ML18004A255
| ML18004A255 | |
| Person / Time | |
|---|---|
| Site: | Harris  |
| Issue date: | 06/24/1986 |
| From: | Zimmerman S CAROLINA POWER & LIGHT CO. |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| RTR-REGGD-01.095, RTR-REGGD-1.095 NLS-86-210, NUDOCS 8607010093 | |
| Download: ML18004A255 (13) | |
Text
REQUL RY INFORMATION DISTR IBUT SYSTEN (R IDB)
ACCESSlON NBR: 8607010093 DOC. DATE: 86/06/24 NOTARlZED:
NQ DOCKET FACIL: 50-400 Sheav on Hav v is Nucleav Powev Plant>
Unit 1>
Cav olina 05000400 AUTH. NANE AUTHOR AFFILIATION ZINl'1ERNAN> S. R.
Cav olina Power Sc Light Co.
RECIP. NAl'>E REC XPlENT AFFILIATION DENTQN> H. R.
OFFice oF Nuclear Reactor Regulation>
Div ec tov (post 851125
SUBJECT:
Fovwav'ds contv ol v oom habitability analysis> pev'ompliance w/Reg Guide
- 1. 95. Clav iFication necessary due to new inFo received
+rom vendors v e control v oom damper closuv e times>
chlov ine detection v esponse times Zc envelope leakage.
DISTRIBUTION CODE:
A003D DOPIER RECEIVED: LTR i ENCL g SIZE: /0 TITLE: QR/Licensing Submittal:
Supp 1 1 to NUREQ-0737<Generic Ltr 82-33)
NOTES: Application Fov permit renewal filed.
05000400 RECXPIENT ID CODE/NANE PWR-A ADTS PWR-A EICSB PWR-A PD2 LA BUCKLEY,B PWR-A RSB INTERNAL: ADN/LFNB NRR BWR ADTS NRR PWR-B ADTS NRR/DSRQ El'1RXT NRR/DSRQ/RSIB RCN2 EXTERNAL: LPDR NSXC COPIES LTTR ENCL 1
2 2
1 1
1 1
1 0
1 1
1 1
1 1
1 1
1 1
RECIPIENT ID CODE/NAiitE PWR-* EB PWR-A FOB PWR-A PD2 PD PWR-* PSB IE/DEPER/EPB NRR PAULSQN> W.
NRR/DHFT/NTB N
EIB FILES NRC PDR COPIES LTTR ENCL 1
1.
1.
7 7
1 3
3 1
1 1
1 1
1 1
1 TOTAL NUBBER QF COPIES REQUIRED:
LTTR 32 ENCL 31
I p.
b g I'K
SNK Carolina Power & Light Company SERIAL: NLS-86-210 tIN 24 Ijg Mr. Harold R. Denton, Director Office of Nuclear Reactor Regulation United States Nuclear Regulatory Commission Washington, DC 20555 SHEARON HARRIS NUCLEAR POWER PLANT UNIT NO.
1 - DOCKET NO.50-000 CONTROL ROOM HABITABILITYANALYSIS
Dear Mr. Denton:
Carolina Power R Light Company (CPRL) submits clarification and justification for the Shearon Harris Nuclear Power Plant (SHNPP) compliance with Regulatory Guide 1.95.
These clarifications are necessary due,to new information received from our vendors on control room damper closure times, chlorine detection response times, and a revised estimate of control room envelope leakage by our Architect/Engineer (Ebasco).
These clarifications do not affect, the, health,and safety of the. public or the ability of our control room staff to safely shut down the SHNPP.
The attachment provides an analysis which demonstrates that the plant operators have ample time to don breathing apparatus in the unlikely event of a chlorine tank accident either onsite or offsite. Revised FSAR pages which reflect this analysis willbe submitted in a future amendment.
If you have any questions, please contact Mr. David C. McCarthy at (919) 362-2010.
Yours very truly, SRZ/GAS/pgp (3962GAS)
Attachment S.
'erman Manager Nuclear Licensing Section cc:
Mr. R. A. Benedict (NRC)
Mr. B. C. Buckley (NRC)
Mr. G. F. Maxwell (NRC-SHNPP)
Dr. 3. Nelson Grace (NRC-RII)
Wake County Public Library Bb07010093 8 000400 FiDR AQQCK 0>
pgR 411 Fayetteville Street
+ P. O. Box 1551
~ Raleigh, N. C. 27602
ATTACHMENTTO NLS-86-210 ANALYSISOF CONTROL ROOM HABITABILITYFOLLOWING POSTULATED CHLORINE RELEASE ACCIDENTS (3962GAS/pgp)
1.
INTRODUCTION As a result of review of'vendor'test data and other design documents, CPRL has
,'stimated the anticipated leakage'f'.the SHNPP control room HVAC envelope.
This calculation, performed in accordance with AEC RRD Report NAA-SR-10100, was done to verify that the control room conformed to the specifications of the NRC Regulatory Guide(RG) 1.95, Rev. 1:
Protection of Nuclear Power Plant 0 erators A ainst an Accidental Chlorine Release.
Table 1 of RG 1.95 enables a simple determination of compliance with the control room habitability requirements following a postulated accidental release of chlorine for any one of the six control room types listed.
The results of the leak rate calculations, as well as test data on the Anacon chlorine detectors which are installed in the control room outside air intake, indicate that the SHNPP control room does not conform to any of these six types, precluding the use of Table 1 to determine compliance.
A site-specific analysis was performed using the model described in RG 1.95.
This model is embodied in Ebasco's TOXCHM computer
- program, which is an enhanced version of the program originally obtained from James Wing of the NRC staff.
A detailed description of the
, original TOXCHM model can be found in NUREG-0570, Toxic Va or Concentrations in the Control Room Followin a Postulated Accidental Release, by James Wing.
,..The..present. analysis. uses. a.recently revised..version.,of,theTOXCHM..model, which
".employs a'more realistic.method:of;calculating the dispersion of;the vapor evolving from
'.leaking "tank. Therefore, results of'these calculations cannot be properly compared to
'hose of the earlier studies.
2.
METHODOLOGY 2.1 Assumed Control Room Characteristics 2.1.1 Infiltration rate of control room envelope The isolated infiltration rate was calculated as follows:
R
=
(L+ r) 60/V 2
R
=
Infiltration rate after isolation 0.102 vol/hr L
=,'eak rate under 1/8" w.g. pressure 315 cfm (Note I) r
=
Infiltration rate due to opening and closing of doors 10 cfm (Note 2)
V
=
Free air volume of control room envelope 71,000 ft (3962GAS/pgp)
NOTES:
1.
This is a very conservative estimate.
The control room leakage test has not been completed at this time but is expected to be greater than 0.06 vol/hr (71 cfm).
2.
This represents an additional
. conservatism since RG 1.78 position C.10 only requires a 10cfm ingress/egress allowance for control rooms without airlocks.
This method of calculating the air exchange rate for an isolated control room was recommended by the NRC AuxiliarySystems Branch in response to a telephone inquiry on" December 10, 1985.
This is the same method of'calculating inleakage that is used in the habitability evaluation for an isolated control room following a radiological accident.
2.1.1 Detector response The chlorine detector probes were assumed to produce a'ignal corresponding to 10 percent of their ultimate readings within 20 seconds of. being exposed to a given, constant chlorine concentration.
This assumption is based on the report of the qualification tests on the Anacon detectors, performed by the Acton Environmental Testing Corp., which lists 20 seconds as the acceptable response time for the probe to register 10 percent of the full-scale reading.
(This interpretation was confirmed by the
.detector. manufacturer.)
.Therefore, since the detectors were assumed to have an. alarm
,trip-.point;,of'5 ppm, it was assumed that the alarm would be tripped 23.2 seconds after
" the chlorine concentration at the air intake reached 50 ppm.
(The electronic processing of the signal can take up to 3.2 seconds.)
Note: the actual alarm trip point will be set less than 5 ppm, and the response time field verified.
2.1.2 Control room parameters The following is a summary of the control room parameters employed in the current analyses:
Height above ground of fresh air intake Outside air exchange rate (normal operation)
Isolation valve closure time (tech spec)
Infiltration rate after isolation Response time of chlorine detector probe (to 50 ppm chlorine concentration)
Electronic processing time of signal 50 ft
.887 volumes/hour 15 sec .102 volumes/hour 20 sec 3s2 sec 2.2, Meteorological Parameters
., Regulatory. Guide 1.95 specifies that atmospheric dispersion calculations be performed by the methods described in RG 1.78, which, in turn, specifies that the worst five percentile meteorology be assumed for accident evaluation (as is done for radiological accidents).
Unlike the case of radionuciide releases, it is difficultto determine ~ariori which of the likely combinations of wind speed and stability class would result in the worst consequences for a given chemical release accident.
Therefore, the calculations performed analyzed or considered likely combinations, based on the meteorological data that had been collected at the SHNPP site, and then determined if the control room was habitable following accidents occurring under all but the worst five percentile conditions.
The calculations used the tables in the SHNPP
- FSAR, speci fically Table 2.3.3-13, which lists the point frequencies of stability class and lower level wind (3962GAS/pgp)
speeds for the years 1976-1978.
This table was selected since it spanned the longest contiguous time period and since the frequencies are presented with much greater precision than in the table for the later period (1979-1980).
The CPRL Meteorology
- Section has confirmed. that this table accurately presents the data described in it and that this data is consistent with corresponding data collected in later years.
The way in which this data was used is explained below in the discussion of the analyses of the different postulated accidents.
Since the consequences of a chlorine release accident for a given wind speed and stability class worsen with increasing temperatures, the highest plausible temperatures, taken from the, FSAR meteorological data tables, were used for each combination of wind speed and stability class.
Each such data set, comprising a given value of stability class, wind speed, air and ground temperatures, and radiant energy flux, used in these analyses, is referred to as a "met case" in the subsequent discussion.
2.3 Accident Scenarios 2.3.1 Onsite Chlorine Releases The following input data was used in analyzing postulated failures of the onsite chlorine storage tank:
..Maximum amount of chlorine release
'istance,to control room from release point Compass direction Distance to remote detectors from release point Separation of remote detectors Height of detector probes 55 tons 1580 feet 65'7 feet 07 feet 2 feet The amount of the release is set equal to the maximum capacity of the tank.
The distances are from the center of the tank car.
The distance to the control room is actually to the outer wall of the RAB and was scaled from the site plan; the direction is from the tank car to the control room.
This direction is almost in the center of the ENE wind sector.
The distance to the remote detectors was calculated by adding the position of the detectors with respect to the site plan (W 28.39') to the position of the railroad spur on which the tank car is parked (E 029).
The following table, abstracted from FSAR Table 2.3.3-13, lists the percent frequency of winds with the given velocities occurring during the listed stability classes in the ENE direction.
Wind Speed (mph)
Class G
F E
D C
B A
0-.75 0.0168 0.02805 0.1202 0.008010 0.
0.
0.
~75 305 1.190 0.6052 0.5931 0.3326 0.02000 0.01603 0.000007 305 705 0.00007 0.06011 0.0127 0.8535 0.1923 0.1162 0.1603 7.5-12.5 0.
0.
0.09217 0.1202 0.03206 0.01202 0.00809 12.5-18.5 0.
0.
0.000007 0.
0.
0.
0.
(3962GAS/pgp)
The following table, abstracted from FSAR Table 2.3.3-13, lists the percent frequency of winds with the given velocities occurring during the listed stability classes in the NNW direction.
Wind Speed (mph)
Class 0-.75
~75 305 305 705 7.5-12.5 12.5-18.5 G
F E
D C
6 A
.3126
.02000
.008010
.008010 0.
0.
0.
.8896
.5009
.0889
.3526
.02805 0.
'01202
.03606
.1960
.7213
.9537
.1603
.2280 0.
.000007
.1322
.569
.1563
.1122
.3005 0.
0.
.02805
.03606
.02000
.01603
.03606 Combinations of wind speed and stability class having non-zero frequencies were considered in these analyses, with each wind speed range being represented by its mid-
. point..For. examplethe.highest:.wind speed:.for. Class G was assumed -to be 5.5 mph, the
.middle of the highest-wind. speed range for.this class.
2.3.2 Offsite Chlorine Releases The only known offsite sources of chlorine that require considerations are the following:
A tank truck traveling on US Highway 1 and carrying 20 tons of chlorine (the maximum loading for such a truck).
The accident is assumed to occur at the point nearest the
- plant, which is 6,965 feet away in the north-northwest direction.
2.
A railroad tank car being transported on the tracks of the Seaboard Railroad carrying 90 tons of chlorine (the maximum loading of a tank car).
The accident is assumed to occur at the point nearest the plant, which is 10,700 feet away, also to the north-northwest.
2.0 Method of Analysis The. TOXCHM model predicts the chlorine concentration both outside and inside the
~ control'room as a function of time from each accident.
Control room operators are
~
~ assumed to don self-contained breathing apparatus within two minutes after the high chlorine concentration alarm is actuated, as specified in RG 1.78.
If the chlorine concentration in the control room does not exceed the IDLH value at this time, the control room is assumed to satisfy the habitability requirement of General Design Criterion 19 of 10 CFR 50, Appendix A.
(IDLH is the acronym for Immediately Dangerous to Life or Health and is defined as a "maximum level from which one could escape within 30 minutes without any escape-impairing symptoms or any irreversible health effects."
A joint NIOSH-OSHA study group has established 25 ppm as the IDLH level for chlorine.)
(3962GAS/pgp)
Three different types of releases from each chlorine source were considered:
l.
instantaneous rupture of entire tank, 2.
";, escape of liquid chlorine through a hole in the bottom of the tank, and 3.
escape of gaseous chlorine through a hole in the top of the tank.
In all cases, the chlorine was assumed to be stored at ambient temperature at a pressure equal to its vapor pressure at that temperature.
2.0. I Offsite Releases For offsite releases, the standard TOXCHM model was used to predict the consequences of an instantaneous tank rupture.
To analyze the second accident type, the gradual escape of liquid chlorine, the TOXCHM computer program was used to calculate the leak f
q, g
g 1
1 d
d pd
~P' En ineers'andbook.
The fraction that was instantaneously vaporized was calculated in the manner described in NUREG-0570 for a total release.
The released vapor plume was assumed to immediately expand in two dimensions, just as the vapor puff was assumed to expand in three dimensions in NUREG-0570.
The remainder of the analysis of a liquid leak proceeds the same way as for the
..instantaneous release
.discussed.above...To,.cover
.the entire spectrum of plausible
. 'ccidents,'leaks through five different holes,;varying in size from 0.1 inch to 6 inches in diameter, were analyzed for.each chlorine source.and each combination of wind speed and stability class.
(A leak through a hole greater than 6 inches in diameter cannot be modeled as a steady stream its effects willbe more properly represented by a total tank rupture.)
In the third accident type, the gaseous leak rate was also calculated using formulas found in Perr
's Chemical En ineers'andbook.
The remainder of the analysis proceeds the same way as for the gradual liquid release discussed above, except that the range of hole sizes was I/O inch to six inches.
3oint frequency table data for the NNW sector (similar to the data for the ENE sector, listed above) was examined and the met cases with non-zero frequencies were considered.
In practice, it was only necessary to consider met cases with stability classes D - G.
Cases with classes A C include the same wind speed ranges since these classes produce lower concentrations for the same wind speeds than, say, class D, such cases need not be analyzed if analyses for class D met cases showed that no hazard existed.
2.0.2 Onsite Releases
- Onsite releases were analyzed in a manner similar to that used for the offsite cases.
First, the control room was assumed to be in the normal operations mode at the time of the accident, and the alarm and isolation signai was assumed to be generated by the local detectors (mounted in the control room fresh air duct).
In those cases in which these detectors did not insure control room habitability, a second analysis was performed.
First, the concentrations at the position of the probe were calculated, assuming that the chlorine was passing exactly midway between the two probes.
(This is a
very conservative assumption, since the chlorine concentration decreases rapidly as the distance from the center line of the trajectory increases.)
These data were then used to determine the time at which the control room was isolated and the time when the operators could be assumed to have donned their breathing apparatus.
Finally, the (3962GAS/pgp)
control room concentrations were re-calculated, assuming that the control room was isolated before the chlorine reached the RAB (this assumption was verified by checking the appropriate result from"the prior calculation),'and the time at which the operators
.were.-assumed to don "breathing "apparatus'-was compared to the time the chlorine concentration reached the IDLH level.
3.
RESULTS 3.1 Onsite Releases 3.1.1 Instantaneous Rupture of Tank The analysis showed that for all rupture scenarios the chlorine concentration at the remote detectors exceeded the trip point within one second of the accident and that the control room was isolated before the arrival of the chlorine at the fresh air intake.
Note: This complies with RG 1.95 in that control room isolation is accomplished before chlorine concentration at the isolation valve reaches 5 ppm.
The control room operators can thus be expected to don their breathing apparatus before the chlorine concentration reaches the IDLH level.
The following table lists the 15 met cases that were used in the analysis.
The frequencies represent the normalized annual average joint frequencies for the ENE sector.
(The frequencies do not add up to 10096 because cases involving stability
- ..classes,A -.C,were. omitted, as discussed above.)
Case 1
2 3
5 6
7 8
9 10 11 12 13 10 15 Stability Class G
G G
F F
F E
E E
E E
D D
D D
Wind speed (mph) 0.0 2.1 5.5 0.0 2.1 5.5 0.0 2.1 5.5 10.0 15.5 0.0 2.1 5.5 10.0 Frequency (96) 7.53 21.58 0.72 0.51 11.66 1.09 2.17 10.72 7.06 1.67 0.07 0.10 6.01 15.02 2.20 Time of toxicity (min:sec) 00:10 8:10 3:12 03:00 8:05 3:13 02:00 8:05 3:20 2:02 2:50 CIAO 8:05 3:30 3:25 Cases 1 - 11 are likely to occur only at, night. It was, therefore, assumed that both the ground.and the air temperatures were 86 F, the highest likely nighttime temperature at the SHNPP site.
Cases 12 - 15 could occur in the daytime.
For these
- cases, the air temperature was assumed to be 100 F, while the ground was 122 F.
These temperatures represent extreme conditions and are, therefore, highly conservative.
For case 10 only, the chlorine concentration in the control room reached the IDLH level before the operators could don their breathing apparatus.
However, the probability of that combination of wind speed and stability class is only 1.6796.
Since RG 1.78 only requires that the ninety-five percentile meteorology be considered, this case can properly be excluded.
In fact, cases 3, 6, 10 and 11 can be jointly excluded, since the sum of their probabilities is only 3.5596.
In the remaining cases, the operators have at (3962GAS/pgp)
least three minutes and 20 seconds from the time of release to don their breathing apparatus.
The alarm willbe actuated within approximately 20 seconds of the accident,
~ and an additional two minutes 'and 56 seconds are available for the donning of the
." apparatus prior to exceeding the IDLH level.
3.1.2 Liquid Leaks In the case of a postulated leak through a 0.1 inch diameter hole in the bottom of the tank, chlorine will escape at approximately 30 pounds per minute (the actual rate will vary with the ambient temperature).
The chlorine concentration in the control room will not reach the IDLH level during the first two
- hours, the longest time steady meteorological conditions can be expected to persist.
For a 0.3 inch hole, the operators will have time to don their breathing apparatus before the IDLH level is reached except in met case 11 (see Table above).
Since that case only occurs.0796 of the time, it can be properly dismissed from further consideration.
In both of the above-mentioned leak scenarios, the local detectors sufficed to isolate and alarm the control room. For larger
- leaks, the combination of remote and local detectors was found to.provide adequate protection in the 15 met cases.
3.1.3 Gaseous Leaks
... The., local. detectors.,provide:.sufficient;protection from;gaseous leaks through holes of
-.I/O;inch~to'2;5 inches in. diameter. in.the.15.met cases..For leaks through larger holes, the remote detectors will alarm.;and'isolate the control room in enough time to enable the operators to don breathin'g.apparatus before the IDLH level is reached for the ninety-fifth percentile meteorological conditions.
3.2 Offsite Releases For offsite accidents, the met cases representing very low wind speeds were omitted from consideration.
Since the highway is over one mile and the railroad about two miles from SHNPP, vapor borne by the wind at.0 miles per hour would take much longer to reach the site than constant meteorological conditions can be reasonably expected to persist.
Case 1
2 3
5 6
7 8
9 10 11 Stability Class G
G F
F E
E E
E D
D D
Wind S eed(m h) 2.1 5.5 2.1 5.5 2.1 5.5 10.0 15.5 2.1 5.5 10.0 Fre uenc (96) 13.52
.507 7.67 2.98 7.03 10.96 2.01
.026 5.36 10.09 8.607 Time of Toxicity (Min:Sec) 57:20 22:55 59:00 00'00 1 10'00 Concentration (ppm) 2 Min.
After Alarm 7.69 27.0 3.78 10.5 2.72 3.89 5.25 5.35 2.00 2.51 2.80
~Limit Not Reached (3962GAS/pgp)
3.2.1 Railroad Tank Car In the case of the instantaneous rupture of the railroad tank car, only for met case 2 did the chlorine concentration. exceed the IDLH level before the operators could don their breathing apparatus.
Since that case occurs much less than 596 of the time, it can be properly excluded from consideration.
In other cases, the maximum concentration at the time the breathing apparatus is donned is 10.5 ppm, well below the IDLHof 25 ppm.
In the entire spectrum of liquid and vapor leaks, either the IDLH level in the control room is never reached, or the chlorine detectors in the air intake isolate and alarm the control room in time to adequately protect the operators.
3.2.2 Tank Truck In credible accidents, the chlorine detectors in the air intake will isolate and alarm the control room, enabling the operators to don breathing apparatus before the concentration reaches the IDLH level. The results of this analysis are bounded by 3.2.1.
0.
CONCLUSIONS 1.
The relatively simple model of detector response used in the present analyses assumes no response until the chlorine concentration reaches 50 ppm (i.e., ten times the assumed setpoint).
However, field tests show comparable response times with much lower chlorine concentrations.
The very steady conditions of wind speed and direction necessary to maintain a constant low ambient concentration are unlikely to persist long enough to build a toxic concentration in the control room.
21 The SHNPP control room habitability systems offer the operators adequate protection from..design.basis. chlorine.release..accidents.,and, therefore,.satisfy the intent of RG 1.95. --The hypothetical.'case of;the:outside,'chlorine"concentration being just below
.the effective alarm level of the detector,;and yet causing the operators to be incapacitated, cannot happen for the following reasons:
31 Chlorine can be detected by odor at a concentration of.31 ppm.
It is highly unlikely that the operators would passively wait while the concentration gradually built up to 25 ppm.
(3962GAS/pgp)
~g r
0