Decision Support Tools for Life Cycle

Management of Concrete Bridges

Zoubir Lounis
 Outline

• Introduction
• Network & project management models
• Life cycle performance models
• Life cycle cost analysis models
• Conclusions
 Introduction

• Bridge = multi-component system

– slab, stringers, barrier walls, bearings
– girders, piers, abutments
– correlation of component performance

• Multiple failure modes

– loss of strength/collapse: bending, shear, punching shear,
fatigue, etc....
– loss of serviceability: excessive cracking, deformation….
– loss of functionality: inadequate traffic capacity, clearance
 Introduction

• Causes of deterioration & failure

– aggressive environment/corrosion
– increased truck load & volume of traffic
– inadequate design,construction and maintenance
– scour, foundation settlement

• Consequences of deterioration/failure
– reduced safety, serviceability and service life
– reduced level of service
– increased risk of fatalities/injuries
– increased maintenance and user costs
– increased environmental impact
 Introduction

• Management of bridges = complex problem

– owner has thousands of bridges/structures
– bridges deteriorate with time
– about 80,000 bridges in Canada: 45% of bridges > 35 years
– > 50% of bridges: structurally/functionally deficient
– bridge investment backlog = $10 billion
– limited allocated funds
– additional concerns (e.g. environmental sustainability)
 Introduction

• Life cycle of a bridge

Design Construction Use Deterioration

Maintenance Inspection

Replacement Rehabilitation Deterioration

Demolition/ Deterioration
Failure

Disposal
 Introduction

• Maintenance options:
– no maintenance/do nothing
– do nothing + bridge posting
– inspection (visual, NDE, destructive)
– repair
– rehabilitation/strengthening
– improvement
– replacement
– closure of lane(s) or bridge
 Introduction

Maintenance 1 Management Strategy #1

Performance

Limit state
Service life 1 Service life 2

Life cycle
Time (years)

Maintenance 2 & 3
Management Strategy #2
Performance

Limit state

Service life 3 Service life 4 Service life 5

Residual life
Life cycle Time (years)
 Introduction

• Life cycle = Time period for analysis

– design life of bridge/component (30-75 years)
– investment analysis period specified by the owner

• Service life = Time to reach a maximum acceptable level

of damage or limit state = Time to failure
• Residual life = Remaining life at end of life cycle
• Life cycle ≠ Service life
 Introduction

• Objectives of life cycle management of bridges

– identify critical bridges/components
– determination of optimal times for maintenance
– identification of optimal maintenance strategies
– extend service life of bridges
– minimize life cycle costs
– determine required funding over life cycle
– optimize the design of new bridges
 Introduction

• Need models to predict for component/system/

bridge/network
– deterioration
– safety, serviceability, functionality
– service life
– life cycle cost
 Project Management Model

• Analysis of a single bridge/system/component

• Considerable data needs (material properties, loading,
environment, etc.)
• Predict life cycle performance and costs of different
design and maintenance alternatives
• Assess safety/risk of failure
• Optimize maintenance actions for each project
 Network Management Model

• Analysis of groups of bridges/systems/components

• Limited data needs (condition ratings at two points in time)
• Predict life cycle performance and cost of network
• Prioritization of projects for maintenance
 Integration of Network &
Project Management Models

Condition Rating Data

Quantitative Data

Network-Level Analysis
Project-Level Analysis

Qualitative Deterioration Models

Quantitative Deterioration Models

Output:
• Estimates of life cycle performance Service Life Prediction
• Required maintenance budgets
• Project prioritization
Life Cycle Performance
Assessment

Life Cycle Cost

Analysis
 Modelling of Life Cycle
Performance

• Qualitative deterioration models

– life cycle performance of networks of bridges/systems
– preliminary life cycle performance of a
bridge/system/component
– performance indicator = “condition rating”
– initial estimates of life cycle costs
 Modelling of Life Cycle
Performance

• Quantitative deterioration models

– detailed life cycle performance of one
bridge/system/component
– performance indicators = load, resistance, safety
margin, etc.
– required for safety evaluation of critical bridge elements
– improve initial estimate of life cycle cost
 Modelling of Life Cycle
Performance

• Bridge performance
– resistance/condition degrade with time
– load increases with time
– considerable scatter/uncertainty in load and resistance
– performance (safety, serviceability, rating): time-dependent
and highly uncertain
• Sources of uncertainty
– environmental exposure
– material properties (concrete, reinforcing steel)
– loading (magnitude, time of occurrence)
– structure design (quality control)
– physical models of structural response
– acceptable damage/failure criteria
 Qualitative Deterioration
Models

• Progressive accumulation of irreversible damage

– reinforcement corrosion
– traffic loading
– freeze-thaw cycles
– fatigue, etc.
• Consideration of uncertainty
• Compatible with existing discrete rating of bridges
 Qualitative Deterioration
Models

Condition Rating Description

6 Excellent Condition

5 Good Condition

4 Fair Condition
3 Poor Condition
2 Urgent Condition
1 Critical Condition
 Qualitative Deterioration
Models

• Bridge deterioration is time-increasing and progressive:

cumulative damage (CD)
• Considerable scatter/fluctuations from mean values of
performance

Bridge condition is a stochastic process

• Future bridge condition depends only on:

– present condition (not on entire history)
– stress cycle (environment, load, material)
Damage accumulation is a first-order
Markovian process
 Qualitative Deterioration
Models
p66 6 probability of being in
state 5 given the present
p65 state is 6 (1 jump) within
a transition period
p55
5 p64
p54 probability of being in
state 4 given the present
probability of
remaining p44 4 state is 6 (2 jumps)
within a transition period
in same
p43
state within a
transition period 3
p33
p32
2
p22
p21
p11 1
 Qualitative Deterioration
Condition Rating Models

6
5
4
3 Probability mass
2 functions of
condition rating
1

0 t1 t2 t3 Time
 Qualitative Deterioration
Models

• Focus on reinforced concrete bridge decks

– weakest links (most deteriorated elements)
– deck maintenance costs = 1/3 to 1/2 projected bridge
maintenance costs

• Development of CD model for bridge decks

– historical field performance data (at least 2 points in time)
– explanatory variables
• highway class
• environmental exposure
• average daily traffic
• average daily truck traffic
 Qualitative Deterioration
Models

6 5 4 3 2 1
6 0.65 0.32 0.03 0.00 0.00 0.00
5 0.00 0.88 0.09 0.02 0.01 0.00
4 0.00 0.00 0.85 0.10 0.03 0.01
3 0.00 0.00 0.00 0.88 0.08 0.04
2 0.00 0.00 0.00 0.00 0.93 0.07
1 0.00 0.00 0.00 0.00 0.00 1.00

Transition probability matrix of RC bridge decks for 3 years transition period

 Qualitative Deterioration
Models

6
Benign
Environment
5
MTQ Condition Rating

4 Low
Environment

MTQ Data
3
Moderate
Environment
2
Severe
Environment
1
0 20 40 60 80 100

Age (years)
 Quantitative Deterioration
Models

• Failure: Load > Resistance

– max. moment > flexural resistance
– max. chloride content > corrosion resistance
– deformation > limit
– damage > limit

• Limit state: Load = Resistance

• Service life = Time to reach limit state

 Quantitative Deterioration
Models

Load
Resistance
Resistance
Initial resistance
fRo(r)

Failure
fS(s)

Load

Time
 Quantitative Deterioration
Models

Load Resistance
Relative frequency

Probability of
failure

Mean Load < Mean Resistance Load

Resistance
 Quantitative Deterioration
Models

• Failure: P (Load > Resistance) > Pfmax

• Service life: Time at which P (Load > Resistance) = Pfmax

 Quantitative Deterioration
Models

Probability Time variation of probability of failure

of Failure

Limit state
Pfmax

Pf(0)

Service Life Time (years)

 Quantitative Deterioration
Models

• Load, resistance, service life are modelled as:

– random variables
– stochastic processes
• Prediction of probability of failure
– advanced first-order reliability theory
– crossing theory
 Quantitative Deterioration
Models

Failure

Delamination/
spalling
Damage

Surface
Internal cracking
cracking
Corrosion

Chloride contamination Time

Service Life
 Quantitative Deterioration
Models

• Six (6) limit states/failure modes are modelled

– critical chloride contamination of deck
– onset of corrosion of reinforcing steel
– internal cracking
– surface cracking
– spalling/delamination
– failure (excessive damage)
• Service life = Time to reach any one of 6 limit states
Quantitative Deterioration
Models

Internal Cracking

Surface Cracking

Spalling

Delamination
 Quantitative Deterioration
Models

Corrosion-induced deterioration of RC bridge deck

 Quantitative Deterioration
Models

• Mechanistic deterioration models

– Ficks’s 2nd law of diffusion for predicting chloride ingress
into concrete
– concept of “threshold chloride” for corrosion initiation in
reinforcing steel
– behavior of deck under corrosion-induced stresses modelled
as thick-walled cylinder under internal pressure
 SLAB-D Software

• SLAB-D: Decision
support software for
service life analysis of
RC bridge decks

• 2 modules in SLAB-D:
– Service Life Prediction
(SLP) module
– Life Cycle Cost Analysis
(LCCA) module
 Quantitative Deterioration
Models

MCS 10000 sim.

1.0
MCS 100000 sim.
0.9
SLAB-D
0.8
Corrosion Probability

0.7

0.6 At 40 years, the probability is:

MCS 10000 = 0.90
0.5 MCS 100000 = 0.90
SLAB-D = 0.92
0.4

0.3

0.2

0.1

0.0
0 10 20 30 40 50
Time (years)
Quantitative Deterioration
Models
 Life Cycle Cost Analysis

• Agency (Direct) Costs

– design cost
– construction cost
– inspection cost
– MR&R costs

• User (Indirect) Costs

– travel delay costs
– vehicle operating costs
– accident costs
 Life Cycle Cost Analysis

F
• Present value life cycle cost
0 1 2 3 4 (n-1)

(n)

F
P P equivalent to F
P =
(1 + r ) n

• Discount rate: To convert future costs incurred at different times to

equivalent present value costs
• P = principal that will give the required amount F in (n) years
assuming a constant interest rate r
• If r is high, lesser significance is given to future expenditures
 Life Cycle Cost Analysis

Initial Agency Costs

HPC deck = 30% more expensive than NPC deck

 Life Cycle Cost Analysis

LC = 30 years Present Value Life Cycle Costs

r = 3% $291/m2

$231/m2

HPC deck = 26% less expensive than NPC deck

 Conclusions

• Life cycle bridge management is formulated as a

two-level management process
• Stochastic cumulative damage models are developed
for network-level management
• Stochastic mechanistic models are developed for
project-level management
• Life cycle cost analysis is used to select cost-effective
maintenance strategies
 Conclusions

Effective life cycle management plan

+
Adequate financing
=
Sustainability of Canada’s bridge
infrastructure
 Acknowledgments

• Alberta Infrastructure and Transportation

• Cement Association of Canada
• City of Hamilton
• City of Ottawa
• City of Winnipeg
• Engineered Management Systems Inc.
• Federal Bridge Corporation Ltd.
• Manitoba Department of Transportation and Government Services
• Ministère des Transports du Québec
• Nova Scotia Department of Transportation and Public Works
• Public Works and Government Services Canada
• Region of Durham