Evaluating Effect of Preservation Treatments on Pavement ...
Evaluating Effect of Preservation Treatments on Pavement Performance and Service Life
James Bryce & Gonzalo Rada Amec Foster Wheeler National Pavement Preservation Conference Nashville, TN October 13, 2016
Outline • Background • Performance measures • Modeling the effects of preservation Performance
Service Life Life Cycle Costs
• Conclusions
Background • Based on research conducted under NCHRP Project 14-33 to: Identify and/or develop pavement performance measures that consider contributions of preservation to performance, service life, and life-cycle costs Prepare guide document to facilitate implementation of measures by highway agencies
Background, Cont’d • Preservation treatments are applied to: Preserve an existing roadway
Slow future deterioration Maintain and improve its functional condition
• No substantial increase to structural capacity of
pavement Performance measures provide means for assessing effectiveness of preservation
Background, Cont’d • Considerable amount of literature Mamlouk and Dosa (2014) show performance of chip seal
is function of initial condition and climate Carvalho et al. (2011) compared control and treatment sections from SPS-3 study in LTPP database Pierce and Muench (2009) evaluated long-term effect of DBR in State of Washington Data does exist to support performance evaluation (LTPP and State PMS)
Background, Cont’d
Performance Measures Defined as: Metrics that reflect degree of achievement of pavement asset towards meeting specific goals
Evaluation of current condition of pavements Long-term trends in pavement condition
Assessment of decisions made to achieve specific goals (e.g., minimize LCC)
Performance Measures, Cont’d • Many measures in current use Individual distresses Composite indexes Cost based e.g., Asset Sustainability Index, etc.
Others e.g., friction, Remaining Service Interval, etc.
Performance Measures, Cont’d • For the purposes of this presentation Demonstrate; cracking, roughness (IRI), rutting (AC) and faulting (PCC) Why? Measured by most agencies, Required HPMS elements,
NPRM resulting from MAP-21 Captures many decision factors
Modeling Effects of Preservation Treatments • Assessed: Immediate change in condition
Changes in performance
•
Use data to assess effects of preservation on service life and LCC
• Data from State agencies and LTPP program will be presented
Modeling Effects of Preservation, Cont’d Distress
Change in condition modeled as a function of initial values of performance measures
Distress
Time
Time
Differences
Distress
Time
Considerable variance in both dependent and independent variables Initial Condition
Used Deming Regression to account for this
Modeling Effects of Preservation, Cont’d Accounting for errors in condition data
Modeling Effects of Preservation, Cont’d Examples: assessing change in IRI
(a)
(b)
Diamond Grind W/ DBR Figure 1. Change in roughness from; (a) diamond grind, andDiamond (b) diamond Grind grind with DBR
Modeling Effects of Preservation, Cont’d Examples: assessing change in IRI
Modeling Effects of Preservation, Cont’d Effect of variance in data
Modeling Effects of Preservation
Distress
Time
Distress
Time
Time
Each Figure on Left = Single Pavement Segment = Individual Distress Measurement = Regression Line = Distress Growth Rate for Single Pavement Segment = Growth Rate as a Function of Initial Condition
Calculate the slope of the regression line for each segment, shown as in the figure on the right
Distress Growth Rate Over Time (Estimate of Performance)
Distress
Change in performance – function of many variables
Initial Condition
Performance Measure Value
Modeling Effects of Preservation, Cont’d • Robust regression used to model distress growth over time for each segment 9 Used to account 8 7 for potential 6 outliers 5 4 3 Data
2
Linear (Data) Residuals
1 0 0
1
2
3 Year
4
5
6
Modeling Effects of Preservation, Cont’d Examples: assessing change in rutting performance
Modeling Effects of Preservation, Cont’d Examples: assessing change in performance
Transverse Crack Growth Rate as a Function of Many Variables
Modeling Effects of Preservation, Cont’d Examples: assessing change in rutting Rut Growth Rate Following Thin Overlay Using LTPP Data Function of precipitation, freeze-thaw cycles, Average Temperature ESALS & Structural Number
Modeling Effects of Preservation, Cont’d Roughness (IRI) Pavement Type and Preservation Treatment
Initial Cond. Change
Asphalt Pavements Thin Asphalt Both Overlay
Cracking (at least one cracking Type)
Rutting
LongTerm Perf.
Initial Cond. Change
LongTerm Perf.
Initial Cond. Change
LongTerm Perf.
LTPP
Both
Both
Both
State DOT
Chip Seal
None
LTPP
Both
Both
None
None
Micro Surfacing
None
DNA
Both
DNA
DNA
DNA
Modeling Effects of Preservation, Cont’d • Changes in immediate condition and performance can be used to assess: Effect of preservation on service life Effect of preservation on life cycle costs
Example using State data to calculate change in service life and LCC following a thin overlay – using only IRI as performance measure
Modeling Effects of Preservation, Cont’d 180 160 IRI Threshold
140
IRI (in/mile)
120 100
Immediate Condition Jump
80 60
Improvement in Performance
40
Service Life Extension
Control (Untreated Pavement) Preserved Pavement
20 0 10
12
14
16 18 20 22 24 Effective Pavement Age (Years)
26
28
30
Modeling Effects of Preservation, Cont’d • IRI performance model from agency 𝐼𝑅𝐼 𝑡 = 40𝑒 0.05𝑡 • Change in condition calculated using agency data • No change in performance found from data • Compare service life and equivalent annual uniform costs for 3 pavements with differing initial IRI values: 85, 100 and 115 in/mile
Modeling Effects of Preservation, Cont’d 220
Life Extension Pavement 3
200
IRI (in/mile)
180 Life Extension Pavement 1
160
Life Extension Pavement 2
140 Threshold
120 100
80 Control Pavement Pavement 2
60 40
10
15
Pavement 1 Pavement 3
20 25 Effective Pavement Age (years)
30
Modeling Effects of Preservation, Cont’d Initial Effective Pavement IRI Age When OL is (in/mile) Applied (years) 85 15.1 100 18.3 115 21.1
IRI After Overlay (in/mile) 72.5 74.0 75.5
𝑟∗ 1+𝑟 𝑡 𝐸𝑈𝐴𝐶 = 𝐶𝑜𝑠𝑡 1+𝑟 𝑡 −1 r = discount rate (3 percent) t = analysis period (varies)
Effective Pavement Effective Age Life Age After Overlay When IRI of 120 Extension (years) in/mile is Reached (years) 11.9 25.9 3.9 12.3 28.9 6.7 12.7 31.0 9.1
𝐸𝑈𝐴𝐶85 = $20,715 𝐸𝑈𝐴𝐶100 = $12,506 𝐸𝑈𝐴𝐶115 = $9,577
Conclusions • Many agencies collect and store data required to assess effectiveness of preservation. Other sources of data (e.g., LTPP) can be used to supplement agency data • Analysis of pavement condition data requires techniques not traditionally in pavement literature • Definition and implementation of performance measures, including development of models, are key steps for evaluating effects of preservation treatments
Thank You! James Bryce, Ph.D. Senior Consultant, Amec Foster Wheeler, 12000 Indian Creek Court, Beltsville MD, 20705 1 (301) 210 5105 ext. 36 Email: [email protected]
James Bryce & Gonzalo Rada Amec Foster Wheeler National Pavement Preservation Conference Nashville, TN October 13, 2016
Outline • Background • Performance measures • Modeling the effects of preservation Performance
Service Life Life Cycle Costs
• Conclusions
Background • Based on research conducted under NCHRP Project 14-33 to: Identify and/or develop pavement performance measures that consider contributions of preservation to performance, service life, and life-cycle costs Prepare guide document to facilitate implementation of measures by highway agencies
Background, Cont’d • Preservation treatments are applied to: Preserve an existing roadway
Slow future deterioration Maintain and improve its functional condition
• No substantial increase to structural capacity of
pavement Performance measures provide means for assessing effectiveness of preservation
Background, Cont’d • Considerable amount of literature Mamlouk and Dosa (2014) show performance of chip seal
is function of initial condition and climate Carvalho et al. (2011) compared control and treatment sections from SPS-3 study in LTPP database Pierce and Muench (2009) evaluated long-term effect of DBR in State of Washington Data does exist to support performance evaluation (LTPP and State PMS)
Background, Cont’d
Performance Measures Defined as: Metrics that reflect degree of achievement of pavement asset towards meeting specific goals
Evaluation of current condition of pavements Long-term trends in pavement condition
Assessment of decisions made to achieve specific goals (e.g., minimize LCC)
Performance Measures, Cont’d • Many measures in current use Individual distresses Composite indexes Cost based e.g., Asset Sustainability Index, etc.
Others e.g., friction, Remaining Service Interval, etc.
Performance Measures, Cont’d • For the purposes of this presentation Demonstrate; cracking, roughness (IRI), rutting (AC) and faulting (PCC) Why? Measured by most agencies, Required HPMS elements,
NPRM resulting from MAP-21 Captures many decision factors
Modeling Effects of Preservation Treatments • Assessed: Immediate change in condition
Changes in performance
•
Use data to assess effects of preservation on service life and LCC
• Data from State agencies and LTPP program will be presented
Modeling Effects of Preservation, Cont’d Distress
Change in condition modeled as a function of initial values of performance measures
Distress
Time
Time
Differences
Distress
Time
Considerable variance in both dependent and independent variables Initial Condition
Used Deming Regression to account for this
Modeling Effects of Preservation, Cont’d Accounting for errors in condition data
Modeling Effects of Preservation, Cont’d Examples: assessing change in IRI
(a)
(b)
Diamond Grind W/ DBR Figure 1. Change in roughness from; (a) diamond grind, andDiamond (b) diamond Grind grind with DBR
Modeling Effects of Preservation, Cont’d Examples: assessing change in IRI
Modeling Effects of Preservation, Cont’d Effect of variance in data
Modeling Effects of Preservation
Distress
Time
Distress
Time
Time
Each Figure on Left = Single Pavement Segment = Individual Distress Measurement = Regression Line = Distress Growth Rate for Single Pavement Segment = Growth Rate as a Function of Initial Condition
Calculate the slope of the regression line for each segment, shown as in the figure on the right
Distress Growth Rate Over Time (Estimate of Performance)
Distress
Change in performance – function of many variables
Initial Condition
Performance Measure Value
Modeling Effects of Preservation, Cont’d • Robust regression used to model distress growth over time for each segment 9 Used to account 8 7 for potential 6 outliers 5 4 3 Data
2
Linear (Data) Residuals
1 0 0
1
2
3 Year
4
5
6
Modeling Effects of Preservation, Cont’d Examples: assessing change in rutting performance
Modeling Effects of Preservation, Cont’d Examples: assessing change in performance
Transverse Crack Growth Rate as a Function of Many Variables
Modeling Effects of Preservation, Cont’d Examples: assessing change in rutting Rut Growth Rate Following Thin Overlay Using LTPP Data Function of precipitation, freeze-thaw cycles, Average Temperature ESALS & Structural Number
Modeling Effects of Preservation, Cont’d Roughness (IRI) Pavement Type and Preservation Treatment
Initial Cond. Change
Asphalt Pavements Thin Asphalt Both Overlay
Cracking (at least one cracking Type)
Rutting
LongTerm Perf.
Initial Cond. Change
LongTerm Perf.
Initial Cond. Change
LongTerm Perf.
LTPP
Both
Both
Both
State DOT
Chip Seal
None
LTPP
Both
Both
None
None
Micro Surfacing
None
DNA
Both
DNA
DNA
DNA
Modeling Effects of Preservation, Cont’d • Changes in immediate condition and performance can be used to assess: Effect of preservation on service life Effect of preservation on life cycle costs
Example using State data to calculate change in service life and LCC following a thin overlay – using only IRI as performance measure
Modeling Effects of Preservation, Cont’d 180 160 IRI Threshold
140
IRI (in/mile)
120 100
Immediate Condition Jump
80 60
Improvement in Performance
40
Service Life Extension
Control (Untreated Pavement) Preserved Pavement
20 0 10
12
14
16 18 20 22 24 Effective Pavement Age (Years)
26
28
30
Modeling Effects of Preservation, Cont’d • IRI performance model from agency 𝐼𝑅𝐼 𝑡 = 40𝑒 0.05𝑡 • Change in condition calculated using agency data • No change in performance found from data • Compare service life and equivalent annual uniform costs for 3 pavements with differing initial IRI values: 85, 100 and 115 in/mile
Modeling Effects of Preservation, Cont’d 220
Life Extension Pavement 3
200
IRI (in/mile)
180 Life Extension Pavement 1
160
Life Extension Pavement 2
140 Threshold
120 100
80 Control Pavement Pavement 2
60 40
10
15
Pavement 1 Pavement 3
20 25 Effective Pavement Age (years)
30
Modeling Effects of Preservation, Cont’d Initial Effective Pavement IRI Age When OL is (in/mile) Applied (years) 85 15.1 100 18.3 115 21.1
IRI After Overlay (in/mile) 72.5 74.0 75.5
𝑟∗ 1+𝑟 𝑡 𝐸𝑈𝐴𝐶 = 𝐶𝑜𝑠𝑡 1+𝑟 𝑡 −1 r = discount rate (3 percent) t = analysis period (varies)
Effective Pavement Effective Age Life Age After Overlay When IRI of 120 Extension (years) in/mile is Reached (years) 11.9 25.9 3.9 12.3 28.9 6.7 12.7 31.0 9.1
𝐸𝑈𝐴𝐶85 = $20,715 𝐸𝑈𝐴𝐶100 = $12,506 𝐸𝑈𝐴𝐶115 = $9,577
Conclusions • Many agencies collect and store data required to assess effectiveness of preservation. Other sources of data (e.g., LTPP) can be used to supplement agency data • Analysis of pavement condition data requires techniques not traditionally in pavement literature • Definition and implementation of performance measures, including development of models, are key steps for evaluating effects of preservation treatments
Thank You! James Bryce, Ph.D. Senior Consultant, Amec Foster Wheeler, 12000 Indian Creek Court, Beltsville MD, 20705 1 (301) 210 5105 ext. 36 Email: [email protected]
