ML19284A664

From kanterella
Jump to navigation Jump to search
Responds to NRC 790206 Ltr Re Violations Noted in IE Insp Rept 50-285/79-01.Corrective Actions:A Procedure Change Will Be Written in Order to Incorporate QC Requirements
ML19284A664
Person / Time
Site: Fort Calhoun Omaha Public Power District icon.png
Issue date: 02/28/1979
From: Short T
OMAHA PUBLIC POWER DISTRICT
To: Madsen G
NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
Shared Package
ML19284A661 List:
References
NUDOCS 7903150165
Download: ML19284A664 (3)


Text

.

l .

(_pe .

v./ :

" ' ~

Omaha Public Power District 1623 HARNEY s OMAHA. NEHRASKA 60102 s TELEPHONE SJ6-4000 AREA CODE 402 February 28, 1979 FC-310-79 Mr. Glen L. Madsen U. S. Iluelear Regulatory Commission Office of Inspection and Enforcement Region IV 611 Ryan Plaza Drive Suite 1000 Arlington, Texas 76011 Re ference : OIE Report 50-285/79-01, dated February 6, 1979

Dear Mr. Madsen:

In reply to the above-referenced inspection report, the Omaha Public Power District herein submits comments to indicate the status of the reported iten of non-compliance.

In fraction Appendix F,10 CFR 50, Criterion V, requires that activities af-fecting quality shall be prescribed by documented procedures and shall be accomplished in accordance with these procedures. This requirement is incorporated in Section 17.5 of the OPPD Quality Assurance Program.

s Surveillance test procedure ST-ICI-2, Incore Detector Alarm Limits, requires verification that the incore detector high alarm linits are less than or equnl to the calculated alarn limits.

Contrary to the above, nine incore detector high alarn limits were not verified to be less thmi or equal to the calculated alarm limits during the performance of ST-ICI-2 cn January 3,1979, and eight of this nine had high alarm limits which were greater than the calculated alarn limits.

Discussion Technical Specification 2.10.3 requires that incore detector alarm linits be set so that proper tonitoring of reactor Kv/Ft is assured.

Recommended alarm settings are generated by an off-site cceputer during the determination of a reactor core power distribution. Inputs to the alarn calculation are detector voltage signals that have had appropriate uncertainties applied.

7so31sD/67

Mr. Glen L. Madsen February 28, 1979 Page Two On January 3,1979, the initial cycle 5 alara limits were set into the plant computer following a power distribution obtained at approxi-tately 655 power during power ascension testing. Personnel error re-sulted in erroneous alarm settings on eight of nine incore detectors involved.

A sunnary of the recommended and actual alarm settinco is as fol-lows:

High Alarn Settings (IIV)

Detector Reco= ended As Left 14-1 h8.83 49.0 18-1 87.lh 91. h 18-2 93.25 101.1 18-3 83.76 100.2 18 h 79.76 95.3 33-1 91.Eh 10h.0 33-2 101.08 104.5 33-3 100.22 95.0 33 4 95.28 10h.5 Linits were set or left higher than recommended on eight of the nine detectors. Their discrepancies were discovered during normal rurveillance performed on January 19, 1979, after a power distribu-tion taken at a nominal 1005 reactor power.

Detector 1k-1 was checked but not reset from its cycle h value.

By procedural definition the setting should be equal to or below the recommendation. While the alarm was not set correctly, this amounted to an error of .12 (IIV) or 0.2h5 of the recommendation in the case of detector lb-1. This percentage would not have been a significant detriment to the ability of the incore system to monitor the Kw/Ft LCO.

The discrepancies on detector strings 18 and 33 resulted when detector 18 was updated twice, leaving string 18 set at the limits re-co mended for 33. Detector 33 was not updated frot cycle h. The error occurred in the correlation of detector designation numbers from the off-site computer with those peculiar to the plant computer. By coin-cidence 'off-site' detector 18 is 'on-site' detector 33.

However, even if all nine of these detectors were considered to be inoperable from the standpoint of Kw/Ft monitoring in addition to 12 other detectors considered to be inoperable because of unreliable signals, the result is that 21 out of 112 detectors vould not be avail-able for monitoring Kw/Ft. Since that is equivalent to 18.755 of the total, the incore detector system satisfies the criteria of Technical Specification 2.10. 3 regarding number of operational detectors. In addition, the requirements on operable detector distribution were satis-fled at all times.

r .

Mr. Glen L. Madsen February 28, 1979 Page Three Additional assurance of reliable conitoring is evident by consider-ing that the methcdology employed in generating alarn limits effectively uses all detectors to monitor the peak core location. That is , any significant change in power distribution will cause all detector signals to increase toward their alarm limits.

Inoperable alares on eight of these nine detectors were not signi-ficant with respect to nuclear safety, but only from the standpoint of procedural non-compliance.

Response

(1) Corrective steps which will be taken to avoid further non-compliance.

(a) In order to assure personnel accuracy is maintained during tasks such as the highly repetitive changing or verifying of the 112 incore detector alarn limits , cross checking is necessary.

A procedure change will be written by March 15, 1979, to Surveillance Test ST-ICI-2, Incore Detector Alarm Limits ,

in order to incorporate Quality Control requirenents so as to implement an independent review of the alarm litits.

(b) The Reactor Engineer and alternates have been reinstructed on the importance of being thorough and effecting timely input on the plant review process.

(2) Date when full compliance will be achieved.

The Omaha Public Powar District is presently in full compliance.

Corrective action was taken on January 19, 1979 Sincerely, -

(?. /),./F.w\

4

\

T. F. Short Assistant General ManaCer TES/RLA/LTr:/LJD:jen ec: LeBoeuf, Lamb, Leiby & MacRae 1757 "II" Street, II. W.

Washington, D. C. 20030