ML19249F113
ML19249F113 | |
Person / Time | |
---|---|
Site: | Three Mile Island |
Issue date: | 01/31/1972 |
From: | GENERAL PUBLIC UTILITIES CORP., RESEARCH CORP. OF NEW ENGLAND |
To: | |
References | |
NUDOCS 7910100432 | |
Download: ML19249F113 (137) | |
Text
Atmospheric Diffusion Experiments with SFs Tracer Gas at Three Mile Island Nuclear Station Under Low Wind Speed Inversion Conditions 1407 301 4
PO ER U COMPANIES t32 7910100(.y
(/3 ,15 -),cx,, /
,s, ..>
ATMOSPHERIC DIFFUSION EXPERIMENTS WITH SF6 TRACER GAS AT THREE MILE ISLAND NUCLEAR STATION UNDER LOW WIND SPEED INVERSION CONDITIONS Pickard, Lowe, and Associates, Inc.
The Research Corporation of New England General Public Utilities Service Corporation January 1972 1407 002
TABLE OF CONTENTS Section Page
1.0 INTRODUCTION
1 1.1 General 1 1.2 Experimental Procedure in Brief 2 1.2.1 General 2 1.2.2 Experimental Phases 3 1.2.3 Scheduling of Tests 4 1.3 Participants 4 2.0
SUMMARY
OF RESULTS 5 2.1 General 5 2.2 Phase 1 5 2.3 Phase 2 6 2.4 Phase 3 7
3.0 CONCLUSION
S 8 4.0 THEORY AND MODELS 9 4.1 Diffusion Models 9 4.2 Diffusion from a Point Source (Phase 1) 10 4.2.1 The Gaussian Diffusion Equation 10 4.2.2 The Sector Average Equation 11 4.2.3 Directional Frequency Model 11 4.3 Diffusion in the Wake of Large Structures (Phase 2) 12 4.3.1 Building Wake Correction to the Gaussian Equation 12 4.3.2 Sector Averaging with Building Wake Included 13 1407 003
e* s
- TABLE OF CONTENTS (continued)
Section Page 5.0 EXPERIMENTAL TECHNIQUE 15 5.1 General 15 5.2 The Suitability of SF 15 5.3 SF6 Detection Technology 16 5.4 Sampling Equipment 16 5.5 Release Equipment 17 5.6 Sample Analysis 17 6.0 ON-SITE METEOROLOGICAL DATA 19 6.1 General 19 6.2 Temperature Data 19 6.3 Wind Data 19 6.4 Smoke Candles
7.0 DESCRIPTION
OF PHASE 1 TESTS (OPEN FIELD SITE) 21 7.1 General 21 7.2 Summary and Results of the Five Phase 1 Tests 21 7.3 Individual Tests 22 7.3.1 Test 2 22 7.3.2 Test 3 23 7.3.3 Test 4 -
25 7.3.4 Test 5 26 7.3.5 Test 6 27 7.4 Conclusions 29 8.J DESCRIPTION OF PHASE 2 TESTS (REACTOR SITE) 30 8.1 General 30 8.2 Sut. unary and Results of the Five Phase 2 Tests 31 1407 004
. . . +
TABLE OF CONTENTS (continued)
Section Page 8.0 DESCRZDTION OF PHASE 2 TESTS (REACTOR SITE) (continued) 8.3 Individual Tests 32 8.3.1 Test 7 32 8.3.2 Test 8 34 8.3.3 Test 9 36 8.3.4 Test 10 38 8.3.5 Test 11 39 8.4 Conclusions 39 9.0 VERTICAL CONCENTRATION PROFILES 41 9.1 General 41 9.2 Phase 1 Open Field Vertical Measurements 41 9.3 Phase 2 Vertical Measurements in the Building Wake 42 9.3.1 Tests 7 and 8 43 9.3.2 Test 9 44 9.3.3 Test 10 45 9.3.4 Test 11 45 9.4 Phase 3 Time Averaged Vertical Measurements 45 9.4.1 General 45 9.4.2 Test 12 46 9.4.3 Comparison of Maximum Concentration 46 With Models 9.5 Conclusions 47 10.0 MOBILE OFF-SITE TRAVERSES 48 10.1 General 48 10.2 Test 10 Road Traverse 48 10.3 Test 12 Road and River Traverses 49
-111-1407 005
e <
TABLE OF CONTENTS (continued)
Section Page 10.0 MC A > 3FF-SITE TRAVERSES (continued) 10.4 Correlation of Results with Models 49 10.4.1 Test 10 Traverse 49 10.4.2 Test 12 Traverse 50 10.5 Conclusions 51 APPENDIX A A-1 APPENDIX B B-1 REFERENCES 1407 006
-iv-
LIST OF TABLES Table Page 1 Weather Condition Summary Table 52 2 Summary of Models 53 3 Pasquill Stability Classes Based on Wind Data 54 4 Pasquill Ltability Classes Based on Temperature Data 54 5 Positions of Sampling Tanks: Phase 1 55 6 Summary of Basic Test Data: Phase 1 56 7 Concentration Calculations Based on Model IP 57
("AEC/DRL AT Model") : Phase 1 8 Concentration Calculations Based on Model 2P 58
("Slade ce Model"): Phase 1 9 Concentration Calculations Based on Model 3P 59
(" Split c Model") : Phase 1 10 Concentration Calculations Based on Model 4P 60
(" Sector Average Model") : Phase 1 11 Concentration Calculations Based on Model SP 61
(" Directional Frequency Model") : Phase 1 12 Summary of Phase 1 Results 62 13 Positions of Sampling Tanks: Phase 2 63 14 Summary of Basic Test Data: Phases 2 and 3 64 15 Concentration Calculations Based on Model IW 65
(" AT Wake Model") : Phase 2 16 Concentration Calculations Based on Model 2W 66
("Slade ce Model with Wake Correction") : Phase 2 17 Concentration Calculations Based on Model 3W 67
(" Split c Wake Model") : Phase 2 18 Concentration Calculations Based on Model 4W 68
("AEC/DRL AT Wake Model") : Phase 2 19 Concentration Calculations Based on Model SW 69
(" Sector Average Wake Model") : Phase 2
-v-1407 007
a LIST OF TABLZS (continued)
Table Page 20 Summary of Results Using North Tower Data: Phase 2 70 21 Summary of Results Using South 100 ft Tower Data: Phase 2 71 22 Vertical Concentration Profiles: Test 3 72 23 Vertical Concentration Profiles: Test 6 72 24 Vertical Conce.tration Profiles: Test 8 73 25 Vertical Concentration Prcfiles: Test 9 73 26 Vertical Concentration Profiles: Test 10 74 27 Vertical Instanta sous Concentration Profiles: Test 11 74 28 Vertical Average Concentratien Profiles: Phase 3 (Test 12) 74 29 Concentration Calculations Using Wake Models for Test 12 75 30 Summary of Wake hodel Performance for Traverses: 76 Tests 10
-vi-1407 008
a LIST OF FIGURES Figure Pm 1 Lateral Diffusion, c , versus Downwind Distance 77 from Source for Pasq'uill's Stability Classes 2 Vertical Diffusion, o,, versus Downwind Distance 78 frem Source for Pasculll's Stability Classes 3 K Isopleths for Reactor Complex. (a) Downwind 79 Release (b) Top Release (c) Upwind Release 4 Sampling Tank Assembly 80 5 Tracer Gas Release Apparatus 81 6 Strip Chart Record from Gas-Leak Detector 82 7 Three Mile Island Site Area 83 8 SF6 Concentrations and Wind Direction Durations: 84
- Test 2 9 SF6 Concentrations and Wind Direction Durations: 85 Test 3 10 SF 6 Concentrations and Wind Direction Durations: 86 Test 4 11 SF6 Concentrations and Wind Direction Durations: 87 Test 5 12 SF6 Concentrations and Wind Direction Durations: 88 Test 6 13 SF6 Concentrations and Wind Direction Durations: 89 Test 7 14 South Tower Wind Direction Durations: Test 7 90 15 SF6 Concentrations and North Wind Direction 91 Durations: Test 8 16 South Tower Wind Durations: Test 8 92 17 SF6 Concentrations and North Wind Direction 93 Durations: Test 9 18 South Tower Wind Direction Durations: Test 9 94
-vii- ,
LIST OF FIGURES Figure Pa,3,e 19 SF6 Concentrations and North Wind Direction 95 Durations: Test 10 20 South Tower Wind Durations: Test 10 96 21 SF6 Concentrations and North Wind Direction 97 Durations: Test 11 22 douth Tower Wind Durations: Test 11 98 23 Building Structures Profile as Seen from the 99 North and West 24 dalloon Locations and North Wind Direction 100 Durations: Phase 3 (Test 12) 25 South Tower Wind Direction Durations: 1 01 Phase 3 (Test 12) 26 Average Vertical Concentration Profiles: 102 Phase 3 (Test 12) 27 Concentrations Along Rt 441: Test 10 Road 103 Traverse 28 North and South Tower Winds: Test 10 Traverse 104 29 of f-site Downwind concentrations: Test 12 Road 105 Trave rs e 30 off-site Downwind Concentrations: Test 12 River 106 Traverse 31 North and South Tower Winds: Test 12 Trave.ses 107 32 Comparison of Measured Downwind Concentrations 108 with Model SW (" Sector Average Wake Model")
A-1 Example of Work Sheets Usea to Compute Tank A-7 Concentrations b-1 Example of Wind Speed and Direction Charts: B-4 North Tower 3-2 Example of Wind Speed and Direction Charts: B-5 South Tower
\h0
-viii-
LIST OF FIGURES Figure Pg B-3 Example of Wind Speed and Direction Charts B-6 from the 30 ft Weather Measure 3-4 Example of 150 f t-25 f t AT Strio Chart B-7 1407 011
-ix-
SECTION 1.0 INTRODUCTION ,,
1.1 General Weather data collected f or several years at the Three Mile Island Nuclear Station site indicate considerable meander during low wind speed conditions. These conditions, which occur about 5% of the time, are ob-served on the wind direction recordings primarily during nighttime. Since many of these data are taken at very low wind speeds, near the threshold of the direction _ vane, it is important to determine whether the meander ob-served is truly representative of the wind conditions or is just an inac-uracy of the measurement due to characteristics of the vane.
Because of this potential inaccuracy, it has been the practice in ecent reactor licensing cases to assume poor dif f usion conditions both vertically and horizontally when these conditions exist during low wind speed conditions at nigh t. Since it is required that the design basis acci-dent diffusion conditions be such that site boundary concentrations are exceeded no more than 5% of the time, theae conditions (which are usually assigned Pasquill F or G stability) play a significant role in the determin-ation of accident dif fusicn es timates ,
if the meander is real, however, the amount of effluent which would reach a stationary receptor during an accident would be considerably lower than the amount computed using a stable plume codel in the Gaussian equa-tien, as is customary in reactor licensing cases. It is the intent of this experiment to determine whether concentrations can be predicted conserva-tively but accurately during low wind speed inversion conditions, using meteorological data collected at the site. If the prediction procedure is validated, it is expected that these wind conditions would not appear among 1407 012
those which yield the highest 5% of predicted concentrations.
The SFe gas tracer experiment discussed in this report was designed to accurately measure average concentrations during nighttime inversion and low wind speed conditions during a 45-minute release of tracer gas. Measure-ments were made in the " free field" and in the vicinity of large plant struc-tures. Results are compared to predictions made using various models to determine the appropriate model for use with the Three Mile Island site weather data.
1.2 Experimental Procedure in Brief 1.2.1 General The basic procedure involves the continuous release of a suitable tracer substance under controlled and monitored ditions at a given point on the site. This substance is then collected at constant rates in evacu-ated tanks for at least a 45-minuce time period at various locations around the release point.
The requirements for the tracer were that it had to: (1) be essential-ly inert with respect to the environment; (2) be nontoxic, (3) have a low background in the site area; (4) be noncendensable; (5) be nonparticulate; (6) be non-buoyant in air; and (7) be detectable in poor visibility condi-tions at very low concentracions. Sulfur hexafluoride (SF 6) at very low concentrations meets all these requirements, and a detector having a thresh-old sensitivity of less than one-tenth of a part per billion is commercially available.
Each tank assembly consisted of an evacuated 16 liter chamber fitted with a vacuum gauge and constant flow rate apparatus. Gas samples were collected by permitting the tanks to fill simultaneously for a fixed period
_2- 1407 013
of time. The tanks were returned to the field laboratory still partially evacuated, and then brought to ambient pressure by admitting clean bottled air. The resulting SF6 tank concentrations was then measured with an Analog Technology Corporation Tracer-Gas Leak Detector. This detector was also used to measure elevated samples and to measure off-site downwind con-centrations during mobile traverses.
Wind measurements were made at three locations during each test, and vertical temperature difference v,i measured at one location. Supplemen-tary wind information was also obtained by releasing smoke candles and recording visual observations during the test.
1.2.2 Exnerinental Phases The experimental program included three phases. In Phase I, an open field was used, with the tracer being released as a " point source" in the center of a 300 foot radius circular grid with 18 sample tanks spaced around the circumference. Elevated samples were also taken using a balloon system. The purpose was to measure concentrations at the samplers without the influence of building obstructions. Five tes;s were conducted in Phase Phase 2 involved the release of gas in the wake of large structures already present in the nuclear plant complex. Releases were made near the Unit 1 containment building at or near grade level. These tests simulate as close as practicable the actual conditions that would exic. in the event of the accidental release of radioactive material during periods of low wind soeed, inversion conditions. Ideally the sample tanks should have been located at 2000 ft, which is the site exclusion boundary; however, the presence of the river made this impractical. The tanks were located as far 1407 014 as possible from the release point, limited by the river's edge and other structures, 14 approximately an 800 ft circle. Vertical concentration pro-files were again measured usine a balloon during each Phase 2 test. F ce tests werc conducted in Phase 2.
The Phase 3 test was dedicated to collecting time-integrated eleva ed samples using four helium filled balloons to support tubing which terminated at several elevations. Integrated samples over a 45-minute period were col-lected in evacuated bottles connected to the tubing at grade level. Mobile traverses off-site were conducted during two of the tests in Phases 2 and 3.
1.2.3 Scheduling of Tests Daily weather forecasts prepared specifically for the site by a private weather service were used for scheduling experiments. Personnel were put on alert if 'he prediction was for near calm conditions. The final decision for a test was made just prior to the scheduled time. Table 1 lists the test dates and summarizes the weather conditions.
1.3 Participants This experiment was funded by the Metronclitan Edison Company (Met Ed),
Pennsylvania Electric Company and Jersey Central Power and Light Company, who are the owners of the Three Mile Island Nuclear Station. On behalf of General Public Utilities Service Corporation (GPU), Pickard, Lowe and Associ-ates, Inc. (PLA) directed the project in consultction with Dr. James Halitsky (University of Massachusetts). The Research Corporation of New England (TRC) was retained to develop the techniaues for, and to conduct, the field measure-ments. Project planning, experimental design and data analysis was provided by Keith Woodard (PLA), Dr. Halitsky, George F. Collins (TRC), and George Kunder (GPU). The Metropolitan Edison Company, who will operate the nuclear station, provided additional personnel and equipment in the field where neces-sary.
1407 015
SECTION 2.0 SLTIARY OF RESULTS 2.1 General All tests were conducted during low wind speed inversion weather con-ditions, and the expected meander was observed during most of the tests.
Measurements of pertinent weather parameters made at several locations on-site during the tests served as input to a series of models developed for predicting the tank concentrations. The "best" models were found to be those which accounted for the wind meander. Comparison of results with the model commonly used in reactor licensing showed the licensing model was very conservative. Following is a discussion of the results for each phase.
2.2 Phase 1 Table 1 summarizes the weather conditions during the Phase 1 tests. In almost all cases the tracer was detected over more than a 150 arc, demon-strating that the meander recorded on the wind instrument was real. In general, locations of the naximum concentrations corresponded to measure-ments of wind di. _.ian persistency during the tests.
Five point source or "P" models described in Section 4.2 were cou-pared for each test. A summary of the Phase 1 model predictions versus measured concentrations is included in Table 12. The model which is com-monly used in licensing cases, referred to as Model IP ("AEC/DRL AT Fbdel"),
over-predicted the concentrations by an average factor of 21. The best
_5_
1407 016
model, Model 4P (" Sector Average Model"); which takes into account the meander effect, had an average X f 1.27 for the Phase 1 max/Xmodel rati series. This constitutes excellent agreement for a diffusion study of this nature. Model 3P (" Split c Model") was the next best (conservative) per-former, while Model 2P ("Slade og Model"), underpredicted (non-conservative) by almost a factor of two for these weather conditions.
One further comaprison is of note. Using the AEC Safety Guide 4 mete-orology of Pasquill F and 1.0 m sec-1 in the standard Gaussian equation (Model IP), the results of Phase 1 tests (which had wind speeds less than 1.0 m sec-1) were overpredicted by an average factor of 5.8.
2.3 Phase 2 As shown in Table 1, low wind speed inversion conditions prevaile. for tests conducted in the vicinity of the reactor building complex. Phase 2 measured concentrations were considerably lower than in Pi.ase 1 due to the aerodynamic turbulence of the buildings e.nd greater distance to the samp-1ers. The meander effect was observed in several tests and contributed, along with the building wake effect, to the very low measured concentra-tions.
Five building wake or "W" diffusion models were tested in Phase 2 (see Section 4.3) with the results being more difficult to correlate. In two of the tests (Tests 7 and 8) the models predicted sample concentrations very well; however, in the remaining three tests, the "best" models over-estimated concentrations by a substantial factor. The "best" model was, however, the one which accounted for the meander condition as well as for the wake effects.
It is believed, as discussed later in this report, that for several 1407 017
tests the building effect caused the maximum concentrations to occur above the tanks during very stable conditions. This belief was substantiated by qualitative visual observations of smoke plumes where it was commonly noted that, despite the low wind speed inversion conditions , the smoke was initi-ally transported vertically in the region of the reactor building, and or.ly af ter gaining considerable altitude did the plume acquire a horizontal tra-jectory. Quantitative evidence of this plume behavior was provided by a series of instantaneous vertical profiles of SFb cc ncentrations measured from the ground to a height of 200 f t which show that the gas is initially distributed vertically by the building wake.
2.4 Phase 3 decause it was suspected that the maximum concentrations may have occurred above the ground samplers, it was decided to obtain time integrated samples of the vertical plume concentrations. This was accomplished in the single Phase 3 test where average concentrations at several elevations up to 250 ft were measured at four radial locations on the circumference of the 800 ft grid. Results showed a marked increase in concentration with height.
However, the maximum measured concentration was lower than the predictions of all models.
Since the n, ximu:. concentration was alof t at a radial distance of 800 ft, it was considered necessary to determine concentrations at the site boundary and beyond. To accomplish this, a series of mobile traverses were made generally downwind to a distance of three miles. These results further validated the model proposed for use during low wind spced inversion conditions at Three Mile Island.
1407 018 SECTION
3.0 CONCLUSION
S The diffusion of the sulfur hexafluoride tracer gas over flat terrain within the valley and in the turbulent wake of large structures during low wini speed inversion conditions was shown to be satisf actorily described by mouels which account for plume meander. Model IP ("AEC/DRL AT Model") which is the common Gaussian eor.cion (corrected for wake effects) with Pasquill sta-bility categories based on vertical temperature structure, is overly conserva-tive during these conditions. This experimental program validated the use of Model SW (" Sector Average Wake Model"), which is a more appropriate, yet con-servative, model for prediction of diffusion during periods of Icw wind speed inversion (nighttime) conditions.
1407 019
SECTION 4.0 THEORY AND MODELS 4.1 Diffusion Models The purpose of this section is to test various diffusion models for comparison with experimental results. In particular, it is desired to find an capression which predicts x/Q: where ,, is the downwind maximum ground level concentration for a given release rate Q. This expression should be based in weather parameters collected during the test in the same manner as in the site weather program so that the expression can be applis' using previously recorded data.
Many models have beer, developed in the past by experimenters for the prediction of diffusion. These are based on wind speed and some indication of atmospheric turbulence. The predictors of turbulence have included time of day, radiation to or f rom the ground, horizontal and/or vertical wind fluctuations, combinations of speed and horizontal fluctuations, and verti-cal temperature difference. In this section, models based on wind speed, wind direction fluctuation and vertical temperature difference are examinea.
A summary of all models considered in this study is included in Table 2.
Diffusion models which are generally accepted employ a Gaussian equa-tion in which atmospheric turbulence is expressed in terms of the standard deviations of plume concentrations both vertically (using o ) and horizon-Curves which give o and c as a function of distance cally (using ay). v z for six Pasquill diffusion categories are shown in Figures 1 and 2. The G categorv used by the AEC/DRL ic represented by a curve located the same dis-tance below F as E is above F or. the curves.
Several methods have been proposed for selection of the appropriate curves. Slade, in Meteorology and Atomic Energv , l proposed the use of the
- 1407 020
standard deviation of azimuthal wind direction angle ( 0) according to Table 3. For licensing cases, AEC/DRL has proposed the use of vertical temperacure difference (AT) alone for selection of dif fusion categories as given in Table 4. Others have used a combination of og and AT.
Models selected for this study are described below.
4.2 Diffusion from a Point Source (Phase 1) 4.2.1 The Gaussian Diffusion Equation For a ground-level continuous point source, the equation which yields centerline ground-level concentrations is:
0 x =
(1) eua c where X is the centerline surf ace concentration (in parts per part), Q is the source strength (release rate of SF6) in m3 sec -1, u is the average horizontal wind speed in m sec-1, and yo and cz are the horizontal and vertical dispersion coefficients from Figures 1 and 2 respectively, in meters.
Equation (1) is used to obtain the first three point source models (designated P) as follows:
Model iP ("AEC/DRL LT Model"): Uses equation (1) with the AT groups given in Table 4 to define both o and o z.
y Model 2P ("Slade 6 M del"): Uses equation (1) with the a, groups given in Table 3 t6 define o and o .
y z Model 3P (" Split c Model"): Uses equation (1) with the AT groups of Table 4 and the ce groups of Table 3 to define c z and ay respectively.
1407 021
4.2.2 The Sector Average Equation If equatici (1) is integrated over all values of y and divided by the length ot arc of angle o (in radians) at distance x, we obtain the average concentration along the arc of:
( 1l2
_ Q f -
(2) ue xv The tourth model is defined as follows:
Model 4P (" Sector Average Model"): Uses equation (2) with 0 equal to the maximum wind direction meander (range) during the saLple period, and c based on the AT groupE of Table 4.
4.2.3 Directional Frequency Model Model 4P assumes that the plume meanders within the boundaries of the measured angular range. However, a variation of this maximum range model, also applicable to long-period stationary sampling, has been suggested.
This fifth model assumes that the sampler is in the direct path of the plume for some f raction of the total sampl;ng interval and that the remain-der of time the plume makes no contribution to the measured concentration.
The fraction of time is taken from the recordings of wind direction during the test interval. The relevant equation for the peak concent ation is equation (1) modified by f, where f is the highest frequency of wind in any 10* sector during the test. The equation is as follows:
x, =
(3) ue,c Q22
The fifth model is specified as follows:
Model iP (" Directional Frequency Model"): Uses equation (3) with e and a based on AT in Table 4.
4.3 Diffusion in the Wake of Large Structures (Phase 2)
When material is released near large structures, the turbulence created by the wake provides additional mixing. Tests in wind tunnels have formed a basis for model development; however, few field experiments have been run, especially during low wind speed inversion conditions. Based on wind tunnel tests and theory, the following models were considered for comparison with the results of Phase 2 tests.
4.3.1 Building Wake Correction to the Gaussian Ecuation.
To account f or building wake ef f ects, equation (1) is modified as follows:
x = (4) u (rcy cz + cA) where c = the shape f actor, and A = the cross sectional area of the building (m2 ),
The AEC in Saf ety Guide Number4 4 uses the same relationship with a shape factor c = 0.5 and building area equal to the smallest vertical cross section area of the reactor containment building. The AEC also imposes the restriction that the concentration calculated according to equation (4) not be less than 1/3 of that calculated according to equation (1).
There is some question as to whether c should be assigned the value of 1407 023
0.5. Dr. J. Halitsky has published isepleths of non-dimensio-alized concen-tration (K ) based on wind tunnel experiments with a typical .or complex as described in Meteorology and Atomic Energy,2page 251 ana re. ..oduced as Figure 3 herein. K is related to the shape factor c by:
1 v z c = g A
C Using a sampler distance of 800 ft and a containment diameter of 150 ft, the non-dimensionalized distance, from the center of the containment to the sam-pler is 5.4 Therefore, from Figure 3, the value of K was about 0.5 value at the samplers. The turbulence intensity in the wind tunnel at the mid height of the model was about 4%, corresponding to a Pasquill F stability. Therefore, o,e at the sampler was about 50 m2 . Inserting.these values into equation (5) with a TMI containment area A = 2000 m2 yields e es 2.
This leads to two variations to the point source model forming four wake models (designated with a W) as follows:
Model 1W ("6T Wake Model"): Uses equation (4) with cA based on area of containment with c=2, and 4T groups in Table 4.
Model 2W ("Slade e, Model Uses equation (4) with cA based with Wake Correcti6n"): on area of containment, c=2, and og group in Table 3.
Model 3W (" Split a Wake Uses equation (4) and both Tables Model"): 3 and 4 as for Model 3P, cA based on area of containment and c=2.
Model 4W ("AEC/DRL AT Wake Uses equation (4) with cA based Model"): on containment area, c=0.5 and 4T from Table 4. Maximum wake correc-tion is 3.
4.3.2 Sector Averaging with Building Wal.e Included Since the meandering wind effectively disperses the released matcrial 1407 024
in a wide area, a model was developed which would take both the building wake and the meander into account. The basic equation in a Gaussian form was ug-3 gested by Davidson and appears on page 112 of Meteorology and Atomic Energv as follows:
Q v2 h2 x = exp - + ,
. (6) ue I I 2I 2 2I 2 y z y z For ground level releases h = 0. If this is integrated over all values of y and divided by the sector arc length 0 at the sample distance x as in equation (2) above, the equation becomes:
o(2):,2 x =
, (7) uI x0 where I = a 2 , 92 g and 0 =
maximum wind direction meander (radians). ,The fifth building wake model is defined as follows:
Model SW (" Sector Average Wake Model"): Uses equation (7) with c=2, A=2000 and a from Table 4.
1407. 025
SECTIO:s 5.0 EXPERIMENTAL TECH'iIQUE 5.1 General The first use of sulphur hexafluoride (SF6 ) as a gas tracer took place at the Connecticut Yankee Atomic Power Plant in 1963 and 1964 (Collins et al, 1965)S. At that time elaborate laboratory analysis was required to measure the collected sample concentration. The gas chromatographic technique used was refined and reported on again by Turk et al, 19686 . The study at Three Mile Island introduces significant improveme,ts in both the sampling and analytical techniques.
The SFc sampling and analytical procedure is described in detail in Appendix A. Following ic background information attesting to the suitability of SF6 as a tracer gas.
5.2 The Sultability of SF6 As discussed in Section 2, SF6 is ideally suited to gas tracer experi-ments. To quote Turk et a1 6 "SFg is particularly useful because it is amenable to ultra sensitive analysis by electron-capture detection, is con-venient to handle and dispense into air, is odorless and noa-toxic, is chemi-cally and thermally : table, and does not usually occur in significant concen-trations in outdoor air". Additional advantageous properties are its low solubf.lity in water and its density which is somewhat heavier than air.
The Three Mile Island site environs were thoroughly checked for SF6 background concentrations. It was found that some switchgear was leaking which was repaired. After this repair, no further problem with background was encountered.
1407 026
5.3 S.f.6 Detection Technology The principal of electron capture is briefly described by Collins et a1 , 5 The measurement of SFE concentrations was recently vastly simplified by the introduction on the market of a portable tracer-gas leak detector manufactured by Analog Technology Corporation. This instrument, accore ng to the manufac-turer, responds to SFg with a sensitivity of greater than one part in 101k.
The Model 112B has a continuous sampling mode and a discontinuous or colum-nar mode for high-sensitivity measuremen:s. The latter mode was used for tank. sampling in this study. Two such instruments were on-site during each test.
The availability of this gas-leak detector greatly facilitated the measurement of sample concentrations since san.ples could be analyzed direct-ly by inserting the detector probe into the collection tank. It was found that concentrations could be relisbly measured ovar a range of values from 0.01 to about 3000 parts per billion. This was verified by shop calibration using standard SF6 sources prepared especially by the laboratories of Axton-Cross.
An additional feature is that the battery-operated instrument could be utilized to measure real-time instantaneous concentrations of SFE present in the air. This feature was used in the field to locate the plume, to conduct mobile. traverses, and to obtain both vertical profiles of SF6 c ncentrations.
5.4 Sampling Equiement The sampling tanks used were first described by A. Turk6 . However, it was found that more reliable measurements could be taken if a vacuum gauge were installed at the end opposia.e to the constant flow controller (see Appendix A). The complete tank assembly is shown schematically in Figure 4.
1407 027
The sampling method was improved after several trial runs to achieve a calibrated flow on each tank of*0.2 liters min-1 which would provide a 45 minute sample interval. Twenty such tank assemblies were available for each test. In both Phases 1 cno 2, cighteen of these tanks were arranged as close as possible to the circumf eret.ce of a circle and were spaced 20*
apart.
5.5 Release Equipment The SFs, which was released at the center of the grid, was discharged at a constant rate through a rotameter. A schematic of *.he release system is shown in Figure 5. The gas was released for about 15 minutes prior to sample tank opening to establish steady-state cenditions. Samplers were all opened simultaneously at the established time. ihe rotameter was con-tinuously checked by the person assigned to the release point, and periodic entries were made in a log including comments on general weather conditions, the measured flow rate of gas and the trajectory of smoke f rom smoke candles released periodically during the test.
5.6 Sample Analvsis After the tanks were shut off, they were returned to the field lab, brought to a pressure of 1 atmosphere, and sampled. Accurate determination of the sa;:nple volume was made, as indicated in Appendix A, for use in accur-ately finding the sample concentration.
The gas detectors could be set at any one of four sensitivity ranges.
The output was tead in volts, and then converted to concentration in ppb from calibration curves . An example of the gas analyzer trace produced by the strip chart recorder is shown in Figure 6. The voltage produced varies linearly with the sample concentration sucn that only one known concentra-h
tion of SF6 was required to draw the calibration curve. When possible, s second contentration was used to increase the reliability. Each of the recorders was calibrated on all four ranges at the TRC labs using standard concentration sources.
1407 029 SECTION 6.0 ON-SITE METEOROLOGICAL DATA 6.1 General Measurements of pert 'aent meteorological parameters were recorded at several locations on the Three Mile Island site throughout the test. These data were redaced for periods coinciding with the tests and the results are summarized in Table 1 and in Aopendix B. Data reduction techniques are also discussed ._n Appendix B. Details of the monitoring program are discussed below.
6.2 Temperature Data There are two permanent micrometeorological towers installed on the site, one north and the other south of the power plant (see Figure 7). The north tower is instrumented with thermistors housed in Geotech aspirated radiation shields which provide values of the temperature difference between the 150-foot and the 25-foot levels. The average value of AT during each test, expressed in degrees Centigrade per 100 meters, served as the basis for determining the Pasquill Stability Category f rom Table 4 A typical recor-der trace of the temperature difference data is shown in Appendix B.
6.3 Wind Data Wind speed and direction were recorded at three locations during the tests. For the Phase 1 tests, a Weather Measure Model W1034-540 low thresh-old recording wind system with three-cup anemometer, 0-10 mph full scale speed range, and a 0-540 degree directional range was used. This instrument was installed to record the wind speed and direction at the 30 ft level above the SF6 release point (see Figure 7) . The Weather Measure instrument yielded the direction data which were used in conjunction with Table 3 for 1407 030
Phase 1 models.
For Phases 2 and 3, the wind data were obtained primarily from the 100 ft level on the north tower. This information was supplemented by wind records from the 100 ft level on the south tower. Data from both towers are summarized in Appendix B. Both tower locations are shown on Figure 7.
The equipment used for these installations are Beckman and Whitley short vane anemometers with a starting threshold of 0.6 mph; the recorders are manufactured by Esterline-Angus.
Examples of typical recorder traces of the wind data and a description of the data reduction appears in Appendix B.
6.4 Smoke dandles It was also found that smoke candles were an important source of addi-tional wind information, yielding not only surface wind direction, but also providing some qualitative indication of plume behavior. For example, dur-ing the Phase 2 tests in the building wake, the smoke exhibited a marked tendency to be transported vertically; candles ignited at the reactor build-ing base (i.e., at the SF6 release point) often revealed the presence of wind currents which carried the smoke upward along the reactor wall and over the roof. Where relevant, the observers' comments of the smoke plume behavior are included in the discussions of the individual tests.
20 1407 331
SECTION
7.0 DESCRIPTION
OF PHASE 1 TESTS (OPEN FIELD SITE) 7.1 General In this portion of the study the site consisted of a relatively level plowed field south of the construction area. The 18 sampling tanks were sus-pended from stakes five ft above ground arranged approximately in a circle centered on the 30 ft Weather Measure tower which was the SF6 release point.
The location of the Weather Measure is shown in Figure '.i The exact positions of the sampling tanks relative to the release point are given in Table 4.
7.2 Sammary and Results of the Five Phase 1 Tests Table 6 is a summary of the basic test data including observed and derived weather parameters from the 30 ft tower. These data were used in the prediction calculations shown for Models IP through SP in Table 7 through 11 respectively. Wind direction range divided by 6 was used to predict a,)
which is conservative compared to the calculated values of o also g shown in Table 6. Table 12 summarizes and compares the results of each model with the Phase 1 test results. Note that Test 1 served as a shakedown test and did not yield meaningful data.
Figures 8 through 12 inclusive are polar projections centered on the SF6 release point for each of the 5 tests. The " range rings" are designed to be used with three different scales. As a distance scale, they indicate the location of the sampling tanks from the release point. Their second function is to indicate the concentrations of SF6 in ppb measured in the collected samples. Note that this is an exponential scale, with marked values of 100, 400, 900 and 1600 pob. In essence, the concentrations are 907 032
depicted in a bar-graph form where each bar extends a radial distance out-ward from the tank position corresponding to the concentration. Note that, because the sample tanks are not located on a perfect circle, the concen-tration around the release point is indicated bv the end of the bar and not the length of the bar. The increasing width with length of the concentra-tion bars are intended merely to dramatize the variations in the values, and have no quantitative meaning.
The third use of the concentric rings is to provide a linear scale for the time duration,in minutes, of each wind direction during the course of the sampling time. These data are indicated by straight lines which emanate from the center. Angular resolution was chosen to be 10*. Solid segments indicate that the corresponding wind speeds were measurable by an anemometer, while dashed nortions are used for periods of calm. Dotted lines denote data inferred froa vi sal observations of smoke candle plumes and the wind vane.
The bearing at which a given wind duration is drawn indicates the direction from which the wind was blowing.
7.3 individual Tests
- 7.3.1 Test 2 Jn the basis of the vercical temperature profile, Test 2 was the most stable test in Phase 1 with an observed increase in temperature with height of 4.26 C/100 m. liowever, as was the case for all five tests in this phase, the wind raeandered through more than 150' which would indicate there was considerable lateral spread.
liigh es t sample concentrations were measured at tank positions 13 (260*),
14 (280*) and 15 (300*) where the values were 2610, 1797 and 840 pob respec-tively (see Figure 8). This is in good agreement with the wind direction 1407 033
data. Both the 120* and the 100* duratiens of 6.7 min and 8.3 min respec-tively fit the concentrations well at positions 14 and 15. The highest con-centration, which was at oosition 13, was downwind from the 80* wind for only 1.7 min. However, if the 90* wind duration is added to this value, the total duration becomes 10 min. The overall average speed was 0.62 m sec-1, the highest of any test in this chase.
High sample concentrations were measured at all positions between bearings of 200* and 20*. The average sample tank concentration for this sector was 711 pub, and the maximum tank concentration was a factor of 3.7 higher. This quantity is referred to hereafter as the peak-to-mean ratio. Although this was the highest for this phase, it must be noted that (as shown in Table 5) the sample grid for this test was not a uniform circle as for the remainder of Phase 1 tests.
Considering the prediction models for downwind concentration (Table
- 12) Model IP ("AEC/9RL /T Model") fit tne data poorest of all with a value for x /x model f 0.096, while Model 3P (Solit a Model") gave the best agreement with a value for the ratio of 0.7.
7.3.2 Test 3 This test took place with a lapse rate of 2.97 C/100 m and again the direction meandered through more than 165*. Although the highest concen-tration, 1788 onb, was measured at position 17 (340*), concentrations of over 100 pob were measured in all the samples between bearings of 100* and 20* inclusive, a range of 280* (see Figure 9). This is in qualitative agreement with the high variability of the horizontal wind direction, which ranged from 60* to 235* , but does not explain the concentrations measured at positions 4 through 11. The wind record indicates no compone.nt toward T407 M4
8 .
- these positions. However, the longest duration was 10.2 min for the 106' wind, which is in direct line with the peak concentration.
Careful examination of the wind record indicates that, of the total sampling period of 50 nin, a calm condition prevailed for about 36 min.
Thus, usefulness of the corresponding measured wind vane directions, shown by dashed radials in Figure 9, is in question. Transcribed below are the notes made by field observers on the behavior of smoke plumes released at the test site.
Time (EDT) Observed Smoke Plume Behe vior First Observer Locstion: SF6 Release Point 0430 Toward W.
0445 Toward NW.
0500 Very little movement.
0505 Drift touard C, then toward SE.
Previous smoke puff did 180* turn.
0525 Toward W, NW.
Srcond Observer Location: North of Release Point 0425 Drifting to W.
0448 Plume rose upward and with very slight drift to N, not very far.
0452 Smoke sitting just N of SF6 release point, estimated height 100-150 ft.
Smoke cloud not moving.
0456 Plume rises to about 25 ft, then drifts very slowly N.
0506 Smoke drifting slowly to SE.
0517 Drifting to W.
0525 Plume drifting W.
0530 Plume sitting about 75 ft W of SF6 release point.
~ '-
1407 0 %
s , s These co:xnents tend to explain the concentrs. tion pattern on the polar diagram.
Turning now to the models, the maximum observed concentration of 1788 ppb yielded a Xm /XMd f 1.28 f r '4cdel 4F ("Sec;or Average Model") . This iaodel showed best agreement, not only in Test 3. but in Tests 5 and 6 as well.
7.3.3 Test 4 The lapse rate was again stable at 2.83 C/100 m and the meander of 175* was the largest measured. Based on the previous discussion cf meander effects,significant concentrations would be expected to occur at many samp-ling points, and this was indeed the case (see Figure 10). With the excep.-
tion of location 16 (81 ppb), values exceeded 100 ppb between bearings of 120* and 340* . This high-concentration are was 60* narrower than for Test 3, but the sector peak-to-taean concentration here was somewhat lower at 2.7 than for Test 3 which had a value or 3.4 The two liighest concentrations were at positions 8 (567 ppb) and 10 (430 ppb). These are reasonab1v well accounted for by the winds at 350*
(6.0 uin duration) and at 360* (13.5 min). Ilowever, there were a number of sample concentrations (locations 18 throu;;h 7) for which there were no corresponding winds. The average wind speed was only 0.19 m sec-1 and for 27 min out of the 45 min sampling period the wind speed was calm. Thus, there is reason to suspect the corresponding recorded wind directions. The record of smoke candle observations which follows does indicate the presence of a westerly wind on at least one occasion.
g hb
Time (EDT) Observed Smoke Plume Behavior Location: SF6 Release Point 0340 Straight up. Drifted a little to the E. Clear skies , but considerable low fog.
0410 Towards E at surface. Rose about 45 ft, turned to SW.
0428 Smoke right on ground.
0445 Dense fog.
The peak concentration predicted by Model IP ("AEC/DRL AT Model") was very much higher than observed, with Xmax/Xmodel " * *
- provided by Model 2P ("Sladego Model") with a ratio of 0.64. This was the only test for which this equation was best; however, Model 4P was very close at 0.6.
7.3.4 Test 5 This test was conducted under less stable conditions than others with a lapse rate of 0.7 C/100 m. The wind again displayed considerable mean-der over 167* of arc.
The polar diagram (Figure 11) reveals notable concentrations in a narrow sector boundei by positions 8 and 12 with the peak of 390 ppb at 220' (position 11). The peak-to-mean ratio was 3.4.
The prevailing wind directions sre well correlated with the high measured concentrations. The overall average wind soeed was 0.15 m sec-1, the lowest of all the tests. Supplementat ;acke candle observations follow.
= ,
- Time (EDT) Observed Smoke Plume Behavior First Observer Location: SF6 Release Point 0355 Drift toward NW to location 15.
0407 Drift toward W, WSW.
L '4 2 J Modtrately strong SW wind. Smoke moved rapidly toward SW.
v 15 Dead calm.
Second Observer Location: Near Tank 15 0408 Drifted slowly southward.
0428 Drifted to SW quadrant. Winds have been quite calm since 0400. From 0428 to 0438 winds toward SW at less than 1 mph.
Considering the prediction models, Model 4P (" Sector Average Model")
was "best" wich X max /Xmodel " *
- 7.3.5 Test 6 As was the case in other tescs, conditions were stable with a lapse rate of 2.05 C/100 m, and again the wind meandered through about 160'.
This last test in Phase 1 yielded a fairly uniform concentration pat-tern in a 180* sector bounded by tank positions 10 and 1 (see Figure 12).
The peak-to-mean ratio was 2.5, the lowest thus far.
The wind direction durations in the polar diagram agree well with the concentration bars. The dotted radials represent visual observations of smeke candle plumes , when recorded winds were questionable. The average wind speed was 0.37 m sec-I. Observations made during the test follow.
- 2 7-g
Time (EDT) Observed Smoke P:ume Behavior 2020 Drifted to SE, then to SE to N. 2100 Smoke drifts due W for about 30 ft, then to SW. 2102 Smoke goes uoward for about 10 ft, then drifts to S. Cups on tower not moving. 2108 Plume still intact and standing just SSW of tower. 2112 Smoke plume is on road W of tower and drifting N. 2120 Drifting W. Winds calm on top of tower. Plume drifted out over river. 2135 Drifting N. 2139 Drifting to WSW. 2143 Drifting into tower area. Smoke odor is detectable. Winds on tower still remain calm. 2210 Plume moving to SW. The peak concentration of 534 ppb was almost exactly predicted by Model 4P (" Sector Average Model") . The ratio of x , /y- del = 1.02 was the best obtained in the entire study. This good agreement may be explained by study-ing Table 10 in conjunction with Figure 12. The value of 9 used in the Sector model was 162* which almost agrees with the 180' sector width through which significant concentrations were measured. Furthermore, this model assumes a uniform dispersion of gas throughout the sector. As was noted earlier, Test 6 did indeed have a very low neak-to-mean ratio. 1407 039
7.4 Conclusions For each test the peak measured cencentration was compared to that predicted by five diffusion models. Best agreement was obtained using Model 4P (" Sector Average Model") which takes into account plume meander. This is consonant with the pre;eding discussion. Model 4P had an average value of the ratic x /X model f r this phase of 1.27. The average value for Model IP ("AEC/DRL AT Model") was 0.046. It thus may be concluded that, although very stable conditions accompanied by low wind speeds tend to inhi-bit the rapid dispersal of contaminants, the wide fluctuations in the hori-zontal wind direction cannot be neglected. It is of interest to compare the measurements with the meteorological conditions specified fr the first 8 hours in the AEC Safety Guide 4, i.e., Model IP with Pasqt.ill F Dif fusion and 1.0 m sec-1 wind speed. The ratio Xmax//model f r this case varied between 0.06 and 0.31 with an average of 0.17 (see Table 12). Thus, the AEC model overpredicts the measured con-centrations b; an average factor of 5.8. Looking briefly at the other models, Model 3P (" Split c Model") is a c.ignificant improvement over Model 1P, but even here the concentrations are overestimated by a factor of more than 3. Model SP (" Directional Frequency Model") performed about the same as Model 3P (" Split c Model") . On the other hand, Model 2P ("Slade cg Model")-which assumes the same value of the Pasquill stability category for m asz used for cr (based y on Range /6 per Table 3), underestimated the downv..id concentrations by about a factor of 2. Although Model 2P underestimated concentrations for these tests, it is not considered that it would behave in this manner for higher winds or in craes where there is no meander. 1407 040
SECTION
8.0 DESCRIPTION
OF PHASE 2 TESTS (REACTOR SITE) 8.1 General The release point for Phase 2 testinn was noved near the large struct-ures couprising Unit 1. The purpose of this arrannement was to attempt to sinulate, and hence assess the ef fects of, an accidental release of radio-active naterial. The Reonetry of the pcwer plant site plays an in-portant role in diffusing released naterial as will becone evident in the discussion conparing the results of this phase to those of Phase 1. The sanpling tanks were now substantially further f rom the SF6 release point than in previous tests. Distances fron the center of the reactor building are provided in Table 13. Because t.he radial at a bearing of 60* (position 3) passes through cooling tower B, an additional tank, pos-ition 3A, was added (and is indicated in all future polar projection figures). Note that, in Phase 2, the distances of the sampling tanks from the center of the grid were not the sane as their distances from the SF6 release point. The grid center was the center of the Unit 1 reactor building, while the tracer gas release point was variable from test to test. The distance used in the nodels was always taken from the center of the reactor building. Because of the larger radius of the sanpling grid, t .ke desirability of depicting the SF6 release point and the balloon location (used for vertical profiles) relative to the structures, a slightly nodified polar display was used to present Phase 2 results as shown in Figures 13, 15, 1407 04I
. =
17, 19 and 21. The radial distance scale is decreased; however, the scales for the sanole concentrations and the wind direction durations rernin unchanged f ron those used in Figures 8 through 12. Direction data fron the south 100-foot tower are included in Figures 14, 16, 18, 20 and 22 which follow the polar display for each test. 8.2 Su==arv and Re_s_ults of tie,five t Phase 2 Tests The basic test data are sunnarized in Table 14. Weather conditions, summarized in Table 1, show that there was an inversion and low wind speed during all tests in Phase 2. To assure detectable concentrations of SF6 in the sanpling tanks with the additional dilution due to the wake and increased distance, the tracer gas release rate was increased above the Phase 1 rate. For Tests 4 through 6 the source strength 0 was set a t 1. 59 x 10* n3 s e c'l . Although this rate was doubled for the first Phase 2 test (Test 7), the highest tank sanple concentration was only 63 ppb. The release rate was again doubled in Test 8 to a value of
- 6. 34 x 10' n 3sec-1 The renainder of the tests used approximately this release rate.
Tables 15 through 19 contain the calculations which conpare the peak concentrations predicted by the wake nodels (Models lW through 5W) to those actually observed during Phase 2 testing. Tables 15 through 19 are based on north 100-foot tower data. Table 20 compares the results of each of the nodels using north 100-foot tower data. Table 21 is a sun-mary comparison of models using the south 100-foot tower data. A de-tailed tabulation of south tower data appears in Anpendix B. _ ,1_ 1407 042
8.3 Individual Tests 8.3.1 Test 7 The tracer eas release location for both Tests 7 and 8 was on the ground at the eastern wall of the turbine building, as indice:ed in Figures 13 and 15 for the respective tests. Test 7 was characterized by a 4.4 C/100 m increase in temperature with height, and had the snallest value of Phase 2 direction ranne (31*). This was atypical of the meander condition which usually prevailed. The polar display of sample concentrations and wind direction durations (Figure 13) shows that sinaificant concentrations were confined to a narrow 60* sector which includes positions 2, 3A, 3, 4, and 5 in con-sonance with the wind direction. The peak-to-nean ratio, i.e., the ratio of the highest neasured value to the ncan value is 2.3. For Test 7, the lonP.est duration of wind direction bearing 210* is displaced downwind by 30* f ron the location of the naximun concentration. However, because of the aerodynanic effects of the structures,it is difficult to extrapolate the surf ace wind patterns f rom those on the tower. Perhaps nore relevant is the following transcription of visually observed smoke candle plunes released at the tracer gas release site. Also included are connents by a second observer on the winds at the Weather 'feasure tower, at the south end of the site. Tine (EDT) Observed Snoke Plune Behavior First Observer Location: SF6 Release Point. 0155 Smoke start. Smoke path erratic; start in westerly direction noving N & S finally bouncing off building and toward E. 1407 043
Tine (LDT) n _bserved snoke Pluna Behavior First Observer - Location: SF6 Telease Point. 0200 Most of snoke in area between building and transforner. 0210 Sneke stayed in vicinity of buildinz. 0219 McVed slowly in N or NE direction. sone snoke odor now noticeable. 0220 Snoke noving in a southerly direction; reached height of reactor building then to E and then to N. What air novenent there is feels like fron the S. 0230 Snoke release is in northerly direc-tion initially for maybe one ninute, then straight up to top of reactor building. Looping back to ground, spreading around buildine in N and S directions. Wind now noticeable from the E. 0240 Snoke release directly to reactor building up the side and lcoping back to Rround; localized. 0250 Smoke release clinbed straicht up si/2 of reactor building; looping back to ground and noved in north-erly direction. Second Observer Location: Weather 'feasure Tower 0155 Snake drif ting to NNE. Finds on (l'eather "easure) tower f ron SSW. 0210 Snoke drif ted to NE, then to N. 0214 Winds on tower fron SW. 0220 Snoke noving to N. Winds on tower from S. 0223 Sneke still drifting to N. 0230 Snoke drifting to NNE, then to NE. 1407 044
Time (EDT) (continued) Observed Snoke Plune Behavior Second Observer Location: weather Measure Tower, at Phase 1 Gas Release Point. 0235 Sncke drifting to NE. 0240 Drifting to N. 0248 Snoke drifting to NNE. With the exception of ?bdel 4W (AEC/DRL LT Uake Model") the Phase 2 models described in Section 4.3 (as shown in Tables 15 t".. rough 20) predicted in the range 0.80 to .92. Using the south tower values of x ax/X nodel data, Table 21 shows that with the exception of Model 4W, the models'under-predicted. This is attributed to the higher windspeed than was observed at the north tower. Fbdel 49 overpredicted (was conservative) by a factor of about 40. 8.3.2 Test 8 The lapse rate for this test was 0.78 C/100 m, representing only a slight inversion. This ias acconpanied by a direction range of 56' and a wind speed of 1.79 n. sec'I the highest for all tests.
, The pattern of ob-served concentrations is shown in Figure 15, here it is seen that tho high-est concentrations were grouped in a fairly large sector, with values of over 10 ppb subtending 120* of are between positions 13 and 1 respectively.
The peak-to-nean ratio within this sector was 2.1. The pattern of wind directions fron the north tower agrees reasonably well with the location of the high concentration sector, but the samples at positions 2 through 14 inclusive, which include the peak value, show no associated winds fron the north tower. However, as shown on the south tower wind rose (Figure 16), vinds were considerably more easterly which would explain the peak values in the tanks on the west side of the grid.
)k0 04
Visual smoke plune observations appear below for the period including the test (2305 to 2350 EDT). Time (EDT) Observed Snoke Plume Behavior First Observer Location: SF6 Release Point. 2250 Snoke initially noved down road in northerly direction and then swung around building out of sight. 2305 Snoke started to head S, then shifted to N noving around build-inn as before. 2320 Snoke release in northerly dir-ection and around building. 2335 Release to north and around building. 2345 Snoke toward building and. up to top; general swirling and then went to the N. Second Observer Location: Weather Measure tower. 2305 Snoke rose wtraight up (Weather Measure) touer and drif ted slowly W. 2308 Smoke still rising up tower and slowly drifting to SE. 2320 Snoke drifting to NW. 2324 Drifting N. 2335 Drifting NNE. 2339 Snoke still e rif ting N. 2345 Snoke drifted N. in the nax /X nodel With the exception of Model 4U, all nodels predicted X range of 0.84 to 1.57 with Model SW, which accounts for the observed meander, yielding the higher ratio. Model 4W ("AEC/DRL iT Uake 'todel"), although it overestinated by a f actor o ? 9, perforned nuch better in Test 8 than in any 1407 046
of the other tests in this series. In general, the above observations hold true using the south tower data in the nodels. 8.3.3 Test 9 The tracer gas release point was relocated for Test 9 to a point inside of the incomplete dicsel generator building about 40 f t from the face of the reactor building at a bearing of 340* (see Figure 17). An inversion of 5.2 C/100 m existed for Test 9, and the direction range was 165*, which was the largest in Phase 2. The average wind speed was 0.9 m/sec. The effect of wind neander is notable in Figure 17 which shows significant sample concentrations at positions 2, 3A, 5, 8, and 13-17. The pattern of prevailing wind directions .is in agreenent with this. The longest duration wind, at 20*, is only counterclockwise 20' out of phase with the location of the peak concentration at position 11. The overall peak-to-mean ratio is 3.6. For the 140' high concentration sector bounded by positions 10 and 17 the ratio is 2.2. Supplemental visual observations on smoke plume behavior follows . Time (EDT) Observed Snoke Plume Behavior First Observer Location: SF6 Release Point. 0315 Initially snoke drif ting from the SF, then drif ting around the west side of the reactor building, then rising to the top of the reactor building, turning to the S, and drifting downwind of the reactor building. 0330 The snoke behaved the sane as at 0315, except that the plume did not rise as high. Some drift to the W.
- 3 6- )h0
Time (EDT) Observed Snoke Plume Behavior First Observer Location: SF6 Release Point. 0345 Good vertical rise with little horizontal movement at first. The plune rose the height of the reactor building and then slowly drifted to the S. 0400 Smoke dispersion pattern similar to that at 0345, slow rise followed by drift to tha S. 0410 The snoke rose and drifted to the SP over and around the west face of the reacter building. Second Observer Location: Feather Measure tower. 0330 Snoke drifted to SW. . 0345 Smoke rose to a height of approxi-mately 40 ft and drifted NW. 0348 Smoke drifted to the N for about 100 ft, then rose and drifted W. 0350 Smoke was stationary about 125 f t N of the (Weather Measure) tower. The smoke still persisted at about 250 ft NNW of the tower at 0352 at a height of approximately 50 ft and was drifting slowly to the NNW. The diffusion ecuation which most closely predicted the maximum measured concentration for Test 9 was Model SW (" Sector Average Wake 'bdel") with a X nax/X nodel f 0.679. Model SW was also best with south tower data as shown in Table 21. kk
8.3.4 Test 10 The release point was the sane for both Tests 10 and 11, i.e., 26 feet above grade on the roof of the auxiliary buildine. This point was 10 feet from the edge of the reactor buildine at a bearing of 240*. For this test tenperature increased with height at a rate in excess of 11.6 C/100 m, which represents full scale on the recorder, indicatinn extren-ely high stability. The wind was observed to have a direction ranee of
~
35' and the wind speed was 0.6 n sec , lowest of all phase 2 tests. The sanpling grid here was slightly nodified in an attempt to obtain a vertical wind profile. Positions 9 and 10 vare not used; instead, sanplers were located on top of the 7-foot high instrument shed for the north weather tewer, and at the 100 and 150 foot levels of the north tower. How-ever, no significant concentrations were neasured at this location due to a wind shift during the test. Significant sanole concentrations were noted at positions 1 through 8 with the maxinun of 3.4 ppb at positions 3 (Figure 19) . This 140' sector had a peak-to-nean ratio oY 2.4. The wind pattern was predoninatly f rom the west, in agreement with the concentrations. All of the nodels substantially overpredicted the neasured maxinun downwind concentration. Models IP, 2U, 3U, and SW all performed about the same with a X' /X of about 0.013. Use of south tower data in the nax nodel nodels showed even poorer nodel correlation with test results. This first case of extrene over-prediction of the nodels is believed due to the plune renaining aloft at the sannle tank locations as discussed later. 1407 049
8.3.5 Test 11 Tenperature diff erences were again full scale at 11.6 C/100 m and the wind meandered over a 60* are with a wind speed of 0.87 n see- . It should be noted that winds at the south tower neandered over a 175* arc, as shown on Figure 22 The tracer nas release point was also unchanged from the previous test. The sanpling grid was altered such that positions 3 and 10 were not used; instead, sannlers were located at the 100 ft levels of both the north and south towers. As shown on the polar diaeran (Figure 21) very low concentrations were neasured. Values above background were neasured at locations 11 (0.26 ppb) and 14 (0.12 pph) and values of between 1-2 ppb were measured at positions 12, 13 and the north tower. nbserved concentrations were in agreenent with the winds which were essentially from the east. As in Test 10, the models overpredicted the results. The "best" nodel again was SW uith a X nax/Xnodel f nly 0.0167. Model 4W ("AEC/DRL AT Wake 'fodel") predicted concentrations 3000 times higher than measured for both tests 10 and 11. South tower data produced sinilar results, as seen in Table 21. 8.4 Conclusions This phase denonstrated the powerfull effect of the building wake in reducing concentrations. Again, as in Phase 1, the effect of meander was observed to disperse the tracer over a wide arc in nost cases. Model SW (" Sector Average Fake Model") was the "best" predictor for nost of the tests. It was also "best" on the averane when conpared with the averane perfornance of the other nodels. This was found to be true for both north and south tower data (see Tables 20 and 21). For the north tower, the Xnax/Xnodel was 0.62 and for the south rover it e,3 0.90. 1407 050
The reason for poor model correlation in tests 10 and 11 nay be due to the failure of the nas to diffuse downward (after its initial rise due to the buildinn wake effect) due to the extrene inversion conditions exist-ing during the test. On nany occasions the snoke was seen to rise up the sides of the reactor building and over the top, a height of 165 ft above the ground. If, as expected, the SF 6f 11 wed the smoke trajectory con-centrations less than the naximum would be detected by the surface sanpling grid. Measurenents of concentrations aloft which confirn this phenomenon were made, and are discussed in Sections 9 and 10. 1407 051
SECTION 4.0 VERTICAL CONCENTRATION PROFILES 9.1 General A major objective of these diffusion experinents was to obtain sone understanding of the vertical distribution of plune concentrations during stable weather conditions. The data were used to explain observed results obtained f ron Rcound level sanplers, and served as a basis for validating the diffusion nodels at distances beyond the fixed sanple locations. 9.2 Phase 1 Goen Field Vertical Meas _urements A Kytoon balloon was introduced in Tests 3 and 4 to obtain concentra-tion data aloft. This was accomplished by drawing air samples through tubing attached to the winch line for real-tine sanpling by a leak detector at the ground. For both of these tests, the balloon was located near position 6. The concentrations neasured at the ground and at the 30 ft level are given in Table 22 for Test 3. As expected, concentrations alof t were, with one exception, lower than at the ground. In Test 4, lighter tubing was used such that the ballocn rose to 100 ft. However, the wind did not blow in the direction of thic balloon and no concentrations were measured. Because of the linited lif t capacity of the Kytoon, it was replaced fron Test 5 on by a Kaysan balloon with a useful lif t of the order of 15 lbs. For Test 5 and 6, two lines of tubing were attached to the winch cable; one at the balloon tether, and the other 100 f t. below. The balloon was raised to an altitude of 200 ft which pernitted sanpling at the 100 and 200 f t levels. By lowering the balloon, additional data were obtained at the 50
*~
1407 052
and 100 ft levels. 3y continuously lowering and raising the balloon, con-centrations could be monitored at 50 ft increr.ents up to 200 ft. Although the balloon location was such that there was no SF6 detected on Test 5, vertical profiles were obtained in Test 6 with the balloon located near position 16. The results shown for Test 6 in Table 23, as well as thv. , which follow, were obtained by plotting concentrations as functions of time for each available level. These points were then joined to form a set of curves. Finally, vertical profiles were obtained by nicking times which-included as many sample levels as possible. Concentrations for remaining levels were internolated whenever possible. It is clear that concentrations are smaller aloft than at the ground, although the variation with height is irregular. The value at the ground (tank 16 below the balloon) was 111 ppb, '.hereas, the peak short-time values aloft were in the range of 8.5 to 27. 9.3 Phase 2 Vertical 'feasurements in the Buildine Wake Attempts to obtain vertical concentration profiles were of limited success in Phase 1 primarily because of the inability to predict wind direction from one 10-minute period to the next. However, in the wakes of buildings, wind tunnel experiments have shown that near the buildings there was a good chance of being in some part of the plume, even for wind directions almost oblique to the sample location. Therefore, in Phase 2 (with the exception of Tests 7 and 8), the balloons were kept relatively close to the building comolex. During several Phase 2 tests, vertical measurements were taken close to the building to ascertain if the gas did indeed rise and distribute itself
- 4 2-1407 053
over the buildine cavity, durine low wind sneed inversion conditions. If this vere found to he the case, the use of convercional buildine wake con-pensated ecuations for such atnoseheric conditions could be reasonably justified. nbservations of snoke and neasurenents of concentrations alof t showed that for all six tests in Phases 2 and 3, the plune was radically affected by the buildinP wake. A discussion of these tests follows. Finure 23 shows a vertical cross section of the plant structures as viewed fron the north and from the west. Location of the 9F6 release points are indicated for each test. Q.3.1 Tests 7 and R For Tests 7 and R, it was decide? that sannline time could he saved by attaching four lenRths of tubine to the winch cable at 50 ft intervals: this allowed the balloon to renain fixed at the 200 f t le. vel. However, Test 7 yielded no data because the wind direction was away f ron the balloon loc-ation. The vertical profiles obtained in Test 8 are documented in Table 24. In this series, surface data at the base of the balloon were also taken to provide five levels in all. The balloon location in Test 8, indicated in FiRure 15 by the solid det marked B, was at about the sane bearing as erid position 17, but was 60 neters farther fron the reactor center. However, despite this 25% areat-er distance fron the release point, concentrations aloft were occasionally a substantial fraction of the 48 ppb neasured at the ground at position 17. For exanple, at 2339 EDT the 100-foot level showed a concentration of 25.5 ppb, almost the sane as that neasured on the ground. At 2349, values changed only slowly with height, while at 2350 the 100 -foot value was almost- twice
) - 4 3-
that at the surface. At 2353, the lowest value was at the surface; this tendency persisted through 2355 when the 100-foot value was larger than at the ground by a facter of 3. Th2 naxinun concentrations were 7.8 ppb at 50 ft; 25.5 ppb at 100 ft; 13.0 ppb at 150 ft; und 6.5 ppb at 200 ft. Thus xground x,yoft ratios were in the rance of 1.9 to 7.4, senewhat lower than for Test 6 where they ranged fron 4.1 to 13. Thus, there seems to be some tentative evidence to indicate greater transport alof t of the tracer gas in Test 8 as a result of aerodynanic ef fects of the structures than in Test 6 which was carried out in the open field. 9.3.2 gs t 9 For Test 9, a new technique was introduced which was intended to substantially reduce the need for interpolating profile data. Plenun chanbers were fitted such that four sanples could be collected sieultaneously using vacuun pumps, and retained for analysis with the leak-detector in the following few ninutes. This necessitated a decrease in the number of sampled levels to four: surface, 50, 125 and 200 ft. As shown in Figure 17, the balloon in Test 9 was located about 57 meters fron the center of the reactor building at a bearing of 300*. This was close to the release point, in contrast to Test 8 where the balloon was uell beyond the grid perineter. The profiles obtained are niven in Table 25. Here we find evidence of high concentrations aloft. The nost striking are: (1) the values of 89.1 ppb at 50 f t (0340 EDT); (2) the cradual increases with height to naxina of 17.3 ppb and 7.8 ppb at 125 ft (0400 EDT and 0405 EDT respectively), and (3) the value of 30.0 ppb at 50 f t (0415 EDT). These high values aloft support th: suRgestion nade earlier that the buildine wake did
)kQ}
induce rapid vertical transport of the naterial. This is in nood agree-nent with the snoke plune observations in which the snoke was noted to rise up and over the roof cf the reactor building. 9.3.3 Test 10 The balloon was nuved again and placed at a bearing of about 25' halfway between the release peint and grid location 1. Its position is indicated in Figure 19 and the results are given in Table 26. It is seen that concentrations of about the sane nagnitude as at the ground were ob-served at elevations up rn 125 ft. 9.3.4 Test 11 This last test in Phase 2 was unique f a that an attenpt was made to nove the balloon location while the test was in proeress to follow changes in the prevailinn wind direction. Three locations were used as indicated in Figure 21 by the solid dots narked B , B2 and 3B . Position 1 yielded data 3 inadequate for profiles, and one profile each resulted from positions 2 and 3. These data are sunnarized in Table 27. The profile for 0444 at position 2 shows a large increase f ron 0.057 ppb at the ground to 123 ppb at 200 ft. The profile for 0514 also shows a steady increase of concentration with height. 9.4 Phase 3 Time Averaged Vertical 'teasurenents 9.4.1 General Because of the frustrations involved in trying to place the balloon a d vertical sanple apparatus within the plune boundary, an "all out" attenpt was nade to sanple the plune vertically. Four balloons were positioned in a 180* arc, each one having sanole tubes which terninated at the followinn
- 4 5-1407 056
i . heights: surface, 75, 150 and 250 ft. An improvenent in the vertical san-ple techniaue was nade to enable collection of integrated samples alof t. This was done by fitting each balloon apparatus with 16 liter evacuated sam-pie bottles connected to each of the four tubes. The tubes were prepurged and were all the same length such that during the 45 minute period each would draw simultaneous sanples fron each level. 9.4.2 Test 12 Only one test (Test 12) was conducted using the configuration des-cribe.1 a'>ove. The weather conditions were characterized by an inversion of
-1 3.14 C/100 n with wind neander of 75* and wind speed of about 0.91 m sec .
The sample locations are shown on the polar diagran of Figure 24. A south tower wind rose is given in Figure 25. Results for each position are given in Table 28 and shown graphically in Figure 26. The balloon position 3 profile indicated a naximun of 2.26 ppb at the 150 foot level. At position 4, the concentration steadily increases with height to a maximum of 32.4 ppb at 250 ft. Positions 1 and 2 were outside of the plume and yielded no in-formation. 9.4.3 Concarison of ?taximum Concentration With *todels The naxinun value of 32.4 ppb at position 4 was compared with the five wake models described in Section 4.3. A sunnary of this conparison using both north and south weather tower data is given in Table 29. The range of X m h/ del f r m dels lW, 2W, 3" and SW with north tower wind data was 0.19 to 0.67. ?todel 5" (" Sector Average Wake 'todsl") was "best" based on the north tower data, with a ratio of 0.67. Based on south tower data, the models also overpredicted, with !todels 1W, 2F and 3W being "best". Model SW over-907 057
predicted considerably due to the direction range of only 11' neasured on the south tower. The X 3x/Xnodel ratios for 'bdel 44 (AEC/DRL AT Wake Model") were 0.02 and 0.04 for the north and south towers, respectively. 9.5 Conclusions The vertical concentration profiles lend support to the conclusion that, in the presence of low wind speed inversion conditions, the aerodynanic ef-feet of structures is to induce inital vertical transport of the diffusing naterial to altitudes conparable to the height of the nearest building, with subsequent advection downwind with the greatest concentration at some higher elevation. For Test 12, where average tank sanples were collected over a 45 nin-nax/Xnodel ute sanple period, the X was less than 0.67 for all nodels even though the naxinun was at an elevated positirn. Thus, the nodels overpre-dicted concentrations at the 800 f t distance with Siodel SW (" Sector Average Wake 7bdel") being the "best". The concentration predicted by the Safety Guide 4 nodel (identical to ? odel 4W) using Pasquill F and 1.0 m/see yields a X ,g/Xmodel f 0.046 when compared to the maxinun Test 12 concentration of 32.4 ppb. Therefore, it is concluded that the Safety Guide 4 model is very conservative.
- 4 7-1407 058
SECTION 10.0 MOBILE OFF-SITE TRAVERSES 10.1 General One of the major objectives of the experiment was to validate a diffu-sion model which could be used to predict site boundary concentrations. Since models were validated at a distance of 800 ft, and the site boundary is at 10GJ ft, a series of mobile traverses were conducted to test the behavior of the models at distances beyond 800 f t. During two tests, a road vehicle was equipped with a leak detector and recorder; and in one of these tests, a boat was also used. For each traverse the vehicle was driven along local highways downwind of the release point taking instantaneous readings of concentration at various locations. 10.2 Test 10 Road Traverse . Following the tank sampling portion of Test 10, the achile SF6 analyzer was mounted in a truck and samples were taken for almost tv.a hours off the site. During this period, release of gas was continued at 1/2 the initial flow rate, or 3.17 x 10-4 m 3 sec -l. The locations where readings were taken and the corresponding time and concentrations are plotted in Figure 27. The prevailing wind speeds and directions on the north and south towers during the traverse period are shown in Figure 28. As shown in Figure 27, concentrations wers observed to be relatively uniform along Route 441 over about a two mile stretch. The wind direction traces are consistent with those observations. Thus, the SF6 was spread over a wide area at a distance of about 2000 ft. _o_ 1407 059
10.3 Test 12 Road and River Traverses For Test 12, the wind was blowing generally from the south. This required the use of a boat for close-in measurement and a car for measure-ments at greater distances downwind. The concentrations and locations for the road traverse are shown in Figure 29 and in Figure 30 for the river traverse. North and south tower winds are shown in Figure 31. The SF6 release rate was 3.17 x 10-4 m3 sec-1 10.4 Correlation of Results with Models 10.4.1 Test 10 Traverse From Figure 27, the maximum concentration measured was 1.87 ppb at 0631 EDT. This value is compared (in Table 30) with predicted values using the "W" models in Table 2 and the meteorological conditions measured at the north and south towers during the traverses. As shown in Table 30, Model SW (" Sector Average Wake Model") at 2000 f t resulted in an g,x/Xmodel f 0.26 using the north tower data. The use of the same model at the 800 ft distances showed generally poor, correspondence of observed versus predicted tank concen-trations at ground level. This is shown in Table 20 for Test 10 where X m
/
X model was 0.013. However, it has been demonstrated that the maximum concentration in the plume is generally aloft at the 800 ft distance. This readily explains the poor correspondence. At greater distances, general streamline descent in the far wake is to be expected, and the concentrations rear the ground should more closely approximate those predicted by the models. This was evident for Test 10 at 2000 ft where the g /Xmodel was 0.26. However, insufficient data were available to determine if the maximum concentrations were really at ground level. If the plume was still somewhat aloft at a 2000 ft distance, 1407 060
the pocrer correspondence could be attributed to this condition. T1.e agreement for Test 10 at 2,000 ft is not as good (with the same model) as was observed at 800 ft for the 250 ft level in Test 12 (0.67) but is still considered reasonable for a test of this nature and it is still on the conservative side. 10.4. 2 Test 12 Traverses P001 CdkIW$ vk, Figure 32 shows a calculated curve of concentration versus distance, using Model SW (" Sector Average Wake Model") with the approximate wind con-ditions which existed during Test l2 (see Table 14) . The experimental obser-vations made during the Test 12 land and river traverses are taken from Figures 29 and 30, and are plotted in Figure 32. The maximum value for Test 10 at a distance of 2,000 f t and the tank samples taken at the 250 f t level during Test 12 are included in Figure 32. It is seen that the Model SW predicted curve is in excellent agreement with the measured concentrations. Some scatter is noted but it must be remembered that the calculated curve is intended to represent an average concentration over a time period of approximately 45 minutes while mobile observations represent approximate-ly 5 second samples. According to Turner7 an inverse one-fifth pouer law describes variation of concentration with sampling time. The ratio of con-centrations corresponding to a decrease of sampling time from 45 minutes to 5 seconds results in an increase in concentration by a factor of about 3.5. It is seen that all the individual observations at distances greater than 2,000 meters fall within a factor of 3.E of the Model SW curve in Figure 32. The group of river observations rhow concentrations markedly less than predictions made by the model. Those observations were taken at about the same distance from the plant as the road traverse observations in Test
- 10. It was concluded in that discussion that the plume may still have been aloft at that distance and this explanati;n nay serve here with the levels 1407 061
of concentrations observed. Note that all of the observed data points fall below the computed curve for the Safety Guide 4 meteorology (Pasquill F stability and 1.0 m sec-1) shown in Figure 32. 10.5 Conclusions It is concluded that Model SW is an accurate and conservative method of predicting the maximum concentration at a specified distance downwind of a source near a reactor building surface without specifying the height at which it occurs. It is conservative, of course, to assume that the maximum occurs at the ground even though it probably remained aloft for distances that ranged to at least 1000 meters during the tests. 1407 062
Table 1 Weather Condition Summary Table Time of Wind
- Wind Day Speed AT Direction Phase Test Date (EDT) (m/sec) (OC/100 m) Range 1 2 8/25/71 0500 0.62 4.26 150 3 9/08/71 0445 0.20 2.97 168 4 9/09/71 0400 0.19 2.83 175 5 9/23/71 0415 0.15 0.70 167 6 9/24/71 2125 0.37 2.05 162 11** 7 10/06/71 0205 1.12 4.40 31 8 10/08/71 2305 1.79 0.78 56 9 10/13/71 0330 0.90 5.20 165 10 10/15/71 0420 0.60 11.60+ 35 11 10/16/71 0400 0.87 11.60+ 60 III** 12 11/12/71 0035 0.91 3.14 75
- All AT data from north tower.
** Wind data from north tower.
1407 063
Table 2 Sununary of Models PHASE 1 MODELS Model 1P ("AEC/DRL AT Model") : Uses equation (1) with the AT groups given in Table 4 to de-fine both a and a . y z Model 2P ("Slade ga Model"): Uses equation (1) with the groups given in Table 3 to define ay and a . Model 3P (" Split a Model") : Uses equation (1) with the AT g.roups of Table 4 and the ag groups of Table 3 to define a, and a respectively. Model 4P (" Sector Average Model"): Uses equation (2) with 6 equal to the maximum wind direction meander (range) during the sample period, and a zbased on the AT groups of Table 4. Model SP (" Directional Frequency Model"): Uses equation (3) anda and a, based on AT in Table 4? PHASE 2 MODELS Model 1W ("AT Wake Model") : Uses equation (4) with cA based on area of containment with c=2, and AT groups in Table 4. Model 2W ("Slade ag Model With Wake Uses equation (4) with cA based Correc tion") : on area of containment with c=2, and a group in Table 3. g Model 3W (" Split a Wake Model") : Uses equation (4) and both Tablec 3 and 4 as for Model 3P, cA based on area of containment and c=2. Model 4W ("AEC/DRL AT Wake Model"): Uses equation (4) with cA based on containment area, c=h and AT from Table 4. Maximum wake correction is 3. Model SW (" Sector Average Wake Model"): Uses equation (7) with c=2, A=2000 anda g from Table 4. 1407 064
Table 3 Pascuill Stability Classes Based on Wind Data
~
Slade Pasquill Standard Deviation of the Stability Class Horizontal Wind Direction, A og > 22.50 B 22.50 > 6 L 17.5 C 17.5 >0 0 L 12.50 D 12.5 > 00 1 7.5 E 7.50 > 0 L 3.80 r 3.80 > on Table 4 Pasquill Stability Classes Based on Temperature Data AEC/DRL Pasquill Stability Vertical Temperature Class Gradient T (C0/100 m) A -1.9 > AT B -1.7 > AT > -1.9 C -1.5 > AT > -1.7 D -0.5 > AT > -1.5 E +1.5 > AT > -0.5 F +4.0 > AT > +1.5 C AT > +4.0 f407 065
Tcble 5 Positions of Sampling Tanks: Phase 1 Distance From Center Sampling Tank Position of Grid (Meters)* Number Bearing (Degrees) Test 2 Tests 3-6 1 20 183 (600) 101 (330) 2 40 190 (625 101 (330) 3 60 158 (518) 101 (330) 4 80 116 (382) 101 (330) (East) 5 100 96 (314) 101 (330) 6 120 94 (310) 98 (322) 7 140 119 (390) 101 (330) 8 160 134 (440) 101 (330) 9 180(South) 150 (492) 94 (310) 10 200 171 (560) 98 (320) 11 220 165 (540) 88 (290) 12 240 120 (393) 101 (330) 13 260 94 (310) 98 (322) (Wes t) 14 280 94 (310) 101 (330) 15 300 109 (357) 107 (350) 16 320 143 (470) 98 (320) 17 340 146 (478) 101 (330) 18 360(North) 174 (570) 101 (330)
- Values in parentheses are equivalent feet.
1407 066
Test 6 Summary of Basic Test Data: Phase 1 Pasquill SF6 30 ft. 30 ft. Stability Pasquill Start Release Wind _ Lapse Direction Group Stability Test Test Duration Rate, Q Speed, u Rate, AT Range Based On B sed On Number Date EDT (min) (m3 sec -l) (m sec-l) ( C/100 m) 0 Range /6 00* AEC/DRL AT _ Range /6 2 8/25/71 0500 50 2.38(-4) 0.62 4.26 150 25.0 37.30 C A 3 9/08/71 0445 50 2.38(-4) 0.2 2.97 168 28.0 49.4 F A 4 9/09/71 0400 45 1.59(-4) 0.19 2.83 175 29.2 46.70 F A 5 9/23/71 0415 45 1.59(-4) 0.15 0.7 167 27.8 40.10 E A 6 9/24/71 2125 45 1.59(-4) 0.37 2.05 1620 27.0 55.40 F A
- Computed (See Appendix B)
- TM CD
-J ~
C1) CB
-J
Table 7 Concentration Calculations Based on Model IP ("AEC/DRL AT Model")r Phase 1 Observed AEC/DRL Somple Tank Distance From Pasquill _ Xmax Sample SF6 Release Stability Q u o cz X X X Test (ppb) Number Point (m) Category (m3 sec -1) (m sec-1) (m) (m) 7odel ppb) max / model 2 2610 13 94 G 2.38(-4) 0.62 3.0 1.5 27,200 0.096 3 1788 17 101 F 2.38(-4) 0.20 4.7 2.3 35,000 0.051 4 567 8 101 F 1.59(-4) 0.19 4.7 2.3 24,600 0.023 5 390 11 88 E 1.59(-4) 0.15 5.9 2.8 20,400 0.019 6 534 l 18 101 F 1.59(-4) 0.37 4.7 2.3 12,700 0.042 di Average 0.046 7' Note: ' Model IP: Q X " model _ nuo o y z oy, o, based on AT and Table 4. N CD
- -I O
CD
Table 8 Concentration Calculations Based on Model 2P ("Slade 00 Model"): Phase . Observed Distance Slade Sample Tank From SF6 Pasquill _ Sample Release Stability u Cy z Xmodel Xmax Q Test (ppb) Number Point (m) Category (m3 sec -1) (m sec-1) (m) (m) (ppb) Xmax/Xmodel 2 2610 13 94 A 2.38(-4) 0.62 22.0 12.0 463 5.64 3 1788 17 101 A 2.38(-4) 0.20 23.0 13.0 1270 1.41 4 567 8 101 A 1.59(-4) 0.19 23.0 13.0 891 0.64 5 390 11 88 A 1.59(-4) 0.15 21.9 11.0 1400 0.28 6 534 18 101 A 1.59(-4) 0.37 23.0 13.0 457 1.17 dn Average 1.83 i Note: Model 2P:
=
x Q ' model _ nu o a y z 4:= a ,o both based on R/6 and Table 3. CD y z N CB
Table 9 1 Concentration Calculations Based on Model 3P (" Split o 'iodel"): r Phase 1 Slade AEC/DRL Observed Distance Pasquill Pasquill Sample Tank From SF6 Stability Stability _ G Xmax Sample Release Category Category Q u y Oz Xmodel Test (ppb) Number Point (m) For oy For Oz (m3 see-1) (m sec-1) (m) (m) (ppb) Xmax/Xmodel 2 2610 13 94 A G 2.38(-4) 0.62 22.0 1.5 3700 0.70 3 1788 17 101 A F 2.38(-4) 0.20 23.0 2.3 7160 0.25 l q; 4 567 8 101 A F 1.59(-4) 0.19 23.0 2.3 5040 0.11 5 390 11 88 A E 1.59(-4) 0.15 21.9 2.8 5500 0.07 6 534 18 101 A F 1.5-(-4) 0.37 23.0 2.3 2590 0.21 Average 0.27 Note: Model 3P: Q X _. model nu o og , a a== a based on R/6 and Table 3 O
-J Oz based on AT and Table 4.
CJ
- N O
Table 10 Concentration Calculations Based on Model 4P (" Sector Average Model"): Phase 1 Observed AEC/DRL Sample Pasquill _ Stability u az Xmax Test (ppb) Xavg) (ppb Category (m sec-1) (m) 0 Xmodel Xmax/Xmodel Xavg/Xmodel 2 2610 710 G 0.62 1.5 150 832 3.14 0.85 3 1788 478 F 0.20 2.3 1680 1400 1.28 0.34 4 567 290 F 0.19 2.3 1750 944 0.6 0.31 5 390 113 E 0.15 2.8 1670 1180 0.33 0.10
, 6 534 209 F 0.37 2.3 1620 524 1.02 0.40 . , _ _
8 Averages: 1.27 _ 0.40 Note: Mode] 4P: X d2/n model = uxo 0 z a based on Table 4 (AT). z A N CD N O M
Table 11 Concentration Calculations Based On Model SP (" Directional Frecuency Model") : Phase 1 Observed Distance f S ample From SF Pasquill Fraction 6 _ Xmax Release Stability Q u y O z Of Total Xmodel Test (ppb) Point (m) Category (m3 sec-l) (m sec-1) (m) (m) Interval (ppb) X max /Xmodel 2 2610 94 G 2.38(-4) 0.62 3.0 1.5 0.17 4510 0.58 3 1788 101 F 2.38(-4) 0.20 4.7 2.3 0.20 7010 0.25 4 567 101 F 1.59(-4) 0.19 4.7 2.3 0.13 3200 0.18 5 390 88 E 1.59(-4) 0.15 5.9 2.8 0.19 3880 0.10
; 6 534 101 F 1.59(-4) 0.37 4.7 2.3 0.066 835 0.64 ___
Average 0.35 Note: x = X X f model model IP s O N O N ' rs)
Table 12 Summary of Phase 1 Results max / model Observed Model Model Sample IP 2P Model 4P Model SP Xmax "AEC/DRL "Slade c M del 3P " Sector " Directional Model IP With Test (ppb) AT Model" Model" 0 "Splito !!odel" Average Model" Frequency Model" Type F & 1 m sec~I 2 2610 0.096 5.64 0.7 3.14 0.58 0.312 r 3 1788 0.051 1.41 0.25 1.28 0.25 0.255 4 567 0.023 0.64 0.11 0.60 0.17 0.121 5 390 0.019 0.28 0.071 0.33 0.10 0.061 6 534 0.042 1.17 0.21 1.02 0.64 0.114 Averages: 0.046 1.83 0.27 1.27 0.35 0.173 Model IP: Table 7 Model 2P: Table 8 Model 3P: Table 9 Model 4P: Table 10 Model SP: Table 11 . -.s N CD N O N trJ
Table 13 Pesitions of Sampling Tanks: (Phase 2) Distance from Center Sampling Tank of Grid * (Meters)** Number Bearing (Degrees) Tests 7-11 1 20 244 (800) 2 40 238 (780) 3 60 149 (490) 3A(See Text) 60 317 (1040) 4 80 244 (800) (East) 5 100 244 (800) 6 120 259 (850) 7 140 244 (800) 8 160 244 (800) 9 180(South) 244 (800) 10 200 244 (800) 11 220 244 (800) 12 240 244 (800) 13 260 204 (670) (West) 14 280 177 (580) 15 300 186 (610) 16 320 201 (660) 17 340 244 (800) 18 360(North) 244 (800) For Phase 1, the grid location was the field about 0.5 mi. south of the actual site; there, the center of the grid corresponded to the SF6 release point. For Phase 2, however, the grid was centered on the exact center of the northernmost reactor building; now the SF6 release point was not the center of the grid. See the individual polar projection figure for each test for the release location.
** Values in parentheses are equivalent feet.
1407 Cf74
Table 14 Summary of Basic Test Data: Phases 2 and 3 Pasquill SF 100 ft Pasquill Stability 6 Wind
- Stability Group Start Release Rate Speed Lapse Rate Group Based On Test Test Duration Q u 6T Range 00** Based on Slade Number Date EDT min. (m3 sec-1) (m sec-1) (OC/100 m) {0) Range /6 (0) AEC/DRL AT Range /6 7 10/06/71 0205 45 3.17(-4) 1.12 4.4 31 5.2 7.6 G E 8 10.08.71 2305 45 6.34(-4) 1.79 0.78 56 9.3 14.2 E D 9 10/13/71 0330 45 3.17(-4) 0.90 5.2 165 27.5 67.8 G A
~
10 10/15/71 0420 45 6.34(-4) 0.60 11.6 + 35 5.8 9.6 G E 11 10/16/71 0400 45 7.93(-4) 0.87 11.6 + 60 10.0 18.2 G D 12 (Phase 3) 11/12/71 0035 45 6.34(-4) 0.91 3.14 75 12.5 19.1 F C
- From north tower at 100 ft
** Computed, see Appendix B an, 4,-
C~) N C~.) N LJ1
Table 15 Concentration Calculations Based On Model IW ("AT Wake Model"): Phase 2 Distance Pasqui.ll Observed From Stability Sample Tank Center Category _ Xmax Sample of Grid Based On Q u* Test (ppb) No. (m) AEC/DRL AT (m3 sec-1) (m sec-1) Cy(m) oz(m) Xmodcl (ppb) Xmax/Xmodel 7 63.0 3 149 G 3.17(-4) 1.12 4.6 2.3 70.2 0.898 8 71.0 14 177 E 6.34(-4) 1.79 10.4 5.8 84.5 0.84 9 7.6 11 244 G 3.17(-4) 0.90 7.2 3.4 86.4 0.088 10 3.4 3 149 G 6.34(-4) 0.60 0.6 2.3 262 0.013 11 1.59 13 204 G 7.93(-4) 0.87 0.87 2.9 225 0.007 Model IW: Xmodel Q , u (ncy o g+ cA) a y,0 z based on Table 4 (AT). ,,, cA = 4000 .C> C)
- Based on north 100 ft tower data.
W C_> _a .i CP
Table 16 Concentration Calculations Based On Model 2W ("Slade ce Model with Wake Correction"): Phase 2 Pasquill* Stability Observed Distance Category Sample Tank From Based On Xmax Sample Center of Slade Q _,u* Gy cz Xmodel Test (ppb) No. Grid (m) 00 (m3 sec -I) (m sec-I) (m) (m) (ppb) Xmax/Xmodel 7 63.0 3 149 E 3.17(-4) 1.12 9.0 5.0 68.3 0.922 8 71.0 14 177 D 6.34(-4) 1.79 14.2 8.8 80.6 0.881 9 7.6 11 244 A 3.17(-4) 0.90 55.0 41.0 31.8 0.239 10 3.4 3 149 E 6.34(-4) 0.60 9.0 5.0 255 0.0133 11 1.59 13 204 D 7.93(-4) 0.87 16.5 9.8 202 0.0078 Model 2W: Q Xmodel " 0 EI (uc y 2 + cA) o y, o g both based on Table 3 (Range /6) cA = 4000
- Based on north 100ft tower data, s
5 CD N O N N
Table 17 Concentration Calculations Based on Model 3W ("Splito Wake Model"): Phase 2 Pasquill Pasquill Observed Distance Stability Stability Sample Tank From SF6 Category Category _ Xmax Sample Release Based On Based On Q u Oy az Xmodel Test (ppb) No. Point (m) Slade 00 AEC/DRL AT (in 3 sec-I) (m sec-l) (m) (m) (ppb) Xmax/Xmodel 7 63.0 3 149 E G 3.17(-4) 1.12 9.0 2.3 69.6 0.905 8 71.0 14 177 D E 6.34(-4) 1.79 14.2 5.8 83.2 0.854 9 7.6 11 244 A G 3.17(-4) 0.90 55.0 3.4 76.8 0.099 10 3.4 3 149 E G 6.34(-4) 0.60 9.0 2.3 260 0.0131 11 1.59 13 204 D G 7.93(-4) 0.87 16.5 2.9 220 0.0072 Model 3W: t m Ymodel u (noy c z + CA) oy based on Table 3 og based on Table 4 cA = 4000 ap,
- Based on north 100 ft tower data C'?
--a i C'
c>
~\
CD
Table 18 Concentration Calculations Based on Model 4W ("AEC/DRL AT Wake Model"): Phase 2 Pasquill Observed Distance Stability Sample Tank From Category Cy U Xmax Sample Center Of Based on Q _u* z Amodel Test (ppb) No. Grid (m) AEC/DRL 4T (m3 sec-1) (m sec-l) (m) (m) (ppb) Xm ax/Amodel 7 63.0 3 149 G 3.17(-4) 1.12 4.6 .3 28,400 0.022 8 71.0 14 177 E 6.34(-4) 1.79 10.4 5.8 623 0.114 9 7.6 11 244 G 3.17(-4) 0.90 7.2 3 . :. 1,530 0.005 4 10 3.4 3 149 G 6.34(-4) 0.60 4.6 2.3 10,600 0.0003
?
11 1.59 13 204 G 7.93(-4) 0.87 6.4 2.9 5,210 0.0003 Model 4W: Q Xmodel Li (noy az + cA) c y ,oz based on Table 4 (AT) cA = 4300 .
- Based on north 100 ft tower data.
P= CD N O N W
Table 19 Concentration Calculations Based On Model SW (" Sector Average Wake Modal"): Phase 2 Pasquill Observed Distance Stability Sample From Category Xmax Center Of Based On Q _u* Oy Uz Xmodel Test (ppb) Grid (m) AEC/DRL AT (m 3 sec-1) (m sec-I) (m) (m) 0* (ppb) Xmax/Xmodel 7 63.0 149 G 3.17(-4) 1.12 4.6 2.3 31 78.6 0.802 8 71.0 177 E 6.34(-4) 1.79 10.4 5.8 56 45.3 1.57 9 7.6 244 G 3.17(-4) 0.90 7.2 3.4 165 11.2 0.679 as 10 3.4 149 G 6.34(-4) 0.60 4.6 2.3 35 260 0.013
?
11 1.59 204 G 7.93(-4) 0.87 6.4 2.9 60 95.4 0.016 Model SW: Q,s/2/n Xmodel " n uEz O ' I '=1f az +1 n a based on Table 4 (AT) z .m. b cA = 4000 c;s -
- Based on north 100 ft tower data.
O3 CD
Table 20 Summary of Results Using North Tower Data: Phase 2 max / model Observed Model 2W Model SW Sample Model IW "Slade 0 M del Model 3W Model 4W " Sector Model 4W With Xmax "AT Wake With Wake " Split o "AEC/DRL Average Type F & 1 Test Date (ppb) Model" Correction" Wake Model" AT Wake Model" Wake Model" (m sec-1) 7 10/06/71 63.0 0.898 0.922 0.905 0.022 0.802 0.039 8 10/08/71 71.0 0.84 0.881 0.854 0.114 1.57 0.030 9 10/13/71 7.6 0.088 0.239 0.099 0.005 0.679 0.01 10 10/15/71 3.4 0.013 0.0133 0.013 h 0.0003 0.013 0.001 11 10/16/71 1.59 0.007 0.0078 0.0072 0.0003 0.017 0.0007 Averages: 0.369 0.413 0.376 0.028 0.621 0.016 Model IW: Table 15 Model 2W: Table 16 Model 3W: Table 17 Model 4W: Table 18 Model SW: Table 19 Q N O CO
Table 21 Summary of Results Using South 100 ft Tower Data: Phase 2 Xmax/xmode l OSserved Model 2W Model 5W Sample Model IW "Slade c9 Model Model 3W Model 4W " Sector Xmax "AT Wake With Wake " Split c "AEC/DRL Average Test Date (ppb) Model" Correction" Wake Model" AT Wake Model" Wake Model" 7 10/06/71 63.0 1.23 1.32 1.25 0.03 2.31 1, 8 10/08/71 71.0 0.61 0.715 0.641 0.083 1.73 7 9 10/13/71 7.6 0.063 0.173 0.071 0.0036 0.446 10 10/15/71 3.4 [ 0.0032 0.0037 0.003. 0.00008 0.0076 11 10/16/71 1.59 0.0045 0.005 0.0045 0.00019 0.031 Averages: 0.383 0.443 0.394 0.023 , 0.9 N O M Q CD N
Table 22 Vertical Concentration Profiles: Test 3 SF6 Concentration (ppb)* Time (EDT) Ground 30 ft. 0507 8.5 14.6 0512 13.6 1.9 0519 1.8 0.53 0523 0.93 0.40 0527 0.28 0.11 0533 0.13 0.058
- Underscored values are interpolated from graphs of instantaneous concentration vs.
time. Table 23 Vertical Concentration Profiles: Test 6 SF6 Concentration (ppb)* Time (EDT) 50 ft 100 ft 150 ft 200 ft 2150 6.4 11.2 2.6 12.7 2153 27.3 12.0 2.7 9.5 2156 6.0 12.8 2.2 7.2 2200 0.88 1.1 1.8 1.4 2203 0.21 0.53 3.6 0.36 2215 5.7 6.5 8.5 2.9 2218 4.5 2.4 6.2 3.5
- Underscored values are interpolated from graphs of instantaneous concentration vs. time.
1407 083
Table 24 Vertical Concentration Profiles: Test 8 SF6 Concentration (ppb)* Time (EDT) Sfc. i 50 ft. 100 ft. 150 ft. 200 ft. 2337 33.7 1.0 4.5 5.8 6.5 2339 29.0 1.8 25.5 13.0 3.4 2341 24.0 3.3 6.2 4.2 1.8 2344 15.0 7.8 0.72 0.55 0.48 2347 1.5 1.4 0.20 0.1 0.13 2349 0.35 0.38 0.25 0.1 0.05 2350 0.16 0.22 0.29 0.1 0.06 2353 0.08 0.20 0.41 0.11 0.11 2355 0.18 0.40 0.54 0.12 0.16 2357 2.5 0.75 0.68 0.15 0.18 2358 2.5 1.05 0.76 0.17 0.2
*Unoerscored values are interpolated from graphs of instantaneous concentration vs. time.
Table 25 Vertical Concentratien Profiles: Test 9 SF6 Concentration Profiles: (pph) ! Time (EDT) Sfc. 50 ft. 125 ft. 200 ft. 0335 0.02 0.20 0.16 1.01 0340 2.2 89.1 0.18 0.07 0353 12.1 4.4 0.64 0.46 0400 2.2 2.9 17.3 0.15 0405 0.73 2.2 7.8 0.42 0415 5.0 30.0 1.82 3.6
- Underscored values are interpolated from graphs of instan-taneous concentration in time.
1407 084
Tabic 26 Vertical Concentration Profiles: Test 10 SF Concentration (ppb) 6 Time (EDT) Sfc. 50 ft 125 ft 200 ft 0449 0.25 0.086 0.028 0.027 0455 1.0 1.0 1.62 0.15 0503 6.0 1.0 0.27 0.18 0512 1.65 1.39 1.53 0.11 0521 1.36 1.29 0.61 0.071 0530 1.63 0.84 0.73 1.0 Table 27 Vertical Instantaneous Concentration Profiles: Test 11 Time (EDT) Sfc. 50 ft 125 ft 200 ft Position 2 0449 0.057 0.10 4.9 123.0 Position 3 0514 0.16 0.18 0.40 0.64 Table 28 Vertical Average Concentration Profiles: Phase 3 (Test 12) l SF Concentration (ppb)* 6 Location Sfc. 75 ft i 150 ft 250 ft 1 0.12 M 0.10 M 2 0.10 0.08 0.11 M 3 0.46 1.72 2.26 0.46 4 0.13 4.0 17.8 32.4
- M denotes missing data.
1407 085
Table 29 Concentration Calculations Using Wake Models for Test 12 Xmax/Xmodel Model 2W Model SW Observed _ Model IW "Slade O M del Model 3W Model 4W " Sector Maximum Weather u T "AT Wake With Wake " Split o "AEC/DRL . Average Test (ppb) Tower (m suc-l) R0 ( C/100 m) Model" Correction" Wake Model" AT Wake Model' Wake Model* 12 32.4 North 0.91 75 3.14 0.193 0.219 0.200 0.021 0.668 12 32.4 South 1.80 110 3.14 0.382 0.382 0.382 0.042 0.194 5 O N O CJO CP
Table 30 Summary of Wake Model Performance for T'raverses: Tests 10 Distance Pasquill Pasquill ' of Xmax Stability Stability Model 2W Model SW Model 4W From Category Category Oaserved "Slade ce Model 3W Model 4W " Sector With Release Tower Wind Based On Based On Xmax Model IW Model "Splito "AEC/DRL Average Pasquill Point Data Speed u Slade AEC/DRL Location "AT Wake With Wake Wake AT Wake Wake F& Test (meters) From (m sec-1) R/6 AT (ppb) Model" Correction" Model" Model" Model* 1 m sec 10 620 North 0.5 A G 1.87 0.0129 0.217 0.019 0.0032 0.26 0.0065 10 620 South 0.2 A G 1.87 0.0051 0.086 0.0078 0.0013 0.0E4 0.0065 h CD N O N
- e 10'
---t ttrg-- I -+ ,, -1 --g f !, ,, f- l-
- t 5 . _.-
-)
3f, q . 7 ,7
]-
7 g r ;- "A t- ,7 4
' I B.i 1: ,
F,- e Ot ,a 2 , , DI ' ' T*
. s 10' >
f f :*,~~Ed,N- b;.- m; L; D S d
' f 3 '
W // i/ / 1I I I
! /// '
El/ I I
$ 2 I 7 e 10 ~ -f' e
[ M f_ e
', [ _,s-/-/ ,'
j - N _ i A- EXTREMELY UNSTABLE , . S
/ 3_' ~ --p/ .
i
, ' 8 C- -MODERATELY 1
I SLIGHTLY UNSTABLE UNSTABLE 0- NEUTRAL
' ' I
l l l l l,
~ ~~ -i i ~'I E - SLtGHTLY STABLE ' ' ' ll ' I l
I F - MODERATELY STABLE 2 - ~ - I ii;ll1 I l l i ' il i i
, i .
i'; l l 80, 1----.
~7a !. i ' .I 1. - 9 1 T-e i., .
i 1 1 i i i i
.v* ; . ~ + -
2 5 10 2 5 10' 2 5 10* 2 5 10 OISTANCE FROM SOURCE tml Figure 1 versus Downwind Distance Lateral from Source Diffusion, for Paso%u,ill's Stability Classes f,[ $ $k 1 N o 1407 088
8 3 s10 2 l /' !/ l /
'3 . 1 1
l ,ii>i f
/
i 1, # w,
, ,/
5 - ~
/f * / ' / / .- ~ ~j ) /l 'J 'E i / /
p y* / j r
/ - )/
o . - E s ! / \ 3 Y
.I.___ [ )_ > [
1 , - [,7 .- R # E' /' ' ' g 5 , 7~4 , ,' 7,'0 [,' ,I , j e ;
$ 3 I i l I // , 7 $ l I I'j , j F g 2 - - - ~
f 0 C 6 l1 - h8-10
._T -
2
- B-A - EXTREMELY UNSTABLE MODERATELY UNSTABLE
[f a / Z["I b' If I i,l 1 e '+ 3 C - SLiGMTLY UNSTABLE 0 - NEUTRAL i I lI i1 i 1 I I E - SLIGHTLY ST'.3LE
/~ i[i l, l.
t l l l' [ l F - MODERATELY STABLE I l ! 10 2 ii n! , i .. 10 2 5 10 8 2 5 IO" 2 5 10' OISTANCE FROM SOURCE tm) Figure 2 Vertical Diffusion, z, versus Downwind Distance from Source for Pasquill's Stability Classes
)'
D**}D i o o Ju. o Ju 1..
*}9 (fu a 1407 089
. .....,u.., . . a . u . .. . . 'm[
f I O (a)
'# .+y .......o..e . , . .. , u o ..
i .**
.. .. .s r ,c i. , I
____ g y __ i ,
. C - c --. , , y
- s o
(b) {/ ,
.- ........u... ..o.n....,,, C ,; .o*'h 5
M/ \'\ _9_ . _ . _ . _ ', h [_[h_p\ _._.L_ \
'z >j, O 2'b[\Nj O -
l . o (c) Figure 3 Kc Isopleths for Reactor Complex. (a) Downwind Release (b) Top Release (c) Upwind Release 1407 090
VG STAINLESS STEEL TANK \
+MBI V
e V QOV 16 LIT E RS NIV LEGEND QOVs QUICK OPENING VA' VE CFC CONSTANT FLOW CONTROLLER NIV NEEDLE INLET VALVE VGi VACUUM GAUGE V VALVE Figure 4 Sampling Tank Assembly i407 09I V RELEASE POINT v 7 e ,g
& J L"
- NEEDLE l:
VALVE 9 f ROTAMETER i
- 2 SF
[ , , , , \, , , TABLE e > Figure 5 Tracer Gas Release Apparatus 1407 092
i , i ' 4 i e i O ( Q ' Qd4. =. g y 5 E 7 8 9 10 6 l J i i i WsN/i i i
. . . ._. L A i L ,.__.._. , SF i
6 i l I ! ! i
....._t_.-
I I I i ' i l j j t& 6F f l i
! l @NlYT --
j
'l t . _ . . . . _ \
i i i
.. OTHER CASES ,
F '
% Sr6 t-7q D/
i i l [
. .. Sr _.] I e Legend i
l'_ .
~
l i Humbers are sampling tank numbers j l
'g 1----- - r ; Gas Leak Detector Range i
I ' R: Sensitivity 1 R 2 Sensitivity 2 O L p', M ,4 R 3 Sensitivity 3 ' E-. ~~~~T YF 6 .l i i i Figure 6 Strip Chart Record From Gas-Leak Detector
~
907 00
~
4 . l l d n d i a Hl s R I E V g'5 I R + Z
-- C 6
q 6 d b -
\N n
l a A I s y N O~ e
~PA l
l e H 6 h S E U Q S R
,MU d n -
a S ls I ile M e e r h mm T sQ y$ r Y
- HE r @ oE{k2 8D$ w I %r ] h upyQ m-t i m o$ 5> *
,.l.N #g~ 1{E2 h ?.
bi NOM QO.3-
e .
- 10 l
1600
- 8. a~2; s 900 /
N s% '"
/ ~
m ,/r %'N\
~ .. ... .
gg " Sippb 4
@u g a 7 % 6 80* ' % 8 4 M (
3 0.6 68 33 @ 240* @ 12 0*
/ \
220* 14 0* c ,, ,...
\/ .. e.,, ,,, \ ' TEST 2 j 6 ' "'*"$ul " s " er'UY'"'" " ... ggE0MBE um es
tras O 10 g
- m ,.
e 20-e,900 / \#
\ ,
320* o , 8 EN90*/
\ *ee 94 300 / h400 p -
x g '~ g f /
'as @ 300rf '" % tr ,, 80* ~ *2 h Q ,'l i g b 's \ =
s s j
*** ,o, I , s'N s%
260*~
@ i \ \ ~~~, ' s oo* / s } g 's @
240
$84 / \ \ \ 838 \ , ' f g %
JSe @ @ 474 \ 4s4 1
/
220* " 40* l
\ ' /\
zoo N ,,;o._ ____.__nco '
/ \ \
I '
\ TEST 3 SF6 Concentrati n Durat ons yes.d3Direction @h2Ihk0$0 096 1407
s ,
- N 1600 go 340* 20
\
4 500 e k [ 88c 320' Oo \ 4o./
#d 400 ,,3 6 j
OO g 60* 10 0 g' 3s 4
@r$m @ N 174 /. @ . .6 280 % 32 0 0*'
26
- = * ,-
\ 33 23e l
j
~ 260*' @ \ %
ion '
\g dc 24s 220 240' , \
120* ter 4s 220- 43. i - I - 200 l 160*
/ \ TEST 4 l
SF6 Concentrations and Wind Directi Durations: Test 4
-86~ 10 l0 ih T Y h\
Le m M X( a .t/AL 1407 097
,o isoo s i i \
seem
'
- f"/'s sg f[ /
%,#s* g ) L 10.' %h / / N* , - \
o* r M y e %
/ / %,%o i ,9 i ~ ,og0S \ g \ l , /
u 2eo' .. 7 is gll ,- 8,. - g#8 g \' # ' * %,m
~
OA j 2ss- o g o.2 ~ ,00
- N e )40* 643 120 / N .
22o-
,200* / N slo # o i g o i. ie.. u,.o o,is ""' I MF5 Figure 11 SF6 Concentrations and Wind Direction q m- m ' n Durations : Test 5 ; 'w o c a d 1407 098
1600 l 10 3 \ 4 . 20 s e e 340
/
900 \ '8
/ \\0s r 40 x
32
/ s 400 266 N 6 p+ ..
s
- C, .- N ./
@ ns(poe) 300 in 10 0 48 O O = / ' 280 , . .
8 0* " f oa = 9
- .- ')-
+ @ ? g .
I ~ ............ ....... . . . . . . . . . . . 260 " ~ i g' 256 @ p
/
2 k, 230 @
/ ,'
12 0
'N 220* ,/'
e40' N i.
/ 180*
- 8. / '
\
I TEST 6 Figure 12 SF6 Concentrations and Wind Direction @ Durations: Test 6 3 [
.. hh a407 099
10 1600 l p O o
\ ,g 300*N 340 l 20" N \ \ \ %f$ \
AC' s 40( 320 g /\ G 400 N h 9 / eo-s N p , 100 eB 3 i , ef - 260 ~ tw @) . [@T[ l ,e n' 'e 6 &
; Die 4.
t
.260*'
( A h # 2'y / o @ 0.2 ) 120* 240
/
g South g ower *
/ fg ,. 140 220* I6 \
160 l o is less than 0.15 ppb 180' \ TEST 7 1 0 800 1600 2400 scales rodlot distancs (tes!) Figure 13 SF6 Concentrations and Wind Direction , , DJrati0 tis: Test 7 1407 iOg "
SOUTH TOWER WIND DATA TEST 7 N 6 OCT 71, 0205-0250 EDT 2 MINUTES 4 6 13 12 8 Figure 14 South Tower Wind Direction ' Durations: Test 7
\01 1A07
O* 10 1600 I 20 e 340* In#900 \ North A
\ 1, Tower [ N /
7 9 320' 53 s
\ k 6A / N, N
p / 60' 300*
\
[ = 10 0 4s g 34 12tpob) m
^
27 / \ O.8 80' ' f,jp/ I 7 [p, s e t / @ ; MOO -
\
260* , 7 x ,
~ @ t.2 ,
i4
'*@2..@i. ,
12 f 240' g South Tower j 220'
/ \
7- ,. w-i TEST 8 j O 800 1600 2400 scale: radict distance (f eet) Figure 15 SF6 Concentrations and Wind Direction Durations : Test 8 M N los 6 5 4 2 2 4 4 - SOUTH TOWER 6 MINUTES WIND DATA TEST 8 8 OCT 71, 2305-2350 EDT Figure 16 South Tower Wind Durations: Test 8 9 07 \03
O* go 1600 20 l 340* 20 N [ 4 / g e 900 \ f:3rth / ower 40'
,/
6 400 g 60* N k 4 10 1 280*s g _ D2
~ roy' 5 " L, 0; '*)
260"
@ @@se '
N 240* South g Tower \ 7 220*
!\/ 200* \ \
16 0 *
/ 18 0* \ TEST 9 0 800 1600 2400 Scale radial distance (feet)
Durat a s Te t 9
\hol \DA
SOUTH TOWER 3 WIND DATA N TEST 9 13 OCT 71, 0330-0415 EDT 13 4 4 4 2 2 2 2 2 Y 2 2 MINUTES Figure 18 South Tower Wind Direction Duration: Test 9 1407 105
0* iO 1600 I .P g 900 ,
.4 20- /
e / 8
\ % AE "e*" \/ /"
x\s4007 ::'; 8 ,::l x f
'O e
68 N 3 s k 7 N 4 10 0 16 @ @,,, ' \,
- o. @R @^ - 8 0* '
2ev ,, ,
,4 @#["1-_ p ' O.9 l
13 - E- f W 7 260 ' @Io.7 D] *
@ b@ - @ '.6 s
n 12 7 240* N g South Tower 140* 22g
/ \
200' l 16 0* 180 \ O is less than 0.15 ppb / l 0 800 1600 2400 scale s radial distance (feet) Figure 19 SF6 Concentr'ations and North Wind Direction Durations: Test 10 _\hol
\b
SOUTH TOWER WIND DATA
^ TEST 10 N 15 OCT 71, 0420-0505 EDT 6
6 2 - 6 2 MINUTES 10 Figure 20 South Tower Wind Durations: Test 10 1407 107
1600 O* i y A 20
'*k 340*
4 f N 4 A :" /
)Sk l
9 00 Q r - N G>
- 4 /
10 0 li
- p. e3 N_@c @A 80 -
280 ~ 'P .n t f,3g/
= sat 3, /@ ',a 4 9 5p2g F( '- 5 ,g 6 J
4 f.o (ppb) g
/ ,9Q 10 0 260 '
gg
* @;gp @ ^
C
- U ' ,,0 N
g South 220 o ( 00')
/ \
l
'S 1ess than o.is pp3 /200 N ,80. / *\ T E S T l l 800 1600 2400 scale: radial distance (f e et)
Figure 21 SF6 Concentrations and North Wind Direction Durations : Test 11 108 9 07
N 6 N 4 4 2 2 4 2 2 4
/
2 MINUTES SOUTH TOWER WIND DATA TEST ll 16 OCT 71, 0400-0445 EDT Figure 22 South Tower Wind Durations: Test 11
'g kh
M 6 N 4 4 2 2 4 2 2 4 2 MINUTES I SOUTH TOWER WIND DATA _ TEST II 16 OCT 71, 0400-0445 EDT Figure 22 South Tower Wind Durations: Test 11 jkh'
M-N PLAN 8 Turbine Building a Fuel " g Reactor Handling 9 Building Building p 200 -
/%
10, 11 g 12
+-E V h 150 -
h2
' ~
Reactor g Test 7 & 8 Building a Turbine Test W -+ y . Building 10, 11 ; 12 I O I i GRADE O 100 200 300 400 NORTH ELEVATION p 200 - g 3 150 - ; O 10 0 J g - Reactor
~ "I Fuel Handling 50 - Test 10, 11 & 12 Building e
E O '9 ' ' ' ' O 10 0 200 300 400 500 WEST ELEVATION
+N S-+
Figure 23 Building Structures Profile as Seen From the North and West
\\ ~
10 l f 348 20*
\ Towe N/
320' 40*
\ !
6 60* 300* [ j ,
.l.. '
I , j B2
' ' ' 100*.
260*' ' r BI 3 e
/
N 120' 240* f 12 / g South
/ Tower /
220*
\/ w ,eO s 'S { \
I TEST 12 0 800 1600 2400 scale: radial distance (feet) Figure 24 Balloon Locations and North Wind Direction Durations: Phase 3 (Test 12)
-100-w2n
SOUTH TOWER WIND DATA TEST 12 VERTICAL PROFILES N 12 NOV 71, 0047- 0132 EST 2 MINUTES 2 6 12 16 Fiaure 25 Soutli Tower Wind birection Durations: Phase 3 (Test 12)
-101- 1407 \\3
250 M g B 200 - B 4 p g 150 + 0 78 2 N W rB I 52 / E 100 .- o, TEST 12 50 '- O ': ' ' ' ' ' ' O 5 to 15 20 25 30 35 SF 6 CONCENTRATION (ppb) Figure 26 Average Vertical Concentration Profiles: Phase 3 (Test 12)
-102- \hol \\k
j i 0635 0700 0702 l 0.04 ~~0.3 W s i2I x td }0703 0.9 o E 0632 0638 0705
# 0.4 0.7 0.9 0640 0707 " 1.0 0.3 Nb '
0631 0641 0708
) 0742 Cpl i ~
nmu (, 3.6 0643 0710 0 C(
# V 0.5 V 7 %b J 06Y -
C 4 g 0.8 I 0607 52 O[ C w064H722
>. 5.2 C 0.7 0.4 Legend I <- ;f0740 2.5 88, 3 % 0646 0733 Time g W 0730 /
0612 0.2 O M , 0647
<n
[ p- 1.3 U T ;/ en 2 0649
# # 0.04 s E 0738 M 0.08 -
LEGEND ' w 0735 RB- UNIT I REACTOR BLOG. l C - NATURAL. DRAFT COOLING TOWERS 0618 N,S- NORTH AND EOUTH MET. TOWERS 0.03 0737 WM- WEATHER MEASURE TOWER (PHASE ! /) SITE ) V M Figure 27 Concentrations Along Rt 441: Test 10 Road Traverse . 2
-103- \401 \\S
,. VERTICAL _, ,, TRAVERSES y' PROFILES DIRECTION 235 160 160 190 2,00 235 200 205 240 225 --* N N ~ ~ > ~ > - - * - .
270 285 290 285 280 265 280 260 265 275 320 325
\\p n \ $ y / ,o y y, p
_ 2.0
= 4 _.
3 NORTH TOWER -
' 5 ;-
E E SPEED - 'O"
$ 2 _
E E k 0.5 E 2 _ m m m. m 3 O k2 o O O w 4' 040 3 '0'450" 0 510 0530 0550 0 6"10" 0630" 0650
^
0710 0'7'3'0 O
' 55 W .9 .5 .8 .6 1.0 .5 .7 .9 .7 .5 y $[. co E TIME (EDT) y @ DIRECTION $ 5 215 250 255 260 230 210 245 250 250 230 270 275 270 12 5 19 5 13 5 230 2
a g 7's* / / s ' ' ," -' ~ -* % 3(30 i si %295 ( >240 225 340 r
=
Z
._ 2.0 $ 84 _ SOUTH TOWER $ l.5 __
h r i o
- E 2 '
e
& - ! WIND SPEED <O 5 MPH o = =
@ o40420 l '_ c e4-c c c 0500 6 e e c05206 e o e05406 e e c06006 o-e-c 6 e e 0640 c 6 e c0700 04-o-c0720c 6 e o0740 ch> o g u - 0450 TIME (E DT) 0620
$ CD Be
0316 _
.'7 7 -
0314 \. , i '
.r ~ '~~~ 0 } 0 09 i .X } ~ $ v .. m S 0313 ,iJ/ \ ' '\ , %) .y i ,) 0.4 ./I ,
h
/ .,F ,' N ', 0306 \ ~
p!!h . 0311 0.6 h1 0.4 0305
/ ,. n.' ./
f
* N N(b ET 1.>s*
d t p 0301
'Y'g , [; .. y . . ~2 0254 '~-, \ . J :. - ~'""""' r- 3 s ' ~ .' k'hh :- 03 ,?- .l u. :Eg 9.05 f_0322. . t i*. , , , 5W I.
- . h J ' '
4 22 s ' 0256 j' 7' ' < 's
, , , ;f s7 -~-
M 1*3 i *
.~ ~0249 .
7 l
'"\ \ -'*i ,
3.2 '_ M ' f ,, m' ! ji ' ' t; 0334 0335 . ' 0247 / 2.0 2.0 'E 3.4 ' Legend p; ,...1.... - i , i( - ~ . . 0337 0338 0245 T ' Time '
.- .) ,
2.1 2.2 ['%' # T.Ti
- Ms .r .
ppb .
' ' ,' _Th/
a 2 0331 t ' y' 1.5 0325 ', ,9 2 0.1 " N - - /, y 4< vf;. . " 0328 -9 M e*-
'W 1 ' .M -> 0230 .N\ 0 . -
A 3 p,' N- .:.z%.",*i' %fu, 's "p-q *
s t t , ,
I s 0240 e 's ' c-mr .J.y. -g i i . :- 2 f2 ' - a * \,
- - - .x j
j>i ,
'g , - e, / - 'i \\ . -)
2 G... y,am. ' "- ". 0233 2.;- s
</yg.
p 7 . Y e , /'
- r s / {'s.ws." = 1;, scei. - -
[.il \t.
-w - x i \, ,e , -mm ~"s M/s _ j..g I:n , P.= 4, . , 'a .y p s 0236 9- .- ,t; *? .. p -_ b 0 ;' Ogr-< "i.5 \}
V QCi ics
\3,\ . jge -
_a j o.g ,; *
,jjj .t .:: ' f. h. j , k,. \('.-
M'
*$ 's 0235 'p%' l . ,%'f ) .b' k~n ' \% t -
Q N 1- 1. , ~, -l , '!
. ,; -- v , D x . g \ . , - . .. . . ... . oT _- . \ ' n ~..
3} ~ 0 0.5 150 s c- 's ' . s
" . 6'\ _ .
Mile Figure 29 Off-site Downwind Concentrations: h Test 12 Road Traverse O
@Rg -105-(A01 \\1
0301
' 0250 0.03 i 0313 0303 = -o .o 0255 /
0 03
\ * / 0.02 0310 0 31 = :::
0318 / M s s
/ 0250 0252 0.2 5
0.6 / [ 0249 1.0 0.3 0245 10 0248 0.4
\
[ 0243 024T O 2 0.6 c Legend ( ] I RB @ 0 Time ppb S , 0j t o0 c c
'A hI m
OWM i ('
\M s \ \
m \ E , 1
~ .E LEGEND RB- UNIT I PEACTO5 BLDG. i C - NATURAL DRAFT COOLING TOWERS N,S- NORTH. AND SOUTH MET. TOWERS WM- WEATHER MEASURE YOWER (PHASE I /
SITE ) Figure 30 Off-site Downwind Concentrations: Test 12 River Traverse
-106- \fo1 \\8
.; VERTICAL ,' i; TRAVERSES ,i PROFILES .
DIRECTION ISO 15 0 16 5 16 0 185 185 19 5 19 0 17 5 18 0 160 14 5 170 l10 14 5 17 0 ;75 18 0
\\III III "
I 1 \ I '\ l " I
= NORTH TOWER _
G -
. - 2:
i g 4 - SPEED y
~
i $ _ 9 - E g 2 - - '
$ .' S g. 'g w I I I I l l /l L 5.2 05 O Ew I45 135 135 14 0 10 5 15 0 15 0 170 180 160 150 14 5 16 5 15 0 16 0 15 5 155 16 0 ~
0 ic
\ *N \\ \\I I $\\$\$$ $$
8 _s
<I 6 G SOUTH TOWER ___y g g .. < N -
3 $ - 2 g I 4 _ SPEED _ cOc0 _ l
=
2
~
E J> I I I I I I
@ 010 0 0120 014 0 0200 0220 0240 0300 0320 0340 TIME (EST)
W 9 O
100,p 9
% _ .-. . . . . _ . _ - .- _ _ -_ .. . +m _ a. . . . . i.
s m-5 - r.4 ~. ... .,. -,. . .
.s - r~ . . , , - - .~.
Legend
, ;i,t + ++t+tt- -++t+ *t- 't b -t- -t- -t + -* r- r\ t H-r t+td t++- +++r t--+-
l'!FMH MtHF +t-H HH HF ++ + -
- H H-N+2 t* ei- & -~*
- - m nu m. = m m . . . . . _. m: .; m .-a u
- Maximum Test 10
-d-t.u__ m: _m. .u- n. : :r: n =
- .m; 2 Road Traverse m.33I:. g:
q = : .: .n._ n= n t q n-. t= u: y:.: ir
._ na : .: = 2 : + ._ .: *t - m- $ Test 12 Road Traverse -+-++,- $t c %--+ .x :; : ~+, + :ITT p)"; ++ , = + .,. . - + 4 -r:- p.;p i -4. - ". .:-- c 0 Test 12 River Traverse , ,- .,1 t: .
3 s - -*+++ e ' 1
- ~ "-
- ATest 12, Balloon
_ - - - _ w.. _ = _ =W.; ---~
=u =.r u =v ==x= ---n+ = n ..s. ==.
_ Position 4 at height m' w, % ,.
~
Wnn *
- e of 250 ft 2
- ~
j Predicted usine N y g Safety Guide 4 i wg ,,
"~
q: ~ g q q ..
", N, model u \ i ,
t+t-1 i , 1 i , ,r ,o , . . ,o, , ,, . i , , .. i i . .s .m , , i i . , i i n s. , i i i. , , 10J ' " i"w T' ' ' '"< ' i ' " ' ' i
..+- , , + + s +4 N k
[ gll T I I ] T l J T g i i l i ?1II s ii t i i 1 i iij i i 'f1 i i , X i i i ' i i 4.. ai. i i n in o' i m' i' i i','. "i '". ii i i '"- ' i i x' ' ' ' -
- g ., mg . - -
- o. --
3 Predicted using - - y .i. ' Model SW - [ as
,(" Sector Average , 'x ' , 1 m
g - _W.ske Model") .
, =. -
08 O ","u***1.. m e 1 i . 1 i" O I 1 -I I
' 1 I r' I I f ii 'sN I 4 1 ' 1 l L I Ij l { i X3 1] t3 +l3 [ l , 1 - t l 3 g i 3 i l .j i g i i I la4 61 g le . I E ,ii 3 j e f i ! 1. ' I ' 4 I l l *O g \ l pi[' q l l i 6 t I44i f
- 1. Q, i ' i o "u i ' 6
- w i i - 3 "'- ' + i ' 'ii '" "
~+ .
m
^' '
8 m s 3 1 3 if I l 31 1 i i ] I I 1. I!. I g i i e i i a i 1 i e i i 1 I+'i f<t tie i i1 a I I , 3I I I i6,, 4 3 i I Illi 1, i I ( . ..y iij, i 4ie i ;y3 g i i a i i ;
.+ .- -- ~. a .. , -__. __
7 _. .- - . g
~
r - k 1 I i I1 I i I . f11 l' 1 I L I I I i 1I 1 1 n- ; ; m t..c -- + - - ' In: - - - - - -
** '7 u -
- q. . ._
2m- +"- I 1 1 1 , Y i 1 I I I 11 al 11 - if 'l T Il l' i ,
! I f 1
3 i + 3 1 1.J i i I i i i ti la 4 . e (16' e , , i i 6 0.1, ' ' ' ' I A - ' ' ' = - ' " + 2 3 4 6 6 7 8 9 to P2 3 4 SP6 7 8 9 to 2 3 100 1,000 10,000 m D'q Wm D i Approximate Downwind Distance (m) a g 6 ,
.,\
Figure 32 Comparison of Measured Downwind Concentrations With Model SW (" Sector Average Wake Model")
-106- 1407 120
.= , APPENDIX A E6 Sampling and Analvtical Procedure 1407 121
LIST OF SYMBOLS C s True concentration of SF6 in the collected sample. C g Concentration of SF6 in the tank (diluted sample) . F Sampling flow rate into tank through needle Valve. P Sampling period (time). R Dilution ratio of sample. V Volume corresponding to maximum evacuation of the sampling tank. V "Make-up" volume, added after sample is collected. V, Volume of collected sample. V Total (net) volume of tank (tank capacity) .
~
g407 (22~ A-1
- APPENDIX A & Samoling and Analvtical Procedure A.1 INTRODUCTION As outlined previously in Sections 1.1 and 5, SF6 tracer gas is released under controlled conditions; resultant concentrations are deter-mined at 18 discrete circumferential points based on air samples collected in evacuated tanks over a finite sampling period. The apparatus and pro-cedures are described in detail below.
A.2 APPARATUS A1D PROCEDURES A.2.1 The Samuling Tanks The sampling tanks used are low pressure oxygen cyclinders with a nominal internal volume of 16 liters. Although 18 such tanks are used for any given test, 20 are readied such that 2 extra are available as spares. Each tank is fitted at one end with a 0-30 in-of-mer-ury vacuum gaoge and a valve. The other end is fitted with a Moore Products Model do. 63SU Constant Diff erential Flow Controller, a fine-adjustment needle valve, and a quick-release valve. (All fittings are rendered air-tight with Teflon tape.) (See Figure 4). A.2.2 Tank Capacity The first step was to accurately determine the actual capacity of each tank, V . This was accomplished by filling each tank (without fit-tings) with water by submersion. The tank was then slowly allowed to drain. The compensating air entered the opposite end through a wet test meter which thus yielded the net volume. The average capacity of the
^~ \
tanks is 16.1 liters each. , A.2.3 Samoling The samples of air are obtained by first evacuating the tanks to a pressure of 1 to 2 inches of mercury, and then slowly allowing the tanks to fill through the needle valve at a controlled rate. It was determined experimentally that, if about 8 in or more of vacuum is maintained in the tank, the cylinder will fill at a constant flow rate. If this rate is set at 0.20 liters min-1, the corresponding sampling time is in excess of one hour, However, to provide a margin of safety, a sampling period of about 45 minutes was used. The flow rate was set on each tank first at the TRC lab , and then again at the TMI site as a check before each test; this is done by evacua-ing the tanks and monitoring their filling through a 0-1 liter min-1 range rotameter. Although the product of the flow rate and the sampling period should yield the volume of the sample, a more accurate approach is taken as follows. A.2.4 Sample Volume First the volume corresponding to a fully evacuated tank, V , is measured using the wet test meter. This differs from the net volume, V , because of the residual air which remains after the evacuation. However, at the end of the sampling period, the tanks are returned to the on-site lab in a still partially evacuated state. In order to ensure against possible leakage which might introduce SF6 from the contaminated environ-mental air, the cylinders are immediately filled with " clean" bottled air. A-3 \
This "make-up" volume, V ,, is also measured with the wet test meter. Since a direct connection between the tank and the nitrogen bottle might cause the sampling cylinder to over-pressurize, a balloon is used as an intermediate. We may now calculate the sample volume from the expression V s
= V e -V m . (1)
The amount of dilution of the sample is thus given by the dilution ratio: V R = y . (2) s A.2.5 Tracer Gas Concentration We come now to the determination of the quantity of SF6 in the tank. After the cylinder has been brought to zero vacuum, the vacuum gauge appar-atus at one end of the tank is removed. (This leaves only a 1/4-in opening to the room air, which permits negligible exchange during the short measure-ment time.) A probe is then inserted into the cylinder such that sample air is drawn into an Analog Technology Corp. Model 112B Tracer-Gas Leak Detector. This instrument produces a voltage output which is recorded in an Esterline Angus 'lodel T1718 strip chart recorder. The detector offers four sensitiv-ity ranges which adequately span the concentrations of SF6 encountered in this study. Operated in the " column mode", the minimum detectable concen-tration is 0.01 pub. Each tank is measured twice on each of two detector-recorders such that instrument malfunction can be immediately discovered; this also provides a real-time check on possible operator error. Both of the detectors used in this study were calibrated by the tanu-facturer. However, the instruments were recalibrated at the TRC labs using A-4 \k
purchased concentrations of SF6 certified correct at 2.5, 250, and 2500 ppb. A.3 PURGE PROCEDURES Following measurements of all tank concentrations, the contents were purged using clean bottled air until the gas detector probe measured a back-ground level in the tanta. The valves were closed and the tanks stored until the next test. A4 SAMPLE CALCULATION Total tank capacity, V = 16.1 liters Volume evacuated, V = 14.2 liters e _ Preset flow rate = 0.20 liters min-1 Make-up volume, V = 5.3 liters Therefore. the sample volume V = V -V s e m
= 14.2 - 5.3 liters (1) = 8.9 liters.
dow, the dilution ratio V R = y s 16.1 ( 2^) 8.9
= 1.8 .
Thus, if the tank concentration, C , of SF6 is 50 ppb, then the true concentration of the sample is C = RC s t
= 1.8 x 50 pob (3) = 90 opb. ] \.
If the sample period P was 50 minutes, the flow rate F through the needle valve would be V F = p 8.9 liters g) 50 min
= 0.18 liters min-l.
lience, this indicates that the flow rate as initially adjusted in the lab was set (0.20 - 0.18 = ) 0.02 liters min-1 too low. Figure A-1 shows one of the work sheets that were used during the tests to perform the abeee calculations. A-6 1407 \27
*
- o Location C MS U M- #44A I E. Test No.
O Tank No. Position No. / __ Flow Rate O .1O liters / min. Tankcapacity/f'N iters Time Vacuum (Inches Eg)
* (Take Reading, OPEN Valve)
Y:ET 2SL4 5:of 2z 5: If /t & 5: sf /f S: 3f 12-(CLOSE Valve, Take Reading) Vacuum at time makeup air added: Ib inches Hg. Time makeup air added: b-Volume corresponding to initial vacuum: / ./ litera r Volume of makeup air: [. A 1*_ters Sample Volume: *b liters SF concentration in tank: 6
/o3 ppd Dilution ratio: /'
True SF C ncentrati n 6
/ ppb Figure A-1 Example of Work Sheets Used to Compute Tank Concentrations A-7 i407128
,. , e APPENDIX B Reduction of Meteorological Data 1407 129
. . s APPENDIX B Reduction of Meteorological Data On-site meteorological measurements are described in Section 6.0. Wind at 100 feet above ground at both the north and south tower is measured con-tinuoisly by Beckman and Whitley short vane anemometers. Typical traces are shown in Figures B-1 and B-2. Wind at 30 ft above ground was recorded only during the Phase 1 SF6 release periods from a Weather Measure Model W1034-540 anemometer situated in the center of the Phase 1 sampling grid. This is the same location as.the SF6 gas release point for Phase 1 tests. A typical trace is shown by Figure B-3.
Temperature differences between 150 and 25 feet on the north tower are measured by matched thermistors housed in Geotech aspirated radiation shields and recorded continuously. A typical AT trace is shown in Figure B-4. Average wind speed (u), average wind direction (9), and directional range (e) were taken over each miner chart division (Figure B-3). When the chart drive was non-uniform, periodic time checks were made manually and a linear time scale assumed between marks. Average one-minute values were estimated from the Analog chart. Direction data were also extracted in a similar manner. During periods of calm or chart drive problems correc-tion for true direction was made on the basis of smoke plumes from Federal H-C 3-minute smoke candles. North and south tower wind data (used primarily with Phase 2 and 3 tests) were extracted every minute. From these data, values of average direction and average speed were computed during the test interval. The total wind direction meander (or range 6) and the standard deviation o g were also computed. Values of c were also computed from the bivane B-1 1407 130
. e located on the north weather tower.
Temperature diff erence between 150 and 25 feet was averaged over the entire period of SF6 sampling (usually 40-50 minutes). Since this difference is expressed in *F per 125 ft, it was multiplied by 1.45 to convert to *C/100m. All pertinent extracted weather data from the three tower locations are presented in Table B-1. 1407 13I B-2
g < d APPENDIX B FIGURES B-3 1407 132
,
- w #
TS eaa -- gq_; -.7 _g_ ._
-,_ _g- 3 % _ w y- , _+_._. g_w- t._,g._%_3_.,__.. - g - -- J T k. . __ g _.f - - HH M- b '[% _+. - .h -_wy _'y _ + y . :M T=T =T .4, W $ - 'g *w -v+,_.- _+_.-:1.*---*--*-+-1:T r-rcc__+c_n(~c= -CM ^~
- = +
e - =- :==.v -. %=o
= ~- & & '
_ ._.-_v__,.
.t w = " 'k .W&'-Y~W'4yn --- a_-* C.c_mg==e---- - :: '* 'N D 01reetion *"Y9_l_%
4 -+-L+--._
.-@5 g . y_ N A.4_]4 2 E WEEED @ _.g _ y_ _ 4 _, g_ _
1 t m--+-
-+ 2. - - . =r p_3~.x. "V:- - .. _ m.+ ---, 4- -+ - - + = =.A_- - _, .-g. ..- -~ . . : .t- a
_m-- =
=- , - +_ -L__ _4. w =t-:=_ ,
y.
"I --M__---- + t ~ + -+-- i.=24EI J=Ir Qt- -~t= Et=1 +4 t=-+ __. - _ . _2 . . _ .
r- -r r E g F.)
.m __4____,. =. _r- _ + -+ t_,__ ,_L ._. - + - - -. +-
_4=f'_=.
==
_~ s
-m -t - =E_ % ._ . : -- -4
- a. -_ y-_ =y+= --: :_ --- x. y =--
4 =_--
== -_ z =_- _,
_g _ s-- - 5
-+
3=.
-= -_. - _x = ==_
p_ .--x 2 _ r-
-= 22 ._m -- .= . == .- n ==. ..
_m % =- -
-_ny-- -
3=
}~ =1t== 5.='= = == =_ =E _:
3 ~
-4 = =-= i= =:_ _TE. _- .2 ~T .= ,=== =1 m-- ) i}.ri. JRr. m -. n .a - -m . = - ._: :. --=. := . . _ , =b :_--.5.= _::b. . ... *w g @Mait$NR*ekhWMJR r r--
3.- : t Tk- , . , -~t-g=
- - 1I, Q MPM _..r .
at=-t-t _ ,_s -+- =:tN, - n=. t=. -,=n. n -,=.,n~~ ,_. _ . . -= 3 .. . .,_,- - :-
~~_.: w=1- q . > - __r: - ,--+ --l y- I: - m-f-i :.-121==f=^ --- f-fi f =J=f : = ~ ~.: -~ '~. Q ~=f ~f ~ f i' lr un ' - ^ --1;EEf 'El 1: 4 A = -g==r ~[F ~g-g.iff=gl21=fg-1 f- 2-.._-
x: d't=-h- a
,= }D'f. l-*--(~i a:. --- ~-t : 2 .--- 'f~'
g__ & QE-^
=c -Wt I;/El ]:. / / ~ D Lf1 ]-~ff.~~j .j- ^==r l-~ * ~~ ~ * * ..,....l-_...,:.-. - --r ~2= / / f f = i I T / w 4^~' n-s---/ ~ ~~- ~J=~== '_-j -
f-g n t. >i --/ =j:=i -- - - . ~l
;. ;_. . E _
D; : E73fY~ if=ht&M $-@. htF j.-j -jbff/ l L./6/ r 6 :-:: : :/222:< E SEP 241971 a E. _1 1 1 1 1 1: ; r - g ; 1 1 1 g , g v2_.~. v_<_ _ u _ ._c\... u. _ _ 1 1 1 1 1 1 1 1 1 1 y 1__g_ 4 _ 1 1 1 1 1
\
1 1 1
\
1 1 1 1 1 1 1 1_
\ \
1 -t North Tower 1 \ \ ' i 1 \ \ '
\-\ ' \ 100 ft Wind Speed i
1 1 1 i 1 1 1
\ \ s 1 \
1 1 1
\ \ --t- +- Scale: 0-60 mph 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i i i i i ' i 1 1 \ \ \ \ i i ' -' '
s 1 .
-H i t-,
p_g._ g mr u, t-1 1 i i
; ; 1 i i i t g ; i , i, i, i i 1 1 1
1 1, i ! 1 1 i i i t"t"t-i i, i i > , i, i, 1 1 t-- i i t i i i i i i s i i i, 1 1 1 1 i i 1
; ; ; i i ; ; i i, i i 1 1 . 1 1
i 1 1 l 1 i t-- , 1 i 1 - -t-- +t--- - - 1 j .g i i
- ; - _ x _n. . .r . + _ - -l l l - ; , l '
I I l f i l I 1 I I I I I f I I I I f I ,I I ' J 1--+--4_l p [ I I F ! I
~ & I > l I I I 1
l i l 1 1 I 1 1 1 1 I T I I I I 1 I i 1 l l
; I i
i f__ __, g_, p__Z-I 1 > > 1 1 i
; > < 1 1 l 1 l l l > 1 1 ,
l l 1 i l 1 +---+--t-- l"" "t-i l ' ' l l l 1 1 l l l l '_ ' l l l l l l---l---I 1 l- l ' I I i 1 I i 1 2 I
' _1 1 ' ' ' 1 l l '
l l l ' ' l l l ' l u -- ' 1 4 '- l-4 1 l l l ' l l l ,
'~~ ,nl/ i , , . . . ' '
l 1 l l 1 i l 1 -+-- > l 1 -l , t---r - M l "t " il
+--' l l I l , l , 1 l 1 l l k-/-- , l l ' ' ~ l 1 I I J 1 I 1 1 wI - - ~
s l 1
' l - _ -- , _ - e __. _ g_..]
1 E l
- ~ - ,
g SEP 241971 g a
- 3 " 'T D** a ae c , e!J . .. :'
Figure B-1 Example of Wind Speed and Direction Charts: North Tower
'l407 133
, s . a TCU
\ \ \
090 - - i i 4 4.u. . -+ - 360 t 1
- 1 l ~
i !'
. ..r i . e 270 - t* . j ! l ! ; ! >
t ! .
. .: . j ... . . _ . . . i l 1 . i
- y
; - . a . . . .. _ . .. 3 8o - .
i ! ,
's ! . l. , ; . . . i + .. - = i.. . .: .-.i l l. i / i South Tower -
opo i .
~
i 100 ft Wind Direction / / l . -
,- .i - _ .. . .
7 i _ . _-. - :n . / l _
~
_E SEP 241971 E a ___
=
Mi %-~T.T _ 1-- E _\~ iN:.'-1.k:. g."AT ?G'."M.7__T
._.y 1 ' %.MM x \ ' . M. \ ' .'M .
I- 4.__\ 7 '; s 'g _+,_'gT ,- -t+x 3 1
\, 1 \- ..; t; :: \_c _p:*C . . t ?f-- r , _ w .2C.7_MQ::x_._w_} x , m.
s _u_a 1 i z._.z_. ; I: _ _ _ -
\ '1 '1 1 \ 1 ---k--'- 1 1 . { [._ " .I'..b. Z ..^ l!_.-*l- $. i Y~i k - t-Y " 1' I - - ~ -'
1 i i i
.. ) . in:t ' d_ .C.$--* +-inh.M ' .t--~.$-4--+--+ I I i .. . ... . .v ...y_. .4.. .- ~ ' ~ ~ '. ~ 4 ~ I '. ^[. _.'._d>~ i 4 .. ., 4 ,.. ' ! .I ! !
I y 'I i
..t _ . .y : .r.- i :q4 1 ::~.t.--:. -1 :.-ny_r i _.._4_4 j - ,- '* " - + -
- t . g T;=:n.4 >
l 1 1
'4 : ,T'- l l ",t- 1 > .t T i j - * - - - * - -
- f. _.. To.. j:.: r. +._.V'~.f_~ . . . ._+ . f_._.1[.:_t. :.f,---f. _, g_
t-1:.-- . . ,., South Tower
...l,.-.n. ^
- j : +-" J- *- -+--+-- 1
. }t-l * '~
1
! , l l ~ ,
00 ft Wind Speed r
- ..l.
- cale: 0-60 mph 7Z 3, _-f - t- -t_;: 7. ._/.'...f "
- = ..a-_a __-'.:"t.- tm n , _ _ i ,z
/
J 7.- :--"--l--)-f.;r- ~ 7:---t-- +--ft? 'r l, l::;c: nj . 1=
- =C"-::l_ :1._ :f : ?.:= # ! w > ~
; s ; z -,'y:- -fr-- .? .;f: .2. ^j- -. M m / . ?- . / w.: Z -t-.n-.
s-,-t
-t- -_1 _ ' ~a=. _ ~ r - W l =m l ~
SEP 241971 E E m co j Figure B-2 I J j Example of Wind Speed and Direction Charts: South Tower B-5
~
6 .1 N. . s' , sg, -e?? ':
*b ., '
t' 23 M !
; ; i. : : , . i, w , Ac. ,! ! . ,.,,, ; i . =
h., .- :i-. '. -s, . .. 3,
.llljiN: ' %
ti
. ; .v n , ' . -t 51, N ' V.
Nq - l' j e'i,
.: ;- g i y[! . .i 'l! ,'[f Q.U'0f' k ! '!! ll : {'. +# .i-}' R - \s. ' N.. 'N s. % j; ' f j';. i , W ,;i - ' # ~ ';
ll l i i: +i . lit l i : h. . !
' s. . N ji %; ;, /- _._
N[..l.,6 i ;, i .', \+;!,l,;' i t--
- < ,_ 'i>- .
O s' 's' s N' ~ ---- ~ i .
' .i ii.' .li t .> ; !..i ; :t . / N .. ' ,.'.~- .. 'N, '
l 8 i t: i / N.l
'3' ,.ix ,i i6. , :? . . . . ' t 4 *t ;l 'i j ! l.J,,,. , . .i - -
s . _,. N 't .. . ,.D
'! il .. i ,N , ' ! ' ,; ', .i-Li___L -. - 1 I '. . . i. 10 nag;' s,. .
g! ,
!l : . 'l ;; ;-
oN %L,h. )h Mls! AV g
~ +I i;,i:j 't. i q'!;' W N llH -
ix- N
..f.
t t- .i. i. ' i
,; ! i '&
4
% y .s . - ' .' --
N ..n-
', / ; ;l ' .44- ...
M,W*Wb'.
, .. . ii %la. l! 'w..
l! @t.ij-t]il ii ..N
~
M j N 3,-
- r. a. eI s,4
.-j. n. x %:. . ;pa.i t; . ii ]' y?;';T; }+l ii - j, _ <' ': ;'k 4, 6 %m!! Qhq . g;~; -i-p , ~
s,,,l':d.4
!. N- , i s r T ' - f 'N N N s
N
,:l. q.1 .; - 4. g-N- i -- ~
l ,
.- ai . t .,,
K; ~
; t= 1 - .- : ,2 ' . 'N lhs ~.;. ii'r- '
j' .%,. 3/
'N N \ , j l v
- 1. -
8 NSj$- l, f . i 8 N a!. AN .
-' N .A.. ^
N -
.l 9 1 !' , ,,,-' s ,, ' !
s .,- -
.,l . p -i i ,,,- qh\ ...,' ;a- N.ss . .. h_/ I - 1 -
g w .- s
._~ '~ g, . _/ - , '. ,l -
N N *I .1 .
' i/i ,. 'N @l @Q W'-
I
~
l ' Y' 4
~ ~l .,,
M - Hb . l- '
- N.N,..,
.;4 i
i i .l..{ 7 I i. N, tg . -- g f. l ,, - s, N i - s . . _ . ._9 . j . .
.i . ,,,.
x s ! r- ~ . , , - - i. s l - t ,
.- ,;, 4y - ..
q , i
.I- ~- ~ .1. I - - i ! '6 J !N(' s Ns ' ' - %' ~l NN s iI'T s y ~j - q ,- .
o: P
,i-N .
g
!~.. ! .p ioo 3 /S$yg y' g N x. - '5N ,,
N
,N s, , /beho iM s, $,. W.
g . -- .-
/
x.
's s ':-fS,- ' ' q% --l
- j. s ., - j. , l.. . -
j -
- - ~.
Figure B-3 Example of Wind Speed and Direction Charts from the 30 ft Weather Measure ! B.-6 1407 0 5
- ~ ~[ ~ ~~ } .___ ..
~~ ~~
__: _3 a
'-~ ~ - ~
i __ _ _q r __ pp. p ____
._ .)_j_ 4_ . .__ g l .______, _ q _. #L1 a .h ! ^ Ti 20 - _4 _
- g. 4p ip: rz
- g. , w. cx
~
_]' ! i
,( i -f! fhf- - -} _ ', %g L N-- -j y _4 -
_-} - w -
;(, ,
ES - '
~
ir , -- me, ' ___ pr p .m 1 _ _ _ _ _ 7p go ' O_
~ =}, ==- --- - . _ - - .= =.- = = . - --
2 c. , __g .
. ___ g _ _ _ _ _ _ . . _ _.. _ g . _ p.._ .. g - __.
_ _ _ _ _ _p "_3 _-~J . _ _ ___1~- a
}___ ~ _ _ _ ~.__1 - - _ep -- f .~ ~~__.h l
w Figure B-4 Examole 150 ft-25 ft AT Strip Chart G 1407 136 B-7
h .. t s Table B-1 Weather Data Summary 30 ft Weather Measure (South Field) North 100 ft Tower South 100 ft rower Gb8
# R # Ne R,/o, T Test Speed 8 8 8 8 8 Speed 2 0 0 8 'e amb RI! Speed 8 8 8 8 "8 2 0.62 132 150 37.3 4. 0 1.25 4.26 119 43 10.7 4. 0 29 7. 4 3. 9 46.6 70.0 1. 2 93 95 26.6 3. 6 3 0.20 142 168 49.4 3. 4 1.65 2.97 93 95 26.6 3. 6 28 6.1 4. 6 66.8 76.3 1.51 84 213 74.1 2. 9 4 0.19 41 175 46.7 3. 7 0.58 2.83 276 190 48.1 3. 6 LW LW LW 67.8 76.6 1. 2 270 112 19.4 5. 8 5 0.15 9 167 40.1 4. 2 1.20 0. 7 34 30 6. 3 4. 8 18 4. 2 4. 2 DD DD 0.98 0 65 12.5 5. 2 , 6 0.37 131 162 55.4 2. 9 0.76 2.05 102 195 73.3 2. 7 LW LW LW 52.5 56.5 0.67 266 140 47.2 3. 0 E 2. 0 1.12 7 210 85 NR NR 4. 4 196 31 7. 6 4.1 31 6. 3 4. 9 62.0 85.0 1.54 195 65 14.2 4. 6 8 1. 5 100 130 NR NR 1.79 0.78 130 56 14.2 4. 0 26 5.1 5.1 50.0 60.0 1. 3 80 85 22.7 3. 8 9 1. 0 95 70 NR NR 0. 9 5. 2 40 165 67.8 2. 4 LW LW LW 43.5 91.5 0.65 19 150 63.0 2. 4 10 NM NM NM NM NM 0. 6 11.6 282 35 9. 6 3. 6 LW LW LW 50.8 76.5 0.15 251 82 21.6 3. 8 11 NM NM NM NM NM 0.87 11.6 97 60 18.2 3. 3 LW LW LW 50.8 89.8 0.56 11 175 54.3 3. 2 12 NM NM NM NM NM 0.91 3.14 178 75 19.1 3. 9 60 15.1 4. 0 15.0 67.0 1. 8 151 11 3. 8 2. 9 Note: NM = No measurement taken NR = Not reduced BD = Bad data LW = Wind speed too slow for response s
CD N N
REFFRENCES
- 1. Slade, D. H., Editor, 1968: Meteorology and Atomic Energy, 1968.
AEC, TID-24190, CFSTI, p. 153.
- 2. Ibid, p. 251.
- 3. Ibid, p. 112
- 4. Safety Guide 4, 1970: Safety Guides for Water Cooled Nuclear Power Plants. U. S. AEC, Division of leactor Standards, Washington.
- 5. Collins, G. F., F. E. Bartlett, A. Turk, S . M. Edmonds and H. L. Mark, 1965: A Preliminary Evaluation of Gas Air Tracers. Jour. Air Polln. Control Assoc., 15,, 1965, p. 109-112.
- 6. Turk, A., S. M. Edmonds, H. L. Mark, and G. F. Collins, 1968: Sulfur Hexaflouride as a Gas-Air Tracer. Env. Sci. and Tech., 2, p. 44-48.
- 7. Turner, D. B.: Workbook of Atmospheric Depression Estimates, PHS Publication No. 999 AD-26,1967, p. 37.
1407 138}}