Impact Assessment Of Hydro - ciri.illinois.educiri.illinois.edu/sites/default/files/Tandon Presentation for CIRI.pdf · TxDOT Funded Study •Problem •Approach •Framework Adopted

Impact Assessment Of Hydro-Meteorological Events On Texas Pavements And Development Of

Resiliency Strategy

Overview• Presentation in Two Parts:• Research Results from Study Funded by Texas

Department of Transportation• Proposed Future Work for funding from DHS

2

TxDOT Funded Study•Problem•Approach•Framework Adopted•Climate Model Data•Vulnerability Assessment of the Pavements•Adaptation Methods•Probabilistic Analysis•Conclusions

3

Why Hydrometerological Events are problem?

Rutting result of change in temperature and soil moisture

4

Flooding: wetter surface, stripping increases, water enters pavement

structures

Hurricane Harvey Damage

Approach• Future Cliamte Projections for Regional and Local

• Parameters Impacting Pavement Infrastructure

• Framework for Evaluating the Performance of the Pavements

• Vulnerability Assessment of Pavements: – Current and projected variability of climate change

– Evaluating effects of climatic change on the performance of pavement structures

– Nature and severity of each parameter

– The sensitivity of roads to climate change

• Adaptation Methods for Climate Change

• Probabilistic Analysis 5

6

• Special Report on Emission Scenario (IPCC AR4, 2000)

• Representative Concentrative Pathways (IPCC AR5, 2014)

• Statistical Downscaling• Dynamic Downscaling

Emission Scenarios

Conduct Probabilistic

Analysis

Downscaling Methods

• Temporal: Daily, 3-hourly• Spatial Resolution: 50x50 km

Climate Data Source used in this StudyClimate Source/ Downscaling method/

Emission scenarioParameters/ Resolution

North American Regional Climate Change Assessment Program

(NARCCAP) uses Dynamic downscaling and is based on

Emission Scenario: SRES A2 .

Temporal resolution: Daily (maximum and minimum surface air temperature)

3-hourly (Climate Parameters: precipitation, surface air temperature, cloud fraction, wind

speed, relative humidity)Spatial resolution: 50x50 km

7

Climate ModelsNARCCAP includes six RCMs and four GCMs namely:

RCMs:

1) Hadley Regional Climate Model Version 3 (HRM3)

2) Regional Climate Model 3.0 (RCM3)

3) The Canadian Regional Climate Model (CRCM)

4) Experimental Climate Prediction Center Regional Spectral Model (ECPC)

5) Mesoscale Meteorological Model Version 5.0 (MM5I)

6) The Weather Research and Forecasting Model (WRFG)

GCMs:

1) The Hadley Centre Climate Model (HadCM3)

2) Community Climate System Model (CCSM)

3) The Canadian Global Climate Model (CGCM3)

4) The Geophysical Fluid Dynamics Laboratory (GFDL) model

8

9

Mean Annual Temperature

Historical Climate (1979-2015)

CRCM-CGCM3 Future(2038-2070)

10

0

10

20

30

40

50

60

70

80

1960 1980 2000 2020 2040 2060 2080

Mea

n A

nnua

l Tem

pera

ture

(°F)

Time Period (yrs)

Existing Temperature (Pavement ME)Model Simulate Current (CRCM-CCSM)Model Simulate Future (CRCM-CCSM)Bias corrected Model Simulate CurrentBias corrected Model Simulate Future

Bias-Correction for TemperatureFort Worth

11

Bias-Correction for PrecipitationCorpus Christi

0

10

20

30

40

50

60

1960 1980 2000 2020 2040 2060 2080

Mea

n A

nnua

l Pre

cipi

tatio

n (in

)

Time Period (years)

Exixting Precipitation (Pavement ME) Model Simulate Current (CRCM-CCSM)

Model Simulate Future (CRCM-CCSM Bias Corrected Model Simulate Current

Bias Corrected Model Simulate Future

According to George E. P. Box

“All models are wrong but some are useful”

12

0

20

40

60

80

100

120

Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec

Mon

thly

Mea

n Te

mpe

ratu

re (°

F)Fort Worth

CRCM-CGCM3 HRM3-HADCM3 MM5I-CCSM

MM5I-HADCM3 HRM3-GFDL CRCM-CCSM

Existing ECP2-GFDL RCM3-CGCM3

RCM3-GFDL WRFG-CGCM3 WRFG-CCSM

0

10

20

30

40

50

60

70

80

Amarillo Austin Corpus Christi Dallas El Paso Fort Worth Houston Lubbock Mc Allen San Antonio

Mea

n A

nnua

l Tem

pera

ture

(°F)

Pavement ME Climate MM5I-HADCM3 MM5I-CCSMHRM3-HADCM3 HRM3-GFDL CRCM-CGCM3CRCM-CCSM ECP2-GFDL RCM3-CGCM3RCM3-GFDL WRFG-CGCM3 WRFG-CCSM

0

20

40

60

80

100

120

140

160

180

200

Amarillo Austin Corpus Christi Dallas El Paso Fort Worth Houston Lubbock Mc Allen San Antonio

Mea

n A

nnua

l Pre

cipi

tatio

n (in

.)Pavement ME Climate MM5I-HADCM3MM5I-CCSM HRM3-HADCM3HRM3-GFDL CRCM-CGCM3CRCM-CCSM ECP2-GFDLRCM3-CGCM3 RCM3-GFDLWRFG-CGCM3 WRFG-CCSM

0

2

4

6

8

10

12

14

16

Amarillo Austin Corpus Christi Dallas El Paso Fort Worth Houston Lubbock Mc Allen San Antonio

Mea

n A

nnua

l Win

d Sp

eed

(mph

)Pavement ME Climate MM5I-HADCM3MM5I-CCSM HRM3-HADCM3HRM3-GFDL CRCM-CGCM3CRCM-CCSM ECP2-GFDLRCM3-CGCM3 RCM3-GFDLWRFG-CGCM3 WRFG-CCSM

0

10

20

30

40

50

60

70

80

90

100

Amarillo Austin Corpus Christi Dallas El Paso Fort Worth Houston Lubbock Mc Allen San Antonio

Mea

n A

nnua

l Rel

ativ

e H

umid

ity (%

)Pavement ME Climate MM5I-HADCM3MM5I-CCSM HRM3-HADCM3HRM3-GFDL CRCM-CGCM3CRCM-CCSM ECP2-GFDLRCM3-CGCM3 RCM3-GFDLWRFG-CGCM3 WRFG-CCSM

18https://www.youtube.com/watch?v=zJe2-W5JKok

19Rut Depth in Asphalt Concrete Layer

Ride QualityIRI (International Roughness

Index)

IRI is calculated from longitudinal profile measured with a road profiler in both wheelpaths. The average IRI of the two wheelpaths is reported as the roughness of the pavement section.

20

Conduct Probabilistic

Analysis

IH 30 Frontage Road for Process Demonstration

• AADTT: 828

• Percentage of Trucks: 3.6%

• Growth rate 1.86 % for 20 years

• Annual average daily traffic (AADT) of 22,990

21

Subgrade

Asphalt Concrete

Cement Treated Base

4.0”

15.0”

Semi Infinite

Layer Material Properties

Asphalt Concrete Type D, PG 70-22

Cement Treated Base 120 ksi

Subgrade 4.5 ksi

Influence of Climate Models on Pavement Performance

22

60

70

80

90

100

110

120

130

140

150

160

0 2 4 6 8 10 12 14 16 18 20

IRI(

in./m

ile)

Pavement Age(years)

CRCM_CCSMCRCM_CGCM3ECP2_GFDLECP2-HADCM3HRM3_GFDLHRM3_HADCM3MM5I_CCSMMM5I_HADCM3RCM3_CGCM3RCM3_GFDLWRFG_CCSMWRFG_CGCM3Pavement ME Climate

Maintenance

Range variation due to different Climatic Predictions

8.5 9.5

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 2 4 6 8 10 12 14 16 18 20

AC

Rut

ting(

in.)

Pavement Age(years)

CRCM_CCSMCRCM_CGCM3ECP2_GFDLECP2-HADCM3HRM3_GFDLHRM3_HADCM3MM5I_CCSMMM5I_HADCM3RCM3_CGCM3RCM3_GFDLWRFG_CCSMWRFG_CGCM3Pavement ME Climate

Threshold Value

Early Maintenance with climate change projections

15.8

Influence of Extreme Event on Pavement Performance

Assumptions:

• Extreme rainfall will occur at an interval of 1, 2, 3, 5, 10, and 15 years of service life of the pavements.

• Saturation will last at a time for 7.5 days or 15 days or one month or two months after occurrence of these events

25

26

270.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 5 10 15 20

Bas

e+S

ub

gra

de

Ru

ttin

g (

in.)

Pavement Age (years)

Pavement ME Climate

2 Month

1 Month

15 Days

7.5 Days

28

290

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

0 2 4 6 8 10 12 14 16 18 20

Subg

rade

Mod

ulus

(psi

)

Pavement Age (yrs)

Pavement ME Climate

1 Month

30

Adaptation Methods• Adaptation Methods for Climate Change

– Increasing the Thickness of AC Layer

– Binder Change

– Changing Mix Type

– Increasing the Thickness of AC Layer and Changing Binder Grade

• Adaptation Methods for Extreme Events

– Improving Subgrade Modulus

• Adaptation to Extreme Event and Climate Change

– Improving Subgrade Modulus and Increasing the Thickness of AC Layer

31

Changing Material Type with Pavement ME

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0 5 10 15 20

AC

Rut

ting

(inch

)

Pavement Age (yrs)

SMA PG 70-22SP-C PG 70-22Type C PG 70-22Type D PG 70-22CC_SMA PG 70-22CC_SP-C PG 70-22CC_Type C PG 70-22CC_Type D PG 70-22

0

20

40

60

80

100

120

140

160

0 5 10 15 20IR

I (in

/mile

)Pavement Age (yrs)

SMA PG 70-22SP-C PG 70-22Type C PG 70-22Type D PG 70-22CC_SMA PG 70-22CC_SP-C PG 70-22CC_Type C PG 70-22CC_Type D PG 70-22

With changing climate the average increase in AC rutting is 0.18 in. for any material type

Increase in IRI is about 5 % for each material type with changing climate

Changing Thickness3.0

3.5

4.0

4.5

5.00

20

40

60

80

100

120

140

160

0 2 4 6 8 10 12 14 16 18 20

PSI

IRI (

in/m

ile)

Pavement Age (yrs)

Type D PG 70-22CC_Type D PG 70-22_1 inCC_Type D PG 70-22

Maintenance

8.3 9.49.1

Monte-Carlo Simulationsfor

Probabilistic Analysis

34

Climate Models

Future Simulation (1938-2070)

(°F)

Current Simulation (1971-2000)

(°F)

Bias Corrected

(°F)

CRCM-CCSM 69.33 64.38 75.01CRCM-CGCM3 66.04 61.60 71.73ECP2-GFDL 58.36 54.92 64.05ECP2-HADCM3 59.37 55.62 65.05HRM3-GFDL 67.94 63.03 73.62HRM3-HADCM3 69.16 65.23 74.85MM5I-CCSM 69.09 65.06 74.77MM5I-HADCM3 66.70 62.30 72.39RCM3-CGCM3 64.38 60.35 70.06RCM3-GFDL 60.56 57.04 66.24WRFG-CCSM 63.95 59.02 69.63WRFG-CGCM3 63.87 61.04 69.55

35

Climate Models

Future Simulation (1938-2070)

(inches)

Current Simulation (1971-2000)

(inches)

Bias Corrected

(inches)

CRCM-CCSM 19.76 20.57 23.71CRCM-CGCM3 24.80 24.40 29.76ECP2-GFDL 30.98 36.12 37.16ECP2-HADCM3 31.02 31.58 37.22HRM3-GFDL 35.44 37.95 42.52HRM3-HADCM3 34.18 33.67 41.00MM5I-CCSM 21.05 21.86 25.25MM5I-HADCM3 28.22 27.10 33.86RCM3-CGCM3 28.98 30.08 34.77RCM3-GFDL 35.96 36.10 43.14WRFG-CCSM 22.58 20.62 27.09WRFG-CGCM3 22.02 21.61 26.42

Annual Average Mean Temperature

Annual Average Mean Precipitation

Pavement ME Mean Temperature: 66.48 °F Pavement ME Mean Precipitation: 34.16 inches

36

Using Bias-corrected Range for Temperature & Precipitation

Findings• Selected model simulation showed a change in future

• All prediction models simulated an increase in mean annual precipitation, which leads to premature failure

• Change in environmental condition adversely impacts the pavements functionality by reducing the service life of pavements

• Increasing the thickness of the asphalt concrete layer, binder grade, and mix type are some of the options for mitigating impact of hydrometerological events

37

Issues• TxDOT manages roughly 73,000 miles of highways.

• Assuming two lane highway, an increase in 1” layer thickness would cost roughly 20 billion dollars to mitigate impact of environmental conditions.

• Ignoring impact of extreme events is not an option either.

• Identifying and enhancing resiliency of critical pavement infrastructure is an option.

38

39

Part IIDepartment of Homeland Security

Funding Request

From Standard Materials Book (Campbell, Flake C. (2008). Elements of Metallurgy and Engineering Alloys. ASM International

Behavior After Strain Hardening

Beginning of 19th CenturyResiliency

Linkov, Igor & Bridges, Todd & Creutzig, Felix & Decker, Jennifer & Fox-Lent, Cate & Kröger, Wolfgang & Lambert, James & Levermann, Anders & Montreuil, Benoit & Nathwani, Jatin & Nyer, Raymond & Renn, Ortwin & Scharte, Benjamin & Scheffler, Alexander & Schreurs, Miranda & Clemen, Thomas. (2014). Changing the resilience paradigm. Nature Climate Change. 4. 407-409. 10.1038/nclimate2227.

Resiliency

42

43P. Bocchini, M. Asce, D.M. Frangopol, D.M. Asce, T. Ummenhofer, T. ZinkeResilience and sustainability of civil

infrastructure: toward a unified approach

44

Resiliency of Critical Transportation Infrastructure

• Accumulate Information on Infrastructure (Transportation Asset Management)

• Identification of critical components of infrastructure through Simulation of Scenarios, traffic patterns, and experience.

• Develop a Resilience Assessment Strategy for Selection and Mitigation

• Develop Multi-criteria Decision Making Framework for Evaluating and Selecting Efficient and Environmentally Friendly Alternative

45

46

Adopted from TRB Report 551 on Transportation Asset Management

Transportation Asset

Management Framework

47

Critical TransportationInfrastructure

48

Simulations to Identify Critical Infrastructure

49Roberto Guidotti, Hana Chmielewski, Vipin Unnikrishnan, Paolo Gardoni, Therese McAllister & John van de Lindt (2016) Modeling the resilience of critical infrastructure: the role of network dependencies, Sustainable and Resilient Infrastructure, 1:3-4, 153-168, DOI: 10.1080/23789689.2016.1254999

Network Dependencies

50

Urban Road Efficiency Model by Ganin et al. (2017)

51

52

Resilience Assessment Strategy Proposed by New Zealand Transportation Agency

53

Resilience Assessment Strategy Proposed by New Zealand Transportation Agency

54

Resilience Assessment Strategy Proposed by New Zealand Transportation Agency

55

T. P. Bostick; E. B. Connelly; J. H. Lambert; I. Linkov2018: Resilience science, policy and investment for civil infrastructure

Decision Model – To identify an effective decision

making model which can handle

multiple criteria (subjective as well as

objective)

– To develop a group decision model

Decision Model

Multi Criteria Group Decision Model

Decision Model: Multi Criteria Decision Model

• Analytical Hierarchy Process (AHP) is traditionally used for decision making in Civil Engineering

Pros Cons57

Group Decision Model

Current Practice

• Arithmetic or Geometric

Mean of individual

decisions.

• Manipulating individual

decisions.

Pros Cons

58

Group Decision Model • Data Envelopment Analysis (DEA) based preference

aggregation method (commonly used in business

community) can be selected for group decision.

𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑗𝑗 = �𝑘𝑘=1

𝑛𝑛

𝑢𝑢𝑘𝑘 Ѳ𝑗𝑗𝑘𝑘

𝑆𝑆𝑢𝑢𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑡𝑡

�𝑘𝑘=1

𝑛𝑛

𝑢𝑢𝑘𝑘Ѳ𝑗𝑗𝑘𝑘 ≤ 1 𝑆𝑆 = 1,2, … ,𝑛𝑛

𝑢𝑢1 ≥ 2𝑢𝑢2 ≥ ⋯ ≥ 𝑛𝑛𝑢𝑢𝑛𝑛

𝑢𝑢𝑛𝑛 ≥ Ɛ =2

)𝑝𝑝𝑛𝑛(𝑛𝑛 + 1

Pros ConsPros Cons

Linear Programming Model 59

Group Decision Model • A new constrained optimization method

αPSO is proposed by Takahama and Sakai

(2004) which is a combination of the α

constrained method and PSO. Pros

• Simplistic Approach.

• Optimizes in minimal time frame.

• Can be used for many engineering problems.

Cons

• Not applicable in a constrained environment.

60

Decision Model

Group Decision Model

H-PSO

DEAAHP

For the study a decision model that integrates the following things

can be used:

• Analytic Hierarchy Process (AHP) (handles multiple criteria)

• Fuzzy logic (lessens the drawbacks of AHP)

• Data Envelopment Analysis (handles group decision)

• Hybrid Particle Swarm Optimization (handles the linear

programming model and minimizes running time).

61

62

63

Thanks TxDOT, DHS, and CIRI for the

Support