance by the NRC, and 2) an Annex that will contain examples of hcw various portions of the plan will be implemented and will give other general information about the facility. The Annex will not be incorporated as a condition of the license and will not be used as a basis for inspection.

The Plan shall include a section that provides a general discussion of how the licensee's MC&A program satisfies the general performance objectives of 10 CFR 74.31(a). The description shall include information on the plant, the process, and the key features of the MC&A system including plant layout, types of internal controls, and how the performance of the system will be monitored.

This section should be written sufficiently general to give the licensee flexi-bility to make procedural or operational changes that do not decrease the level of safeguards performance without informing the NRC.

The licensee will have the flexibility to change procedures referenced or equipment specified in the Annex without notifying the NRC as long as the MC&A system performance documented in the Plan is not degraded. The material given in the Annex should be cross-referenced to the Plan, and it should be clearly evident how it relates to the affirmations stated in the Plan or to the requirements in the LEU Reform Amendments. The Plan should also clearly state that references to the Annet do not incorporate the Annex into the Plan commitments unless otherwise indicated.

1-1

2.0 ACCEPTANCE CRITERIA FOR CONFIRMING THE PRESENCE OF SNM, RESOLVING INDICATIONS OF MISSING MATERIAL AND AIDING THE INVESTIGATION AND RECOVERY OF MISSING MATERIAL 2.1 RULE 74.31(a) General Performance Objectives.

Each licensee who is authorized to possess and use more than one effective kilogram of special nuclear material of low strategic significance, excluding sealed sources, at any site or contiguous sites subject to control by the licensee, other than a production or utilization facility licensed pursuant to Part 50 of this chapter, or operations involved in waste disposal, shall implement and maintain a Commission approved material control and accounting system that will achieve the following objectives:

(1) Confirm the presence of special nuclear material; and (2) Resolve indications of missing material; and (3) Aid in the investigation and recovery of missing material.

2.2 INTENT Each licensee subject to this proposed rule is expected to implement and maintain an MC&A system that will be capable of achieving the above general performance objectives.

It is intended that the licensee will be able to verify the presence of the SNM for which the licensee is responsible at least annually. Verification involves confirmation of the presence of the material.

An accurate record system is essential.

It is also intended that the licensee will investigate and determine the causes of indications of possible losses of significant amounts of SNM. An indication may be an excessive inventory dif-forence, an allegation of theft, evidence of a breach in the physical security barrier, or unauthorized entry into buildings or areas in buildings containing accessible SNM accompanied by other factors such as tampering with equipment or missing items, or the inability to locate material called for in routine oper-ations. The licensee's investigations of possible losses are expected to either locate the missing material or reach a conclusion based on objective evidence as to the most probable cause for the indicated loss.

If theft or diversion is the assigned cause of a loss, the licensee will be expected to provide any available information to help identify the loss by material type and quantity and the time period over which the loss could have occurred in order to aid in the recovery of the SNM.

2.3 AFFIRMATIONS The licensee must make the following affirmations with respect to confirm-ing the presence of SNM, resolving indications of missing material, and aiding the investigation and recovery of missing material:

2-1

An MC&A system will be developed and maintained that is capable of annually confirming the presence of all SNM, with the possible exception of waste materials stored in holding accounts.

An expeditious investigation will be promptly initiated for all indications of significant losses of SNM and allegations of theft of SNM.

Information will be provided to aid in the investigation of indications of missing material and recovery of SNM in the event of a loss, theft, or unauthorized diversion.

A cause or probable cause that is based on objective evidence will be assigned for all indications of possible loss that are investigated.

The results of investigations of alleged thefts or of indications of loss in which the amount of SNM exceeds a quantity of U-235 established by the NRC on a site specific basis will be reported to the appropriate NRC Regional Office.

2.4 INFORMATION TO BE INCLUDED To be complete, the Plan should contain the following information:

A description of the MC&A system defining its key components and their purposes.

A description of the licensee's plan for investigating indications of missing material.

In addition, enough information regarding the licensee's approach for resolving indicated losses should be presented to permit an evaluation of the plan and determination if the affirmations can be achieved.

The bases for deciding to initiate a loss investigation; some categories of material that the licensee may wish to treat separately are:

1) tamper-safe sealed items or sealed sources, 2) unsealed items in storage locations, 3) unsealed items not in storage, and 4) bulk material within processing equipment.

The bases for resolving an indication of missing SNM or terminating an investigation of a possible loss.

2.5 ACCEPTANCE CRITERIA A judgment that the licensee's Plan is acceptable will be based on the i

following criteria:

Based on the commitments and details provided in the MC&A Plan, the licensee can be expected to achieve the general performance objectives of the rule.

The licensee's investigation of indications of a loss of SNM can be completed expeditiously but no longer than 60 calendar days after the indicated loss is discovered even if a physical inventory is required.

I 2-2 l

i

Generally speaking, allegations of theft or indications of loss involving less than 500 grams U-235, need not be subject to investigations unless requested by the NRC.

The licensee's Plan commits to expeditions investigation of SNM losses.

For incidents involving an alleged theft or an indication of missing SNM accompanied by evidence that a theft may have occurred, the investigation shall be initiated within eight hours. All investigations shall be docu-mented with the level of effort taking into account the seriousness of the event.

The bases for initiating a loss investigation include (1) an allegation of theft, (2) an indication that a theft has occurred, (3) an inventory dif-ference equal to or exceeding an action threshold based on detecting with 90% probability the quantity established by the NRC on a site specific basis, (4) failure to locate an item after normal procedures for location of the item have been completed, and (5) any knowledge or other positive indication of a loss of SNM.

However, an investigation may be discontinued if the indicated loss quantity is less than the site specific Detection Quantity and there is no tangible evidence that a theft or less may have occurred or that a recurring loss situation is continuing.

In the event of an indication of missing SNM, the MC&A system will provide an estimate of the amount of SNM involved.

In addition, when possible, the licensee should identify the physical form of the material and the time interval within which the loss, if any, would have occurred.

No investigation relative to an indication of a loss or theft of SNM exceeding the site specific Detection Quantity (established by the NRC) should be declared as completed but unresolved without first conducting a shutdown, cleanout inventory in which all unsealed material is remeasured for SNM content.

Additional criteria for loss indications involving excessive inventory differences are given in Section 5, and for item losses are given in Section 6.

2.6 QUESTIONS AND ANSWERS 1.

Q:

What constitutes evidence that a theft of SNM has occurred?

A:

Indications that a building or storage enclosure has been broken into may not be sufficient reason to conclude a theft may have taken place.

However, if the indication is accompanied by other evidence such as missing SNM, damaged SNM containers, tampering with equipment usable for transferring SNM, etc., the event should be treated as a possible theft or attempted theft.

2.

Q:

What kinds of information might licensees be expected to provide to aid in the investigation and recovery of missing SNM?

A:

It is not intended that licensees be required to implement any special procedures to enable " localization" of a diversion of SNM.

2-3

However, information will usually be readily available to the licensee from production records, accountability data or routinely prepared reports that can facilitate the investigation of missing material.

Examples of the kinds of information that may aid investigations and timely recovery of the SNM are:

Data that led the licensee to determine that a loss or theft had occurred; Any abnormal events that may have contributed to or caused the loss; The interval during which the loss could have occurred; The amount of material and material form involved in the loss; and The names of the people who could have been responsible for the loss.

3.

Q:

Why must the licensee's Plan show the capability to initiate an investigation within eight hours in the event of incidents involving an alleged theft or an indication of missing SNM accompanied by evidence that a theft may have occurred?

A:

A commitment on the part of the licensee to initiate investigations of these kinds of loss indicators without delay is important so that physical evidence needed for resolution is not lost and events are fresh in the memory of possible witnesses.

If law enforcement agen-cies or other authorities are brought in to investigate or recover the material, time may be an essentici factor in collecting factual information and, in the event that a theft actually occurred, prompt apprehension of the thief and recovery of the material is important.

Also, public concern may be an issue in some events, and considera-tion needs to be given to show that efforts are being taken promptly to protect public health and welfare.

4.

Q:

What is meant by annually when confirming and verifying the presence of SNM7 A:

Verification involves physical confirmation of the presence of SNM at least every 12 months.

Reasonable alternative approaches are:

(1) state that no more than 13 calendar months (e.g., Jan. 17, 1985 to Feb. 17, 1986) will elapse between any two consecutive inventories, no more than 25 calendar months will elapse between any three consecutive inventories, and no more than 36 calendar months will elapse between any four consecutive inventories; (2) establish a date for starting the inventory that will be met within plus or minus 30 days; (3) start the inventory in the same calendar quarter of each calendar year but no more than 13 months will elapse between the start of two consecutive inventories.

2-4

3.0 ACCEPTANCE CRITERIA FOR MANAGEMENT STRUCTURE 3.1 RULE 74.31(c) System Capabilities.

To meet the general performance objectives of paragraph 74.31(a), the material control and accounting system must include the capabilities described in paragraphs (c)(1) through (8) of this section.

The licensee shall:

(1) Establish, document, and maintain a management structure which assures clear overall responsibility for material control and accounting functions, independence from production responsibil-ities, separation of key responsibilities, and adequate review and use of critical material control and accounting procedures.

3.2 INTENT The intent of this paragraph is to require licensees to implement a management structure that permits effective functioning of the MC&A system, and assures that the MC&A program performance will not be adversely affected by the management structure.

Documentation, review and approval of the procedures, and the assignment of the key functions to specific positions eliminates ambiguities about what is to be done by whom.

The management structure is meant to separate key MC&A functions from each other in order to provide overchecks that increase MC&A system reliability and counter defeat of the system through deceit and falsification.

It is also meant to free MC&A manage-ment from conflicts of interest with other major functions such as production.

3.3 AFFIRMATIONS The licensee must make the following affirmations with respect to manage-ment structure:

Responsibility for the overall MC&A system management will be assigned to a position that is separate from production responsibilities or any other responsibilities that may give rise to conflicts of interest.

The responsibility for each MC&A function is assigned to a specific posi-tion in the organization, and the organization is structured in a way that the key functions are separated or overcheck one another.

The position descriptions will be made available in writing to the personnel affected.

The facility organization and the MC&A policies and procedures are documented.

All critical procedures will be reviewed and approved prior to their implementation.

3-1

Management policies will be established, documented and maintained to ensure that all critical MC&A procedures are adhered to including i

measurement procedures used for accountability purposes.

- 3.4_ INFORMATION TO BE INCLUDED i

To be complete, the Plan should contain the following information:

l

}

' A description of the management structure giving the functional respot si-bilities of each organizational unit showing that the MC&A organization is independent of potentially conflicting responsibilities, and that the i

. organization allows separation or overchecks of key MC&A functions.

A j

_ chart showing the relationships between the functions should be included.

The Annex should contain an organization chart and descriptions of duties, authorities, and responsibilities for each key MC&A position.

A descripti:on of the organizational relationships of persons who perform key MC&A functions including performing measurements and measurement control activities, preparing source documents, performing shipping and receiving activities, and conducting inventories.

A description of the policies, instructions, procedures, duties, responsi-bilities, and delegations of authority.in sufficient detail to demonstrate the separation or overchecks built into the MC&A system.

A summary of the review and approval plan for critical MC&A procedures.

Criteria for identifying _ critical procedures.

3.5 ACCEPTANCE CRITERIA A judgment that the licensee's Plan for establishing, documenting, and maint.nining the management structure'is-~ acceptable will.be based on the follow-ing criteria:

.The management structure and MC&A policies and procedures will identify-authorship, approval authorizations, and the effective date.

The responsibilities and authorities vor each position assigned a function having a significant impact on SNM control and accounting (including all positions authorized to control SNM movement, source data generation, measurement control, and data analysis) are clearly defined in a written

. position description that defines ~the responsibilities for that position.

l

- - The background experience required for each position assigned an SNM control and accounting function will correspond to the duties required of that position.

t 4

. The descriptions of the management structure and assignment of duties and 4

authorities show that those responsible for each MC&A function will have

' sufficient authority to perform the function in the intended manner.

4 3-2 m

-m,-,ee

-.,,, _-.n-

- -~-----

-,n.,

e

--eg y

~n--c r----m s,----#

9

~_

I The MC&A organization will be separate from the production organization and will also be separate from organizations that generate source data if practical; otherwise,. independence of the functions will be attained by suitable controls.

Examples of suitable controls are complete separation

-of the central accounting functions from the SNM handling functions; overchecks of shipping, receiving, and waste disposal documents-by MC&A before SNM is moved; and overchecks by authorized personnel of measurement data and other source data, calculations, and data analyses.

'The MC&A management responsibility will be delegated to an individual at an organizational level sufficient to assure independence of action and i

objectiveness of decisions.

No two key MC&A functions will be assigned to the same person unless adequate checks and balances are provided. As a consequence of this i

criterion:

Individuals who generate source data, such as performing measurements, preparing analytical reports, and performing measurement control or shipping and receiving activities, will not i

perform any accounting or record control functions unless overchecks of the work are provided to prevent falsification of both source data

-and accounting records, and No individual may have the sole authority to overcheck,. evaluate performance or audit information for which that individual is l

responsible.

J Critical MC&A procedures and all changes to them which directly affect the licensee's ability to detect the loss of SNM or resolve indications of missing SNM as per $74.31(a) will undergo technical review by cognizant

-members of the staff and will be approved by line management directly 4

affected and by a level of management above the level responsible for 1

executing the critical procedures (but not beyond onsite plant management).

l All critical procedures should be maintained current and show for each

, procedure (1) the latest revision number, (2) date issued, (3) who pre-pared the procedure, and (4) who approved the procedure.

3.6 QUESTIONS AND ANSWERS

~

1.

Q:

What are the primary responsibilities of the MC&A organization manager?

A:

The MC&A' manager has overall responsibility for:

. Adequacy, completeness and effectiveness of the MC&A program, The adequate measurement of all SNM receipts, shipments, waste discards, and material on hand at the time of inventory, Recording and reporting of SNM status, 3-3

.__.,,o_.

.,...._,,__--3

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

...m ray-w~,...,m

,c-m

.~,...,,-...-_,_..,__.4

Management'and effective use of approved MC&A systems, proce-

+-

dures and results including recommending corrective actions and confirming that they have been carried out,

> Control and approval of all physical inventories of SNM including evaluation and reconciliation of inventory values, ids, estimated standard error of the ids, and approvals of reports adjusting inventory records, Performance of technical evaluations and analyses of measurement capabilities including equipment and methods, Reviewing new or changed operations to determine if they involve previously unreviewed safeguards considerations and determining that safeguards plans meet regulatory requirements, and Coordinating MC&A activities with other organizations performing

-safeguards activities.

2.

Q:

-What is meant by " clear overall responsibility"?

A:

Responsibilities should be assigned in a way that no ambiguity exists i

regarding who is responsible for a task or function.

This may be done by making assignments by name or position title where each posi-tion is currently assigned or delegated to an individual.

Responsi-bilities should not be assigned to an organizational unit but rather-to specific positions within that unit.

"Overall responsibility" means responsibility for assuring that certain defined tasks and/or duties, that comprise a major organiza-tional function, are performed satisfactorily, even though such duties may be performed by others.

" Direct responsibility" on the other hand pertains to duties performed by the individual himself (or herself).

3.

Q:

What are the " key responsibilities" which must be separated?

A:

The tasks for material measurement and material accountancy must be separated at some point so that one person does not perform both data 3-generation and evaluation tasks and thereby control all shipper /

receiver difference and/or inventory difference information all the way from raw data taking through evaluation.

In addition, separation of responsibilities must be preserved in the independent assessments as described under the discussion of $74.31(c)(8).

4 4.

Q:

When may a conflict of interest occur?

A:

A conflict of interest may occur when a performance objective of one of the functions assigned to an individual could be contrary to a performance objective of another of that individual's functions.

Examples are:

3-4

~

When a production operator or supervisor has responsibility for both maintaining effective material accounting and meeting production schedules; or When an individual can control decisions having the possibility of personal gain, such as when authorized to evaluate the per-formance of the work or organization for which that person is responsible; or When an individual has authority over two functions that should normally serve as overchecks, such as generating source data as well as maintaining material accounting records.

5.

Q:

What are some examples of overchecks that serve to protect the MC&A system against falsifications or errors?

A:

Examples of appropriate overchecks are:

Review of the measurement data and calculations by another person; Maintaining a duplicate copy of all source data and transfer forms under controls separate from the accounting function; Double entry bookkeeping; Separating computer software development and modification from the data input capability; Separating the power to authorize a discard of SNM from the authority for hands-on access; and Performing independent audits.

6.

Q:

What are " critical MC&A procedures," and what specific functions do they include?

A:

Critical MC&A procedures are those procedures which, if not performed correctly, could prevent the licensee from satisfying the two objec-tives of (1) confirming the presence of special nuclear material and (2) resolving indications of missing material and aiding in the investigation and recovery of missing material.

Hence, critical procedures encompass the establishment of basic MC&A system policies, determination of the physical inventory, determination of inventory differences and shipper / receiver differences including their resolu-tion, detection of a loss, maintaining a measurement control program for all key measurement systems, and determination of the total MC&A measurement error.

i 3-5

7.: Q:

What are some examples of MC&A procedures that would not be " critical procedures"?

A:

Examples would be (1) a measurement procedure that (when taking into account its measurement uncertainty and the quantity of SNM to be measured during an inventory period) would not be dealing with a key measurement system, and (2) procedures that define ~the format for reporting source data, calibration data, or measurement control data.

3-6 1 ___

4.0 ACCEPTANCE CRITERIA FOR MC&A MEASUREMENTS 4.1 RULE 74.31(c)(2)

Establish and maintain a measurement system which assures that all quantities in the material accounting records are based on measured values.

74.31(c)(3)

Follow a measurement control program which a..,ures that measure-ment bias is estimated and significant biases are eliminated from inventory difference values of record.

74.31(c)(4)

In each inventory period, control total material control and accounting measurement uncertainty so that twice its standard error is less than the greater of 9 kilograms of U-235 or 0.25%

of the active inventory, and assure that any measurement per-formed under contract is controlled so that the licensee can satisfy this requirement.

4.2 INTENT The intent of the rule is that all record quantities (for both uranium and U-235) are to be based on measurements.

The measurement systems used to deter-mine these quantities are to be controlled by a formal measurement control program.that results in a total measurement standard error that is within the limits stated in the rule.

The program will also provide bias estimates to be used in correcting the ID and shipper / receiver measurements for significant measurement biases.

Periodic monitoring is needed to assure that biases and random error variances are controlled within limits necessary to satisfy the requirements of the rule.

4.3 AFFIRMATIONS The licensee must make the following affirmations with respect to account-ability measurements and measurement quality:

A measurement system will be established and maintained for all SNM l

receipts, removals and inventory items, and all quantities of SNM in the material accounting records will be based on measured values.

A measurement control program will~be followed by which all measurement l

biases associated with key measurement systems, are estimated and any significant biases will be eliminated from ID values and shipper / receiver differences.

l-Measurement systems that are the key contributors to the total measurement i

standard error will be identified.

The list will be reviewed annually and updated as necessary..These will be considered as key measurement systems i

4-1 i

and their standard deviations will be monitored and controlled by the measurement control program.

The estimate of the measurement contributions to the SEID will be traceable to the appropriate measurement error data and to the calibration standards used.

The total measurement error will be controlled so,that twice its standard error for a material balance period will be less than the greater of 9 kilograms of U-235 or 0.25% of the active U-235 inventory in the inven-tory period.

(NOTE:

"Twice the standard error associated with total measurement error" was previously denoted as "LEID".)

The measurement systems will have adequate calibration frequencies, suffi-cient control of biases, and sufficiently small standard deviations to achieve the requirements of S74.31(c)(4).

A measurement control program will be used by the licensee and contractors to assure that the quality of the measurements will be maintained on a level consistent with the regula-tory requirements.

4.4 INFORMATION TO BE INCLUDED To be complete, the Plan should contain the following information:

Identification of the measurement systems that are expected to be key contributors to the total measurement standard error.

The measurement methods and their applications should be described.

The expected standard deviation of each measurement system and the quantity of material to be measured by each system should be described in the Annex.

A description of the approach to be used for monitoring and controlling measurement error standard deviations of the key measurement systems and for estimating biases of all such measurement systems including the bases for:

Setting frequencies of replicate measurements for monitoring standard deviation and standards measurements for estimating biases and per-forming recalibrations; Assigning standards to be used for calibrations and bias determinations; Setting control limits for determining when a measurement system is out of control (i.e., warning limits and action limits for standard measurements and replicate measurements);

Deciding which measurement data will be corrected for an estimated bias; and Determining when records will be adjusted for measurement biases.

The controls over measurements performed by contractors and how they are included in the licusee's system.

4-2

1 An example in the Annex showing the estimation of the measurement standard error for a' typical inventory difference based on material flows and inventory components and the variances due to random effects and bias estimates expected for the plant.

The example should include the equa-tions for calculating measurement standard deviations and bias estimates, I

a description of the error model, and the error propagation method used in the calculation of the total measurement standard error.

The example I

should also show which measurement systems are key contributors to the l

total measurement standard error.

i If the licensee does not intend to calculate the measurement contribution i

.of the estimated SEID for each inventory period, the approach should be described for assuring that the measurement contribution used (based on previously accumulated data) is valid and meets the requirements of 74.31(c)(4).

4.5 ACCEPTANCE CRITERIA A judgment that the licensee's Plan for accountability measurements and measurement quality is acceptable will be based on the following criteria:

The licensee identifies key measurement systems in the Plan (and non-key systems in the Annex) which are used to generate SNM quantities for accountability purposes.

A measurement system being defined as any

. instrument or device, or combination of devices, used to derive (1) a uranium element concentration, (2) a U-235 or U-238 quantity, (3) a U-235 j

enrichment or isotopic distribution, (4) a bulk material mass, or (5) a bulk material volume.

The licensee's description of the MC&A measurements and measurement control plans show that the measurement systems that are the key con-tributors to the total measurement standard error will be routinely monitored by the measurement control program.

The set of key measurement systems identified by the licensee is estimated to account for at least 90% of the total measurement contribution to standard error, and includes t

any measurement system that is normally utilized to measure a total SNM l

quantity, during an inventory period, in excess of 25% of the active inventory quantity.

The measurement control program provides for periodic recalibrations and routine determinations of the random effects and biases of at least the

-key measurement systems for materials accounting.

The program should be in accordance with the following principles:

The standard deviations of measurement systems are estimated from replicate data from measurements made in the same manner as made routinely on typical items and samples.

If standard deviations are based en replicate measurements of reference standards (for NOA or mass measurement systems), the licensee collects data that demon-strate that the standard deviation estimates do not differ signifi-

.cantly from those based on replicated routine item measurements.

(NOTE:

Random analytical chemistry measurement uncertainties must be i

i 4-3 i

l

~

.i

' derived from the measurement.of replicate process samples rather than 1

based on multiple measurements.of a standard.)

All-reasonable and probable sources of measurement error, such as the effects of sampling,-instruments, and environmental-factors are accounted for in estimating the standard deviations, either directly as experimental variables in an analysis of variance or by being

^

" included in the sample of measurement control' data from which the standard deviations are determined. The licensee will show that it is reasonable to expect that the measurement systems that could signifi-(

cantly impact the ID will be in control.so that the total measurement standard error will be within the limits st.ited in the rule at the Linventory time.

Bias tests are made by measurements of represent 7tive standards whose assigned values are traceable to national measurement systems.

The control standards should closely resemble the unknowns to which the method is applied, and the measurement procedures and conditions of measurement should closely resemble those of typical measurements made on unknowns.

The recalibration frequencies are to be compatible with the expected t

stability of the measurement systems.. All calibrations should be made with standard reference items or materials, or with standards in which traceability to a standard. reference material is maintained.

Recali-brations for all measurement systems should be performed at frequencies compatible with widely established.or licensee-demonstrated stability of c

each particular system.

The reference standards used for recalibrations need not be representative of the unknowns to be measured by the system unless it is a point-calibrated, bias-free system, in which case the calibration standard (s) must be representative.

A record of bias estimates for.each key measurement system (not defined as

a. bias-free, point-calibrated system), as. derived from control standard measurements, will be maintained.

The basis for determining the time frame associated with each significant bias should be provided (so that the quantity of measured SNM to which that bias applies can be determined).

Bias corrections (expressed as grams U-235) will be derived for each significant bias (based on the quantity of measured SNM to which the bias applies).

The Plan should confirm that each significant bias is either applied as a correction to the inventory record or.as a direct adjustment to inventory difference, and if both options are to be utilized, the Plan should also provide decision criteria for determining the proper bias correction methodology. Generally, it will be sufficient to simply maintain a bias adjustment account to be applied to ids, except where shipments or subsequent material balances require that bias adjustments be made to individual lots or items.

(See Tingey 1983)

A' bias correction for a single key measurement system should be considered as being "significant," and thus applied (directly and/or indirectly) as a correction to the ID if such a bias is statistically significant at the 95% confidence level, and if either (or both) of the following are also true:

4-4

I.

Applying the correction would cause the ID to exceed the Detection Threshold value, or II. The biasr is greater than 0.0100% relative and also impacts the ID value by more than 1000 grams U-235.

Additionally, the net algebraic sum (expressed as grams U-235) of all statistically significant (95% confidence level) biases (from key measurement systems not defined as bias-free) that have not been applied as a correction under condition I and/or II, above, are considered to be significant and should be applied as a net correction to the ID if either (or both) of the following are true:

III. Applying such correction would cause the ID to exceed the Detection Threshold value, or IV.

The net correction impacts the ID value by more than 5.00% of the site-specific Detection (Goal) Quantity, or 10,000 grams U-235, whichever is larger.

The proposed schedules and frequencies of measurements for measurement control are designed so that the estimates of standard deviations and measurement biases will be based on measurement control data collected under the same measurement circumstances and over a span of time that corresponds with the time span of the SNM accounting measurements to which the standard deviations and bias estimates will be applied.

Redetermina-tions of the standard deviation and bias of each key measurement procedure will be made periodically.

The frequency of such redeterminations should be a function of the system's stability, but should be at least every four months.

Pooling of data from previous determinations of the standard deviation is used only if statistical tests show the standard deviations do not differ significantly, and further provided the pooled data does not include any data generated more than 24 months prior to the redetermination of such bias or standard deviation.

The effort expended by the licensee in monitoring and controlling the bias of each measurement system and the standard deviation of each key measure-ment system is shown to be consistent with its impact on ID and the total measurement standard error.

The number of degrees of freedom for estimat-ing the measurement standard deviation may be graded according to its contribution to the total measurement standard error.

Warning limits for a change in bias (for those systems that are not point-calibrated, bias-free) will be set at the 0.05 level of significance. Warning limits at the.05 level are optional, however, for the monitoring of replicate data (for standard deviations).

If control data exceed this limit the individual responsible for the measurement control program will be notified (this normally should occur within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />) and a data review will be initiated to find the cause and corrective action will be taken when appropriate.

Such reviews and corrective actions sould be completed and documented within 2 weeks.

4-5

9 Action limits for both standard deviation and bias will be set at the 0.001 level of significance for all key measurement systems, except that no bias control limits are needed for bias-free, point-calibrated systems.

If. control data exceed this limit, measurement values obtained since the generation of the last within control data point shall not be used for MC&A purposes until corrective action and resolution is completed, and the system is back in control below the 0.05 level, with the minor exceptions of those cases that fall within the exception criteria stated in the NRC generic letter of June 17,1983 (Guidance for Out-of-Control Situations).

I

'(It should be noted that other criteria for initiating corrective action relative to potentially out of-control measurement systems may be accepted where it can be demonstrated that the licensee's capability to meet the 0.25% of active inventory requirement for measurements will not be l

jeoparized.

The approach to be used for bounding the total measurement standard error for a typical material balance period, as shown in the Annex, meets the following criteria:

All. reasonable and probable sources of measurement error affecting ID are to be included, t

l Any assumed measurement standard deviations are shown to be reasonable.

They may be shown to be reasonable by comparison either with records of past performance data from the licensee's facility or published measurement performance in similar applications, as reported in such sources as Rogers (1983) or Reilly and Evans (1977).

i (Records showing these data must be available to the NRC).

The calculation of the total measurement standard error is performed l

'in accordance with a recognized error propagation method.

Such methods have been published by Jaech (1973), Lumb and Tingey (1981);

and the IAEA (1977).

The licensee will confirm the accountability measurements supplied by a contractor will be controlled by a measurement control program, and that t

the licensee will confirm by an audit every two years that the contractor's measurement control program is adequate.

If the licensee does not plan to use an actual calculated total measurement standard error, the licensee will have current measurement control program data or values based on engineering estimates (if control data is not available) to justify the standard deviations for the measurement systems.

The licensee will be able to show that the measurement control program includes the key measurement systems.

The predicted range of variation of the active inventory is based on i

production scheduling commitments and previous experience, and the assumed ranges of standard deviations of the key measurements are based on historical data for those measurement systems currently under the licensee's measurement control program or, for applicants, from

(

publications such as Rogers (1982) or Reilly and Evans (1977).

l l

4-6

4.6 QUESTIONS AND ANSWERS 1

s 1.~

Q:

What is mear.t by the total measurement standard error?

A:

The total MC&A measurement uncertainty is the contribution to the SEID of a material balance due to the measurement variances.

l The measurement variances should include all sources of vari-ability of the measurement and sampling processes, in::1uding random and systematic effects due to such factors as instrument I

and analyst differences, time of day, time during an inventory period, and variations in environmental conditions.

-2.

Q:

Must the total measurement standard error for the facility be calculated at each physical inventory?

A:

No.

The licensee has two options.

Under option one, the licensee would calculate the measurement uncertainty at each physical inventory.

Under option two, the licensee would make an analysis of the total measurement uncertainty only when the new plan is submitted and when changes are made to the process or to the measurement systems used which would invalidate the previous analysis.

For the second option, the licensee would

' also identify a sufficient set of measurement parameters which will be monitored to assure that the 0.25% goal is still being achieved by the major systems that contribute to the measurement uncertainty.

3.

Q:

Is there any limit on how well any single key measurement system must be controlled?

A:

Yes, out-of-control limits must be utilized, and the overall performance level of 0.25% of the active inventory must not be compromised by a system whose performance is not being adequately monitored by control measurements of sufficient frequency.

4.

Q:

Why is there a requirement that the standard deviations and biases must be redetermined periodically?

A:

A change of the measurement standard deviation or departure from the calibration point may occur gradually for a variety of l

reasons, and sometimes it occurs abruptly.

To detect these changes before too much uncontrolled or biased measurement data l

has been acquired for which remeasurements are not always l

possible, the state of statistical control must be checked periodically. Most methods should be checked at least monthly and some may need weekly or daily checks.

Lengthening the time intervals between determinations of. bias and standard deviation r

should be justified by sequential monitoring data, such as con-l trol chart records, showing that the method is stable over such time intervals.

In any event, no more than four months should go by without a redetermination of bias and standard deviation for key systems that are utilized more or less continuously.

Of s

I 4-7

course, methods used only periodically for SNM accountability analyses need only be monitored during the time periods actually used.

5.

Q:

Must measurements made by contractors be included in the 0.25%

goal?

A:

Yes, to the extent that they would have to be covered if the licensee were performing the measurement.

The licensee is responsible for monitoring the contractor's measurement control program to assure that when uncertainty from those measurements-are combined with the licensee's own measurement uncertainty, the overall goal is still met. Also, the contractor's measure-ment program should be included in the independent performance assessment.

6.

Q:

Why has the concept of active inventory been used?

A:

The concept of active inventory is used to establish a perfor-mance level for control of the measurement uncertainty.

Since this performance bound is intended to reflect normal plant operating conditions, it is not the intent of the definition that activities be conducted just to increase the active inven-tory value.

7.

Q:

Why is active inventory defined the way it is?

'A:

Active inventory as defined in $74.4 is used to identify that material which has an opportunity to contribute to the inventory difference and measurement uncertainty.

Common terms are excluded because when a term shows up more than once in the active inventory calculation, has the same value, and that value is derived from the same measurement, that term can not contri-bute to the ID or measurement uncertainty (e.g., an item on beginning and ending inventories will just cancel itself in the ID calculation and does not add to the uncertainty in ID).

8.

Q:

If a chemical titration method is used for measuring the uranium concentration of two unlike material forms (e.g., UNH solutions and dirty 002 scrap), is the method regarded as one or two measurement systems?

A:

For this particular example we are dealing with two measurement systems because (1) the sampling technique differs for the two materials, (2) one control standard (or set of like standards) could not be representative of both material types, and (3) the initial sample aliquoting and pretreatment steps would differ

-for the two material types.

Generally speaking, individual measurement systems are defined by their unique combination of the following five parameters:

(a) measurement device or equipment utilized, (b) standard (s) used for calibration, (c) standard (s) used for control, (d) sampling technique and equipment utilized (if any), and (e) subsampling (aliquoting) and/or sample pretreatment (if any).

4-8

5.0 ACCEPTANCE CRITERIA FOR PHYSICAL INVENTORIES 5.1 RULE I:

74.31(c)(5)

Unless otherwise required to satisfy Part 75 of this chapter, i

perform a physical inventory at least every 12 months and, within 60 days after the start of the inventory, reconcile and adjust the book inventory to the results of the physical inven-tory, and resolve, or report an inability to resolve, any inven-tory difference which is rejected by a statistical test which has a 90% power of detecting a discrepancy of a quantity of U-235 established by NRC on a site-specific basis.

5.2 INTENT The principal method of confirming the presence of SNM is to perform a physical inventory and compare it to the book inventory.

If all components of the inventory are based on measured quantities and if all SNM is included, the expected difference between the book inventory and the physical inventory is zero.

In any actual case, the size of the estimated inventory difference depends on measurement errors, the accuracy of the records, and nonmeasurement contributors.

The intent of $74.31(c)(5) is to require licensees to perform annual physical inventories and resolve any inventory difference which exceeds its detection threshold.

5.3 AFFIRMATIONS The licensee must make the following affirmations with respect to physical inventories:

Unless otherwise required by Facility Attachments that satisfy 10 CFR Part 75, a physical inventory will be performed at least every 12 months

)

and will be used as the basis for reconciling and adjusting the book inventory which will be done within 60 days after the start of the physical inventory.

The inventory procedures will be clearly written and will be reviewed and approved by the individual responsible for the conduct of the inventory.

The: individual responsible for the conduct of the inventory either will be free from potential conflicts of interest or will be overchecked suffi-ciently to prevent compromising the validity of the physical inventory.

l The inventory listing will include all SNM possessed (on the inventory l

date), with the possible exception of waste materials assigned to holding i

l accounts (in accordance with DOE /NRC Form-741 instructions) and residual l

holdup that can be regarded as zero by definition, and that all such j

listed SNM quantities shall be based on measurements (for SNM quantity).

l l

5-1 1

i L

Within 60 days after the start of an inventory, the inventory difference will be resolved or the inability to resolve an inventory difference, which is rejected by a statistical test that has a 90% power of detecting a discrepancy of a quantity of U-235 established by the NRC on a site-specific basis will be reported to the NRC.

Discrepancies in the identity, quantity or location of items, objects or containers of SNM that are detected during the physical inventory will be corrected.

~

Inventory difference values will be corrected for (1) accounting adjust-ments resulting from prior period activity, and (2) significant biases that have not been previously taken into account.

(Note:

See the definition of significant bias in Chapter 4.)

Adjustments made to reconcile the book inventory to the physical inventory will be in accordance with standard accounting practices and will be traceable and auditable in the licensee's records.

Corrections for k1own significant biases in the physical inventory values will be made before inventory accounts are changed to reconcile them to the physical inventory values.

In the event an estimated ID indicating a discrepancy greater than the quantity of U-235 established by tha NRC on a site-specific basis is ni t resolved wichin the 60 day reconciliation perica, the innstintion wlil be continued if requested by the NRC.

The results of physical inventories and of investigations and resolution actions following an excessive estimated ID will be recorded and will be auditable.

5.4 INFORf1ATION TO BE INCLUDED The Plan should contain the following information:

A description of the procedures for performing physical inventories containing sufficient detail to demonstrate that they will provide valid inventories.

The description will include a general outline of how:

Inventory functions are organized and how the functions are separated; Inventory teams are assigned and instructed in the use of uniform procedures; Source data is obtained, verified and transmitted; Inventory forms are controlled; Item counts verify each item while preventing counts of any item more than once; 5-2

Cut off procedures for production and material movement are handled; and Thoroughly the process vessels, equipment and piping need to be cleaned out or drained down, and how residual holdup is treated.

A description of special item storage and handling or tamper-indicating methods used so that the recorded SNM content can be assumed to be valid at inventory time without remeasurements.

This should include a descrip-tion of the type of storage facilities and any special features (i.e.,

locked doors, seals, hardening or short residence times) that facilitate detection or prevent unauthorized removals of SNM from items.

Actual facilities and procedures in use at the time of the preparation of the Plan should be described in detail in the Annex.

A description of how item identities are confirmed and how tampering with the contents of items will be detected or prevented.

The basis for determining which items are to be measured for SNM content at physical inventory time and justification of any proposed alternatives to measurement of any SNM included in the inventory (whether determined previously or at the time of inventory--e.g., the U-235 contained in liquid waste materials where an average enrichment factor is applied to

,. m the measured uranium concentration).

If statistical sampling (for deter-

.'.[$

mining which items are to be remeasured) is proposed as an alternative

-.5; method for inventory verification, the Plan must describe the sampling jd-plan.

The description should include:

.f W

7., -

The method of classifying (stratifying) the types of items to be

sampled,

'., e; e -

How the sample sizes (i.e., number of items) will be calculated, 1

The method or methods used for random selection of items to be Ch3 included in the statistical sample,

c791

e The quality of the measurement methods used to verify original

.f !:.

measurement values, ycf,;

m..

The plans for reconciling discrepancies between original and

[f remeasurement values, and when additional tests and remeasurements T.r'$.90 would be performed, and

,jf~

l y '.?.

The basis for discarding an original SNM quantity (for an item on pu.:

inventory) and replacing it with a remeasurement value.

h

^

In addition, a list of typical inventory strata, and an elaboration on plans for investigating and reconciliation of indications of material FQ shortages should be given in the Annex.

fie

. ;x !

• w-M.

The basis for deciding when the element and isotope factors for items, objects, or containers will be measured directly for inventory and when

it.

they may be based on or derived from other measurements.

If inventory 177

q,=

5-3 MJ.8

? g @

d.k.;','

a. ~ :.,

' -)

1 B

=

5W values are based on or derived from other measurements, the Annex will M

describe how the values and their uncertainties are determined, and the Q

basis for their validity, and will include any supporting historical g

information used.

g A description of how the book inventory (for both U and U-235) is recon-G ciled and adjusted to the physical inventory.

The basis for setting the action threshold for an excessive ID at which an I

investigation will be initiated in order to achieve a 90% power of detect-ing a discrepancy equal to the quantity of U-235 established by the NRC on jg a site-specific basis.

The description should include the source of the data and the statistical method used to determine the estimated standard i

error of the ID.

The Annex should give all relevant formulas used in the computations and a detailed example of determination of the test statistic and the detection threshold.

p The Annex should contain a description of a typical inventory composition for the plant, the formulas for calculating the inventory difference, and i:

the resolution procedures to be used.

The basis for determining the non-measurement contribution to the total k

standard error of the inventory difference.

4 5.5 ACCEPTANCE CRITERIA A judgment that the licensee's Plan for physical inventories is acceptable will be based on the following criteria:

A The inventory plan will contain sufficient detail to demonstrate that the physical inventory process will be organized and coordinated so that all i

persons are instructed in the use of uniform procedures of checking SNM

^

quantity and recording observations.

The inventory procedure will ensure that no SNM currently on hand (with the possible exception of holding account waste materials and residual holdup following cleanout that can be regarded as zero by definition) will be omitted and no quantity will be counted more than once.

The inventory procedures provide for confirming the presence of all items by direct observation and the presence of all quantities of SNM either by direct measurement or an acceptable alternative.

The proposed b

alternatives to remeasurement satisfy one of the following criteria:

J IB The SNM content is verified by statistical sampling and measurement of representative items, objects or samples of the material.

The 7

total overall sampling plan shall support the capability for detect-e ing any loss in excess of the pre established Detection Quantity with 5

90% (or better) probability.

The Detection Quantity will be estab-lished on a site-specific basis.

p The previous measurement results are accepted because assurance is 2

provided that the SNM content could not have changed since they were gj-'

made.

Examples of alternatives for satisfying this are use of 5-4 2

P

n

-i iw tamper-indicating seals (or encapsulation) or containment by storage in controlled access enclosures.

(See Question 1 of Section 5.6.)

]

Residual holdup after cleanout or draindown may be estimated if the

=

estimate is based on previously measured values, and it is periodic-ally verified or validated.

=

For material whose SNM content has been previously measured, and

$i there is no likelihood of any significant change in the uranium con-IF centration (or weight fraction) or in the uranium enrichment due to 1

such factors as oxidation, change in moisture content, commingling 6

with materials of different enrichment or different composition, sg etc., the previously determined uranium concentration and U-235 4

enrichment may be accepted without verification of SNM content g

provided:

1) the gross weight and/or net weight of all items within

=E the population is confirmed by (a) a 100% reweighing of all such items or (b) a reweighing sampling plan, or 2) access to the items n,

has been controlled.

g Any previously measured, but unsealed or unencapsulated SNM (that is on hand at the time of inventory) which is to be introduced into subsequent processing steps prior to inventory reconciliation should be remeasured or have its prior measurement value confirmed (by an acceptable alternative)

(

before the subsequent processing is initiated.

2 As an alternative to direct measurement (relative to establishing the 2

initial measurement or accountability value for a given item, batch, 2

etc.), the SNM concentrations and isotope factors used may be determined

=

from historical averages, controlled input specification values or empirical relationships where such values or relationships are periodically tested, and their uncertainties or bounds have been determined.

m As an additional alternative to remeasurement at inventory time, a program of routine process monitoring will be acceptable when the combination of the process monitoring program and the inventory procedures will achieve

=

the same level of loss detection capability as that provided by a physical y

inventory with all items either tamper-safed or remeasured.

5 The information in the Plan shows the detection threshold for an excessive ID will result in a 90% (or better) probability of detecting a discrepancy r

equal to or larger than the U-235 quantity established by the NRC on a site-specific basis.

(This quantity is referred to as the Detection Quan-d tity.)

In general, the licensee may assume the ID distribution approxi-mates a normal distribution, and therefore:

]9 Detection Threshold = Detection Quantity - 1.3 s q

where s is the SEID.

The estimated SEID value should be confirmed or updated annually.

Acceptable methodology for estimating the measurement 2

error contribution to the SEID by error propagation is found in Jaech R

(1973) and IAEA (1977).

Special attention is given to inclusion of all

]

1

-d 5

j s-s M

applicable and measurable sources of error to avoid underestimating the SEIO.

ids will be corrected (either directly, or indirectly by correcting the accounting records of individual items) for known accumulated measurement biases that are significant (as defined in Chapte, 4).

Whenever it is r

necessary to adjust an ID for significant biases, it will be the bias adjusted (i.e., corrected for bias) ID value that is used as the basis for detecting a loss equal to or greater than the site specific Detection Quantity.

The Detection Quantity for the plant will be proposed by each licensee and will be no greater than 1.3% of throughput for facilities involved in chemical processing, such as UFs conversion or scrap recovery, or 0.9% of throughput for facilities where material only undergoes physical changes, such as pressing U02 powder into pellets or loading pellets into fuel rods.

In either case, these quantities need be no smaller than 25 kilo-grams of U-235.

The throughput will be established from either historical experience or predicted future activity and throughput will be defined as the greater of " additions (of receipt material) to process" or " removals (of final product) from process" during a 12-month inventory period.

The resources and level of effort to be committed to the investigation of an excessive ID will be sufficient to reassess the results of the physical inventory, the accounting records, and the measurement control program data; to confirm the relevant calculations and data analysis; and to carry out searches for unmeasured inventory such as holdup and discards. The submittal shows the investigation will be completed within 60 days (except in extenuating circumstances) after beginning the physical inventory.

(See Question 6 of Section 5.6.)

The sample size of randomly selected items, representing a statistical sample of a given stratum that is to be verified on the basis of measure-ment of the selected items, should be determined by methods designed to detect a loss or removal of a substantial quantity from the stratum in question with a probability of 90% or greater. What constitutes a

" substantial quantity" may vary with each stratum (i.e., each class of items) and will be a function of the stratum size.

(See Question 5 of Section 5.6.)

5.6 QUESTIONS AND ANSWERS 1.

Q:

What kinds of protection will permit prior measurement data for items to be used at inventory time?

A:

Encapsulated items such as fuel rods and containers or process vessels protected with wire or paper seals provide adequate capa-bility to detect tampering.

It is not necessary that seals be accounted for if controls are applied to limit the use of the seals to designated individuals.

Storage of items in controlled areas will also protect against tampering.

The storage areas should be in locations where unauthorized removal of SNM from items would be 5-6

I obvious to other workers or where access is controlled and limited to a small number of authorized people.

2.

Q:

How well does the licensee have to resolve an excessive estimated ID?

A:

During the two-month resolution period, the licensee will investigate any excessive ID and should be able to reconcile it to a level neces-sary to satisfy the performance level stated in S74.31(c)(5).

If the licensee is unable to resolve the ID, then a full report of the investigation must be submitted giving a probable cause and a rationale for the cause or causes listed.

In addition, the licensee shall be prepared to perform a shutdown reinventory if requested by the NRC.

3.

Q:

Can you give an example of how the 90% power of detection can be calculated to satisfy S74.31(c)(5)?

A:

In general a normal distribution can be assumed for ID, and the estimated ID is compared against a detection threshold value where:

Detection Threshold = Detection Quantity - 1.3 (SEID).

If the esti-mated ID is larger than the detection threshold then resolution action is required.

For current levels of active inventory, most licensees should be able to show that they satisfy this condition with simple calculations which bound their capabilities rather than by making detailed exact calculations.

4.

Q:

In performing the annual inventory, may the presence of all items be determined by a statistical sampling method?

A:

No; 100% of the items must be accounted for by direct verification.

5.

Q:

Can you discuss the requirements of an acceptable inventory sampling plan or give an example?

A:

Generally, all that is required of items that must be verified for SNM content (i.e., those items not exempted from remeasurement by 74.31(c)(6)) is a reweighing of each item (for which the previous SNM measurement was based on weighing, sampling and analysis), or an NDA remeasurement (for those items originally measured by NDA).

The previously determined SNM content of all items reweighed can be accepted if there are no statistically significant discrepancies between original weight values and their corresponding reweighings.

For any reweighed items that reveal a weight discrepancy, a complete remeasurement based on weighing, sampling and analysis must be per-formed, and the remeasurement values (for those items showing weight discrepancies) will be used for inventory listing and accounting purposes.

In order to accept a mere reweighing as confirmation of SNM content, however, there must be a basis for concluding that there is no likelihood of a change in uranium element concentration or in enrichment.

In addition, the items to be verified can be chosen by statistical sampling with a sample size sufficient to ensure that the overall 5-7

system has at least a 90% probability of detecting a discrepancy equal to or greater than the pre-established detection goal.

Any one of several published sampling plans could be applied.

A variables sampling method for inventory verification is described by Stewart (1970), Hough et al. (1974), Piepel and Brouns (1981) and the IAEA (1977).

Piepel and Brouns also describe an inventory measurement method using a statistical sampling based on the method of Levy and Lemeshow (1980).

{

The verification methods involve estimation of the sample size required from an inventory to achieve a pre-established gcal standard deviation for the sum of the differences between the new (inspectors) and previous (operators) SNM measurements on the sampled items.

On the other hand, the inventory measurement method involves estimation of the sample size required to achieve a pre-established optimum goal standard error for the total inventory of the facility and an alloca-tion of the sample size to individual, uniform strata to achieve a minimum inventory standard error.

Verification methods based on an attribute test, which are somewhat simpler than the variables sampling methods, can also be used. One way to apply the method is as follows:

Group or stratify the inventory into fairly uniform item sizes.

Calculate a sampling fraction for each stratum by the method based on a binomial distribution with an acceptance number of zero; 9

y=1-p where:

n = number of items in the sample N = total number of items in the stratum

= the desired probability of not detecting a discrepancy of g/x items missing from the population N (for this case, p = 0.1) g = the pre-established detection goal for the inventory verification.

A reasonable value for g is the site-specific detection goal for ID times the total U-235 in the stratum divided by the total U-235 of active inventory.

i = the average SNM content of an item (if the range of item 5-8

sizes in a stratum is greater than about 25%, the largest item size should be used instead of the' average)

(The above formula gives conservative values for n.

A more optimum formula is:

n=(N-h)(1-$1/d) and d = g/x Note also that the detection goal may be applied independently to each stratum without loss of detection capability.)

Select the n items by a random selection procedure and weigh each item (or measure it by NDA if the previous measurement was by NDA).

Regard the difference between the remeasured and previous results dj = xjy i2

-x as a reject if d exceeds 4 times the standard deviation of d.

j If no rejects are detected in the n items, accept the previously measured values for all items in the stratum.

If one or more rejects are detected, investigate and reconcile the discrepancy, which may require remeasurements of both net weight and the element factor and correction of the records, and proceed with additional tests of the faulty stratum.

The additional tests may include further sampling (using a Stage 2 sampling plan) or remeasurement of all N items and reconciliation of all signifi-cant differences.

6.

Q:

Must the investigation of an excessive inventory difference be completed within 60 days after the start of the physical inventory?

A:

Generally yes; however, it is recognized that instances may arise where additional time is needed for resolution of an excessive ID.

In those instances, a request for an extension of time, including appropriate justification, should be forwarded to NRC ahead of the end date for the 60 day period.

An example of where this may occur would be a need to conduct a shutdown cleanout inventory.

5-9

j 7.

Q:

What are some examples of sources of variance that may increase the standard error of the inventory difference above that estimated from measurement control data?

A:

There are two principal sources of variance in ID:

the process variance and the measurement variance. The process variances may be caused by fluctuations in process holdups, fluctuations of unmeasured process losses, and operator errors in recording and transmitting data.

Examples of measurement variations not always accounted for in measurement control program data are between-laboratory effects, sampling errors caused by inadequate control of blending or changes in the sample before analysis, and uncorrected fluctuating biases.

Measurement variance can also be understated if operators and analysts improve the quality of their measurement performance when they knowingly are making control standard or replicate measurements, or if they (on their own initiative) perform extra control measurements and only report the better results.

i i

5-10

m...

f 6.0 ACCEPTANCE CRITERIA FOR ITEM CONTROL

' 6.1 RULE 74.31(c)(6)

Maintain current knowledge of items when the sum of the time of existence of an item, the time to make a record of the item, and the time necessary to locate the item exceeds 14 days.

Store and handle, or subsequently measure, items in a manner so that unauthorized removals of substantial quantities of material from items will be detected.

Exempted are items individually containing less than 500 grams of U-235 up to a total of 50 kilograms of U-235, solutions with a concentration of less than 5 grams of U-235 per liter, and items of waste destined for burial or incineration.

6.2 INTENT LThe intent of this.section is to require licensees to implement item control procedures that protect against unauthorized and unrecorded removal of material from items and that enable timely location of an item.

Items ara known quantities of SNM in well-defined containment such as canisters, or fixed units, e.g., fuel assemblies.

In order to promptly locate a given item, suffi-cient current information must be recorded.

These item control requirements apply only to items for which the total length of time that the item exists including the length of time it takes to locate it after having been recorded will exceed 14 days.

For example, if an item has been in existence for 10 days, the licensee must be capable of making a record of the item and locating the item within 4 days to be exempt from the current knowledge requirement.

Conversely, if it takes the licensee 4 days to make a record of the item and locate it, then an item which will exist for less than 10 days need not be formally entered into the current knowledge system.

Also in order to eliminate the need to keep records for insignificant quantities of SNM, waste containers and most solutions are exempt.

The licensee is permitted some flexibility in controlling. items by allowing exemption of up to 50 kilograms U-235 in items containing less than 500 grams U-235 each.

6.3 AFFIRMATIONS The licensee must make the following affirmations with respect to maintaining current knowledge of items and detecting unauthorized removals:

A record system will be established and maintained to provide a knowledge of the current status of items for all items for which the sum of the time of existence of the item and the time required to locate the item, includ-

-ing any time required to complete or update the records of the item, would exceed 14 days.

For items subject to this commitment the item control and records system will provide the capability to promptly locate and confirm the existence of any specific item or group of items upon demand.

The item record system will be secured in such a manner that the record of an 6-1

item's existence cannot be destroyed or falsified without a reasonable probability of detection.

Each item (unless it is an exempted item) will be stored and handled in a manner that enables detection of or provides protection against unauthor-ized or unrecorded removals of SNM; otherwise knowledge of the SNM content will be assured by seals or maintaining the item as a sealed source (i.e.,

as encapsulated material).

All incidents involving missing or compromised items or falsified item

. records will be investigated.

(A compromised item is one for which there is evidence of tampering or which is found outside its assigned controlled access area.)

The contents of a compromised item or an unsealed, unencapsulated item located after it has been missing will be reconfirmed by measurements (i.e., by NDA or by weighing, sampling and analysis).

6.4 INFORMATION TO BE INCLUDED To be complete, the Plan should contain the following information:

A description of the item control and accounting measures that provide capability to locate an item when required.

The estimated times and sup-porting rationale to complete and update the item control records and to complete the major steps for locating items should be given in the Annex.

A description of the storage and handling measures to be used to protect against or detect unauthorized removals of SNM from an item.

A description of the item record system showing how items are identified for each material type and/or each type of container, what information is recorded, the mechanism for entering source data in the system, how changes in location or the destruction of items are recorded, and how protection against simultaneous removal of an item and all records of that item is provided.

The types of items and categories or classes of material in items that the licensee proposes to exempt from the item control requirements and the rationale for such exemptions.

A description of the controls that assure that any items originally expected to meet the 14-day exemption criteria, but later fail to meet it, are promptly entered into the item control record system.

A brief description of the approach to ensure that the combined quantities of U-235 in exempted items each containing less than 500 grams of U-235 will remain under the total allowance of 50 kilograms of U-235.

J 6-2

6.5 ACCEPTANCE CRITERIA A judgment that the licensee's Plan for the control of items subject to the requirements of S74.31(c)(6) is acceptable will be based on the following criteria:

The licensee's planned item record system will uniquely identify items.

The records will include information on the chemical form, quantity of material (element and/or isotope), physical description, identification label or number, and location.

The system will provide reasonable assur-ance of detecting falsification or destruction of records of an item's existence. Groups of items that are produced, stored, processed, or otherwise handled together as a unit, such as a batch or sublot of mate-rial, may be exempt from the requirement for identification of each item if the group of items will be uniquely identified and stored as a separate group under conditions such that group identity, composition and quantity will be maintained constant.

The record of the status of an item can be completed or updated in suffi-cient time to allow the licensee to meet the requirements for promptly locating an item.

The licensee has shown in the Annex that the location information to be kept on an item can provide the capability to locate and confirm the presence (if requested by the NRC) of (1) any specifically identified item within 3 calendar days, or (2) for one or more items from a group of items within 14 calendar days when identified by only one of its attributes (chemical or physical form, physical description, quantity of SNM, loca-tion or item number).

For items that will not be remeasured at inventory time, the item control procedures will provide reasonable assurance that the SNM contents given in the records are valid and that unauthorized removal of SNM from the item has not occurred.

Remeasurement is not needed if the SNM content of the item was measured previously and reasonable assurance is provided that the SNM content has not subsequently changed.

How such assurance can be provided is discussed in Section 5.5 and 5.6.

The previous measurement of the SNM content of every item will be made directly, but the element and isotope fractions may be determined indirectly or by sampling procedures including:

Analysis of composites or measurements of representative items, objects or samples selected by statistical sampling; or Use of concentration and enrichment factors determined from histori-cal averages, controlled input specification values or empirical relationships where such values or relationships are periodically tested, their uncertainties or bounds have been determined to be within about 2% of the factor value, and where diversions with mate-rial substitution are improbable.

However, heterogeneous materials, such as ammonium diuranate, may not be assigned common factors unless the quantities are small, such as less than about 500 g of contained U-235 (per material type, per inventory period).

Justifications for any materials assigned common factors without batch-by-batch verifi-cation analyses should be presented in the Annex.

6-3

The Plan makes provisions for determining the total quantity of U-235 in items exempted from control on the basis that they contain less than 500 grams U-235 each.

Items exempted from item control procedures will satisfy one of the following conditions:

Items that will be in existence for only a short period of time, as defined in $74.31(c)(6);

Waste destined for burial or incineration; Solutions containing less than 5 grams U-235 per liter; or Items individually containing less than 500 grams U-235 each up to a total of 50 kilograms U-235.

A current accounting will be maintained of the total quantity of SNM con-tained in items that are exempted from item control.

The accounts will identify the quantities by material type category and separate listing by material type category of controlled and exempted material will be feasible.

6.6 QUESTIONS AND ANSWERS 1.

Q:

What is an item?

A:

An item is created any time a finite quantity of SNM is contained in a fixed volume.

All items, however, do not necessarily need to be included in the current knowledge requirement.

Items which exist for a short period and will soon be destroyed for further processing may not need to be formally recorded.

Items that are going to be in existence for less than 14 days and which can be located promptly when necessary are exempt from the current knowledge requirement.

However, if the time to update the item records plus the time to locate the item combined with the time since the item was created exceeds 14 days, the formal item controls must be applied.

In addi-tion if an " exempted" item is held longer than expected and it become's evident that the 14-day limit will be exceeded, the item must be reclassified as a " controlled" item.

2.

Q:

Do items have to be tamper-safed according to Regulatory Guide 5.15?

A:

No.

Tamper-safing is not required as long as a valid inventory is taken with reasonable assurance that the value previously assigned to an item is still correct.

For items where it may be questionable whether the item is still intact, such as items sitting around for a long period of time, alternative measures to tamper-safing may be used (e.g., increased surveillance of ite_ms, seals which do not necessarily satisfy Regulatory Guide 5.15, and controlling access to items).

6-4

f 3.

Q:

Why is there a limit on the item control system exemption?

A:

All waste destined for burial or incineration and solutions containing less than 5 g/l of U-235 are exempted from the item control system without a limit.

The limit only applies to other items individually containing less than 500 grams of U-235.

This was intended to allow the licensee to have flexibility in exempting unattractive theft targets.

The type of items exempted will depend upon a licensee's particular operation.

However, to assure that a large portion of the licensee's inventory is not exempted, an overall limit was established.

4.

Q:

Does the requirement for maintaining current knowledge of items apply when a physical inventory is being taken?

A:

Yes.

The current knowledge capability should be able to assist in the assessment of an alleged material theft where such an allegation could be received at any time.

5.

Q:

How well does current knowledge have to be assured?

A:

Two of the criteria for acceptance of a licensee's item control plan will apply.

First, item records must be kept separate'from the item so that an unauthorized removal would not go undetected.

Thus, while totally separate centralized records are not required, the items and records should be stored such that it would be difficult to remove an item and all records of that item or to remove a record and all items on that record.

Second, items must be stored and handled in a manner such that unauthorized removals of SNM from them are difficult or are likely to be detected.

Items of SNM where the current material content may be subject to question should be remeasured for SNM content at least at inventory time.

Examples of such material are:

1) material stored in an area where unexplained discrepancies have occurred, 2) material 6hich has received no control during the inventory period, and

3) material in a container with a broken seal.

6-5 i

]

7.0 ACCEPTANCE CRITERIA FOR RESOLVING SHIPPER / RECEIVER DIFFERENCES 7.1 RULE 74.31(c)(7)

Resolve, on a shipment basis and, when required to satisfy Part 75 of this chapter, on a batch basis, shipper / receiver differences that exceed both twice the combined measurement standard error for that shipment and 500 grams of U-235.

7. 2 INTENT Tae intent of this section is to require the material control and account-ing sy. stem to promptly detect and resolve all significant shipper / receiver differences. The shipper / receiver value is important in determining when shippers' and receivers' values are acceptable for establishing the book inven-tory of the facility.

7.3 AFFIRMATIONS The licensee must make the following affirmations with respect to resolving shipper / receiver differences:

A capability will be provided for timely detection of significant discrepancies in SNM shipments.

Etch shipment received will be inspected for loss or damage to the ccntainer or seals to determine if SNM could have been removed.

If the irtegrity of a shipping container is questionable, the presence of all items that were packaged in the container will be verified.

Confirmatory measurements of the quantity of SNM received in each shipment will be performed and the shipper / receiver difference will be tested for statistical significance.

Occurrences of significant shipper / receiver differences and missing items will be reported to the shipper promptly.

Measurement results for shipments and receipts will be corrected for biases that are significant at the 0.05 level (i.e., for any bias that exceeds two times the standard error associated with a mean).

A significant shipper / receiver difference, e.g., one that exceeds both twice the combined measurement standard deviation for that shipment and 500 grams of U-235, will be promptly investigated and resolved on a ship-meat basis or on a batch basis when required to satisfy 10 CFR 75.4(d).

3ignificant shipper / receiver differences will be reported to the NRC either as being unresolved or with the conclusions of the resolution procedures.

7-1

7. 4 INFORMATION TO BE INCLUDED To be complete, the Plan should contain the following information:

A description of the licensee's plans for detecting and investigating significant shipper / receiver differences, including an estimate of the time required to complete the investigation.

The Annex should describe the procedures for inspecting shipments, and the measurement methods and their uncertainties that will be used to confirm SNM in the shipment.

A description of the statistical test of hypothesis for determining if a shipper / receiver difference is significant.

An example should be given in the Annex showing a typical calculation and statistical test to determine if a shipper / receiver difference is significant.

7.5 ACCEPTANCE CRITERIA A judgment that the licensee's Plan for resolving shipper / receiver differ-ences is acceptable will be based on the following criteria:

Each shipping container will be inspected within 3 working days after receipt.

If the integrity of a container is questionable, the presence of all items that were packaged in the shipping container will be confirmed within 3 working days after receipt.

Confirmatory measurements of scrap shipments will be performed by the receiver to determine the amount of U-235 received within a time period consistent with the accountability needs of the shipper.

The test for significance of a shipper / receiver difference will be based on hypothesis tests.

The planned investigation procedure for significant shipper / receiver differences is sufficiently cc:aprehensive to ensure that the difference will be resolved.

Comprehensiveness is sufficient if the licensee shows the capability to verify records, resample, perform remeasurements, estab-lish liaison with the shipper, provide samples to a referee laboratory, and perform the statistical analysis needed to evaluate the measurements.

For those cases where the shipper's limit of error data is not available, such limit of error should be assumed as being equal to that of the receiver for the purpose of calculating the combined mersurement standard error.

7.6 QUESTIONS AND ANSWERS q

1.

Q:

Where is a " batch" defined?

A:

The definition for a batch is given in 10 CFR 75.4(d).

2.

Q:

Does " resolving shipper / receiver differences on a shipment basis" mean per DOE /NRC Form-741 or per truckload?

A:

A shipment basis means per D0E/NRC Form-741.

i 7-2

3.

Q:

What constitutes resolution of a shipper / receiver difference?

A:

A shipper / receiver difference is resolved _when the shipper and the receiver agree on the quantity of SNM transferred in the shipment.

The value agreed upon is then the accountability value used by both the shipper and receiver.

In the event that the parties do not reach an agreement, the NRC must be notified.

4.

Q:

What is the basis for accountability measurements by a scrap recovery plant?

A:

Scrap recovery plants carry scrap receipts at shipper's values until they have dissolved the scrap and can sample at their accountability tank. The accountability value is the sum of the SNM in solution and in residues.

5.

Q:

Where does one obtain guidance on hypothesis tests for shipper / receiver differences?

A:

Discussion of shipper / receiver difference tests is found in ANSI Standard N15.17 (1974); Jaech (1973); Johnston, Brouns and Stewart (1982), and Regulatory Guide 5.28 (1974).

l i

7-3

8.0 ACCEPTANCE CRITERIA FOR PERIODIC ASSESSMENT OF THE MC&A SYSTEM 8.1 RULE 74.31(c)(8).

Independently assess the effectiveness of the material control and accounting system at least every 24 months, and document management's action on prior assessment recommendations.

8.2 INTENT The intent of this section is to require management to periodically review the performance of the MC&A system and evaluate its overall effectiveness.

It is intended that the reviews will be performed by knowledgeable, technically competent individuals free from conflicts of interest and that the deficiencies

~

will be brought to the attention of the plant management so that the deficien-cies will be corrected.

8.3 AFFIRMATIONS The licensee must make the following affirmations with respect to inde-pendent assessment and evaluation of the MC&A system effectiveness:

r The capabilities and performance of the MC&A system will be reviewed and its effectiveness will be independently assessed at least every 24 months.

That is, the nominal elapsed time from the completion of one review /

assessment to the completion of the next will not exceed 24 calendar months.

The review and assessment will be performed either by qualified individuals from outside or qualified individuals from inside the facility l

organization whose work assignments and positions within the organization

-will not impair their ability to make objective judgments of the MC&A l

system capabilities and performance.

Each overall review and assessment will be conducted and completed in a time frame that is short with respect to the time for changes to have occurred in the MC&A system and will include any such changes made during l

the time the review / assessment is being conducted.

The completion date for any given review / assessment is defined as the date l

when the team submits its final written report (of findings and recommen-dations) to plant management. The start sfate is the first day in which i

l one or more team members actually inspect records and/or interview MC&A personnel, and such start date will be documented.

The assessment team leader will have no responsibility for managing or performing any of the MC&A system functions.

The results of the assessment and recommendations for corrective action, if any, will be documented and reported to the plant manager and other I

8-1 1

managers affected by the assessment. Management will review the assessment report and take the necessary actions to correct MC&A system deficiencies.

Such corrective actions (if any) that pertain to daily or weekly activities will be initiated within 40 calendar days following the submittal of the review / assessment final report.

Management's response to recommendations from the review and assessment, including any corrective actions ordered by management and the expected time frame for completing such actions, will be documented within 30 days following the submittal of the team's report.

8.4 INFORMATION TO BE INCLUDED To be complete, the Plan should contain the following information with respect to independent assessments:

A description of the plans for conducting the independent assessments of the MC&A system including:

A statement of the objectives of the assessment program; The scope of the reviews to be performed giving the areas of the MC&A system to be covered and the level of detail required; How the audit will be conducted and the areas that will be treated by observation; testing; review of data, procedures or records, etc; The policies for selecting the assessment team listing the required qualifications, restrictions and limitations to be applied in order to assure independence and objectivity of the team; and A description of the sampling plan used for selecting records to be audited as part of the assessment.

The organizational positions responsible for initiating the assessment, approving the membership of the assessment team, and implementing the recommendations of the assessment.

The-Annex will include a typical assessment schedule showing time and effort to be devoted to different areas to indicate the expected intensity L

of the assessment.

i 8.5 ACCEPTANCE CRITERIA d judgment that the licensee's program for independent assessment of the MC&A system is acceptable will be based on the following criteria:

The periodic assessments will be comprehensive and sufficiently detailed

~

to enable the assessment team to rate the MC&A system effectiveness, capability, and performance by comparison with the expected and required performance. The overall objectives will be to determine that the MC&A l

system, as designed and implemented, is continuing to meet the overall 8-2

i

. safeguards objectives and to identify weaknesses or deficiencies in the system design or performance that may need correcting.

The areas to be reviewed will encompass the entire MC&A system and the level of detail of the reviews will be sufficient to ensure that the assessment team has adequate information to make reasoned judgments of the MC&A system effectiveness which includes:

Organizational effectiveness and management responsiveness to possible SNM losses; Staff training and competency to carry out MC&A functions; Soundness of the material accounting records; Capability to promptly locate items; Timeliness and effectiveness of shipper / receiver difference ~evalua-tions and resolution of excessive shipper / receiver differences; Soundness of the inventory taking procedure; Effectiveness of the measurement control program in monitoring the key measurement systems and its sufficiency to meet the requirements for controlling standard error of the ID; Capability to confirm the presence of special nuclear material; and Capability to resolve indications of missing material and aid in any investigations pertaining to missing SNM, and provide information that would aid in the recovery of missing material.

Every record or report need not be examined, but, by using generally accepted auditing principles, representative samples of each type of record and report must be checked. The selection of samples of the records to be examined should be random.

l Personnel assigned to the assessment team will have an understanding of l

the objectives and the requirements of the MC&A system and will have l

sufficient knowledge and experience to be able to judge the adequacy of the parts of the system they are asked to review. The team will have authority to investigate any aspect of the MC&A system and will be given i

access to all information needed to do this.

l The team members may be selected from the facility staff or from outside l

the facility, but an individual member will not participate in the assess-i l

ment of the parts of the MC&A system for which that person has direct responsibility. Also, an individual will not assess the parts of the system that are the responsibility of team members assessing his or her l

area.

The entire MC&A system will be reviewed and evaluted during each single assessment (to be completed within an elapsed time that is short relative 8-3

i I

l to the time between changes in the MC&A system and is demonstrated to be able to include any such changes made during the review / assessment).

Cenducting two or more assessments during a 24-month interval, in which only part of.the MC&A system is covered in each, is not deemed acceptable.

That is, reviewing a single component of the MC&A system at 24-month intervals is not very meaningful unless a knowledge is obtained of how well the other components interact.

Piecemeal review and evaluation is permissible, however, if each subsystem is covered at 12-month (or less) intervals.

The leader of the assessment team will have no responsibilities for performing or managing the functions being assessed.

The responsibility and authority for the assessment program and for initiating corrective actions will lie:

1) at least one level higher in the organization than the MC&A manager, or 2) the on-site plant manager.

8.6 0UESTION AND ANSWER 1.

Q:

What should be included in the assessment of each key MC&A area?

A:

The assessment should include an independent review of the MC&A

-system in each key area which provides information concerning the soundness, adequacy, and actual application of material controls and assurance of satisfactory accounting procedures and practices. The assessment should' emphasize the system's performance and effective-ness rather than technical compliance with the Plan. The reliability of data and information within the MC&A system should be evaluated and assessment made of the quality of the source data and records.

The assessment should include recomputation of quantities, tracing of selected transactions through the accounting process, and a thorough review of corrections and adjustments.

The responsibilities and procedures for corrective action should be defined and the response of management to the recommendations resulting from the assessment should be documented. The assessment must also encompass SNM accountability measurements and measurement control performed by a contractor (or off-site) laboratory.

l*

2.

Q:

Why does the criteria refer to assessment teams, as opposed to an I

individual assessor?

i A:

In general a team of individuals with expertise in the many varying aspects of MC&A would be essential to an effective audit / review.

i However, if a single individual possessed all of the necessary quali-fications and would not be subject to conflict of interest situations, such an individual would be deemed acceptable for performing the audit / review.

l l

8-4

9.0 ACCEPTANCE CRITERIA FOR RECORDKEEPING 9.1 RULE 74.31(d)- Recordteeping.

Each licensee shall establish records that will demonstrate that the requirements of paragraph (c) of this section have been met and maintain these records for at least 3 years unless a longer retention time is required by Part 75 of this chapter.

9.2 INTENT The intent of this section is to require the establishment, maintenance, and protection of a recordkeeping system that will demonstrate that the system capability requirements described in S74.31(c) have been met.

Records are to be retained for at least 3 years thereby providing a means for assessing the performance of the MC&A system and inspecting for compliance with regulations.

9.3 AFFIRMATIONS-The licensee must make the following affirmations with respect to recordkeeping:

A record system will be established and those records necessary to show that the MC8 A system requirements of S74.31(c) have been met will be main-

'tained for tt least 3 years. The records referred to in $$75.22 and 75.23 and generated while the licensee was under IAEA safeguards will be retained for at least five years by the licensee.

Records of the

-following w 11 be maintained current and will be retained for at least three years:

l Management structure, MC&A job descriptions, and MC&A policies and procedJres; Accounting source data records (a);

Records of shipments and receipts and investigations of significant shipper / receiver differences plus the information used to resolve them; Measurement data for receipts, shipments, discards, and inventory; 1

)

i:

(a) The account.ing source data normally consist of shipping and receiving forms, physical inventory forms, and the forms used for initially recording neasurement and measurement control data.

After an item is destroyed, the item location record needs to be retained for an additional 14 days but then may be destroyed.

l l

l 9-1

L l

Calibrations of measurement systems, measurement control data, bias estimates, and the statistical analyses of the measurement control data;.

Data used to demonstrate that measurement system performance achieves the standard deviation limits required by $74.31(c)(4);

)

_ Physical inventory listings and inventory work sheets; Calculations'of detection thresholds for excessive ID; Calculations of the standard error of the estimated ID and information used to reconcile excessive ids; Reports of investigations and resolution of indications of loss of

-~

s SNM;~and i'

The results of independent assessments and management action taken to j '

i-

correct.any deficiencies identified.

Sufficient protection and redundancy of the record system will be provided so that an act of record alteration or destruction will not eliminate the capability to provide a complete and correct set of SNM control and accounting information that could be used to confirm the presence of SNM, resolve indications of missing material, or aid in the investigation and j

recovery of missing material.

- Ready traceability will.be provided for all SNM transactions from source data to final accounting records.

9.'4 INFORMATION TO BE INCLUDED j

To be complete, the Plan should contain the following information:

A general description of the MC&A record system giving the recordkeeping l

-policies and defining the~ data and information to be routinely recorded, which documents need authorizing signatures, and the kinds of records to-be retained in order to meet the recordkeeping requirements.

The Annex should include a flow chart showing how data flows from the source docu-ments to final accounting reports and typical forms and report formats used throughout the MC&A system.

l

.A description of any records system design features that facilitate i

traceability and enable auditing.

L

- A description of the overchecks for preventing or detecting missing or I

falsified data and records, ensuring completeness of the records, and y

locating data discrepancies and errors.

The plan for reconstructing lost or destroyed SNM records.

A description of any access controls used to assure that only authorized persons can update and correct records.

L 9-2 i

l t

A description of record accessibility.

9.5 ACCEPTANCE CRITERIA A fudgment that the licensee's Plan for recordkeeping is acceptable will be based on the following criteria:

The submittal shows that key material accounting and original source data documents and relevant reports and documents will be retained for three years or as long as needed to show continuing compliance with 10 CFR Part 74.

(For example, records of the organizational structure cannot be destroyed if the current structure is still the same.)

The source data will be retained in its original form until the physical inventory and any subsequent ID investigations have been completed.

After this time, any readable facsimile is acceptable for the remainder of the required retention period.

All other records may be retained as hard copy, microfiche, permanent computer readable forms, or other permanently readable forms.

The retained records and reports will contain sufficient detail t'o enable NRC inspectors to determine that the control and accounting of the SNM has complied with the requirements of S74.31(c) and has met the general per-formance objectives of S74.31(a).

The records will be retrievable, sufficiently complete and detailed to permit auditing all parts of the MC&A system, and traceable back to origi-nal source data.

The records of the data that is the basis of the estimated SEID will permit traceability to the sources of the variances due to calibrations, bias adjustments and random effects in the measurements. These records may be summaries of calibrations, bias tests, and variance monitoring data or control charts.

The record system will have sufficient redundancy to enable reconstruction of lost or missing records so that knowledge of the SNM inventory is always available.

The primary records, as contrasted with duplicate or backup records, will be provided security against computer memory failure (dump), fire or water damage, vandalism, and access by unauthorized persons.

All retained MC&A records are to be readily accessible, in order to meet I

1 time restraints relative to their use.

Snme examples are given.

In general, the record retention system should possess the capability to retrieve records used for measurement control or accountability within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> if the record was generated within the past 12 months, and within 7 calendar days if generated more than 12 months ago.

Physical inventory listings should be available within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for the latest two inventories.

Item control records should be retrievable in time to satisfy the criteria in section 6.5 9-3

i I

i Overchecks or other controls including access controls for updating and correcting records will be provided to prevent or detect errors in the records that would affect inventory difference and item location.

9.6 QUESTIONS AND ANSWERS 1.

Q:

What is an acceptable record?

A:

The following are characteristics of an acceptable record:

Permanence for the applicable record keeping period; The capability to determine the location of all pertinent data; Retrievability in a time period necessary to fulfill the capabilities; Readability of retrieved data; and Traceability to provide an audit trail.

2.

Q:

What are some examples of overchecks or controls that would protect inventory difference or item location records from errors or losses?

A:

Some examples are:

Minimize the number of people authorized to make data entries; Use verification methods for data entry for shipments, receipts, waste discards, and item records (item records may be verified by random sampling rather than on a 100 percent basis if the error rate is low);

Overcheck calculations, at least by random sampling; Use double entry bookkeeping; Retain duplicate records in a separate secure location so that a single individual or a single event cannot alter both sets of records; and Use two person inventory teams.

9-4

REFERENCES ANSI N15.17.

1974Property "ANSI code" (as page type) with input value "ANSI N15.17.</br></br>1974" contains invalid characters or is incomplete and therefore can cause unexpected results during a query or annotation process.. Concepts and Principles for the Statistical Evaluation of Shipper-Receiver Differences in the Transfer of Special Nuclear Material.

American National Standards Institute, New York, NY.

IAEA.

1977.

IAEA Safeguards Technical Manual, Part F - Statistical Concepts and Techniques.

IAEA-174, International Atomic Energy Agency, Vienna, Austria.

Jaech, J. L.

1973. Statistical Methods in Nuclear Material Control.

TID-26298. National Technical Information Service, Springfield, VA.

Johnston, J. W., R. J. Brouns and K. B. Stewart.

1982. Resolution of Shipper-Receiver Differences.

NUREG/CR-2820.

Pacific Northwest Laboratory, Richland, WA.

Levy, P. S., and S. Lemeshow.

1980.

Sampling for Health Professionals.

Wadsworth, Inc., Belmont, CA.

Piepel, G. F. and R. J. Brouns.

1982. Stati:tical Sampling Plans for Prior Measurement Verification and Determination of the SNM Content of Inventories.

NUREG/CR-2466 (PNL-4057). Pacific Northwest Laboratory, Richland, WA.

Regulatory Guide 5.15.

1974.

Security Seals for the Protection and Control of Special Nuclear Materials. Nuclear Regulatory Commission, Washington, D.C.

Regulatory Guide 5.28.

1974.

Evaluation of Shipper-Receiver Differences in the Transfer of Special Nuclear Materials. Nuclear Regulatory Commission, Washington, D.C.

Reilly, T. D. and H. L. Evans.

1977. Measurement Reliability for Nuclear Material Assay.

LA-6574.

Los Alamos Scientific Laboratory, Los Alamos, NM.

Rogers, D. R., (ed.).

1983. Handbook of Nuclear Safeguards Measurement Methods. NUREG/CR-2078. Monsanto Research Corp., Mound Laboratories, l

Miamisburg, OH.

l l

Stewart, K. B.

1970.

"A Cost Effectiveness Approach to Inventory Verification."

, Safeguards Techniques. Vol. II, pp. 382-410:

Proceedings of a Symposium, Karlsruhe, 6-10 July 1970.

IAEA-SM-133/59, IAEA, Vienna, Austria.

Tingey, F. H.

1983. Estimating and Applying Bias Correction to Nuclear Material Accounting Data, NUSAC Report 877, NUSAC Inc., Reston, Va.

l Tingey, F. H., C. J. Barnhart and R. F. Lumb.

1983.

Resolving the Components l

of Process Variability and Estimating the Uncertainty of the LEID.

NUSAC Report No. 752, Rev. 1.

NUSAC, Inc., Reston, VA.

l Tingey, F. H., and R. F. Lumb.

1981. Mathematical Derivation of tim Automated Material Accounting Statistics System (AMASS), NUSAC Report 591, NUSAC, Inc.

Reston, Va.

R-1

• U.S. CCVERNMENT PRINTING OFFICEe 1985.W1 721 e 20059 t

AIRC PO#es 35 U.S. NUCLE 1W A E LULLTORY COMMISSION I AEFORT NUMBER fAss gned er T/DC edd Vet No. sf aard REG-1065 3$

BIBLIOGRAPHIC DATA SHEET Revision 1 SEE nNsTRUCTIONS ON TMt RtytRSE

3. TITLE AND $uSTITLE 3 LE Avt 9LANIL Accept ce Criteria for the Low Enriched Uranium f

Reform endments

,o,, f o,,Co.,L,T,o f

l

, EAR MONm February /

1985

,,.u f,,OR,,,

O ATE REPORT ISSUED C. W. Emei

, G. E. Gundersen, C. J. Withee MOy'.,

j vEAa April' 1985

7. Pt ~FORMcNG ORGANIZ AT NAME AND MAILING ADDRE55 ff**cr dele Come; T

u Division of Sa uards Office of Nuclea aterial Safety and Safeguards U.S. Nuclear Regul' ory Commission Washington, DC 205 to SPONSORING ORGANiiAT,04 NAME AND W NG ADORE 55 (facsweele Codes ita YVPE OF REPORY Same as item #7 Technical e PERIOD COVERED flac'es*e 8erest 13 SUPPLEMtNT ARv NOTES I3 AS$TR ACT (200 e.res or wsJ This report documents a standard fo t su ested by the NRC for use in preparing fundamental nuclear material control an as required by the Low Enriched Uranium Reform Amendments (portions of 10 CFR 74). The report also describes the necessary contents of a comprehensive p' 3 and provides example acceptance criteria which are intended to comunicate acce a'

means of achieving the performance capabilities of the Reform Amendments By ing the suggested fonnat, the license applicant will minimize administrati prob s asscciated with the submittal, review and approval of the FNMC plan. Pre ration o the plan in accordance with this format will assist the NRC in evaluating t e plan an standardizing the review and licen-sing process. However, conformanc with this 9 ance is not required by the NRC.

A license applicant who employs a ormat that pro des an equal level of completeness and detail may use their own fo. t.

4 j g g y g, gag;giy y 14 DOCUMENT ANakvsil - ate vwORDS DESCRiPTOR$

Reform Amendments 4(

Unlimited Low Enriched Uranium Special Nuclear Material j

LowStrategicSignificance[

' 5 " c'a'" c $$ 'c^'*N EricTassified

..DiNTi.. R zo,iN

~0 0 NRM.

L Thrs !.00rt)

Unclassified

)

17 NuYSER OF PAGES 18 9R>Ct

UNITED STATES '

FOU2TH CLASS MAIL NUCLEA3 RE'IULATCRY CIMMISSISN P 8'^GE,6 FEE 3 PilD u

WASHINGTON, D.C. 20566 wasa o.c.

m PERMIT No. G 47,

O L

- OFFICIAL BUSINESS-OL

' PENALTY FOR PRIVATE USE, 8300

@~

y.

~O Om

.m "ip.-

-Z-Om O3

-4 m

3-D-

m O

2

-4 2m r"O&

nZ 3:

OZm O

-C-2D Z

C

'E mm

'1 O'

m.

E E,

miZ l OEm; Z'

-l in

>l 3

r- '

(D a

b..