ML20236B698

Insp Rept 70-0371/88-12 on 881205-09.Violations Noted.Major Areas Inspected:Measurement Control & Statistics & Followup on inspector-identified Items

ML20236B698
Person / Time
Site:	07000371
Issue date:	01/27/1989
From:	David Jones, Keimig R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20236B660	List: ML20236B657 ML20236B693 ML20236B698
References
70-0371-88-12, 70-371-88-12, NUDOCS 8903210232
Download: ML20236B698 (7)

Text

_ _ _ _ _ _ - --. _

1 4 . .

3

U.S. NUCLEAR REGULATCRY COMMISSION REGION I

.l Report No: 70-371/88-12 Licensee: UNC, Incorporated

'UNC Naval Products 67 Sandy Desert Road Uncasville. Connecticut 06382 Docket No: 70-371 License No: SNM-368 Facility Name: UNC Naval Products '

Inspection Conducted: December 5-9, 1988 Inspector: .. For i/27k9 D. Jones, Region II Statistician D&te S/igned Approved by: .k. ejl mg R.' R. Keimig,- ChTef, Sifeguards-1/27!K9 Date' Signed Section, Division of Radiation Safety and Safeguards

SUMMARY

Scope: The areas examined during this routine _ unannounced inspection included measurement control and statistics, and follow-up on ,

inspector identified items.

Results: Within the areas inspected, the following violations were identified:

Failure to establish written material control and accounting procedures, paragraph 2, Failure to correctly calculate additions to and removals from material in process, paragraph 2, Failure to include the lab waste analytical measurement system in the measurement control program, paragraph 2. '

8903210232 890130 P, PDR ADOCK 07000371 p' .

,- C PDC I.

} ,

b' DETAILS

1. Persons' Contacted Licensee Employees

• T.:Collopy, Manager, Nuclear Material Control

• R. Gregg, Director, Technical Services

• M. Kelly, Statistical Specialist

• J. L'Heureux,' Specialist, Nuclear Material Control S. Ververis, Supervisor, Chemistry Laboratory The inspectors .also interviewed several other licensee employees.

.NRC Personnel

• A. Della Ratta, Safeguards Auditor, Region I

• G. Gundersen, Safeguards Physical Scientist

~*P. Ting, Chief, MC&A and PS Section, ONMSS

.*M.-Williams, Safeguards Statistician, ONMSS

• Attended exit interview .

2. Measurement Control and Statistics (85209) j The licensee's statistical control programs for calculating inventory difference (ID) and its associated limit of error, (LEID) and for.

monitoring _the quality of SNM accounting measurements were inspected. The inspection consisted of an examination of the licensee's written procedures for performing the statistical calculations, the data bases used to evaluate the statistical parameters, and the statistical methods  ;

applied in the evaluation of the data bases. The performance of the programs during the July 15 through September 16, 1988 material balance period was inspected.

.a. Procedure Review-The licensee's procedures for performing LEID calculations (NMC-005), I monitoring measurement performance (NMC-003) and evaluating shipper-receiver differences (NMC-007) had not been approved by written signature of the Nuclear Material Control Manager. These were three additional examples of the apparent violation issued in Inspection Report 70-371/88-10 for the inspection conducted on October 11-14, 1988. The-licensee verbally committed to have these three procedures

. approved by written signature, as well as the procedure (NMC-004) identified in the above referenced apparent violation, by March 15, 1989. This issue will be tracked as an inspector follow-up item (88-12-01). The review of these three procedures revealed that they 1

~

~

1 I

contained narrative descriptions of the statistical calculations rather than the mathematical models of the statistical parameters )

being evaluated. During the inspection the licensee verbally committed to revise the procedures, by March 15, 1989, to include the mathematical models of the statistical parameters. This issue wil.1 be tracked as an inspector follow-up item (88-12-02).

Pursuant to 10 CFR 70.51(c) the licensee is required to establish, f maintain, and follow written material control and accounting '

procedures. Three examples of the licensee's failure to comply with this requirement were identified. (1) A procedure for calculating additions to and removals from " material in process" had not been established. (2) The licensee's procedure (NMC-003) for monitoring the quality of SNM accounting measurements did not address the establishment of controls for replicate measurements of process materials. (3) A procedure had not been established for responding to occurrences of control data exceeding the control limits for analytical measurement systems. This is an apparent violation of NRC

]

requirements (88-12-03), j

b. Inventory Differences and Limit of Error The licensee's calculations of the inventory difference (ID) and its associated limit of error (LEID) for the July 15 through '.

September 16, 1988 material balance period were inspected. The inspection consisted of an audit of the material balance summary, a review of the mathematical model of the LEID, and an examination of the coefficients used in the model to propagate measurement error variances through the material balance statement. The material quantities of beginning inventory, receipts, shipments, ending inventory and their adjustments, as reported on the SNM Physical Inventory Report (NRC Form 327), were compared to the licensee's official accountability records and were substantiated. Pursuant to 10 CFR 70.51(e)(4)(iv), the licensee had calculated the additions to and removals from material in process during the material balance period. However, those calculations were found to have been performed incorrectly in that the quantities of material included in those calculations did not meet the definitions of " additions to material in process" and " removals from material in process" as specified in 10 CFR 70.51(a). In the licensee's nomenclature, unused fuel means finished fuel which was taken from the storage vault to the processing area during the morning but not used and returned to the vault in the afternoon for overnight storage. The licensee included the quantities of unused fuel in the calculations  !

for both additions to and removals from material in process. This resulted in both quantities, additions and removals, being overstated. Also the calculation of removals from material ir.

process included the quantity of fuel charges generated rather than ,

the quantity of elements produced. The licensee's failure to j correctly calculate the additions to and removals from material in 1

m3 r , ,

a c 1 p

~

ll , ,

e

, 3 ,

4

[

1 l ,

q i

process is 'an ' apparent. violation' of NRC requirements'(88-12-04), {

y'

^

Duringethe inspection the licensee verbally committed to submit, by l February.1,1989 a revision of the Fundamental Nuclear Materiak ]

Control (FNMC) Plan which will incorporate plant-specific definitions {

of the quantities of material to_be included in the. calculations of )

both additions to and removals from material in process. The ',

. licensee also indicated that, subsequent to NRC approval of the FNMCe  !

Plan revision, the approved definitions will be incorporated into a-written procedure for performing the calculations..This issue.will be tracked as an inspector = follow-up item (88-12-05).

The licensee's calculations of the coefficients, i.e. the material.

s quantities used in the LEID calculations,.for the following measurement systems were inspected: weight and elemental analysis s

of finished fuel, and nondestr'uctive' analysis of scrap.and waste materials (sludge, chips, rags, and filters). The coefficients used ,

for the scrap.and w'aste materials were found to be accurate but the coeff.icients used for finished fuel weight and elemental analysis' were found to have been overstated by approximately 84 Kgs V and 6.5 '

j kgs.U respectively. These discrepancies were a direct result of the 4 incorrect calculations of additions to and removals from material in process noted above. This issue will be discussed further in 1 paragraph 3 below. j i

c. Measurement Quality )

The licensee's program for monitoring and controlling the quality of accountability measurements was inspected. The inspection consisted

- of an examination of the licensee's written procedures.for evaluating measurements, responding to out-of-control measurements, and i

. calculating revised control limits. The performance of the program during'the July 15 through September 16, 1988, inventory period was reviewed. Control charts were inspected and control limit .;

calculations were verified.

Controls charts were established and maintained as specified in the FNMC Plan and in written procedure'NMC-003, Revision 01, " Procedures for Calculating Control Charts." The procedure covered calculation and revision of control limits, the data base used, tests for data normality and randomness, and preparation of control charts. The procedure covered only standard measurements, however, and did not address replicate control. There was no procedure for controlling replicate measurements. This is an apparent violation of NRC requirements and is included as one of the three examples listed in ,

paragraph 2a above (see item 88-12-03). {

i Procedure NMC-003 included instructions for calculating control  !

limits, however the mathematical model was not included. Failure to l include mathematical models in procedures was identified as an issue i during the inspection conducted during the week of October 11, 1988.

l

g 4 t

5 The licensee has agreed to revise this procedure to include the mathematical model for calculating control limits. This issue will be tracked as part of inspector follow-up item 88-12-02. In addition, the procedure had not been approved by written signatures.

This is an example of an issue identified as an apparent violation during the October inspection and will be tracked as part of 'q l inspector. follow-up item 88-12-01. The licensee made a verbal l commitment to correct both of these problems by March-15, 1989.

Paragraph 4.5.1 of the FNMC Plan outlined the appropriate methods for evaluating and responding to out-of-control measurements. The specific written procedure for handling out-of-control measurements of bulk standards was included on Form QA-017, Revision 6, " Nuclear Material Control, Verification of Scale Precision." Measurements of J laboratory standards were charted in the lab and out-of-control measurements were treated as specified in Paragraph 4.5.1, however there was no written procedure for responding to out-of-control measurements of lab standards. This is a an apparent violation of NRC requirements and is included as one of the three examples listed in paragraph 2a above (see item 88-12-03).

Control charts for standard measurements and for replicate measure-ments performed during the July 15 - September 16, 1988 inventory period were examined. Standards were measured and controlled for all v U and U-235 analytical systems except spectrophotometric analysis which was used to determine the uranium concentration of lab waste solution. The system is calibrated prior to measuring lab waste solution samples, however, control limits for monitoring the standards associated with the system was estimated from standards measured by titration (modified Davies and Gray) instead of from calibration or control standards associated with spectrophotometric analysis. Lab waste solution samples were not replicated, so replicate U and U-235 analyses were not performed or monitored for lab waste solution. Failure to include spectrophotometric analysis of lab waste solutions in the program for monitoring the quality of SNM accountability measurements and to generate current data on the performance of the system is an apparent violation of NRC require-ments (88-12-06).

While reviewing the measurements systems subject to control, the inspector noted that systematic error associated with U analysis for NDA measurement systems was consistently less than the systematic error associated with U-235 analysis. Since the U value associated with an item measured by NDA is obtained by applying a %U-235 factor to the measured U-235 value, the U error should be greater than the U-235 error. That is, variance (90)= variance (9 0-235) + variance (average enrichment). The licensee is in the process of reviewing the calculation of factors pursuant to new requirements specified in '

Iim

6  :

i 10 CFR Part 74 as they apply to high enriched facilities. These new  !

requirements will be implemented in the new future. Since the error  !

associated with the factor depends on the data base used to calculate I the factor, the licensee has agreed to review the error estimate  !

along with the review of requirements for calculating factors. This  !

issue will be tracked as an inspector follow-up item (88-12-07). i

3. Follow-up on Inspector Identified Items (92701)

(0 pen) Inspector Follow-up Item (85-14-02) Revision of method used to determined coefficients employed in LEID calculations. During the i inspection conducted on September 30 - October 4, 1985, it was found 1 that the coefficients used in LEID calculations had been overstated for finished fuel measurement systems. During this inspection it was found i that, as a result of inaccuracies in the calculations of additions to  :

and removals from material in process, those coefficients were again  !

overstated. Therefore the licensee verbally committed to document a rigorous derivation of the coefficients for each measurement system and to- . '3 incorporate the appropriate mathematical models of those coefficients  ;

into the procedure LEID calculations by March 15, 1989. This item will  !

remain open pending NRC review of the derived models and the revised procedure.

(Closed) Inspector Follow-up Item (85-14-01) Revision of methods used to compute add 1tions to and removals from material in process. As a result l

of the findings listed in paragraph 2b above, this item is closed.

4. Exit Interview (30703)

The inspection scope and findings were summarized on December 9, 1988 with those persons indicated in paragraph 1 above. The areas inspected were described and the inspection findings listed below were discussed in detail. The licensee acknowledged the inspection findings and took no exceptions.

Item Description and References 88-12-01 Inspector Follow-up Item - Three additional examples of procedures not having been approved as in previous item 88-10-01 (Paragraph 2a).

88-12-02 Inspector Follow-up Item - Revision of statistical procedures to include mathematical models (paragraph 2a).

1 l

L_________________. _ _ _ _ _ _ .__ _ . . _

!1 i i ,

7 Item Description and References' E 88-12-03 Violation . Failure to. establish written.

, material control.and. accounting procedures,

~t hree examples (paragraph 2a).

!88-12-04: Violation -~~ Failure to correctly. calculate -

additions-to'and removals;from. material in process (paragraph 2b).

88-12-05 . Inspector Follow-up Item .' Revision'of FNMC Plan to define' additions to and removals from

. material:in process (paragraph 2b).

88-12 Violation - Failure to: include' lab waste-analytical system in the measurement controlf program (paragraph 2c).

l 88-12-07 . Inspector Follow-up Item - RevisionLof method used to calculate NOA measurement error' variance.

-Licen'see management was informed that- of the two inspectorL follow-up

.. items discussed in paragraph 3, one was considered. closed'and one remained open.

li a

l L___!____________ . _ _ . _ _ _ _ . .

_ -l