Project Assessment Manual

SHRP2 R23
Using Existing Pavement in Place and Achieving Long Life
Project Assessment Manual
June 29, 2013

0
Project Assessment Manual Topics

Topic
Section 1 Introduction
Explanation as to why this manual was developed
How to use the manual
Assessment data categories
Overall Assessment Scheme
Section 2 Pavement Distress Survey
Purpose
Measurement methods
Analysis tools
Section 3 Pavement Rut Depth and Roughness
Purpose
Measurement methods
Analysis tools
Section 4 Nondestructive Testing via the Falling Weight Deflectometer (FWD)
Purpose
Measurement method
Analysis tools
Section 5 Ground Penetrating Radar (GPR)
Purpose
Measurement method
Analysis tools
Section 6 Pavement Cores
Purpose
Measurement method
Analysis tools
Section 7 Dynamic Cone Penetrometer (DCP)
Purpose
Measurement method
Analysis tools
Section 8 Subgrade Soil Sampling and Tests
Purpose
Measurement methods
Analysis tools
Section 9 Traffic Loads for Design
Purpose
Measurement method
Analysis tools
Page
3
16
21
36
53
54
62
67
Project Assessment Manual Topics (continued)

Topic
Section 10 Construction Productivity and Traffic Impacts
Purpose
Measurement methods
Analysis tools
Section 11 Life Cycle Assessment (Environmental Accounting)
Purpose
Measurement methods
Analysis tools
Section 12 Miscellaneous Material Properties
Page
73
103
117
Section 1
Introduction
1.1 Why This Assessment Manual?
This assessment manual was prepared to aid the process of renewing existing pavements so
that long lives can be achieved. To achieve this goal a systematic collection of relevant
pavement-related data is needed. Further, such data needs to be organized to maximize the
usefulness in pavement decision-making process. To that end, this manual will help.
The types of data collection contained in this manual range from basic information such as a
distress survey to insights on traffic impacts. The last section provides information on life
cycle assessments (environmental accounting). This type of assessment is receiving
increasing usage and is likely to be widely applied in the future.
1.2 How to Use the Manual

The use of the manual is to compliment the design tools developed by the SHRP2 R23 study.
The types of data critical for making pavement-related decisions are described along with
methods (analysis tools) for organizing the information for decision-making. It is not
assumed that all data categories will be collected or assessed for a specific renewal project.
Rather, the manual is designed as a reference document that provides information relevant
to all renewal strategies considered in the SHRP2 R23 project.
1.3 Assessment Data Categories

There are 10 categories of data contained in this manual. These are:
Pavement distress surveys
Pavement rut depths and roughness
Nondestructive TestingFalling Weight Deflectometer
Ground Penetrating Radar
Pavement cores
Dynamic Cone Penetrometer
Subgrade soil sampling and tests
Traffic Loads for Design
Traffic impacts
Life cycle assessment
Each data category is structured much the same, namely by (1) the purpose for collecting the
data, (2) applicable standards, definitions and data organization recommendations, and (3)
analysis tools.
3
1.4 Overall Assessment Scheme

The overall assessment scheme performed by the user can range from rather basic information
about the existing and proposed pavement structure to substantially more detailed data and
analyses. The basic scheme is illustrated in Figure 1.1.
Identify Pavement Type for

Existing Pavement Structure
Identify Distress Categories
2
3
Collect other pavement related

data and conduct analyses
Select Renewal Pavement Type

Option
Apply Recommended Renewal

Actions and Design
Figure 1.1 Outline of Assessment Scheme

The first three boxes (1 through 3) shown in Figure 1.1 are addressed in this assessment
manual with that information being applied to the processes shown in the last two boxes (4
and 5). The types of input data include the distress types associated with the existing
pavement structure, characterization of future traffic (in terms of ESALs and ADT), subgrade
characterization (strength or stiffness) and more.
Section 2
Pavement Distress Survey
2.1 Purpose
This section overviews the use of a pavement distress survey for aiding pavement
assessment decisions.
2.2 Measurement Methods

This subsection is used to describe definitions and standards applicable for pavement
distresses and provides a way to organize such information.
(i) Pavement Distress Measurements: ASTM D6433-07 Standard Practice for Roads and
Parking Lots Pavement Condition Index Surveys.
(ii) Distress Identification Manual for the Long-Term Pavement Performance Program:
FHWA-RD-03-031, June 2003.
(iii) Discussion
Pavement distress data can be used for numerous purposes but three are noted: (1)
establish pavement reconstruction, rehabilitation, and maintenance priorities, (2)
determine rehabilitation and maintenance strategies, and (3) predict pavement
performance. This type of information is a key element for decision-making associated with
pavement renewal options.
McCullough (1971) provided a detailed description of three basic pavement distress groups,
associated modes, and examples as shown in Table 2.1. Most all distress survey schemes
use a subset of fracture, distortion, and/or disintegration.
Upon closer inspection of Table 2.1 for flexible pavements, two of thesefracture and
disintegration are responsible for most pavement rehabilitation and maintenance actions.
More specifically these can be categorized by fatigue, transverse cracking, and
stripping/raveling. Tables 2.2, 2.3, and 2.4 provide templates for flexible pavement distress
data collection. It is assumed that cores will be an integral part of the pavement distress
examination hence locations would logically be organized by mileposts or other appropriate
location referencing system. For multilane highways, this information can be collected for
the design lane or all lanes in one directionas per project requirements.
Table 2.1 Distress Groups (after McCullough, 1971)

Distress Group
Fracture
Distress Mode
Cracking
Examples of Distress Mechanism

Excessive loading
Repeated loading (i.e., fatigue)
Thermal changes
Moisture changes
Slippage (horizontal forces)
Shrinkage
Spalling
Excessive loading
Repeated loading (i.e., fatigue)
Thermal changes
Moisture changes
Distortion
Permanent
Deformation
Excessive loading
Time-dependent deformation (e.g., creep)
Densification (i.e., compaction)
Consolidation
Swelling
Frost
Faulting
Excessive loading
Densification (i.e., compaction)
Consolidation
Swelling
Disintegration
Stripping
Adhesion (i.e. loss of bond)

Chemical reactivity
Abrasion by traffic
Raveling
and Scaling
Adhesion (i.e. loss of bond)

Chemical reactivity
Abrasion by traffic
Degradation of aggregate
Durability of binder
The following distress types should be measured and recorded if present on the existing
pavement:
Flexible Pavement Distress (definitions from or modified after LTPP Distress Manual, Miller
and Dellinger, 2003):
1. Fatigue cracking: Occurs in areas subjected to repeated traffic loadings (wheel paths).
Can be a series of interconnected cracks in early stages of development. Develops into
many-sided, sharp-angled pieces, usually less than 0.3 m on the longest side,
characteristically with a chicken wire/alligator pattern, in later stages.
2. Transverse cracking: Cracks that are predominantly perpendicular to the pavement
centerline.
3. Stripping or raveling: Wearing away of the pavement surface caused by the dislodging
of aggregate particles and loss of asphalt binder. Raveling ranges from loss of fines to
loss of some coarse aggregate and ultimately to a very rough and pitted surface with
obvious loss of aggregate. This study expands the definition to identification of
stripping/raveling in the surface layer to include stripping that may be occurring in lower
HMA layers in the pavement structure.
Rigid Pavement Distress for JPCP, JRCP, and CRCP (definitions from or modified after LTPP
Distress Manual, Miller and Dellinger, 2003 with the exception of ASR cracking):
1. Pavement Cracking: Pavement cracking includes all major types of cracks that can
occur in a slab. This can include corner breaks, longitudinal and transverse cracking as
defined by Miller and Dellinger, 2003. Corner break cracks intersect the adjacent
transverse and longitudinal joints at approximately a 45 angle. Longitudinal and
transverse cracking are parallel and transverse to the centerline, respectfully.
2. Joint Faulting: Joint faulting is the difference in elevation across a joint or crack.
3. Materials Caused Distress: (1) D-Cracking: Closely spaced crescent-shaped hairline
cracking pattern; occurs adjacent to joints, cracks, or free edges; dark coloring of the
cracking pattern and surrounding area; sometimes referred to as durability cracking, and
(2) Alkali-Silica Reactivity (ASR) Cracking: Cracking of the PCC which can be easily
confused with D-cracking or shrinkage cracking.
4. Pumping: Pumping is the ejection of water from beneath the pavement. In some
cases, detectable deposits of fine material are left on the pavement surface, which were
eroded (pumped) form the support layers and have stained the surface.
5. Punchouts: The area enclosed by two closely spaced (usually < 0.6 m) transverse
cracks, a short longitudinal crack, and the edge of the pavement or a longitudinal joint.
Also includes Y cracks that exhibit spalling, breakup, or faulting.
2.2.1 Pavement Distress Data Templates
The templates for specific pavement distress types follow.
Table 2.2 Template for Flexible Pavement DistressFatigue Cracking

Location
Depth
Distress
(milepost) HMA Base
Fatigue Cracking
(in)
(in)
Severity2 Extent1
Depth of Fatigue Cracks4
(measured from the pavement
surface)
Low
Moderate
High
Notes:
1. Extent of fatigue cracking is based on % of wheelpath areas.
2. Severity of fatigue cracking is low, medium, and high. (1) Low = None or only a few connecting cracks; cracks are not
spalled or sealed; pumping not evident, (2) Moderate = Interconnected cracks forming a complete pattern; cracks
may be slightly spalled; cracks may be sealed; pumping is not evident, and (3) High = Moderately or severely spalled
interconnected cracks forming a complete pattern; pieces may move when subjected to traffic; cracks may be sealed;
pumping may be evident. The severity definitions are from the LTPP Distress Identification Manual (Miller and
Bellinger, 2003).
3. Record extent for each level of severity.
4. Depth of fatigue cracks can be full depth or top down cracking. This should be determined by the use of pavement
cores.
Low Severity
Moderate Severity
High Severity
(Source: Pavement Interactive)
(Source: N. Jackson)
Figure 2.1 Illustrations of Fatigue Cracking Severity Levels
Table 2.3 Template for Flexible Pavement DistressTransverse Cracking

Location
Depth
Distress
(milepost) HMA
Base
Transverse Cracking
2
1
(in)
(in)
Severity
Extent
Depth of Transverse Cracks
(measured from the pavement
surface)
Low
Moderat
e
High
Notes:
1. Extent of transverse cracking is based on the number of cracks per 100 ft.
2. Severity of transverse cracking is low, medium, and high. (1) Low = Unsealed cracks with a mean width 6 mm;
sealed cracks with sealant material in good condition and with a width that cannot be determined, (2) Moderate =
Cracks with mean widths 6 mm and 19 mm; or any cracks with a mean width 19 mm and adjacent low severity
random cracking, and (3) High = Cracks with a mean width of 19 mm; or cracks with a mean width 19 mm and
adjacent to moderate to high severity random cracking. The severity definitions are from the LTPP Distress
Identification Manual (Miller and Bellinger, 2003).
3. Record extent for each level of severity.
4. Depth of fatigue cracks might be full depth of the HMA or top down cracking. This can only be determined by the use
of pavement cores.
Moderate Severity
Moderate to High Severity High Severity
(Source: WSDOT)
Figure 2.2 Illustrations of Transverse Cracking Severity Levels
Table 2.4 Template for Flexible Pavement DistressStripping/Raveling

Location
Depth
Distress
(milepost) HMA Base
Stripping/Raveling
(in)
(in)
Extent
Full depth stripping/raveling or
(% of surface area) confined to the wearing surface only?
Observation must be based on cores.
Note: Severity levels are not applicable for stripping. Either it exists or does not.
Photo source: WSDOT
Figure 2.3 Illustration of Raveling

Using Table 2.1 again, the most important Jointed Plain Concrete Pavement (JPCP) distress
types which initiates PCCP renewal actions are fracture (slab or pavement cracking),
distortion (faultingtypically at transverse contraction joints), and disintegration which
includes materials caused distresses of D-cracking and ASR cracking. These are shown in
Tables 2.5, 2.6, 2.7, and 2.8. Tables 2.9 and 2.10 apply to CRCP and composite pavements.
Table 2.5 Template for Rigid Pavement DistressJPCP or JRCPPavement Cracking

Location
Depth
Distress
(milepost) PCC Slab
Base
Pavement or Slab Cracking
1
(in)
Type Thick (in) % Slabs with Multiple
Comments
2
Cracks
Notes
1. Three types of base underlying PCC: (1) Granular Base, (2) Cement Treated Base, or (3) Asphalt Treated Base.
10
2. Percentage of slabs with two or more pavement cracks.
Examples of PCC Slab Multiple Cracks

Photo sources: PI and J. Mahoney
Figure 2.4 Illustrations of PCC Slabs with Multiple Cracks
Table 2.6 Template for Rigid Pavement DistressJPCP or JRCPFaulting

Location
Depth
Distress
(milepost) PCC Slab
Base
Faulting
1
(in)
Type
Thick
Avg Fault Depth (in)
Comments
(in)
Note
Average Fault 0.25 to 0.5 in.
Average Fault 0.5 in.
Figure 2.5 Illustrations of Various Levels of Joint Faulting

11
Location
(milepost)
Table 2.7 Template for Rigid Pavement DistressD-Cracking

Depth
Distress
PCC
Base
D-Cracking
1
2
Slab
Type Thick (in) Severity
Extent3
Comments
(in)
Low
Moderat
e
High
Notes
2. Severity of D-cracking is low, medium (moderate), and high. (1) Low = D-cracks are tight, with no loose or missing
pieces, and no patching is in the affected area, (2) Moderate = D-cracks are well-defined, and some small pieces are loose
or have been displaced, and (3) High = D-cracking has a well-developed pattern, with a significant amount of loose or
2
missing material. Displaced pieces, up to 0.1 m , may have been patched.
3. Extent is based on the amount of cracks or joints that exhibit D-cracking. This definition of extent is different than used
by LTPP.
Low Severity
Low Severity
High Severity
(Source: PI and C.L. Monismith)
Figure 2.6 Illustrations of D-Cracking Severity Levels
12
Table 2.8 Template for Rigid Pavement DistressASR Cracking

Location
Depth
Distress
(milepost) PCC Slab
Base
ASR Related Cracking
(in)
Type1 Thick (in)
Does ASR Cracking
How was ASR detected
Apply to this Pavement?
or measured?
Yes or No
Note
1. Three types of base underlying PCC: (1) Granular Base, (2) Cement Treated Base, or (3) Asphalt Treated
Base.
Early State of Cracking
Advanced Stage of Cracking
Figure 2.7 Illustrations of ASR Cracking Severity Levels
Table 2.9 applies to Continuously Reinforced Concrete Pavement (CRCP). A critical distress
for CRCP is punchouts (which falls under fracture in Table 2.1).
Location
(milepost)
Table 2.9 Template for Rigid Pavement DistressCRCPPunchouts

Depth
Distress
PCC
Base
Punchouts
Slab
Type1
Thick (in) No./mile
Comments
(in)
Note 1: Three types of base underlying PCC: (1) Granular Base, (2) Cement Treated Base, or (3) Asphalt
Treated Base.
13
Advanced Stage for a Punchout

(Source: PI and FHWA)
Figure 2.8 Illustration of a CRCP Punchout
Location
(milepost)
Table 2.10 Composite Pavement Distress1

Depth
Distress4
HMA
PCC
Describe
Comments
Surfacing PCC PCC Slab
condition of
Base
Base
2
3
(in)
surface
course
Type Thick (in.) Type Thick (in)
Poor Condition
Very Poor
Condition
Notes:
1. Composite pavement definition assumes a flexible (HMA) layer overlies PCC.
2. Three types of PCC pavement: (1) JPCP, (2) JRCP, or (3) CRCP.
4. Distress is broadly defined for composite pavements. The only initial information available to the user is the surface
condition which can include a range of distress typesmost likely cracking.
Other PCCP distress types can be important and such information collected and used; however,
the distress types in the preceding tables were judged as the most critical for pavement
renewal decision-making.
2.2.2 Drainage Conditions
An assessment of the existing pavements subsurface drainage is important in making
pavement renewal decisions. The following factors, if observed, suggest that subsurface
drainage may be an issue and corrective actions needed for the renewal design process:
Pumping
PCC joint or crack faulting
Standing water in shallow ditches
Use of cement stabilized base under PCC.
14
2.3 Analysis Tools

How pavement distress data is specifically used in the renewal decision-making process is
covered in a separate project report.
2.4 References
McCullough, B.F. (1971), "Distress Mechanisms-General," Special Report No. 126, Highway
Research Board, National Academy of Sciences, Washington, DC.
Miller, J.S. and Bellinger, W.Y. (2003), Distress Identification Manual for the Long-Term
Pavement Performance Program (Fourth Edition), Report FHWA-RD-03-031, Office of
Infrastructure Research and Development, Federal Highway Administration, McLean, Virginia,
June 2003.
Stark, D. (1994), Handbook for the Identification of Alkali-Silica Reactivity in Highway
Structures, SHRP-C-315, Strategic Highway Research Program, Washington, DC, originally
printed in 1994 but updated. http://leadstates.transportation.org/asr/library/C315/index.stm#f
15
Section 3
Pavement Rut Depth and Roughness (Profile)
3.1 Purpose
This section overviews the use of pavement rut depths and roughness for aiding pavement
assessment decisions.

This subsection is used to describe definitions and standards applicable for pavement rut and
roughness measurements.
(i) Rut Depth Measurements
NCHRP Synthesis 334 (McGhee, 2004) notes that 46 State DOTs collect automated rut depth
measurements almost always associated with roughness measurements. McGhee (2004) and
SHRP (1993) define rut depth as the longitudinal surface depressions in the wheel paths.
Figure 3.1 will helps to define lateral locations of a typical highway lane (from AASHTO, 2001).
Figure 3.2 shows how rut depths are measured with automated equipment.
Figure 3.1 Sketch Illustrating Wheel Paths and Between Wheel Path
(from McGhee, 2004 and AASHTO, 2001)
16
Figure 3.2 Rut Depth Measurements (from McGhee, 2004 and AASHTO, 2000)
(ii) IRI Measurements
McGhee (2004) defines pavement roughness as the deviation of a surface from a true planar
surface with characteristic dimensions that affect vehicle dynamics and ride quality. ASTM
E1926-08 (Standard Practice for Computing International Roughness Index of Roads from
Longitudinal Profile Measurements) defines International Roughness Index (IRI) as the
pavement roughness index computed from a longitudinal profile measurement using a
quarter-car simulation at a simulation speed of 80 kph (50 mph). Further, ASTM E1926 notes
IRI is reported in either meters per kilometer (m/km) or inches per mile (in/mile).
3.3 Analysis Tools

Some of the analysis tools available include allowable rut depths and recommended IRI levels
are shown in Tables 3.1, 3.2, and 3.3.
A study done in Wisconsin (Start et al, 1998) found for state highways with speed limits greater
than 45 mph, hydroplaning related accidents significantly increased when rut depths were 0.3
inches or greater. State DOTs such as the Washington State DOT use a rehabilitation trigger
level of 0.4 inches (10 mm). The Texas DOT notes in their Hydraulic Design Manual that water
depths of 0.2 inches or greater and Fwa (2006) a rut depth of 0.5 inches or more can create the
potential for hydroplaning. Thus, rut depths less than or equal to 0.5 inches appear to be
reasonable trigger level range for rehabilitation decisions.
17
Table 3.1 Typical Maximum Rut Depths

Pavement Type
Texas DOT
[concern about hydroplaning]
Wisconsin Hydroplaning Study
(Start et al, 1998)
Washington State DOT
Fwa (2006)
[based on hydroplaning]
Shahin (1997)
[from the PAVER Asphalt Distress
ManualPavement Distress
Identification Guide for AsphaltSurfaced Roads and Parking Lots]
Maximum Rut Depth, inches (mm)

0.2
(5)
0.3
(7.6)
0.4
(10)
0.5
(12.5)
Low
0.25 to 0.5
(6 to 13)
Medium
0.5 to 1.0
(13 to 25)
High
1.0
( 25)
Table 3.2 FHWA IRI Criteria (from FHWA, 2006)

Ride Quality
Terms
Good
Acceptable
Not Acceptable
All Functional Classifications

IRI, inches/mi
PSR Rating
(m/km)
< 95
Good
(< 1.5)
170
Acceptable
( 2.7)
> 170
Not Acceptable
(> 2.7)
18
Table 3.3 Earlier FHWA IRI Criteria (FHWA, 1999)

Ride Quality
PSR Rating
IRI, in/mile
National Highway
Terms
(m/km)
System Ride
Quality
Very Good
4.0
60
( 0.95)
Good
3.5 to 3.9
60 to 94
Acceptable
(0.95 to 1.48)
between 0 and
Fair
3.1 to 3.4
95 to 119
170 in./mile
(1.50 to 1.88)
Mediocre
2.6 to 3.0
120 to 170
(1.89 to 2.68)
Poor
Less than
2.5
170
acceptable
( 2.68)
170 in./mile
The IRI criteria used by the FHWA have evolved as illustrated by review of Tables 3.2 and 3.3. In
1999, the most detailed breakdown, suggests that IRI values of less than 60 inches/mile are
quite good and greater than 170 inches/mile poor. Interestingly, many newly paved HMA
projects typically have IRI values close to the 60 inches/mile value. Eventually, the FHWA
simplified their criteria as shown in Table 3.2.
A study conducted on Seattle area urban freeways using driver in-vehicle opinion surveys
(Shafizadeh and Mannering, 2003) confirmed that motorists find pavements with IRI values less
than 170 inches/mile acceptable as to ride quality (85% acceptable). The paper concluded that
that there was no evidence to change federal IRI guides (in essence those shown in Table 3.3).
3.4 References
AASHTO (2000), Standard Practice for Determining Maximum Rut Depth in Asphalt
Pavements, AASHTO Designation PP38-00, American Association of State Highway and
Transportation Officials.
AASHTO (2001), Standard Practice for Quantifying Cracks in Asphalt Pavement Surfaces,
AASHTO Designation PP44-01, American Association of State Highway and Transportation
Officials, April 2001.
FHWA (1999), 1999 Status of the Nations Highways, bridges, and Transit: Conditions and
Performance, Report FHWA-PL-99-017, Federal Highway Administration, November 1999.
FHWA (2006), 2006 Status of the Nations Highways, Bridges, and Transit: Conditions and
Performance, Federal Highway Administration,
http://www.fhwa.dot.gov/policy/2006cpr/chap3.htm
19
Fwa, T. (2006), The Handbook of Highway Engineering, Taylor and Francis Group, CRC Press.
McGhee, K. (2004), Automated Pavement Distress Collection Techniques, Synthesis 334,
National Cooperative Highway Research Program, Transportation Research Board.
Shafizadeh, K. and Mannering, F. (2003), Acceptability of Pavement Roughness on Urban
Highways by Driving Public, Transportation Research Record 1860, Transportation Research
Board.
Shahin, M. (1997), PAVER Distress Manual, TR 97/104, US Army Construction Engineering
Research Laboratories, Champaign, IL, June 1997.
Start, M., Jeong, K., and Berg, W. (1998), Potential Safety Cost-Effectiveness of Treating Rutted
Pavements, Transportation Research Record 1629, Transportation Research Board.
SHRP (1993), Distress Identification Manual for the Long-Term Pavement Performance
Project, Strategic Highway Research Program, National Research Council.
Texas DOT (2009), Hydroplaning, Hydraulic Design Manual, Texas DOT, March 1, 2009.
20
Section 4
Nondestructive Testing via the Falling Weight Deflectometer
4.1 Purpose
This section overviews the most commonly used Falling Weight Deflectometer (FWD) in use and
how it can be used to aid pavement assessment decisions.
4.2 Measurement Method

This subsection will briefly overview impact (or impulse) pavement loading. The device
described is the Dynatest FWD. This device can obtain measurements rapidly and the impact
load is easily varied.
All impact load NDT devices deliver a transient impulse load to the pavement surface. The
subsequent pavement response (deflection) is measured. Standard test methods include:
(i) ASTM D4694-96: Standard Test Method for Deflections with a Falling-Weight-Type Impulse
Load Device
(ii) ASTM D4695-03: Standard Guide for General Pavement Deflection Measurements
The significant features of ASTM D4694 include: (1) the force pulse will approximate a
haversine or half-sine wave, (2) the peak force of 11,000 lb must be achievable by the loading
device, (3) the force-pulse duration should be within range of 20 to 60 ms with a rise time in
range of 10 to 30 ms, (4) the loading plates standard sizes are 300 mm (12 in.) and 450 mm (18
in.), (5) the deflection transducers, which are used to measure the maximum vertical
movement of the pavement, can be seismometers, velocity transducers, or accelerometers, (6)
the load measurements must be accurate to at least 2 percent or 160 N ( 36 lb), whichever
is greater, (7) the deflection measurements must be accurate to at least 2 percent or 2 m
( 0.08 mils), whichever is greater. Note that 0.08 mils = 0.00008 inch and 2 m = 0.002 mm,
and (8) a precision guide in ASTM D4694 notes when a device is operated by a single operator
in repetitive tests at the same location, the test results are questionable if the difference in the
measured center deflection (D0) between two consecutive tests at the same drop height (or
force level) is greater than 5 percent. For example, if D0 = 0.254 mm (10 mils) then the next
load must result in a D0 range less than 0.241 mm to 0.267 mm (9.5 to 10.5 mils).
21
(iii) Dynatest FWD

The Dynatest FWD is the most widely used FWD in the US. The device is trailer mounted and
uses deflection sensors that are velocity transducers. By use of different drop weights and
heights this device can vary the impulse load to the pavement structure from about 1,500 to
27,000 lb. The weights are dropped onto a rubber buffer system resulting in a load pulse of
0.025 to 0.030 seconds. The standard load plate has a 300 mm (11.8 in.) diameter.
Locations for the seven velocity transducers vary. From ASTM D4694 the number and spacing
of the sensors is optional and will depend upon the purpose of the test and the pavement layer
characteristics. A sensor spacing of 12 in. is frequently used. A number of State DOTs have used:
Distance from the
center of the load
plate (in.)
0
8
12
24
36
48
The Strategic Highway Research Program (SHRP) sensor spacing with the 11.8 in. load plate are:
Distance from the
center of the load
plate (in.)
0
8
12
18
24
36
60
22
4.3 Analysis Tools

This subsection will focus on straightforward analysis tools that can be applied to FWD
deflection results.
4.3.1 Description of Available Analysis Tools for Flexible Pavements

This subsection will be used to describe three data assessment tools: (1) maximum deflection,
(2) the Area Parameter, and (3) a simplified method for calculating subgrade modulus.
The use of selected indices and algorithms provide a "picture" of the relative conditions found
throughout a project. This picture is useful in performing backcalculation and may at times be
used by themselves on projects with large variations in surfacing layers. Deflections measured
at the center of the test load combined with Area values and E SG computed from deflections
measured at 24 in. from the center of the load plate are shown in the linear plot to provide a
visual picture of the conditions found along the length of any project (as illustrated by data
from a rural road in Figure 4.1).
Figure 4.1 Illustrations of FWD Deflection Data Summarized by the Three Types of Data
23
The deflection data in Figure 4.1 is normalized data in that the measured deflections are
calculated for a 9,000 lb load. The modulus determination was based on the deflection 24 in.
from the center of the load plate.
Table 4.1 provides general information about conclusions that can be drawn from the FWD
parameters of Area and D0.
Table 4.1 General Information about the Area and D0
Generalized Conclusions*
FWD Based Parameter
Area
Maximum Surface
Deflection (D0)
Low
Low
Weak structure, strong subgrade
Low
High
Weak structure, weak subgrade
High
Low
Strong structure, strong subgrade
High
High
Strong structure, weak subgrade
(i) Maximum pavement deflection (D0)

The maximum pavement deflection can vary widely for different pavement structures and
throughout the day as its temperature changes. D0 ranges can be grouped into the following
broad and approximate categories (Table 4.2):
Table 4.2 D0 Ranges
Maximum Surface
Deflection (D0) Level
Generalized
Conclusions
Approximate D0
(in.)
Low Deflections
Strong structure
0.020
Medium Deflections
Medium structure
0.030
High Deflections
Weak structure
0.050
(ii) Area Parameter

The Area Parameter represents the normalized area of a slice taken through any deflection
basin between the center of the test load and 3 ft. By normalized, it is meant that the area of
the slice is divided by the deflection measured at the center of the test load, D 0. Thus the Area
Parameter is the length of one side of a rectangle where the other side of the rectangle is D 0;
hence, the Area Parameter has units of inches.
24
The Area equation is:

A = 6(D0 + 2D1 + 2D2 + D3)/D0
where
D0 = surface deflection at center of test load,
D1 = surface deflection at 1 ft,
D2 = surface deflection at 2 ft, and
D3 = surface deflection at 3 ft.
The maximum value for Area is 36.0 and occurs when all four deflection measurements are
equal (not likely to actually occur) as follows:
If, D0 = D1 = D2 = D3 then, Area = 6(1 + 2 + 2 + 1) = 36.0 in.
For all four deflection measurements to be equal (or nearly equal) would indicate an extremely
stiff pavement system (like portland cement concrete slabs or thick, full-depth asphalt
concrete.)
The minimum Area value should be no less than 11.1 in. This value can be calculated for a onelayer system which is analogous to testing (or deflecting) the top of the subgrade (i.e., no
pavement structure). Using appropriate equations, the ratios of
D1 D2 D3
D0 , D0 , D0
always result in 0.26, 0.125, and 0.083, respectively. Putting these ratios in the Area equation
results in Area = 6(1+ 2(0.26) + 2(0.125) + 0.083) = 11.1 in. Further, this value of Area suggests
that the elastic moduli of any pavement system would all be equal (e.g., E1 = E2 = E3 = ). This
is highly unlikely for actual, in-service pavement structures. Low area values suggest that the
pavement structure is not much different from the underlying subgrade material (this is not
always a bad thing if the subgrade is extremely stiff). Typical Area values are shown in the Table
4.3.
Table 4.3 Typical Area Values
Pavement Structure
Area Parameter (in.)
PCCP Range
24-33
Sound PCC
29-32
27+
Thick HMA ( 9 in. of HMA)
23
Medium HMA ( 5 in. of HMA)
17
Thin HMA ( 2 in. of HMA)
Chip sealed flexible pavement
15-17
Weak chip sealed flexible pavement
12-15
25
(iii) Subgrade Modulus

An NCHRP study [Darter, et al, 1991] which revised Part III of the AASHTO Pavement Guide
recommended that the following equation be used to solve for subgrade modulus:
MR = P(1 - 2)/( )(Dr)(r)
where
(Eq. 4.1)
MR = backcalculated subgrade resilient modulus (psi),

P = applied load (lbs) from the FWD,
Dr = pavement surface deflection a distance r from the center of the
load plate (inches), and
r = distance from center of load plate to Dr (inches).

Using a Poisson's ratio of 0.40, Equation 4.1 reduces to
MR = 0.01114 (P/D2)
(Eq. 4.2)
MR = 0.00743 (P/D3)
(Eq. 4.3)
MR = 0.00557 (P/D4)
for sensor spacing of 2 ft (610 mm), 3 ft (914 mm), and 4 ft (1219 mm).
(Eq. 4.4)
If a Poisson's ratio of 0.45 is used instead for the same sensor spacing, the equations become:
MR = 0.01058(P/D2)
(Eq. 4.5)
MR = 0.00705 (P/D3)
(Eq. 4.6)
MR = 0.00529 (P/D4)
(Eq. 4.7)
Darter et al (1991) recommended that the deflection used for subgrade modulus determination
should be taken at a distance at least 0.7 times r/ae where r is the radial distance to the
deflection sensor and ae is the radial dimension of the applied stress bulb at the subgrade
"surface." The ae dimension can be determined from the following:
ae =
where
a2 + D
EP
MR
ae = radius of stress bulb at the subgrade-pavement interface,

a = NDT load plate radius (inches),
D = total thickness of pavement layers (inches)
EP = effective pavement modulus (psi), and
26
MR = backcalculated subgrade resilient modulus.

For "thin" pavements, ae ~ 15 in. and "medium" to "thick" pavements, ae ~ 26 to 33 in.
Thus, the minimum r is usually 24 to 36 in. (recall r > 0.7 (ae)).
Typical subgrade moduli are shown in the Table 4.4 below (after Chou et al, 1989):
Material
Dry, psi
Clay
Silt
Silty or
Clayey Sand
Sand
Silty or
Clayey
Gravel
Gravel
15,000
15,000
20,000
Table 4.4 Typical Subgrade Moduli

Subgrade Moduli and Climate Condition
Wet No
Wet - Freeze
Freeze, psi
Unfrozen, psi
Frozen, psi
6,000
6,000
50,000
10,000
5,000
50,000
10,000
5,000
50,000
25,000
40,000
25,000
30,000
25,000
20,000
50,000
50,000
50,000
50,000
40,000
50,000
4.3.2 Examples of Analyses of FWD deflection basins for Flexible Pavement

The following deflection basins shown in Table 4.5 were obtained with a Dynatest FWD. The
pavement temperature at the time of testing was 46F (8 C). The deflection basins for the four
FWD drops and normalized to 9,000 lb are shown below:
FWD Load (lb)

16,987
12,070
9,406
6,186
Normalized to
9000 lb.
D0
27.07
21.28
17.53
12.33
16.59
Table 4.5 Example FWD Deflection Data

Deflection (mils)
D8
D12
D24
21.55
18.60
11.27
16.98
14.62
8.67
13.95
11.98
7.01
9.77
8.31
4.65
13.24
11.34
6.58
The pavement structure at the time of FWD testing was:

HMA: 6.0 in. and the HMA layer exhibited some fatigue cracking.
Granular Base (sandy gravel): 18.0 in.
27
D36
7.33
5.56
4.45
2.88
4.18
D48
5.28
3.98
3.23
2.05
2.99
Subgrade: Silt (ML) with a wide seasonal variation in water table depth. The soil is frost
susceptible and this area can have substantial ground freezing. At the time of testing the
spring thaw had occurred about one month earlier.
(i) Requirements
Analyze the available data to characterize the overall structure and estimate the layer
properties (moduli) using only the information provided above.
(ii) Results
Maximum surface deflection
The maximum surface deflection = 0.01657 in. for a pavement with 6 in. of HMA. This value
suggests a low pavement deflection.
Subgrade Modulus (closed form equations) from the AASHTO Guide (1993)
MR = P(1- 2)/( )(Dr)(r)
= 9000(1-0.452)/( )(0.00418) (36)
15,200 psi
Check r 0.7(ae), OK.
The pavement subgrade modulus for a ML silt is better than average.
Area Parameter
Area = 6(D0 + 2D12 + 2D24 +D36)/D0
= 6(0.01659 + (2)(0.01134) + (2)(0.00658) + 0.00418)/0.01659
20.5 in.
This Area Parameter is low for this thickness of AC. Thus, the Area value suggests a
weak pavement structure but not extremely so.
(iii) Detailed Project Data Example
Table 4.6 below summarizes deflection data that was collected on a portion of an actual
project. The project was about 5 miles in length and FWD testing was performed every 250
feet but only four of FWD locations are shown (these locations were also coring sites). The
average pavement temperature at the time the FWD data was collected was 46 F to 50 F. The
timing of the survey was about 1.5 to 2 months after the spring thaw in this area.
28
As shown in the Table 4.6, the normalized D0 deflections range from about 9 to 36 mils.
Deflections less than about 30 mils are considered normal. The HMA thicknesses varied
between 4.6 to 5.3 inches with an average of 5.2 inches which constitutes a medium
thickness of HMA (refer back to Table 4.2).
The Area values shown in the table suggest weak HMA, but not necessarily extreme weakness
due to stripping. Table 4.7 illustrates typical theoretical Area values for various uncracked HMA
thicknesses which aids this type of comparison.
Table 4.6 FWD Deflections, Area Value, and Subgrade Modulus
Core
Location (MP)
207.85
Load
(lbf)
16,940
12,086
9,421
6,218
Normalized Values
208.00
16,987
12,070
9.405
6,186
Normalized Values
208.50
16,829
12,245
9,533
6,297
Normalized Values
209.00
16,305
11,737
9,247
6,154
Normalized Values
Deflections (mils)
D0
31.30
24.21
19.45
13.19
18.39
27.04
21.26
17.52
12.32
16.57
14.92
11.65
9.61
6.73
9.01
59.25
46.14
36.93
25.00
35.51
D8
26.18
20.31
16.38
11.26
15.51
21.53
16.97
13.94
9.76
13.23
11.89
9.29
7.63
5.35
7.17
48.58
37.52
29.80
19.88
28.66
D12
23.19
18.11
14.57
9.92
13.78
18.58
14.61
11.97
8.31
11.34
10.23
7.95
6.53
4.49
6.10
42.52
32.56
25.63
16.77
24.61
29
D24
13.78
10.35
8.11
5.12
7.60
11.26
8.66
7.01
4.65
6.57
5.91
4.49
3.62
2.40
3.39
21.30
15.59
11.77
7.28
11.42
D36
9.09
6.81
5.28
3.39
5.00
7.32
5.55
4.45
2.87
4.17
3.19
2.13
1.81
1.26
1.69
9.53
6.69
4.96
3.03
4.84
D48
6.65
4.96
3.98
2.83
3.82
5.28
3.98
3.23
2.05
2.99
2.28
1.73
1.30
0.87
1.26
5.12
3.58
2.68
1.73
2.64
Area
Value
(in)
MR
(psi)
21
14,358
20
16,534
19
32,198
19
9,572
Table 4.7 Typical Theoretical Area Values for Uncracked HMA

Approximate Area Parameter (in.)
HMA Thickness (in.)
Normal Stiffness Low Stiffness
2
17
16
3
19
18
4
21
19
5
23
21
6
24
22
7
26
22
8
26
23
9
27
24
10
28
24
A quick, slightly more formal check of the pavement structure is to compare the actual Area
value to see if it falls within the range (normal to low stiffness), above this range (above normal
stiffness) or below this range (below normal stiffness). This comparison is shown in Table 4.8.
Table 4.8 Comparison of Area Value and Acceptable Area Value Range
Core
Location
207.85
208.00
208.50
209.00
HMA
Thickness
(in.)
5.3
6.0
4.7
4.6
Actual Area
(in.)
Above, Below or
Within Range
21
20
19
19
Within
Below
Below
Below
4.3.3 Description of Available Analysis Tools for Rigid Pavements

Rehabilitation of portland cement concrete pavements is not straightforward. To provide a
more consistent analysis process, the load transfer efficiency should be checked with FWD
obtained deflection data if the pavement type is jointed plain concrete pavement (JPCP).
(i) Load Transfer Efficiency
When a wheel load is applied at a joint or crack, both the loaded slab and adjacent unloaded
slab deflect. The amount the unloaded slab deflects is directly related to joint performance. If
a joint is performing perfectly, both the loaded and unloaded slabs deflect equally.
30
Joint performance can be evaluated by calculating load transfer efficiency (LTE) across a joint or
crack using measured deflection data. The concept of joint load transfer efficiency is illustrated
in Figure 4.2. Load transfer efficiency can be calculated using the following equation:
LTE = ( U/ L)(100)
LTE = load transfer efficiency, percent
U = the deflection of the unloaded slab, mils
L = the loaded slab deflection, mils.
Joint efficiency depends on several factors, including temperature (which affects joint opening),
joint spacing, number and magnitude of load applications, foundation support, aggregate
particle angularity, and the presence of mechanical load transfer devices.
As mentioned, temperature plays a major role in determining joint effectiveness. In general,
the lower the temperature, the lower the load transfer efficiency. Load transfer efficiency is
reduced because joints open during cooler weather, reducing contact between faces. Joint
load transfer efficiency has also been shown, in both laboratory and field studies, to decrease
with increasing load applications. However, this impact is lessened for harder aggregates. The
aggregate characteristics play a more significant role after many load applications.
To test the approach side of a joint or crack, the FWD loading plate is placed in front of the
joint, with the other velocity transducers located across the joint. The leave side of the joint is
tested by placing the loading plate at the joint edge on the leave slab with an extra velocity
transducer mounted behind the loading plate across the joint. The concept of slab approach
and leave sides and of transverse joint testing are illustrated in Figure 4.3.
L
= Deflection loaded slab

U
= Deflection unloaded
slab
U
Figure 4.2 Illustration of Joint Load Transfer Efficiency
31
2
1
3
1
3
Figure 4.3 Locations of FWD Load Plate and Deflection Sensors for Determining Load
Transfer Efficiency
The percentage load transfer can vary between almost 100% (excellent) to near 0% (extremely
low). AASHTO (1993) notes that load transfer restoration should be considered for transverse
joints and cracks with load transfer efficiencies ranging between 0 to 50%. It has been observed
for numerous in-service jointed PCC pavements that load transfer efficiencies of 70% or greater
generally provides good joint or crack performance.
4.3.4 Backcalculation
Backcalculation is the process by which pavement layer moduli are estimated by matching
measured and calculated surface deflection basins. This is done via a computer program and
there are a number of these available in the US. It is likely that within a specific state there is a
preferred backcalculation software package to use.
32
The general guidelines which follow are broad in scope and should be considered
rules-of-thumb.
(i) Number of Layers
Generally, use no more than 3 or 4 layers of unknown moduli in the backcalculation process
(preferably, no more than 3 layers). If a three layer system is being evaluated, and questionable
results are being produced (weak or low stiffness base moduli, for example), it is sometimes
advantageous to evaluate this pavement structure as a two layer system. This modification
would possibly indicate that the base material has been contaminated by the underlying
subgrade and is weaker due to the presence of fine material. Alternatively, a stiff layer should
be considered if not done so previously (see below). If a pavement structure consists of a stiffer
layer between two weak layers, it may be difficult to obtain realistic backcalculated moduli. For
example, a pavement structure which consists of deteriorated asphalt concrete over a cement
treated base.
(ii) Thickness of Layers
Surfacing. It can be difficult to accurately backcalculate HMA or BST moduli for bituminous
surface layers less than 3 in. thick. Such backcalculation can be attempted for layers less than 3
in., but caution is suggested.
In theory, it is possible to backcalculate separate layer moduli for various types of bituminous
layers within a flexible pavement. Generally, it is not advisable to do this since one can quickly
be attempting to backcalculate too many unknown layer moduli (i.e., greater than 3 or 4). By
necessity, one should expect to combine all bituminous layers (seal coats, asphalt concrete,
etc.) into one layer unless there is evidence (or the potential) for distress, such as stripping, in
a HMA layer or some other such distress which is critical to pavement performance.
Unstabilized Base/Subbase Course. Thin base course beneath thick surfacing layers (say
HMA or PCC) often result in low base moduli. There are a number of reasons why this can
occur. One, a thin base is not a significant layer under a stiff, thick layer. Second, the base
modulus may be relatively low due to the stress sensitivity of granular materials. The use of a
stiff layer generally improves the modulus estimate for base/subbase layers.
(iii) Subgrade
If unusually high subgrade moduli are calculated, check to see if a stiff layer is present. Stiff
layers, if unaccounted for in the backcalculation process, will generally result in unrealistically
high subgrade moduli. This is particularly true if a stiff layer is within a depth of about 20 to 30
feet below the pavement surface.
33
(iv) Stiff Layer

Often, stiff layers are given fixed stiffness ranging from 100,000 to 1,000,000 psi with semiinfinite depth. This, in effect, makes the subgrade a layer with a fixed depth instead of the
normally assumed semi-infinite depth. What is not so clear is whether one should always fix the
depth to stiff layer at say 20, 30, or 50 feet if no stiff layer is otherwise indicated (i.e., use a
semi-infinite depth for the subgrade). The depth to stiff layer should be verified whenever
possible with other NDT data or borings.
The stiffness (modulus) of the stiff layer can vary. If the stiff layer is due to saturated conditions
(e.g. water table) then moduli of about 50,000 psi appear more appropriate. If rock or stiff
glacial tills are the source of the stiff layer then moduli of about 1,000,000 psi appear to be
more appropriate.
(v) Layer Moduli
A few comments about layer moduli are appropriate.
Cracked HMA Moduli. Generally, fatigue cracked HMA (about 10 percent wheelpath cracking)
is often observed to have backcalculated moduli of about 100,000 to 250,000 psi. What is most
important in the backcalculation process, assuming surface fatigue cracking is present, is to
determine whether the cracks are confined to only the immediate wearing course or penetrate
through the whole depth of the HMA layer. For HMA layers greater than 6 in. thick, cracking
only in the wearing course is often observed and the overall HMA layer will have a substantially
higher stiffness than noted above (at moderate layer temperatures of say 75 to 80F).
Base and Subbase Moduli. Typical base and subbase moduli are shown in Table 4.9.
Table 4.9 Typical Unstabilized and Stabilized Base and Subbase Moduli
Material
Typical Modulus (psi)
Modulus Range (psi)
Crushed Stone or Gravel Base
Crushed Stone or Grave Subbase
Sand Base
Sand Subbase
Material
Lime Stabilized
Cement Stabilized
Unstabilized
35,000
10,000 to 150,000
30,000
10,000 to 100,000
20,000
5,000 to 80,000
15,000
5.000 to 80,000
Stabilized
Compressive
Typical Modulus
Modulus Range (psi)
Strength (psi)
(psi)
< 250
30,000
5,000 to 100,000
250 to 500
50,000
15,000 to 150,000
70,000
20,000 to 200,000
500
< 750
400,000
100,000 to 1,500,000
750 to 1250
1,000,000
200,000 to 3,000,000
1,500,000
300,000 to 4,000,000
1250
34
Subgrade Moduli. Typical subgrade moduli were previously shown in Table 4.5.
(vi) Backcalculation Summary
Performing backcalculation of pavement layer moduli is part science and part art; thus,
experience typically will improve the estimated results. It is advisable to initially work with
someone who has solid experience doing backcalculation or take a short course on the topic
assuming one is available. It will take only a few projects along with experience from others to
become well informed on this powerful assessment technique.
4.4 References
AASHTO (1993), AASHTO Guide for Design of Pavement Structures, 1993, American
Association of State Highway and Transportation Officials, Washington, DC.
Chou, Y. J., Uzan, J., and Lytton, R. L., "Backcalculation of Layer Moduli from Nondestructive
Pavement Deflection Data Using the Expert System Approach," Nondestructive Testing of
Pavements and Backcalculation of Moduli , ASTM STP 1026, American Society for Testing and
Materials, Philadelphia, 1989, pp. 341 - 354.
Darter, M.I., Elliott, R.P., and Hall, K.T., (1991) "Revision of AASHTO Pavement Overlay Design
Procedure," Project 20-7/39, National Cooperative Highway Research Program, Transportation
Research Board, Washington, D.C., September 1991.
35
Section 5
Ground Penetrating Radar
5.1 Purpose
This section describes Ground Penetrating Radar (GPR) technology and presents an overview of
the most common applications of both air coupled and ground coupled GPR systems for aiding
in pavement assessment decisions.

This section briefly describes the two types of GPR and the basic principles of operation.
The standard references for GPR applications in highways are:
AASHTO PP 40-00 Standard Recommended Practice for Application of Ground Penetrating
Radar to Highways
ASTM D6087-08 Standard Test Method for Evaluating Asphalt Covered Concrete Bridge Decks
using Ground Penetrating Radar
ASTM D6432- 99 (2005) Standard Guide for Using Surface Ground Penetrating Radar Method
for Subsurface Investigation
(i) Air Coupled GPR systems
A typical commercially available 2.2 GHz air-coupled Ground Penetrating Radar unit is shown in
Figure 5.1. The radar antenna is attached to a fiber glass boom and suspended about 5 feet
from the vehicle and 14 inches above the pavement. This particular GPR unit can operate at
highway speeds (70 mph); it transmits and receives 50 pulses per second, and can effectively
penetrate to a depth of around 20 to 24 inches. All GPR systems include a distance measuring
system and many of the new systems also have synchronized/integrated video logging, so the
operator can view both surface and subsurface conditions. GPS is also included in many new
systems for identifying problem locations
The advantages of these systems are the speed data collection which does not require any
special traffic control. The GPR generate clean signals which without filtering are ideal for
quantitative analysis using automated data processing techniques to compute layer dielectrics
and thickness. These systems are also excellent for locating near surface defects in flexible
pavements.
The disadvantages are a) the limit depth of penetration, b) not ideal for penetrating thick
concrete pavements and c) the most popular operating frequency (1GHz) is now subject to FCC
restrictions in the US.
36
Figure 5.1 Air Coupled GPR systems for highways
(ii) Ground Coupled GPR systems

As shown in Figure 5.2, a whole range of different operating frequencies are available for
ground coupled GPR systems. The selection of the best frequency for a particular application
depends on the required depth of penetration. As the name implies these antennas have to
stay in close contact to the pavement under test.
The advantage of these systems is their depth of penetration, several of the lower frequency
systems can penetrate 20 feet under ideal conditions. The higher frequency systems are
superior for many concrete pavement applications such as locating both reinforcing steel and
sub-slab defects such as voids or trapped moisture. The disadvantage of these is the speed of
data collection, when towed behind a vehicle the maximum speed is around 5 mph. The signals
are also noisy and filtering is required. Substantial training is required to clean up and interpret
ground coupled GPR data.
37
Figure 5.2 Ground coupled systems, 1.5 GHz on left, lower frequency antennas
with control unit on right
5.3 Analysis Tools

All GPR systems send discrete pulses of radar energy into the pavement and capture the
reflections from each layer interface within the structure. Radar is an electro-magnetic (e-m)
wave and therefore obeys the laws governing reflection and transmission of e-m waves in
layered media. At each interface within a pavement structure a part of the incident energy will
be reflected and a part will be transmitted. It is normal to collect between 30 and 50 GPR
return signals per second, which for high speed air coupled surveys could mean on trace for
every 2 to 3 feet of travel. The captured return signal is often color coded and stacked side by
side to provide a profile of subsurface conditions, this is analogous to an X-Ray of the
pavement structure. Examples of this will be given later. However with air coupled signals as
described below these signals can also be used to automatically calculate the engineering
properties of the pavement layers.
5.3.1 Air Coupled GPR system
A typical plot of captured reflected energy versus time for one pulse of an air coupled GPR
system is shown in Figure 5.3, as a graph of volts versus arrival time in nanoseconds. To
understand GPR signals it is important to understand the significance of this plot.
38
Figure 5.3 Captured GPR reflections from a typical flexible pavement
The reflection A0 is known as the end reflection it is internally generated system noise which
will be present in all captured GPR waves. The more important peaks are those that occur after
A0. The reflection A1 (in volts) is the energy reflected from the surface of the pavement and A 2
and A3 are reflections from the top of the base and subgrade respectively. These are all
classified as positive reflections, which indicate an interface with a transition from a low to a
high dielectric material (typically low to higher moisture content). These amplitudes of
reflection and the time delays between reflections are used to calculate both layer dielectrics
and thickness. The dielectric constant of a material is an electrical property which is most
influenced by moisture content and density, it also governs the speed at which the GPR wave
travels in the layer. An increase in moisture will cause an increase in layer dielectric; in contrast
an increase in air void content will cause a decrease in layer dielectric.
The equations to calculate surface layer thickness and dielectrics are summarized below:
1
1
A1 / Am
A1 / Am
(Eq 1)
39
where
= the dielectric of the surface layer,
A1 = the amplitude of surface reflection, in volts

Am = the amplitude of reflection from a large metal plate in volts (this represents the
100% reflection case, see Figure 5.1 for the metal plate test)
h1
cx t1
(Eq 2)
where h1 = the thickness of the top layer

c = a constant speed of e-m wave in air (5.9 ins/ns two way travel)
t1 = the time delay between peaks A1 and A2, (in ns)
Similar equations are available for calculating the base layer dielectric and thickness. This
calculation process is performed automatically in most operating systems with the end user
simply getting a table of layer properties.
In most GPR projects several thousand GPR traces like Figure 5.3 are collected. In order to
conveniently display and interpret this information color-coding schemes are used to convert
the traces into line scans and then stack them side-by-side so that a subsurface image of the
pavement structure can be obtained. This approach is shown below in Figure 5.4.
40
Figure 5.4 Color Coding and stacking individual GPR images

The raw GPR image collection is displayed vertically in the middle of Figure 5.4. This image is
for one specific location in the pavement. The GPR antenna shoots straight down and the
resulting thickness and dielectric estimates are point specific. The single trace generated is
color coded into a line scan using the color scheme in the middle of Figure 5.4. In the current
scheme the high positive reflections are colored red and the negatives are colored blue. The
green color is used where the reflections are near zero and are of little significance. These
individual line scans are stacked so that a display for a length of pavement is developed. Being
able to read and interpret these images is critical to effectively using GPR for pavement
investigations, to locate section breaks in the pavement structure and to pinpoint the location
of subsurface defects.
An example of a typical GPR display for approximately 3000 ft by 24 inches deep of a thick
flexible pavement is shown in Figure 5.5. This is taken from a section of newly constructed
thick asphalt pavement over a thin granular base. In all such displays the x axis is distance (in
miles and feet) along the section and the y axis is a depth scale in inches.
41
Figure 5.5 Color-Coded GPR Traces

The labels on this figure are as follows A) GPR files being used in analysis, B) Main Pull down
menu bar, C) button to define the color coding scheme, D) Distance scale (miles and feet), E)
end location of data within the GPR file (1mile and 3479 feet), G) depth scale in inches, with
the zero (0) being the surface of the pavement, F) Default dielectric value used to convert
the measure time scale into a depth scale. The important feature of this figure are the lines
marked H, I and J these are the reflection from the surface, top and bottom of base
respectively. This pavement is homogeneous and the layer interfaces are easy to detect.
When processing GPR data the first step is to develop displays such as Figure 5.5. From this it is
possible to identify any clear breaks in pavement structure and to identify any significant
subsurface defects. The intensity of the subsurface colors is related to the amplitude of
reflection, therefore areas of wet base would be observed as bright red reflections (I).
For many applications a black/white coding scheme is selected. This is widely used when
reviewing data collected with ground coupled GPR systems. An example of the grey scale for
the pavement shown in Figure 5.5 is shown below in Figure 5.6.
42
Figure 5.6 Similar data to Figure 5.5 presented as a grey scale
All of the commercially available software packages, produce both a color display of subsurface
condition such as Figures 5.5 and 5.6 together with a table of computed layer thicknesses and
dielectrics which is usually exported to Excel. A typical table is shown below in Figure 5.7,
where E1 and Thick 1 are the top layer dielectric and thickness.
Trace
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
Feet
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
Time1
1.6
1.5
1.5
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.5
1.5
1.5
Time2
3.2
3.3
3.4
3.4
3.5
3.5
3.6
3.6
3.6
3.6
3.6
3.6
3.6
3.6
Time3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Thick1
3.8
3.7
3.6
3.4
3.5
3.4
3.4
3.3
3.4
3.5
3.5
3.5
3.6
3.6
Thick2
6.1
6.1
6.4
6.4
6.5
6.6
6.7
6.7
6.4
6.6
6.5
6.4
6.5
6.4
Thick3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
E1
6.2
6.2
6.2
6.3
6.2
6.2
6.2
6.2
6.3
6.2
6.3
6.1
6.1
6.0
E2
10.0
10.3
9.9
10.1
10.2
10.3
10.4
10.6
11.3
10.6
11.3
11.6
11.3
11.4
Figure 5.7 Tabulated thicknesses and dielectric values from GPR data
43
E3
11.1
11.5
10.8
10.9
11.3
11.4
11.9
11.8
12.5
12.0
12.4
12.8
12.4
12.6
5.3.2 Examples of Analysis of GPR data for Flexible Pavements

When planning to incorporate the existing pavement as part of a new pavement structure it is
critical to have good information on the existing subsurface layer thicknesses and layer types. A
few DOTs maintain good pavement layer data bases, but this is not always the case, most
DOTs often have poor information on existing layer thicknesses. Often maintenance activities
significantly alter the as-constructed pavement structure in localized areas and these activities
are often not captured in existing data bases.
One popular method of rehabilitating old flexible pavements is by the use of full depth
reclamation (FDR) and chemical treatment to incorporate and stabilize the existing pavement
to form a solid foundation layer for the new pavement structure. However several major
problems have occurred during construction, or poor pavement performance has occurred
because of the failure to account for the variability of the existing pavement in the design
phase. Lab designs are based on testing at localized sampling locations, sections which are
either too thick or too thin have been documented to cause problems. GPR can help in this
area.
It also must be recalled that processing FWD as described in Chapter 4 requires information
about the thickness of the asphalt surface layer. GPR can provide substantial help in analyzing
and explaining FWD deflection data.
Three case studies are presented below to demonstrate how GPR can assist in upfront flexible
pavement evaluations.
(i) Thickness profiling for an FDR application
In many FDR applications the purpose is to treat the existing pavement to create a stable
uniform pavement foundation layer for the new pavement structure. In most FDR design
samples are taken from the existing pavement and taken back to laboratory to determine the
optimal level of either cement or asphalt stabilization to reach a specified target strength. It is
therefore important to know that the sampling location selected is representative of the overall
project. It is also important to assess if the selected design will be appropriate when variations
in layer thicknesses occur.
Figure 5.8 shows variations in asphalt layer thickness for a FDR candidate. At the sample
location the structure was 5 inches of asphalt and 10 inches of granular base. Based on lab test
results the plan was to recycle to a depth of 10 inches blending 50% base with 50% existing
base with 3% cement. However from a review of Figure 5.8 the 5 inches of HMA is common on
this highway with several noticeable exceptions. The first 800 feet only has 3 inches of asphalt
this is not thought to be a concern. However for about 2000 feet of this project the total HMA
thickness is over 12 inches. From previous experience the 3% cement treatment does not work
with 100% RAP. In these locations it will be necessary to modify the construction plan. In these
locations 5 inches of the existing HMA was milled and replace with 5 inches of new flexible
44
base. In that way the FDR process can continue and in all locations the as designed 50/50 blend
can be treated with cement.
Figure 5.8 Surface thickness variations from GPR profiling on FM 550
(ii) Defect detection prior to pavement rehabilitation

In many cases the long life of the existing flexible pavement can be achieved by simply adding a
structural overlay to the existing structure. This process works well provided there are no
major defects in the existing HMA layer or flexible base layer. GPR has shown that it can detect
stripping problems in HMA layers and areas where the exiting base layer is holding moisture. It
must be recalled that GPR traces are collected frequently at 2 3 feet intervals so very precise
location of deflects is possible. The GPR color coded profile shown in Figure 5.5 is from a thick
HMA section with no defects. This should be contrasted with GPR profile shown below in
Figure 5.9. This again is a thick HMA section, but in this case there are strong reflections from
within the HMA and very strong reflections from the bottom of the layer. The red/blue
reflections from within the HMA are associated with deteriorated areas where moisture is
trapped. When these deteriorated areas are close to the surface they can severely impact long
term performance.
The presence of defects in either HMA or base layers can be easily detected by GPR, their
severity will then need to be confirmed by localized coring. This is valuable input to the
pavement designer who has to make a decision whether they impact the future anticipated
performance of the proposed section. If they are very localized, then full depth milling can be
used in these areas.
45
Figure 5.9 Using GPR to identify defects in surface and base layers
(iii) Section uniformity

With many older pavements, particularly those involving some form of pavement widening the
existing pavement structure can be very variable. It is important to identify the different
structures in order to explain the cause of current conditions and to design future repairs.
Such a case is shown below in Figure 5.10. This is a 1.8 mile section and entire section had all
received a thin overlay and so surface condition was very similar. However the first part of the
section was performing poorly. A GPR surface was undertaken and from the display it is clear
that this section has three distinct pavement structures. Structure A was a thin HMA pavement
over a flexible base, Structure B was thick HMA and Structure C was a road built on top of an
existing roadway. This type of subsurface mapping can clearly help designers with their
rehabilitation designs
46
Figure 5.10 Using GPR to map subsurface variability
5.3.3 Examples of Analysis of GPR data for PCC pavements

The most popular applications of GPR in evaluating concrete pavements when making
pavement rehabilitation decisions are a) measuring slab thickness, b) detecting the presence
and depth of reinforcing steel and c) identify problems beneath the slab such as voids or
trapped moisture. In several instances especially for steel detection the ground coupled
systems performed better than the air coupled. The high frequency ground coupled systems
such as the 1.5 GHz unit shown in Figure 5.2 can give more focus and better target resolution
than air coupled units. Several case studies are shown below.
(i) Rebar detection
The GSSI handbook on Radar Inspection of GPR has some very good examples on rebar
detection. The figure below shows the typical GPR signature obtained over reinforcing steel.
There is a hyperbola shape and the top of the hyperbola is the location of the steel. The surface
of the concrete is the direct couple signature and the depth between the surface and the top
of the hyperbola is the depth of concrete cover. GSSI also claim that the size of the rebar can
be determined by the shape of the hyperbola.
47
Figure 5.11 Ground Coupled GPR signals from steel in concrete (GSSI)
Moving the GPR antenna slowly across the surface of the concrete it is possible to map
different layers of steel and the bottom of the concrete slab as shown in Figure 5.12
Figure 5.12 Mapping multiple layers of steel in concrete (GSSI)
(ii) Void detection

Detecting thin air voids with air coupled GPR is often problematic, and furthermore even very
thin voids is very detrimental to slab performance. Controlled studies have found that air voids
of less than 0.75 inches thick cannot be readily detected with air coupled GPR. However if the
voids are larger or if they are moisture filled then they can readily be detected. An example of
a GPR color profile for an 8 inch PCC slab with water filled voids is shown below in Figure 5.13.
The strong reflections (red areas) are locations of trapped water.
48
Figure 5.13 Mapping sub-slab water filled voids with GPR
(iii) Deep Investigations of sub slab conditions with GPR

The lower frequency ground coupled GPR can be used to investigate deep beneath concrete
pavements to identify changes in support conditions and possibly to help explain the
occurrence of surface distress. Figure 5.14 shows the color profile from a 400 MHz ground
coupled system. The entire pavement system and changes in pavement support can be
observed. The transverse rebar can be seen towards the top of the figure. The steel is more
closely spaced in the left of the figure. The anomaly on the left is a culvert. The bottom of the
slab is indicated. There is a clear change in subgrade support at the top of the subgrade
showing the transition from a cut to a fill area.
49
Figure 5.14 Mapping concrete pavements structure with GPR
50
5.4 Implementing GPR technology for Pavement Evaluation

GPR is in excellent technology for inspecting pavements when pavement rehabilitation
decisions are being made. Many case studies have been presented over the past two decades
but widespread implementation of the technology has been painfully slow. There are several
factors causing this and these will be discussed in this section, but the main factors are;
1) The FCC ban on 1 GHz air coupled systems in 2002 (these units can be purchased in
any country worldwide except the US). For the past decade most air coupled GPR
systems have been performed with systems built before 2002. Only recently have
commercial systems become available such as GSSIs 2.2 GHz as shown in Figure 1.
2) A lack of understanding about what GPR can and cannot do, in many cases the
technology was oversold
3) Inadequate data processing software and a lack of end user training
Agencies undertaking GPR implementation should be aware of the following issues which must
be resolved before GPR can be implemented as a routine pavement inspection tool; these
include
1) Need for GPR hardware specifications
2) Need for data collection software specifications
3) Training/Specifications for data collection activities
4) Specifications/Software for processing and interpreting GPR signals
5) End User training
6) Specifications for output formats and data storage system
Several DOTs and consultants have implemented GPR technology in-house (for example FDOT
and TxDOT and others), but most agencies get GPR services from consultant companies.
Selecting the best vendor can also be a problem.
(i) Obtaining GPR services
The AASHTO publication has a short section with recommendations for agencies on hiring GPR
consultants. In initiating contracts the agency has to be convinced that;
a) the consultant has quality equipment, ask them to run their equipment against the
performance specs (which are available), and
b) the consultant has good data processing skills, references from existing customers will help
here. GPR interpretation should never be done without taking limited field verification cores
early in the project. If the project is for layer thickness determination or for defect detection it
should be simple to set up a verification system early in the project.
51
(ii) Barriers to GPR Implementation

In addition to the FCC requirements there are also several common misconceptions that must
be overcome before any agency will adopt GPR technology; these are
a) GPR is only for layer thickness determination: My state has good as-built records so we do not
need GPR
As stressed throughout this report GPR is much more that a thickness measuring tool. It
provides information on the quality of existing structures and helps to explain the causes of
pavement distresses. Distresses are often associated with moisture ingress into pavement
layers. GPR signals are highly sensitive to moisture in any layer.
b) GPR systems are too expensive
A complete air coupled system described in this section costs around $100,000 for the
complete turnkey system, including vehicle. Ground coupled systems cost approximately
$60,000. With the cost of pavement rehabilitation activities these cost are minimal with the
cost of rehabilitating sections of Interstate pavement. GPR are substantially less than other
nondestructive testing equipment such as FWDs
c) GPR is a black box which is impossible to understand
Not true. The basics of GPR are very simple. The key here is that agency personnel should
attend training schools to get to understand this technology. Even if the plan is to initiate GPR
work through consultants the agency personnel need to have a basic understanding of what
this technology can and cannot do.
d) Our first experience with GPR was disappointing
This is often true. In the early 1990s a host of companies sold GPR services. They sometimes
made extensive claims on GPRs potential and their ability to successfully interpret the signals.
Many claimed to be able to find thin voids beneath concrete pavements often to disappoint the
DOT when validation field cores were taken. In some cases the vendors did not have adequate
software or interpretation skills. The key here again is training for end user agency personnel.
The AASHTO publication also is a good source to identify applications that have a high
probability of success
e) When the agency initiates a GPR program a host of vendors make claims about their
capabilities and it is impossible for the agency to judge their merits.
This is often true. But it can be overcome by firstly training of end user agency personnel. Also
as with any new technology field verification of any predictions must be a critical part of any
program. GPR will not eliminate coring but it will greatly reduce the number of cores
5.5 References
GSSI Handbook for Radar Inspection of Concrete, Aug 2006, www.geophysical.com
52
Section 6
Pavement Cores
6.1 Purpose
This section overviews the use of pavement cores and how they can be used to aid pavement
assessment decisions. Much of pavement analysis and understanding stem from knowledge of
layer thicknesses, types of materials, and condition.

This subsection briefly overviews both the frequency of sampling and organization of data from
pavement cores. Pavement cores not only reveal much about the existing pavement structure
but also allow for use of the DCP. Knowing the HMA layer thickness to within inch is essential
in assuring a more accurate prediction of layer moduli if a backcalculation procedure is used.
The number of cores obtained will depend on project specific conditions; however, a
reasonable rule-of-thumb is to obtain a core at every 5th or 10th FWD test location. If the
pavement thicknesses are found to vary substantially (not probable but this can be the case),
then cores should be obtained at every FWD test location.
Typical core diameters are either 4 or 6 inches.
6.3 Analysis Tools

This subsection will focus on how to organize pavement core data to aid decision-making.
Core data should be organized similar to the example data shown in Table 6.1 below.
Additionally, the location of each core in the lane should be recorded (such as centerline, left
wheelpath, between wheelpath, right wheelpath, outside pavement edge).
Core
Location
(milepost)
207.85
208.00
208.50
209.00
Table 6.1 Organization of Pavement Core Data

Depth
Comments
HMA
Base (Cores should be taken frequently at cracks, if they exist,
to determine if the crack is full-depth or partial-depth)
(in)
(in)
5.3
18.0
Core taken at a crack, crack is full depth
6.0
18.0
Core taken at a crack, core not intact
4.7
12.0
Core taken at a crack, crack is full depth
4.6
12.0
Very fatigued, core broke into several pieces
53
Section 7
Dynamic Cone Penetrometer
7.1 Purpose
This section overviews the Dynamic Cone Penetrometer (DCP) and how it can be used to aid
pavement assessment decisions.

This subsection describes the dynamic cone penetrometer device. The standard test method is:
(iii) ASTM D6951-03: Standard Test Method for Use of the Dynamic Cone Penetrometer in
Shallow Pavement Applications
From ASTM D6951: This test method is used to assess in situ strength of undisturbed soil
and/or compacted materials. The penetration rate of the 8-kg DCP can be used to estimate insitu CBR (California Bearing Ratio), to identify strata thickness, shear strength of strata, and
other material characteristics. The 8-kg DCP is held vertically and therefore is typically used in
horizontal construction applications, such as pavements and floor slabs. This instrument is
typically used to assess material properties down to a depth of 1000-mm (39-in.) below the
surface. The penetration depth can be increased using drive rod extensions. However, if drive
rod extensions are used, care should be taken when using correlations to estimate other
parameters since these correlations are only appropriate for specific DCP configurations. The
mass and inertia of the device will change and skin friction along drive rod extensions will
occur.
The 8-kg DCP can be used to estimate the strength characteristics of fine- and coarse-grained
soils, granular construction materials and weak stabilized or modified materials. The 8-kg DCP
cannot be used in highly stabilized or cemented materials or for granular materials containing a
large percentage of aggregates greater than 50-mm (2-in.). The 8-kg DCP can be used to
estimate the strength of in situ materials underlying a bound or highly stabilized layer by first
drilling or coring an access hole.
A sketch of a standard DCP is shown in Figure 7.1.
54
Figure 7.1 Sketch of the Minnesota DOT DCP (from Minnesota DOT, 1993)
55
7.3 Analysis Tools

DCP test results are typically expressed in terms of DPI which is the vertical movement of
the DCP cone produced by one drop of the hammer. This is expressed as either
mm/hammer blow or inches/hammer blow (Minnesota DOT, 1993).
(i) Basic correlation
A common correlation with DCP data is to estimate the California Bearing Ratio (CBR) of
unstabilized materials in a pavement structure. A correlation developed by the U.S. Army
Corps of Engineers (Webster et al, 1992) is:
log CBR = 2.46 1.12 log(DPI) or CBR = 292/DPI1.12
where DPI = mm/blow
Table 7.1 shows typical CBR and DPI ranges for three soil types (Minnesota DOT, 1993):
Table 7.1 Soils Types, CBR Values, and DPI
Soil Type
CBR Range
DPI Range
(%)
(mm/blow)
Clay (CL)
15-127
1-14
Sand (S-W)
14-39
6-15
Gravel (G-W)
47-95
2.7-5
[The table was modified by the authors of this document so that the DPI and CRB
correlation matched.]
(ii) Typical results
Burnham (1997) described an extensive set of DCP measurements on the subgrade soils and
base materials used in the various test sections at the Mn/Road facility. These are
summarized in Table 7.2. Following this work, the following DPI limits were recommended
for use by MnDOT personnel when analyzing DCP results for rehab studies.
Silty/clay materials: DPI 25 mm/blow
Select granular materials: DPI 7 mm/blow
Class 3 Special gradation materials: DPI 5 mm/blow
56
Table 7.2 Minnesota DCP Results Following Placement of the Base Course
Material
DPI Avg (mm/blow)
DPI Avg (mm/blow)
DPI Avg (mm/blow)
(Std Dev)
(Std Dev)
(Std Dev)
0-12 inches depth
12-24 inches depth
24-36 inches depth
Clay/Silt
11
21
21
Location 1
(3)
(7)
(7)
Clay/Silt
14
18
16
Location 2
(6)
(5)
(5)
Clay/Silt
12
20
15
Location 3
(5)
(7)
(7)
Sand
5
5
6
(2)
(1)
(2)
Base
4
3
3
Course
(2)
(1)
( 1)
[DPI average values were rounded to the nearest whole number.]
(iii) Subgrade stability
The Illinois DOT (1982, 2005) has used the DCP to check the subgrade stability. The purpose
of this is straightforwardthey want to know if the subgrade is stable enough to avoid
excessive rutting and/or shoving during and following construction activities. The subgrade
IBV (Immediate Bearing Value) can be estimated from the DPI. The IBV is similar to the CBR
except than IBV testing is conducted on a 4-inch molded sample instead of the CBRs 6-inch
samplefurther, the penetration test for determining the IBV is conducted immediately
after compaction instead of waiting 96 hoursthus IBV and CBR are similar but not
identical (Illinois DOT, 2005). Figure 7.2 shows the relationship between unsoaked CBR
(actually IBV), DPI, and required thickness of remedial measures. Remedial measures can
include the addition of granular backfill or subgrade modification such as lime stabilization.
The Illinois DOT DCP results and those from the Minnesota DOT broadly agree in that
subgrade DPI values greater than 25 mm/blow are of concern.
57
Figure 7.2 DCP Based Thickness Design for Granular Backfill and Subgrade Modification for
the Illinois DOT (figure from Burnham, 1997 but checked against Illinois DOT, 2005)
(iv) Use of DCP data in renewable decisions
The Texas Transportation Institute developed guidelines for the Texas DOT as to conditions
suitable for rubblizing existing rigid pavements (Figure 7.3). The High Risk portion of the
figure implies the pavement is not a good candidate for rubblization since the supporting
base and subgrade is excessively weak. Figure 7.3 is similar to but modified from similar
guidelines developed for Illinois (Figure 7.4). Figure 7.4 is of interest since data obtained by
Sebesta and Scullion (2007) for US 83 in Texas are plotted by total pavement thickness
versus DCP derived CBR values.
The DCPCBR correlation used in Texas is the same as the one described in 6.3(i) which
was originally done by the US Army COE.
58
Figure 7.3 Rubblization Selection Chart Developed by TTI

(from Sebesta and Scullion, 2007)
59
Figure 7.4 Illinois Rubblization Selection Chart with Data from US 83 (Texas)
(from Sebesta and Scullion, 2007; original Illinois DOT criteria from Heckel, 2002)
60
7.4 References
Burnham, T. (1997), Application of Dynamic Cone Penetrometer to Minnesota Department
of Transportation Pavement Assessment Procedures, Report No. MN/RC 97/19, Minnesota
Department of Transportation, St. Paul, MN, May 1997.
Heckel, L (2002), Rubblizing with Bituminous Concrete Overlay10 Years Experience in
Illinois, Report IL-PRR-137, Illinois Department of Transportation, April 2002. [The report
contains an appendix entitled Guidelines for Rubblizing PCC Pavement and Designing a
Bituminous Overlay. This appendix contains a similar version of Figure 6.4.]
Illinois DOT (1982), Subgrade Stability Manual, Policy Mat-10, Illinois Department of
Transportation.
Illinois DOT (2005), Dynamic Cone Penetrometer, Pavement Technology Advisory PTA-T4,
Bureau of Materials and Physical Research, Illinois Department of Transportation, February
2005.
Minnesota DOT (1993), User Guide to the Dynamic Cone Penetrometer, Office of
Minnesota Road Research, Minnesota DOT.
Sebesta, S. and Scullion, T. (2007), Field Evaluations and Guidelines for Rubblization in
Texas, Report No. FHWA/TX-08/0-4687-2, Texas Transportation Institute, December 2007.
Webster, S., Grau, R., and Thomas, W. (1992), Description and Application of Dual Mass
Dynamic Cone Penetrometer, Instruction Report GL-92-3, Waterways Experiment Station,
US Army Corps of Engineers, May 1992.
61
Section 8
Subgrade Soil Sampling and Tests
8.1 Purpose
This section is used to overview selected elements associated with subgrade soils and what
information is needed for pavement assessment decisions. Much of pavement analysis and
understanding stem from knowledge of layer thicknesses, types of materials, and condition.

This subsection will show both the types of tests and frequency of sampling associated with
subgrade soils. A summary of these tests is contained in Table 8.1.
Subgrade Test
Soil
Classification
California
Bearing Ratio
Resilient
Modulus-Laboratory
Resilient
Modulus
NDT
Table 8.1 Summary of Typical Subgrade Tests

Standard Test Method Purpose of Test
ASTM D2487-00
Soil classification is basic information which can
Standard Classification be used to estimate various design-related
of Soils for Engineering parameters. The required tests for classification
Purposes (Unified Soil
can be used for other determinations
Classification System)
(gradation, Atterberg Limits)
ASTM D1883-07e2
Straightforward test for determining relative
Standard Test Method
shear strength of the subgrade soils. CBR can
for CBR of Laboratorybe estimated from a laboratory test or through
Compacted Soils
correlations with devices such as the DCP
(Section 7). Caution is needed since laboratory
and field produced CBRs can have quite
different moisture conditionshence results.
AASHTO T307 Standard If subgrade soil samples are available,
Method of Test for
laboratory resilient modulus determinations
Determining the
can be made. Triaxial testing is expensive and
Resilient Modulus of
the results a function of careful sample
Soils and Aggregate
preparation.
Materials
ASTM D4694-96
The preferred test apparatus for
Standard
nondestructive testing of pavement structures
Test Method for
is the FWD (see Section 4). Straightforward
Deflections with a
methods for estimating MR are available
Falling-Weight-Type
(Section 4) or backcalculation procedures allow
Impulse Load Device
up to 3 pavement layers to be estimated.
62
8.3 Analysis Tools

Questions that need to be answered for the project assessment about subgrade soils include
the following:
How well do the subgrade soils support the existing pavement structure?
Are the subgrade soils frost susceptible (if the project is located within a potential freezing
zone)?
Are the subgrade soils subject to expansion and contraction (such as expansive clay soils)?
Are groundwater issues associated with the project site?
(i) Support for Existing Pavement Structure
The support for the existing pavement structure can be estimated through a combination of
laboratory or nondestructive testingbut most likely it will be NDT. A set of FWD deflection
basins, pavement coring, and DCP measurements is generally sufficient along with use of the
analysis tools provided in the preceding sections.
(ii) Frost Susceptibility
Both sophisticated and very straightforward soils tests are available for estimating the
likelihood of subgrade soil frost susceptibility. The basic issue is the potential for the creation of
ice lenses under the existing pavement and the resulting loss of support when it all thaws out.
When ice lenses form in frost susceptible soils, large volume changes can occur (just liquid
water changing to ice increases the its volume by 9%). A illustration of ice lenses in pavements
in shown in Figure 8.1.
A basic approach for assessing frost susceptibility is based on gradation and it has been in use
for almost 80 years. Casagrande noted the following in 1932 [reference for this content is
Terzaghi and Peck, 1967]:
Under natural freezing conditions and with sufficient water supply one should expect
considerable ice segregation in non-uniform soils containing more than 3% of grains smaller
than 0.02 mmNo ice segregation was observed in soils containing less than 1% of grains
smaller than 0.02 mm, even if the groundwater level is as high as the frost line.
To determine the percent passing 0.02 mm requires a hydrometer test. A reasonable
approximation of 3% passing 0.02 mm is about 7% passing a 0.075 mm (No. 200 sieve).
Another tool which can aid decisions about the potential frost susceptibility of a subgrade soil is
to use the US Army Corps of Engineers classification system for frost design (NCHRP Synthesis
26, 1974) as shown in Table 8.2.
63
Figure 8.1 Formation of Ice Lenses in a Pavement Structure
64
Table 8.2 US Army Corps of Engineers Frost Design Soil Classification (after US Army, 1990)
Percentage finer
Typical soil types under
Soil
Type
Frost Group
than 0.02 mm
Unified Soil Classification
by weight
System
Nonfrost (a) Gravels including
0 to 1.5
GW, GP
Susceptible
crushed stone and
(NFS)
crushed rock
Potentially Frost Susceptible Soils
PFS
(a) Gravels
Crushed stone
Crushed rock
1.5 to 3
GW, GP
(b) Sands
3 to 10
SW, SP
S1
Gravelly soils
3 to 6
GW, GP, GW-GM, GP-GM
S2
Sandy soils
3 to 6
SW, SP, SW-SM, SP-SM
F1
Gravelly soils
6 to 10
GM, GW-GM, GP-GM
3 to 10
GW, GP, GW-GM, GP-GM
(a) Gravelly soils
10 to 20
GM, GW-GM, GP-GM
(b)Sands
6 to 15
SM, SW-SM, SP-SM
F1
F2
F3
F4
Gravelly soils
(a) Gravelly soils
>20
GM, GC
(b)Sands, except very

fine silty sands
>15
SM, SC
(c) Clays, Pl > 12
CL, CH
(a) All silts
ML, MH
(b)Very fine silty sands
>15
SM
(c) Clays, Pl < 12
CL, CL-ML
(d)Varved clays and

other fine-grained,
banded sediments
CL, ML, and SM;

CL, CH, and ML;
CL, CH, ML, and SM
65
(iii) Expansion and Contraction

If these types of soils are present, attempt to answer the following:
Were the subgrade soils previously treated with materials such as lime?
Is the profile of the existing pavement stable?
(iv) Groundwater Issues
Groundwater issues if apparent, may require the geotechnical engineer to sort out.
8.4 References
Terzaghi, K. and Peck, R. (1967), Soil Mechanics in Engineering Practice, John Wiley and Sons,
1967.
US Army (1990), Design of Aggregate Surfaced Roads and Airfields, Technical Manual TM 5822-12, Department of the Army, September 1990.
66
Section 9
Traffic Loads for Design
9.1 Purpose
This section overviews the use of basic traffic information to estimate design loadings for pavement
design. The fundamental parameter that will be estimated is Equivalent Single Axle Loads (ESALs).

This subsection overviews the kind of traffic information needed to quickly estimate future ESALs.
(i) Tire Loads and Terminology
Typical truck and bus axles are shown in Figure 9.1 which illustrates single and tandem axles with
either single or dual tires.
States generally have regulations limiting allowable load per inch width of tire. This tire load
limitation varies from a high of 800 lbs/inch to a low of 450 lbs/inch. The primary impact of such
state laws has to do with the use of dual or single tires on a specific axle and steer axles.
Single Axle with Dual Tires
Single Axle with Single Tires
Tandem Axles with Dual Tires

Tandem Axles with Single Tires
Figure 9.1 Illustration of Typical Axle and Tire Configurations
67
(ii) Typical Federal and State Axle Load Limits

Typical federal and state axle load limits are:
Single axles: 20,000 lbs
Tandem axles: 34,000 lbs
Total truck gross weight: 80,000 lbs
(iii) FHWA Bridge Formula
A major additional limitation on US trucks and buses is the FHWA Bridge Formula. The total gross
weight in pounds imposed on the pavement by any group of two or more consecutive axles on a
vehicle or combination of vehicles shall not exceed that weight calculated by use of Equation 9.1
below. The bridge formula is needed since an individual set of bridge design computations cannot
be done for every type of truck that may use highways.
Bridge designers use a standard design vehicle for estimating critical stresses, strains, or deflections
in a bridge structure. This vehicle is designated HS-20-44 and has been referred to as an umbrella
loading. Federal law requires its use in bridge design for the Interstate system.
In effect, the bridge formula helps to ensure bridges are not "overstressed" due to almost infinite
number of truck-axle configurations and weights.
W = 500(NL/(N-1) + 12N + 36) Eq. 9.1
Where
W = maximum weight on any group of two or more consecutive axles to the nearest 500 lb,
L = distance between the extremes of any group of two or more consecutive axles, ft, and
N = number of axles in the group under consideration.
To illustrate, an example is a 5 axle truck with a 51 ft. separation from the steer axle to the rear
portion of the back tandem. If you wish to know the total vehicle allowable gross weight via the
Bridge Formula, then W = 500(5(51)/(5-1) + 12(5) + 36) = 80,000 lb.
(iv) Repetitions of Wheel Loads and ESALs
We must be able to convert wheel loads of various magnitudes and repetitions ("mixed traffic") to
an equivalent number of "standard" or "equivalent" loads for design purposes. The most
commonly used equivalent load is 18,000 lb (80 kN) equivalent single axle loads (normally
designated ESAL). The ESAL standard axle load is used in the AASHTO "Guide for Design of
Pavement Structures."
Wheel load equivalency has been one of the most widely adopted results of the AASHO Road Test
(1958 to 1960), i.e., to relate relative damage attributed to axles of different type (single and
tandem) and weight. Highway design in most states is based on the ESAL traffic input anticipated
over a future 10 to 50 year period.
68
The relationship between repetitions is not arithmetically proportional to the axle loading. Instead,
a 10,000 lb single axle needs to be applied to a pavement structure many more than 1.8 times the
number of repetitions of an 18,000 lb single axle to have the same effect in fact, more than 12
times. Similarly, a 22,000 lb single axle needs to be repeated less than half the number of times of
an 18,000 lb single axle to have an equivalent effect. A sample of ESAL load equivalency factors is
shown in Table 9.1.
Table 9.1 Sample of AASHTO Equivalency Factors
Axle Type
Axle Load
ESAL Load Equivalency
(lbs)
(lbs)
Factors [from AASHTO,
1993]
Single axle
2,000
0.0003
10,000
0.118
14,000
0.399
18,000
1.000
20,000
1.4
30,000
7.9
Tandem axle
2,000
0.0001
10,000
0.011
14,000
0.042
18,000
0.109
20,000
0.162
30,000
0.703
34,000
1.11
40,000
2.06
50,000
5.03
A basic element in estimating the future ESALs for a specific project is to forecast the truck and bus
volumes for the design (and analysis) period. Once this is done, LEFs in various forms can be
applied to the forecast volumes and summed.
A complete forecast will include the 13 FHWA vehicle classes (which are not the same vehicle
classes as used by vehicle manufacturers). These classes are shown in Table 9.2.
69
FHWA
Vehicle
Class
Class 1
Class 2
Class 3
Class 4
Class 5
Class 6
Class 7
Class 8
Class 9
Class 10
Class 11
Class 12
Class 13
Table 9.2 FHWA Vehicle Classes

Vehicle Class Description
Motorcycles (Optional) All two- or three-wheeled motorized vehicles.

Typical vehicles in this category have saddle type seats and are steered by
handle bars rather than wheels. This category includes motorcycles, motor
scooters, mopeds, motor-powered bicycles, and three-wheel motorcycles.
This vehicle type may be reported at the option of the State.
Passenger Cars All sedans, coupes, and station wagons manufactured
primarily for the purpose of carrying passengers and including those passenger
cars pulling recreational or other light trailers.
Other Two-Axle, Four-Tire Single Unit Vehicles All two-axle, four tire,
vehicles, other than passenger cars. Included in this classification are pickups,
panels, vans, and other vehicles such as campers, motor homes, ambulances,
hearses, and carryalls. Other two-axle, four-tire single unit vehicles pulling
recreational or other light trailers are included in this classification.
Buses All vehicles manufactured as traditional passenger-carrying buses
with two axles and six tires or three or more axles. This category includes only
traditional buses (including school buses) functioning as passenger-carrying
vehicles. All two-axle, four-tire single unit vehicles. Modified buses should be
considered to be a truck and be appropriately classified.
Two-Axle, Six-Tire, Single Unit Trucks All vehicles on a single frame including
trucks, camping and recreational vehicles, motor homes, etc., having two axles
and dual rear wheels.
Three-Axle Single Unit Trucks All vehicles on a single frame including trucks,
camping and recreational vehicles, motor homes, etc., having three axles.
Four or More Axle Single Unit Trucks All trucks on a single frame with four or
more axles.
Four or Less Axle Single Trailer Trucks All vehicles with four or less axles
consisting of two units, one of which is a tractor or straight truck power unit.
Five-Axle Single Trailer Trucks All five-axle vehicles consisting of two units,
one of which is a tractor or straight truck power unit.
Six or More Axle Single Trailer Trucks All vehicles with six or more axles
consisting of two units, one of which is a tractor or straight truck power unit.
Five or Less Axle Multi-Trailer Trucks All vehicles with five or less axles
consisting of three or more units, one of which is a tractor or straight truck
power unit.
Six-Axle Multi-Trailer Trucks All six-axle vehicles consisting of three or more
units, one of which is a tractor or straight truck power unit.
Seven or More Axle Multi-Trailer Trucks All vehicles with seven or more
axles consisting of three or more units, one of which is a tractor or straight
truck power unit.
70
A somewhat simplified scheme for summarizing the 13 vehicle classes in Table 9.2 is to group all
truck and bus traffic into three groups or units as shown in the Table 9.3.
Table 9.3 Simplified Truck and Bus Groups
Groupings of FHWA Vehicle Classes
Simplified Vehicle Categories

Single Units which includes
Single Trailers which includes
Multi-Trailers which includes
(i)
(ii)
(iii)
(iv)
Buses
2 axle, 6 tire single units
3 axle single units
4+ axle single units
(FHWA Class 4)
(FHWA Class 5)
(FHWA Class 6)
(FHWA Class 7)
(i)
(ii)
(iii)
(i)
(ii)
(iii)
4 axle single trailer

5 axle single trailer
6+ axle single trailer
5 axle multi-trailer
6 axle multi-trailer
7+ axle multi-trailer
(FHWA Class 8)
(FHWA Class 9)
(FHWA Class 10)
(FHWA Class 11)
(FHWA Class 12)
(FHWA Class 13)
9.3 Analysis Tools

This subsection will focus on how to organize ESAL data so that an overall ESAL estimate for the
design period can be made.
Table 9.4 shows typical ESALs per vehicle according to the groupings in Table 9.3. The ESALs per
vehicle were developed by a State DOT and appear to be typical for US truck traffic. They may
appear to be low but the values are averages which include empty backhauls
Table 9.4 ESALs per Vehicle for Simplified Vehicle Groups
Simplified Vehicle
Categories
FHWA Classes
Average ESALs per Vehicle
Single Unit Trucks
4, 5, 6, 7
0.40
Trucks with Single Trailers
8, 9, 10
1.00
Trucks with Multi-Trailers
11, 12, 13
1.75
1.60
Buses (1/2 full)
Thus, if you estimated that a specific highway has daily (one-way) 1,000 single unit trucks, 2,000
trucks with single trailers, and 500 trucks with multi-trailers and no buses, then the daily ESALs
would be [1,000(0.4) + 2,000(1.00) + 500(1.75)] = 3,275 ESALs per day or about 1,200,000 ESALs per
year. The annual value can be scaled up to the design period with a suitable growth rate (typically 2
to 3 percent).
71
The approximate annual ESALs were estimated for various levels of ADT and are shown in Table
9.5. The primary (and major) assumptions were the typical truck (average over FHWA Classes 4
through 13 based on actual weigh data) has an average ESAL = 1.0. Further, the range of truck
percentages as a function of ADT is typically 5 to 10%. Given these assumptions, the following can
be used as a guide.
Table 9.5 Approximate Annual ESALs as a Function of ADT
Average Daily Traffic
Approximate Annual ESALs
10,000
250,000
25,000
750,000
50,000
1,500,000
100,000
2,500,000
The annual ESALs can vary significantly from the values in Table 9.5. Examples include routes with
large numbers of buses which generally have higher ESALs per vehicle than an average truck (often
by a factor of two to three or more). The percentage of trucks in the total traffic can vary
significantly and particularly so in the vicinity of seaports. Further, rural interstate routes typically
have higher percentages of trucks than urban located interstates along with differing mixes of
trucks. Urban areas will have higher single unit truck percentages than rural locations.
9.4 References
AASHTO (1993), AASHTO Guide for Design of Pavement Structures, 1993, American Association
of State Highway and Transportation Officials, Washington, DC.
72
Section 10
Construction Productivity and Traffic Impacts
10.1 Purpose
This section overviews the various methods for determining construction productivity and traffic
impacts of pavement and roadway construction. Traffic impacts can often make up the largest
societal cost associated with a paving project sometimes being an order of magnitude more than
the agency cost to build/rehabilitate the pavement. An early understanding of productivity and
potential traffic impacts can assist the project in determining the most advantageous construction
timing, project sequencing (staging) and lane closure scenarios. Often, full roadway closures (in
contrast to repeated partial closures) over longer periods of time (e.g., full weekends or multiple
days instead of nighttime only closures) can prove to be the least costly alternative if user costs are
properly accounted for in construction planning.

Traffic impacts are typically quantified by user delay with typical metrics being (1) total user delay,
(2) total user cost associated with delay, (3) maximum vehicle queue length, and (4) maximum time
in vehicle queue. Usually, the goal of minimizing traffic impacts is interpreted to mean minimizing
the total user cost attributable to the existence of the project workzone. Other important
considerations (e.g., accident/incident minimization, avoidance of certain public event days that
generate high traffic, etc.) may cause the ultimate traffic impacts to be somewhat greater than the
optimal minimum. Nonetheless, it is useful to estimate, as accurate as practical considerations
allow, the minimum traffic impact scenario for pavement construction. Generally, this estimate
uses the following six basic actions:
1. Determine construction productivity. This involves estimating the productivity of basic
construction processes involved in the project such as demolition crew speed/efficiency, dump
truck number/capacity, paver speed and materials manufacturing plant productivity. It also
involves estimating mobilization/demobilization times, concrete cure time, hot mix asphalt
cooling time and traffic control setup time. There may be several estimates of each depending
upon the construction scenarios being investigated.
2. Measure existing traffic. While an actual time history is best (e.g., from loop detector
information or manual counts) average daily traffic (ADT) can be used and hourly traffic
volumes can be developed multiplying ADT by typical hourly distribution factors for the type of
roadway being analyzed.
3. Estimate the fraction of traffic that will cancel their trips and the fraction of traffic that will
use detour routes during the construction. At best, these will be rough estimates unless more
sophisticated models are used. These estimates are also highly dependent on the publicity
given the roadway work. Values can be obtained from:
a. Agency experience with similar closures and similar publicity in the past.
b. A general literature review of similar traffic closures.
73
4. Develop construction scheduling (staging) alternatives. This involves determining the number,
duration and sequence of lane closures required to complete the project. As the traffic impact
analysis progresses it is often necessary to refine these alternatives. Strong consideration
should be given to scheduling alternatives that result in workzone traffic capacity greater than
traffic demand during the hours of work. Essentially, this results in little or no user cost
attributed to the roadway work. However, such scheduling alternatives may not exist or be
feasible from a construction productivity and/or constructability standpoint. Any number of
lane closure scenarios can be considered but it is helpful to at least investigate the following
four scenarios:
a. Partial night closures: closure only during night hours with light traffic where each roadway
direction is still open although with reduced capacity in at least one direction. These
closures are often the first considered since they tend to minimize traffic impacts by only
closing lanes when traffic is the lightest. However, they may not provide the lowest user
costs because mobilization/demobilization can take up a large percentage of total closure
time resulting in low overall productivity. In some scenarios, it may not be possible to make
any meaningful progress in a short nighttime closure. Even if partial night closures cannot
be used for mainline paving, they are often useful for pre-paving work (e.g., PCC panel
sawcutting, restriping lanes, milling HMA, etc.).
b. Full night closures: same as above but with at least one roadway direction fully closed.
These may involve detouring an entire direction, counterflowing traffic on one side of a
highway or using a pilot car to alternate traffic directions in one lane. Full night closures are
sometimes required to do such things as set up counterflow traffic on one side of a roadway
or accomplish dangerous overhead work such as overpass demolition/placement.
c. Partial day closures: closure only during day hours with where each roadway direction is
still open although with reduced capacity in at least one direction. These closures are often
the first considered for lightly trafficked roadways where user delay is unexpected even
with some lanes closed. If traffic delays are minimal, day closures can improve safety by
providing better visibility and encountering fewer impaired drivers than night work and
reduce construction costs by avoiding overtime pay. However, they may not provide the
lowest user costs because mobilization/demobilization can take up a large percentage of
total closure time resulting in low overall productivity. In some scenarios, it may not be
possible to make any meaningful progress in a short day closure.
d. Full day closures: same as above but with at least one roadway direction fully closed. These
may involve detouring an entire direction, counterflowing traffic on one side of a highway
or using a pilot car to alternate traffic directions in one lane. Full day closures are usually
only feasible for lightly trafficked roadways or roadways with large capacity detour routes
that do not add significantly to commute time.
e. Partial or full weekend continuous closures: closure starting Friday evening after peak hour
traffic and ending Monday morning before peak hour traffic. The typical scenario is a 55hour weekend closure starting at 9 or 10 p.m. Friday night and ending at 4 or 5 a.m. on
Monday morning. The long closure time allows for better productivity because
mobilization/demobilization takes up a smaller fraction of total closure time and, more
importantly, because construction crews generally get better and faster in their work given
74
a longer working window. Weekends are typically preferred because weekend traffic is
usually more discretionary (leading to more canceled trips and less total user delay) and
sometimes lighter than weekday traffic.
f. Partial or full week-long continuous closures: closures that are maintained continuously
over an entire week (168 hours). While it may not be known if any closure windows will
extend over a week or more, estimating this alternative will generally allow estimation of
longer closure windows with reasonable accuracy. For instance, the productivity for a 3week closure is roughly, but not exactly (due to mobilization/demobilization times), three
times the productivity of a 1-week continuous closure.
5. Model traffic using the tool of choice (see Analysis Tools Section). This modeling will result in
an estimate of total user cost for the roadway project. In general, larger projects on major
routes warrant more modeling while smaller projects on minor routes can often be estimated
sufficiently using spreadsheets.
6. Apply FHWA Interim Report Life-Cycle Cost Analysis in Pavement Design (Walls and Smith
1998) standards to estimate user delay cost. This report provides reasonable values for user
time (Table 10.1). This table is in 1996 dollars and should be adjusted to current dollars using
the Consumer Price Index (CPI). A simple calculator is available from the U.S. Bureau of Labor
Statistics (BLS) at: http://www.bls.gov/data/inflation_calculator.htm. Multiplying these values
by total delay for each class of vehicle gives an estimate of total workzone user delay cost.
Table 10.1 Recommended Values of Time (from Walls and Smith 1998)
Vehicle Class
Passenger Vehicles
Single-Unit Trucks
Combination Trucks
$ Value per Vehicle Hour (1996 dollars)

Value
Range
$11.58
$10 to 13
$18.54
$17 to 20
$22.31
$21 to 24
(i) General Guidance

The following general guidance for traffic impacts comes largely from the guidance documents
listed in the references portion of this section:
Closure Scenarios
Productivity is usually much higher and worker safety is greater with longer, more complete
closures, e.g., full closures, weekend closures, etc. (FHWA 2003)
The public is generally very accepting of full closures or a few longer duration closures as an
alternative to lengthy schedules of night or day closures (FHWA 2003).
As a work zone remains in effect for a longer period of time (e.g., over several days or
several weekends) the fraction of drivers either canceling their trips or taking the detour
75
route is likely to decrease as drivers become used to the situation or determine that a trip
can no longer be put off.
Detour routes may experience several times their normal traffic volumes (Lee and Harvey
2006; Lee et al. 2001). It may be prudent to improve detour route capacity through
additional lanes, a temporarily reversible lane, signal retiming or other improvements
(FHWA 2003).
For major highway jobs, the construction of one lane usually requires a second adjacent
lane for access. This means either using an existing wide shoulder (e.g., 10 ft shoulder) if
one exists or closing a second lane (Lee 2008).
For major highway jobs, if the lane under construction has more than one major activity
underway on it simultaneously (e.g., demolition and paving) a second access lane will likely
be needed to avoid stationary trucks in the adjacent lane (Lee 2008).
Avoid creating work zones with live traffic on both sides (e.g., in the middle lanes in one
direction). These generally do not leave workers a safe exit from the work zone if it is
compromised.
It may be better to use a simpler lane closure plan that is more easily understood by the
public even if it does not result in minimum modeled user delay.
Contracting
Lane rental or time-based bonus/penalty contracts should have a clear clause describing
how to address changed conditions or any situation where the owner wishes to add work
that impacts productivity (Lee et al. 2007). Often, contractors plan to spend more money
than the contract price in order to finish early and receive the bonus. In this scenario,
without bonus payments the contractor will lose money.
Contracts that contain bonus/penalty amounts for speed and quality should balance these
amounts so that it does not become advantageous to sacrifice one bonus to get the other
(Muench et al. 2007). For instance, of a maximum quality bonus/penalty is $3,000 but the
maximum speed bonus/penalty is $100,000 then in some scenarios it may be logical to
sacrifice a small quality bonus for a large speed bonus.
Productivity
The slowest process in a reconstruction project is often demolition (Lee et al. 2007). If
several processes are being done simultaneously, demolition will most often control the
overall productivity.
The rate at which dump trucks can be filled by an excavator or milling machine is relatively
consistent from job-to-job (Lee et al. 2007). Therefore, the best estimate is often what
happened on the previous job. If no local information is available, Lee et al. (2007) provides
good baseline estimates.
Production rate is often controlled by access to the construction site and allowances made
for traffic (e.g., temporary off-ramps in work zones, separation between work zone and
traffic).
76
Workzone Capacity
Work zone capacity is highly variable and only moderately predictable. Work zone capacity
can be affected by the number of lanes open, intensity of work, the presence of ramps,
fraction of heavy vehicles, lane width, lateral clearance, work zone grade and more.
Highway Capacity Manual (HCM) procedures are very rough but the Highway Capacity
Manual (HCM) (TRB 2000) does suggest 1,600 passenger cars per lane per hour (pc/ln/hr)
be used as a baseline for short-term work zones. Typically this number is adjusted
downward based on other factors and can be as low as about half the original value.
The more a work zone can be physically and visually separated from traffic (e.g., semipermanent barriers like jersey barriers or k-rails instead of traffic cones or barrels) the
greater the work zone traffic capacity.
Incidents (i.e., accidents, stalled vehicles, etc.) are one of the largest contributors to work
zone user delay because there are fewer lanes (if any) that traffic can use to bypass the
incident. Dedicating resources (e.g., incident response vehicle, video cameras, variable
message boards, traffic management center) to reduce incidents and clear them more
quickly can be a cost effective way to minimize user delay (FHWA 2004).
Publicity
Roadway work and closure publicity can be effective in drastically reducing traffic during
work zone closures. Often, several-mile long queues predicted using normal traffic volumes
never materialize because many drivers cancel their trips or alter their routes.
Even if a local public information campaign is effective it may still be difficult to get closure
information to travelers or freight carriers out of the local area who plan on using the
affected roadway.
10.3 Analysis Tools

This subsection overviews some of the more popular methods for determining traffic impacts for
pavement construction projects and factors that influence the choice of tools. Some key
considerations when selecting tools:
How much detail is needed? Work zone characteristics, desired outputs and the stage of
planning/design/construction will influence tool choice. Often a simpler tool, with less detail is
adequate.
Is the tool calibrated to the local area? If not, results may still be useful, however accuracy may
be less than expected or needed.
Is the tool stochastic or deterministic? Construction productivity and traffic can be highly
variable and difficult to predict. While a deterministic model can provide a single number it is
better to provide a reasonable range of answers in order to capture the variable nature of
productivity and traffic.
How much detail does the tool produce? Some tools can only estimate traffic impacts over one
24-hour period while some can estimate over much longer time periods. Some tools can only
estimate delay on an hourly basis while some can estimate them in much smaller time
77
increments. Some tools make estimates using one single days traffic input while others are
able to account for daily, weekly and monthly traffic variations.
(i) Analysis Tools: Construction Productivity
Construction productivity tools discussed are: manual methods, standard estimating software and
Construction Analysis for Pavement Rehabilitation Strategies (CA4PRS).
Manual method. Demolition and paving productivity estimates can be made manually by
comparing productivities of the constituent processes and identifying the limiting factor. There are
a few references to help in paving productivity calculations. The National Asphalt Pavement
Association (NAPA) publishes Balancing Production Rates in Hot Mix Asphalt Operations (IS 120),
which contains a step-by-step guide for determining HMA paving productivity. Several companies
also offer custom printed asphalt productivity slide rules that paving companies can purchase and
brand to be given out to potential customers.
Estimating software. Most estimating software (e.g., Bid2Win, HeavyBid) assists users in
calculating the productivity of construction processes.
CA4PRS. A Microsoft Access-based software tool that can be used to analyze highway pavement
rehabilitation strategies including productivity, project scheduling, traffic impacts, and initial
project costs based on input data and constraints supplied by the user. The goal is to help
determine roadway rehabilitation strategies that maximize production and minimize costs without
creating unacceptable traffic delays. As of 2009, all state transportation departments have free
group licenses for CA4PRS.
(ii) First Order Productivity Estimates
In the early planning stages of a project it may be useful to quickly determine rough construction
productivity based on a few known parameters. This section displays productivity graphs produced
using CA4PRS with most inputs being held constant a typical values. The purpose of these graphs is
only to give a rough estimate of typical productivity. CA4PRS should be used to produce more
accurate numbers based on actual site-specific parameters for use in any project planning. In
general, most inputs were fixed except for the trucking rates (i.e., removal of demolition from the
site and delivery of paving material to the site). Thus, the 95% confidence intervals seen are mostly
dependent on these delivery rates. In all cases, a 10-mile stretch of two lanes was analyzed (20
lane-miles total). As with all that data input values, this length of highway and total lane-miles has
some influence on productivity. Tables 10.2 through 10.7 show input parameters use in CA4PRS to
generate Figures 10.1 through 10.9. Estimates are given for:
Remove-and-replace with PCC. Remove the existing pavement and replace with the same
depth of new PCC pavement. Productivity is estimated for sequential operations (only one
major operations demolition or paving is occurring on the jobsite at any one time) and
concurrent operations (both major operations demolition and paving are occurring on the
jobsite at once, with the appropriate space in between). One lane is paved at a time. Sequential
operations require one additional lane shut down for construction access, while concurrent
78
operations require two additional lanes shut down for construction access. Calculations were
made for both screed paving (slower) and slipform paving (faster).
o Screed paving. Using fixed forms and a screed, this paving is usually slower. Assumes 7.5 yd3
agitating mixers arriving at 10 trucks/hr and only one demolition crew.
o Slipform paving. Using a slipform paver, this paving is usually faster. Assumes 8.5 yd 3 end
dump trucks arriving at 17 trucks/hr and two demolition crews.
Remove-and-replace with HMA. Remove the existing pavement and replace with the same
depth of new HMA pavement. The roadway lanes being paved are fully shut down, only one
paver with a 12-ft wide screed is used and HMA is paved in lifts. Lifts are generally 3 inches
thick with the exception of the top two lifts, which are either 2 or 1.5 inches thick. A lift is paved
for each lane across before the next lift is paved on any lane.
Mill-and-fill with HMA. Remove a predetermined thickness from the existing pavement with a
HMA milling machine then replace the same thickness with new HMA. The roadway lanes being
paved are fully shut down, only one paver with a 12-ft wide screed is used and HMA is paved in
lifts. Lifts are generally 3 inches thick with the exception of the top two lifts, which are either 2
or 1.5 inches thick. A lift is paved for each lane across before the next lift is paved on any lane.
Crack, seat and overlay. Crack and seat the existing PCC pavement then overlay with HMA. The
roadway being paved is fully shut down, only one paver with a 12-ft wide screed is used and
HMA is paved in lifts. Lifts are generally 3 inches thick with the exception of the top two lifts,
which are either 2 or 1.5 inches thick. A lift is paved for each lane across before the next lift is
paved on any lane.
Unbonded PCC overlay. Prepare the surface of the existing PCC pavement then overlay with
PCC that is not bonded to the existing pavement. This is essentially like the remove-andreplace with PCC without the demolition component.
79
Table 10.2. CA4PRS Input Values for Remove-and-Replace with PCC

Input
Value
Distribution/Comments
Activity Constraints
Mobilization
1.0 hours
None - Deterministic
Demobilization
2.0 hours
Base Paving
none
N/A
Demo-to-PCC Paving Lag
1.0 hours
Triangular (min = 0.5 hrs, max = 1.5 hrs)
Times for Sequential
Method
2.0 hours
Times for Concurrent
Method
Resource Profile
Demolition Hauling Truck
Rated Capacity
18.0 tons
9 yd3 of a 15 yd3 truck filled w/2.0 tons/yd3 material
Trucks/hr/team
10 trucks
Triangular (min = 8 trucks, max = 12 trucks)
Packing Efficiency
1.0
Number of Teams
1.0
1 team for screed paving, 2 teams for slipform
2.0
None Deterministic
Team Efficiency
0.90
Triangular (min = 0.85, max = 0.95)
Base Delivery Truck
None
N/A (no base material)
Batch Plant
Capacity
500 yd3/hr None Deterministic
(set high to ensure plant is not the limiting activity)
Number of Plants
1
Concrete Delivery Truck
Capacity
7.5 yd3
N/A
Trucks per Hour
The first rate is for screed paving and the second is
for slipform paving.
10/hr
Triangular (min = 8/hr, max = 12/hr)
13/hr
Packing Efficiency
1.0
Paver
Speed
5 ft/min
Number of Pavers
1
Schedule Analysis
Construction Window
see graphs
Section Profile
see graphs Note: no base material included in graphs
Change in Roadway
No Change
Elevation
Lane Widths
12 ft.
Curing Time
12 hours
Working Method
see graphs
80
Table 10.3. CA4PRS Input Values for Remove-and-Replace with HMA

Input
Value
Mobilization
1.0 hours
Demobilization
2.0 hours
Base Paving
none
N/A
Demo-to-HMA Paving Lag 1.0 hours
Half Closure Traffic Switch 0.5 hours
Resource Profile
Rated Capacity
18.0 tons
Trucks/hr/team
10 trucks
Packing Efficiency
1.0
Number of Teams
1.0
Team Efficiency
0.90
Paver
None
Non-Paving Speed
15 mph
Batch Plant
Capacity
Number of Plants
1
HMA Delivery Truck
Capacity
18 tons
N/A
Trucks per Hour
12/hr
Packing Efficiency
1.0
Schedule Analysis
Construction Window
see graphs
Section Profile
see graphs Top two lifts are 2 inches each, all other lifts are 3
inches each. Paver moves at 0.6 mph for top two
lifts and 0.5 mph for all other lifts.
Change in Roadway
No Change
Elevation
Shoulder Overlay
Pre-Paving Shoulder overlays are not accounted for
Curing Time
12-hours
Working Method
see graphs
Cooling Time Analysis
User Spec. Time calculated in MultiCool and manually entered
Lane Widths
No. of Lanes
2
Lane Widths
12 ft each
81
Table 10.4. CA4PRS Input Values for Mill-and-Fill with HMA

Input
Value
Mobilization
1.0 hours
Demobilization
2.0 hours
Mill-to-HMA Paving Lag
1.0 hours
Resource Profile
Milling and Hauling
Number of Teams
1.0
Team Efficiency
0.90
Milling Machine
Class
Large
Material Type
AC-Hard
Efficiency Factor
0.90
Hauling Truck
Rated Capacity
18.0 tons
Trucks/hr/team
13 trucks
Packing Efficiency
1.0
Batch Plant
Capacity
Number of Plants
1
HMA Delivery Truck
Capacity
18 tons
N/A
Trucks per Hour
12/hr
Packing Efficiency
1.0
Paver
None
Non-Paving Speed
15 mph
Schedule Analysis
Construction Window
see graphs
Section Profile
see graphs Lifts are between 1.5 and 3 inches. Paver speeds
are 0.5 to 0.6 mph.
Change in Roadway
No Change
Elevation
Shoulder Overlay
Curing Time
12-hours
Working Method
see graphs
Lane Widths
No. of Lanes
2
Lane Widths
12 ft each
82
Table 10.5. CA4PRS Input Values for Crack, Seat and Overlay
Input
Value
Mobilization
3.0 hours
Demobilization
2.0 hours
Resource Profile
Paver
None
Non-Paving Speed
15 mph
Batch Plant
Capacity
Number of Plants
1
HMA Delivery Truck
Capacity
18 tons
N/A
Trucks per Hour
12/hr
Packing Efficiency
1.0
Schedule Analysis
Construction Window
see graphs
Section Profile
see graphs Top two lifts are 2 inches each, all other lifts are 3
inches each. Paver moves at 0.6 mph for top two
lifts and 0.5 mph for all other lifts.
Change in Roadway
No Change
Elevation
Shoulder Overlay
Curing Time
12-hours
Working Method
see graphs
Lane Widths
No. of Lanes
2
Lane Widths
12 ft each
83
Table 10.6. CA4PRS Input Values for Unbonded PCC Overlay

Input
Value
Mobilization
3.0 hours
(longer time accounts for surface preparation)
Demobilization
2.0 hours
Base Paving
none
N/A
0 hours
No demolition occurs
Times for Sequential
Method
0 hours
No demolition occurs
Times for Concurrent
Method
Resource Profile
High numbers are a work-around to make
demolition take essentially no time
Rated Capacity
100.0 tons None - Deterministic
Trucks/hr/team
100 trucks None - Deterministic
Packing Efficiency
1.0
Number of Teams
100.0
Team Efficiency
1.00
Base Delivery Truck
None
Batch Plant
Capacity
Number of Plants
1
Concrete Delivery Truck
Capacity
7.5 yd3
N/A
Trucks per Hour
10/hr
Packing Efficiency
1.0
Paver
Speed
5 ft/min
Number of Pavers
1
Schedule Analysis
Construction Window
see graphs
Section Profile
see graphs Note: no base material included in graphs
Change in Roadway
No Change
Elevation
Lane Widths
12 ft.
Curing Time
12 hours
Working Method
see graphs
84
Table 10.7. MultiCool Input Parameters for HMA Options
Input
Value
Constant Inputs in All Scenarios
Start Time
1000, 7/15/2010
Environmental Conditions
Ambient Air Temp.
60F
Average Wind Speed
5 mph
Sky Conditions
Clear&Dry
Latitude
38 North
Existing Surface
Material Type
Granular Base
Moisture Content
Dry
State of Moisture
Unfrozen
Surface Temp.
60F
Mix Specifications
Mix Type
Dense Graded
PG Grade
64-22
Delivery Temp.
300F
Stop Temp.
140F
Lift Thicknesses
3 inches of HMA total
2 lifts of 1.5 inches each
3 lifts of 2 inches each
3 lifts of 2 inches, 1 lift of 3 inches
2 lifts of 1.5 inches 3 lifts of 3 inches
85
Thickness of Pavement Constructed (inches)
14
12
10
10-hour Night Closure (not feasible)

55-hour Weekend Closure
168-hour 24/7 Continuous Week-Long Closure
0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Lanes-MIles Constructed in One Closure
Figure 10.1. Productivity estimates for remove-and-replace with PCC (fixed form) using sequential
operations. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Note:
this option is not feasible using 10-hour night closures.
86
14
12
10

2
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Figure 10.2. Productivity estimates for remove-and-replace with PCC (slipform) using sequential
operations. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Note:
this option is not feasible using 10-hour night closures.
87
14
12
10
4
2
0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Figure 10.3. Productivity estimates for remove-and-replace with PCC (fixed form) using concurrent
operations. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Notes:
(1) this option is not feasible using 10-hour night closures, (2) doing demolition and paving
concurrently results in significantly higher productivities than doing them sequentially.
88
14
12
10
4
2
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Figure 10.4. Productivity estimates for remove-and-replace with PCC (slipform) using concurrent
operations. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Notes:
(1) this option is not feasible using 10-hour night closures, (2) doing demolition and paving
concurrently results in significantly higher productivities than doing them sequentially.
89
14
12
10-hour Night Closure
10

8
0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
Figure 10.5. Productivity estimates for remove-and-replace with HMA. Solid lines indicate averages
and dashed lines indicate 95% confidence intervals.
90
6

4
0
0.0
5.0
10.0
15.0
20.0
25.0
Figure 10.6. Productivity estimates for mill-and-fill with HMA. Solid lines indicate averages and
dashed lines indicate 95% confidence intervals.
91
10
9

6
5
4
3
2
1
0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
Figure 10.7. Productivity estimates for crack, seat and overlay. Solid lines indicate averages and
92
14
12
10

0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Figure 10.8. Productivity estimates for PCC unbounded overlay. Solid lines indicate averages and
93
14
Crack, Seat and Overlay (CSOL)

12
Remove-and-Replace with HMA

Remove-and-Replace with PCC (Fixed Form)
10
Remove-and-Replace with PCC (Slipform)

8
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
Figure 10.9. A productivity comparison of PCC remove-and-replace (both fixed form and slipform),
HMA remove-and-replace and crack, seat and overlay (CSOL).
(iii) Analysis Tools Traffic Impacts
There are a number of analysis tools available to assist in work zone traffic impacts estimation. The
FHWA divides these tools up into six broad categories (Hardy and Wunderlich 2008 and
summarized in Table 10.8):
1. Sketch-planning tools. Specialized models designed for work zone analysis. These models
can vary from simple spreadsheet calculations to general delay estimation tools. Typically,
models are deterministic and based on simple queuing equations or volume-to-capacity
relationships from the Highway Capacity Manual (HCM). Such simple estimation tools are
often adequate for work zone delay estimation.
2. Travel demand models. Forecast future traffic demand based on current conditions, and
future predictions of household and employment centers (Alexiadis et al. 2004). Travel
demand models are usually used in large regional planning efforts. In work zone analysis
they can help predict region-wide impacts of extended roadway closures (e.g., closing a
freeway for several months). It is not likely that a travel demand model would be built for
94
3.
4.
5.
6.
the specific purpose of work zone traffic analysis. Rather, an existing model may be used if
available and warranted.
Traffic signal optimization tools. Used to develop signal timing plans. These can be useful if
a temporary signal is used or if signals are retimed to accommodate work zone traffic or
increased detour route traffic.
Macroscopic simulation models. Based on the deterministic relationships of traffic speed,
flow and density (Alexiadis et al. 2004). These models treat flow as an aggregate quantity in
a defined area and do not track individual vehicles. They are useful for modeling larger area
impacts of work zones because of their aggregate nature.
Mesoscopic simulation models. Represent relative flow of vehicles on a network but do not
model individual lanes or vehicles. These models are between macroscopic and microscopic
models in detail. These models can simulate both large geographic areas as well as specific
corridors. They do not, however possess the detail to model more detailed strategies such
as signal timing. These models require large amounts of data.
Microscopic simulation models. Simulate the movement of individual vehicles. These
models require large amounts of data and can get unwieldy when attempting to simulate a
large network. Often these models can provide animated output which can clearly
communicate to decision-makers and the public what the potential traffic impacts of
modeled actions will be.
Table 10.8. Traffic Model Types for Work Zone Traffic Impacts
Examples
Strengths
Weaknesses
HDM, QUEWZ-98,
Low cost, specific to
Limited modeling
QuickZone, CA4PRS
work zones, fast
ability, not well
supported
Travel demand
EMME/2, TransCAD,
Can model large areas Low detail, cannot
TRANSIMS
model short term work
zone effects
Signal optimization
PASSER, Synchro
Models signal timing
Does not model other
and coordination
things
Macroscopic
BTS, KRONOS,
Can model large areas Low detail, cannot
METACORE/METANET,
model short term work
TRANSYT-7F
zone effects
Mesoscopic
CONTRAM,
Good compromise
Data intensive
DYNASMART,
between macro- and
DYNAMIT, MesoTS
micro models
Microscopic
CORSIM, VISSIM,
Can model small
Data intensive
PARAMICS
details, good
communication tool
Model Type
Sketch-planning
95
The most appropriate modeling approach depends upon (Hardy and Wundurlich 2008):
Work zone characteristics. The expected level of impact a work zone will have on travelers
including the geographic scale of affected area and complexity of the road network within
this area.
Transportation management plan strategies. The means by which traffic will be managed
including such items as lane closures, full roadway closures, lane shifts, counterflow traffic,
night/day work, detours, weekend work, etc.
Data availability and quality. The type, amount, accuracy and timeliness of available data.
Agency resources. The owner agencys funding, technical staff, and schedule.
Work zone performance measures. The performance measures selected by the owner
agency to quantify traffic impacts. Typically this is some form of delay (in minutes or cost)
either in total (total delay/cost) or peak (longest queue, longest wait).
Since the use of modeling tools beyond sketch-planning tools will almost surely require traffic
expertise beyond the pavement profession, further discussion is limited to a few sketch-planning
tools that may be of use: QuickZone and CA4PRS. Both of these tools can provide meaningful traffic
impact estimates for a relatively small money and time investment.
QuickZone 2.0. A Microsoft Excel-based tool (requires Excel 97 as a minimum) that estimates work
zone traffic impacts. It allows the user to input a node-and-link network (see Figures 10.10 and
10.11) then assign traffic counts to that network. It can coarsely simulate traffic variations between
days of the week, and months of the year by applying multiples to standard average daily traffic
(ADT) inputs. It can simulate multiple lane closures over time, model traffic over an entire week
(Figure 10.12) and display various traffic impact metrics (Figure 10.13). These capabilities are
helpful because they allow QuickZone to show difference in traffic impacts between nights and
days, weekends and weekdays, and seasons (e.g., summer vs. fall work). The user guide explains
the algorithm QuickZone uses to estimate delay and user cost but specific equations are not listed
or discussed. QuickZone is inexpensive (about $200) but is getting relatively old (version 2.0 was
released in 2005) without any significant upgrade or support beyond a user guide. Simple scenarios
with just a few links and nodes are relatively easy to simulate, however more complex scenarios
become cumbersome due to tedious data entry and difficult input troubleshooting if outputs are
suspect.
96
Figure 10.10 A simple network that works quite well in QuickZone.
Figure 10.11. A complex network simulation in QuickZone (I-5 in the Seattle, WA area is shown).
This network simulation exposed several program bugs, was unwieldy to process and required
tedious troubleshooting to make operational. This level of complexity is not recommended.
97
After Case Queue Length (Miles) for Inbound Direction from Phase Monday 8am to Saturday
midnight
20.00
18.00
16.00
Queue Length (Miles)
14.00
12.00
10.00
Reconstruction
area (70-73.64)
8.00
6.00
4.00
2.00
0.00
Time of Day
Figure 10.12. Unedited QuickZone 2.0 simulation output chart for a one-week time period. Note
that the automatic graph labeling on the horizontal axis is unreadable, however this can be
corrected by editing the graph in Excel.
Period with highest delay
in After Case
2 - Sunday
midnight to
Thursday 8am
Phase
Direction
Inbound
Day/Time
Sunday 21:00
Total Project User Cost ($)
Baseline
After
Total
Max Queue (Miles) Max Delay (min) Passenger Cars

Truck
Detour Econ/Misc
0
0
$0
$0
$0
$0
34.13
778.23
$22,036,199
$1,851,536
$0
$0
34.13
778.23
$22,036,199
$1,851,536
$0
$0
Total
$0
$23,887,735
$23,887,735
Figure 10.13. QuickZone 2.0 summary tables showing available traffic impact metrics.
CA4PRS. A Microsoft Access-based software tool that can be used to analyze highway pavement
rehabilitation strategies including productivity, project scheduling, traffic impacts, and initial
98
project costs based on input data and constraints supplied by the user. The traffic impacts analysis
portion of CA4PRS (Labeled Work-Zone Analysis in the software) can simulate 24 hours of traffic
through a defined work zone. Work zones are defined by the number of lanes closed, closure
duration and work zone capacity (Figure 10.14). Traffic can be entered by hourly count or ADT can
be entered and then distributed over 24 hours using hourly factors. CA4PRS can simulate one lane
closure scenario over a 24-hour period. Longer closures are estimated by multiplying the results of
one 24-hour analysis by the total number of closures. The 24-hour simulation limit using only one
traffic count makes it difficult to account for longer closures (e.g., over several weeks or months)
where traffic flow is likely to change over time (e.g., weekday vs. weekend or summer vs. fall).
Output is similar to that of QuickZone (Figures 10.15 and 10.16). Currently, the CA4PRS user
manual does not explain the delay estimation algorithm it uses. As of 2010, CA4PRS development is
ongoing and licenses for state DOTs are free. CA4PRS only models traffic in the work zone and does
not model any wider network.
Figure 10.14 CA4PRS Work-Zone Analysis input screen.
99
Figure 10.15 CA4PRS Work-Zone Analysis summary results screen showing available traffic impact
metrics.
Figure 10.16 CA4PRS Work-Zone Analysis hourly traffic results graph showing demand vs. capacity.
100
10.4 References
Alexiadis, V.; Jeannotte, K. and Chandra, A. (2004). Traffic Analysis Toolbox Volume I: Traffic
Analysis Tools Primer. FHWA-HRT-04-038. Federal Highway Administration (FHWA), U.S.
Department of Transportation, Washington, D.C.
Federal Highway Administration (FHWA). (2003). Full Road Closure for Work Zone Operations: A
Cross-Cutting Study. http://ops.fhwa.dot.gov/wz/resources/publications/FullClosure/CrossCutting/its.htm
Hardy, M. and Wunderlich, K. (2008). Traffic Analysis Toolbox Volume VIII: Work Zone Modeling and
Simulation A Guide for Decision Makers. FHWA-HOP-08-029. Federal Highway Administration
(FHWA), U.S. Department of Transportation, Washington, D.C.
Hardy, M. and Wunderlich, K. (2008). Traffic Analysis Toolbox Volume IX: Work Zone Modeling and
Simulation A Guide for Analysts. FHWA-HOP-09-001. Federal Highway Administration (FHWA),
U.S. Department of Transportation, Washington, D.C.
Lee, E.B. and Ibbs, C.W. (2005). Computer Simulation Model: Construction Analysis for Pavement
Rehabilitation Strategies. Journal of Construction Engineering and Management, 131(4), p. 449458.
Lee, E.B.; Harvey, J.T. and Thomas, D. (2005). Integrated Design/Construction/Operations Analysis
for Fast-Track Urban Freeway Reconstruction. Journal of Construction Engineering and
Management, 131(12), p. 1283-1291.
Lee, E.B.; Lee H.; and Ibbs, C.W. (2007). Productivity Aspects of Urban Freeway Rehabilitation with
Accelerated Construction Journal of Construction Engineering and Management, 133(10), p. 798806.
Lee, E.B.; Lee, H. and Harvey, J.T. (2006). Fast-Track Urban Freeway Rehabilitation with 55 h
Weekend Closures: I-710 Long Beach Case Study. Journal of Construction Engineering and
Management, 132(5), p. 465-472.
Lee, E.B.; Roesler, J.R.; Harvey, J.T. and Ibbs, C.W. (2001). Case Study of Urban Concrete Pavement
Reconstruction and Traffic Management for the I-10 ( Pomona, CA) Project. University of California,
Berkeley Institute of Transportation Studies, Pavement Research Center. Report for the Innovative
Pavement Research Foundation (IPRF), Falls Church, VA.
Muench, S.T.; Ozolin, B. Uhlmeyer, J. and Pierce, L.M.(2007). Rapid Concrete Panel Replacement in
Washington State: Lessons Learned. International Conference on Optimizing Paving Concrete
Mixtures and Accelerated Concrete Pavement Construction and Rehabilitation, 7-9 November 2007,
Atlanta, GA
101
Lee, E.B.; Roesler, J.R.; Harvey, J.T. and Ibbs, C.W. (2002). Case Study of Urban Concrete Pavement
Reconstruction on I-10. Journal of Construction Engineering and Management, 128(1), p. 49-56.
National Asphalt Pavement Association (NAPA). (1996). Balancing Production Rates in Hot Mix
Asphalt Operations , IS-120. NAPA, Landham, MD.
Walls, J. and Smith, M.R. (1998). Life-Cycle Cost Analysis in Pavement Design. Report No. FHWA-SA98-079. FHWA, Washington, D.C. http://isddc.dot.gov/OLPFiles/FHWA/013017.pdf
Lee, E.B.; Ibbs, C.W. and Thomas, D. (2005). Minimizing Total Cost for Urban Freeway
Reconstruction with Integrated Construction/Traffic Analysis. Journal of Infrastructure Systems,
11(4), pp. 250-257.
Maze, T.H.; Schrock, S.D. and Kamyab, A. (2000). Capacity of Freeway Work Zone Lane Closures.
Mid-Continent Transportation Symposium 2000 Proceedings.
http://www.ctre.iastate.edu/pubs/midcon/Maze.pdf.
Kim, T.; Lovell, D.J. and Paracha, J. (2001). A New Methodology to Estimate Capacity for Freeway
Work Zones. Presentation at the 2001 Transportation Research Board Annual Meeting, Washington
D.C. http://www.workzonesafety.org/files/documents/database_documents/00675.pdf.
Borchardt, D.W.; Geza, P.; Sun, D. and Ding, L. (2009). Capacity and Road User Cost Analysis of
Selected Freeway Work Zones in Texas. FHWA/TX-09/0-5619-1. Texas Department of
Transportation, Austin, TX. http://tti.tamu.edu/documents/0-5619-1.pdf.
Lee, E.B. (2008). CA4PRS User Manual (Version 2.1). University of California, Berkeley, Berkeley, CA.
Transportation Research Board (TRB). (2000). Highway Capacity Manual (HCM2000). TRB, National
Research Council, Washington, D.C.
Federal Highway Administration (FHWA). (2004). Intelligent Transportation Systems in Work Zones:
Keeping Traffic Moving During Reconstruction of the Big I, a Major Interstate-Interstate Interchange
in Albuquerque. FHWA, U.S. Department of Transportation, Washington, D.C.
102
Section 11
Life Cycle Assessment (Environmental Accounting)
11.1 Purpose
This section overviews a method for determining the inputs and outputs of a pavement system that
are relevant to the environment. This can include, but is not limited to: energy use, water use,
emissions, raw materials and human health impacts. This method, called life cycle assessment
(LCA) is essentially an environmental accounting protocol. LCA results can be used as part of the
decision-making process when determining the appropriate pavement
rehabilitation/reconstruction strategy. For instance, if an owner-agency must comply with a
greenhouse gas (GHG) reduction mandate, options resulting in less GHG may be considered more
favorably. Often, but not always, environmental accounting results tend to agree with life-cycle
assessment results in pavement construction scenarios.
In the future, it is likely that energy and emissions associated with roadway construction, or any
industry, will be scrutinized more carefully. Green house gas (GHG) emissions are likely to be
subject to a cap-and-trade scheme in the U.S. and are increasingly being addressed through the
National Environmental Policy Act (NEPA) as recent White House Council on Environmental Quality
(CEQ) guidance shows (Sutley 2010). As this scrutiny increases, there will likely be more tools to
help in analysis. It also seems plausible that once industry has a fair idea what energy, emissions
and other resources are associated with roadway construction, it will begin to adopt (either
voluntarily or by regulation) efficiency standards associated with these items similar to what has
happened with the automobile industry (i.e., fuel efficiency standards), power generation (i.e.,
clean energy portfolio requirements) and even toilets (i.e., maximum allowable flow).

(i) Introduction
A LCA attempts to identify inputs and outputs of a system that are relevant to the environment
from its inception to its ultimate disposition. This means that a LCA includes everything from
gathering raw materials to the point at which those materials are returned to the environment.
This collection of all processes from cradle to grave allows LCA to provide a cumulative total of
inputs and outputs (e.g., energy, emissions, water, use, etc.) for a final product and the
environmental impacts associated with those inputs and outputs. The resulting environmental
impacts of these cumulative inputs and outputs is assessed and results can be used to compare
alternatives and improve the system. The International Standards Organization (ISO) outlines a
systematic four phased approach:
1. Goal and scope. Define the reasons for carrying out the LCA, the intended audience,
geographic and temporal considerations, system functions and boundaries, impact
assessment and interpretation methods.
103
2. Inventory assessment. Quantify life cycle energy use, emissions, and land and water use for
technology use in each life cycle stage.
3. Impact assessment. Estimate the impacts of inventory results.
4. Interpretation. Investigate the contribution of each life cycle stage, technology use
throughout the life cycle and include data quality, sensitivity and uncertainty analyses.
LCA in general, and for pavements in particular, is still in a relatively early stage of development
and thus common practices are still developing and available data can be sparse. This presents
problems when using LCA as a decision support tool; especially when comparing alternatives.
Results using different data sets, methods and practices can be an order of magnitude different for
the same analyzed pavement section. Common issues with LCA include:
Data sources. Often LCA data come from a select few databases such as the U.S. Life-Cycle
Inventory Database (from NREL), ecoinvent, ELCD database, etc. These are generally reviewed
for accuracy/errors and can help standardize information for use in LCAs. However, data usually
come from many different sources beyond these ranging from personal observation to national
databases, which can lead to problems when comparing one LCA with another. For instance,
the CO2 associated with hot mix asphalt (HMA) production is not a universal constant, but
rather it varies depending upon plant type, components and manufacturer, aggregate moisture
content, fuel type, amount of reclaimed asphalt (RAP) included, asphalt binder grade, crude oil
source, regional electricity mix, etc. While databases of national averages can lead to some
consistency in results between LCAs, they often do not provide the detail necessary to
distinguish between process changes (e.g., using warm mix asphalt or not), especially at the
local project level. At the very least, a LCA should clearly identify its data sources.
Missing data. There are many industrial processes where some, if not all, relevant data are not
known, recorded, or made available for public use. For instance, the amount of fugitive dust on
site associated with pavement construction is not generally known. Or, the exact chemical
make-up of an asphalt modifier may be a trade secret that the manufacturer is not willing to
divulge.
Outdated data. Sometimes, data exist but are outdated. Over time, processes change,
equipment improves, raw material sources change, etc.
Data specificity. While general average data may be more readily available or lead to more
consistency between LCAs, it often does not contain the detail needed to distinguish between
two alternatives being considered. For instance, the EPAs AP-42 document contains average
emissions data for asphalt plants; however, it assumes only an average amount of RAP being
used at the plant. Therefore, if this data is used it cannot distinguish between a mix using all
virgin materials and one using 25 percent RAP, for instance.
Setting boundaries. A LCA that attempts to account for all processes associated with a system
can quickly become intractable. For example, one could account for the slipform paver and its
energy use and emissions associated with a concrete pavement. One could also account for the
energy and emissions associated with manufacturing that slipform paver. However that leads to
potentially considering the energy and emissions associated with the manufacture of the
machines that made the paver and so on. For another example, one can choose whether to
include the effect of pavement stiffness on the rolling resistance it offers to vehicles that travel
104
on it. Reduced rolling resistance over the life of a pavement may lead to substantial energy
savings when summed over the millions of vehicles that may use the road. However, one might
question whether including this effect alone is a realistic assessment of the pavements impact
on vehicle operations. Because of these boundary issues (what processes are included and
excluded), every LCA has a defined boundary (that should be explicitly stated) that details which
processes are included and which are not. Inclusions and exclusions are often not consistent
between LCAs and can be controversial.
Procedural practices. Most LCAs generally follow ISO 14040 and ISO 14044. However, these
standards are still quite generally written and leave much room for interpretation. No set of
more precise LCA procedures exist for pavements. For instance, ISO 14044 says that feedstock
energy (energy associated with burning the material in a product when that material is not used
as an energy source but could be) should be included in the analysis. In essence, on must assign
the energy that is involved in burning the asphalt cement in a HMA pavement to the energy use
of that pavement even though that asphalt cement is almost certainly never going to be
burned. This is significant since asphalt has, as a significant amount of energy stored in it
(Santero, 2009).
Despite these limitations, LCA can still provide meaningful results and aid the project decision
process.
(ii) LCA Methods
There are two main methods typically used for LCA: the process-based approach and the Economic
Input-Output (EIO)-based approach. Both methods are acceptable for performing LCAs although
each has its strengths. Each method is briefly discussed here.
Process-based LCA. A selected system is chosen and defined so that it meets a set of desired
requirements (e.g., a pavement structure to meet traffic, environmental and structural
requirements). This system is then broken down into separate processes (e.g., aggregate
production, cement production, concrete transport, etc.) whose energy requirements and
emissions can be quantified. Further contributory processes can be defined and analyzed (e.g.,
manufacture of the aggregate crushers used in aggregate production) but at some reasonable point
a boundary must be established beyond which no downstream contributory processes are
considered. The location of this boundary is an important part of a LCA because it may significantly
affect the results. Ultimately, boundary locations are somewhat subjective, which can lead to
difficulty in comparing one LCAs results to another. Process-based LCAs are desirable because they
can be done in enough detail so that they include processes that can differentiate between two
options (e.g., using warm mix asphalt or not). They are problematic because of the subjective
boundary and difficulty in obtaining data on specific processes.
Economic Input-Output LCA (EIO-LCA). EIO-LCA overcomes the subjective boundary issue and data
availability issue by basing process and their relationships on national economic input-output
model. An EIO model divides the economy of a country into industry-level sectors which represent
individual activity in the selected economy, and depicts the economic interaction of industries
105
(sectors) in a nation (or a region) by showing how output of each sector is used as input for other.
The system boundary is inherently the whole countrys entire economy. Interactions are
represented by monetary value in a matrix form, called Economic input-output table (I-O table).
The data stored in the table are collected by public agencies (ex: Department of Commerce) during
a certain time period (usually 5 years). This conveniently avoids collecting individual process data
and sets a consistent boundary (the nations economy). EIO-LCA can be problematic because it uses
aggregate data, which can be inconsistently aggregated or does not contain enough detail to
differentiate between two options (e.g., using warm mix asphalt or not).
(iii) Typical Values

There have been a number of documented pavement LCAs in the past decade or so that can
provide valuable information on typical values. Muench (2010) reviewed 12 pavement LCA
papers/reports that documented 66 assessments of actual or hypothetical roadways and found:
System scope. Most LCAs tend to address the pavement structure only and not include other
road features (e.g., striping, guardrails, etc.). Analysis periods are usually 40-50 years.
Relation of roadway construction to traffic use. A good rule-of-thumb is: the energy expended
in initial construction of a new roadway is roughly equivalent to the energy used by traffic on
the facility over 1-2 years.
Relation of roadway construction to operations. Operations are defined as those equipment,
actions and operations that happen on a routine basis necessary to ensure proper and safe
roadway use. They include items such as lighting, traffic signals, de-icing, sanding, drawbridge
actions, toll booths, etc. Construction energy ranges from about 25-100 percent of operations
energy.
Total energy use. It can be loosely stated that energy expenditures per lane-mile of pavement
are typically on the order of 3-7 TJ depending upon the pavement section, maintenance
activities and LCA scope.
CO2 emissions. It can be loosely stated that CO2 emissions per lane-mile of pavement are
typically on the order of 200-600 tonnes depending upon the pavement section, maintenance
activities and LCA scope.
Contribution of roadway construction components. The following general statements are
reasonable:
Materials production accounts for about 60-80 percent of energy use and 60-90 percent of
CO2 emissions.
Construction accounts for less than 5 percent of energy use and CO2 emissions.
Transportation associated with construction accounts for about 10-30 percent of energy
use and about 10 percent of CO2 emissions.
Maintenance activities account for a broad range of about 5-50 percent of energy and CO2
emissions.
106
11.3 Analysis Tools

At present, there are few tools available to help the non-specialist conduct a meaningful pavement
LCA, however several efforts are underway to develop such tools. This section briefly overviews the
few existing tools.
EIO-LCA. An online tool from Carnegie Mellon Universitys Green Design Institute (www.eiolca.net)
that uses the EIO method to report U.S. economic sector averages of economic activity,
greenhouse gases, energy, toxic releases and water use for different processes (Figure 11.1).
Answers for specific sectors can be obtained quickly, however there is not enough detail to
distinguish between processes within a sector (e.g., using warm mix asphalt or not).
Figure 11.1. Output screen of the EIO-LCA online tool showing greenhouse gases associated with $1
million of economic activity in sector #230230 (Highway, street, bridge, and tunnel construction)
using the 1997 Industry Benchmark Model for producer prices.
PaLATE. A Microsoft Excel-based tool from the University of California, Berkeleys Consortium on
Green Design and Manufacturing that allows the user to input pavement construction and
materials parameters and calculates life-cycle energy use and a number of life-cycle emissions
parameters. It is primarily built on the EIO-LCA method, but uses the process approach for a few
items. PaLATE contains numerous errors in process data, computation, and physical input
107
parameters. These errors are significant enough to cause results to be incorrect by orders of
magnitude in some cases, thus rendering PaLATE essentially useless.
CHANGER. A computer software program from the International Road Federation (IRF) that
calculates the life-cycle CO2 emissions associated with pavement construction. It uses a processbased method and has been analyzed and validated by the Traffic Facilities Laboratory (LAVOC) of
the Swiss Federal Institute of Technology (Ecole Polytechnique Fdrale de Lausanne - EPFL). At
present it only reports CO2 emissions but can do so for pavement, earthwork, and clearing and
grubbing. The IRF plans to expand this tool to address the entire roadway (i.e., beyond just the
pavement to include signs, striping, guardrail, etc.).
Project Emissions Estimator (PE2). An online application developed at Michigan Technological
University (Mukherjee and Cass, 2012). PE-2 is a web-based
(http://www.construction.mtu.edu/cass_reports/webpage/plca_estimator.php) tool that can be
used to estimate and benchmark the CO2 footprint of highway construction projects. (Mukherjee
and Cass, 2012). The underlying inventory for PE-2 comes from 14 highway projects in Michigan,
and the estimator has features that are tailored to use with Michigan DOT project information.
Athena Impact Estimator for Highways. An LCA computer program roadway construction, use and
rehabilitation (Athena Institute, 2013). It is available for free on the Athena website
((http://www.athenasmi.org/our-software-data/impact-estimator-for-highways). Athena appears
to be based on Canadian data and is designed for use in Canada.
Roadprint. An online application developed with funding from Federal Lands Highway (FLH), the
Oregon Department of Transportation, and the Washington State Department of Transportation
(Lin, 2012). Currently available at http://clients.paviasystems.com/wfl, this online tool performs a
process-based LCA on pavement materials production, materials transport, and on-site
construction. It does not address any environmental inputs or outputs associated with vehicles
driving on the pavement surface. Roadprints development is comprehensively documented (Lin
2012) and there are plans to expand it to address the entire roadway (i.e., beyond just the
pavement to include signs, striping, guardrail, etc.). Roadprint is available for use for free and is
designed to produce LCA results in about 15 minutes for users who have basic knowledge of road
construction and no knowledge of LCA processes.
(i) Example LCA and findings using Roadprint
This section describes a case study where actual construction project data was used as input data
for a LCA conducted with Roadprint. Several potential changes to the construction materials are
investigated to determine their effects on energy consumption and greenhouse gas emissions over
a 50-year analysis period.
Project description and data. A local collector road in Kailua, HI is scheduled for repaving. The work
involves removing 4 inches of HMA with a milling machine and an additional 2 inches of base
material then replacing it with two layers of HMA: a 4-inch base course and a 2-inch surface course.
108
This job is equivalent to 11.87 lanes-miles (assumed lane width is 12 feet) of paving. Initial
construction quantities are as follows:
Surface course: 9,516 tons of HMA
o 5.5% asphalt by total weight of mix
o No recycled material in the mix
Base course: 18,790 tons of HMA
o 5% asphalt content by total weight of mix
o 10% glass cullet by total weight of mix
Milling: 79,386 yd2 of 6-inch deep milling
Assumed processes. A 2-inch mill-and-fill for surface renewal is assumed every 10 years (year 10,
20, 30, and 40). Quantities for each mill-and-fill quantities are assumed as follows:
Surface course: 7,913 tons of HMA
o 5.5% asphalt by total weight of mix
o No recycled material in the mix
Milling: 79,386 yd2 of 2-inch deep milling
Materials locations. Materials for both the initial construction and mill-and-fills come from the
following locations:
Aggregate, HMA and RAP: a local quarry 6 miles from the job site.
Asphalt: a local asphalt terminal 30 miles from the job site.
Glass cullet: assumed to come from a source 30 miles from the job site
Ground tire rubber: assumed to come from a source 30 miles from the job site
Results. Results from this analysis are (Figure 11.2):
62.34 TJ of life cycle energy consumption (5.25 TJ/lane-mile).
o Including 141.03 TJ of feedstock energy (at 40.20 MJ/kg of asphalt cement) this total
rises to 203.37 TJ of life cycle energy consumption (17.13 MJ/lane-mile).
3,482 Mg (tonnes) of life cycle CO2 equivalent emissions (293 tonnes/lane-mile).
109
Figure 11.2. Example Roadprint output graphs showing total energy consumption (left) and global
warming potential in terms of CO2e (right).
Table 11.1. Example Roadprint output table showing an overall contribution analysis by impact
assessment categories.
Energy
Consumption With
feedstock
Energy (GJ)
Unit
Material Production 200381
99%
Equipment 1415
1%
Transportation 1574
1%
203370
Total
Energy
Consumption w/o
feedstock
Energy (GJ)
59353
95%
1415
2%
1574
3%
62342
Global Warming
Potential
GWP(CO 2 Mg-E)
3247
93%
115
3%
120
3%
3482
Acidification
Kg SO2
14042
704
3805
76%
4%
21%
18550
110
Photochemical
Smog
Kg NOx
9484
85%
1063
10%
553
5%
11100
Eutrophication
Human Health Criteria Air
Kg PO4
milli - DALYs/Kg
964
123
67
84%
11%
6%
1153
287
19
52
80%
5%
15%
358
Table 11.2. Example Roadprint output table showing a contribution analysis on materials
production. The second energy (GJ) column assess contributions while ignoring feedstock energy.
Material Production
Energy (GJ)
Energy (GJ)
GWP(CO2 Mg-E)
Material
Value
%
Value
%
Value
%
HMA/WMA 41465.3 20.7%
41465.3
70% 1835.4
56.5%
PCC
0.0
0.0%
0.0
0% 0.0
0.0%
Virgin Aggregate 3818.4
1.9%
3818.4
6% 242.6
7.5%
Sand and Gravel
0.0
0.0%
0.0
0% 0.0
0.0%
Bitumen 13951.0
7.0%
13951.0
24% 1089.6
33.6%
Feedstock 141027.4 70.4%
0.0
0% 0.0
0.0%
Cement
0.0
0.0%
0.0
0% 0.0
0.0%
RAP/RAC to plant
0.0
0.0%
0.0
0% 0.0
0.0%
Aggregate substitutes 118.7
0.1%
118.7
0% 79.6
2.5%
Steel
0.0
0.0%
0.0
0% 0.0
0.0%
Rap/RAC Collection
0
0%
0
0%
Table 11.3 Example Roadprint output table showing a contribution analysis on transportation.
Transportation
Energy (GJ)
GWP(CO2 Mg-E)
Material
Value
%
Value
%
HMA/WMA 906.3
57.6%
69.3
57.6%
PCC
0.0
0.0%
0.0
0.0%
Virgin Aggregate
0.0
0.0%
0.0
0.0%
Sand and Gravel
0.0
0.0%
0.0
0.0%
Bitumen 238.0
15.1%
18.2
15.1%
Cement
0.0
0.0%
0.0
0.0%
RAP/RAC to plant
0.0
0.0%
0.0
0.0%
Aggregate substitutes 225.7
14.3%
17.2
14.3%
Steel
0.0
0.0%
0.0
0.0%
Rap/RAC Collection 204.5
13.0%
204.5
13.0%
Alternative investigation. Several alternatives to the previous baseline scenario were investigated
to determine potential changes if these alternatives were pursued. These options were:
All virgin material: Remove the glass cullet from the base course.
Glassphalt base. Include 10% glass cullet by weight of mix in place of base layer aggregate.
This is the baseline scenario calculated previously.
15% surface / 15% base RAP: Include 15% RAP in the surface and base courses.
20% surface / 40% base RAP. Include 20% RAP in the surface course and 40% RAP in the
base course. This is the maximum RAP percentage allowed by Hawaii DOT specifications.
WMA. Use warm mix asphalt instead of hot mix asphalt. This assumes a 15% reduction in
energy and CO2 emissions from the HMA manufacturing process only.
111
SMA. Use a stone matrix asphalt surface course at 6.5% asphalt by total weight of mix that
allows a surface life of 17 years. This results in resurfacing at years 17 and 34 only.
Quiet pavement. Use an open-graded friction course to reduce tire-pavement noise. This
mix is at 10% asphalt by total weight of mix (20% of the asphalt binder is ground tire rubber)
and the surface life is assumed to be 8 years. This results in five total resurfacings during the
50-year analysis period
Ultimate. Use a combination of a SMA surface course, no glass in the base course, 40% RAP
in the base course and warm mix asphalt for both courses in order to reduce the energy and
greenhouse gas (GHG) footprint to the maximum extent possible given current standards.
Figure 11.5 through 11.6 shows the percentage change from the baseline practice in terms of
energy consumption.
Figure 11.3. Life cycle energy consumption for the current practice and eight alternate scenarios for
the example LCA.
112
Figure 11.4. Life cycle CO2 equivalent emissions for the current practice and eight alternate
scenarios for the example LCA.
Figure 11.5. The percentage change from the baseline value of energy consumption for a number
of alternate scenarios for the example LCA.
113
Figure 11.6. The percentage change from the baseline value of CO2e emissions for a number of
alternate scenarios for the example LCA.
General conclusions. Some general conclusions that can be reached for this example are:
Extending service life can be the biggest single influence in energy used and CO2 emitted by
the pavement. The biggest single improvement came with the use of SMA and the assumed
increase in surface life from 10 to 17 years.
Often, a combination of options can produce an even greater savings in energy used and
CO2 emitted by the pavement. Several of the options can be combined on one project to
provide even bigger savings.
The inclusion or exclusion of the glass cullet makes very little difference in energy used and
CO2 emitted by the pavement. While the inclusion of glass cullet was, at one time,
mandated in Hawaii (where practical), it inclusion has little impact on energy and
emissions.
The use of quiet pavement (and its assumed shorter surface life of 8 years) results in the
consumption of more materials and a corresponding increase in energy and emissions.
Therefore, in this scenario the use of quiet pavement represents a sacrifice in energy and
emissions in order to achieve less tire-pavement noise.
11.4 References
Athena Institute. (2013). Impact Estimator for Highways: User Guide. Athena Institute, Ottawa, ON.
Lin, Y.Y. (2012). Eco-decision Making for Pavement Construction Projects. Ph.D. dissertation.
University of Washington, Seattle, WA.
114
Muench, S.T. (2010). Roadway Construction Sustainability Impacts: A Life Cycle Assessment Review.
Transportation Research Record 2151, TRB, National Research Council, Washington, D.C., pp.
36-45.
Mukherjee, A. and Cass, D. (2012). Project Emissions Estimator: Implementation of a Project-Based
Framework for Monitoring the Greenhouse Gas Emissions of Pavement. In Transportation
Research Record, No. 2282. Transportation Research Board of the National Academies,
Washington, D.C. pp. 91-99.
Santero, N. (2009). Pavements and the Environment: A Life-Cycle Assessment Approach. Ph.D.
dissertation, University of California, Berkeley. Berkeley, CA.
Sutley, N.H., Chair, Council on Environmental Quality. (18 February 2010). Memorandum for Heads
of Federal Departments and Agencies.
Some pavement LCAs
Athena Institute. (2006). A Life-Cycle Perspective on Concrete and Asphalt Roadways: Embodied
Primary Energy and Global Warming Potential. Report to the Cement Association of Canada.
Carpenter, A.C., Gardner, K.H., Fopiano, J., Benson, C.H., Edil, T.B.. (2007). Life Cycle based risk
assessment of recycled materials in roadway construction. Waste Management 27, pp. 14581464.
Chui, C-T., Hsu, T-H., Yang, W-F. (2008). Life cycle assessment on using recycled materials for
rehabilitating asphalt pavements. Resources, Conservation and Recycling 52, pp. 545- 556.
Horvath, A. (2003). Life-Cycle Environmental and Economic Assessment of Using Recycled Materials
for Asphalt Pavements. University of California Transportation Center, Berkeley, CA.
Huang, Y., Bird, R., Bell, M. (2009). A comparative study of the emissions by road maintenance
works and the disrupted traffic using life cycle assessment and micro-simulation.
Transportation Research Part D 14, pp. 197-204.
Huang, Y., Bird, R., Heidrich, O. (2009). Development of a life cycle assessment tool for construction
and maintenance of asphalt pavements. J. of Cleaner Production 17, 2009, pp. 283-296.
Mroueh, U-M, Eskola, P., Laine-Ylijoki, J. (2001). Life-cycle impacts of the use of industrial by
products in road and earth construction. Waste Management 21, 2001, pp. 271-277.
Rajendran, S., Gambatese, J.A. (2005). Solid Waste Generation in Asphalt and Reinforced Concrete
Roadway Life Cycles. J. of Infrastructure Systems 13(2), pp. 88-96.
Stripple, H. (2000). Life Cycle Inventory of Asphalt Pavements. IVL Swedish Environmental research
Institute Ltd report for the European Asphalt Pavement Association (EAPA) and Eurobitume,
2000.
115
Stripple, H. (2001). Life Cycle Assessment of Road: A Pilot Study for Inventory Analysis, Second
Revised Edition. IVL Swedish Environmental Research Institute Ltd report for the Swedish
National Road Administration.
Tramore House Regional Design Office. (2006). Integration of the Measurement of Energy Usage
into Road Design. Final Report to the Commission of the European Directorate-General for
Energy and Transport. Project Number 4.1031/Z/02-091/2002, 2006.
Treloar, G.J.; Love, P.E.D. and Crawford, R.H. (2004). Hybrid Life-Cycle Inventory for Road
Construction and Use, J. of Const. Engr. and Mgmt. 130(1), pp. 43-49.
Weiland, C.D. Life Cycle Assessment of Portland Cement Concrete Interstate Highway Rehabilitation
and Replacement. Masters Thesis, University of Washington, Seattle, WA, 2008.
Zapata, P., Gambatese, J.A., (2005). Energy Consumption of Asphalt and Reinforced Concrete
Pavement Materials and Construction. J. of Infrastructure Systems 11(1), pp. 9-20.
116
Section 12
Miscellaneous Material Properties
12.1 Purpose
This section provides summaries of material properties that are relevant in designing pavement
renewal options.
12.2 Material Properties

Table 12.1 shows typical layer moduli for several material conditions. Table 12.2 shows information
about rubblized PCC and Table 12.3 about crack and seat renewal.
Table 12.1 HMA Pavement Typical Moduli and Ranges of Moduli
Material
Modulus Range (psi)
HMA (temperature dependent)
50,000 to 4,000,000
Cracked HMA Range
50,000 to 500,000
Cracked HMA (10% of wheelpathslight
100,000 to 250,000
to moderate fatigue cracks)
Pulverized HMA
40,000
Table 12.2 PCC Pavement Rubblization Typical Moduli and Ranges of Moduli
Material
Value or Property
Ratio of rubblized PCC elastic modulus/original
0.05
PCC slab elastic modulus
Slab modulus range prior to rubblization
Range: 3,000,000 to 7,000,000 psi
Typical slab modulus
4,000,000
Rubblized PCC Modulus
Range: 40,000 to 700,000 psi
Typical Rubblized PCC Modulus
50,000 to 150,000 psi
Table 12.3 Crack and Seat and Break and Seat Renewal
Material
Value or Property
Typical modulus of crack and seated PCCP
200,000 psi
Modulus of crack and seated PCCP
Range: 200,000 to 800,000
Modulus of break and seat PCCP
Range: 250,000 to 2,000,000 psi
117
12.3 References
Buncher, M., Fitts, G., Scullion, T., and McQueen, R. (2008), Development of Guidelines for
RubblizationFinal Report, Airfield Asphalt Pavement Technology Program Project 04-01, Asphalt
Institute, May 2008.
Sebesta, S. and Scullion, T. (2007), Field Evaluations and Guidelines for Rubblization in Texas,
Report No. FHWA/TX-08/0-4687-2, Texas Transportation Institute, December 2007.
118

Project Assessment Manual

Uploaded by

Copyright:

Available Formats

Project Assessment Manual

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Project Assessment Manual

Uploaded by

Copyright:

Available Formats

SHRP2 R23

Using Existing Pavement in Place and Achieving Long Life

Project Assessment Manual

June 29, 2013

Project Assessment Manual Topics

Project Assessment Manual Topics (continued)

1.2 How to Use the Manual

1.3 Assessment Data Categories

1.4 Overall Assessment Scheme

Identify Pavement Type for

Identify Distress Categories

Collect other pavement related

Select Renewal Pavement Type

Apply Recommended Renewal

Figure 1.1 Outline of Assessment Scheme

2.2 Measurement Methods

Table 2.1 Distress Groups (after McCullough, 1971)

Examples of Distress Mechanism

Adhesion (i.e. loss of bond)

Adhesion (i.e. loss of bond)

Table 2.2 Template for Flexible Pavement DistressFatigue Cracking

(Source: Pavement Interactive)

(Source: Pavement Interactive)

Figure 2.1 Illustrations of Fatigue Cracking Severity Levels

Table 2.3 Template for Flexible Pavement DistressTransverse Cracking

Moderate to High Severity High Severity

(Source: Pavement Interactive)

(Source: Pavement Interactive)

Figure 2.2 Illustrations of Transverse Cracking Severity Levels

Table 2.4 Template for Flexible Pavement DistressStripping/Raveling

Photo source: WSDOT

Figure 2.3 Illustration of Raveling

Table 2.5 Template for Rigid Pavement DistressJPCP or JRCPPavement Cracking

2. Percentage of slabs with two or more pavement cracks.

Examples of PCC Slab Multiple Cracks

Table 2.6 Template for Rigid Pavement DistressJPCP or JRCPFaulting

Average Fault 0.25 to 0.5 in.

Average Fault 0.5 in.

(Source: Pavement Interactive)

(Source: Pavement Interactive)

Figure 2.5 Illustrations of Various Levels of Joint Faulting

Table 2.7 Template for Rigid Pavement DistressD-Cracking

(Source: PI and C.L. Monismith)

Figure 2.6 Illustrations of D-Cracking Severity Levels

Table 2.8 Template for Rigid Pavement DistressASR Cracking

Early State of Cracking

Advanced Stage of Cracking

Figure 2.7 Illustrations of ASR Cracking Severity Levels

Table 2.9 Template for Rigid Pavement DistressCRCPPunchouts

Advanced Stage for a Punchout

Figure 2.8 Illustration of a CRCP Punchout

Table 2.10 Composite Pavement Distress1

2.3 Analysis Tools

3.2 Measurement Methods

3.3 Analysis Tools

Table 3.1 Typical Maximum Rut Depths

Maximum Rut Depth, inches (mm)

Table 3.2 FHWA IRI Criteria (from FHWA, 2006)

All Functional Classifications