Table D-1 Examples of Similar Damage from Straight Winds, Hurricanes, and Tornadoes

DOE-STD-1020-2002

Table D-1 Examples of Similar Damage from Straight Winds, Hurricanes, and Tornadoes

Type of Damage

Winds

Hurricanes

Tornadoes

Windward wall collapses

Mobile home, Big Spring,

A-frame, Hurricane Diana

Metal building, Lubbock

Inward

Texas 1973

1984

Texas 1970

Leeward wall or side wall

Warehouse, Big Spring,

Commercial building,

Warehouse, Lubbock,

collapses outward

Texas 1973

Hurricane Celia 1970

Texas 1970

Roof

Warehouse, Joplin,

Motel, Hurricane Frederick

School, Wichita Falls,

Missouri 1973

1979

Texas 1979

Eaves

Mobile home, Big Spring,

A-frame, Hurricane Diana

Metal building, Lubbock,

Texas 1973

1984

Texas 1970

Roof corners

Residence, Irvine,

Residence, Hurricane

Apartment building,

California 1977

Frederick 1979

Omaha, Nebraska 1975

Wall corners

Metal building, Irvine,

Flagship Motel, Hurricane

Manufacturing building,

California 1977

Alicia 1983

Wichita Falls, Texas 1979

Internal pressure

Not applicable

Two-story office building,

High School, Xenia, Ohio

Cyclone Tracey, Darwin,

1974

Australia 1974

A somewhat arbitrary, but quantitative approach is used to determine if a particular DOE

site should be designed for tornadoes. Hazard assessments for both straight winds and tornadoes

for each DOE site are presented in Reference D-2. The intersection of the straight wind and

tornado hazard curves determines if tornadoes should be included in the design and evaluation

criteria. If the exceedance probability at the intersection is greater than or equal to 2x10-5,

tornadoes are a viable threat at the site. If the exceedance probability is less than 2x10-5,

straight winds control the design or evaluation criteria. The concept is illustrated in Figure D-1

(using fastest mile basis for illustration). Straight wind and tornado hazard curves are shown for

Oak Ridge National Laboratory (ORNL) and Stanford Linear Accelerator Center (SLAC). The

SLAC curves intersect at an exceedance probability of approximately 2x10-7, indicating that

tornadoes are not a viable threat at the California site. On the other hand, the intersection of the

ORNL curves is at 6x10-5 suggesting that tornadoes should be included in the design and

evaluation criteria. Design wind speeds for the 25 DOE project sites were selected on this basis.

However, this methodology is now superceeded by adopting latest information for straight winds

and hurricanes in ASCE 7-98. However, at few sites previous site specific data still governs.

For tornadoes, assessment is to be made by new methodology developed at LLNL (Reference D-

13).

D.3

Load Combinations

The ratios of hazard probabilities to performance goal probabilities (risk reduction factor)

for the Performance Categories in Table D-2 are an approximate measure of the conservatism

D-3