Performance-based assessment and

design of buildings
(the PEER approach)
LESSLOSS Final Workshop
Risk Mitigation for Earthquakes and
Landslides
Helmut Krawinkler
Stanford University, California, USA

19-20 July 2007

Hotel Villa Carlotta, Belgirate (VB) - Italy
Priority 1.1.6.3 - Global Change and Ecosystems / European Integrated Project GOCE-CT-2003-505488
19-20 July 2007 Belgirate,
Italy

LESSLOSS Final Workshop

Krawinkler

Measures of
Performance - PBEE
Forces and deformation?

Yes, but only for engineering calculations

Intermediate variables
Not for communication with clients and community

Communication in terms of the three Ds:

Dollars (direct economic loss)

Downtime (loss of operation/occupancy)
Death (injuries, fatalities, collapse)

Quantification

Losses for a given shaking intensity

Losses for a specific scenario (M & R)
Annualized losses
With or without rigorous consideration of uncertainties

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Vision of PBEE
1.

Complete simulation

2.

Defined performance
objectives

Joes
Beer!
Food!
3.

Quantifiable
performance
targets

Annual
probabilities of
achieving them

Informed owners

Joes

Joes
Beer!
Food!

Beer!
Food!

Sources: G. Deierlein, R. Hamburger

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Evolution of PBEE
O
P
E
N

Base
Shear

O
P
E
N

O
P
E
N

Collapse
Onset

Damage
Threshold

Deformation

PBEE yesterday
PBEE today
0

0.0

IO

LS

25%
0.0001

1
7
0
19-20 July 2007 Belgirate, Italy

50%
0.001
30

CP

FEMA 356 Performance Levels

100%
0.01

0.25

180
LESSLOSS Final Workshop

$, % replacement
Casualty risk
Downtime, days
Krawinkler

The PEER framework

equation - 1999

Curse?

Blessing

v (DV ) = G DV DM | dG DM EDP | dG EDP IM | d ( IM )

Impact

Performance (Loss) Models and Simulation

19-20 July 2007 Belgirate,

Italy

LESSLOSS Final Workshop

Hazard
Krawinkler

PBEE Probability
Framework Equation
v (DV ) = G DV DM | dG DM EDP | dG EDP IM | d ( IM )
Impact

Performance (Loss) Models and Simulation

Hazard

IM Intensity Measure
EDP Engineering Demand Parameter
DM Damage Measure
DV Decision Variable
(DV) Probabilistic Description of Decision Variable
(e.g., Mean Annual Probability $ Loss > 50% Replacement Cost)
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Performance-Based
Methodology
Collapse & Casualties

Decision
DecisionVariable
Variable

Direct Financial Loss

Downtime

Damage
DamageMeasure
Measure

Engineering
Engineering
Demand
DemandParameter
Parameter

drift as an EDP

Source: G. Deierlein
19-20 July 2007 Belgirate, Italy

Intensity
IntensityMeasure
Measure
LESSLOSS Final Workshop

Krawinkler

Performance-Based
Methodology

Intensity
IntensityMeasure
Measure
19-20 July 2007 Belgirate, Italy

Mean Annual Freq. of Exceedance, Sa

Engineering
Engineering
Demand
DemandParameter
Parameter
MEAN SPECTRAL ACC. HAZARD CURVE -- T = 1.8 sec.
Van Nuys, CA, Horizontal Component

10
1
0.1
0.01
0.001

Medina & Krawinkler

0.0001
0

0.2

0.8 Workshop
1.0
LESSLOSS Final

0.4
0.6
Spectral Acceleration Sa(g)

Krawinkler

IM (e.g., Sa(T1))

Incremental Dynamic
Analysis

Individual records
Median
84%

IM Hazard curve
(annual freq. of exceedance)

EDP (e.g., max. interstory drift)

EDP ( y ) = P [EDP y | IM = x ] | d IM ( x ) |
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Performance-Based
Methodology
Performance Assessment types (ATC-58 definitions):

Decision
DecisionVariable
Variable

Intensity-based: Prob. performance, given intensity of ground motion

Scenario-based: Prob. performance, given a specific eq. scenario
Time-based:
Prob. performance In a specific period of time
Damage Fragility Curves:

P($>x | DM)

1.0

Drywall partitions

E[Loss | EDP]
1.4

Drywall Partitions with Metal Frame

1.0

1.2

0.8

Damage
DamageMeasure
Measure

Mean Loss Curve:

Cost Functions:

Drywall Partitions with Metal Frame

P(DM>dm | EDP)

0.8

1.0

0.6
0.4
DM 1

0.2

0.6

0.8

0.4

0.6
Tape, Paste & Repaint
Replacement of gypsum boards
Partition replacement

0.2

DM 2
DM 3

0.0
0.000

0.0

0.005

0.010

0.015

0.020

0.0

EDP (IDR)

0.4

0.8
1.2
1.6
Cost of Repair / Cost New

2.0

2.4

0.4
0.2
0.0
0.000

0.005

0.010

0.015

0.020

0.025

0.030

EDP (IDR)

Aslani & Miranda

Intensity
IntensityMeasure
Measure
19-20 July 2007 Belgirate, Italy

Mean Annual Freq. of Exceedance, Sa

Engineering
Engineering
Demand
DemandParameter
Parameter
MEAN SPECTRAL ACC. HAZARD CURVE -- T = 1.8 sec.
Van Nuys, CA, Horizontal Component

10
1
0.1
0.01
0.001

Medina & Krawinkler

0.0001
0

0.2

0.8 Workshop
1.0
LESSLOSS Final

0.4
0.6
Spectral Acceleration Sa(g)

Krawinkler

Testbed Building

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

De-aggregation of
expected annual loss
Example: Van Nuys Testbed Building

Structural
12%

Collapse
29%

Non-collapse
71%

Non-structural
88%

Source: E. Miranda
19-20 July 2007 Belgirate,
Italy

LESSLOSS Final Workshop

Krawinkler

Design Decision Support

Hazard
Domain

Structural System Domain

Mean IM-EDP Curves

Mean Hazard Curve

EDP = Max. Interstory Drift

10/50

EDP = Max. Floor Acceleration

Loss Domain
Expected $Loss

Mean Subsystem Loss Curves

NSDSS
EDP = Max. Interstory Drift

19-20 July 2007 Belgirate, Italy

NSASS
EDP = Max. Floor Acceleration

LESSLOSS Final Workshop

Zareian & Krawinkler (2005)

Krawinkler

Assessment of
Collapse Potential
NORM. STRENGTH VS. MAX. STORY DUCT.

N=9, T1=0.9, =0.05, =0.03, =0.015, H 3, BH, K 1, S1, NR94nya

20

[Sa(T1)/g] /

Vy

Non-degrading system
Degrading system

15
10

Collapse
Capacity

5
0
0

19-20 July 2007 Belgirate, Italy

10

si,max

15

LESSLOSS Final Workshop

20

Krawinkler

Modeling of
Deterioration

UCI G12 OSB

UCI G12 OSB

Fy=8.2 kips, y=0.45 in, s=0.047, c=-0.081, u=1.94, c/y=5.44

Pinching Model, =0.5, Fy=8.2 kips, y=0.45 in

10
8
6
4
2
0
-2
-4
-6
-8
-10

Load (kips)

Load (kips)

s=0.047, c=-0.081, c=1.94, c/y=5.44, s=270, c=270, k=, a=270

-4

-2

10
8
6
4
2
0
-2
-4
-6
-8
-10
-4

-2

Displacement (in)

19-20 July 2007 Belgirate, Italy

Displacement (in)

LESSLOSS Final Workshop

Krawinkler

Collapse Capacity for a Set

of Ground Motions
M AX . STO RY DUCTILITY vs. NO RM . STRENG TH
N=9, T 1 =0.9, =0.05, K 1 , S 1 , BH , =0.015, Peak-O riented M odel,
s =0.05, c / y =4, c =-0.10, s =8 , c =8 , k =8 , a =8 , =0, LM SR

10

[Sa(T1)/g]/

8
6
4
2

Individual responses

0
0

10
20
M axim um Story Ductility O ver the H eight, s,m ax

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

30

Krawinkler

Collapse Fragility
Curve
Obtaining the collapse fragility curve (MRF)
N = 8, T1 = 1.2, = 0.17, Stiff & Str = Shear, SCB = 2.4-2.4, = 0.05
p = 0.03, pc/p = 5, = 20, Mc/My = 1.1

Probability of Collapse

0.75

0.5

Data points

0.25

Collapse fragility curve

0
0

0.5

1.5

2.5

3.5

4.5

IM[Sa(T1)]

19-20 July 2007 Belgirate, Italy

Zareian & Krawinkler (2004)

LESSLOSS Final Workshop

Krawinkler

Implementation of
Methodology

ATC-58 Guidelines for Seismic

Performance Assessment of
Buildings
ATC-63 Recommended
Methodology for Quantification of
Building System Performance
TBI Tall Building Initiative
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

ATC-58
Seismic Performance
Assessment of Bldgs.

Present emphasis is on damage

assessment (PACT program)
Next stage will be collapse and
casualty assessment
Next step will be downtime
Final step will be putting the pieces
together
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Importance of Damage
State Fragility Functions
P(DM|EPD)
1.0

Small
cracks
only
Small
cracks
only

0.8
0.6

5/8" Gypsum partition wall with 3-5/8" Wall

Frame

Severe damage to
gypsum board and
distorsion of metal frame

0.4
0.2

Wide cracks in gypsum boards

0.0
0

0.005

0.01

0.015

0.02

0.025

EPD (IDR)

Source: E. Miranda
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

ATC-58 relies heavily

on Fragility specs.

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Damage State 0

Probability of Requiring a MOR

Fragility Functions for

Beam-Column Joint

Damage State 2

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 0.0

P[DM:EDP]
MOR 0
MOR 1
MOR 2
MOR 3
MOR 4
1.0

2.0

3.0
4.0
Drift (%)

5.0

6.0

Example of compilations for:

RC Beam-Columns, Joints, Walls
Interior Partitions
Laboratory benches & equipment

Damage State 4

19-20 July 2007 Belgirate, Italy

Ceiling & MEP Systems

LESSLOSS Final Workshop

Krawinkler

Example Application

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Example Application
PACT Results

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

ATC-63 Project
Objectives
Primary Create a methodology for
determining Seismic Performance Factors
(R-factor, Cd-factor, overstrength factor)
for different lateral-force-resisting
systems
Secondary Evaluate a sufficient number
of different lateral-force-resisting systems
to provide a basis for Seismic Code
committees to develop more rational
Seismic Performance Factors that will
more reliably achieve the inherent
earthquake safety performance objectives
of building codes
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Example IDA Results and

Collapse Margin Ratio

(4-story, SDC D, space frame with 30-foot bays)

Sa(T=0.81s) [g]

MC = 2.5
(2.77/1.11)

Median Collapse SCT = 2.77g

0
0

MCE SMT = 1.11g

0.05

0.1

0.15

Maximum Interstory Drift Ratio

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Source: G. Deierlein
Krawinkler

Collapse Fragility with

Modeling Uncertainty
1

Cummulative Probability of Collapse

0.9

MC

0.8
0.7
0.6
0.5
0.4
0.3
0.2

Empirical CDF
Lognormal CDF (RTR Var.)
Lognormal CDF (RTR + Modeling Var.)

0.1
0
0

0.5

1.5

Sa

19-20 July 2007 Belgirate, Italy

2.5
g.m.

3.5

4.5

(T=1.0s) [g]

LESSLOSS Final Workshop

Source: G. Deierlein

Krawinkler

Exercised on Several
Seismic-Force-Resisting
Systems
Reinforced-Concrete
Structures

4-Story SMF, IMF and OMF

12-Story IMF/OMF and Shear Wall (Core Wall)

Parametric Study of RC Frames

1, 2, 4, 8, 12 and 20 stories

Wood Structures:

Space vs. perimeter configurations

Townhouse Superior, typical, poor quality

Apartment Superior, typical and poor quality

Other (Japanese Home, Templeton Hospital)

Autoclaved Aerated Concrete (AAC) Test Structures

Steel Structures:

4-Story (RBS) SMF (IMF, OMF)

19-20 July 2007 Belgirate, Italy

Source: C. Kircher
LESSLOSS Final Workshop

Krawinkler

Index Archetype
Configuration (4-Story)

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Tall Building
Initiative

Source: R. Klemencic
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Source: J. Maffei
Krawinkler

Source: R. Klemencic
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Whats different about

these buildings?
High-performance materials
Framing systems not satisfying
code prescriptive limits
Non-prescriptive designs are
accepted in the code by
demonstrating at least equivalent
seismic performance.
UBC 1629.10.1, 1605.2, 104.2.8
after MKA
Source: J. Moehle
19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Tall Building
Initiative
Identify performance objectives
Ground motion selection and scaling
Effects of GM selection and scaling on
response
Guidelines on modeling and acceptance
criteria
Input ground motions for tall buildings with
embedded foundations
Dissemination and consensus building

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Gaps in Knowledge

For damage assessment:

Fragility curves for damage in structural

and nonstructural components
Consequence functions and loss curves
Effects of correlations

For downtime assessment

Length of downtime
Scenario dependence of consequences

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Gaps in Knowledge

For collapse prediction and life safety

Better analytical modeling rules for

incorporation of all deterioration and brittle
failure modes at the component level
Modeling of propagation of local collapse
Incorporation of intangible contributions to
collapse capacity
Relationship between collapse and
casualty rate

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler

Concluding remarks 1999

Performance based engineering is here to stay
It enforces a transparent design/assessment approach
Much more emphasis must be placed on $ losses and
loss of
function (downtime)

Performance based design should be reliability based

We have a long road ahead of us

19-20 July 2007 Belgirate, Italy

LESSLOSS Final Workshop

Krawinkler