woodplc.com
NCHRP Project 14-38: Guide for Timing of Asphalt-Surfaced Pavement Preservation
Gonzalo R. Rada, Ph.D., P.E.
AASHTO MaC Meeting – Charlotte, NCJuly 24, 2018
Presentation Outline
1. Project background2. Formulation of timing framework3. Development of timing framework4. Summary & conclusions
2
Project Background
Pavement Preservation & Performance
4
Pavement Age
xx Threshold Condition
Preservation TimingPreservation timing is defined as point in time when benefits and costs are optimized• Majority of approaches based on Cost-Benefit
analysis – definitions of benefit and cost vary ► Require estimation of pavement performance and
expected costs
• Timing is affected by condition and non-condition factors► Factors that affect pavement
performance affect timing
5
Cos
t / B
enef
itTime
Project Objectives
To develop guide for identifying timing for preservation of asphalt-surfaced pavements considering:
► Condition factors► Pavement condition at time of treatment► IRI, cracking, rutting, etc.
► Non-condition factors► Treatment type► Pavement structure► Subgrade soil► Traffic► Climate
6
Formulation of Timing Framework
Literature Review, Survey & Interviews
• Objective:► To gather relevant information relating to
timing of pavement preservation treatments► Timing of treatments► Definition of costs and benefits► Pavement performance after treatment
• Findings:► Preservation timing problem lends itself to
CBA► Agency pavement management goal is
providing acceptable LOS for lowest cost► Leverages existing agency practices such as
life cycle cost analysis
8
Defining Costs
► Approaches:► Net present value ► Equivalent uniform annual costs
► Components:► Agency costs – e.g., treatment and traffic control costs► User costs – e.g., delays and safety costs associated with
crashes► Life cycle cost analysis approaches
Timing framework needs to accommodate multiple cost approaches and components
0
20
40
60
80
100
0 10 20 30 40
Pave
men
t Con
ditio
n
Time
Condition Threshold
Life Extension
Change in Performance
Defining Benefits
► Approaches:► Service life extension► Change in performance
► No preference cited in literature to either approach, but…
► Important advantages to “change in performance” approach
Recommended Timing Framework
► Performance and benefit models
► Cost models
► Optimization function
► Uncertainties
DefineIndicators and cutoff valuesPerformance curves and cost functionUncertainties on or of each inputEffect of preservation on performance & costs
For time t = t + 1, predict pavement condition
Calculate distribution of benefits (i.e., the range of possible benefits
accounting for uncertainties)
Calculate distribution of costs (i.e., the range of possible costs
accounting for uncertainties)
Calculate objective function using Monte Carlo simulation
Store expected value, 5th, and 95th
percentile values
Cutoff values reached?
Output: Preservation Timing
Time t = 0
YesNo
Development of Timing Framework
Pavement Data
Agency No. of Years of Condition Data
Thin AC OL
Chip Seal Micro-surfacing
Slurry Seal
MD-SHA 15 VDOT 8 KSDOT 30+ (entire database) ITD 15 UTDOT 3 TXDOT 10 OHDOT 30+ (entire database) TNDOT 16 MEDOT 16 LADOTD 15 LTPP 15
indicates data received Blank cell indicates no data were provided
Other Data
• Developed climate and subgrade resilient modulus databases► Climate data for every county from NASA’s Modern-Era
Retrospective Analysis for Research Application (MERRA) Climate database
► Subgrade soil from Natural Resources Conservation Service (NRCS) maps and further development in NCHRP Project 9-23A
120 ° W
110 ° W 100
° W 90
° W 80
° W 70
° W
30 ° N
40 ° N
50 ° N
30
40
50
60
70
80
Tem
pera
ture
in D
egre
es F
ahre
nhei
t
120 ° W
110 ° W 100
° W 90
° W 80
° W 70
° W
30 ° N
40 ° N
50 ° N
5
10
15
20
25
30
35
Mod
ulus
(tho
usan
ds o
f Pou
nds
per S
quar
e In
ch)
Pavement Performance Models
► Example models developed using State DOT and LTPP data plus default databases
► Models developed for more frequently used treatments and measures (based on NCHRP 14-33)
IRI Rutting Transverse Cracking
Fatigue Cracking
NWPLong.
CrackingThin Overlay
Chip Seal
Microsurfacing
Slurry Seal
Immediate Change in Condition
• Trends generally consistent across DOTs
0 0.1 0.2 0.3 0.4 0.5 0.6
Rutting before Overlay (inches)
-0.2
0
0.2
0.4
0.6
Rut
ing
befo
re m
inus
Rut
ting
Aft
er O
verl
ay (
inch
)
LTPP DataAgency 1 DataAgency 2 DataLine of Maximum ImprovementRut Depth After = 0.1 Inch
Change in Performance
► Chip seals do not have immediate effect on IRI, but… they do affect IRI growth
IRI growth rate = f(….)
• Mean annual freeze thaw cycles
• Mean annual air temperature
• IRI before treatment• Subgrade modulus• Mean annual
precipitation• Average annual daily
trafficMeasured Values (inches/mile/year)Pr
edic
ted
Valu
es (f
inch
es//m
ile/y
ear)
-5 0 5 10 15 20
-5
0
5
10
15
20
Adjusted r2 = 0.32
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8Correlation Between Factor and Predicted Value
Mean Annual Number of Freeze Thaw Cycles
Preservation Indicator
Subgrade Resilient Modulus
IRI Before Chip SealMean Annual Air Temperature
Mean Annual Precipitation
AADT
Benefit Definition
► Approaches:► Service life extension► Benefit area = Area A / (Area A + Area B)
Time (Years)
Perf
orm
ance
M
easu
re V
alue Control Curve
Treatment Curve Case 1
Treatment Curve Case 2
Area B
Service life extensionThreshold condition
Cost Models
► Cost data provided not consistent► Contract cost versus expected treatment cost versus
expected preservation cost (i.e., preservation cost independent of treatment type)
► Attempted to assess variation in cost with pavement condition ► Kansas data show increased cost with worsening condition
for thin overlays
Flexible approach chosen to accommodate costs in timing framework
Optimizing Costs and Benefits
Approaches – minimizing:1. Cost / benefit ratio2. Distance “d” from “hypothetical” optimal solution
Most Benefit
Least Benefit
Lowest Cost
Highest Cost
d
Distance approach recommended because CB can be adjusted independently and their contributions weighted separately
Uncertainties
► Performance models – computed during model development effort
► Cost models – estimates to be developed based on information available to agency
IRI Before Overlay (inches/mile)0 50 100 150 200 250
Impr
ovem
ent i
n IR
I (in
ches
/mile
)
-100
-50
0
50
100
150
Improvement = 0.63*IRI B e f o r e - 31
RMSE = 17
DataBest Fit Robust Regression
Immediate Change in IRI
Summary & Conclusions
Timing Framework
• Example performance models that explain effects of preservation were developed for more frequently used treatments and measures
• Benefit was defined as area encompassed by treated and untreated curves
• Flexible approach to costs was selected to accommodate multiple cost approaches and components
• Distance function recommended for optimizing costs and benefits
• Uncertainties can be estimated for performance models, but qualitative approach required for costs
Implementation Guide
Year0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Obj
ectiv
e Fu
nctio
n V
alue
0.5
1
1.5
2
2.5
3
50 t h Percentile
5 t h Percentile
95 t h Percentile
Select Performance MeasuresActivity 1
Establish Performance ModelsActivity 2
Establish Cost ModelsActivity 4
Select Preservation TimingActivity 5
Estimate Preservation BenefitsActivity 3
Process
Output
woodplc.com