Probabilistic seismic hazard analysis of civil infrastructure - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 5

due to earthquakes (Cetin et al. , 2009), can be incorporated into the assess-
ment to quantify the extent of geotechnical consequences probabilistically
(similar extension from PSHA to PSRA). A key fact is that PSHA produces
the fundamental input information (i.e. seismic hazard curve or earthquake
scenarios) to such advanced analyses.
This chapter gives an overview of the simulation-based PSHA methodol-
ogy by summarising a step-by-step analysis procedure and current develop-
ments/issues in this area. The use of the simulation-based approach is
highlighted, because it is versatile and extendable to facilitate other
advanced earthquake engineering applications. This is particularly impor-
tant in the context of the PBEE methodology. Then, two example applica-
tions/illustrations will be discussed: one is related to liquefaction hazard
analysis, while the other is related to seismic risk analysis. The chapter con-
cludes by mentioning future trends and suggested improvements in PSHA.

1.2 Simulation-based probabilistic seismic hazard

analysis (PSHA)

1.2.1 Methodology
The main objective of this section is to introduce a basic PSHA methodol-
ogy to a broader audience who is not familiar with PSHA. In particular, an
emphasis is given to a Monte Carlo simulation approach. To facilitate this,
a case study for western Canada is used as an illustration in the following.
PSHA evaluates the expected seismic intensities corresponding to differ-
ent probability levels. A rational approach to achieve this is to take all
possible scenarios into account, and sum the effects in the assessment by
refl ecting their relative contributions. Based on Total Probability Theorem
(and some assumptions/simplifi cations), the overall integration/summation
process can be decomposed into several modules. Mathematically, the
annual rate that a specifi c ground motion level is exceeded ν GM ( ≥ gm ) is

given by:

⎡ ⎤

p q

∑
model

∑

( ) =

∫∫

( ) ( )

ν ≥ gm w λ P GM ≥ gm m r f , m r , dd m r ,

⎢
⎢

⎥
⎥

GM i Mij , MRij , ,

⎣ ⎦

, i = 1 zone , j = 1

Ω mr , ij

[1.1]

where the subscript indices i and j are for the model/assumption (1 to p )
and for the source zone (1 to q ), respectively; w i is the weight of a model/
assumption i (note: w i sums to 1); λ M , ij is the annual occurrence rate of

potentially damaging earthquakes (e.g. events with magnitudes greater than
4.5); P ( GM ≥ gm | m , r ) is the probability of ground motion parameter