ML20058L564
| ML20058L564 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 08/02/1990 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20058L563 | List: |
| References | |
| NUDOCS 9008070284 | |
| Download: ML20058L564 (5) | |
Text
-
p %g ENCLOSURE
[f
+
. J.
UNITED $TATES NUCLEAR REGULATORY COMMISSION W ASHINGTON, D. C 2MM o
g e.e s SAFETY EVALVATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION AMMONIA DETECTION CAPABILITY FOR CONTROL ROOM OPERATOR PROTECTION TROJAN NUCLEAR PLANT DOCKET NO. 50-344 1.0 INTRODUCTig In response to NRC staff concerns regardin control room habitabilit followingapostulatedaccident,PortlanddeneralElectricCompany(yPGE, the licensee) proposed by letter dated March 2, 1982 to install additional protective measures to counter a potential release of a monia in the vicinity of Trojan Nuclear Plant that might incapacitate personnel in the control room.
By letters dated March 31, 1983, February 24, 1984, and February 12, 1985, PGE informed the NRC staff of measures being taken to monitor for amonia, and described operational problems with the amonia detectors, and continuing troubleshooting and corrective efforts being made by PGE.
By letter dated September 30, 1986, PGE provided results of an analysis to determine whether instrumented amonia detection capability is actually necessary. The analysis concluded that use of ammonia detectors was not necessary for protection of control room personnel. To facilitate the NRC staff. review of this analysis, the staff requested by letter dated January 11, 1988, that PGE provide the results of detector performance observations that had been completed since their September 30, 1986 submittal.
By letter dated llarch 2, 1988, PGE responded with the infor-mation requested.
2.0 EVALUATION The licensee's updated FSAR Section 6.4 identifies anhy( ius amonia as a potential hazard to control room personnel in the event sf an accidental spill in the Trojan Plant vicinity. A potential source of a monia from a spill is the iall line located on the east side of the Columbia River and used by Burlington Northern and Union Pacific for industrial transportation.
Accordingly, in 1982, amonium detectors were installed to protect against this postulated event.
Following installation, there has been a long period of unsuccessful efforts to make the detectors operate properly. The original malfunctioning detectors were replaced with new detectors.
Despite efforts by onsite vendor representatives, the new detectors were eventually sent back to the manufacturers in February 1985 for troubleshooting before reinstalling them. Despite all reasonable efforts, the detectors are currently unreliable, will not hold calibration, and are generally undependable for their intended function.
900807o284 900002 PDR ADOCK 05000344 P
During this period of unsuccessful effort to render the detectors operable, the licensee undertook a study to determine whether the a conia detectors were actnlly necessary for protection against an amonia spill i
event. Using NURN/CR-1741, "Models for the Estimation of Incapacitation i
Times Following Exposure to Toxic Gases or Vapors' Ion system the licensee performed an analysis for the existing control room ventilat Also, using the guidance given in NUREG/CR-1741, anhydrous ammonia is identified as a concentration-dependent substance, rather than a dose-dependent substance, in its physiological effect on human performance.
That is, incapacitation from exposure to anhydrous amonia is much more strongly cependent upon concentration during exposure than upon the total
)
dose over a prolonged period.
Be'ow a certain threshold concentration, amonia will not cause incapacihtion even with prolonged exposure, because the body can negate thr potentially toxic effects of the substance.
The model used to calculate the concentration of a monia in the control room atmosphere in the event of a spill is consistent with NURE6-0570, i
- Toxic Vapor Concentration in the Control Room Following a Postulat 4 Accidentel Release." The base: and assumptions used to determine 9 concentrations in the vicinity of the spill are consistent with 1..e guidance in Regulatory Guide 1.78 and NUREG-0570. Dilution due to wake effects from the plant buildings is conservatively modelled ano calculated using accepted methodology.
Concentrations in the control room atmosphere were calculated conservatively using normal control room ventilation.
l The anal T0XGAS (ysis was performed using the Bechtel sta ard computer code NE314). Conservative assumptions were used. The normal control room HVAC was used rather than using the control room emergency i
ventilation system in the recirculation mode to exclude the postulated i
ammonia plume from the control room. Also, no credit is taken for detection and warning of the event until the control room operators detect the odor in the control room, l
Using the results of the model for concentration, the time to first i
l detection of odor and the time to incapacitation were calculated.
Results of the calculation show that, without control room isolation and without operator action, it will take 4 minutes, 3 seconds for the operators to become unconscious from the calculated level of 25 ppm. Since amonia has a sharp irritating odor at lower levels of concentration, incapacitation time can be measured from the time the odor threshold is reached.
There are 3 minutes, 48 seconds of lapsed time from the odor threshold level (at which time operators begin to don breathing apparatus) until incapacitation l
levels are reached if no action is taken.
Using the normally applied i
I assumption of 2 minutes allowed time for donning breathing apparatus after detection of the presence of amonia, there is ample time for operators to don breathing apparatus without instrumented amonia detection capability.
According to the licensee comitment contained in their letter dated September 30, 1986, human detection will be employed in place of instrumented amonia detectors, along with administrative controls and training of control room personnel, to provide protection of control room personnel from the effects of an amonia spill.
4
3.0 CONCLUSION
The staff has completed its review and finds that operable instrumented amonia detection capability is not necessary for protection of control room personnel in the event of a spill of anhydrous amonia in the vicinity of the Trojan plant. The licensee will instead provide the protection afforded by human detection, together with appropriate administrative measures and training, and accessible suitable breathing apparatus that can be donned by control room personnel within two minutes after onset of human detection of the toxic gas.
Principal Contributor: Roby Bevan Dated:
August 2, 1990 r
t L
~
.