CE 5603 SEISMIC HAZARD ASSESSMENT

LECTURE 1: INTRODUCTION

By : Prof. Dr. K. nder etin

Middle East Technical University
Civil Engineering Department

Course Outline

2/47

Course Outline

3/47

Why Assess Seismic Hazard Levels ?

"Every once in a while, something bad happens as a result of an earthquake, and
probabilistic seismic hazard analysis is the basis on which one reckons how often
bad happens at some place of interest" (Cornell)

Seismic Hazard vs. Seismic Risk

"In 1910, the Seismological Society of America identified the three parts of the
earthqake problem that merit study: the event itself (when, where and how
earthquakes occur), the associated ground motions, and the effects on structures.
These are still the fundamental elements in evaluating earthquake risk." (McGuire,
2004)
"Seismic hazard analysis is the foundation of seismic risk and/or seismically
induced failure analysis of facilities; therefore seismic hazard analysis can be used
to determine risk-based seismic design and construction practice. Seismic hazard
analysis can also be developed and used as an end in itself, generally in the form
of seismic hazard maps." (Cornell)

Seismic Hazard vs. Seismic Risk

More specifically, "risk" is identified as the product of "hazard" and "vulnerability".
Along with more formal definitions of "hazard" and "risk", a clear distinction
between the frequently confused words in terminology can be made using specific
examples:

Seismic Hazard vs. Seismic Risk

If an engineer is utilizing a building code in his/her design, the seismic coefficients
and the design spectrum to be used are the results of a "hazard" analysis, i.e. it
categorizes the zones in a specific area prone to earthquake effects. A hazard map
doesn't define whether your structure to be built is safe or not against failure, nor it
tells the order of expected damage.

Seismic Hazard vs. Seismic Risk

Imagine a residence, built in accordance with seismic regulations, at a distance of
5 km from the active fault. It'll potentially be prone to higher "hazard" than a similar
structure located at 50 km away from the fault. However, this does not necessarily
mean that the residence located far away is under less "risk". Given the conditions
that the latter is very poorly constructed, it will probably be under higher "risk",
although located in a zone of lighter "hazard". The same conditions apply vice
versa.

Seismic Hazard vs. Seismic Risk

Imagine two identically constructed (same structural system, same construction
quality) structures at adjacent lots, having the same in-situ conditions. Building A
inhabits residents, however Building B is the headquarters of an international
software company. Two buildings are under the same "hazard" level, and expected
to undergo similar light damage, however the calculated annual income loss due to
temporary repair and operational loss at Building B makes it susceptible to higher
"risk". Similar examples are valid for other types of structures such as tunnels,
dams and transportation infrastructure.

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Supporters of probabilistic and deterministic philosophies in seismic hazard
analyses have been almost continuously debating, sometimes ironically, on the
proper selection of hazard levels and calculation of ground motions, since the
probabilistic methodology first sprung back in late 1960's. Here is a pinch from two
opposite views:

10

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Probabilistic View Supported
"The deterministic approach to seismic hazard analysis DSHA, seems to be very
different animal from PSHA. DSHA deals with fascinating !!! things like MCE
(Maximum Credible Earthquake), or SSE (Safe Shutdown Earthquake), or OBE
(Operating Basis Earthquake) terminology plied in the large dams nuclear reactor
trades.
In the heyday of DSHA, back in 1960's, and 1970's when there were still
acknowledged wise men in the Earth and Engineering Science related to
earthquakes, one or may be a few of these people would decide the
MCE/SSE/OBE's and where they would be likely to occur. A little cookbook ground
motion estimation would then ensure and bingo!- seismic design criteria. This
doesn't sound like much serious things like nuclear reactors, and people have
been on the lookout for something better ever since. This, of course is PSHA,
essentially the only other game in town."
11

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Probabilistic View Supported

12

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Almost Pure Determinisim: Krinitzsky's Position
Ellis Krinitzsky, a core supporter of deterministic methods in seismic hazard
analysis defends his opinions with the following main items listed:

"Earthquakes do not occur randomly in space and time"

"Using Poissons models is wrong (lack of memory)"
"The magnitude recurrence relation is so uncertain that it becomes meaningless"
"Probabilitic seismic assessment is a bad science"
"Only deterministic approaches (maximum earthquake) are reasonable"
Krinitzsky's and other supporters' deterministic point of view in depth can be
accessed from numerous publications, where a few are listed here as suggested
references:

Krinitzsky (1993a), Krinitzsky (1993b), Krinitzsky (1993c), Krinitzsky (2002a),

Krinitzsky (2002b), Mualchin and Krinitzsky (2003), Hatheway (2003), Castaos
13
and Lomnitz (2002).

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Almost Pure Determinisim: Krinitzsky's Position

14

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Keeping a Balance Between
Criticism on probabilistic seismic hazard analysis is primarily focused on points
such as misuse of probabilistic tools, lack of adequate data for proper handling of
uncertainties and ambiguities in applying expert opinions. However, as Bommer
(2003) contributes, the debate regarding the relative merits and shortcomings of
deterministic and probabilistic approaches are likely to continue for many years. A
big step ahead taken towards a healthier debate will probably be an agreement on
the clear definitions of hazard assessment concepts and terms.

15

Probabilistic vs. Deterministic

Seismic Hazard Assessment
Keeping a Balance Between
More fair criticism towards probabilistic seismic hazard assessment can be made
upon identifying the true source of problematic issues. Probabilistic methodology
provides the neccessary tools for handling the uncertainties associated with the
earthquake and seismic hazard assessment phenomenon. However, only proper
use of these available tools will lead to an acceptable and sound analysis
workflow.
Specific examples for most of the criticism towards probabilistic methods can be
given as; misinterpretation of magnitude recurrence relations of earthquakes
based on instrumental data obtained from records aged slightly over a century,
misuse of logic trees for combining results from analyses having different distance,
magnitude definitions, and several others. While some of these issues are
currently under research to come up with more representative analysis methods,
an inevitable amount that can not be overlooked has to do with building up a
common terminology that bridges probabilistic and deterministic methodologies.
McGuire (2001), also shows very good demonstrations of complementary roles of
16
deterministic and probabilistic methods for decision making purposes.

Probabilistic vs. Deterministic

Seismic Hazard Assessment
References
McGuire, R. K., (2004). Seismic Hazard and Risk Analysis, EERI Monograph
Series, No. 10, Earthquake Engineering Research Institute.
Cornell's monograph supply full reference details (also not sure about the
reference of paragraphs in quotes)
Krinitzsky, E. L., (1993a). Earthquake probability in engineering Part I: The use
and misuse of expert opinion, Engineering Geology, 33, 257-288.
Krinitzsky, E. L., (1993b). Earthquake probability in engineering Part II:
Earthquake recurrence and limitations of Gutenberg-Richter b-values for the
engineering of critical structures, Engineering Geology, 36, 1-52.
Krinitzsky, E. L., (1993c). The Hazard in Using Probabilistic Seismic Hazard
Analysis, Civil Engineering, November, 60-61.
Krinitzsky, E. L., (2002a). How to obtain earthquake ground motions for
engineering design, Engineering Geology, 65, 1-16.
Krinitzsky, E. L., (2002b). Epistemic and aleatory uncertainty: a new shtick for
probabilistic seismic hazard analysis, Engineering Geology, 66, 157-159.
17
Engineering, 21, 377-384.

Probabilistic vs. Deterministic

Seismic Hazard Assessment
References
Mualchin, L., Krinitizsky, E. L., (2003). A new and defective regulation in California
for protecting critical buildings from earthquakes, Engineering Geology, 69, 415419.
Hatheway, A. W., (2003). How to obtain earthquake ground motions for
engineering design, 67, 403-404.
Castaos, H., Lomnitz, C. (2002). PSHA: is it science?, Engineering Geology, 66,
315-317.
Bommer, J. J., (2003). Uncertainty about the uncertainty in seismic hazard
analysis, Engineering Geology, 70, 165-168.
McGuire, R. K., (2001). Deterministic vs. probabilistic earthquake hazards and
risks, Soil Dynamics and Earthquake Engineering, 21, 377-384.

18