Seismic risk analysis using Bayesian belief networks - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 196

3 2000

2.5

1975

2

1950

1. 5

1925

1

1900

0.5

0 1875

(a) (c)

34

35 35

Ye s
No
Unknown

Ye s
No
Unknown

30 30

26

25

25

20

20

15

15

10 10

7

4

3

5 5

2

1 1

1 1

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0

0

AT C ‐ 13 damage state

AT C ‐ 13 damage state

(b) (d)

7.14 Seismic hazard and performance modifi ers for the decrease in
demand: (a) spectral acceleration; (b) lateral ground movement; (c)
year of construction; (d) deterioration from previous earthquake.

7.5.5 Damage quantifi cation using BBN

The aggregation through the hierarchical structure is illustrated below. The
illustration is carried out for a Building ID = CDMG087-CG-213, with the

number of stories N 3. During the 1994 Northridge earthquake, the build-

ing was identifi ed as damage state of 'Moderate'. From the walk down
survey, the building performance modifi ers summarized in Table 7.10.
The intermediate step will be computed to obtain the value for building
vulnerability. Finally, selecting the values for SSH

=

0.769 g, the building
damage is computed. Snapshots of the BBN results at each step are shown
in Fig. 7.15. Results in Fig. 7.15 show that, the probabilities associated with
ID, DC, BV and BD are:

=

• Increase in demand, ID: (ID VL , ID L , ID M , ID H , and ID VH ) = (0.20, 0.20,

0.20, 0.20, 0.20).