ML20147D520
| ML20147D520 | |
| Person / Time | |
|---|---|
| Site: | Vallecitos File:GEH Hitachi icon.png |
| Issue date: | 05/31/1977 |
| From: | Fujita T ARGONNE NATIONAL LABORATORY |
| To: | |
| References | |
| NUDOCS 7812200053 | |
| Download: ML20147D520 (25) | |
Text
. . _ _ _ _ - - _ . . . . __ . . - . . . . . .- -- - _ __
.4- .
.-s o
REVIEW OF SEVERE WEATHER METEOROLOGY at GENERAL ELECTRIC COMPANY _l VALLECITOS, CALIFORNIA 1 l
l by T. Theodore Fujita Professor of Meteorology The University of Chicago May 31,1977 Under Contract No. 31-109-38-3731 Argonne National Laboratory 9700 South Cass Avenue Argonne, Illinois 60439 7812200051 7 1
- ,_ _y gy,.ywr7 w - v.--ar*t--- -+-r'
'+ - I
.i 'f Review of Severe Weather Meteorology at General Electric Company Vallecitos, California by T. Theodore Fujita Professor of Meteorology The University of Chicago 1 INTRODUCTION ,
The General Electric Company, Vallecitos, California, is located in the Vallecitos Valley at 37 37' N and 121 50' W (Refer to Figure l).
D o
S mMN LIVERMORE
/ 1 g 'og v( .s Vallecitos Valley 5 4 G8 Roc FREMONT j l
3820 2000 0 0
- C o \ D
- - I W )kM(h _ A f 1
Figure 1. General Electric Company and vicinity. The Company is located in the Vallecitos Valley, 510 ft MSL. Height contours were drawn at 500 ft intervals.
i
4-2 ll %
According to Pautz (1969){1) the SELS Log reported no occurrence of 50 kt or greater wind storms and 3 tornado cases within the one-degree box of latitudes and longitudes which includes the Company site. It is the purpose of this review to determine the intensity of severe weather events which could affect this location.
Both straight-line winds and tornadoes are regarded as prime phenomena, because no herricane has ever been reported in this part of the country.
This site is located in Region II of WASH-1300, ne calculated tornado windspeed by five-degree squares for 10-# per year probability is 280 mph.( '
2 STRAIGHT-LINE Vf Q .
Straight-line w xcur more frequently than tornadoes, but their interpre-tation and evaluation are difficult. " Climatological Data" include four stations from which the fastest-mile windspeeds are available, nese stations are Oakland, San Francisco, S.F. Airport, and Stockton (see Figure 1A).. -
Most of the winds during the fastest-mile periods are from southerly directions.
SSE to SW winds were the strongest in the Bay areas. These winds occur predominantly during the cold season, October through February.
Shown in Table 1 are the maximum fastest-mile winds by year during the 26-year. period, 1950-75. He San Francisco urban winds are missing in the years 1963 and 1973-75 while the Stockton data became available at the beginning of 1964 The mean windspeed at the S.F. Airport is predominantly higher than the three other values, ne airport winds are probably higher due to the open exposure of the anemometer. De S. F. Airport winds are, therefore, normalized to those of San Francisco City simply by multiplying the ratio of mean winds from the City and the Airport. Normalized speeds are tabulated in ( ).
II)Pautz, Maurice E. (1969):. Severe Loca1 Storm Occurrences, 1955-1967 ESSA Tech. Memo WBut FCST 12
( ) WASH-1300 by Markee, E. H. , Jr. , J. G, Beckerley, and K. E. Sanders (1974):
Technical Basis for Interim Regional Tornado Criteria. U. S. Atomic Energy Commission, Office of Regulation,
. - . _ _ . - ._- j
._; _ _ . . . _ _ _ _ . - _ . , __,..._.,I
. _ - ~_- . . _ _ . - . - -.
if - '$
i Table 1. 'The maximum fastest-mile windspeed in mph by year during the,26-year period, 1950-75. From Climatological Data,1950-1975. ' Speeds in ( ) are normalized to those .
I of San Francisco City. For locations, see next page.
Year- 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 Oakland 43 48 46 49 38 46 32 46 43 40 San Francisco 43 42 46 41 35 42 35 34 40 40 S. F. Airport 51 48 50 55 48 53 40 47 48 48 (normalized) (41) (39) (40) (44) (39) (43) (32) (38) (39) (39).
Stockton -- -- -- -- -- -- -- -- -- --
Year 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 .
Oakland 41 40 35 40 46 38 35- 35 35 32 San Francisco 36 31 38 -- 36 47 36 40 35 34 S. F. Airport 52 39 50 60 59 51 48 49 45 47 (normalized) (42) (31) (40) (48) (48) (41) (39) (39) (36) (38) l Stockton -- -. -- -- 40 44 39 46 33 32 1970 1971 1972 1973 1974 1975 mean Windepeed i se 0641and 35 31 30 31 33 32 38.5 mph San Francisco 34 33 34 -- -- -- 37.8 mph S. F. Airport 41 40 32 39 39 41 46.9 mph (normalized) (32) (32) (26) (31) (31) (33) (37.8 mph) i Stockton 39 36 31 35 32 35 36.8 mph , l The seasonal variation of the occurrence of maximum windspeeds is presented in Table 2 . It is evident that the occurrences are centered in the cold months when CalHornia, in general, experiences precipitation in advance of cyclones from the Pactuc. 1 Table 2. Frequency of the maximum fastest-mile windspeed (mph) by month, 1950-75.
Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Oakland 3 2 3 4 3 11 San Francisco .5 1 3 1 2 1 2 1 6-S. F.~ Airport 7 3 3 1 1 1 1 3 1 5 Stockton 3 2 2 1 2 2 Total 18 8 5 7 2 3 2 0 1 11 5 24 m
owg -ew- -
-v-=rw ww- eww--a,-- te e c tw m ap g- we -g-,ev-m,ev9.= m w +-m.. ,>=p-,y .a-,y3 -,5
4 - 5 t. 1 1
50 MILES l
I
~
Stockton e
hSan Francisco, City f
S.Egirport Oakland 10 MLES ,
I
)
Figure 1A, Location of four climatological stations in relation to the General Electric site at Vallecitos, California. Topo-graphic features within 10-mile circle is shown in Fig.1, i
- . _- . _ . . . _ . - . - . .- - =.. - - -
4-4 Windspeeds of peak gusts are higher than those of fastest mile, because the duration of peak gust is considerably shorter than the period of the fastest-mile wind.
In this paper the former is regarded as being 25% larger than the latter. Namely, Peak Gust T 1.25 Fastest-mile Speed.
The probabilities of the occurrence of the maximum windspeed should be defined differently from those of tornadoes. For tornadoes, the National Weather Service lists all storms, even if they occur on the same day or even a few minutes later, ne maximum fastest-mile speed is listed in " Climatological Data" by month and by year. %ere is no mention as to how often the maximum speed occurred within one month or one year. Furthermore, the period of the straight-line winds, ,
especially the ones caused by continental cyclones, is long, lasting hours or days.
There will be numerous maxima during such a long period.
We should, therefore, define the following terms:
Fastest-mile day. . . . . the day on which the speed occurred.
Fastest-mile month .. the month in which the speed occurred.
Fastest-mile year.. .. the year in which the speed occurred. ,
nese are similar to the term Tornado day. . . . . . . . . the day in which one or more tornadoes occurred.
In these cases the number of occurrences within the stated period is not important, ne probabilities of the fastest-mile year can be computed by p ,
Number of years in whi specific speed or larger speed occurred Total number of years used in statistics where P. denotes the occurrence probability per year.
The probabilities in Table 3 were computed by combining the observation years at the four stations together: '26 years at Oakland and S. F. A1.. port; 22 years at San Franciso; and 12 years at Stockton. The observation years total 86
" Climatological Data". Pubitcation of NOAA, published monthly with an Annual Summary May be obtained from Environmental Data Service, National Climatic Center, Federal Building,- Asheville, North Carolina 28801
~
4-5 ,
. l Table 3. Frequencies of maximum fastest-mile of the year 1950-75. Data includes Oakland, San Francisco, City, S.F. Airport normalized to City, and Stockton.
Equivalent peak gust is estimated to be 25% larger than i fastest-mile speed. .
Fastest-mile Corresponding Years of Cumulative Probability of the' year peak gust occurrence years per year )
49 mph 61 mph 1 1 0.01 48 60 3 4 0.05 ,
47 59 1 5 0.06 46 58 6 11 0.13 45 56 0 11 0.13 44 55 2 13 0.15 l 43 54 4 17 0.20 42 53 3 20 0.23 -
41 51 4 24 0.28 40 50 9 33 0.38 l 39 49 8 41 0.48 ;
38 48 5 46 0.53 I 37 46 0 46 0.53 36 45 4 50 0 58 35 44 11 61 0.70 34 43 4 65 0.76 33 42 4 69 0.80 32 40 8 77 0.90 31 39 7 84 0.98 30 38 1 85 0.99 29 36 0 85 0.99 28 35 0 85 0.99 1 27 34 0 85 0.99 l 26 33 1 86 1.00 25 31 0 86 1 00 4
4-6 s
- 3. TORNADO FREQUENCIES During the 26 years,1950-75, a total of 29 tornadoes occurred within 144-miles of General Electric Company. The frequencies were 1.1 tornadoes per year (Refer to NSSFC Tornado Tape).
A breakdown of the frequencies for every 23-mile-range intervalis shown in Table 4. By oisregarding the area of the Pacific Ocean, included within range circles centered at General Electric, an attempt was made to express tornado frequencies as a function of the range. As shown in Figure 2, the best-fit parabolic curve is N = 0. 0018 R* ,
l
- where N is the numbur of tornadoes within the range, R, in miles.
The number of tornadoes, N, can be prorated by multiplying the ratio of land and water areas, thus N' = N land area within R This prorated number represents the tornado frequency if the entire area within Since the water area is up to any range from General Electric had no water areas.
32% of the total circular area, prorated numbers within some ranges are significantly higher than actual numbers. For prorated numbers, we obtain the best-fit parabola expressed by N, = 0.0025 R' where N, denotes the prorated tornado number or frequency.
Shown in Table 5 are the frequencies by month as listed in the 1950-75 NSSFC There is a tape. Apparently the tornado season begins in January and ends in May.
small peak in October, however.
Table 4. Frequencies of tornadoes within 144 miles from General Elsetric Company as a function of the distance from the Company. Based on NSSFC tape, 1950-75.
40-60 60-80 80-100 100-120 120-125 Distance n. miles 0-20 20-40 46-69 69-92 92-115 115-138 138-144
- s. miles 1-23 23-46 7 3 5 5 7 1 Frequency 1 1
4-7 9
~
4o ti 30 N, = 0.0025 R*
(Water area prorated ) ,
t 20
/*
l o
- o g N = 0.0018 R*
8 0 40 60 80 10 0 12 0 I40 O 20
- RANGE IN MILES Figure 2 Cumulative number of tornadoes as a function of the distance from General Electric Company Based on the 29 tornadoes in NSSFC tape, 1950-75.- Open circles represent actual number and painted circles are prorated numbers.
4-8 Table 5. Frequencies of' tornadoes within 144 miles from .
General Electric Company by month. Based on NSSFC tape,
, 1950-75.
Month Jan Feb Mar Apr may Jun Jul Aug Sep Oct Nov Dec Frequencies '3 3 3 12 1 2 0 0 0 4 1 0 The tornado frequencies by year, according to the NSSFC tape, had peak years: six in 1958 followed by four in 1967. In some years no tornadoes occurred within 144 miles from the Company. These years were 1950, 1952, 1954, 1956,1960, 1961,1963,1966,1968,1970,1973, and 1974 (see Tab!3 6).
Table 6. Frequencies of tornadoes within 144 milt,s from
. General Electric Company by year. Based on NSSFC tape, 1950-75.
Years 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 Frequencies 0 1 0 1 0 1 0 1 6 1 Years 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 Frequencies 0 0 1 0 2 1 0 4 0 1 Years 1970 1971 1972 1973 1974 1975 Frequencies 0 3 3 0 0 3 e
9 4-9 e ,
- 4. DESCRIPTION OF TORNADOES Three data sources--Report of the Chief of the Weather Bureau, Monthly Weather Review, and C1hnatological Data--were used in assessing tornadoes which were reported since 1916 Tornado frequencies by decade show a gradualincrease until the end of the 1940s. Thereafter, the frequencies by decade are more or less constant; about one tornado per year within a 140-mile range from General Electric Company (see Table 7).
For some reason, no tornado was reported in the 1940s within 140 miles from the Company, although 2 occurred elsewhere in California. Apparently, tornado activities were very low in California in that decade. .
Of 34 tornadoes during the 61-year period, 1916-76, 17 were rated as F 0, 10 as F1, and 7 as F 2 Two tornadoes, possibly three, originated as waterspouts off the Pacific Coast.
Two tornadoes, identified as Pacific Tornado and Merced 46 NE Tornado, occurred on the western slope of the Sierra Nevada at 4,000 to 5,000 ft MSL.
Most tornadoes occurred in San Joaquin Valey near Highway 99 which goes through high population areas between Sacramento and Fresno. It is likely that some rural tornadoes in the Valley are not reported to the National Weather Service.
Within the narrow coastal basin between parallel mountain ranges therb were reports of occasional tornadoes, especially in the cold season between November and February. Due to the high population density most Bay-area tornadoes were likely to have been reported (see Figure 3).
Shown in Table 8 is a list of 34 tornadoes, numbered according to the range from General Electric Company. An account of each tornado is also presented.
Table 7. Frequencies of, tornadoes within 140 m' ?s from General Electric Company by decade.
Decades 1916-19 1920s 1930s 1940s 1950s 19s 1970-76 Total F0 0 0 3 0 3 6 5 17 F1 0 2 0 0 3 2 3 10 F2 0 0 1 0 4 1 1 7 Total 0 2 4 0 10 9 9 34
4 - IO s
Table 8. List of tornadoes between 1916 and 1976 which were reported within 140 miles from General Electric Company.
Number Range Year - mo - Day F Length Name of tornado
~
1 20 mi 1951 11 2 3.0 mi SUNNYVALE l 2 21 1951 11 1 3.0 SAN JOSE 3 26 1969 16 0 0.1 MODESTO 20W 4 30 1958 01 1 0.3 S.F. AIRPORT 5 35 1971 24 0 0.3 STOCKTON l 6 39 1970 27 1 1.5 STOCKTON 7 40 1967 31 2 0.3 STOCKTON 8 42 1953 27 1 10 MODESTO 9 43 1965 01 0 0.3 CORRALITOS 10 47 1926 05 1 4.0 ESCALON 11 47 1933 16 0 0.1 LODI l 12 56 1958 01 2 2.0 TURLOCK 13 60 1933 16 2 0.3 DELHI 14 60 1936 26 0 0.2 SON 0mA COUNTY 15 66 1921 26 1 02 SACRAMENTO 16 67 1975 05 1 0.3 LOS BANOS 17 69 1967 22 0 0.1 mERCED 20SW 18 74 1933 16 0 0.2 CARmICHAEL ;
19 78 1958 10 2 15.0 SEBASTOPOL 20 82 1972 15 0 0.1 SACRAMENTO 11NE 21 84 1964 10 1 1.0 WINDSOR 22 90 1967 22 0 01 FRESNO 40W 23 99 1959 17 0 0.5 FT. ROSS 24 107 1967 21 1 0.2 MADERA 5SW 25 107 1958 12 2 05 MADERA 26 109 1972 01 0 01 PACIFIC 27 111 1972 15 0 0.1 MERCED 46NE 28 118 1975 05 0 03 FRESNO 29 119 1970 28 1 30- CRIDLEY 30 122 1958 03 1 1.0 CUALALA 31 126 1962 22 0 0.2 FRESNO 32 127 1957 19 0 1.0 FRESNO 33 127 1964 21 0 1.0 FRESNO 34 136 1975 13 2 0.3 F0WLER
- - -- -- e,- , ,,--e,w-w w ,rer ms +== wax-,
4 - II eFo GFI $F2 TORNADOES 90 29 gg0
\00 0 e26 6
21 18e 23 315 l9 6' 0 -
14 to $10 -
- 5
! 22 24 28 **33 e 3 3$ @1 4
/
Figure 3. Locations of tornadoes during the 61-year period, 1916-1976, within 140-mile range from General Electric Company. Concentric circles denote ranges 20, 40, 60, 80,100,120, and 140 miles, ,
4 - 12 l
1 NO. 1 SUNNYVALE TORNADO (Jan. 11, 1951). F PP = 2,1,4 l After damaging Los Altos residential area tornado left skipping l path eastward to El Camino Real. Then it went through Sunnyvale business district. Southern Pacific Railroad depot, Westinghouse and
\Voolridge plants were hard hit. This tornado was associated with a cold front with 78 (fastest mile 47 mph) gust at San Francisco Airport, 23 miles NW of the tornado. 1 NO.2 SAN JOSE TORNADO (Jan. 11,1951) F P P = 1,1,2 A tornado moved castward passing near City Hall (old location).
Roofs ripped off; trees forced down; structures damaged. .
NO.3 MODESTO 20W TORNADO (Cet. 16, 1969) F PP = 0,0,0 Pilot reported a spout in sight approximately 20 miles west of Modesto picking up dirt from ground. No report of wind received from observers on the ground.
NO.4 S. F. AIRPORT TORNADO (Apr.1,1958) F P P = 1,0,2 l Damage confined to narrow path across air base. Extreme wind lasted for a few seconds. Some structural damage to hangar and two other buildings. Several parked automobiles damaged by flying debris.
NO.5 STOCKTON TORNADO (Apr. 24, 1971) F P P = 0,0,0 Very small tornado west and northwest of Stockton caused minor damage to crops.
NO.6 STOCKTON TORNADO (Apr. 27, 1970) F P P = 1,1,1 ,
During a thunderstorm a funnel cloud touched down at three points, moving ESE. Trees were blown over, some of them damaging automobiles.
NO. 7 STOCKTON TORNADO (May 31, 1967) F P P = 2,0,2 A boathouse was lifted into the air and dropped on several boats; the roof was thrown across the channel, damaging other boats.
9 9
a ,
4 - 13
, '.i MODESTO TORNADO (Apr. 27, 1953) F P P = 1,1,1 I NO.8 A twiner dem oU.shed a 32 stanchion dairy barn, damaged other structures. '. 2d upror ad trees. Twister was erratic, passing over some districts without dantage. Power lines broke in 3 places. .
Nd. 9 CARRALITOS TORNADO (Apr.1,1965) F PP = 0,0,2 Newspaper reports minor damage to buildings and trees.
NO. 10 ESCALON TORNADO (Apr, 5,1926) F P P = 1,2,3 A tornado moved north causing much damage to buildings and fruit trees. The path was several miles long and 1/4 mile wide.
i NO. 11 LODI TORNADO (Mar. 16, 1933) F PP = 0,0,0 Garage shifted on foundation, instrument shelter overturned.
'l NO. 12 TURLOCK TORNADO (Apr.1,1958) F P P = 2,1,1 A tornado moved NE across a poultry farm. Demolished fences; l lifted 2 buildings. Deposited concrete piers and trusses on top of other I buildings. f l
i NO.13 DELHI TORNADO (Mar. 16,1933) F PP = 2,0,0 ;
House and garage demolished, fruit trees uprooted. Damage to
$2,000 )
NO. 14 SONOMA COUNTY TORNADO (Dec. 26, 1936) F PP = 0,0,0 Rural property damaged.
NO.15, SACRAMENTO TORNADO (Oct. 26, 1921) F P P = 1~,0,1 A tornado moved NE causing $17,000 property damage to roofs J and buildings under construction. Several persons were slightly injured. i NO. 16 LOS BANOS TORNADO (Apr. 5,1975) F P P = 1',0,1 Storm damaged several homes and sheds, uprooted trees ar.d scattered debris over a large area.
I
4 - I4
. g NO.17 MERCED 20 SW TORNADO (Apr. 22, 1967) F FP = 0,0,O Pilot reported sighting tornado.
\
NO. 18 CARMICHAEL TORNADO (Mar. 16, 1933) F PP = 0,0,0 Tornado-like winds uprooted rose bushes, torn blackberry ,
vin 6s,. property damaged. l NO.19 SEBASTOPOL TORNADO (Jan. 10, 1958) F PP = 2,3,2 A tornado moved from Bodega Bay on the Pacific Coast to between Sebastopol and Santa Rosa. Trees uprooted; houses and ranch buildings damaged; a few small outbuildings demolished completely.
'Ihe path extended ENE across 200 to 500 ft hills. l NO. 20 SACRAMENTO 11NE TORNADO (Oct. 15, 1972) F P P = 0,0,0 Tornado reported 11 miles northeast of Sacramento.
NO. 21 WINDSOR TORNADO (Nov. 10, 1964) F P P = 1,1,1 Small tornado damaged buildings, trees, billboards, and utility lines.
NO. 22 FRESNO 40W TORNADO (Apr. 22,1%7) F P P = 0,0,0 Pilot reported sighting tornado moving toward south.
NO. 23 FORT ROSS TORNADO (Feb. 17, 1959) F PP = 0,0,1 Small tornado moved NE. Shingles ripped from 12 x 54 ft area of roof. Fences destroyed and branches twisted off. Heavy rain and lightning.
NO. 24 MADERA SSW TORNADO (Apr. 21, 1967) F PP = 1,0,1 Funnel cloud obst.:ved from 5:00 to 5:30 PM. Cloud touched ground 5:30 ta 5:40 damaging small buildings. Storm moved to northeast.
NO. 25 MADERA TORNADO (Mar. 12, 1958) F P P = 2,0,0 A tornado moved NE across a golf course. A white funnel about 1,000 ft long was preceded by hall and rain. Heavy shake roofing ripped off; supporting posts ripped off concrete footings. A heavy bench carried into air and demolished.
., 4-l5
\
NO. 26 PACIFIC TORNADO (Oct.1,1972) F P P = 0,0,0 Tornado with hall accumulated to 3 inches on ground.
NO 27 MERCED 46NE TORNADO (Oct. 15, 1972) F P P = 0,0,0 A tornado reported 46 miles northeast of Merced.
NO. 28 FRESNO TORNADO (Apr. 5,1975) F PP = 0,0,3 Storm struck on western outskirts of Fresno in a thinly populated area. Minor damage to houses and sheds.
NO 29 ORIDLEY TORNADO (Jun. 28, 1970) F P P = 1,1,3 Small tornado skipped along over a path of about 3 miles.
Several hundred orchard trees were damaged or uprooted, and several barns suffered damage.
NO. 30 OUALALA TORNADO (Feb. 3,1958) F P P = 1,1,1 A waterspout landed and moved northward across Del Mar Ranch. 3 large cypress trees were uprooted. A cone-shaped funnel 150 to 200 ft high was followed by thunderstorm with hail.
NO. 31 FRESNO TORNADO (Mar. 22, 1962) F PP = 0,0,0 A funnel cloud and damage on a farm reported, No eyewitnesses.
NO. 32 FRESNO TORNADO (May 19, 1957) F P P = 0,1,0 Tornado first observed as a vertical funnel moved ENE with roaring noise. It ripped shingles from homes and uprooted Almond tree.
NO. 33 FRESNO TORNADO (Jan. 21, 1964) F P P = 0,1,0 Tornado moved NE through north part of Fresno, damaging several homes and minor structure.
NO. 34 FOWLER TORNADO (Mar. 13, 1975) F P P = 2,0,0 A steel shed weighing over one ton lifted over 200 ft and dropped 100 yds away. Roofing and rafters torn. Pilot reported tornado but no ground sighting of funnel due to rain.
4-16 5 TORNADO PROBABILITIES The NSSFC tape reveals that the mean damage area of the 22 tornadoes with known. path characteristics is 0.01 sq. mile per tornado. 'Ihe statistics are based an the 1950-75 tornadoes within a 144-mile radius of the General Electric Company.
Applying this mean tornado area to the total. number of 29 tornadoes, we estimate the total damage area to be 0.01 x 29 = 0.29 sq. mile in 26 years. 1
'Ihe area within the 144-mile radius from the Company, excluding lake areas, is 44,300 sq. miles. The tornado probability from these data is tornado area ,
0.29 = 2. 52 x 10*7 per year.
land area x years 44,300 x 26 i
Although the definition of the tornado area is rather vague, it probably represents the area affected by F0 to F1 or stronger winds.
I The DAPPLE (Damage Area Per Path Length) METHOD developed by Abbey and Fujita (1975)is capable of computing tornado probabilities as a function of the F-scale damage categories, which can be converted into windspeeds (see Table 9).
Using the DAPPLE METHOD the area of specific windspeed can be computed by the product Windspeed area = Path length x DAPPLE where path length denotes that of specific intensity tornado.
Table 9. Range of F-scale windspeeds and their weighted mean values. (Refer to Abbey, Robert (1977): Risk probabilities associated with tornado windspeeds. proc.
of Symp. on Tornadoes, Assessment of knowledge and im-plications for men.)
F scale F0 F1 F2 F3 F4 F5 Range of windspeed 40-72 73-112 113-157 158-206 207-260 261-318 mph Weighted mean speed 59 92 131 177 227 276 mph
4-I7 Since F-scale assessments assume an accuracy of or.a scale, tornadoes are classified into three categories (instead of six) for' DAPPLE computation purposes.
These categories are VIOLENT (F 4 and F 5), STRCMG (F 2 and F 3), and WEAK (F0 and F 1). DAPPLE values were then obtained based on 147 tornadoes of April 3-4,1974 Superoutbreak (see Table 10).
No DAPPLE values have been computed for the California tornadoes, because _
practically no survey data for DAPPLE are available. The use of Table 1 will, therefore, overestimate the damage areas because Midwestern tornadoes are usually larger than California tornadoes. ,
l l
I Table 10. 0 APPLE in miles as a function of tornada windspeed.
i Values were obtained from 147 tornadoes of April 3-4, 1974 Superoutbreak. I Windspeeds WEAK (F0&F'1) STRONO (F2&F3) 50 mph 0.074 mile 0.43 mile 100 0.00'28 0.062 150 0.000052 0.0098 200 0.000000 0.0012 '
250 0.000000 0.000087 We have to realize, on the other hand, that California tornadoes have not been fully reported and assessed. 'Ihe reasons being the weakness in tornado intensity, relatively low public awareness, etc. 'Ihe use of Midwestern DAPPLE may as well offset the low reporting efficiency, resulting J' a reasonable estimate of damage areas.
4 - ja l .
Path lengths of the two-category tornadoes, strong (F 2 & 3) and weak (F0 & 1), were obtained from Table 8 The measurements were made for each of the annular' rings, 20 miles apart. The cumulative path lengths were, then, obtained to determine the total path length within each radius from the Company site (see Table 11).
Table 11. The total path mileages of toinadoes within seven ranges from General Electric Company.
WEAK (F0 and F1) TORNADOES Ranges 20 40 60 80 100 120 140 miles 1916-49 0.0 0.0 4.6 50 50 5.0 5 0 mi 1950-76 0.0 52 6.5 6.9 8.3 12.0 15.2 Total 0.0 5.2 11.1 11.9 13.3 17.0 20.2 STRONO (F2) TORNADOES Rangen.. 20 40 60 80 100 120 140 miles 1916-49 0.0 0.0 0.3 0.3 0.3 0.3 0.3 ml 1950-76 3.0 33 5.3 20.3 20.3 20.8' 21.1 Total 30 3.3 5.6 20.6 20.6 21.1 21 4 l
t
4 - l9 The DAPPLE values and path lengths can be used in computing the damage area for each windspeed. The damage area .of a 100 mph windspeed, for instance, is computed as A ioo = Path length of strong tornado x 0.062
+ Path length of weak tornado x 0.0028 where A ooi is the area of 100 mph or stronger windspeed (see Table 12).
Table 12. Damage areas within seven ranges from General Electric Company. Areas were obtained by multiplying the path mileage in Table 11 by Dapple in Table 10.
A 27 year period, 1950-76 was used because data in earlier years are, apparently, not representative.
Range Land area Windspeed (damage) area in sq. mi.
(miles) (sq.mi.) 50 mph 100 mph 150 mph 200 mph 20 1,100 1.290 0.186 0.029 0.000 40 4,600 1 804 0.220 0.032 0.000 60 9,100 2.760 0.347 0.052 0.006 80, 15,100 9 240 1.278 0 199 0.'024 100 22,300 9.343 1.282 0 199 0.024 120 31,000 9.832 1.323 0.205 0.025 140 41,000 10.198 1.351 0 208 0.025
~~
t
4- 20 l , ,
The tornado probabilities were computed based on the 27-year period, 1950-76
'Ihe formula used is ,
Damage area of specific vdndspeed Land area x 27 years In computing probabilities windspeeds were varied: 50,100,150 and 200 mph.
Meanwhile, the ranges of probability computations were varied between 20 and 140 miles at 20-mile intervals (see Table 13).
It is a difficult question to select a proper range of probability computations.
'Iheoretically, the range must be small enough to preserve characteristics of the Company site. If we choose an excessive radius, the area could include Sierra -
Nevada which is entirely different from the Company site.
If we choose a 20- to 40-mile range, the number of tornadoes is so small that a statistical assessment will not be feasible.
In view of these difficulties, a 100-mile range was selected for assessment of the tornado risk applicable to the General Electric Company. Note that this area is considerably smaller than the 5-degree boxes used in WASH-1300 Table 13. Probabilities of tornado windspeed within 7 ranges from General Electric Company. Based on 27-year data, 1950-76.
Range Probabilities (miles) 50 mph 100 mph 150 mph 200 mph 20 4 34 x 10~' 6 26 x 10-' 9.76 x 10-7 0.00 40 1.45 x 10-8 1.77 x 10" 2.58 x 10-7 0.00 60 1 12 x 10** 1 41 x 10-* 2.11 x 10**
2.44 x 10
80 2 26 x 10 3.13 x 10** 4.88 x 10 5.89 x 10**
100 1 55 x 10 2 13 x 10-* 3.31 x 10*7 3.99 x 10
120 1.17 x 10 1.58 x 10-* 2 45 x 10-7 2 99 x 10**
140 9.21 x 10** 1 22 x 10~' 1.88 x 10~7 2 26 x 10
- 4-2I l' .
6
SUMMARY
AND CONCLUSIONS Results of the foregoing computations of wi" Speed probabilities are summa-rized in Figure 4, which includes three curves:
A. Prd) ability of fastest-mile year B. Probability of peak gust assumed 1.25 times the fastest-mile speed C. Tornado probability based on the 27-year data
'Ihe figure reveals that the speeds of straight-line winds are higher than tornado winds when the probability is greater than about: 10-' per year.
Table 14 gives the maximum windspeeds corresponding to the return periods of :ne to 10 million years. It is obvious that tornadoes become important when the probability decreases below one in 1,000,000 years.
Table 14.
Maximum windspeeds expected at the General Electric Company site as a function of the probability per year. (A)... fastest-mile speeds of straight-line -
winde. (B)... gust speed computed as 125% of A. (C)...
tornado windspeeds based on 1950-74 data.
Probabilities Return periods Windspeeds in miles per hour (A) (B) (C) 10' 1 year 38 47 -
10" per year 10 47 59 -
10 100 53 66 -
10-8 1,000 59 74 -
10" 10,000 66 83 -
10'8 100,000 72 90 -
10 1,000,000 - - 122 10 10,000,000 - - 178
, c
- ... 4-22 e
1 1
l The three-category probabilities used in this site analysis are
. High probability . . . . . . . . 10-3 per year ;
Low probability . . . . . . . . 10"* per year Remote probability . . . . . . '10-7 per year .
It should be noted that 10-7 per year is used in determining the design-basis tornadoes for nuclear power plants in WASH-1300 For secondary or short-life structures, one may wish to use maximum wind-speeds corresponding to the low- or even high-probability category. .
The storm characteristics in Table 15 were computed for these three proba-I bilities. If one wishes to protect a structure against the high probability (10~3 ) I 1
storm, a 74-mph gust should be used as the design criteria.
Table 15. Characteristics of storms corresponding to three (10"*),probability (categories, and high 10-3 per year).remoto (10-7 ),
Air density at low the Company site, 510-ft MSL, is assumed 1 2 kg/m 3 and
- radius of maximum wind, 100m for computational purposes.
Probabilities (per year) Remote (10-') Low (10-') High (10-s )
Storm types Tornado Tornado Straight wind maximum total speed (mph), b 178 122 74 gust Translational speed (mph),T 36 24 -
Tangential speed (mph) 142 98 -
Total press. drop (mb) 48.3 23.0 -
(psi) 0.70 0 33 -
Rate of pr. drop (mb/sec) 7.8 25 -
(psi /sec) 0.11 0 04 .
4-23 200 250 mph lo, 0 _ 59 10 0 15 0
~
10,
$ \ STRAIGHT-LINE WINDS 1
48 STATES PEAK GUST from l o
' O 1955 -72) y 1 /(SELS LOG E
E cn 10" j c E
- E 10" A\
'3 L
to--
\ \ .
io"
\C TORNADOES
\ '00
<9, '4, 7
lo" %
lo-'
N Figure 4. Probability of straight-line and tornado winds, ne probability of straight-line gust is higher for windspeeds less than 100 mph. Above 100 mph, the probability of tornado wind exceeds that of straight-line winds.
C
\
..~.
- , 4 - 24 e e
'Ihe translational speed, T, was assumed as T = 0. 20 V., ,
where V e is the maximum total speed along the radius of maximum wind, r. .
Under the, assumption of the combined Rankine vortex and cyclostrophic wind equation, the total pressure drop was computed from P, = p V.a ,
where p at the 510-ft MSL Company site was assumed to be 1.2 kg/m3 . !
The rate of pressure drop was computed from dP , p V.' and dr = Tdt dr r.
or dP T gy a l
dt r.
which is identical to Equation (3 ) of WASH-1300 It is recommended that the team of structure analysts determine the proper probability category to be applied to each structure or portions of structure of the General Electric Company. Table 15 will, then, be used to determine the charac-teristics of the design-basis storm.
l V
T. Theodore Fujita v
'1 l
j