ML20112D927

Joint Intervenors Proposed Findings on Joint Contention IV Re TLDs

ML20112D927
Person / Time
Site:	Harris
Issue date:	01/09/1985
From:	Runkle J JOINT INTERVENORS - SHEARON HARRIS
To:
Shared Package
ML20112D930	List: ML20112D927 ML20112D946
References
CON-#185-044, CON-#185-44 OL, NUDOCS 8501140547
Download: ML20112D927 (4)

Text

'bb January 9, 1985 E r ZEC s,u UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION N0.22 BEFORE THE ATOMIC SAFETY AND LICENSING BOARD (FFicE 9jy, uDC4L T'i'd ** $ERViCi*

ER4NCu In the Matter of )

)

Carolina Power & Light Company and ) Docket No. 50-400 OL North Carolina Eastern Municipal )

Power Agency )

)

(Shearon Harris Nuclear Power Plant) )

JOINT INTERVENORS' PROPOSED FINDINGS ON JOINT CONIENTION IV (THERM 0 LUMINESCENT DOSIMETERS)

A hearing was held on this and other safety-related contentions in Apex and Raleigh, North Carolina, in October and November, 1984. Applicants filed proposed findings on December 22, 1984. Other safety-related contentions will be covered by Wells Eddleman, for himself and for the Joint Intervenors.

Proposed Findings

1. Joint Contention IV as litigated in the hearing was as follows:

"whether the TLDs and measuring equipment and processes to be used at the Harris facility can measure occupationaldoses with sufficient accuracy to comply with the NRC regulations." Tr. 2218 (Telephone Conference of August 10, 1984).

2. [FindinEs 10 through 14 by Applicants summarize the procedural history of this contention. Finding 13 oversimplifies the Board's ruling of April 13,

-1984, and we would delete the second sentence).

3. [ Finding 15 by Applicants).

4. [ Finding 16 by Applicants}.

~

8501140547 850109 PDR ADOCK 05000400 0 PDR Y__.._-__----

Page 2

5. [ Finding 17 by Applicants].

6. Applicants will use Panasonic Model UD-802 AQ TLDs at the Harris plant to perform routine radiation monitoring of personnel. The objectives are twofold: the first is to assess the quarterly dose to individuals so as not to violate NRC regulatory limits, and the second is to compile cumulative dose to the individual for official exposure record-keeping purposes.

Browne, p. 3 - 4.

7. Maximum doses for a calendar quarter are established in 10 C.F.R.

Q20.101(a) as follows: (1) whole body; head and trunk; active bloodforming organs; lens of eyes; or gonads.. 1k rems; (2) hands and forearms; feet and ankles.. 18 3/4 rems; (3) skin of whole body.. 7 rems. {20.101(b) allows dosages inexcess of these under certain limited conditions. 620.102 restricts the licensee in exceeding the @20.101 limits further depending on a determination of prior doses the individual has received.

8. 10 C.F.R. $20.407(b) requires the Applicants to report on an annual basis the number of individuals in each of a series of exposure ranges.

The doses under 1 rem have approximately a quarter-rem increment while those greater than 1 rem are in terms of integer values in rem units.

9. [ Finding 22 by Applicants with the word " prestigious" in the second sentence deleted].

10. [ Finding 23 by Applicants].

11. The performance standard was then changed to P + S < L, rather than P + 2S < L. Tr. 6607 (Plato). As adopted, the ANSI performance standard assures that virtually every dosimeter tested by a particular processor must be accurate to 50% (or to 30% in the case of accident level categories).

s Page 3

12. NUREG/CR-2891 (Performance Testing of Personnel Dosimetry Services, introduced as Board Exhibit 2) lists at pp. A.46 - A.47 various sources of uncertainty not included in the performance evaluation. They are as follows:

(1) dependence of dosimeter response on radiation energy for a given type of radiation and geometry of radiation incidence; (2) dependence of dosimeter response on angle of radiation incidence for different types of radiation and different radiation energies; (3) dependence of response on ambient temperature and humidity; (4) time intervals between dosimeter issue, irradiation, and processing or readout; (5) dependence of response on visible and ultraviolet light; (6) position of the badge on the human body; (7) possible bias in the performance of an open test, that is, a test carried out with the knowledge of the processor.

If, as the Applicants' witness testified, these listed sources are minimized, the Applicants should be able to meet the ANSI performance standard of 0.5, steps provided that/are also taken to minimize dosimeter variability and clerical errors. Additionally, the Applicants' overall quality assurance program should include: (1) detailed written procedures for the performance of all routine dosimetry operations; (2) formal training and qualification for all operating personnel; and (3) formal supervisory review of all quality control records. Browne, pp. 21 - 24.

13. The ANSI performance standard (P + S < L) was designed to give 50%

accuracy with a 95% confidence level; in other words, for each dose read off of a TLD, it is very likely that the actual dose is somewhere within i of that reading. To meet the actual maximum doses allowed in the NRC regulations, and again be at a 95% confidence level the regulatory limit was not exceeded,

Page 4 one must take the regulatory limit and multiply it by two-thirds. This algorithm is numerically determined with the following fornula:

R + .5R = X, where X is the regulatory limit and R is the maximum value which can be read from the TLD with 50% accuracy and still not exceed X Thus, if the annual limit is 5 sams, the maximum permitted a reading from the a TLD meeting / ANSI performance standard of 50% would be 3.33 rems. Similarly, Applicants would not very likely (95% confidence level) exceed the quarterly limit of it rems if the maximum permitted a reading from a TLD was .833 rems.

14. In conclusion, given the inaccuracies present in the state-of-the-art TLDs, Applicants should be required to have a maximum value from the read TLDs which would make it very likely that the regulatory limits were not exceeded. As the Applicants have expressed their commitment to the ANSI performance standard of 50% accuracy (and 30% accuracy at accident levels) with a 95% confidence level, the maximum value from the read TLDs should take those levels of accuracy into consideration.

Respectfully submitted, John Runkle Counsel fot Joint Intervenors--Joint Contention IV (TLDs) 307 Granville Road Chapel Hill, NC 27514 919/942-7935, 942-0600 January 9, 1985

-