ML20004G123

Forwards Responses to NRC 810529 Request for Addl Info Re Fsar.Info Will Be Incorporated Into Next Revision to FSAR

ML20004G123
Person / Time
Site:	Callaway
Issue date:	06/23/1981
From:	Bryan J UNION ELECTRIC CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
ULNRC-456, NUDOCS 8106290230
Download: ML20004G123 (27)

Text

.

s UNION ELECTRIC COMPANY seOS GRATIoT STREET ST. Louis. Missoun JOHN K. BMYAN

, mox 34e June 23, 1981 Mr. Ilarold R.

Denton Director of Nuclear Reactor Regulation U.

S.

Nuclear Regulatory Commission Washington, D.C.

20555

Dear Mr. Denton:

ULNRC-456 DOCKET NUMBERS 50-483 AND 50-486 CALLAWAY PLANT, UNITS 1 & 2 FINAL SAFETY ANALYSIS REPORT

Reference:

NRC letter dated May 29, 1981 signed by R.

L.

Tedesco The referenced letter requested additional information concerning the Callaway Plant FSAR.

Transmitted herewith are responses to questions in the referenced letter.

This information will be formally incorporated into the Callaway Plant FSAR in the next revision.

This information is hereby incorporated into the Callaway Application.

Ver truly yours, John \\

Bryan {?@y h

r%

NCS/afg

m.

=.

STATE OF MISSOURI )

)

SS CITY OF ST. LOUIS )

John K. Bryan, of lawful age, being first duly sworn upon oath says that he is Vice President-Nuclear and an officer o_f Union Electric Company; that he has read the foregoing document and knows the content thereof; that he-has executed the same for and on behalf of said company with full power and authority to do so; and that the facts therein stated are true and correct to the best of his knowledge, i.nformation and belief.

By -

L h ~ FA h'" RI' Eryan(

jice Presidenh

' Nuclear SUBSCRIBED and' sworn to before me this 23rd day of June, 1981 hAsw 0C kbf BARGARfJ. PFAFf # l' h0TARf PUBLIC. STATE OF MISSOURI MY COMMIS$10N EXPIT-ES APRIL 22,1985 ST. LCuts COUNTY.

cc:

Glenn L.

Koester Vice President Operations Kansas Gas & Electric P.O.

Box 208

~ Wichita, Kansas 67201-John E. Arthur Chief Engineer Rochester Gas &' Electric Company 89 East Avenue Rochester, New York 14649 A. V.

Dienhart Vice President Plant Engineering and Construction Northern States Power 414 Nicollet Mall Minneapolis, Minnesota 55401 Donald T. McPhee Vice President Kansas City Power and Light Company 1330 Baltimore Avenue Kansas City, Missouri 64141 Gerald-Charnoff, Esq.

Shaw, Pittman, Potts & Trowbridge 1800 M.

Street, N.W.

Washington, D.C.

20036 Nicholas A. Petrick Executive Director SNUPPS S Choke Cherry Road Rockville, Maryland 20850 W.

Hansen Callaway Resident Office U.S. Nuclear Regulatory Commission RR#1 Steedman, Misscuri 65077

.r7 7

y

~

4 I'

T

ITEM 451.1C:

The frequency o f'i' lightning strikes is not presented in the discussion of severe weather and P-extreme meteorol'ogical conditions in Section 2.3.1 of the FSAR.

Provide seasonal and annual estimates of lightning strikes to safety-related structures at the site considering the " attractive area" of the structures.

A suggested reference for this type of-analysis is J.

L.

Marshall, Lightning Protection, 1973.

See also Section 2.3.1,of the Wolf Creek FSAR for a discussion of exppeted lightning strikes to ground as a function of number of thunderstorms.

5

RESPONSE

T) a frequency of lightning strikes to an area is felated to the number of thunderstorm days in that

. area.

In order to characterize the expected frequency of lightning strikes to the area of the Callaway Plant, data from Columbia, Missouri regarding the average number of thunder-storm days over a 30-year period were used.

These data were presented in Table 2.3-1 of the FSAR-and are summarized below:

SEASON THUNDERSTORM DAYS Winter (January through 1 arch) 5 Spring (April through June) 22

. Summer (July through September) 22 Fall (October through December)

_ 6_

Annual Total 55 The following discussion, which estimates the number of lightning strikes to safety-related structures at the site, was developed following the methodology presented by J.

L.

Marshall in Lightning Protection, published in 1973.

The

" attractive area" of the structures was determined for a lightning strike with an electrical current magnitude of 20,000 amperes, which corresponds to the current magnitude of 50 percent of lightning flashes.

The attractive area (A) of a structure is:

A = LW + 4H (W + L + uH) where:

L = structure length, meters; W = structure width, meters; and H = structure height, meters.

D 451.lC-1

The groupirig of safety-related structures that maximizes the attractive area is composed of five structures.

These are the reactor building, control

building, auxiliary building, diesel generator building, and fuel building.

For simplicity, this grouping has been assumed to have the following dimensions:

L = 99.1 m W = 91.4 m H = 63.4 m The assumed dimensions are the maximum linear dimensions of this grouping and, thus, maximize the attractive area of the structures.

These dimen' ions yield an attractive ' area of s

0.108 km2 The number of lightning strikes to earth per thunderstorm day per square kilometer (N ) is given by:

E NE= (0.1 + 0.35 sin x) (0.40 + 0. 20) where:

x = the geographic latitude.

Using the approximate plant latitude of 38047',

the value of NE calculated from the above couation is NE 0.128.

Then, the number of lightning

=

strikes per square kilometer per year is:

thunderstorm days ashes NE x 55

= 7.04 year km2 year Since the safety-related structures of interest have an attractive area of 0.108 km2, the proba-bility is that there will be:

flashes flashes 7.04 x 0.108 km2 = 0.76 km2 ye r year or one lightning flash every 1.32 years (480 days).

From the Wolf Creek FSAR, it was seen that the number of flashes to ground per square mile per ye_ar is between 0.05 and 0.8 times the number of thunderstorm days per year.

For the Callaway Plant area, this is between 3 and 44 lightning strikes per square mile per year, or between 1 and 451.lC-2 l i

'I

17 lightning strikes per square kilometer per year.

The number previously calculated (7.04 lightning strikes per square kilometer per year) falls within this range.

The seasonal estimate of lightning strikes to safety-related structures considering their attractive area is presented below:

NUMBER OF FLASHES SEASONS FOR SEASON PER SEASON ONE FLASH Winter 0.07 14.5 Spring 0.30 3.3 Summer 0.30 3.3 Fall 0.08

12.1 REFERENCES

Dames & Moore, 1980, Final safety analysis report, Wolf' Creek, for Kansas Gas &

Electric Co.

Marshall, J.L.,

1973, Lightning protection:

John Wiley and Sons.

]

e I

451.1C-3 l

,. j

ITEM 451.2C:

The tornado statistics presented in Section 2.3.1.2.6 are based on a regional data base that ended in 1971.

Identify any tor: ados that have occurred in the vicinity of the site since

1971, and provide estimates of the intensity (maximum wind speed) and path area of each.

Compare the annual tornado strike probability for this period with strike probabilities determined for previous periods of record (see pages 2.3-6 and 2. 3-7).

RESPONSE

The publication Storm Data, published by the National Oceanic and Atmospheric Administration (NOAA), was consulted to obtain information concerning tornado strikes in the vicinity of the site for the period 1972 t,h rough 1980.

The area comprising this vicinity was assumed to include Callaway County and the seven-county area surrounding Callaway County.

The counties investigated are Audrain, Boone, Callaway, Cole, Ga nonade, Montgomery, Osage, and Warren counties.

The tornados recorded in these countins are shown below, along with an estimate of the path area of each.

No estimate of the maximum wind speed that occurred was available from this source.

In order to provide some indication of the intensity of the tornado, an estimate of property and crop damage is included, also obtained from the NOAA publication.

The parameters of the design basis tornado for the Callaway Plant, which were obtained f rom Regulatory Guide 1.76 (1974), are shown in Section 2.3.1.2.6.2 of the FSAR Addendum.

PATil PATH ESTIMATED LOCATION LENGTil WIDTil DAMAGE *

(COUNTY)

DATE (km)

(m)

PROPERTY CROPS Boone 09/07/72 0.2 46 4

Boone 03/13/73 11.3 46 4

Boone 05/26/73 4.8 46 5

Callaway 07/20/73 1.6 46 3

7 05/12/80 40.2 46 4

4 a ay Montgomery 05/12/80 1.6 91 5

2

• Storm damages are placed in nine categories:

1 - Less than $50 6 - $500,000 to $5 million 2 - $50 to $500 7 - $5 million to $50 million 3 - $500 to $5,000 8 - $50 million to $50L million 4 - $5,000 to $50,000 9 - $500 million to $5 billion 5 - $50,000 to $500,000 l

451.2C-1 1

For this 9-year period, six tornados were recorded.

Using the method.of Section 2.3.1.2.6.1 and assuming that the eight-county area corres-ponds to the 1-degree, longitude-latitude square, the annual strike probability (Pg) is computed

-from the data period 1972 through 1980:

Ps =

= 5.01 x 10-4 This figure is comparable to the strike proba-bility computed using the Pautz (1969) data where Ps = 7.5x10-4 and is somewhat less than the value determined by Poultney (1973) where the annual strike probability was found to be 1.21x10-3,

REFERENCES:

Pautz, M.E.,

1969, Severe local storm occurrences 1955-1967.

Office of Meteorologlcal Operations, Silver Spring, Maryland, ESSA Technical Memo, WBTM FCST 12.

Poultney, N.E.,

1973, The tornado season of 1972.

Weatherwise,- 26, 22-27.

U.S.

Department of Commerce, 1972-1980, Storm data.

NOAA, Washington, D.C.,

volumes 14 through 22.

c l

451.2C-2

-r--

ITEM'451.3C:

Describe the procedures used for determining the

. meteorological conditions which would result in

'the minimum heat transfer rates (Tables 2.3-13 and i

2.3-14) and the greatest evaporation from the retention pond (Table 2.3-15) for design of the ultimate heat sink.

RESPONSE

The follow *ng procedures were used to obtain the meteorological data required for the design of the ultimate heat sink.

All average and extreme values of meteorological parameters were based on 3-hourly data for Columbia, Missouri, for-the 25-year period January 1,

1945 to October 31, 1969.

The. data were obtained from the U.S.

Department of Commerce, National Climatic Center, Asheville, North Caroline, on magnetic tapes in TDF-14 format.

Since many of the calculations are concerned with daily averages of meteorological parameters, an additional data file was compiled consisting of average values of the following meteorological parameters for each calendar day of the data period:

cloud cover, wind speed, dry-bulb, wet-bulb, and dew point temperatures, and relative humidity.

A separate dew point temperature value was calculated for each 3-hourly observation using dry-bulb and wet-bulb values, and the daily averages for the calculated dew point appear in the daily average data file, together with daily averages for the depression of the wet bulb and the depression of the calculated dew point.

The calculated dew point was used in the data analysis instead of the original dew point, since wet-bulb measurements are generally more reliable than direct measurements of dew pcint.

Where 30-day average values of a parameter were required, the 30-day periods were obtained by taking each consecutive day as the beginning of a particular 30-day period.

For example, after June 1 - 30, the 30-day period of June 2 - July 1 was considered rather than July 1 - 30.

The data set of daily average values was used in computing 30-day averages.

In computing averages, a minimum of four 3-hourly average values were considered necessary for a I-i 451.3C-1 f

1. 3

valid daily average, and 15 valid daily, values were used as the minimum for a valid 30-day average.

m After the highest 30-day average period for a given parameter was determined, t.h e daily values of the parameter were determined.

The daily values of the parameter for the period were obtained by listing the required portion of the data set.

The evaporative heat flux for each day within the summer months of June to September was calculated using the. following equation from "An Analytical and Experimental Study of Transient Cooling Pond Behavior" by P.O.

Ryan and D.R.F.

Harleman, Ralph Parsons Lab, MIT Report N_o.

161, January 1973:

4 = f (w) (es - e) 2 where:

4

= evaporative heat flux (Btu /ft / day);

e

= saturation vapor pressure (mm Hg);

s e

= actual vapor pressure (mm Hg) ; and f (w) = 70 + 0.7 w2 wind speed (mph).

where w =

The period of minimum heat transfer was determined by finding the period of highest equilibrium temperature for the retention pond.

The equili-brium temperature, as defined by Ryan and Harleman, is a function of net radiation, wind

speed, ambient temperature, and dew point temperature, in accordance with the 'f ollowing equation:

, $r a f(w) [8Td + 0.255 Ta] - 1600 23 + f(w) (S + 0.255) where:

TE

= equilibrium temperature toward which the pond tends (OF);

2

&r

= net radiation term (Btu /ft / day);

= 4sn + 1.16x10-13 (460 + T )6(1 + 0.17c2);

a

&sn

= net incident solar radiation 2

(Btu /ft / day);

2 (1 - 0.71c ) H x 24;

=

o lj 451.3C-2

1 i

i H

= average daily absorbed solar o

2 radiation for clear sky (Btu /ft / day);

= 68.362-40.982 x sin [2n x (DAY /3 66) +1. 739]

for 390 latitude; DAY

= sequential number of the day of the year beginning with I for January 1 and ending with 365 or 366 for December 31; c

= average cloud cover (in tenths);

2 (Btu /ft / day /mm Hg);

f (w) = wind function

= 70 + 0.70 x ws2; The Brady form of f(w) instead of the Lake !!cfner form [f (w) = 12.4 x ws2]

in the above reference was used because the latter is physically unrealistic and gives excessive values of TE for low wind speeds.

ws

= wind speed (mph) ;

6

= 0.255 - 0.0085 T* + 0.000204 T*2 (mm !!g/0F);

T*

= 1/2 (Ts+T) ( F);

d Td

= dew point (OF) ;

T

= ambient temperature (OF);

and a

Ts

= surface temperature of the pond (OF).

The 3-hourly observations for ambient temperature, wet-bulb temperature, cloud cover, and wind speed were averaged to obtain daily values.

The dew point temperature was then calculat~ed from the ambient temperature and wet-bulb temperature.

The equilibrium temperature was calculated for each day using the above parameters.

An initial value of T was assumed and T*

and 8 were s

l l

calculated.

Then, for given values of &r, Tde i.

l I.

t

(!

f.'

451.3C-3

~ ~ '

~.

Ta, and f(w), the value of TE was calculated.

The difference, (TE Ts)/2, was then used as an improved estimate of the value of T

-and the 3

process was repeated until the difference became less than or equal to 0.50F.

Generally,. the equilibrium temperatures are found within 4 to 5 iterations.

The 40 highest daily equilibrium temperatures were found, as well as the 40 highest 30-day equilibrium temperatures.

The 30-day equilibrium temperature is the average of a 30-consecutive-day period between June 1 and September 30.

REFERENCE:

Ryan, P.J.,

and Harleman, D.R.F.,

1973, An analytical and experimental study of transient cooling pond behavior.

Ralph H.

Parsons Laboratory, Department of Civil Engineering, MIT, Report no. 161 -(January).

h 1

451.3C-4

.e 1

ITEM 451.4C:

Table 2.3-37 of the-FSAR indicates that moderately stable (Pasquill Type F) and extremely stable L-3 (Pasquill Type G)e conditions have persisted for 20- and 19-hour periods, respectively, at the Callaway Site during the period May 1973 May 1975.

Persistence of these stability classes for, periods greater than 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> is very unusual.

Discuss the causes of persistent stability conditions for greater than 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> for Classes F and G.-

Identify the synoptic conditions during the observed periods of persistent F and G stab _ility classes for periods greater than 12 ho"ps, and discuss the possibility of instrument mrifunction.

RESPONSE

Reanalysis of the two F,SAR years 1973-1975 occurred after a portion of the 90-10m delta temperature values, which were misaligned and located in a data file used to calculate stability, were corrected.

The results are shown s

in Table 451.4C-1, and the corresponding instances where F and G stability persisted for a period of greater than 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> are presented below.

These were:

HOURLY STABILITY TIME PERIOD PERSISTENCE F'

05/27/73-1400 to 05/ 28/73-0900 20 F

07/05/73-1900 to 07/06/73-0700 13 G

10/25/73-2100 to 10/26/73-0900 13 F

12/21/73-2000 to 12/22/73-1000 15 F

02/12/74-1800 to 02/13/74-0700 14 F

11/01/74-1900 to 11/02/74-0800 14 Stability persistence time periods ending on 12/22/73 at 1000 and on 11/02/74 at 0800 occurred in advance of a low pressure system.

Prevalent meteorological conditions were cloudy skies with very little surface heating, all of which increased stability of the surrouncing air.

There was no evidence of instrument malfunction.

Stability persistence time periods ending on 07/06/73 at 0700, 10/26/73 at 0900, and 02/13/74 at 0700 occurred during strong high pressure system passage.

Meteorological conditions were clear skies, which prometed radiational cooling and thereby increased stability.

Again, there was no evidence of an instrument malfunction.

451.4C-1

. o TABLE 451.4C.1l 1

~.

4 CALLAWAY GENERATING STATION REFORM, MISSOURI UNION ELECTRIC COMPANY DAMES AND MOORE JOD NO 7677-066-07 DATA PERIOD FROM 5/ 4/73 TO 5/ 4/75 DATE AND TIME OF RUN 06/15/81.

14.34.59.

J NUMBER OF HOURS 4

NUMUER OG PASGUILL STABILITY CLASS i

CONSECU11/E

!!OURS

-A-

-D-

-C-

-D -

-E-

-F-

-G-2 392 198 292 4999 3778 1656 670 3

251 66 113 4050 2919 1130 486 4

153 27 48 3377 2313 777 354 5

91 6

20 2897 1872 531 250 6

50 0

8 2516 1524 371 168 7

27 0

2 2209 1242 260 110 8

15 0

0 1954 1008 176 71 9

7 0

0 1740 814 114 41 10 2

O O

1559 641 73 22 11 O

O O

1413 500 44 10 12 O

O O

1296 382 25 4

i 13 0

0 0

1199 279 16 1

14 O

O O

1112 202 11 0

15 O

O O

1034 140 7

0 16 0

0 0

964 93 5

0 17 0

0 0

900 63 4

0 18 O

O O

843 42 3

0 19 0

0 0

792 32 2

0 20 0

0 0

747 25 1

0 21 O

O O

702 19 0

0 22 O

O O

659 14 0

0 23 0

0 0

621 11 O

O 24 O

O O

586 8

O O

>24 0

0 0

551 5

0 O

i i

671 INVALID HOUN(S).

~

l f

451.4C-la

~

Finally, the most persistent stability time period of 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> occurred between May 27 and May 28, 1973.

Although the stability classes._ determined

~

by the two delta temperature sensors differ, this

.could be accounted for by the weather system that-passed through Missouri over that 2-day period.

During that time period, a slow-moving cold _ front from'a-deep low pressure' system moved through Missouri.

-A low-level inversion dces occur during these ' episodes and causes fog.

Because this slow-moving low pressure system traveled almost dirpetly over Fulton, Missouri, the large spread in delta temperature values was possible.

Fog and li]ht. rain showers were reported from this system oi May 27, 1973 in Springfield, Missouri and Laa ha, Nebraska.

Although the weather map data only reproduce conditions at one time' period on

.May 27, it is probably safe'to assume that fog did occur before the advancing cold front.

.If this is the case, then a low-level inve-sion would have occurred and caused the great di terence in delta temperature values.

The slow movement of 'the system would have caused a persistent F stability for the 60-10m delta temperature.

REFERENCE:

Daily Weather Maps, U.S.

Department of Commerce.

S i

451.4C-2 W

v

]

3, g.

Finally, the most persistent stability time period of 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> ' occurred between May 27'and May 28, L...

1973.

Although the stability classes determined uy the two delta temperature sensors differ, this

~

could be accountad for by the weather system that passed through-Missouri ove r.. tha t 2-day period.

During that time period, a slow-moving cold front.

from a deep low pressure system moved through Missouri.

A low-level inversion coes occur during these epistdes and'causes fog.

Because this slow-moving low pressure system traveled almost ditpctly over Fulton, Missouri, the large spread in delta temperature values was possible.

Fog and 15 lb,t rain showers were reported from this system oi May 27, 1973 in Springfield, Missouri and Taaha, Nebraska.

Although the weather map data only reproduce conditions at one time period on May 27, it is probably safe'to assume that fog did occur before the advancing cold front.

If this is the case, then a low-level inversion would have occurred and caused the great difference in delta temperature values.

The slow movement of the system would - have caused a persistent F stability for the 60-10m delta temperature.

REFERENCE:

Daily Weather Maps, U.S.

Department of Commerce.

i 451.4C-2 I

ITEM 451.5C:

In the discussion of the potential influ6nce of the plant and its facilities on local meteorology 3-(Section 2. 3. 2. 2),, two somewhat different sets of

~

design parameter.c for the natural draft cooling towers are presented (see pages 2.3-23 and 2.3-31).

Clarify the. design characteristics for the natural draft cooling towers, particularly for thel exit diameter and heat rejection rate.

RESPONSE

The design characteristics for the natural draft co ef. i ng towers contain-typographical errors on paces 2.3-23 and 2.3-31 of the FSAR.

The correct pa ameters are:

-7..

Tower height, 555 feet; b.

Diameter of top of tower, 252.7 feet; c..

Heat rejectiori rate, 8.04 x 109 Btu /hr per tower; and d.

Water flow rate, 568,000 gpm per tower.

The heat rejection rate as listed on page 2.3-23 is incorrect, and the tower exit diameter as listed on page 2.3-31 is incorrect.

However, the correct parameter:, listed above were used in the analysis.

e I

451.5C-1

ITEM.451.6C:

In the calculation of. cooling system impacts, wind speed and wind dire'ction measurements at the 60m level were used,to determine conditions repre-s sentative of the top of the cooling tower (at about 170m above the surface).

Discuss the rationale for using. measurements from the 60m level when similar measurements were available from the 90m level.

Also discuss the validity of use of the wind speed power law described on pages 2.3-25'and 2.3-26 to extrapolate from measurements at 60m to represent conditions at 170m.

i

' RESPONSE:

A. # Wind Speed and Wind Direction la the analysis of cooling system impacts, the results of an on-site meteorological monitoring program were utilized.

The information available for use in the cooling tower model (TOWER 1 as described in the FSAR) consisted (in part) of the following:

PARAMETER LEVEL Wind speed / direction 10 m Wind speed / direction 60 m Wind speed / direction 90 m Temperature 10 m Dew Point 10 m AT 90/10 m AT 60/10 m Wind speed and direction measurements were available at three levels, namely, 10, 60, and 9.0 m AGL.

Temperature lapse (AT) measurements were available over two intervals, 90-10 and 60-10 m.

For the analysis of cooling system impacts, wind speed and direction measurements from the 60 m level were used in conjunction with temperature and dew point measurements at the 10 m level at A T measurements over the 90-10 m interval.

The rationale for using wind speed and direction measurements from the 60 m level as opposed to the 90 or 10 m levels was based primarily on compatibility with AT measurements and data recovery.

Inasmuch as the 90-10 m AT measurements span essentially the entire surface

~"

layer (assumed to be the lowest 100 m of the friction laye r), they are ideally suited for the determination of stability in the lowest layers of the atmosphere.

In addition to a more favorable 451.6C-1

i e

e n

data recovery for the 60 m wind measurements, it was felt that it would be more appropriate to us:e 1

wind measur.ements that were bracketed by the temperature lapse measurements rather than to have wind speed and direction measurements at the upper or lower end of the temperature lapse measurement interval.

Presumably, this approach will be more representative of average conditions in the layer over which atmospheric stability was calculated.

B.

Wind Speed Power Law In, order to extrapolate wind speed measurements at ti e,60 m level to represent conditions at the top c

the cooling tower (370 m), a simple power law

'as used.

T' ) power law as used in TOWER 1 was as s

follows:

8 h

Uh=Uo (g) where:

Uh = wind speed at cooling tower height (m/s);

Uo = wind speed at 60 m (m/s);

height of cooling tower (170 m);

h

=

i height of wind sensor (60 m); and h

=

o power law exponent s

=

= 0.25 for unstabic/ neutral conditions

= 0.50 for stabic conditions.

The use of the power law is consistent with current theories on the vertical structure of wind speed in the surface layer.

This formulation has been used by many investigators such as Frost (1948) and Sutton (1953).

Frost estimated that the value of the power law exponent should vary between 0.1 for extremely unstable atmo-spheric conditions and 0.8 for extremely stable conditions.

Inasmuch as the atmosphere rarely exhibits extreme behavior, it is more reasonable to assume values for the exponents that are more representative of typical atmospheric stability conditions.

The values used of s=0.25 (unstable /

neutral) and s=0.5 (stable) are within the range of values used by there earlier researchers.

The l

results obtained with the predictive model TOWER 1 l

should be less sensitive to choice of power law i

451.6C-2

-y ;

exponent -in the power law extrapolation than to 1

the choice of crite'ria. used in the determination 1

of' atmospheric stability.

~~

REFERENCES:

Frost, R.,.1948,-Quart.

J.

Roy. Met.

Soc., vol.

74.

Sutton, O.G.,

1953, Micrometeorology.

McGraw-l-

Hill, New York.

(

/

F 4

T i

4 e

1 l

9 l

451,6C-3 H,

j

ITEM 451.7C:

The discussion of data recovery on page 2.3-49 indicates that data from other levels and intervals were substituted to enhance the data recovery for the primary measurements, i.e.,

wind speed and wind direction at the 10m level and temperature difference between 10m and 60m.

Preliminary analysis of the hour-by-hour meteoro-logical data provided on magnetic tape suggests that about 23% of the primary data for tre combined three-year period (5/73 - 5/75 and 3/78 -

?/79) had to be substituted for the primary measurements.

Discuss the problems with the data collection program which necessitated such a large fraction of substituted data, and indicate what modifications will be made to the operational program to enhance data recovery of the primary meteorological measurements.

Also discuss the difficulties in measuring" precipitation at.the site which necessitated use of precipitation data from Columbia, and indicate the real-time representativeness of Columbia precipitation data for use at the Callaway site.

RESPONSE

The Union Electric Company's meteorological monitoring system consists of Climet wind systems located at the 10, 60, and 90-meter (m) levels; Climet delta temperature systems that measure temperature differences between the 10- and 60-m levels and also the 10- and 90-m levels; an EG&G cooled mirror dew point system at 10 m; a back-up Climet lithium chloride (LICl) dew point system at 10 m; a Climet temperature sensor at 10 m; and a Climet weighing bucket rain gauge at 2 c.

All data as of March 1978 are recorded on Esterline Angus (EA) analog recorders.

The sequential multipoint recorder, EA Model Ell 24E, records the reference temperature, LICl dew point, and both delta temperatures.

(In Phase I of this study, the multipoint also recorded the 90-m deu point.)

Three EA E1102S side-by-side dual-pen analog recorders record the wind speed and wind direction at all three tower levels.

The EG&G cooled mirror dew point is also recorded on a separate EA L110lS analog recorder.

The weighing bucket rain gauge records precipitation on an EA 6016 analog recorder.

Before March 1978, digital data were available to augment the analog data, but with the beginning of the last FSAR data t

451.7C-1 s

collection year, this digital data system was judged unsuitable as a back-up system and, therefore, was not. used in the final esA9 yaar.

The Union Electric 3-year data collection effort has been noteworthy because of the problems that the instruments and recorders have had.

The dual-pen recorders that record wind speed and wind direction have capillary inking pens.

The pens have had a tendency to accumulate ink at the tip; the ink drics, blocking ink flow and preventing data from being recorded on the analog charts.

This occurrence does not take place at all recorders concurrently, and if it does happen at the 10-m primary data level, data from either the 60- or 90-m wind sensors are substituted after the data are adjusted to height.

The multipoint recorder has had numerous break-downs over the 3-year period.

Another multipoint recorder is used if the original recorder is not repairable at the site.

The original recorder is placed on line after being repaired by the manufacturer.

The EG&G cooled mirror dew point, like the multipoint recorder, has been sent back to the manufacturer a number of times for repairs because of failures within the dew point system.

In the event of the EG&G dew point failure,'the LiCl dew point data are substituted until the EG&G dew point is back on line.

The 60-10m delta temperature displayed inter-mittent problems in the first 2 years of data collection.

This problem appeared during periods of high humidity.

Numerous tests were performed on the 60-10m delta temperature system to no avail.

Finally, the problem was traced to a small crack in the tower cabling from the 60-m level.

All tower cabling was replaced and the problem ceased.

When the 60-10m data did appear suspect, it was invalidated and 90-10m delta temperature data substituted.

In addition to instrunent and recorder problems, the Union Electric meteorological tower has been hit by lightning, ice storms, and freezing drizzle.

Lightning has struck the tower at least three times, krocking out all instrumentation.

Freezing drizzle and ice storms have frozen L

451.7C-2

the wind sensors and stopped the sensors from functioning normally.

In-March 1981, heaters were installed on all three levels of wind cansnrn en prevent this icing problem.

The combination of recorder ralfuncticns, sensor malfunctions, and acts of God have worked together, yielding reduced data recovery rates at the 10-m primary level.

Procedures have been implemented to increare the data recovery for all parameters.

These procedures consist mainly of intensified inspection of the monitoring system operating parameters by, Union Electric personnel performing. site checks in order to more quickly identify potential problems and respond with remedial measures._

It is expected that this increased attention to system operation, along with the new tower cablihg and sensor heaters (where appropriate), will increase the valid data recovery.of the meteorological monitoring system.

(

As can be seen in the response to Item 451.8C, data recovery of meteorological parameters has been generally above the 90 percent rate of recovery specified for most such parameters.

Instrument operating difficulties were experienced with the precipitation gauge at the Callaway site.

Since precipitation events can produce significant quantities of precipitation during short periods of time, even short periods of instrument outage can result in serious distortion of the data base.

The Columbia National Weather Service is within 40 km (25 miles) of the Callaway site, and there are no intervening topographic features to suggest the two locations would have different precipi-tation climatologies.

Therefore, it was decided that the Columbia precipitation data were probably more representative of the Callaway site than the short-term data available from the on-site sensor.

Considering the seasonal and annual anomalies that can occur in precipitation data, the Columbia period of record is almost certainly more representative of the Callaway site than any 2-or 3-year period measured on site.

More emphasis has been placed on the careful operation of the on-site precipitation sensor since March 1979.

Except for a 3-month period in

1980, it has been operating at better than 90 percent data recovery.

During that 3-month period, en evaporation study was conducted at the

'451.7C-3

UHS retention pond that included the measurement of precipitation that can be substituted for missing data-during that peri:d.

.1: 0, th:

primary precipitation sensor at the on-site tower is being replaced to provide a more accurate, reliable -data base.

The replacement sensor will use the tipping bucket method of measurement.

This method is considered superior, with respect to accuracy, reliability, and resolution, to the presently used method of determining precipi-tation, which is a weighing bucket.

F Altho' ugh it is r ecogni::ed that in real time, "recipitation data from Columbia may be skewed or differ from that of the Callaway site, such as

_3 a rainstorm should arrive at the two locations at different times or if it should arrive at one and not the other, it is expected that the Columbia data will be comparable to conditions at the callaway site.

Although the data since March 1979 have not been recovered from the strip chart recordings, they are available for making a real-time comparison between Columbia and callaway or a longer-term comparison when a sufficiently large data base is available to average out seasonal and annual anomalics.

I 451.7C-4

?,e

'0y b

ITEM 451.8C:

Describe the status of the-on-site meteorological measurements program since March 1979.

-RESPONSE:

Since March 1979, the on-site meteorological monitoring program has continued to operate.

The instruments _ are checked thrce times per week by Union Electr'c-Nuclear Operations and calibrated quarterly F James & Moore.

The data are recorded on analog recorders.

The strip chart records are reviewed to verify that the data are acceptabic and then archived at the Dames & Moore office in the Chicago area.

Estimated percentage data recovery rates for ea,ch parameter are as follows:

04/79 01/80 01/81 to to to PARMETER 12/79 12/80 02/81 Wind Speed, 10m 90 92 94 Wind Speed, 60m 91 94 98 Wind Speed, 90m 90 92 98 Wind Direction, 10m 89 93 85 Vind Direction, 60m 88 96 98 Wind Direction, 90m 83 91 97 Temperature, 10m 93 94 100 Delta Temperature,60-10m 93 94 100 Delta Temperature, 90-10m-93 90 100 f

LICl Dew Point, 10m 91 94 100 Cooled Mirror Dew Point, 10m 32 58 98 Precipitation, im 94 74 92 4

t' e

D eg 451.8C-1

ITEM 451.9C:

Tables 2.3-66, 2.3-67, and 2.3-68 present terrain /

recirculation correction factors to be applied to a straight-line Gaussian diaparcian na d a

• a better characterize temporal variatioas in meteorological conditions.

These correction factors were estimated based on the results of a variable-trajectory puff advect.fon model using one yaar of hour-by-hour meteorological data from the Callaway site.

Substantial reductions (up to a factor of 100 lower than the straight-line model) are suggested for distances approaching 50 miles.

Discuss the reasonableness and appropriateness of correction factors for receptors at distances greater than about 5 miles from the source developed by use of a variable-trajectory model with only a single source of meteorological data as input.

Alsc discuss the use of site-specific wind speed profiles f o r' this analysis when standard wind speed profiles are assumed for data substitution and cooling tower impact assessments.

RESPONSE

Dames & Moore's variable-trajectory puff advection model, PUFF, was used, along with a straight-line model, in the derivation of terrain / recirculation correction factors (TCFs).

PUFF tracks the advection and dispersion of up to 500 Gaussian puffs across the study area.

New puffs are emitted continuously at 20-minute intervals throughout the year.

ruffs are discarded when they leave the study area, or when they have become so attenuated that they no longer have a significant impact at any receptor Jocation.

The criterion for discarding an attenuated puff is comparison of the puff center X /0 to a user-specified cutoff X/0 value.

In the original analysis, this cutoff was inadvertently set to an inappropriately high value.

The result was that puffs were discarded too quickly, before they could reach the more distant receptor locations.

The PUFF model analysis has been repeated for ground-level release using a more appropriate X /0 cutoff value.

Revisdd TCFs are presented in Table 451.9C-1 for the 10 receptor ring distances used in the PUFF analysis.

As this table indicates, the strong systematic underprediction of PUFF model results in relation to straight-line model results for large source-receptor distances is no longer present.

451.9C-1

TABLE 451.90-1 TERRAIN / RECIRCULATION CORRECTION FACTORS AT TEN STANDARD DISTANCES CROUND-LEVC*. RELEASE L

DASED ON PAY 4.

1974 TO MAY 4.

1975 DATA DISTANCE SECTOR (MILES)

NNE FE ENE E

ESE SE SSE S

SSW SW WSW W

WNW NW NNW N

.25

.99

.87 1.05

.77

.91

.90

.98 1.03

.89 1.03

.99 1.14 1.07 1.02

.95

.99

.75 1.04

.92

.98

.87

.91

.95 99

.94

.93 92 1.10 1.13 1.13

.95

.94 1.06 1.50 1.13 1.03 1.01

.93

.96

1. G 4 1.00

.89

.98 93 1.08 1.01 1.05 1.00

.89 1.00 2.50

4. 11

.41

.93

.80

.91

.91

1. 01

.95

.99 99 1.16

.94 1.10

.96 93 1.03 3.50

.99

.79

.83

.75

.99 1.00

.99

.93

.84 94 1.14 1.03 1.05 95

.87

.97 5.03

.92

.72

.77

.74

.92

.93 1.04

.98

.00 93 1.04

.91 1.03

.94

.77 93 10.03

.95

.75

.80 74

.80

.78

.96

. 62

.78 60

.88

.94

.81

.92

.85

.79 20.00

.91

.64 1.01 65

.72

, 61

.73

.61

.52

.51

.65

.74

.82

.88

.81

.81 35.03

.70

.58

.91

.45

. 53

.47

.65

.37

.34

.32

.39

.55

.82

.74

. 71

.77 50.03

.56 48

. 50'

.28

.42

.28 46

.41

.32

.27

.30 4R.

.64

.51

.49

.65 Cll3 C:l3 4 51. 9 C-la llp=

m

(

The mild overall decrease in TCP values at large downwind distances may be attributed - to plume meander, accounted for in Pl!P P bie r nne in tha straight-line model.

With wind directions varying hourly, plume elements in PUFF actually cover a greater distance before arriving at a given receptor than is assumed in the straight-line model.

They are, therefore, somewhat more attenuated on arrival at the receptor than the straight-line model algorithm would indicate.

The revised PUFF model analysis utilized standard vertical wind speed profiles, rather than the site-specific profiles employed in the original PUFF analysis.

Although the site-specific profiles may more accurately reflect actual conditions at the plant site, the standard profiles were chosen in order to ensure consis-tency with other portions of the final safety analysis.

Revised TCFs will be computed for the mixed-mode case as well as the ground-level case.

The revised TCFs will be logarithmically interpolated to provide TCFs for all downwind distances of interest.

This complete set of TCFs will then be applied to all straight-line model results presented in the FSt,R.

Use of a single meteorological station as'the data source for the PUFF analysis is justified by the absence of severe terrain within the region of interest and by the fact that only long-term average relative concentrations are evaluated.

Absence of s eire r e terrain implies that deviations from straight-line flow that do occur are not strongly systematic.

Effects of random plume meander and mesoscale recirculation on annual average X/O values are adequately represented via PUFF simulations with single-station on-site meteorological input.

l t

s 451.9C-2

-