ML18024A773

Direct Radiation Levels Around Browns Ferry Nuclear Plant - Background Data. Draft Rept Encl

ML18024A773
Person / Time
Site:	Browns Ferry
Issue date:	02/28/1977
From:	Doty R, Jenkins P TENNESSEE VALLEY AUTHORITY
To:
Shared Package
ML18024A772	List: ML18024A771 ML18024A773
References
I-RH-77-1, NUDOCS 7904240303
Download: ML18024A773 (60)

Text

ENCLOSURE 3

~1'ENNESSEE VALLEY AUTHORITY g ~ ~

This Report Contains Preliminary Information Subje'ct to Change or Revision and Is Intended for Use Within TVA Only

+ gT DIVISION OF ENVIRONMENTALPLANNING

f ' ~

DIRECT RADIATION LEVELS AROUiilD BR04%aS FERRY NUCLEAR PLAIT -- OPERATIONAL DATA BY PHILLIP H. JERKINS A'.iD RICHARD L. DOTY

~ TENNESSEE VALLEY AUTHORITY HUSCLE SHOALS, ALABAMA

CONTENTS

~Pa e List of Tables Acknowledgements .

iii iv Sections Introduction Experimental Phase Discussion Summary and Conclusions References

TABLES No.

~Pa e Summary of Data Direct Radia tion Levels Operational Phase Brovns Ferry Nuclear Plant Rate lfeasurements at N 6-3 for Each Instrument Summary of Normalized Rate Heasurements P t;.

0--" '

LAs ~

'" f

~r ~~ ~ E'~~~'~S 6 n

y~ g4V sL EI'~

ahE4 L<~

P' l l; " A

/$

L0C,~ s~ QL4 QA I

<pgsz I ~ t a.v o.wE,~A-$

h

+~ I r3 i HAIQ >~

~

r ST,/oS.qs~ .-~SS'SW C ST Lift S>

(~LI4 i'g~~

AC1QlOWLEDGEHENTS This report is submitted in partial fulfillment of Energy Accomplishment Plan 80 BD'f under terms of Interagency Agreement EPA-IAG-D8-E721, Sub-agreement Number 5, with the Environmental Protection Agency (EPA). The Project Directors are E. A. Belvin and H. R. Hickey, and the Project Officer is G. D'Alessio of the EPA.

INTRODUCTION In this report.results of measurements of direct radiation levels around an operating nuclear power plant are presented. The data were collected between July 1976,, and July 1977, and these data are compared to similar data collected between December 1975, and June 1976. Because the plant was not operating during that earlier period of time, these earlier data could be considered to be background or control data. The earlier data were the subject of 4 previous report, hereafter referred to as "Report l."1 Readers of the present report are encouraged to read Report 1 for inforrha-tion on materials and methods common to 'the two periods of data collection.

Three boiling water reactors make up the power-generating capacity of TVA's Browns Ferry Nuclear Plant (BFNP). Each of the units is capable of producing approximately 1,152 megawatts (Mi) of electricity, making the complex one of the largest nuclear power facilities in the world and an excellent station on which to base studies of the impact of nuclear facility operation. Data

. discussed in this "and the previous report were obtained at BFNP utilizing .

pressurized ionization chambers.

cd~

EXPERIh'ENTAL PHASE Data collected between. July 15, 1976, and July 28, 1977, are presented in.

this report. A discussion of the instrumentation and methods used for the collection of the data can be'found in Report l. A discussion of the measure-ment locations around BFHP, and the terminology used to identify the locations, also can be found in Report 1. The measurement locations are indicated on the site drawing in Figure 1.

f Operations were resumed at BFNP in September 1976; therefore, most of the data presented in this report was collected while the plant was operating.

g total of 1502 rate measurements and 211 integral measurements were made.

All data collection was performed during the day, primarily between the hours of 8:30 a.m. and 3:00 p.m. CST. A summary of these data is presented qgE,p w bIW l

in Table 1. The format of Table (Table 1) except that the table At five of the 83 1 is in Report the same as+a similar table in Report 1 included data collected at night.

locations at which data were collected for Report 1, no 1

data were'collected for this report. However, all 83 locations are indicated in Table 1 to facilitate a comparison between this table and the similar .table in Report l.

Table 1. SUlQfARY OF DATA DIRECT RADIATION LEVELS OPERATIONAL P~SE BROWS FERRY NUCLEAR PLihiT Rate UR/h Inte ral ljR/h Location No. Mean Std. Dev. No. 1fean , Std. Dev.

N 1-1 8 7.79 1 13. 65 N 2-1 16 10. 39~ 1.59 6 11.61 0. 90 N 2-2 6 11.28~ 0.53 6 11.46 0. 62 N 2-3 20 13.37~ 2.08 18 13.78 '1. 56 3-1 2 8.84/ n.oo 0 N

N N

4-2 5-1 ll 18 9.21~

9.97

0. 30 0.26 0

1 10.31 N 5-2 1 92" 0 N 5-3 13 9. Os i 0.16 0 N 6-1 2 9. 89~ 0.17 1 9.85 N 6-2 13 0.20 0 N 6-3 . 208 lo 06/ 0.29 6 10.26 0.25 N 6-4 20 9. 79~ 0.65 0 NNE l-l 8 15.41~

8.41/

6.03 0.99 1 6.76 8.58 0.53 NNE 2-1 16 4 NNE 2-3 16 9.29 0.68 0 NNE 2-4 0 0

.NNE 3-3 2 10. 0.26 0' NNE 3-4 '0 NNE 4-1 0 0 NNE 5;1 25. 2'.46'.06/

0.63 10 8.39 0.22 NNE 6-1 26 0.37 3 8.98 0.17 NE l-l 8, 17.08 7.S2~

6.88 1.17 1 7.51 NE 2-1 2 0 NE 2-2 10 13.36'- 2.89 1 13.91 NE 3-1 13 9. 10 0.56 0 NE 4-1 9 S.17'0.04

0. 46 0 NE 4-2'E 9 0.25 0 5-1 7 10.09+ 0.19 0 NE 6-1 1 9. 79~ '0 Std Dev. = Standard deviation

) 0.

g F

Table 1. (CONTINUED) SlPPQRY OF DATA DIRECT RADIATION LEVELS OPERATIONAL PHASE BROGANS FERRY iilJCLEAR PLANT Location ENE ENE 1-1 2-2 No.

6 14

~tel ~ Hean

12. 33'.

38 R/t Std. Dev.

3.86 0.94 No..

0 0

~/

lnte tfean ra'1 Std. Dev.

~ o<

ENE 2-3 9 ll. 29 3.05 0 ENE 3-2 4 8.74~ 0.58 0 ENE 4-1 31 9.89 0.35 16 10. 02 0.50 ENE 5-1 37 9.8& 0.35 8 9.97 .0.17 ENE 5-2 32 10.30~ 0. 62 2 8.67 2.16 ENE 6-2 21 7.91 .0.32 0 E 2-1 26 9.60'1.20~

1.82 0 E 2-2 8 3.41 0 E 3-1 10 7i61> 0.38 0 E 3-2 8 8.08~ 0.59 1 7.55 E 4-1 30 8.55~ 0.45 14 8.53 0.45 E 5-1 24 10.39/ 0.31 7 10.44 0.24 ESE l-l 0 0 ESE 2-1 7 9. 17' 2.08 0 ESE 3-3. 2 53~ 0.28 0 ESE 4-3. 26 0. 40 8 9.88 0. 40, SE l-l 25 9.65'6.60/

14. 43 0 SE 2-1 37 13.66 3.37 3 13. 86 5. 12";

SE 3-1 20 10. 10~ 0.23 4 10.04 0.21 SSE 1-1 28 ll.21- 5.97 1 8.19 SSE 2-1 36 8.27m 1.19 5 7.77 1.36 S 1-1 27. 11.72 3.27 0 S 1-2 21 7.40 1.68 0 SSW l-l 27 11.21~ , 3.42

'l. 14 0

SSW 1-2 19 9.06~ 0 SW l-l , 27 11.21 3.27 1 10.90 8.96 SW 1-2 . 22 9.62 1.29 1 l-l

~

WSW 25 12.09j 3.57 0 WSW 1-2 21 9.00 0.86 3 9.35 0; W l-l 25 13. 29~ 3.30 0

'3 03'.53

0. 51 W 2-1 20 8.83 0.66 WNC l<<l 30 10.30 2 '24. 84 4.62

'.56

'3.49 7.85 2-1 17 0.44 10 WNW WNW 2-2 36 77'.60 i -

0.95 13 9.15 0.71 NW I-l 17 21.82- 5.83 1 '23.19 NW 2-1 10. 27>'0.

0.68 0 NW 3-1 21" 0 NE 3-2 9.80 0.58 1 10.58

Table l. (CONTINUED) SUlQfARY OF DATA DIRECT RADIATION LEVELS OPERATIONAL PHASE BROWNS FERRY NUCLEAR PLANT Rate gR/h Inte ral gR/h Location Mean Std. Dev. No. Mean Std. Dev.

NW 4-1 0 0 NNW l-l 30 18.51'.85 5. 97 24 19.64 5.62 NNW 2-2 9 1. 19 1 9.52 NNW 3-2 8 8.20~ 0.37 0 NNW 3-3 8 7.60 0.27 0 NNW 4-1 10 9. 91~, 0.57 0 NNW 4-2 6 9. OB'.

0.56 0 NNW 4-3 9 06 0..22 0 NNW 5-1 26 lO.33'-79 0.22 2 10.31 0.08 NNW 5-2 33 0.28 6 9.79 0.19 NNW 5-3 22 9.63~ 0.31 2 9.49 0.11 NNW 5-4 32 8.92~ 0.49 9 9.38 0.48 NNW 6-1 25 10.49 0.29 10.54 0. 14

/[rr>

~

DISCUSSION VARIATION fQ10iC'NSTRPlENTS A discussion regarding variations of measurements due to differences among the instruments was presented- in Report l. A one-way analysis of variance (ANOVA) performed on rate measurements made during the day at location N 6-3 indicated that there were statistically significant differences among the instruments. In the discussion it was recommended that a rigorous calibra-tion should be performed to eliminate or compensate for these differences.

The authors felt that for this to,be done properly all five instruments should be calibrated at the same time. Unfortunately, there was not an occasion since very early in the study when all five instruments were operable.

Therefore, a calibration of the instruments was not performed.

The authors felt that the next best method 'of eliminating or reducing the effect of the differences among the instruments was to normalize the data to that of one instrument. First, it was necessary to determine if the differences in the instruments could be observed from the data collected during the opera-tional phase of the study; therefore, a one-way AXOVA was performed on the data from location N 6-3. The analysi's indicated that there were statistically significant differences among the instruments. The means of the rate measure-meats made at 'N 6-3 by each instrument are presented in Table 2 for the data from both t8%8RHcground phase and the operational phase of the study. The means from the> e'rational phase are larger than those from the background phase, but the', relationships among the five instruments are similar for both 1

sets of data. The numbers appearing under the heading "Normalization Factor"

0 0

Table 2.~RATE HEASUREKNTS AT N 6-3 .FOR EACH INSTRlPiKNT Normalization Ba ck round P na s e 0 erational Phase Nean, Hean, Normaliza tion Instrument No. pR/h ., Factor No. pR/h Factor T-3512 26 9.95 1.00000 36 10.33 1.00000 T-3514 18 9.88 1.00709 34 10.15 1. 01764 T-3513 21 9.81 1.01427 15 10. 04 1. 02874

.T-3517 14 9.58 1.03862 73 9.95 1.03?60 T-3516 18 9 ~57 1.03971 50 9.99 1.03357

ment No. T-3~5 to the means for the other instruments. The values of these

~~p~z ratios do not differ greatly between the two sets of data. It was concluded goal that (1) the~ata provide reasonably consistent estimates of the differences among the instruments and (2) the normalization of the data to that of one instrument is a reasonable approach to reducing the effect of the variation among the instruments.

Each set of data.was normalized to instrument No. T-3512 using the normaliza-tion factors in Table 2. Instrument No. T-3512 was chosen as the basis of the normalization because it appeared to yield consistently higher measure-ments than the other four instruments; therefore, all measurements values that were changed were increased.

e A summary of the normalized data from:both the background phase and the operational phase is presented in Table 3. By comparing the values for the standard deviations in Table 3 to those in Table 1 and in Table 1 of Report 1, it can be seen that in many cases the normalization of the data caused the standard deviation to decrease. Also, in many cases the standard deviation

'I increased. In cases where large values for the standard deviation were observed,'ay greater than 1 pR/h, the noMalization in all but one case caused the standard deviation to increase. Furthermore, the magnitude of the increase appears to be correlated with the magnitude of the standard deviation. This phenomenon. would'e an indication that (1) the normalization factors are not accurate for measurements appreciably greater than 10 IIR/h, (2) .application of v>> '."g" '

the normalization factors skews the distribution of the data, or (3) the

~

~ ~

Table 3. SUlMARY OF NOR'1ALEZED RATE MEASUREKNTS Back round UR/h ~

0 erational pR/h Location No. Mean Std. Dev. No. Mean Std. Dev.

N l-l 10 6.73 0.18 18.06< 7.98 N 2-1 2-2 ll 10

7. 62 8.03 0.47 0.27 8

16 10.71~

i+ 1.62 N 0.48 ll 6 60

'l.

~

N 2-3 7.44 0. 19 20 13.76 2.16 N 3-1 11 8.43 0.30 2 9.09 0.00 N

N 4-2 5-1 12 10 9.87 10.27 0.71 0.25 ll 18 10.

9.41~ 0.29 0.24 N 5-2 9 9.83 0.'43 1 18'0.

19 N 5-3 15 f. 61 0.44 13 9.29. 0.18 N 6-1 25 8.99 0.26 2 10. 0.10 N 6-2 N 6-3 ll 8.08 0.40 0.24 13 12'.

34 0.26

.97 9.95 208 10.33'. 0.27 N 6-4 14 10.29 0.27 -20 97 0.66 NNE 1-1 6.67 0.16 1'iNE 2-1 NNE 2-3 10 10 6.75 7.72 0.25 0.15 16 8 15.78'.68 6.17 1.03 10 16 0.67 NNg 2-4

.NNE 3-3 ll9. 8.05 9.95 0.21 0.21 0

2 9.58'0.54~.

0.21 NNE 3-4 10 8.66 0.32 0 NNE 4-1 11 9.01 0.28 0 NNE 5-1 '13 7.87 0.32 25 8.63~ 0.68 NNE 6-1 15 9.01 0..27 26 0.33 NE l-l 10 6.88 0.09 8 9.27'7.49'.04 7.07 NE 2-1 ll, 7. 52 0.32 2 ~

1.20, NE 2-2 12 7.88 0.16 10 13. 70 . 2.92 NE 3-1 10 8.96 0.64 13 9.32~ 0. 47.

NE 4-1 10 7.44 0.29 9 0.47 8.43'0.32 NE 4-2 10 9.34 0.22 9 0.20 NE 5-1 9 9.46 0.28 7 10.40 0.19 NE 6-1 10 7.98 0.15 1. 10.16 ENE 1-1 9 6;93 0.29 6 12.65 i 4.08 ENE 2-2 ~

9 6.78 0.16 14 0.97 ENE 2-3'NE 10 8.36 0.11 , 9 8.63'1.55'.97 3.16 3-2 10 8.70 0.23 4 0.58 t

a% Std. Dev. Seap+egg.J)eviation gJj

b. One rate measurem$ $t>,from the background phase. was omitted because it 'was made' using instrument >4Q~T-3590 for which there were insufficient data to normalize measurement valuing,to those of instrument No. T-3512.

Table 3. SlkkfARY OF NOR'M.IZED RATE HEASUREHEiNTS (CONTINUED)

Back round UR/h 0 eratlonal uR/h Loca t ion No. Hean Std. Dev. No. Mean Std. Dev.

KNE 4-1 14 9. 70 0.16 31 10.15~ 0.36 EiNE 5-1 13 9. 57 0.29 37 10.09 0.36 KNE 5-2 9 10. 17 0.26 .32 10.59 0.62 ENE 6-2 b

11 8.44 0.48 21 8.13 ~ 0.37 E 2-1 9 7.24 0.12 26 9.85~ 1.84 E 2-2 11 8.32 0.13 8 11. 51~ 3.61 E 3-1 10 7.00 0.21 10 0.39 E 3-2 10 8.55 0.23 8 7.84'.32'.78 0.66 E 4-1 b

17 f. 54 0;24 30 0.44 E 5-1 10. 39 0.19 10.69' 0.30 ESE 1-1 '014

8. 07 0.42 24 0

ESE 2-1 ESE 3-1 ll ll 6.87 7.70 0.36 0.26 7

2 9.43 2.12 0.28 7.66'.86 ESE 4-1 13 10.03 0.17 26'5 0;34 SE 1-1 16 9.75 0.98 17.03 ~ 14.98 SE 2-1 10 7:46 0.31 37 14.07'0.41 3.47 b

SE 3-1 10 9. 77 0.22 20 0.23 SSE 1-1 16 8. 69 0.35 28 11.48~ 6.22 SSE 2-1 12 6. 21 0.20 36 8. 50~ 1.24 S 1-1 16 10.33 0.95 27 12.00 3.44

'S 1-2 16 8.00 1.30 21 7.59~ 1.79 SSW 1-1 16 9. 93 1.70 27 11.48 3.58 SSW 1-2 17 9. 03 0.94 19 9.28j 1.23 SW l-l 16 10. 82 2.60 27 ll. 47; 3.41 SW 1-2 18 9. 69 1.28 22 9. 1.38 MSW l-l 16 1'l. 71 '.71 25 86'2.36 3.71 hiSW 1-2 17 9. 55 1.11 21 9.21'3.60 0.93 W 1-1 16 13. 40 3.00 25 3.44 W 2-1 17 8.96 0.70 20 9.03 0.74 h>h'-l 18 24.63 9.73 30 24.05 10.64 WNW 2-1 14 7.12 0. 48. 17 7.95 0.39 hKW 2-2 15 9.43 0. 50 36 9.83 0.99 Kti 1-1 2-1 ll 10 12.48 8.36 2.09 0.15 17 10

. 22.56'0.

6.10 0.70 Xi'M 3-1 .

NW 3-.2 Qi 4-1 12 10.

1T

'.59

-M~'" 9. 51 9.62 0.46 0.21 0.20

'8

'0 1

10.11 53'0.21, 0.64 NNW l-l 10, '. - 8.22 0.50 30 19. 05 6+ 14";:""'.t'-"'

NQ4 NQ4 2-2 3-2 ll ~. '.38

.11 ~+~

"';;-'.8.40 0.29 0.33 9

8 9.06

'.45 1 11'"

0.38 NNW 3-3 10, 8. 54 0.18 8 7.78~ 0.24

~ Table 3. StDDtARY OF NORttALIZED RATE NEASUREttENTS (CONTINUED)

'~ ~

Back round pR/h 0 eratioaal ttR/h Location No. Mean Std. Dev. No. kfean Std. Dev.

NNW 4-1 10 8. 79 0.34 10 10. 16 0.59 NNW 4-2 10 10. 24 0.26 6 9.33 0.57 NNW 4-3 10 8.83 0.34 9 8.26 0.21 NNW 5-1 15 10.42 0.35 26 10. 57+ 0.19 NNW 5-2 10 9.79 0.20 33 10.06 > 0.27 NNW 5-3 10 10.06 0.19 22 9.88'.13 0.34 NNW 5-4 14 9.88 0.73 32 ~ 0.47 NNW 6-1 14 10.74 0;27 25 10. 75 0.26

systematic ~or due to the variation among the instruments was obscuring some of the vai'Ration that should have been observed in the data. In spite of the uncerga'inties associated with the use of the normalization factors the authors believe that it is more desirable to apply them than to make no atte-pt to reduce the variation among the instruments.

REFERENCES

1. Jenkins, P; H. and R. L. Doty, "Direct Radiation Levels Around Browns 44.

Ferry Nuclear Plant Background Data," I-RH-77-'1, Tennessee Valley Authority, Huscle Shoals, Alabama, February 1977.

~ >> ~

~

~ ~ ~

/

5*>>

lo

.r

~P

~ ~

~>> ~ I ~

~

4>>) g>>2 I

~ 4 g>>I 22 ~ lit~

fo>>l ~

/ fh I . 2.iW.

I eg" P' ~

~ <

pg Q gimp Figure l. Measurement locations - Brows Perry Nuclear Plant 0 )wi'>. I~ C.

O Q~o~p C

COMPARISONS BETWEEN BACKGROIPiiD A'ND OPERATIO'iZL DATA

~a It is expected tLat operations at BFNP'may increase the 4 xposure rate p+Ql at a given location due to direct radiation from radioactive material confined in the plant (e.g., N-l.6 "skyshine") and due to radioactive materi.als in the gaseous effluent. Also, it is expected that such increases in exposure rate would vary, depending upon parameters such as, power levels of the reactors, wind direction, and dispersion, of the gaseous effluent plume. Thus, increases in both the mean and the standard deviation of exposure rate measurements at a given location may be an indication of an effect due to plant operations.

I.

~ . Coimparisons were made between the background data and the 'operational data (presented in Table 3) for each of the measurement locations.:

An F-test (2) was used to determine whether there was a significant difference between the variances of the two data sets. -

The results of the F-tests are'resented in Table 4. Unless otherwise indicated

~

. in Table 0, the F-statistic was .calculated by the follhg3jng equation:

2 .2 F s/s o

s. = standard b'he're 0

deviation'of the operational data.

.,'.~sb r standard devia'tion of the background data.

A t-test (2) wasWhen used to determine whether there was a signi.ficant

'ifference between.the means of the two data sets. The result of the F-test determined whether the variances were assumed to be equal in Mhich case the t-statistic was calculated, or the variances were

a

, Table 4. RESULTS OF TESTS ON MEANS AibD VARIAXCES BETMEEN BACKGROUND AND.OPERATIONAL DATA AT EACH LOCATION A. Locations at which signi ficant increases in both mean and variance were observed.

Back round uR/h Operational uR/h Location Mean Std. Dev. a Mean S td. Dev.

b

,N 1-1 73 0. 18 18. 06 7.98 4. Olbb 1965. 44b N 2-1' 7. 62 0. 47 10. 71 l. 62 7. 20b 11.88b 2"3 7 ~ 44 0.19 13.76 2. 16 12.99b 129.24b NNE 1-1 6. 67 0.16 15.78 6. 17 4.18b 1487.07b NNE 2-1 6.75 0.25 8. 68 1. 03 7.17b 16.97b NNE 2-3 7.72 0.15 9. 58 0.67 10 68b 19.95b NNE 5"1 7. 87 0. 32 8. 63 0.68 4.68b 4.52b NE 1-1 6. 88 0.09 17.49- 7.07 4.24b 6170.98b NE 2"2 7. 88 0.16 13.70 2 '92 ~

6.30b 333.06b ENE 1-1 6.93 0.29 12.65 4. 08 3.43b 197.94b ENE 2-2 6.78 0. 16 8.63 0.97 6.99 36.75b g 2-2'.

ENE 2-3 5-2 8.36

.'9.

10. 17 70 0.11

.0. 16 0.26 11.55 10.15 10.59

3. 16 0.36 0;62 3 '3b 825.26b

5. 8lb 3 'lb

5. 06c

5. 69b E 2-1' 7 24 0. 12 9.85 1 ~ 84 7.19 235.lib E 8. 32 0 ~ 13 11.51 3 ~ 61 2.50 771. 13 E 4-1 7.54 0. 24 8. 78 p. 44 12 '0 3.36

. ZSE 2-1 6. 87 0.36 9. 43 2. 12 3 ~ 17 34. 68b SE 1-1 9. 75 0.98 17. 03 14. 98 2. 42b 233. 65b SE 2-1 7.46 0.31 14.07 3.47 ll. 42 125.30b SSE 1-1

.,'SE 2-1

8. 69 0.35

0. 20

.11 '8

'8.5Q

6. 22 l.

2 ~ 37b

10. 67 315.82b 38.44b

6. 21 24 1-1 10. 33 0. 95 12.00 3.44 2 37b 13'lib mW 1-,1 12.48 2. 09 22 ~ '56 6 10 6. 27b 8 52b

) Ww 2-1 8. 36 0. 15 10. 53 0. 70 9. 59b 21. 78b

%. 3-2 8. 59 0.'21 10.11 0. 64 6. 45b'.57 9e29b NNN 1-1 8. 22 0. 50 19.05 6 ~ 14 150. 80 fg

• a . Std. Dev. S tand'awd Deviation

b. Significant at t+~~ = 0.01 level.

cd Significant a t t.ba.-- ~ 0. 05 leve 1 ~

Tabl.e 4 (continued)

B. Locations a~which significant increases in mean, but not in variance,~ere observed.

Back ro8nd uR/h 0 erational uR/h Location Nean Std. Dev. Mean Std. Dev. t t N 2-2 8. 03 0. 27 11. 60 0. 48 19. 24 ~

3. 16 N 3-1 8. 43 0.30 9.09 0.00 3 ~ Oob 7. 30 N 6-1 8. 99 0.26 10.12 0.10 6 ~ 02b (6. 76)

N 6-2 8. 08 0. 40 9. 34 0. 26 9 ~ 29b (2 ~ 37)

N 6-3 9. 95 0.24 10.33'.27 11.85b 1. 27 NNE 3-3 9.95 0.21 10.54 0.21 3.59 1.00 NHE 6-1 9. 01 0.27 9.27 0.33 2 59b 1. 49 NE 4-1 7. 44 0.29 8.43 0.47 5. 59b 2. 63 NE 4-2 9.34 0.22 10.32 0.20 10.12b (1. 21) i~~= NE 5-1 9.46 0.28 10.40 0.19 7. 60 (2.17)

.ENE 5-1 9.57 0.29 10.09 0.36 4 69b 1.54

'3

~

J SE 3-1 9. 77 0.22 10.41 0 7. 28b '.09 WNM 2-1 7.12 0.48 7.95 0.39 5. 32b (1.51)

NM 4-1 8.79 0.34'0.16 0.59 6. 36b 3.01 9.79 0 '0 . 10.06 0.27 2. 92 1 82 C Locations a't which no significant change in the mean was observed, and at'which effects from the plant were observed during the "background" d'ata coL3ection.

Back round uR/h 0 erational uR/h I ocation Std. Dev. Mean S td. Dev.

S 1-2 -0.77 8.00 1.30 7.59 1. 79 1 ~ 90b SSM .1-1 9.93 1.70 11.48 3.58 1 ~ 91 4. 43

. SSM 1-2 9.03 '.. 0.94 9.28 1.23 0. 68 1. 71 SM 1-1 3.41 1,72 SM 1-2 1-1'fean

10. 82 9 69 .'

60 28 ..

11 ~ 47

'9.86 1. 38

0. 66 0.40 '1 ~ 16

~ ~

'1

~

WSM 1-1 11;71 71 12. 36 3 0.60 1. 87

~ WSM 1;2 9 55~ 1.11 9.21 0.93 <<1; 03 (1. 42)

W 1-1' 13. 40 "'sip@+. 00 13. 60 3.44 0. 19 1 ~ 31 W

WNM 2-1 8.96'

24. 63 .. ~ 70 --., 9.03

~ 73 24. 05 0.74.

10.64 .

0. 29

-0. 19

1. 12

'1 ~ 20.

d. Parentheses indicate that the std. dev. of the operational data, s was egal)er than the 'std. dev. of the background data, sb, and thaP s /s

< ~1 ~

.r ~

P ~ g 7.00 yy; 0.21 7. 84 0.39 6. 00b 3. 45+

10.39 ~ .0. 19 10. 69 0.30 3. 36 2.49

Table 4 (continued)

D. Locations, other than those'n group C, at which no significant change in the jaean was observed.

Back round uR/h'perational uR/h Location Mean Std. Dev. Mean Std. Dev.

N 4-2 9. 87 0. 71 9.41 0. 29 -2. 06 (5.99)

N 5-1 10. 27 0.25 10. 18 0. 24 -0. 94 (1. 09(

N 6-4 10. 29 0.27 9.97 0. 66 -0.43 5. 98b NE 2-1 .7. 52 0.32 8.04 1. 20 0.61 14.06 NE 3-1 8. 96 0.64 9.32 0.47 1. 56 (1.85$

ENE 3-2 ENE 6-2

8. 70

8. 44 0.23 0.48

8. 97

8. 13 0 '8

0. 37. -2. 03 0 ~ 90 .6.36 (1. 68$

E 3-2 8. 55 0.23 8. 32 0. 66 -0. 94 8. 23 ESE 3-1 7 F 70 0.26 7. 66 0. 28 -0.20 .1 ~ 16 ESE 4-1 10.03 0. 17. 9. 86 0.34 <<2.08 4e00b WKV 2-2 NNW 2-2 9.43 8.40 0 50 0 ~ 29

9. 83

9. 06

0. 99

1. 11

.1 '1 1.74 3 92b 14.65 b NW 5-1 10.42 10.57 0. 19 1.53 (3.39~

NNW 5-3 MW 6-1 10.06

10. 74 0.35'.

0. 27 19 9. 88 10.75

0. 34

0. 26 0 11

-1.56 3 '0 (1.08)

E Locations at which a significant decrease in the mean was ob'served.

Ir Sack round uR/h 0 erational uR/h Location Mean Std; Dev. Mean Std. Dev. t N 5-3 9. 61 0.44 9. 29 0. 18 -2. 58 (5. 98)

NN!0 . 3-2 0.33 b

9. 38 8. 45 0.38 -5.69b 1.33 NNW 3-3 8. 54 0.18 7. 78 0.24 -7. 69 1. 78.

NNW 4-2 10. 24 0. 26 9.33 0.57 -3. 69 4. 81 NKJ 4-3 8. 83 0.34 8. 26 0.21 -4. 33 (2. 62)

M 5-4 9. 88 0. 73 9 ~ 13 0.47 -3. 54 b (2. 41)

Table 4 (coniinued)

F. 'Locations at 'which the quantity of "operational" data collected was not suffiwient for the tests.

Back r ound uR/h Operational Me/h Location Mean Std. Dev. Mean Std. Dev.

N 5-2 9. 83 0.43 10. 19 NNE 2-4 8.05 " 0.21 NNE 3-4 8. 66 0.32 NNE 4<<1 9. 01 0. 28 NE 6-1 7. 98 0. 15 10. 16 ESE 1-1 8. 07 0. 42 NW 3-1. 9. 62 0.46 10. 21 mW 4-1 9 ~ 51 . 0.20

0 not assuned to be equal in which case the t'-statistic was calculated.

The results .of the t-tests are presented in Table 4.

Caution must be used in interpreting the results of the F- and t-tests~

because there ~pe several reasons why these tests may not be appropriate in all cases. These tests are only valid for data which are normally distributed, and the data analyzed here may not be normally distributed.

Due to the. uncontrolable nature of the experiment, effects observed may have been caused by sonmPing other than operations at BFNP. Also, the m~gurements were not made in a truly random manner, and anamolies may result due to confounding with other parameters; for example, at a location that is close to the plant and is affected by skyshine 7 the data may not indicate an effect because the times at which 0 measurements were made

~ ~

at that 'location happened to coincide with periods when the plant was not producing power. (This type of problem n,

is most likely to occur where a small number of measurements were made.) Et is felt, however, that as long 'as these problems are recognized, the t- and F-te s ts pr ovide a good me/Bod of systematically comparing the background and operational data.

Based on the results of the t- and F-tests, the 83 measurement

~

'ocations were arranged into six groups as presented in Table'.

Each of. ~

these~ups is discussed below.

~

4f':

I Grou A This group contain]s locations at which both the meanand the standard deviations of the operational data were significantly larger than the means and standard deviations of the background data. In'ost~6'ases, these locations are within 610 m (2000 ft) of the plant, and are like'ly to be affected by skyshine from the (Correlation of exposure rate measurements with power levels 'lant.

of the reactors is discussed in a later section of this report.) Five locations, NNE 5-1, ENE 4-1, ENE 5-2, E 4-1, and NW 3-2, are beyond 610.m (2000 ft) from the plant and are likely to be affected to a extent by skyshine. Therefore, if plant operation's caused the 'esser observed increases in both the means and standard devia~ons at these

~ locations, the effect may have been due to radioactive materials in the gaseous e ffluent plume.

~Grou B Thi.s.'group contains locations at which the means of the operational data were significantly larger than those of the background data, but no significant changes in the standard deviations were

.~ +s of these locations, N'-2,

.observed. Seven

• N 6-.3, N 3-1, NNE 6-1, NE 4-1, NE 4-2, and NE 5-1, are in close proximity to paved, roads.

A possible ex'planation for the effects observed at these locations is J

that, the roads may have been surfaced ( or resurfaced) with a material

~

r-containing a sl g tly elevated concentration of naturally occurr'ing

~

radioactive mate%'ials,

~

thus causing small, but nonfluctuating, increases

'n exposure ~~%

ra'Ee. This may also be true for location WÃl 2-1, which is close to an area which was sur faced.'t each 'of three locations,

N 6-1, N 3-1, and NNE 3-'3, only two "operational". measurements were made; therefore, there is little confidence in the compariaons/ made between the background and operational data for these locations.

(Note that for~ 3-1, the F-statistic was not calculated because the standard deviation of the operational data was suspiciously small

( 00000015 uR/h); therefore,-.both the t- and t'-statistics were reported in Table w.) En spite of the fact that no significant changes in the standard deviations at the locations in this group were observed, operations at BFNP cannot be absolutely eliminated as the cause of the observed increases in the means of the exposure rate measurements.-

~Grou C This group consists of .ten locations at which no changes in the means of the 'exposure'ate measurements were observed, but also at which effects from the plant were observed during the "background"-

da'ta collection. ln Report 1 it was stated that exposure rates at e

locations .in the S, SSW, SW, WSW, and V sectors, plus WK< 1-1. and NW 1-1, were affected by radwaste that was stored in the west corner of the plant. The corn P arisons between the ."bac kg r ound and operational l.t 4

data indicate at all of these locations, except S 1.-1 and NW 1-1, n

that operation of the reac. tors does not cause a significant change in the mean ex ure rate. (S 1-1 .and NW 1-1 were included .in Group A.-) '=~A c

In other words,,'gge exposure rates at these locations are being affected by, plagt operations in 'approximately the same manner as they were before the plant resumed operations.

~ I G~rou D, This group consists of locations at which no changes in the means of the exposure rate measurements were observed, and at which effects from th~ plant were not observed during the "background"

.~~r

~ ~

data collectiony" Three of these locations,'E 2-1, WM 2-2, and NUM 2-2, are extremely close to the plant (within 610 m (2000 ft))

and are close to other locations at which affects from the plant have been observed. Also, significant increases in the standard deviations were observed at these three locations. Based on this information.z it is felt that further measurements at these locations may indicate that they should be included in*Group A.

At four other locations in this group, N 6-4, EttE 3-2, E 3-2, and t ~

ESE 4-1, sign'ificant increases in the standard deviations were observed but'at three of these locations the mean was observed to decrease SH.ghtly, the exception being ENE 3-2. Z Also, at two locations,

.N 4-2 and NVA 5-1, significant decreases in the standard deviations were obseived. Such decreases in the standard deviations are difficult to interpret.

G~rou E This group consists of six locations at which significant, t

decreases in the means of the exposure rate measurements were observed.

P 4 I Five of these locations, N&l 3-2, NhU 3-3, hhW 4-2, NhW 4-3, and HNM 5-4, a'e i'~e cooling tower area. Because these five locations are grouped in t~.'same general area,

.SAC' it is felt that the observed ="=

~ decreases in the ate'ans are not due to coincidence or a'notaalies in the data. A possible explanation is that roads in this area may have been

I

'urfaced (or resurfaced) with gravel which may contain extremely small concentrations of naturally occurring radionuclides, and may even act as a~hield against radiation eiu.tted from natural radio-giga<

active materials in the soil. This is based on (a) measurements made+'uring the "background" phase which indicated that gravel in parking lots near the plant exhibited this shielding effect (see Report l,,

p 35), and (b) these five locations are near gravel roads.

The sixth location in this group, N 5-3, is also near a gravel road; thrgfore, the same explanation may apply to the significant decxease in the mean observed tQQe. However, two other locations are near that same road, but one, N 4-2, was included in Group D and the other, Xiii 5-1, was included in Group A. Therefore, it is difficult to interpret- the deere'ase in the mean observed at N 5-3.

~

At two of the six locations in this group, N 5-3 and NNW 5-4,

'i.gnificant decreases in the standard deviations were observed.

.Again, such decreases in the standard deviations are difficult to interpret. Also, at NNW 4-2 a significant increase in the standard

~

deviation was'observed, in spite of the fact that a significant de'crease in the mean was,observed. This is also difficult to t

interpret<

I P ~

eight locations at which. either a

Garou E Thi%tgeoup consists of one dl'.no ".operatio " .measurement was made. Therefore, it was not C

possible .to perfor m t- and F-tests for these data.

Reference 2 is Ostle, Bernard~>Statistics in Kesearch The iowa State Univer si ty J

Press, Ames, Eoya, 1963, pp L19-124.

PLUME DETECTION METHODS Xn Report egg, two methods were suggested for detecting the effect of a gaseous effluent plume. The finest meth+ was based on a change in the distygbution of the fifty exposure rate readings which w

were averaged to obtain a single rate measurement. Such a change in this distribution may be observed as an increase in the standard deviation of the fifty readings and by a change in the character-istics of a log-normal plot of the-fifty values. The standard deviation of the fifty readings is an estimate of subsampling error, and should not be confused with standard deviations discussed in previous sections of this report.

The subsampling standard deviations associated with the rate measurements made during. the background phase ranged from 0.24 to 0.77 pR/h. Out of 1507 rate measurements made during the operati.onal pose, the subsampling standard deviation was substan-

'tially larger than 0.77 pR/h in five instances. These five standard deviation/ values were 1.07, le64, 2.10, 3.02, and 11.64

}iR(h.

X~

the set of readings whose standard deviation was 3.02 uRgh, y'~ll4L5 the exposure rate read4ags fluctuated both above and below the normal background level. A log-normal plot of these data was essentially a straight line, but the slope of the line was lar'ger.

than that of typical background log-nor'mal plots. It was felt that

. these data could be valid only if the RS-ill system responded to J

abrupt changes in exposure, rate by overshooting the actual change or fluctuat.ion 'and then settling in on the'correct exposure rate g%%%a values.. A bnef experiment was conducted to determine whether

~

I or not the RS~11 does respond in thgfis manner; Xt was found that 4'."

the RS-111 napidly responds to an abrupt change in exposure rate,

~

~

without overshooting the actual change. Further, data from the

BFNP meteorological tower indicated that it is highly unlikely that a plume effect could have been observed at the particular time and lo~tion at which these data were 'collected. It was concluded, therefore, that the observed fluctuation in exposure j~l rate was du&" to an instrument malfunction or some type of interference, and these data were deleted from the data set.

N the set of readings who/ac standard deviation was 11.64,uR/h, the data indicated an abrupt increase from tgical values to Sh~hbbdhh approxim'ately 65 pR/h, followed by a return to typical values.

one of

$ f Qr.w.-r the ejnvironmental monitoring stations which transmits data to BFNP. These data were collected at ENE 5-1, wh'ich is close to one of the environmental monitoring stations. Further, metevrological data indi.cated that it is highly unlikely. that a

'plume effect could gave been observed at this location at the time these data were collected. Therefore, it was concluded that the observed fluctuation in exposure rate was due to radio interference from the environmental monitor, and these data were deleted from the data set.

The two sets of readings whose standard deviations were 1.07 and 1.64 pR/h .were collected at N 6-3. This location is approximately f

9,700 meters (6.0 miles) nofth of the plant, and was originally believed to be far enough from the p'lant to serve as a "control" location. A log-normal plot of the first set of readings (standard deviation = 1.07 pR!h) indicated that only one of the fifty

readings deviated significantly from the straight line associated with a. typi~+background distribution. A log-normal plot of the second set indicated that three of the fifty readings deviated .

si.gnificantl~trom the straight line. Sources of interference I

~ which could; have caused such a response can be postulated; e.g.,'

t

.a radio signal from a passing vehicle on the nearby road. However, the meteorological data indicate the possibility of detecting a plume effect .at this location at the times the data were collected. ~

From the information available it is. impossible to make' firm conclusiont4egarding the source of the observed fluctuations. The data were deleted, from the data set. However, had these data not been deleted, there would have been no changes in the mean and standard deviation for N 6-3 reported in Table 4, and no changes in the results of the t- and F-tests.

The set of readings whose standard deviation was 2.10 pR/h was collected at N 4-2. A log-normal plot of these data indicated that twenty out WE of the fifty readings deviated significantly from the straight line associated with a typical set of data.

From the log-normal plot,,it appears that a second log-normal I t distribution was superimposed over the "background" log-normal distribution. This is what would be expected influencing the exposure rate.

if XRP a plume was

. These data were deleted from the data set, because the observed deviations were initially thought to be due to instrument malfunction or some type of interference. In retrospect, these data should perhaps not ha've been deleted. This location is approximately* 1,070 meters .

t (3500 feet) from the plant, so a plume would be dispersed to a much lesser extent than in the previous case where the lo'cation was approximate'ly 9,700 meters (6.0 miles) from the plant,. Z Also, meteorological data indicated that the prevailing wind, at the time the measurement. was made, was undergoing a shift from blowing

.'..toward th&HNW to blowing toward the NE. The average of the fifty tI readings was .1 .19 pR/h. Had these data'not b'een deleted, the* -

%gg mean and'tag,ard deviation reported in Table 4 far N 4-2 would '- '

~

'e changed d6. rr9o.a41 pR/h and 0.29 pR/h to 9.56 pR/h and 0.58 pR/ht I

respectively. The F-statistic would be changed to 1.50; which

'would no longer indicate a significant difference between the

background and operational standard deviations. The t-statistic would be -1;17, which would indicate no significant difference between the ba~round and operational means; therefore, N 4-2 would still be inc'luded in Group O.

.7 (Note'~M~;- I recommend that the above changes be made in Table 4, and that this last paragraph be rewritten to indicate that the data were not deleted from the data set. A similar change would be required in Table 3 under "o erational" N 4-2 12 9.58 0.58 and in Table 1 N 4-2 12 9.35 0.59 Also, in the 2nd paragraph of EXPERIMENTAL POSE, the "1502 rate measurements" would 'change to "1503 ")

I Except for the five cases discussed above, subsampling standard deviations were observed outside the range of 0.24 to 0.77 pR/h, which was observed during the background phase, in nine instances.

Five of the nine 'were below 0.24 uR/h; specifically, 0.23, 0.23, 0.22, 0.21, and 0.21 ugjh. Four of the nine were above 0.77 pR/h; 0.81, 0.83, 0.89, and 0.93 uR/h. The remaining 1493 subsampling r

standard deviation values were withig the range observed during the background phase.

The second plume detection method is based on a 3 x 3 Latin-square experimental design which is explained in Report l. Integral measurements we~made at two groups of three locations according to this design. The duration of each of the integral measurements i~

was approximate1~ne,hour. The data'are presented in Table 5.

~ " S Ew es C

Unfortunately, meteorological data indicated that the wind was not I

Table 5. Data collected for Latin-square experimental design Garou1

~

~ ~

' Inte ral measurement 'uR/h Observation ANNE 5-1 NNN 5-4 ENE 5-1 8.41 9.04 9.91 8.55 9.30 10. 19

8. 51 9.92 9.83 8.42 9. 41 9. 92

8. 38 10. 18 9.72

'8. 50 9. 03 10. 18

8. 56 8. 86 10. 11

~Grou 2 Observation 4-1 4-1 4-1

~ ESE

9. 56 9.90 E

8. 12 1

9. 75 'NE 10.18 8. 25 3 9. 68 9. 92 8. 17

9. 75 9.95 8. 35

'.55 10.35 9.31 6 9.71 10. 46 9.06 Meteorological data indicated that the wind was blowing toward this location at the time the measurement was made.

~ s

~ blowing toward any of the measurements were being locations in the second group at the times made at, these locations. For the first group; the data. ndicated that the wind was blowing toward NNE 5-1.

w'9 one occasion (observation 4) arid toward NNW 5-4 on three occasions

'n (observations l~g" 5, and 6). Unfortunately, the data indicated that the wind was not blowing toward ENE 5-1 during any of the seven observations mad'e.

A statistical analysis was performed on nine combinations of data from the first group, which at least partially met the criteria of the design. The results of the analyses are presented in Table 6.

In the first column, the specific combinations of data are indicated.

For example, the combination designated as "4,5,2" indicates'hat the 0,

data were as follows:

Observation NNE 5-1. NNW 5-4 ENE 5-1 4 8.42 9.41 9.92 8.38 10.18 9e72

8. 55 9.30 10. 19 The nine combinations were arranged so that measurements which were taken while the wind was, blowing toward the location appear on the main diagonal of the 3 x 3 array of data. The tenth combination listed in Table '6 represents an arrangement of data which places the largest measurement value observed at each location on the main diagonal.

'I This combiriation should produce the largest calculated plume effect

~ ~

that is possible with these data, although

~ ~

it has no physical meaning.

A The second column in Table 6 contains values of P, the estimated Cgg average plume effect, and the third column'ontains values of the A

F-statistic associated with the values of P. In all cases, the a estimated plume effect was less than 0.5 pR/h, and in five cases

~

Observations ~ -h

~P+E/h 4,5,2 0. 383 3. 58 4,5,3 0.167 0.76 4,5,7 0.428 2.17 4,6,2 -0.097 0. 61 4,6,3 -.0. 313 2.80 4,6,7 -0. 052 0.08 4,1,2 -0.033 0.12 4,1,3 -0.250 2. 67 4,1,7 0.012 0. 005 0.490. 3.05

! ~

the estimated plume effect was negative. A negative plume effect

~ has no physical meaning. In no case was the F-statistic significant.

Values of 10.1 and 34.1 for this statistic would be significant at the 0.05 andj'0.01 levels, respectively.> The results of .these analyses are similar to results of analyses of da'ta collected during the background ase, which were presented in Report 1.

CORRELATION OF EXPOSURE RATES AND POWER LEVELS Operati.onal..dna were collected from BFNP for the purpose of evaluating rel'ationships among releases in gaseous effluents, power levels~~eteorological data, and exposure rate data. The operational data consisted of effluent monitor readings and pwer levels of the three units recorded on an hourly basis on log Ll*V sheets. Several person-days'o effort +are expended retrieving the log sheets from storage at BFNP and hand-copying the data corresponding to the periods when exposure rate meas'urements were made. The data were then coded onto comput'r cards. 'he process of verifying the data on the computer cards had just begun when funding far this project was terminated.

Effluent monitoring systems at a nuclear power plant, such as BFNP, t are designed to de'monstrate compliance with regulations, but not, to provide data for a research project such as this one. Therefore, a large percentage of the effluent data was recorded as "less than" a particular lower detection limit. Obtaining meaningful informa-ti'on from these data would have required. considerable addit~onal effort. Because funding for the project'ad been terminated, verification and analysis of the effluent data was abandoned.

However, verification and analysis of the pwer level data was A

continued.

At several of the measurement locations, particularly those close to the plant, one'ould expect to observe an increase in the exposure rate due to skyshine from N-'16 in the turbine building. Further, xe the effect d~p skyshine should be directly related to the power 1 M4w w levels of the victors. In an effort to determine the extent to

~

-~

~

which skyshine".::affected the exposure rate measurements at each t

~ location, the power"level data and exposure rate data were analyzed

~

using two linear regression medels. The first model was

p+ px +z (Model X) wher~Y= observed exposure rate (tdl/h) r

&X total thermal power. of the three units'Qlt)

regression coefficients f = 'error term and the second model was Y

(~ 4 I$ > Xl + (~X2 + (zX3 +F (Model II) where, Y = observed exposure rate (pR/h)

Xl,X2,X3 = thermal pwer level of units 1, 2, and 3, respectively (GNt)

,pl,p2,(,'3 = regression coefficients E= error term The Statistical Analysis System (SAS) was used to merge the exposure-

~regression rate and power-level data sets by date and time, and to perform the analyses.

. Estimates of the regression coefficients and R values for both 'models for each location are presented in Table 7. For some locations, no information is given in .Table 7 becasQe an insufficient quantity of I f~s

~A for the ~ec4~ The values presented in Table 7 m ~

data was aviafable show a general trend of an effect due to skyshine at locations close to.the plant, and little or no effect at locations farther away from I plant. Thxs is, seen by'the large R 2 values at locations cloae to

~

'he the plant, and smaller R 2 values at locations farther away from the p

~

plant. Also, in general, the numerical values and statistical significance h

.of the estimate'8- of the regression coefficients (b 1 for Model X, and bl,b2, and b> .for Hadel XX) are greater at locations close'o the plant..;3 4

'Care must be take5~:-however, in making detailed interpretations of the esults reportedAn Table 7. Anomalous results should be expected when performing analyses such as these; Several indications of anomalous s

~ Table 7. REGRESSION PARAMETERS FOR PREDICTION EQUATIONS RELATING EXPOSURE RATES TO POWER LEVELS odel I Model II bo. 2 b, bl,

/h'GGt b2 b3 Location ~R/h /h GWt R P/5 uR/h'GGt /h GWt R N 1-1 4. 51 2. 188 0. 8916 4. 89 4. 033 1. 238 1. 141 0. 923

'N 2>>1 7.00 0.592 0.8706 6. 96 1. 163 0. 364 0. 294 0. 921 N 2-2 9 ~ 35 0. 258 0. 2679 9. 70 1. 422 -0.140 -0.571 0.952 N 2-3 7.55 0.808 0.6951 7. 07 1. 928 0.712 -0.009 0.8511 N 3-1 N 4-2 9.05 0.056 0.1891 9.42 -0.035 -0. 165 0.189 b 0.5561 N 5-1 10.25 -0.012 0.0251 10.18 -0.055 0.111 -0.074 0.273t

. N 5-2 N 5-3 9.41 -0. 022 0. 1255 9. 44 0. 009 -0.054 -0.027 0.165

~ N 6-1 N 6-2 9.46 -0.017 0. 0369 9. 42 -0. 137 -0.123 0. 181 0. 224 N 6-3 10.39 -0.011 0.0203 10.38 -0.012 0.018 '0.035 0.034 N 6-4 11. 26 -0.'183 0.2868 11.52 '0.235 -0.370 -0.054 0.343 NNE 1-1 5. 20 1". 705 0.9051 5.54 3. 008 1. 364 0.627

'68(

0.917'.183 NNE 2-1 6. 29 '.352 0.7313 5.99 0.424 0 ~ 578 0

..NNE 2-3 8. 11 0.245 0.8055 7 '4 0. 523 0. 207 0.094 0 '10 NNE 2-4 NNE 3-3 HNE 3 "4

. NNE 4-'1 NNE 5"1- 8. 6V: -0. 006 0. 0006 8.79 -0.079 -0. 197 0. 211 0. 235 NNE 6-1 9. 90 a

~-0. 089 . 0. 4110 9 '0 -0.102 0. 050 -0 ~ 166, 0. 57:

.NE 1-1 5. 27,..; .1. 971 0. 9212 5 '2 3 '93 1. 645 0.527 '0.9333 NE 2-1 JgF e.~ ~

NE 2-2 . 6.41 1.093 0.9145 6. 58 1. 571 l. 159 0 '49 0.935 NE 3-1 9. 12 0.026 0.0125 9.28 ~

0.274 .0.169 -0.387 0.802(

a. Significantly different from zero at the ~ = .01 level.

b. Significantly different from zero at the M = .05 level.

bo>

" ~l> J ')

.:(: 4/'<

~

C~/h -,

~ c

-,2 iW f~

NE 4-1 11.35 -0.342 0.3340 7.21 -0.216 -0.277 0. 802 0.

357'.081 NE 4-2 9. 88 0. 058 0.4133 9. 86 0. 032 0.070 0.416.

NE 5-1 10.72 -0.047 0. 6621 10.59 0.051 0.080 -0.203 0 924( ~

NE 6-1 ENE 1-1 7. 42 0.990 0.7400 8. 11 3. 562 0. 906 -1.617 0.805(

ENE 2-2 6. 75 0.330 0.7449 6. 60 0. 691 0. 133 0.184 0.850'.121 ENE 2-3 7.20 0.798 0.9083 6.98 -0.120 0.415 0.934(

ENE 3-2 7.89 - 0.162 0.7647 ENE 4-1 10. 10 0. 009 0. 0044 10. 08 0. 021 0. 038 -0 '26 0,021m ENE 5-1 9. 90 0.031 0.0558 9.89 0. 026 o.o44 0.025 0.057(

tNE 5-2 10. 90 -0.048 0.0399 10. 80 0. 014 0. 116 -0.215 0.1201 ENE 6-2 8. 07 .0. 009 0.0039 8. 03 0.074 0.014 -0. 043 0. 032 i E 2-1 6. 78 0.552 0.8330 6.81 0.772 0.439 0.434 0.846 E 2-2 6. 94 0. 896 0.9173 6.81 0.507 0.124 2. 009 0. 967 E 3-1 7 ~ 30 0. 089 0.5789 7.34 '0.160 0. 106 0.002 0.601 E 3-2 10.00 -0.226 0.5167 9.95 -0.341 -0.183 -0.137 0.580 E 4-1 8. 75 0. 005; 0. 0008 8.78 -0.182 0.120 0.070 0.086.

E 5-1 10.84 -0.022 0,0339 10.79 0.016 0.038 -0.100 0.1147 ESE 1-1 ESE 2-1 ~

7 '6 0.395 0.2169 7.23 -2.635 0. 151 3.076 0.5262 ESE 3-1 ESE 4-1 10.10 -0.040 0.0826 9.87 -0. 111 0.142 -0.077 0.2173 SE 1-1 9.47m~ -4.5O4 0.9177 9.84 -6. 931 0.099 19.180 0.9744 SE 2-1 5. 67 ~y.. '. 231 0.7794 6.45 0. 501 0.581 2.177 0.919j-SE 3-1. 10.31,~- 0.021 0.0860 10.25 -0.059 0.111 0.021 0+817 SE 1-1 8. 51' 1.'613 0. 6820 8. 66 -0. 766 -0. 360 5.362 0.8093

C I

~ < > e Table 7.

@pc T.

6'il. I t".3d'; 1 t. T g /i bl, 'p 2 b~,

", /;) i'.',,

b3 Loc a 1 0~i 1 SSE 2-1 5. 43 0. 481 0. 8109 5.51 . 0.413 0 ~ 414 0.570 0.818 S 1-1. 10. 21 1. 013 0. 8730 10. 21 0. 226 0. 192 2.299 . 0 '06 S 1-2 6. 67 0.514 0.82'06 6.77 0.031 0.276 1.232 0.908 SSW 1-1 9. 65 1. 033 0. 8379 9. 65 0. 396 0.418 2.040 0.857 SSW 1-2 8. 89 0.395 0. 6414'. 96 7.058 -12.202 7.662 0.801 SW 1-1 10. 02 0. 872 0. 6222 10. 14 -0. 187 0.402 2.318 0.654 SW 1-2 9. 24 0. 362 0. 6587 9. 29 -0. 230 0.787 0 ~ 554. 0. 697 WSW 1-1 10. 83 0. 918 0.6187 10. 88 -0. 056 0.112 2.489 0.640 MSW 1-2 8. 93 0. 268 0.4817 8. 97 0 ~ 668 -0.487 0.772 0 '45 M l-l 12;27 0. 797 0 '397 12. 32 -0. 085 -0.379 2.549 0. 582 T 2-.1. 8. 80 0. 245 0.6527 8.81 3.477 -5.339 3.109 0.707 WNW 1-1 21. 19 '.520 0.2260 21.17 2. 856 -1.391" 2. 600 0. 252.'.

. WNW .2"1 7.34 0.089 0.3476 7. 03 0.302 0. 139 -0.056 675 MNW 2-2 8. 96 0.240 0. 7958 8. 94 0.580 0.075 0. 072 0.

844'.050 NW 1-1 8. 34 2.005 0. 8326 8. 17 3. 831 1 ~ 204 0.954

. NW 2-1 9.34 0.196 0. 8390 9. 30. 0. 31.4 0. 227 . 0.079 0. 887 NW 1 NW 3-2 10. 80 -0; 099 0. 1690 10. 84 -0. 039 -'0. 141 -0.'123 0.175 NNW 1-1 5.26 1.997 0.5068 3. 72 5. 261 1.273 0.295 0.742 NNW 2-2 7.22 . 0.305 . 0.8010, 7. 23 0.539 0.424 0.018 0.854 NNW 3-2 8.83 -0.075 0.4125 8.74 -0.171 0.109 -0.118 0.752 NNW

'NW 4-1 NNW. 4-2.

3 "3 '7.6~0.021

10. 65 .-.:=-0. 073

9. 97~~>>'0.

122'.

0.0969 1426 6515 7.60 10.56

9. 64

-0.069

'0.344

-0.126

0. 116 0.024 436 0.027 0.096

-0.564 A).232 0.33, 0.993

~

4 INW 4-3 8.26 0.0003 0.00003 8.29 -0.049 0.045 0.011 0.038

l ( I Gc) 1I.C / g a. I'....r i I 'i (;I'"a 0'i I. s J '~i~%w) j WJ/% ~

'~ ~ 'r ~ > '0 ). J '~ J o()L>ll \' ! ~

REf~'.I 1J '6 I.'Xi 0'-'li .i.. I'""2;.' I (..-.'"..I! I.'..".. LS ~

(.c.~~~u~ ~

Y.o(ii 1 Il 9 1

02~ b3 Location gg

~ I /~~ gl.. t .'.i /)1 ~( 7'it pP~ /I> ~ G<. ":.

NNW 5-1 -0.002 0.0007 10.58 -0.006 0.003 -0.002 0.001 NNW 5-2 10. 35 -0.046 0.2347 10.33 '0.033 0.004 -0.098 0.286 NNW 5-3 10. 26 -0.061 0.2709 10.24 -0.092 -0.033 -0.052 0.277 NNW 5-4 9.37 -0.034 0. 0189 9. 19 -0.098 0. 136 -0.070 0.093 NNW 6-1 '0.95 -0.034 0. 1324 10.86 -0.011 0. 060 -0.114 0.306

s ) ~

results in Table 7 are discussed below:

(1) Estimates of the regression coefficients which are negative have no physical meaning and indicate that the data do not clear@jr show an effect due to skyshine. In some cases, .for example NNW 4-2, Model II, the negative coefficient may be stat&stically significant and the R2 value may indicate a high degree of correlation, but the result should be regarded as anomalous. 'I (2) Some improvement in the R is expected in Model II as compared with Model I; .however, a large difference in the 2

R values, such as in the case of NE 3-1, zkzmXdxha is an indication of an anomalous result.

(3) The values for b 0 are estimates of the exposure rat'e when the reactors are at zero power level, in other words they are estimates of background. Therefore, large discre-E pancies. between these values and measured background values

'reported previously XMell' d' f '

results. Also, large discrepancies between the b 0 values for the two models, such as in the case of NE 4-1, is an indication of an anomalous result.

Of particular interest here are the locations which were listed in Table 4 under Groups A and B,'.e. locations at which the operational me+ exposure rate was significantly greater than the background ~

exposure rate.~ Group A consisted of locations at which incre'ases. in both the mean and standard deviation of the exposure rate were observed,

'thus indicating a possible effect due to the operaton of the'lant.

I ~

~ ~ ~

The informatioq, in Table 7 definitely indicates an effect due to skyshine

~ ~ ~

at 21 of the 27~qcations 4~ in Group A. The six exceptions are NNE 5-1,--

ENE 4-1, ENE 5-2~4~ '."4-l, ESE 2-1, and NW 3-2. Note that this supports.--':".-

the statements yVde in a previous section of this report that Po I

"Five locations, NNE 5-1, ENE 4-1, ENE 5-2, E 4-1, and NW 3-2, are beyond 610 m (2000 ft) from the plant and are likely to be affected to a lesser extent by skyshine.

There~e, if plant operations caused the observed increases in bott'eans and standard deviations, the effect may have been due to radioactive materials in the gaseous effluent p lumen~

It is not clear why the results of the analyses do not indicate a significant effect due to skyshine at ESE 2-1. This location is in close proximity to other locations, specifically E 2-2 and SE 2-1, at which such an e ffeet is indicated. Further monitoring may show a X significant skyshine effect at ESE 2-1.

Group B consisted of locations at which significant increases in the mean exposure rate were observed, but no significant changes in the standard deviation were observed. The information in Table 7 suggests that there may be a weak effect due to skyshine at two 1.ocati ons in Group B, E 3-1 and WNW 2-1. At the fifteen remaining

~ locations in Group B, no significant effect from skyshine is indicated;

(

~ ~

At all of the locations in Group C effects due to skyshine are in'dicated. However, effects due to radwaste stored in this general area make it difficult to interpret the re'suits for these locations; A significant , ~, effect due to skyshine is indicated at two locations

( Id t. c ~ '." "i'g: O,~U. t- - I,, i .-'-.-',. .)

~> '~c in Group D WNW 2-2 and NNW 2-2. Both of these locations are in close proximity to other locations at which a significant effect due to skyshine is indicat'ed.

CONCLUSIONS AND RECOMMENDATIONS The original oBfective of this project was to refine models used to predict doses prom radioactive'aterials in the gaseous effluents c

from'a numlear'ower plant. The project was terminated before this objective was realized. However, a large amount of information is available as a result of the project which characterizes well the external radiation environment in the vicinity of BFNP.

Xn the course of the project, five ionization chamber systems were used to make measurements at 83 locations; there fore, only a small amount of time could be devoted to any one measurement. This made it difficult to determine whether an observed effect in exposure rate was due'o plant operations or due to variation in background.. It was particularly .difficult to.identify an effect due to radioactive materials in the gaseous effluent from BFNP. In only one measurement out of approximat'ely 15'00 was an effect detected which was believed to be due to a gaseous effluent plume.

At the time the funding for the project was terminated, plans were underway to change the approach to making measurements around BFNP.

These plans included the ins'tallation.of tape recorders on the instru-ments so that data could be'ollected on a continous basis and could '.

be analyzed with a .minimum of non-computerized data handling and calculation . Also, four. of the instruments were to "be E mounted at fixed locations where they would collect data continuously. The fifth instrument would serve as a backup to the others, and would be free for .making short-term measurements at other locations. From sets of contir%88s data, plume effects could be much more'asily ",

identified, quajfied; and related to plant'perations and meteorology.

This change inm'pproach sould have resulted in a more efficient us'e of the instruments and Jbuld have increased the chances,: of meeting the original objective of the project.

~ ~

I' t

The results presented in this report showed that plant operations increased the expos)Ge rate at locations in the immediate vicinity of the plant,~thin approximately 610 m (2000 ft), due to radio-active materiaFs .confined in the plant.'ffects on exposure rate due to plant @rations at locations beyond 610 m (2000 ft) from the plant were not clearly shown. )8 However, the results suggested the possibiliity of effects in two general areas. The first is generally to the NW-NNW from the plant, including locations NW 3-2, NNN 4-1, and NNW 5-2. Meteorological data indicate that a major component of the wind frequency is gaaax~ toward Wa this area.

I The second area is generally toward the ENE-E from the plant, including locations ENE 4-1, ENE 5-2, E 3-1, E 4-1, and E 5-1. However, the meteorological data indicate that the wind does not blow toward this area with,a large frequency. A possible third area is generally toward the N-NNE from the plant. A sufficient quantity of data is lacking for a number of locations in this area, but an effect which

)

~ was believed to be due to a plume was observed at N 4-2. Also, an effect was observed at NNE 5-1 which may have been due to a plume or may have been due to interference from the transmitter at the /

nearby environmental monitoring s tation or some other unidenti fied cause.

Recommendations for any future work on this*project are di'scussed below.

A. Furthei work with data already collected:

(1) The .verification and analysis of the data .from effluent monitors could be completed. Due to the .nature of the data "

(large n~pis .of "less thans"), however, this may not produce

':,useful resg<s..'2)

The m6deling of exposure rate as a function of power levels of the reactors could be expanded to include the distance and direction of each 1ocationz from the plant. This would result

. ~ - in an overall model of the exposure rate profile in the, area surrounding BFNP.

due to skyshine B. Pork invoLWng further data collection.

{1) The +proach to making measurements Could be changed as di,scussed above, using fixad locations and tape recorders fcr data collection.

v/

(2) Refinements Should be made in the measurements of radio-active materials in the gaseous effluents.

YJ (3) Consideration Should be given to basing the measurements on one radionuclide, such as one of the noble gases or iodines, rather than a large mixture of radionculides. This would necessarily mean a change from ionization chambers to some other form of monitoring system.

(4) Based on the results presented in this report, the fixed I

~ monitoring stations'entioned in (1) should be located in the two or three general areas discussed above..

( t ~

'w )

l'1