implications of recent fema p-58 methodology …

1

Implications of Recent FEMA P-58 Methodology Advancements

for Resilient BRBF Design

Presented by: Curt B. Haselton, PhD, PEJohn F. O’Connell Endowed Professor @ CSU, Chico

Co-Founder and CEO @ Haselton Baker Risk Group (SP3)

Work by: The Full SP3 TeamEdward Almeter (project lead), Jack Baker, Katie Wade,, D. Jared DeBock,

Shaunt Kojabashian, Mike McGlone, Tracy Rice, and Dustin Cook

SP3 | where research meets practicewww.hbrisk.com

NASCC Steel Conference | April 5, 2019

2

© HB Risk Group

§ Motivation – Current performance and the push for resilient design

§ The FEMA P-58 Methodology for assessing resilience:ü Overviewü Method enhancement for BRBFs

§ Updated resilience predictions for BRBFs§ Resilient design of BRBFs§ Summary and Q&A

Presentation Outline

3

© HB Risk Group





4

© HB Risk Group

Code Design (ASCE7, etc.)§ Safety Goal – Yes (divides for demands by R and results in a

“safe but disposal” building)

“Performance”-Based Design (ASCE 41, SF AB 083, etc.)§ Safety Goal – Yes§ Also enhanced modeling and design scrutiny

“Resiliency”-Based Design (and Risk Assessment)§ Safety Goal – Yes§ Repair Time Goal – Yes§ Repair Cost Goal - Yes§ Also enhanced modeling and design scrutiny

Current Performance and Push for Resilient Design

5

© HB Risk Group


6

© HB Risk Group

§ Safety: Expected to be safe (per 1989 and 1994 experience), but this was not assessed in this study.

§ Losses: 5-25% mean loss for design event and 10-80% loss for rare event (huge range!).

§ Downtime: 6-12 months for design event and up to 2 years for rare event (substantial predicted downtimes!).

§ Why?: High values and huge ranges of performance because resiliency (losses and downtime) are simply not constrained in our code design process.

Hence, we refer to code-compliant buildings as safe but

disposable for a rare event!


7

© HB Risk Group

How can we do better?§ Design buildings not only to be safe, but also to limited

downtime, and possibly also to limit repair cost.§ Two options:

a) Do direct resilient design using targets for downtime/losses and using FEMA P-58 analysis method.

b) Do indirect resilient design using modified building code requirements (e.g. Risk Category IV), but such requirement have not yet been developed (in process).

c) a

§ A§ a


8

© HB Risk Group

Current pushes for resilient design at all levels:§ California Assembly Bill 393 – “functional recovery” code

provisions for California § NEHRP Reauthorization – Congress tasked NIST/FEMA with

recommendations for design for functional recovery§ Cities like San Francisco – considering enhanced design

requirements for tall buildings (to consider functionality)


9

© HB Risk Group





10

© HB Risk Group

§ FEMA P-58 is a probabilistic performance prediction methodology (15 year, $16M+ invested, ~100+ on the team)

§ FEMA P-58 is tailored for building-specific analysis (in contrast to most risk assessment methods)

§ FEMA P-58 output results:• Repair costs• Repair time• Safety: Fat & Injury

FEMA P-58: Overview

11

© HB Risk Group

Site Hazard Structural Responses

Structural Components &

Fragilities

Nonstructural Components &

Fragilities

Building-Specific Vulnerability

Curves

Full distributions of losses and repair

times, and expected annual values.

FEMA P-58 Monte Carlo Analysis

ENGINE

FEMA P-58: Detailed Steps of Method

12

© HB Risk Group

FEMA P-58: Ground Motions

§ Step 1: Define ground motion hazard (with soil)• Option #1: SP3 can provide curve (given an address)

• Option #2: User-specified curve

13

© HB Risk Group

FEMA P-58: Structural Response

§ Step 2: Predict “engineering demand parameters”

• Story drift ratio at each story• Peak floor acceleration at each

floor• For wall buildings, also wall

rotations and coupling beam rotations

Option #1: Response-history structural analysis

Option #2: Statistically calibrated predictive equations (**and we will need to extend these for BRBFs**)

EQ: 11122, Sacomp(T=1sec): 1.02g

Deierlein, Haselton, Liel (Stanford University)

14

© HB Risk Group

FEMA P-58: Component Damage

§ Step 3: Quantify component damageFirst, establish what components are in the building. Types and quantities of can be specified or estimated from building size and occupancy type

Windows Piping

Partitions Structural components

15

© HB Risk Group

We end up with a list of component types, quantities and locations


§ Step 3: Quantify component damage

16

© HB Risk Group

Each component type has a “fragility function” that specifies the probability that a structural demand causes damage (**and we will need these for CoreBrace BRBFs**)


§ Step 3: Quantify component damage

17

© HB Risk Group

FEMA P-58: Consequences of Damage

Fragility functions have been calibrated for hundreds of components from test data, and repair cost and labor has been developed by cost estimators.

Cost per 100 ft. Labor per 100 ft.Cracked wallboard $2,730 24 person-hours

Crushed gypsum wall $5,190 45 person-hours

Buckled studs $31,100 273 person-hoursThese are median values—each also has uncertainty

§ Step 4: Quantify consequences of the component damage (component repair costs, repair times, etc.).

18

© HB Risk Group

FEMA P-58: Building-Level Consequences

Repair costs are the sum of component repair costs (considering volume efficiencies)

Recovery time is aggregated from component damage, but is more complex (mobilization, staffing, construction sequencing, …)

Windows $26,892Partitions $43,964

Piping $5,456

Structural Components

$77,920

… …

Sum = $253,968

§ Step 5: Aggregate to building-level consequences

19

© HB Risk Group

FEMA P-58: Monte Carlo Simulations

For a given building and ground shaking intensity, repeat the following steps 2,000-10,000 times!

1. Simulate each structural response parameter2. Simulate damage to each component3. Simulate repair costs and repair time for each component4. Aggregate to compute total repair cost and recovery time

We can then look at the mean cost, 90th percentile, etc.

§ Step 5: Aggregate to building-level consequences

20

© HB Risk Group

FEMA P-58: Summary

§ Step 1: Site Hazard• Soil and hazard curve

• Ground motions (if needed)

§ Step 2: Structural Responses• Option #1: Structural analysis• Option #2: Predictive equations

§ Step 3: Damage Prediction• Contents • Fragility curves

§ Step 4: Loss Estimation (loss curves & other consequences)

Step 5: Aggregate to building-level consequences

Thousands of Monte Carlo simulations

The simulations provide detailed statistical

information on building performance.

21

© HB Risk Group

8%

70%

16%

0% 3% 3% 0% 0%0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

StructuralComponents

Partitions InteriorFinishes

Cladding Plumbing andHVAC

OtherComponents

Collapse Residual Drift

Loss Contributions by Component Type for a 50 Year Ground Motion

FEMA P-58: Sample Outputs

New 8-story office in Los Angeles

Loss Ratio = 0.04

22

© HB Risk Group


37%32%

7%1% 1% 2% 1%

19%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%




OtherComponents



Loss Ratio = 0.15


23

© HB Risk Group


26%

12%

2% 2% 0% 1% 3%

54%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%




OtherComponents



Loss Ratio = 0.44


24

© HB Risk Group

§ Average Repair Times (REDi, 2013):

0.8 0.9

3.5

0

2

4

6

8

10

12

14

REDi Re-Occupancy REDi Functional Recovery REDi Full Recovery

MO

NTHS

Repair Time Output at a 43 Year Earthquake


25

© HB Risk Group


6.0 6.3

9.2

0

2

4

6

8

10

12

14


MO

NTHS



26

© HB Risk Group


10.7 11.0

13.1

0

2

4

6

8

10

12

14


MO

NTHS



27

© HB Risk Group

FEMA P-58 provides the comprehensive and standardizedbuilding-specific risk assessment.

SP3 software provides a user-friendly software to integrate all steps in a

FEMA P-58 risk assessment.

The initial assessment should take a couple hours and not days or

weeks.

FEMA P-58: Enabling SP3 Software

28

© HB Risk Group





29

© HB Risk Group

Site Hazard Structural Responses

Structural Components &

Fragilities

Nonstructural Components &

Fragilities

Building-Specific Vulnerability

Curves

Full distributions of losses and repair

times, and expected annual values.

FEMA P-58 Monte Carlo Analysis

ENGINE

FEMA P-58 Method Advancements for BRBFs

SP3 Structural Response Prediction ENGINE

“We do the nonlinear dynamic structural analysis for you.”

Component Fragility Database

(132 new fragilities specific to brace geometry and higher

ductility of CoreBrace BRBF data)

30

© HB Risk Group

Component Fragility Database

(132 new fragilities specific to brace geometry and higher

ductility of CoreBrace BRBF data)


31

© HB Risk Group

• Comparison of new fragilities with standard FEMA P-58 fragilities (where the damage state is fracture of the brace requiring replacement)

FEMA P-58 baseline

CoreBrace


32

© HB Risk Group

SP3 Structural Response Prediction ENGINE

“We do the nonlinear dynamic structural analysis for you.”

Specifically predict:– peak interstory drift– peak floor acceleration– residual interstory drift (a big focus)


33

© HB Risk Group


34

© HB Risk Group

Detailed nonlinear dynamic structural modeling, with many

building designs, was used to refine the residual drift model for CoreBrace BRBF buildings.


35

© HB Risk Group

The FEMA P-58 default residual drift model is slightly conservative for

CoreBrace BRBFs (but only slightly).


36

© HB Risk Group

Including a typical gravity system (beam/slab and

shear tab connections) in the nonlinear structural model

shows substantial reduction in residual drifts.




37

© HB Risk Group

Including a moment-connected back-up frame in

the nonlinear structural model shows even more

reduction in residual drifts. The typically designed back-up frames (sized for gravity) were sufficient and they did

not need additional requirements for the back-

up frames.



Including a typical gravity system (beam/slab and

shear tab connections) in the nonlinear structural model

shows substantial reduction in residual drifts.


39

© HB Risk Group

§ We now want to see how the resiliency assessment results have changed based on these method improvements from the past 1.5 year of research and development (use a new mid-rise office building example in SDC D).

§ This comparison study was motivated by a more generalized study by the Applied Technology Council, in which the BRBF results appeared conservative.

FEMA P-58 Updated Risk Assessments

ATC Study Repair Costs for Various Structural Systems (Hooper et al. 2018)

41

© HB Risk Group

Average Repair Costs (percentage building value):


Step 1: Baseline ATC studyStep 2: Design for coupled strength/stiffnessStep 3: Consider gravity system, residual updatesStep 4: More ductile CoreBrace BRBF fragilitiesStep 5: Use moment-connected frame

44

© HB Risk Group

Other Performance Metrics from FEMA P-58 – residual drifts and percentage repairable, many others available…


46

© HB Risk Group

§ To design a building to be resilient:ü Need no structural damage that requires repair (easy for a CoreBrace

BRBF)ü Need minimal residual drifts for no demolition (issue for ductile systems

like BRBFs but can be well controlled using a back-up moment frame)ü Need to either limit floor accelerations and/or design non-structural

components to resist them (BRBF yielding helps)ü Need to limit drifts to prevent other drift-sensitive non-structural

damage

§ Two resilient design options:a) Do direct resilient design using targets for downtime/losses and using

FEMA P-58 analysis method.b) Do indirect resilient design using modified building code requirements

(e.g. Risk Category IV), but such requirement have not yet been developed (in process).

c) For CoreBrace BRBFs: Use the Resilient Design Guide!

Resilient Design of CoreBrace BRBF Buildings

48

© HB Risk Group

§ The FEMA P-58 methodology can provide quantitative resiliency predictions (repair costs and repair times).

§ Much research and development has been done to refine the FEMA P-58 methodology for BRBF buildings (fragilities and residual drifts).

§ The refined methods shows that damage and residual drifts are lower than previously predicted, though can still be an issue for such a ductile system. Even so, residual drifts can be controlled well by using a moment-resisting back-up frame.

§ Resilient design of CoreBrace BRBF buildings can now be done using either FEMA P-58 assessment or the Resilient Design Guide (available from CoreBrace.

Summary and Conclusions

49

© HB Risk Group

Closing and Questions

§ Thank you for your time.§ Our goal is to support adoption of resilience-based design and

risk assessment, and we welcome feedback and suggestions.

§ Time for questions!

Curt Haselton: [email protected], Direct: (530) 514-8980Tracy Rice (HB-Risk admin): [email protected]

www.hbrisk.comSP3 | where research meets practice

implications of recent fema p-58 methodology …

Documents