ML20137K487
| ML20137K487 | |
| Person / Time | |
|---|---|
| Site: | Vogtle  |
| Issue date: | 11/25/1985 |
| From: | GEORGIA POWER CO. |
| To: | |
| Shared Package | |
| ML20137K470 | List: |
| References | |
| NUDOCS 8512030207 | |
| Download: ML20137K487 (15) | |
Text
-
p
.i ,
e EVALUATION OF THE CHLORINE CONCENTRATION AT THE VEGP CONTROL BUILDING AIR INTAKE NOVEMBER 25, 1985 l
l l
l SBA22885Sihajig4 E
f Table of Contents Eagdi
1.0 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . 1 2.0 MODEL DESCRIPTION . . . . . . . . . . . . . . . . . . . . . 2 2.1 Model 1 - Regulatory Guide 1.95 Dispersion Model . . . . . . . . . . . . . . . . . . . 2 2.2 Model 2 - Unobstructed Path, Elevated Intake Model . . . . . . . . . . . . . . . . 3 2.3 Model 3 - Building Wake Model . . . . . . . . . . . . 4 3.0 DETERMINATION OF 5 PERCENTILE DISPERSION CONDITIONS FOR MODELS 1, 2. AND 3 . . . . . . . . . . . . . . . . . . . 7 4.0 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . 10 5.0 DISCUSSION . . .. . . . . . . . . . . . . . . . . . . . . . 12 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 l
1 I
l l
1
g,
1.0 INTRODUCTION
In response to the letter from Elinor G. Adensam of the Nuclear Regulatory Commission (NRC) dated September 6, 1985, the chlorine concentrations estimated to exist at the Plant Vogtle control room air intake due to the rupture of a one-ton chlorine cylinder is provided using the three atmospheric dispersion models and related equations recommended by the NRC staff (Reference 1).
This report describes the three models and summarizes the chlorine concentrations each predicts at the VEGP control room HVAC outside air intakes. A discussion of the findings and the selected dispersion model appropriate to use for the VEGP design basis is presented in Section 5.0.
1 4210t
(
2.0 MODEL DESCRIPTION The majority of chlorine releases experienced to date result in a leakage rate of chlorine of less than one pound of chlorine per second (Reference 2). However, in accordance with Regulatory Guides 1.78 and 1.95, a less probable, more severe accident is postulated which involves the instantaneous, catastrophic rupture of a one-ton chlorine cylinder located at the Nuclear Service Cooling Water (NSCW) chlorine building. Chlorine released by the accident is assumed to move directly towards the closest control room air intake located on the east side of the control building roof. It is conservatively assumed that all the outside makeup air enters the control room at this point.
This evaluation a) uses a model-dependent worst 5 percentile meteorology (which corresponds to a worst 5 percentile x/Q) and b) assumes 25 percent of the chlorine flashes instantaneously to form a puff.
The mass release rate associated with the residual chlorine pool was determined using the methodology described in NUREG 0570 (Reference 3,
- p. 9). The chlorine concentration predicted at the control room air intake was evaluated using the models (1, 2, 3) and equations described in sections 2.1, 2.2, and 2.3.
2.1 MODEL 1 - REGULATORY GUIDE 1.95 DISPERSION MODEL This model assumes that the chlorine puff and the plume disperse in a Gaussian manner as the cloud travels downwind from the point of release.
Pasquill Gifford dispersion coefficients (Reference 4 pp. 102-103) are used to characterize the chlorine concentration distribution for both the puff and plume portions of the release.
Upon reaching the control building, it is assumed that the chlorine undergoes uniform vertical mixing up to the height of the control room HVAC system outside air intakes.
The dispersion equations for Model 1 are as follows:
x (t) = exp
~( -"' Eq. la E"
6.28 (o pg+o)H 2 (oypg + oy)_
and O'(t')
xp g ,,(t) #= f r t' 1 0 Eq. Ib (2r)1/2 uoypgH 4
=0 , for t' < 0 L
2 4210t
o V' e
Where xpuff (t) = chlorine concentration at the intake due to the puff at time t, g/m3 xplume(t) = chlorine concentration at the intake due to the plume at time t, g/m3 Q = mass of the instantaneously vaporized chlorine, g Q'(t') = mass release rate from the pool of boiling chlorine at time t', g/see t = time elapsed since the beginning of the chlorine release, see t' = t - D/u = time adjusted to take into account how long it takes for the plume to reach the intake, see oy pg = Pasquill-Gifford standard deviation in the Y direction, m .
oy = [Q/(7.87 x density of C12 vapor)]l/3 = initial standard deviation of the puff, m u = windspeed, m/sec H = elevation of the Control Room air intake above grade, m D = distance from the release point to the intake, m 2.2 MODEL 2 - UNOBSTRUCTED PATH, ELEVATED INTAKE The second model considered assumes a ground-level release, elevated receptor, and an unobstructed path of travel as the puff / plume moves downwind from the release point. Dispersion in both the horizontal and vertical directions is assumed to be Gaussian. Dispersion coefficients described by Islitzer and Slade (NUREG 0570, p. 22) are used to characterize the chlorine concentration distribution within the puff. Pasquill-Gifford dispersion coefficients are assumed for the plume portion of the release.
The dispersion equations for model 2 are as follows:
0 ~ ' 'g - ~ "g xpuff (t) =
/2 #XP 2 8XP 2 2 Eq. 2a 7.87 (o y +o y) (o, + ay) _
y I z I and xplume (t) =
' exp -
, for t' > 0 Eq. 2b vu o o ~
ypg zPG 2o zPG 4
= 0 , for t'< 0
(
3 4210t
0 1
- 1 where ozPG = Pasquill Gifford standard deviation of the plume in the vertical direction oy= standard deviation of the puff in the horizontal alongwind and horizontal crosswind directions as reported by Islitzer and Slade, m 1
oz= standard deviation of the puff in the vertical direction as reported by Islitzer and Slade, m and other variables have the same meaning as defined for equations la and Ib.
2.3 MODEL 3 - BUILDING WAKE The third model considered is a building wake dispersion model. It assumes that a wake is created by air flow around the VEGP Unit 1 Containment Building. Chlorine gas entrained in the wake becomes uniformly mixed within the wake cavity. Prior to that time, puff and plume dispersion are assumed to follow Gaussian behavior. Dispersion coefficients in the puff portion of the release are evaluated using the relationships given by Islitzer and Slade. Dispersion' coefficients for the plume are derived from the Pasquill Gif ford curves. The dispersion equations for Model 3 are as follows:
x ~ -"
exp Eq. 3a Puff (t) = 2 1/2 2,,
(3)(7.87)(o 2 2(o 2)
+o y )(oz 2+#I2 _
Y and
- plume " u (w a '# 1 pg a p + A/2) 9'
=0 , for t' < 0 where A = the effective cross sectional area (m2 ) of the structure which creates the wake cavity other variables are as previously defined.
Table 1 suamarizes the meteorological input data and assumptions for NRC Models 1, 2, and 3. Table 2 lists the physical properties of chlorine used in the analyses.
4 4210t
TABLE 1 h METEOROLOGICAL MODELING INPUT DATA AND ASSUMPTIONS S
NRC R.G. 1.95 NRC Elevated NRC Building Model Intake Mocel Wake Model
- 1. Assumed Dispersion Worst 5 percentile Worst 5 percentile Worst 5 percentile Conditions x/Q x/Q x/Q
- 2. Pasquill Stability F B F
- 3. Windpseed (m/sec) 0.9 2.5 0.9
- 4. 1 C12 flash 25% 25% 25%
- 5. Ambient Temp 20*C O'C 20*C
- 6. Horizontal Distance, 200 meters 200 meters 200 meters
- spill to intake
- 7. Assumed elevation of 18.4 meters 18.4 meters Not relevant intake (uniform (credit taken (intake at vertical for Z exponent ground level) mixing depth) in dispersion equation)
- 8. X-Section of building Not relevant Not relevant 1722 m*
area for wake calculation
- 9. Dispersion coefficients Pasquill- Islitzer & Islitzer &
for puff Gifford Slade Slade
- 10. Dispersion coefficients Pasquill- Pasquill- Pasquill-for plume Cifford Cifford Gifford 4231t 9
(
TABLE 2 PHYSICAL PROPERTIES OF CHLORINE Value of Parameter Parameter Reference Molecular weight 70.9 g/ mole Ref 3, p 31 Normal boiling point -34.05 *C Ref 7, p 47 Liquid density 1.570 g/cm3 Ref 3, p 31 Liquid heat capacity 0.226 cal /g *C Ref 3, p 31 Heat of vaporization 68.8 cal /g Ref 3, p 31 6
4210t
g-
. 3.0 DETERMINATION OF 5 PERCENTILE DISPERSION CONDITIONS FOR MODELS 1. 2 AND 3 For the accidental release of chlorine at the VEGP, chlorine concentrations at the control room air intake are calculated for model-dependent worst 5 percentile dispersion conditions. For each dispersion model used, the 5%
meteorology was determined by ranking the I/Q values in descending order and by summing up their associated freqeincies from joint frequency tables of wind speed and atmospheric stability until the cumulative frequency became equal to or greater than 5%. The joint frequency table of windspeed and stability for VEGP is given on Table 3. The windspeed, stability category, and X/Q representative of the worst 5 percentile meteorology for each model is given on Table 4.
The methodology described above for determining the 5 percentile X/Q has previously been deemed appropriate and acceptable by the NRC staff (Reference 1). The assumption of 5 percentile meteorology is specified in Regulatory Guide 1.78 (Sections 1.78-4, -5, and -8) and NUREG/CR-3786 (Reference 6), and is considered sufficiently conservative for use in the VEGP control room habitability analysis.
4210t '
a
,r . -- 7 a: ,
O TABLE 3 JOINT FREQUENCY WINDSPEED - STABILITY CLASS FOR VECP (%)
Windspeed Stability Class m/sec A B C D E F G 0.9 0.989 0.207 0.125 2.19 6.13 4.902 3.857 2.5 8.2 1.67 0.929 9.59 18.73 9.27 4.997 4.5 7.8 1.235 0.583 6.86 5.85 0.514 0.082 7.0 1.76 0.229 0.073 1.348 0.739 0.0475 0.017 9.6 0.13 0.0086 0.013 0.104 0.0605 - -
12.8 - - -
0.0086 0.013 - -
14.6 - - - - - - -
l i
a 4210t
- - - , , ,me -
.-- .7 TABLE 4
SUMMARY
OF WORST 5 PERCENTILE METEOROLOGY FOR NRC MODELS 1, 2, 3 NRC Model Stability Windspeed, I/Q, Number Class m/sec sec/m3 1 F 0.9 3.020 x 10-3 2 B 2.5 1.307 x 10-4 3 F 0.9 1.152 x 10-3 m
9 4210t
(-
~
7 4.0 RESULTS The predicted chlorine concentrations at the VEGP control room NVAC intakes as a function of time for Models 1, 2, and 3 are shown on Figure 1.
h l
I 10 4210t
{ $
E.M0 .
LEGEND
. 200,000 O m REG. GUIDE 1.95 MODE L 1, F STABILITY,0.9 M/SEC 0-ELEVATED INTAKE MODEL 2 B STABILITY,2.5 M/SEC Q--BUILDING WAKE MODEL 3 F STABILITY,0.9M/SEC 100,000
_ l 80,000 j
60,000 l
40,000 l
?l 20,000
?'
- r. ';.lll w
f E 6,000 [
c 4,000 I-54 g -
l!
! 2,000 m"
5I l I g s
}
} l EG. GUIDE 1.95 g
40 h\
l l % %, BUILDING WAKE 200 l (a
MODEL i
100 _
} ;
k '!!
A P
() % y ELEVATED INTAKE I
20 l
10 f 100 200 a
300 400 500 700 600 TIME AFTER RELEASE,SEC Note:
Peak concentrations do not occur at the same time for models evaluated !
at different wind speeds.
)
CHLORINE CONCENTRATION AT CONTROL BUILDING AIR INTAKES i 1
FIGURE 1 )
i 11 I
)
[- U 5.0 DISCUSSION Numerical values of chlorine concentrations predicted by the three atmospheric dispersion models differ in magnitude and curve shape. The highest peak concentration is predicted by the building wake model (model 3) and is associated with the puff portion of the release. Predicted concentrations for the plume portion of the release are highest for the Regulatory Guide 1.95 model (model 1). The unobstructed path, elevated receptor model (model 2) yields lower predicted concentrations than either the building wake or Regulatory Guide 1.95 model for both the puff and plume portions of the release.
The configuration of buildings at VEGP is complex. There are numerous obstacles that a chlorine cloud would encounter while traveling between the NSCW chlorine building and the control room HVAC system outside air intakes. Given the complexity of the airflow pattern created by the VECP buildings, it is difficult to predict which of the atmospheric dispersion models will yield more appropriate chlorine concentrations. From the modeling point of view, both the unobstructed path, elevated intake model and the Regulatory Guide 1.95 model do not take into account the physical interactions between air and structures that would generate phenomena that are responsible for turbulence and mixing. The elevated receptor model assumes unobstructed flow which is not appropriate for Plant Vogtle. The Regulatory Guide 1.95 model is very conservative due to the assumption that the chlorine cloud would become uniformly mixed in the vertical direction up to the elevation of the control room HVAC intake.
The building wake model accounts in the physical sense, for the effect of intervening structures. However, the concentrations predicted by the model are considered to be excessive for low windspeed conditions, i.e.,
13 meters /second (see Appendix A of USNRC Regulatory Guide 1.145).
Moreover, there is uncertainty about how much dilution the puff would undergo when entrapped within the wake of the containment building. Based on discussions with the NRC we have assumed a maximum factor of 3 dilution of the puff within the wake cavity. Other assumptions are consistent with accepted NRC guidance.
We believe the building wake model, which yields excessive peak conrentrations due to the assumed value of the dilution factor for the puff and does noc apply at windspeeds less than 3m/sec, is inappropriate.
The elevated intake model results in chlorine concentrations that are lower than those predicted by other models and may not be representative of the VEGP physical configuration. Therefore, based on our evaluation of the three models we conclude that the Regulatory Guide 1.95 model is appropriate for use in evaluating the VECP control room habitability since it predicts the highest concentrations for the long-term, plume portion of the release and a peak concentration that is conservative. Consequently, chlorine concentrations predicted by the Regulatory Guide 1.95 model for 5 percentile dispersion conditions will be used in the subsequent evaluation of chlorine concentrations that could occur inside the VEGP control room.
12 4210t
r r3 REFERENCES
- 1. Meeting on September 18, 1985 between CPC consultants, Y. J. Lin and Ping Wan; and NRC meteorologists Irwin Spickler and James Fairobent -
to clarify models, formulas, and methodology.
- 2. Regulatory Guide 1.95, " Protection of Nuclear Plant Control Room Operators Against an Accidental Chlorine Release", Revision 1. U.S.
Nuclear Regulatory Commission Office of Standards Development.
Washington D.C., 1977.
- 3. EUREG-0570, " Toxic Vapor Concentrations in the Control Room Following A Postulated Accidental Release," U.S. Nuclear Regulatory Commission, Washington, D.C., 1978.
- 4. D. H. Slade, Meteorology and Atomic Energy. TID-24190, U.S. Atomic Energy Commission, Washington, D.C. (1968)
- 5. Regulatory Guide 1.78, " Assumptions for Evaluating the Habitability of a Nuclear Power Plant Control Room During a Postulated Hazardous Chemical Release".
- 6. NUREG/CR-3786, "A Review of Regulatory Requirements Governing Control Room Habitability Systems", Sandia National Labs, Prepared for the Nuclear Regulatory Commission, Washington, D.C.
- 7. Committee on the Safety of Nuclear Installations Organisation for Economic Cooperation and Development, " Physical and Toxic Properties of Hazardous Chemicals Regularly Stored and Transported in the Vicinity of Nuclear Installations", Nuclear Energy Agency, Paris, France, March 1976.
- 8. Regulatory Guide 1.145, " Atmospheric Dispersion Models for Potential Accident Consequence Assessments at Nuclear Power Plants", U.S. Nuclear Regulatory Commission.
- 9. Bird, Stewart, and Lightfoot, Transport phenomena, Department of Chemical Engineering, University of Wisconsin, John Wiley and Sons, New York, New York, 1960.
13 4210t
<