Cost/Benefit Analysis of the

Effectiveness of Crack
Sealing Techniques
Manik Barman, Principal Investigator
Civil Engineering
University of Minnesota Duluth

June 2019

Research Report
Final Report 2019-26

• mndot.gov/research
To request this document in an alternative format, such as braille or large print, call 651-366-4718 or 1-
800-657-3774 (Greater Minnesota) or email your request to [email protected]. Please
request at least one week in advance.
 Technical Report Documentation Page
1. Report No. 2. 3. Recipients Accession No.
MN/RC 2019-26
4. Title and Subtitle 5. Report Date
Cost/Benefit Analysis of the Effectiveness of Crack June 2019
Sealing Techniques 6.

7. Author(s) 8. Performing Organization Report No.

Manik Barman, Jared Munch, and Uma Maheswar Arepalli
9. Performing Organization Name and Address 10. Project/Task/Work Unit No.
Civil Engineering CTS Project #2018011
University of Minnesota Duluth 11. Contract (C) or Grant (G)
No.
1405 University Drive, Duluth, MN 55812
1003325 WO 24
12. Sponsoring Organization Name and Address 13. Type of Report and Period Covered
Local Road Research Board Final Report
Minnesota Department of Transportation 14. Sponsoring Agency Code

Office of Research & Innovation

395 John Ireland Boulevard, MS 330 St. Paul,
Minnesota 55155-1899
15. Supplementary Notes
http:// mndot.gov/research/reports/2019/201926.pdf
http:// mndot.gov/research/reports/2019/201926A.pdf
16. Abstract (Limit: 250 words)
Crack sealing is an important preventive treatment in the pavement preservation program. To achieve
a cost-effective crack seal, it is important to select a proper crack sealing method. While Minnesota
usually seals cracks in asphalt pavements, there is no clear consensus on the most appropriate crack
sealing method for a specific job. This study focused on developing a guideline so that a cost-effective
crack sealing method could be chosen based on pavement type, functional condition, pavement age,
and traffic volume etc. This study includes a literature review, online survey, field performance data
collection and analysis, and development of a guideline. The effectiveness of the crack seals was
determined using a benefit-cost analysis. Two decision trees were developed for choosing the most
appropriate crack sealing method. The first, which can be used in MnDOT’s pavement management
system, needs information such as crack severity, pavement type (new vs overlay), pavement analysis
period and design life, traffic level, and crack seal occurrence number. The second decision tree, which
is a simplified version of the first and can be used by the preventive maintenance crews and requires
less information: crack severity, traffic level, and crack sealing occurrence number.

17. Document Analysis/Descriptors 18. Availability Statement

Asphalt pavements, Pavement cracking, Cracking of No restrictions. Document available from:
asphalt concrete pavement, Guidelines, Benefit National Technical Information Services,
cost analysis, Field tests, Pavement performance Alexandria, Virginia 22312

19. Security Class (this report) 20. Security Class (this page) 21. No. of Pages 22. Price
Unclassified Unclassified 139
COST/BENEFIT ANALYSIS OF THE EFFECTIVENESS OF CRACK
SEALING TECHNIQUES

FINAL REPORT

Prepared by:

Manik Barman
Jared Munch
Uma Maheswar Arepalli

Department of Civil Engineering

University of Minnesota Duluth

June 2019

Published by:

Minnesota Department of Transportation

Office of Research & Innovation
395 John Ireland Boulevard, MS 330
St. Paul, Minnesota 55155-1899

This report represents the results of research conducted by the authors and does not necessarily represent the views or policies
of the Minnesota Local Road Research Board, the Minnesota Department of Transportation or the University of Minnesota.
This report does not contain a standard or specified technique.

The authors, the Minnesota Department of Transportation, Local Road Research Board, and the University of Minnesota do not
endorse products or manufacturers. Trade or manufacturers’ names appear herein solely because they are considered essential
to this report because they are considered essential to this report.
ACKNOWLEDGMENTS

The Research team gratefully acknowledges the financial support provided by the Minnesota Local Road
Research Board (LRRB) for conducting this research. The members of the Technical Advisory Panel (TAP)
were very helpful, and provided valuable suggestions and feedback throughout the duration of this
project; the research team is grateful to them and acknowledge their kind contributions. The
cooperation and support provided the Technical Liaison, Mr. Dan Knapek and several other colleagues
from MnDOT's Materials and Road Research office, especially Mr. Gerard Geib, Mr. Ed Johnson, and Mr.
Paul Nolan are greatly appreciated. The research team also acknowledges the cooperation of Cindy
Voigt of the City of Duluth, Doug Welk of Astech Corp., Trent Nicholson Koochiching County for their
help providing access to several crack sealing projects at different parts of Minnesota. Also, a special
thank goes to the MnDOT Project Coordinators Bruce Holdhusen and Thomas Johnson-Kaiser, and CTS
Research Coordinator Elizabeth Andrews for their time and administrative support.

The research team sincerely acknowledges the contribution of several students and staff of the
University of Minnesota Duluth (UMD) who helped the team at different phases of this project.
TABLE OF CONTENTS
CHAPTER 1: Introduction ....................................................................................................................1

1.1 Objectives and Tasks........................................................................................................................... 3

CHAPTER 2: Crack Sealing Practices and Sealants ................................................................................5

2.1 Crack Sealing ....................................................................................................................................... 5

2.2 Crack Filling ......................................................................................................................................... 5

2.3 Difference between Crack Sealing and Crack Filling........................................................................... 5

2.4 Installation Season .............................................................................................................................. 6

2.5 Crack Sealing Equipment .................................................................................................................... 7

2.5.1 Crack Preparation Equipment ..................................................................................................... 7

2.5.2 Sealant Preparation Equipment .................................................................................................. 9

2.6 Sealant Installation ........................................................................................................................... 11

2.6.1 Sealant Installation in Clean-and-Seal Method ......................................................................... 11

2.6.2 Sealant Installation in Rout-and-Seal Method .......................................................................... 13

2.7 Placement Configurations ................................................................................................................ 17

2.8 Types of Sealants Used in Minnesota ............................................................................................... 21

2.8.1 MnDOT 3719 (Joint and Crack Sealer: Hot-Poured, Crumb Rubber Type) ............................... 22

2.8.2 MnDOT 3723 (Joint and Crack Sealer: Hot-Poured, Elastic Type) ............................................. 22

2.8.3 MnDOT 3725 (Joint and Crack Sealer: Hot-Poured, Extra Low Modulus, Elastic Type) ............ 22

CHAPTER 3: Previous Studies on Crack Seal Performance and Benefits .............................................. 23

3.1 Performance Evaluation of Crack Seals ............................................................................................ 23

3.2 Benefit of Crack Sealing .................................................................................................................... 29

CHAPTER 4: Crack Seal Performance Evaluation in Present Study ...................................................... 31

4.1 Phase 1: Practitioners' Opinions (Online Survey) ............................................................................. 31

 4.1.1 Data Analysis ............................................................................................................................. 34

4.2 Phase 2: Historical Data .................................................................................................................... 41

4.2.1 Interviews .................................................................................................................................. 41

4.2.2 MnDOT’s Pavement Construction Project Log .......................................................................... 42

4.3 Phase 3: Newly-installed Crack Seal Sites ......................................................................................... 52

4.3.1 Performance Analysis for Site E ................................................................................................ 55

4.3.2 Performance Analysis for Site O ................................................................................................ 60

4.3.3 Performance Summary of Newly-installed Crack Sealants ....................................................... 65

4.4 Phase 4: Old Crack Seal Sites ............................................................................................................ 68

4.4.1 Performance Analysis for Old Sites ........................................................................................... 71

4.4.2 Performance Summary of Old Crack Seal Sites ......................................................................... 75

CHAPTER 5: Effectiveness of Crack Sealing Methods.......................................................................... 77

5.1 Performance-effectiveness............................................................................................................... 77

5.1.1 Benefit ....................................................................................................................................... 77

5.2 Cost-Effectiveness............................................................................................................................. 93

5.2.1 Unit Cost of Crack Sealing ......................................................................................................... 93

5.2.2 Pavement Life-Cycle Cost (LCC) ................................................................................................. 95

5.3 The Benefit to Cost (B/C) Ratio ....................................................................................................... 104

CHAPTER 6: Draft Recommendation for the Selection of Appropriate Crack Sealing Method ............ 107

6.1 Variables ......................................................................................................................................... 107

6.1.1 Crack Severity and Crack Width .............................................................................................. 107

6.1.2 Pavement Type and Age .......................................................................................................... 107

6.1.3 Traffic Level ............................................................................................................................. 109

6.1.4 Soil /Subgrade Type................................................................................................................. 109

 6.1.5 Cost factors.............................................................................................................................. 109

6.1.6 Practitioners’ Preference ........................................................................................................ 110

6.2 Crack Sealing Selection Criteria ...................................................................................................... 110

6.2.1 Decision Trees ......................................................................................................................... 113

CHAPTER 7: Summary and Conclusions ........................................................................................... 116

References..................................................................................................................................... 119

Appendix A

Appendix B

Appendix C

Appendix D

Appendix E

Appendix F

Appendix G

Appendix H

Appendix I

Appendix J

Appendix K

Appendix L

Appendix M

Appendix N

Appendix O

Appendix P

Appendix Q

Appendix R
Appendix S

Appendix T

Appendix U
LIST OF FIGURES
Figure 1.1 Photographs Show the Benefits of Effective Crack Seal .............................................................. 2

Figure 1.2 Photograph of a Transverse Crack ............................................................................................... 2

Figure 1.3 Photograph of a Longitudinal Crack ............................................................................................. 2

Figure 1.4. Crack Sealing Methods: (a) Clean-and-seal (b) Rout-and-seal, After Stoikes (2017) .................. 3

Figure 1.5. Part of MnDOT’s Pavement Preventive Maintenance Decision Tree (Crack Treatment)
(MnDOT 2018a)............................................................................................................................................. 4

Figure 2.1: Freeze/Thaw Related Damage Caused by Water Expanding and Contracting in a Crack (after,
Schulte, 2018) ............................................................................................................................................... 6

Figure 2.2: Incompressibles Intrusion Related Damage (after, Stoikes, 2017) ............................................ 6

Figure 2.3: Crack Sealed in Spring and Corresponding Profile Views during Summer and Winter (after,
Johnson, 2000) .............................................................................................................................................. 7

Figure 2.4: Airblaster Being used to Clean a Crack (after, Smith and Romine, 1999) .................................. 8

Figure 2.5: Airlance being used to Clean a Crack .......................................................................................... 8

Figure 2.6: Sandblaster Being Used to Clean a Crack (after, Smith and Romine, 1999)............................... 8

Figure 2.7: Wire Brush .................................................................................................................................. 8

Figure 2.8: Routing Machine (after, Smith and Romine, 1999) .................................................................... 9

Figure 2.9: Melter ....................................................................................................................................... 10

Figure 2.10: Distribution hose with a precision tip ..................................................................................... 10

Figure 2.11: Blotter Application .................................................................................................................. 10

Figure 2.12: Clean-and-seal (a) before and (b) after sealant installation ................................................... 12

Figure 2.13: Cross-Sectional View of Clean-and-Seal (after, Smith and Romine, 1999) ............................. 13

Figure 2.14: Routed Seal Expanding and Contracting Without Failure (after, Stoikes, 2017) .................... 14

Figure 2.15: Crack (a) Before and (b) After Being Routed .......................................................................... 15

Figure 2.16: Routed Crack After Cleaning ................................................................................................... 15

Figure 2.17: Routed Crack (a) After First Sealant Pass and (b) After Second Pass and Blotter Application16

Figure 2.18: MnDOT Recommended Placement Configuration for Rout-and-Seal (after, Smith and
Romine, 1999) ............................................................................................................................................. 17

Figure 2.19 (i) and (ii): Various Sealant Configurations (after, Smith and Romine, 1999).......................... 20

Figure 3.1: Trend of Seal Deterioration over Time (after, Smith and Romine, 1999)................................. 23

Figure 3.2: Various Sealant Failures (after, Al Quadi, 2017) ....................................................................... 26

Figure 3.3: Performance Evaluations of Crack Seals and Sealants (after Al-Quadi et al., 2017) ................ 27

Figure 3.4: Performance Indices of Clean-and-seal (C&S) and Rout-and-seal (R&S) Methods (after Al-
Quadi et al., 2017)....................................................................................................................................... 28

Figure 3.5: (a) Overband Failure and (b) PI Comparison of Overband vs no-overband seals (after, Al-
Quadi et al., 2017)....................................................................................................................................... 28

Figure 4.1: Crack Sealing Performance Data Sources ................................................................................. 31

Figure 4.2: Locations of the Online Survey Responders (Cities: Yellow, Counties: Red, MnDOT Districts:
Green) ......................................................................................................................................................... 33

Figure 4.3: Analysis of the Responses to Question 1, All data .................................................................... 34

Figure 4.4: Preferred Crack Sealing Method (a) City Offices (b) County Offices (c) MnDOT Offices .......... 35

Figure 4.5: Locations Reporting Preference about the Crack Sealing Method: Rout-and-seal (yellow) and
Clean-and-seal (green) ................................................................................................................................ 36

Figure 4.6: Analysis of the Responses to Question 2 .................................................................................. 37

Figure 4.7: Selection Criteria for Rout-and-seal.......................................................................................... 38

Figure 4.8: Selection Criteria for Clean-and-seal ........................................................................................ 38

Figure 4.9: Typical Service Life of Clean-and-seal Method ......................................................................... 39

Figure 4.10: Typical Service Life of Rout-and-seal Method ........................................................................ 39

Figure 4.11: Number of Responders agreed to take the Follow-up Survey ............................................... 40

Figure 4.12: Preference of the Sealing Method for the Responders of Question 5 ................................... 40

Figure 4.13: Average Service Period of Rout-and-seal and Clean-and-seal Projects .................................. 43
Figure 4.14: Histogram – Service Periods of Rout-and-Seal Projects ......................................................... 45

Figure 4.15: Histogram – Service Periods of Clean-and-Seal Projects ........................................................ 46

Figure 4.16: Service Period vs. Traffic (AADT) – Crack Sealing Methods .................................................... 46

Figure 4.17: Service Period vs. Traffic (AADT) - All, Rout-and-seal Projects ............................................... 48

Figure 4.18: Service Period vs. Traffic (AADT) – Below 10,000, Rout-and-seal Projects ............................ 48

Figure 4.19: Service Period vs. Traffic (AADT) - All, Clean-and-seal Projects.............................................. 48

Figure 4.20: Service Period vs. Traffic (AADT) – Above 10,000, Rout-and-seal Projects ............................ 48

Figure 4.21: Service Period vs. Traffic (AADT) – Below 10,000, Clean-and-seal Projects ........................... 49

Figure 4.22: Service Period vs. Traffic (AADT) – Above 10,000, Clean-and-seal Projects ........................... 49

Figure 4.23: Service Period vs. Pavement Surface Layer Thickness, Rout-and-seal Projects ..................... 50

Figure 4.24: Service Period vs. Pavement Surface Layer Thickness, Clean-and-seal Projects .................... 50

Figure 4.25: Service Period vs. Pavement Surface Layer Age, Rout-and-seal Projects ............................... 51

Figure 4.26: Service Period vs. Pavement Surface Layer Age, Clean-and-seal Projects ............................. 51

Figure 4.27: Locations of Newly-installed Crack Seal Field Sites ................................................................ 53

Figure 4.28: Examples of Seal Failures (i) Spalling, (ii) Partial Adhesion Loss, (iii) Full-depth Adhesion
Loss.) ........................................................................................................................................................... 54

Figure 4.29: (i) Cohesion Failure of Clean-and-Seal Crack (ii) Early Signs of Cohesion Failure on Rout-and-
Seal Crack .................................................................................................................................................... 54

Figure 4.30: Seal Performance at Site E after (i) First Winter, and (ii) Second Winter ............................... 57

Figure 4.31: Site E Performance Index of Site E after the Second Winter .................................................. 57

Figure 4.32: Photographic Performance Documentation of Site E. E-1-i Shows Crack E-1 upon being
sealed. E-1-ii and E-1-iii show Crack E-1 at the end of its first winter and during the second winter after
being sealed. This same sequence of photos follows for cracks E-2, E-3, E-4, E-5, E-6, E-7, E-8, E-9, and E-
10. E-10-iv shows the seal performance at the end of the first winter compared to the mid-winter
performance presented in E-10-iii. ............................................................................................................. 60

Figure 4.33: Site O Clean-and-Seal Performance, after First Winter (i) and Second Winter (ii) ................. 63

Figure 4.34: Site O Rout-and-Seal Performance After First Winter (i) and Second Winter (ii) ................... 63
Figure 4.35: Site O Clean-and-Seal Performance Index of Site E after the Second Winter ........................ 64

Figure 4.36: Site O Rout-and-Seal Performance Index of Site E after the Second Winter ......................... 64

Figure 4.37: Additional Seal Failure Types Observed in Old Crack Sealing Projects ................................... 70

Figure 4.38: Site J- E Palm St., Duluth; Crack Sealing Year - 2014............................................................... 71

Figure 4.39: Performance Evaluations of 2014-year Crack sealing Projects .............................................. 72

Figure 4.40: Site D- Sundby Rd., Duluth; Crack Sealing Year - 2015 ........................................................... 73

Figure 4.41: Performance Evaluations of 2015-year Crack sealing Projects .............................................. 73

Figure 4.42: Site Q- 43AE Glenwood to Superior St., Duluth; Crack Sealing Year – 2016 .......................... 74

Figure 4.43: Performance Evaluations of 2016-year Crack sealing Projects .............................................. 74

Figure 4.44: Yearly Average Performance Index (PI): Rout-and-seal Projects ............................................ 75

Figure 4.45: Segment 1-Hwy 22 to 1145 Michigan Street, Hutchinson...................................................... 76

Figure 4.46: Segment 2-1145 Michigan Street to 5th Ave SE, Hutchinson................................................. 76

Figure 4.47: Performance Index of the Rural and Urban Segments – Crack Sealed in the Same Year (2012)
.................................................................................................................................................................... 76

Figure 5.1: PI vs Age of Crack Seal and Benefit Area Calculation (end of first winter) ............................... 78

Figure 5.2: Average PI vs. Age of Crack Seal .............................................................................................. 79

Figure 5.3: Performance Evaluations and Average Benefit Area of Rout-and-Seal Projects ...................... 81

Figure 5.4: Probability Distribution of Rout-and-Seal Service life .............................................................. 84

Figure 5.5: Probability Distribution of Rout-and-Seal Service Life after Second Winter Analysis .............. 85

Figure 5.6: Clean-and-Seal Benefit Area Determination (end of the first winter) ...................................... 87

Figure 5.7: Service Life of Clean-and-Seal method for Different Sites after First Winter ........................... 88

Figure 5.8: Average Benefit Area for Clean-and-Seal method (end of the first winter) ............................. 88

Figure 5.9: Probability Density for Clean-and-Seal Service Life after First Winer Analysis......................... 89

Figure 5.10: Service Life of Clean-and-Seal method for Different Sites after the Second Winter.............. 90

Figure 5.11: Service Life/ Service Period of Crack Seals from Different Sources ........................................ 92
Figure 5.12: A Typical MnDOT Pavement Cross Section (after MnDOT, 2018b) ........................................ 95

Figure 5.13: Typical MnDOT Pavement Structure Thickness Selection Guideline (after MnDOT, 2018b) . 96

Figure 5.14: Benefit/Treatment Cost Ratio of Rout-and-Seal and Clean-and-Seal (Note- ‘benefit’ here is
not in terms of the dollar, so B/C ratio shall not be compared with unity (one)) .................................... 105

Figure 5.15: Benefit/Cost Ratio for Rout-and-Seal and Clean-and-Seal for Different Cases .................... 106

Figure 6.1. Decision tree for selection of crack sealing method ............................................................... 114

Figure 6.2. Simplified decision tree for selecting a crack sealing method ................................................ 115
LIST OF TABLES
Table 1: Possible Sealant Failures and Likely Causes (after Caltrans, 2008) ............................................... 24

Table 2: List of Solutions for Commonly Observed Sealant Failures (after Caltrans, 2008) ....................... 25

Table 3: Service Life Extension among Various Pavement Types (after, Rajagopal, 2011) ........................ 29

Table 4: List of MnDOT Districts, Counties, Cities, and Contractors that participated in the Online Survey.
.................................................................................................................................................................... 32

Table 5: ANOVA Results – Service Period of Rout-and-seal and Clean-and-seal Projects .......................... 44

Table 6: Descriptive Statistics – Service Periods of Rout-and-seal and Clean-and-seal ............................. 45

Table 7: ANOVA Results – Service Period of R&S and C&S with respect to Traffic .................................... 47

Table 8: Documented Cracks in Site E......................................................................................................... 55

Table 9: Performance of Site E after Second Winter .................................................................................. 56

Table 10: Documented Cracks in Site O ...................................................................................................... 61

Table 11: Performance of Site O after Second Winter ............................................................................... 62

Table 12: Summary of Performance Evaluations of Newly-installed Crack Sealing Project Sites .............. 66

Table 13: List of Old Crack Seal Project Sites Inspected during Field Visits ................................................ 69

Table 14: Forecasting of Crack Seal Performance upon First Winter Evaluations ...................................... 79

Table 15: Forecasting of Crack Seal Performance after Second Winter Evaluations ................................. 82

Table 16: Estimated PI vs Service Life for Rout-and-Seal after First Winter ............................................... 83

Table 17: Rout-and-Seal Service Life and Reliability after First Winter ...................................................... 84

Table 18: Rout-and-Seal Service Life and Reliability after Second Winter ................................................. 85

Table 19: Clean-and-Seal Service Life and Reliability after First Winter ..................................................... 89

Table 20: Clean-and-Seal Service Life and Reliability after Second Winter ................................................ 90

Table 21: Service Life of Crack Seals Observed in Different USA States and Canadian Provinces.............. 92

Table 22: Unit Cost of Crack Sealing Treatments Obtained from Previous Bid Letting Abstracts .............. 94

Table 23: Crack Sealing Unit Cost per Lane Mile ........................................................................................ 95
Table 24: Initial Construction Cost Per Lane-Mile, New HMA, Design Life = 20 Years ............................... 97

Table 25: Initial Construction Cost Per Lane-Mile, HMA Overlay, Design Life = 13 Years .......................... 97

Table 26: Initial Construction Cost Per Lane-Mile, HMA Overlay, Design Life = 20 Years .......................... 98

Table 27: Example LCCA per lane mile for a Typical New HMA with 20 years of design life and 35 years
Analysis Period ............................................................................................................................................ 99

Table 28: Example LCCA per lane mile for a Typical New HMA with 20 years of design life and 50 years
Analysis Period .......................................................................................................................................... 100

Table 29: Example LCCA per lane mile for a Typical HMA Overlay with 13 years design life and 35 years
Analysis Period .......................................................................................................................................... 102

Table 30: Example LCCA per lane mile for a Typical HMA Overlay with 20 years design life and 35 years
Analysis Period .......................................................................................................................................... 103

Table 31: Benefit/Treatment Cost Ratio of Rout-and-Seal and Clean-and-Seal Methods ....................... 105

Table 32: Benefit/ Life-Cycle-Cost Ratio of Rout-and-Seal and Clean-and-Seal Methods ........................ 105

Table 33. Crack Treatment Schedule (MnDOT 2018a) ............................................................................. 107

Table 34. Decision worksheet for crack sealing selection criteria ............................................................ 111
EXECUTIVE SUMMARY
Sealing and filling cracks in asphalt pavements are important preventive maintenance treatments
for achieving a good service life. Properly implemented crack sealing and crack filling can minimize the
intrusion of water into the underlying layers of pavements. This can reduce moisture-related damages.
Improperly implemented crack sealing may experience premature failure, can become ineffective in the
reducing intrusion of water into the pavement structure. This may result in severe distresses, like stripping
of asphalt and potholes, etc. In Minnesota, working cracks of asphalt pavements are repaired by sealing,
and non-working cracks are repaired by filling. Working cracks, which mostly run in the transverse
direction, are those that open in the winter and close in the summer due to thermal expansion and
contraction of the surrounding pavement. Non-working cracks are those that do not undergo notable
changes in width between seasons. Non-working cracks mostly develop in the longitudinal direction,
typically due to pavement fatigue failure within the wheel path, or at the lane joint because of the weak
or less-dense asphalt. The most commonly used crack sealing methods in Minnesota to repair asphalt
pavements are clean-and-seal and rout-and-seal. In the clean-and-seal method, cracks are treated by
blowing out the debris and then sealing the cracks with rubber sealant materials. In the rout-and-seal
method, a reservoir is routed, centering it above the existing cracks before pouring the sealants. Between
the above-mentioned two crack sealing methods, the rout-and-seal method is approximately two times
more expensive than the clean-and-seal method and relatively time-consuming. Even though some
studies indicate that rout-and-seal is superior to its counterpart, the actual cost-effectiveness of either of
these seal methods is not yet fully understood. While the application of crack filling for non-working cracks
is a widely accepted protocol, transportation agencies do not have a universally accepted guideline for
selecting the most cost-effective crack sealing method for a specific job. Cities, counties, and MnDOT
districts thereby do not follow a uniform procedure; some agencies prefer rout-and-seal, while others
believe in clean-and-seal. The current study is aligned to provide a solution to the problem mentioned
above so that a guideline is available for transportation agencies to refer to when selecting the most cost-
effective crack sealing method based on the factors that can influence the performance of the seals.

The main objectives of this study are (i) to compare the service life and cost-effectiveness of the
two crack sealing methods mentioned above and (ii) to develop a criterion to select the most appropriate
crack sealing method based on pavement type, functional condition, pavement age, and traffic
characteristics, etc. The abovementioned objectives were accomplished by performing the following
major tasks: (i) literature review on crack sealant practices, (ii) crack sealant performance data collection
and analysis, (iii) performance- and cost-effectiveness analysis, and (iv) development of a
recommendation. Various literature on the crack sealant practices and performances, including research
reports, synthesis, journal articles, and other relevant publications, were reviewed. The crack seal
performance data was collected through an online survey and reviewing the history of crack sealing
projects documented in several pavement construction data logs of the Minnesota Department of
Transportation (MnDOT). Direct field performance data was collected through periodical evaluations of
crack seals at 35 different sites located throughout Minnesota. The performance of the crack seals was
studied by quantifying the performance index of the crack seals. The effectiveness of the crack seals was
studied with respect to a benefit-cost analysis. Two decision trees were developed to guide transportation
agencies when selecting the most appropriate crack sealing method. The major specific conclusions drawn
from the different tasks of this study are listed below:

 Between the two crack seal methods, it was found that rout-and-seal (of transverse cracks) is
more commonly used in Minnesota.
 While there are no uniformly accepted criteria, currently the most commonly reported criterion
for selecting a sealing method is crack/pavement conditions followed by pre-determined
schedules. A good percent of practitioners does not follow any criteria at all.
 Regarding the failures of the crack seals, it was found that most of the crack seals in Minnesota
fail by adhesion and it occurs during the winter season. This failure is being seen more commonly
with wider crack spacing. Crack sealants either do not stretch enough or the adhesiveness
achieved between the sealant and crack face is not strong enough to offer resistance against the
tensile stress generated on the sealant in winter. Some cohesion failures were also observed but
in very limited quantities. A good amount of spalling failure was also observed at some rout-and-
seal sites.
 The crack seal performance data was collected from 35 inspection sites, which were then used to
develop a direct correlation (R2 = 0.95) between the average performance index and age of the
seals. The correlation was useful to determine the average service lives of the rout-and-seal and
clean-and-seal methods.
 It was found that the rout-and-seal and clean-and-seal methods have approximately 4- and 3-year
service lives, respectively. These service lives were determined using the data collected in this
project, as well as other relevant crack seal performance data found in the literature. The short-
term performance for closely spaced cracks on a rural road is, however, identical for both clean-
and-seal and rout-and-seal.
 The life-cycle cost analysis (LCCA) and benefit-cost ratio analysis showed that rout-and-seal is
slightly more effective than the clean-and-seal, due to its longer performance period. However, if
only a short-term benefit is considered, then clean-and-seal may be more cost-effective than its
counterpart.
 Because the difference in benefit-cost ratios between the two crack sealing methods was not
significant, several other decision factors were considered to establish the effectiveness of each
crack seal method. Various factors, such as treatment cost, expected life, ease of operation,
practitioners’ opinion, and traffic level, were considered in addition to the benefit-cost ratio.
 Two decision trees were developed for choosing the most appropriate crack sealing method. The
first one, which can be used for pavement management systems, needs information, including
crack severity, pavement type (new vs. overlay), pavement analysis period and design life, traffic
level, and crack seal occurrence number. The second decision tree, which is a simplified version
of the first and can be used by preventive maintenance crews, needs less information: crack
severity, traffic level, and crack sealing occurrence number (first time, second time or third time).
 In general, the clean-and-seal method was found to be appropriate for high crack severity
conditions. The choice between rout-and-seal and clean-and-seal for low and moderate crack
severity is found to be varied based on pavement age and type, traffic levels. The clean-and-seal
method is found to be appropriate for sandy soil subgrades and the low initial budget scenario;
whereas, the rout-and-seal method is preferred for clayey and silty subgrades, irrespective of
other variables.
CHAPTER 1: INTRODUCTION
The goal of a pavement preservation program is to extend the pavement service life and enhance
the state-wide performance in the most cost-effective way possible. Some of the preventive maintenance
treatments within the framework of the pavement preservation program are crack sealing and crack
filling, chip sealing, fog sealing, thin overlays, rut filling, etc. Johnson (2000) stated that the preventive
maintenance is six to ten times more cost-effective than a “do-nothing” maintenance strategy. The benefit
of preventive maintenance is a substantial saving of life-cycle cost as well as achieving an extended period
of acceptable driving conditions. Among the various preventive maintenance treatments, the crack sealing
and filling are the two most commonly used treatments performed on asphalt concrete pavements in
Minnesota.

Properly implemented crack sealing and crack filling can minimize the intrusion of water into
pavements’ underlying layers and reduce moisture-related damages. Improperly implemented crack
sealing may lead to the development of premature failure, seals become ineffective in reducing the
intrusion of water into the pavement structure, which may results in severe distresses, like stripping of
asphalt and pothole, etc. Figure 1.1 presents two photographs demonstrating the difference in
performance between an effective and ineffective crack seals. In Figure 1.1 (a), it can be seen that water
could intrude into the pavement structure through the failed or ineffective sealant; whereas, the intact
or effective seal in Figure 1.1 (b) was able to stop the intrusion of water.

In Minnesota, the working and non-working cracks are repaired with sealing and filling,
respectively. Working cracks, which mostly run in the transverse direction, are those that expand in the
winter and contract in the summer due to thermal expansion and contraction of the surrounding asphalt.
These cracks, popularly known as thermal cracks, do initiate in the winter months due to the contraction
of the asphalt in the surface layer. Non-working cracks are those that do not undergo notable changes in
width between seasons. Non-working cracks mostly develop in the longitudinal direction typically due to
pavement fatigue failure within the wheel path, or at the lane joint because of the weak or less-dense
asphalt. Figure 1.2 and Figure 1.3 show typical examples of transverse and longitudinal cracks.

In Minnesota, the most frequently used crack sealing methods to repair asphalt pavements are
clean-and-seal and rout-and-seal (Figure 1.4). In the clean-and-seal method, cracks are treated by blowing
out the debris and then sealing the cracks with rubber sealant materials. In the rout-and-seal method, a
reservoir is prepared by centering the existing cracks, and then sealants are poured. Among the two crack
sealing methods mentioned above, the rout-and-seal method is approximately two times more expensive
than the clean-and-seal method and relatively time-consuming. Even though some studies indicated the
rout-and-seal is superior to its counterpart, the actual cost-effectiveness of either of these methods are
not yet fully understood. While the application of crack filling on the non-working cracks is a widely
accepted protocol, transportation agencies do not have a universally accepted guideline for selecting the
most cost-effective crack sealing method for a specific job. Cities, counties, and Minnesota Department
of Transportation (MnDOT) districts thereby do not follow a uniform procedure; some agencies prefer the
rout-and-seal while the others believe in the clean-and-seal methods. Due to this, pavement engineers
often face the challenge of choosing the most appropriate crack sealing method for their job.

1
 (a) Failed Crack Seal (b) Intact Crack Seal

Figure 1.1 Photographs Show the Benefits of Effective Crack Seal

Figure 1.2 Photograph of a Transverse Crack Figure 1.3 Photograph of a Longitudinal Crack

2
The Decision Tree available in MnDOT’s Pavement Preventive Maintenance Guide provides a guidance on
selecting different preventive treatments, as shown in Figure 1.5. The decision tree has provision for the
crack sealing and crack filling treatments; however, it does not provide any guidance to choose between
the clean-and-seal and rout-and-seal. The current study is aligned to provide a solution to the above-
mentioned problem so that a guideline is available to the transportation agencies to refer to select the
most appropriate crack sealing method based on the factors that can influence the performance of the
seals.

(a) (b)

Figure 1.4. Crack Sealing Methods: (a) Clean-and-seal (b) Rout-and-seal, After Stoikes (2017)

1.1 OBJECTIVES AND TASKS

The main objectives of this project are (i) to compare the service life and cost-effectiveness of
different crack sealing methods and (ii) to develop criteria to select the most cost-effective crack sealing
method as a function of pavement type, functional condition, pavement age, and traffic characteristics,
etc.

The abovementioned objectives were accomplished by performing the following major tasks: (i)
reviewing the literature on the crack sealant installation practices and sealant performances; (ii) collecting
and analyzing crack sealant performance data; (iii) conducting performance and cost-effectiveness
analysis; and (iv) developing recommendations. The recommendations include decision trees for choosing
the effective crack sealing method. Two decision trees were developed: one for the pavement
management system, which can be added to the Minnesota Department of Transportation’s (MnDOT’s)
existing ‘Pavement Preventive Maintenance Decision Tree’ shown in Figure 1.5. The second one is a
simplified version of the former and can be used by the maintenance crew.

3
 Current Study

Figure 1.5. Part of MnDOT’s Pavement Preventive Maintenance Decision Tree (Crack Treatment) (MnDOT 2018a)

4
CHAPTER 2: CRACK SEALING PRACTICES AND SEALANTS
2.1 CRACK SEALING

Crack sealing in asphalt concrete pavement is a preventive maintenance practice and is used throughout
the country. The process of crack sealing involves placing a non-permeable rubber sealant into an
existing crack in the wearing course of asphalt concrete pavement. Various construction and installation
techniques are used by different states across the country to achieve the goals of crack sealing.

Applications of crack sealing vary throughout the country. In Minnesota, crack sealing is mostly used on
working cracks. Working cracks are those that open in the winter and close in the summer due to
thermal expansion and contraction of the surrounding pavement. Working cracks typically develop in
the transverse direction. Smith and Romine (1999) defined working cracks as those that meet a 3 mm
movement criteria.

2.2 CRACK FILLING

Crack filling techniques are often used on non-working cracks. Non-working cracks are those that do not
undergo notable changes in width between seasons. Non-working cracks mostly develop in the
longitudinal direction typically due to pavement fatigue cracking or lane joint separation. Transverse
cracks may, however, be referred to as non-working if their spacing relative to each other is close
enough that no significant changes in crack width will occur due to thermal expansion and contraction of
the surrounding pavement. The goal of crack filling is not necessarily to prevent water from entering a
crack but to support the surrounding pavement. The fill does, however, impede some water from
entering cracks.

2.3 DIFFERENCE BETWEEN CRACK SEALING AND CRACK FILLING

While crack sealing and crack filling are two different processes with different goals in mind, it is often
difficult to distinguish between working and non-working cracks in the field. For this reason, some
states, such as Colorado, do not distinguish between the two for all of their in-house sealing projects. In
the Colorado Department of Transportation (CDOT), Trunschke et al. (2014) noted that Colorado uses a
single operation for its sealing contracts that conforms to more of a “filling” activity than a “sealing”
activity. Minnesota does distinguish between the two and often uses different sealing and filling
procedures on various cracks throughout a project. Although there is a little distinction, and often
contradicting views on the differences between crack sealing and crack filling, Smith and Romine (1999)
describe them as follows in the FHWA Materials and Procedures for Sealing and Filling Cracks is Asphalt-
Surfaced Pavements Manual of Practice.

Crack Sealing- “The placement of specialized treatment materials above or into working cracks using
unique configurations to prevent the intrusion of water and incompressibles into the crack.”

Crack Filling- “The placement of ordinary treatment materials into non-working cracks to substantially
reduce infiltration of water and to reinforce the adjacent pavement.”

Due to working cracks constantly undergoing changes in width, sealing requires the use of higher quality
materials and more sophisticated construction equipment and processes; crack sealing, therefore, has a

5
greater cost involved. If performed effectively and in a timely manner, crack sealing can improve
pavement performance and extend its serviceable life. Limiting the volume of water entering a crack
minimizes the risk of freeze/thaw related damage caused by water expanding and contracting in a crack,
as demonstrated in Figure 2.1. Limiting incompressible materials from entering a crack allows free
movement of the road to expand during warm weather. Otherwise, incompressible augment the
compressive forces on the asphalt materials during the summer, as demonstrated in Figure 2.2. By
restricting water and incompressible debris from entering cracks, properly placed seals extend the
pavement life by keeping cracks from progressing in severity.

Figure 2.1: Freeze/Thaw Related Damage Caused by Water Expanding and Contracting in a Crack (after,
Schulte, 2018)

Figure 2.2: Incompressibles Intrusion Related Damage (after, Stoikes, 2017)

2.4 INSTALLATION SEASON

Crack sealing in Minnesota mostly takes place during the spring, summer, and fall months.
Installation season influences the seal performance due to different crack widths at different seasons.
Sealing a working crack while it is partially closed during the summer months allows a minimal amount of
material into the crack. A seal that is placed flush with the surface of the road during the summer stretches
in the winter as the pavement contracts, then returns to flush the following summer. If the winter creates
too large of a drop in air temperatures, cracks experience large opening and cause the adhesive or
cohesive failure of the seal. Likewise, if a seal is placed flush to the pavement surface during the fall or

6
spring, the seal would protrude from the crack. Figure 2.3 demonstrates the effects of crack expansion
(during winter) and contraction (during summer) on a seal placed during the spring and fall months. Due
to the reasons mentioned above, crack sealing is not recommended to be performed during the winter
months. The seal would likely fail quickly due to excessive material loss and seal degradation. The other
reason for not sealing during the winter is that the Minnesota climate presents a great challenge for
proper crack preparation (cleaning and drying) during the winter months.

Figure 2.3: Crack Sealed in Spring and Corresponding Profile Views during Summer and Winter (after, Johnson,
2000)

2.5 CRACK SEALING EQUIPMENT

Crack sealing operation demands a wide range of equipment for sealant preparation, crack
preparation, sealant installation, and sealant finishing. Some of the most common pieces of construction
equipment for crack sealing are summarized below.

2.5.1 Crack Preparation Equipment

Airblaster: Airblasting is done with a high-pressured air compressor placed on a truck with hoses and
wands. High-pressure blasting is fairly effective at removing dust and debris. Its downside is that it is not
effective in drying the crack channel. Air blasting may also be performed with leaf blowers, but most states
do not allow it due to a lack of air velocity and poor cleaning results associated with it. Figure 2.4 shows
an example of air blasting to clean a crack.

Air Lance: Hot air blasting uses air that is heated to a minimum of 1,370o Celsius as defined by the Smith
and Romine (1999). This form of cleaning is effective at removing dirt and debris. It also creates a dry and
hot crack surface for a sealant to bond to. A hot surface will likely create a better bond for the sealant by
activating the binder in the pavement itself. Caution must be used, however, as it is possible to burn the
asphalt concrete pavement with a lance. For this reason, an open flame torch should never be used for
this procedure. Figure 2.5 shows an example of an air lancing to clean a crack.

7
 Figure 2.4: Airblaster Being used to Clean a Crack (after, Figure 2.5: Airlance being used to Clean a Crack
Smith and Romine, 1999)

Sandblaster: Sandblasting is a highly effective way of removing debris and loosened fragments from the
channel of a crack. One pass of the sandblaster should be made on each side of a routed reservoir. The
procedure leaves a smooth and textured surface that is ideal for the sealant to bond to. Sandblasting
consists of a compressed air unit, sandblaster machine, hoses, and a wand. A second pass with an air
compressor is typically necessary to clean any debris that was left during the sand blast. Due to the
number of passes needed, sand blasting requires a great deal of equipment, is labor-intensive, and time-
consuming. Figure 2.6 shows an example of sand blaster being used to clean a crack.

Figure 2.6: Sandblaster Being Used to Clean a Crack Figure 2.7: Wire Brush
(after, Smith and Romine, 1999)

8
Wire Brush: In this cleaning method, power-driven wire brushes are used in conjunction with some form
of compressed air. This combination effectively removes debris from the crack but fails to remove loose
pieces of the asphalt. Wire brushes are available with and without blowers. Some contractors have had
success modifying pavement saws to fit wire brushes in place of the saw blade. Figure 2.7 shows an
example of a wire brush used for cleaning cracks.

Routing Machine: Crack routing is performed by a worker using a router or saw unit mounted onto a cart.
The operator uses his/her eyes and best judgment to follow the path of the crack with the routing or
sawing machine. Smith and Romine (1999) noted that although a saw with a 150-200 mm diameter
diamond blade can follow the meanders of cracks fairly well, the high cutting rate of an impact router
creates smoother reservoir walls with a higher percentage of aggregate area for the sealant to bond to.
Most companies require that employees operating routing equipment wear some type of respirator mask.
The most modern routing machines have air and dust control systems built into them. Figure 2.8 shows
an example of a router in operation.

Figure 2.8: Routing Machine (after, Smith and Romine, 1999)

2.5.2 Sealant Preparation Equipment

Melter: Hot pour sealant is heated in a double-walled heating tank that uses a heat transfer oil, such that
no flames come into direct contact with the tank holding the sealant. The Federal Highway Administration
(FHWA) recommends that the melter should allow an operator to regulate the sealant temperature up to
220oC (428oF). The ideal heated temperature of each sealant material is typically specified on the package
label. Upon being heated, the materials will transform from a solid state to a liquid state. Some melters
have a recirculation feature, which is important to prevent temperature stratification within the tank and
maintain a proper temperature for the sealant being laid into cracks. Upon reaching the desired
temperature, the material is stored in the tank until it is placed into the crack. Figure 2.9 shows an example
of a melter.

9
 a b

Figure 2.9: Melter

Distribution Hose: A melter truck has various distribution hoses connected to the back of it. The sealant
flows through these hoses in order to be applied into cracks. A distribution hose may have a precision tip
or may be equipped with a squeegee type nozzle that shapes the material in addition to applying it into
the crack. A metal distributor may also be used to level off the poured sealant if no squeegee is present.
Figure 2.10 shows an example of a distribution hose with a precision tip.

Figure 2.10: Distribution hose with a precision tip Figure 2.11: Blotter Application

Blotter Applier: Application of a paper blotter material (typically toilet paper), occasionally soapy water,
is used to prevent fresh sealant material from sticking to passing vehicle tires. Figure 2.10 shows an
example of a blotter application.

10
2.6 SEALANT INSTALLATION

According to Smith and Romine (1999), crack sealant installation consists of at least two and up to
five steps. These steps are:

1. Crack cutting (routing, sawing, etc.)

2. Crack cleaning and drying
3. Sealant preparation and application
4. Sealant finishing and shaping
5. Blotting

The two essential steps that every treatment process must consist of include (i) crack cleaning and drying,
and (ii) sealant preparation and application. The following subsections describe the various installation
steps usually adopted in different crack sealing methods: clean-and-seal and rout-and-seal.

2.6.1 Sealant Installation in Clean-and-Seal Method

It is the quickest and simplest form of crack sealing. It involves using a hot air lance or compressed
air to blow debris from a crack, then filling it with a sealant. It is noted that clean-and-seal may often be
confused with filling, but they are two different treatment procedures, mainly differ in the quality of
sealant material used for the treatment. A higher quality sealant material (Type II: 50% extension at -20oF)
with better bonding characteristics is preferred for clean-and-seal (applied on transverse cracks) to
withstand the thermal expansion and contractions during winter. A lower graded sealant material (Type
I: 50% extension at 0oF) would be sufficient for crack filling (applied on longitudinal cracks) due to the
nature of non-working cracks. It is noted that the crack preparation (blowing or air lance) is usually the
same for both treatment methods. It is recommended by Johnson (2000) that clean-and-seal be
performed while cracks are still narrow and during the spring and fall seasons when temperatures are
moderately cool. Figure 2.12 shows a photograph of a candidate transverse crack before and after being
sealed using the clean-and-seal method. It is not advised to implement this sealing method on cracks that
feature secondary cracking or edge deterioration at the crack face. Traffic should be re-routed during the
construction process, and material should be allowed to cure before being re-opened to traffic. If the
pavement must be re-opened immediately, a blotter material should be applied to prevent the sealant
from being picked up by vehicle tires.

According to Johnson (2000), a clean-and-seal method is expected to perform for three years before a
significant amount of materials begin to pull from the side of the crack. The seal will still perform at this
point, as it still prevents some water and solids from entering the crack. Often, a second clean-and-seal
may be applied on an existing clean-and-seal after the original seal begins to pull apart.

2.6.1.1 Equipment Needed

The clean-and-seal method of crack sealing requires the least amount of heavy equipment. Unlike
rout-and-seal and saw-and-seal, no cutting or sawing machinery is required in this method. This method
requires the following equipment:

11
Crack Preparation Equipment (at least one of the following): Airblaster, Air lance, Sandblaster, and Wire
brush, etc.

Material Preparation: Melter

Material Application and Blotting: Distribution hoses, Shaping tools (squeegee, metal distributor), Blotting
paper

a b

Figure 2.12: Clean-and-seal (a) before and (b) after sealant installation

2.6.1.2 Crack Preparation

Cleaning of the crack channel is an extremely important part of the crack sealing process, as a
poorly cleaned crack will have a poor bonding surface for the sealant that is being applied. A high
percentage of seal failures are adhesion failure that occurs due to the application of sealants on poorly
cleaned or moist crack channels. Cleaning can include air blasters, hot air blasters (air lance), sandblasters,
and wire brushes.

2.6.1.3 Sealant Preparation

The two types of sealants typically used to seal cracks in asphalt concrete pavement are hot-
applied thermoplastic bituminous materials and cold-applied thermosetting materials. Both hot-applied
and cold-applied materials may also be referred to as hot pour and cold pour materials. Upon reaching
the desired temperature, the sealant is stored in the tank until it is placed into the crack.

2.6.1.4 Sealant Placement and Finishing

Sealant placement occurs through hoses and wands that extend from the back of the heater truck.
In this placement process, a truck driver stops just past each crack, while a worker operates the hose and
wand to fill the cracks. Trucks will often have multiple hoses and workers following them. For clean-and-
seal, the hot pour sealant is placed directly into and over the crack. The placement can be finished with
the same wand used for application or by the use of a metal distributor. Upon the seal being laid flat

12
against the surface of the pavement, a blotter material may be applied to prevent the sticky seal from
being pulled up by vehicle tires. Typical blotter materials include toilet paper and soapy water. Figure 2.13
shows a cross-section view of sealant placement in clean-and seal crack repair. It is advised that all seals
should have time to cure prior to experiencing tire wear. If the road must be opened immediately, a
blotter material is relied on.

Figure 2.13: Cross-Sectional View of Clean-and-Seal (after, Smith and Romine, 1999)

2.6.1.5 Labor Needs and Cost

Clean-and-seal is the least labor-intensive form of crack sealing practiced in Minnesota. It typically
requires two trucks; one for crack preparation and one for sealing. The trucks will each require a driver.
The crack preparation truck is typically accompanied by 1-3 workers operating the air blaster or air lance.
The melter will also be accompanied by 1-3 workers operating the sealant applicators and blotting
material. A select number of workers are also required for traffic control. Since clean-and-seal is also the
fastest crack sealing method used in Minnesota, it requires a much fewer number of work hours. Johnson
(2000) stated that the unit price for clean-and-seal typically ranges between $0.10 and $0.30 per linear
foot, depending on the size of the project.

2.6.2 Sealant Installation in Rout-and-Seal Method

Rout-and-seal is used on transverse cracks. It involves using a router or pavement saw along the
length of a crack to create a reservoir centered over the existing cracks. The reservoir is then filled with a
sealant. In the MnDOT’s Best Practices for Asphalt Pavement Maintenance Handbook, Johnson (2000)
recommends performing rout-and-seal early in the pavement’s life to be successful. It states that if the
cracks are too badly deteriorated or too wide, consider filling them rather than sealing them. The
handbook does not directly state how wide is too wide, but it does recommend cutting a ¾-inch x ¾-inch
reservoir. The reservoir should have a flush sidewall, that is, the crack should not be so wide or
deteriorated that the saw or routing blade does not constitute the entire area of the reservoir.

13
There are three major benefits of creating the reservoir. One benefit is that the saw routing machine
removes a percentage of deteriorated material from the immediate area around the existing crack. This
reduces the possibility of water entering hairline secondary cracks around the existing crack and causing
continues deterioration. The second benefit of cutting a reservoir is that it creates a very cleanable and
sealable surface. It is much easier for workers to blow debris out of a fresh cut reservoir than an existing
crack with various irregularities in its geometric shape. The consistently clean sidewalls made by routing
are a good surface for a sealant to bond to. The third and largest benefit of routing cracks is that the
reservoir holds a large pool of sealant that has a larger ability to expand and contract with the pavement
due to thermal fluctuations. See Figure 2.14.

Figure 2.14: Routed Seal Expanding and Contracting Without Failure (after, Stoikes, 2017)

MnDOT specifies that rout-and-seal works best when performed in the spring and fall when temperatures
are cool. The MnDOT’s Best Practices for Asphalt Pavement Maintenance Handbook suggests that, if
performed at the right time, a rout-and-seal can be expected to perform for three years before significant
amounts of sealant begin to pull from the side of the reservoir. The seal will still perform after three years,
however, since it still does prevent some incompressible material and water from entering the crack. The
manual states that a project in Ontario has shown that rout-and-seal adds a minimum of two years and
an average of five years to the life of a pavement.

2.6.2.1 Equipment Needed

Rout-and-seal requires additional pieces of equipment when compared to clean-and-seal. The

extra equipment, a router, is used for making routs over the cracks being sealed. The remaining
equipment is as same as the ones required for clean-and-seal method.

14
2.6.2.2 Crack Preparation

The first step of performing the rout-and-seal crack treatment is routing the cracks. Routing
machinery comparable to that shown in Figure 2.8 is used for this. Figure 2.15 shows a crack before and
after being routed.

a b

Figure 2.15: Crack (a) Before and (b) After Being Routed

Upon a crack being routed, air blasters, air lances, or wire rushes must be used to clean the crack prior to
application of sealant. Figure 2.16 shows a routed crack after being air blasted and heat lanced.

Figure 2.16: Routed Crack After Cleaning

15
2.6.2.3 Sealant Preparation

Sealant preparation procedures for clean-and-seal and rout-and-seal are the same, except
different grades of sealant are recommended in these.

2.6.2.4 Sealant Placement and Finishing

After being cleaned, a routed crack is ready to be sealed. For a rout-and-seal, two passes of sealant
applications are made, typically with two different trucks about 5-10 minutes apart. The first pass fills the
reservoir partially full with sealant. The sealant is allowed to seep into the crack as it cools. The second
pass fills the reservoir full and finishes the placement in a band-aid, overband, or other configuration.
Configurations are discussed in section 4.4 of this report. The reason that rout-and-seal is done in two
passes is to allow the finished reservoir to lie flush with the surface of the pavement. Figure 2.17 (a) and
(b) below show a routed crack after the first pass of sealant and after the second pass with blotter paper.
A schematic of the cross-section view of the rout-and-seal sealant application is shown in Figure 2.18.

a b

Figure 2.17: Routed Crack (a) After First Sealant Pass and (b) After Second Pass and Blotter Application

16
Figure 2.18: MnDOT Recommended Placement Configuration for Rout-and-Seal (after, Smith and Romine, 1999)

2.6.2.5 Labor Needs and Costs

The costs involved with rout-and-seal are much higher than those of clean-and-seal, largely due to
increased labor needs. Rout-and-seal requires extra workers to operate the routing machines. Extra labor
is also needed to operate the truck preforming the second pass of sealant into the routed reservoirs. There
is also a cost increase because rout-and-seal requires more sealant material to fill the reservoirs. Johnson
(2000) specified the unit price of a typical rout-and-seal to range between $0.50 and $0.85 per linear foot,
depending on the size of the project; this is approximately two to three times more than what is required
for clean-and-seal method.

2.7 PLACEMENT CONFIGURATIONS

With the chosen crack sealing method, the sealant can be finished in a variety of configurations.
The Smith and Romine (1999) defined the most common configurations as shown below; also Figure 2.19
shows schematic of different sealant configurations.

Flush Fill: In this configuration, the material is simply placed into the existing crack. Excess material is
scraped off.

Simple Band-aid: Material is placed into an uncut crack. The excess material over the crack is shaped into
a flat overband using a squeegee.

Capped: Material is placed into an uncut crack. The excess material is left unshaped to form a “cap”.

Standard Reservoir-and-flush: Material is poured into a routed crack. No excess material is placed outside
of the reservoir.

17
Standard Recessed Band-aid: Material is poured into a routed crack. Excess material is shaped into a thin
overband using a squeegee.

Deep Reservoir-and-flush: Material is poured into a deep reservoir with no overband configuration.

Deep Recessed Band-aid: Material is poured into a deep reservoir and the excess material is shaped into
a thin overband with a squeegee.

Shallow Reservoir-and-Flush: Sealant is poured into a wide reservoir with no excess material.

Shallow Recessed Band-Aid: Material is poured into a wide reservoir and finished with a thin overband
using a squeegee.

Deep Reservoir and Recess (Backer Rod): A backer rod is installed into the bottom of the reservoir to keep
sealant from flowing into the crack or from forming a bond with the bottom of the reservoir. The reservoir
is then partially filled with sealant. No sealant is brought to the surface, in order to prevent material from
sticking to tires.

Deep Reservoir-and-Flush (Backer Rod): Material is poured onto a backer rod that is paced into the bottom
of a deep reservoir. No excess material is placed outside of the reservoir.

Deep Recessed Band-Aid (Backer Rod): Material is poured onto a backer rod placed at the bottom of a
reservoir. The material is then finished with a thin overband at the surface using a squeegee.

18
(i)

19
 (ii)

Figure 2.19 (i) and (ii): Various Sealant Configurations (after, Smith and Romine, 1999)

20
2.8 TYPES OF SEALANTS USED IN MINNESOTA

The type of sealant is a factor of climatic conditions, crack properties (spacing, density, and
orientation), traffic loading, and material availability with respect to cost. The most significant properties
for a crack sealant material to poses are:

a. Durability – The ability to endure traffic loading, tire wear, and climatic variations.
b. Extensibility – The ability of the material to deform as a crack expands.
c. Resilience – The ability of the material to recover after deformation.
d. Adhesiveness – The ability for a material to create a good bond to the wall of a crack.
e. Cohesiveness – The ability for a sealant to not suffer from internal ruptures during
elongation.

The different types of materials used to seal and fill cracks can be categorized into three types:

1) Cold-applied thermoplastic bituminous materials

a. Liquid asphalt (emulsion)
b. Polymer-modified liquid asphalt
2) Hot-applied thermoplastic bituminous materials
a. Asphalt cement
b. Fiberized asphalt
c. Rubberized/Polymerized asphalt
d. Asphalt rubber
e. Low-modulus rubberized or polymerized asphalt
3) Chemically cured thermosetting materials
a. Self-leveling silicone

Typically, cold-applied thermoplastics and low-quality hot-applied thermoplastics are used for filling
procedures. Higher quality hot-applied thermoplastics and silicone materials are used for sealing
operations. As per ASTM D 6690 (2015), rubber and polymer-modified sealants are categorized into four
classes to match low-temperature performance with climate, as shown below:

Type I: Moderate Climates, 50% extension at 0oF

Type II: Most climates, 50% extension at -20oF

Type III: Most climates, 50% extension at -20oF, with other special tests

Type IV: Very cold climates, 200% extension at -20oF

Various state agencies have made modifications to the ASTM D 6690 specifications to better suit their
climatic and traffic conditions. Al-Quadi et al. (2017) developed performance-based test guidelines for
selecting hot-applied thermoplastic sealants. In that study, a new set of testing regulations for sealant
properties was developed and recommended as various provisional AASHTO standards, as provided in
Appendix A.

21
The Minnesota Department of Transportation (MnDOT) uses three categories of sealants as provided
below:

1) MnDOT 3719
2) MnDOT 3723
3) MnDOT 3725

The MnDOT specifications have additional requirements on top of the in-place ASTM D 6690
specification. When compared to various other state transportation departments, these specifications
are much more precise. The physical requirements of these specifications, as defined by the 2016
Minnesota Standard Specifications for Construction (MnDOT, 2016), are provided in appendix A.

2.8.1 MnDOT 3719 (Joint and Crack Sealer: Hot-Poured, Crumb Rubber Type)

MnDOT 3719 is recommended for crack filling of non-working transverse and longitudinal cracks.
The test specification requires that the sealant material be composed of a mix of asphalt and crumb rubber
that is blended into a homogeneous mixture. The sealant is required to be heated in a double-walled
melter with mixing and temperature monitoring.

2.8.2 MnDOT 3723 (Joint and Crack Sealer: Hot-Poured, Elastic Type)

MnDOT 3723 exhibits good adhesion qualities and is recommended to be used in clean-and-seal
and rout-and-seal, where wider reservoirs are needed. The specification requires a sealant to be
composed of a combination of polymeric materials, fully reacted chemically to form a homogenous
compound.

2.8.3 MnDOT 3725 (Joint and Crack Sealer: Hot-Poured, Extra Low Modulus, Elastic
Type)

MnDOT 3725 has low resilience properties and is recommended only for use in transverse rout-
and-seal applications. As with MnDOT 3723, this specification requires that sealant be composed of a
homogenous mixture of fully chemically reacted polymeric materials. The heating of the material shall
take place in a temperature regulated, double-walled tank with mixing.

22
CHAPTER 3: PREVIOUS STUDIES ON CRACK SEAL
PERFORMANCE AND BENEFITS
3.1 PERFORMANCE EVALUATION OF CRACK SEALS

Smith and Romine (1999) suggested that sealant performance can be measured by summing the
lengths of failed segments and dividing by the total length of treated cracks inspected. After multiple field
inspections, a graph can be made that shows the treatment effectiveness over time. Also, by defining a
minimum threshold of effectiveness, the project lifespan can be determined. Figure 3.1 shows such a
graph, with treatment effectiveness (%) on the y-axis and age (months) on the x-axis. In addition, the
author recommended that at least one inspection be performed each year on treated cracks to document
the failure rates and plan for future maintenance operations. It is also recommended that a mid-winter
evaluation shall be conducted in order to evaluate the performance of crack treatments during maximum
crack expansions. Caltrans (2008) provided a guideline on the crack sealant distresses and their likely
cause and remedial measures. See Table 1 and Table 2.

Figure 3.1: Trend of Seal Deterioration over Time (after, Smith and Romine, 1999)

23
Table 1: Possible Sealant Failures and Likely Causes (after Caltrans, 2008)

24
 Table 2: List of Solutions for Commonly Observed Sealant Failures (after Caltrans, 2008)

Al-Quadi et al. (2017) described the most common types of distresses (Figure 3.2) observed in
crack sealants as follows:

(a) Adhesion loss – the loss of bond between the sealant and sidewall of the crack.
(b) Cohesion loss – the loss of bond within the sealant material itself.
(c) Overband wear – the loss of overband material to tires and snowplows.
(d) Tracking – the pull-out of material by vehicle tires and snowplows.
(e) Stone intrusion – the intrusion of rocks into the seal.

25
 Figure 3.2: Various Sealant Failures (after, Al Quadi, 2017)

Performance of more than 200 cracks was evaluated and documented. Representative pavements from
several states (including Minnesota) and a couple of provinces of Canada were considered in the study.
Each crack was evaluated for percentage length of full-depth adhesive and cohesive failures, plus the
percentage length of partial-depth adhesive and cohesive failures. Percent length of overband wear,
percent length of spalling, and the amount of stone intrusion were also documented. Sealant performance
was used to determine the performance index (PI). The performance index is a function of percent full-
depth adhesive loss and cohesive failures (AC) and a percent of partial adhesive and cohesive failures
(PAC) as shown in Equation 1.

PI = 100 – (AC + PAC x 0.5) (1)

It was observed that the most common failure was the adhesion loss that occurred during the winter
months. The performance of various sealant and seal types was monitored over three years. Figure 3.3
shows typical examples of performance evaluations for Minnesota and Ontario (Canada) sites of rout-
and-seal and clean-and-seal for various sealants. It can be seen that the significant drop in PI values for a
majority of the seals occurred between the second and third winters. The clean-and-seal almost
completely failed in all sections after the second winter, where rout-and-seal was still performing with an
acceptable threshold. The exact performance indexes are listed in Figure 3.4.

26
 (a) Minnesota, Rout-and-seal (b) Ontario, Rout-and-seal

(c) Minnesota, Clean-and-seal (d) Ontario, Clean-and-seal

Figure 3.3: Performance Evaluations of Crack Seals and Sealants (after Al-Quadi et al., 2017)

27
Figure 3.4: Performance Indices of Clean-and-seal (C&S) and Rout-and-seal (R&S) Methods (after Al-Quadi et al.,
2017)

In that study, in order to observe the influence of the overband, two roadway sections in New
York (NY) and New Hampshire (NH), were used; both overband and no-overband placement
configurations of sealant were also considered. As shown in Figure 3.5, it was found that the overband
seals out-performed the no-overband seals for all sealant types in most cases.

a b

Figure 3.5: (a) Overband Failure and (b) PI Comparison of Overband vs no-overband seals (after, Al-Quadi et al.,
2017)

Al-Quadi et al. (2017) also studied the effects of plow damage on clean-and-seal overbands in a
roadway at Michigan. The performance index of seals was evaluated over a three year time period. The
Performance Index (PI) was computed using the percent overband failure (OBF) caused by plow abrasion
and sealant tracking, as shown in Equation 2.

PI = 100 -%OBF (2)

28
It was found that the sealant suffered little damage in the first couple of years; however, there was a great
reduction in PI from 2012 to 2013.

3.2 BENEFIT OF CRACK SEALING

Rajagopal (2011) performed a study in cooperation with the Ohio Department of Transportation
(ODOT) to determine the overall benefit of crack sealing. The questions that the study addressed were:

- Do existing crack sealing practices within ODOT enhance pavement performance?

- If so, what is the optimum timing for treatment?
- Does the crack seal extend pavement life?
- Does crack sealing provide cost benefit? If so, to what extent?

Performance of several crack seals, mostly sealed with clean-and-seal technique, was evaluated at
different times to understand the overall benefits of crack sealing. Different types of pavements were
considered in that study as shown in Table 3. Service lives of the control sections (not sealed) were
compared with the sections that were sealed. It can be seen in Table 3 that creak seal could extend the
life of the pavement irrespective of the type of pavements. This study also found out that crack sealing
can enhance a pavement condition rating (PCR) by 5 points. The other significant observation in that study
was that the clean-and-seal could provide the largest five-year condition gain when the sealing is
performed in pavements with a PCR of 66 to 80. Operating outside of this range produced lower condition
gains.

Table 3: Service Life Extension among Various Pavement Types (after, Rajagopal, 2011)

In another study, Hajj et al. (2010) evaluated the performance of various maintenance activities
including crack filling by determining the Present Serviceability Index (PSI) before and after the treatment
until failure. A benefit-cost ratio analysis was used to compare the performance of various test sections
and corresponding treatments. The benefit was measured in terms of area under the performance curve
plotted between PSI and Time in years. The cost per lane-mile included the cost of labor, material, and

29
equipment. This study determined that crack filling is beneficial when performed on pavements that have
a present serviceability index (PSI) of 2.5 or more.

30
CHAPTER 4: CRACK SEAL PERFORMANCE EVALUATION IN
PRESENT STUDY
Crack sealing performance data was collected from multiple sources as shown in Figure 4.1. Since
data from these sources were collected at different times and scenarios, they were categorized as Phase
1, 2, 3, and 4.

Phase 2:
Phase 3:
Historical Data
Newly-installed
(interview & crack seal projects
MnDOT's const.
(site visits)
log)

Phase 1: Phase 4:
Practitioners' Old crack seal
opinions projects
(online survey) (site visits)
Performance
data Sources

Figure 4.1: Crack Sealing Performance Data Sources

4.1 PHASE 1: PRACTITIONERS' OPINIONS (ONLINE SURVEY)

The objective of the online survey was to understand the current practice of crack sealing
methods in the state of Minnesota and to generate a group of practitioners to collect more site-specific
data. A short five-question survey form was distributed to collect information on (i) the most commonly
used crack sealing method, (ii) criteria followed for deciding the suitable method, and (iii) average
anticipated service life for different sealing methods. The online survey was conducted electronically using
‘Google Survey Form’ during the summer of 2017. A copy of the survey form is provided in Appendix B.
The participants of the online survey were state-aid engineers, and personnel from MnDOT districts,
counties, and cities.

A total of 47 practitioners from various parts of the state participated in this short survey. The
locations of these responders are provided in Table 4. A map of the locations of the responders is shown
in Figure 4.2; where yellow, red, and green pins indicate city offices, county offices, and MnDOT districts
that responded, respectively.

31
Table 4: List of MnDOT Districts, Counties, Cities, and Contractors that participated in the Online Survey.

County Offices Cities MnDOT Districts

1. Lincoln County 1. Albert Lea

2. Clay County 2. Burnsville
3. Pennington County 3. St. Michael
1. MnDOT District 7
4. Dodge County 4. Inver Grove Heights
2. MnDOT District 6
5. Sherburne County 5. Coon Rapids
3. MnDOT Metro
6. Hennepin County 6. Austin
4. MnDOT District 8
7. Cottonwood County 7. Crystal
8. Houston County 8. Rosemount
9. Wabash County 9. Shoreview
10. Clearwater County 10. Little falls
11. Mille Lacs County 11. Hutchinson
12. Fillmore County 12. Pipestone
13. Aitkin County 13. Mankato
14. Murray County 14. Redwing
15. Anoka County 15. Bemidji
16. Koochiching County 16. Woodbury
Private Contractors
17. McLeod County 17. Andover
18. Rice County 18. Roseville 1. WSB & Associates
19. Sibley County 19. Oakdale
20. Blue Earth County 20. West St. Paul
21. Pope County
22. Ottertail County

32
Figure 4.2: Locations of the Online Survey Responders (Cities: Yellow, Counties: Red, MnDOT Districts: Green)

33
4.1.1 Data Analysis

Question 1: The first question was aimed at determining the most commonly used crack sealing method
in Minnesota. Figure 4.3 and Figure 4.4 analyze responses for Question 1 in the form of pie-charts. Figure
4.3 shows all the data; whereas, Figure 4.4 shows the data broken into categories by agency type (city,
county office, and MnDOT district). Figure 4.4(a) and Figure 4.4(b) show that counties and cities mostly
use the rout-and-seal method. The pie chart in Figure 4.4(c) shows that out of the four MnDOT districts
participated in the survey; three districts mostly prefer clean-and-seal method over the more expensive
rout-and-seal method. However, since only four districts participated in the survey, it may not be
appropriate to generalize the preference for all the MnDOT districts. Figure 4.5 shows the geographic
locations of the responders according to their preference of sealing method: yellow pins indicate locations
that reported the primary use of rout-and-seal and green pins indicate the locations that reported the
primary use of clean-and-seal. No clear trend is seen by locating geographic positions, however.

Q1. Between different crack sealing methods, which one do you

use more?

15

32

Clean-and-Seal: Rout-and- Seal

Figure 4.3: Analysis of the Responses to Question 1, All data

34
 (a) (b) (c)

Figure 4.4: Preferred Crack Sealing Method (a) City Offices (b) County Offices (c) MnDOT Offices

35
Figure 4.5: Locations Reporting Preference about the Crack Sealing Method: Rout-and-seal (yellow) and Clean-
and-seal (green)

Question 2: This question was asked to understand the selection criteria for choosing a crack sealing
method. Figure 4.6 analyzes the responses obtained for this question. Although responses were varied,
an effort was made to compile the results into different categories. The categories include no criteria, pre-
determined schedule, crack/pavement conditions, pavement age, subgrade material, and budget. It can
be seen that the most common criteria for selecting the crack sealing method are the crack/pavement
conditions, some practitioners although use the pre-determined schedule and pavement age as the
criteria for choosing the sealing method.

36
 Figure 4.6: Analysis of the Responses to Question 2

When responses from questions 1 and 2 were combined, it was observed that those who primarily
use rout-and-seal method, a good number (6 out of 31) of them do not follow any criteria (Figure 4.7).
The responders who answered that they primarily use clean-and-seal method indicated that they most
typically use crack conditions as their method selection criteria (Figure 4.8).

37
 Figure 4.7: Selection Criteria for Rout-and-seal Figure 4.8: Selection Criteria for Clean-and-seal

Question 3: This question was asked to know the typical service life of the seal when installed according
to the clean-and-seal method. The service life is defined as the time between the sealant installation and
failure. Since several responders noted that they only use the rout-and-seal method, the response was
not provided to this question by all the 47 responders. Available responses were grouped into several
ranges and presented in Figure 4.9. These responses were ranged from 2-3 years to over a10-year period.
Note that the variability in the responses could be because no threshold serviceability was defined in the
question. Additionally, some of the responders may not track the performance of crack sealing closely
enough to provide a quantitative assessment.

38
 Figure 4.9: Typical Service Life of Clean-and-seal Method

Question 4: This question was asked to know the typical service life of seals installed using the rout-and-
seal method. Responses were grouped into several ranges and presented in Figure 4.10. These responses
ranged from 2-5 years to over 15 years. In addition to the reasons for variation in the service life
mentioned in Question 3, the performance of the seals is likely influenced by the pavement and traffic
conditions.

Figure 4.10: Typical Service Life of Rout-and-seal Method

39
Question 5: This question was asked to know if the responder would be willing to further participate in
this study by providing site-specific information. A total of 24 responders agreed to provide additional
information, as shown in Figure 4.11. From those 24 who responded “yes,” a total of seven indicated that
they primarily use clean-and-seal, while the remaining 17 reported the primary use of rout-and-seal
(Figure 4.12). A follow-up, in-depth survey was sent to the 24 responders to obtain the performance data
of previous crack sealing projects. The results of the follow-up survey are presented in section 4.2 of this
report.

Figure 4.11: Number of Responders agreed to take the Follow-up Survey

Figure 4.12: Preference of the Sealing Method for the Responders of Question 5

40
Question 6: The purpose of this question was to select some representative field sites for collecting crack
seal performance data from newly sealed projects. A total of eleven project sites, located throughout the
state, were identified where cracks were sealed in the Summer and Fall of 2017. The descriptions of these
crack sealing project sites are presented in Section 4.

4.2 PHASE 2: HISTORICAL DATA

In order to collect more in-depth and site-specific data, a secondary survey form (Appendix C) was
sent out to 24 survey participants those had previously agreed to provide additional information (Online
survey: Q#5). This survey also asked questions on the cost-effectiveness of crack sealing methods. Even
though good responses were received on the cost related questions from this exercise, conclusive
information could not be collected on the crack seal performance. The research team then conducted
telephonic or face-to-face interviews to determine alternate data sources. Based on the feedback of these
interviews, it was understood that MnDOT’s Pavement Construction Project Log could be a good source
to estimate the service period of the crack seals. Accordingly, the historical data of crack sealing projects
were collected and studied to draw a possible conclusion on the service period of the crack seal as a
function of various parameters, such as traffic, pavement surface layer’s thickness and age at the time of
sealant installation, sealant material, and sealing procedure. It may be noted that the ‘service period’ is
defined as the period between the sealant installation and the follow-up surface treatment or major
rehabilitation work, irrespective of the failure of the sealant. The ‘service period’ is different from the
‘service life’; it is the period between the sealant installation and time when the sealant fails, irrespective
of the follow-up surface treatment or major rehabilitation work.

4.2.1 Interviews

A total of ten interviews were conducted, either through telephone or visiting the interviewee’s
work location; a majority of the people being interviewed were maintenance engineers or supervisors.
The interviews were focused on collecting their experiences on the performance of the crack sealing
methods and identifying potential data sources on the performance of the crack seals. The results of the
interviews, however, did not yield a clear distinction between the performance of rout-and-seal and clean-
and-seal methods in terms of their service life. MnDOT District 7 expressed that the clean-and-seal
performs longer than the rout-and seal; according to them, the major drawback of the rout-and-seal is
the adhesion failure. Whereas, MnDOT District 1, the City of Duluth, and the City of St. Michael prefer the
rout-and-seal over the clean-and-seal method; the tearing out of sealant material is the notable failure
with the clean-and-seal method according to them.

It was found that the deciding factors in choosing between rout-and-seal method and clean-and-
seal method vary between the agencies. Some of the factors include the age of the pavement surface
layer, crack width, and also the type of subgrade soil. For example, Sherburne County and the City of
Andover use rout-and-seal on 1- to 2-year old asphalt layers and clean-and-seal on 8-year or older layers.
Beltrami County uses rout-and-seal for pavements with clayey and silty soil subgrades and the clean-and-
seal for sandy soil subgrades. The City of Hutchinson follows the crack width criteria as specified MnDOT
guidelines. Many of these agencies also lay a seal coat on the crack-sealed sections in the same or
following season, or a couple of years later, especially for the important roadways. At the same time,
some agencies like Sherburne County and the City of St. Michael expressed their concerns with seal coat;
they suspect seal coat invites moisture-related distresses, such as stripping.

41
 From the interviews, it can be concluded that a greater number of agencies use rout-and-seal
method than the clean-and-seal method. The choice of sealing method primarily depends on their
experience and other local conditions. Overall, the interviews with the maintenance engineers or
supervisors were helpful to recognize the current practice of crack sealing methods and most importantly
identify another useful data source, such as MnDOT’s Pavement Construction Project Log.

4.2.2 MnDOT’s Pavement Construction Project Log

MnDOT maintains a record of pavement construction projects (MnDOT, 2018b) with details of
year built, project number, type of work along with specification, thickness (if available), and plan of the
project location. A screenshot of such data log is shown in Appendix D. Additionally, MnDOT maintains a
traffic record (MnDOT 2018c) for various routes. A screenshot of traffic data for a county map is shown in
Appendix D. MnDOT also maintains a record of post letting awarded bid abstracts (MnDOT 2018d) under
Bid letting database. These abstracts provide the job description along with cost information. A screenshot
of an awarded abstract search page from the bid-letting website is shown in Appendix D. Also, the data
such as a list of crack sealing projects in last ten years, proposal documents, etc. were obtained from the
MnDOT’s Office of Materials and Research (Paul W. Nolan, Personal Communication). All of these data
sources were initially used to identify the method of crack sealing performed and then to determine the
service period.

4.2.2.1 Identification of Crack Sealing Methods

Rout-and-seal

In order to collect the performance of rout-and-seal projects, a total of 84 projects (Appendix E and
Appendix F) were reviewed from various MnDOT districts. However, only 26 rout-and-seal projects were
finally shortlisted based on the completeness of the dataset as shown in Appendix G and Appendix H.
Note that many of the crack sealing treatments before the year 2000 were noted as crack repair; no
documentation was found to identify the type of crack treatment though.

Clean-and-seal

Unlike the rout-and-seal method, it was difficult to identify the clean-and-seal projects from
construction project log due to lack of information. Hence, the actual bid proposal documents for few
potential projects (State Project # 8821-71/221/242, 8823-273, etc.), collected from MnDOT’s Office of
Material and Research, were used to identify the clean-and-seal projects. The proposals in which the
sealant material Type 3723 was used, cracks were assumed to be sealed with the clean-and-seal method.
Later, the construction plans ( Appendix J) with details of project sites in those proposals were used to
locate the project section.

A total of 12 clean-and-seal projects were identified (Appendix J and Appendix K), constructed
between the year 2007 to the year 2013. These projects did not have any follow-up treatment on record.
The latest available pavement condition ratings (MnDOT 2018e) performed in 2016 indicate that these
project segments are still in good condition.

42
4.2.2.2 Traffic Data

MnDOT’s traffic data maps that are available for different years were used to determine the
representative Average Annual Daily Traffic (AADT) for each project section. This was obtained by
averaging the traffic between the year of crack sealant installation and the year of follow-up treatment
(Appendix L to Appendix N).

4.2.2.3 Service Period

The performance of crack seals was evaluated based on the average service period. Even though
the service period may be influenced by the factors like AADT, thickness, and age of the pavement surface
on which sealant was installed. However, it mostly depends on the maintenance and pavement
preservation plan of the agency, which often is dictated by the budget. Appendix O and Appendix P
present the service periods for the different rout-and-seal and clean-and-seal projects considered.

Figure 4.13 shows the average service period of rout-and-seal and clean-and seal projects. It may
be stated that only 12 clean-and-seal projects were available to compare with 26 rout-and-seal projects;
therefore, a strong conclusion on the service period of the clean-and-seal method (at least from this
exercise) may not be justified. Within the available dataset, the average of service periods for both the
methods was found to be similar, which is close to 6 years. The statistical analysis (ANOVA) conducted
upon the results of service periods of rout-and-seal and clean-and-seal projects also did not show any
significant difference (Table 5). Although this could be an oversimplification, it appeared that MnDOT
usually considers a follow-up treatment approximately after six years of the crack treatment.

10
Average Service Period, Years

8
6.4 6.0
6

0
Rout-and-Seal Clean-and-Seal
Crack sealing method

Figure 4.13: Average Service Period of Rout-and-seal and Clean-and-seal Projects

43
 Table 5: ANOVA Results – Service Period of Rout-and-seal and Clean-and-seal Projects

Anova: Single Factor

SUMMARY

Groups Count Sum Average Variance

R&S 26 167 6.42308 5.29385

C&S 12 72 6 4.54545

ANOVA

Source of Variation SS df MS F P-value F crit

Between Groups 1.46964 1 1.46964 0.29015 0.59344 4.11317

Within Groups 182.346 36 5.06517

Total 183.816 37

*Note: P-value < 0.05 indicates a significant difference between the groups

Table 6 shows the descriptive statistics of service periods of rout-and-seal and clean-and-seal
projects. The service periods of rout-and-seal projects vary between 3 and 11 years; whereas clean-and-
seal projects vary between 5 and 11 years. The service periods, at 95% confidence level, for these projects
were found to be 5.5 - 7.4 years for rout-and-seal, and 4.6 -7.4 years for clean-and-seal. Figure 4.14 and
Figure 4.15 show the histogram of service periods of these methods.

44
Table 6: Descriptive Statistics – Service Periods of Rout-and-seal and Clean-and-seal

Descriptive Rout-and-seal Clean-and-seal

Mean 6.4 6.0

Standard Error 0.451 0.615

Median 6.0 5.0

Mode 6.0 5.0

Standard Deviation 2.301 2.132

Sample Variance 5.294 4.545

Kurtosis -0.701 2.754

Skewness -0.032 2.026

Range 8 6

Minimum 3 5

Maximum 11 11

Sum 167 72

Count 26 12

Confidence Level (95.0%) 0.929 1.355

6
Frequency

0
3 4 5 6 7 8 More
Service Periods, Years

Figure 4.14: Histogram – Service Periods of Rout-and-Seal Projects

45
 10
8
Frequency 6
4
2
0
3 4 5 6 7 8 More

Service Periods, Years

Figure 4.15: Histogram – Service Periods of Clean-and-Seal Projects

Influence of Traffic on the Service Period

Figure 4.16 shows the average service periods of rout-and-seal and clean-and-seal projects with
respect to traffic categories, below and above 10,000 AADT. The average service period for the rout-and-
seal and clean-and-seal are similar though, approximately one year higher for the roads with less than
10,000 AADT. The statistical analysis conducted on different combinations such as within and between
rout-and-seal and clean-and-seal methods did not show any significant difference (P-value > 0.05) at a
95% level of confidence (Table 7). Additionally, no clear correlation was found between the AADT and
service period (Figure 2.12 to Figure 2.17) for both of the sealing methods.

10
Rout-and-Seal

8 Clean-and-Seal
6.9
Service Period, Years

6.4
6 5.4 5.3

0
Below 10,000 Above 10,000
Traffic (AADT)

Figure 4.16: Service Period vs. Traffic (AADT) – Crack Sealing Methods

46
Table 7: ANOVA Results – Service Period of R&S and C&S with respect to Traffic

SS df MS F P-value F crit

Below 10,000 AADT (Rout-and-seal Vs. Clean-and-seal)

Between Groups 1.46261 1 1.46261 0.26264 0.613 4.25968

Within Groups 133.653 24 5.56887

Above 10,000 AADT (Rout-and-seal Vs. Clean-and-seal)

Between Groups 0.04167 1 0.04167 0.01277 0.91226 4.9646

Within Groups 32.625 10 3.2625

Rout-and-seal (Below 10,000 AADT Vs. Above 10,000 AADT)

Between Groups 12.6934 1 12.6934 2.54604 0.12366 4.25968

Within Groups 119.653 24 4.98553

Clean-and-seal (Below 10,000 AADT Vs. Above 10,000 AADT)

Between Groups 3.375 1 3.375 0.72386 0.41479 4.9646

Within Groups 46.625 10 4.6625

*Note: P-value < 0.05 indicates a significant difference between the groups

47
 12 12
Service Period, Years

Service Period, Years

10 10
8 8
6 6
4 4 R² = 0.0766
R² = 0.0055
2 2
0 0
- 20,000 40,000 60,000 80,000 100,000 - 2,000 4,000 6,000 8,000 10,000 12,000
Traffic (AADT) Traffic (AADT)

Figure 4.17: Service Period vs. Traffic (AADT) - All, Rout-and-seal Figure 4.18: Service Period vs. Traffic (AADT) – Below 10,000, Rout-
Projects and-seal Projects

10 12

Service Period, Years

Service Period, Years

8 10
8
6
R² = 0.6582 6
4
4
2 2 R² = 0.0008

0 0
- 20,000 40,000 60,000 80,000 100,000 - 5,000 10,000 15,000 20,000 25,000
Traffic (AADT) Traffic (AADT)

Figure 4.20: Service Period vs. Traffic (AADT) – Above 10,000, Rout- Figure 4.19: Service Period vs. Traffic (AADT) - All, Clean-and-seal
and-seal Projects Projects
48
 12

10

Service Period, Years

R² = 0.2615
8

0
- 2,000 4,000 6,000 8,000 10,000
Traffic (AADT)

Figure 4.21: Service Period vs. Traffic (AADT) – Below 10,000, Clean-and-seal Projects

7
6
Service Period, Years

5
4 R² = 0.4572
3
2
1
0
- 5,000 10,000 15,000 20,000 25,000
Traffic (AADT)

Figure 4.22: Service Period vs. Traffic (AADT) – Above 10,000, Clean-and-seal Projects

49
Influence of Pavement Surface Layer Thickness on the Service Period

Figure 4.23 and Figure 4.24 show the results of service Period vs. pavement surface layer thickness
for rout-and-seal and clean-and-seal projects. No significant correlation between the pavement surface
layer thickness and service period was found either.

12

10
Service Period, Years

6
R² = 0.0939
4

0
0 1 2 3 4 5 6
Pavement Surface Layer Thickness, Inches

Figure 4.23: Service Period vs. Pavement Surface Layer Thickness, Rout-and-seal Projects

12

10
Service Period, Years

4 R² = 0.1585
2

0
0 1 2 3 4 5 6 7
Pavement Surface Layer Thickness, Inches

Figure 4.24: Service Period vs. Pavement Surface Layer Thickness, Clean-and-seal Projects

50
Influence of the Pavement Surface Layer Age on the Service Period of Crack Sealants

Figure 4.25 and Figure 4.26 show the results of service period vs. pavement surface layer age for
rout-and-seal and clean-and-seal projects, respectively. Also in this case, the plots did not show any clear
correlation for both the sealing methods within the available dataset; except for indicating a slightly
declining service period for older pavement surfaces.

12
10
Service Period, Years

8
6
4
R² = 0.0013
2
0
0 2 4 6 8 10
Pavement Surface Layer Age, Years

Figure 4.25: Service Period vs. Pavement Surface Layer Age, Rout-and-seal Projects

12
10
Service Period, Years

8 R² = 0.1226

6
4
2
0
0 2 4 6 8 10 12
Pavement Surface Layer Age, Years

Figure 4.26: Service Period vs. Pavement Surface Layer Age, Clean-and-seal Projects

51
4.3 PHASE 3: NEWLY-INSTALLED CRACK SEAL SITES

The performance evaluations of 11 newly-installed crack sealing sites are presented in this
section. All of these sites were sealed in the summer and fall of 2017. Figure 4.27 shows the locations of
these sites. Yellow, green, and red pins represent the locations of the sites sealed with the rout-and-seal
method only, the clean-and-seal method only, and both the methods. The research team visited these
sites during the sealant installation. Many data, such as sealing procedures, materials, sites, and crack
locations were documented at the project sites. A copy of the form used for data collection is attached in
appendix R.

Following the sealant installation, additional visits were made to most of the 11 sites at various times:
approximately 2 months (end of summer), 6 months (mid-winter), 8 months (end of winter), 12 months
(second summer, and 18 months (during second winter) after the sealant application. On each visit, the
length of the failed sealant and the failure type were documented. These data were used to determine
the performance index (PI) for each sealed crack. The PI equation used in Al-Quadi et al. (2017) study was
slightly modified for the current project to include the spalling failure, which has been noticed in multiple
project sites unlike in other places. As shown in Equation 1, the performance index is a function of total
crack length, cohesion loss, spalling, partial-depth adhesion loss, and full-depth adhesion loss.

𝑃𝐼 = 100 − [(𝑃𝐹𝐷𝐴) + (𝑃𝑃𝐷𝐴 ∗ 0.5) + (𝑃𝑆) + 𝑃𝐶𝐿] (3)

Where:

PI = Performance Index

PFDA = Percent full depth adhesion loss by length

PPDA = percent partial depth adhesion loss by length

PS = Percent spalling by length

PCL = Percent cohesion loss by length

In this equation, percent partial-depth adhesion loss and percent cohesion loss are multiplied by a factor
of 0.5 as it remains unclear if the seal remains water-tight or not by observation with the naked eye.
Percent full-depth adhesion loss and percent spalling are multiplied by a factor of 1.

52
 Figure 4.27: Locations of Newly-installed Crack Seal Field Sites

A sample performance documentation form is attached in Appendix Q. The performance criteria noted in
these forms included plow abrasion, wheel-path flushing, pullout failure, full-depth adhesion loss, partial-
depth adhesion loss, cohesion loss, spalling, and heaving. The majority of the distresses observed in the
field sites included partial-depth adhesion loss, full-depth adhesion loss, and spalling. Examples of the
abovementioned failure types are shown in Figure 4.28. Some cohesion failures were observed in a few
clean-and-seal cracks. Early signs of cohesion failures were observed on a few rout-and-seal cracks;
however, the cohesion loss only extended through the upper portion of routed reservoirs, still allowing

53
the seals to remain watertight. These early signs of cohesion failure were not included in the performance
index calculation. An example of the cohesion failure observed is shown in Figure 4.29.

Original
Crack

(i) (ii) (iii)

Figure 4.28: Examples of Seal Failures (i) Spalling, (ii) Partial Adhesion Loss, (iii) Full-depth Adhesion Loss.)

i ii

Figure 4.29: (i) Cohesion Failure of Clean-and-Seal Crack (ii) Early Signs of Cohesion Failure on Rout-and-Seal
Crack

54
 In addition to recording measurements, efforts were made to take photographs of each crack
during each field visit. Placing these photographs side-by-side allows for a visual interpretation of seal
degradation with respect to time. Photographs could not be collected from every crack in each site visit
because of traffic, weather, and limited visibility of crack ID numbers in some sites due to snow on
pavement and shoulders. The construction history of the pavement surface layer and traffic for these
sites, if available, was obtained from MnDOT databases as described in Section 4.2.2 .

The performance index for every crack of all the project sites was calculated using Equation 3. The
following subsections discuss the performance evaluations of Site E, and Site O. Site E has been selected
for the representation of rout-and-seal sites while Site O has been chosen to represent sites containing
both cracks sealing methods. The data for all other sites are included Appendix S. The data for each site is
presented in tables, graphs, and pie charts. Rows that are highlighted red have been excluded from the
overall analysis of each site, as these cracks are mainly longitudinal, treated with filling, and fail differently
than the primary thermally induced transverse cracks.

4.3.1 Performance Analysis for Site E

Site E is located on TH 200 between miles 176 and 180 which has an AADT of 1,200. The site
contains ten documented cracks. Nine of the cracks were sealed with the rout-and-seal method and one
crack, along the lane joint, left untreated. The sealed cracks ran in the transverse direction and were of
low severity.

All the cracks were sealed with MnDOT 3725 sealant. Three trucks were used in the sealing
operation in that project site. One truck would drive 10-15 minutes ahead of the other two, filling the
routed reservoirs half full of sealant. The second truck would fill the reservoir and apply toilet paper over
the freshly sealed crack. The third truck performed filling on longitudinal cracks in a single pass.

The most recent construction data for this site includes an asphalt surfacing in 1999 followed by
crack sealing in 2001. The next rehabilitation performed on this section of road was a 1.5-inch mill and
3.5-inch asphalt overlay in 2012 on which the crack sealants were installed in 2017. Information on each
of the ten cracks is shown in Table 8.

Table 8: Documented Cracks in Site E

55
 The data on the most recent site visit (second winter) is presented in Table 9 and Figure 4.30
below. In this site, noticeable amounts of adhesion and spalling failures had occurred in the first winter
itself. The performance index vs. seal age is plotted in Figure 4.31. Figure 4.32 shows the photographs of
the cracks during installation, during the first winter, and during the second winter since installation. This
figure indicates that the PI dropped first time during the first winter and then remained the same until
the next winter when it dropped again by 10 to 20%.

Table 9: Performance of Site E after Second Winter

Site E Performance (end of 2nd Winter)

Crack IDcrack lengthLength PDA Length FDA Length Spalling Unfailed length Performance Index
E-1 12 1.5 3 1.5 6 56.25
E-2 12 1.5 3 0 7.5 68.75
E-3 12 2 3 2 5 50.00
E-4 12 0 1 0 11 91.67
E-5 90 90 x
E-6 12 1 3 1 7 62.50
E-7 12 1 3 1 7 62.50
E-8 12 3 0 0.5 8.5 83.33
E-9 12 0.5 0 2 9.5 81.25
E-10 12 3 0.5 0.5 8 79.17
sum 108 13.5 16.5 8.5 69.5 70.60

56
 5
9.5 13.5
5
16.5

69.5 8.5
88.5

Length PDA Length FDA Length PDA Length FDA

Length Spalling Unfailed length Length Spalling Unfailed Length

ii
i
Figure 4.30: Seal Performance at Site E after (i) First Winter, and (ii) Second Winter

Site E Performance (until second winter)

100
90
Performance Index (PI)

80
70
60
50
First winter
40
30 Second winter

20
10
0
0 5 10 15 20
Age (months)

Figure 4.31: Site E Performance Index of Site E after the Second Winter

57
 E-1 E-2

ii iii
ii i iii
i

E-3 E-4

iii iiii ii
i

E-5 E-6

58
 ii ii i
i

E-7 E-8

i iii iii
iii
ii ii iiii

E-9 E-10

59
 ii
ii ii iii iii iv

Figure 4.32: Photographic Performance Documentation of Site E. E-1-i Shows Crack E-1 upon being sealed. E-1-ii
and E-1-iii show Crack E-1 at the end of its first winter and during the second winter after being sealed. This
same sequence of photos follows for cracks E-2, E-3, E-4, E-5, E-6, E-7, E-8, E-9, and E-10. E-10-iv shows the seal
performance at the end of the first winter compared to the mid-winter performance presented in E-10-iii.

4.3.2 Performance Analysis for Site O

Site O is located on CR 10 in Koochiching County, MN, which is a rural section of road. The AADT
is around 100, and no construction history was found. The pavement surface appears to be quite old
though, with crack spacing approximately ranging from 5 - 10 feet. This site consists of 20 documented
cracks, all of which cracks run in the transverse direction. Ten of the twenty were sealed with clean-and-
seal and the remaining ten were sealed with rout-and-seal. Information on each of the twenty cracks is
shown in Table 10. The data on the most recent site visit (second winter) is presented in

60
 Table 11. The performance evaluation of this site for the first and second winter, for both crack
sealing methods, are presented in Figure 4.33 through Figure 4.35. This site experienced minimal failure
in the first winter. However, the following more severe second winter had caused significant expansion
of the crack width and failed many seals as a result. The performance data collected during the second
winter visit of Site O indicates that the clean-and-seal deteriorated much faster than a rout-an-seal. The
increased amounts of failures seen in the clean-and-seal cracks at this site included full and partial depth
adhesion failures as well as cohesion failures. This was the only site where cohesion failure was observed,
and the cohesion failure was only observed on the clean-and-seal cracks.

Table 10: Documented Cracks in Site O

61
Table 11: Performance of Site O after Second Winter

Site O Performance (end of 2nd Winter) (clean-and-seal)

Crack ID crack length Length PDA Length FDA Length Spalling Length Cohesion Unfailed length Performance Index
O-1 12 0 0 0 0 12 100.0
O-2 12 6 0 0 0 6 75.0
O-3 12 8 0 0 0 4 66.7
O-4 12 2 5.5 0.5 1 3 41.7
O-5 12 3 0 0 1 8 100.0
O-6 12 1 0 0 0 11 95.8
O-7 12 2 0 0 0 10 91.7
O-8 12 1 0 0 0 11 95.8
O-9 12 1 0 0 0 11 91.7
O-10 12 2.5 0 0 0.5 9 79.2
sum 120 26.5 5.5 0.5 2.5 85 83.8
Site O Performance (end of 2nd Winter) (rout-and-seal)
Crack ID crack length Length PDA Length FDA Length Spalling Length Cohesion Unfailed length Peformance Index
O-11 12 0 0 0 0 12 100.0
O-12 12 0 0 0 0 12 100.0
O-13 12 1 0 0 0 11 95.8
O-14 12 0 0 0 0 12 100.0
O-15 12 0 0 0 0 12 100.0
O-16 12 1 0 0 0 11 95.8
O-17 12 1 0 0 0 11 95.8
O-18 12 1.5 0 0 0 10.5 93.8
O-19 12 0 0 0 0 12 100.0
O-20 12 1 0 0 0 11 99.5
sum 120 5.5 0 0 0 114.5 98.1

62
 0.5 0 0
0.5

26.5

5.5
0.5
2.5
85
119
i ii

Length PDA Length FDA Length PDA Length FDA

Length Spalling Length Cohesion Length Spalling Length Cohesion
Unfailed length Unfailed length

Figure 4.33: Site O Clean-and-Seal Performance, after First Winter (i) and Second Winter (ii)

00 0 0 0
5.5

114.5
120
i
ii

Length PDA Length FDA Length PDA Length FDA

Length Spalling Unfailed length Length Spalling Unfailed length

Figure 4.34: Site O Rout-and-Seal Performance After First Winter (i) and Second Winter (ii)

63
 Site O Clean-and-Seal Performance (second
winter)
100
Performcane Index (PI)

80

60

40

20

0
0 5 10 15 20
Age (months)

Figure 4.35: Site O Clean-and-Seal Performance Index of Site E after the Second Winter

Site O Rout-and-Seal Performance (second

winter)
100
Performance Index (PI)

80

60

40

20

0
0 5 10 15 20
Age (months)

Figure 4.36: Site O Rout-and-Seal Performance Index of Site E after the Second Winter

64
4.3.3 Performance Summary of Newly-installed Crack Sealants

Similar to the Sites E and O, performance indices were calculated for all other sites. Table 12
provides a summary of the performance indices of all these sites along with available pavement and traffic
information. After the first-year service (one summer and one winter), the average performance indices
for rout-and-seal and clean-and-seal sites were found to be 85 and 90. As anticipated, the results from
the second winter evaluations showed that the performance of the crack sealing is continuing to
deteriorate with respect to time and expedited by the harsh winter. The rates of deterioration, however,
were greater than that was anticipated, which is discussed further in section 5.1 of this report. As the
winter of 2018 - 2019 in Minnesota was very severe (up to -70oF wind chill), many of the cracks considered
in this study expanded beyond their average winter expansion, which had failed many seals prematurely.
The average performance indices for the second winter for rout-and-seal and clean-and-seal are 68.5 and
80.5, respectively. However, it should not be concluded that the clean-and-seal performed better than
the rout-and-seal as the number of cracks sealed with rout-and-seal was much higher than its counterpart.

One of the common failures observed (primarily on rout-and-seal) was spalling. Although it was
not formally addressed as a seal failure for calculating performance in other studies, spalling is
contributing to a large portion of failure in the cracks documented for the current study. This failure likely
occurs due to weakening of the asphalts along the sides of the routed reservoir, probably because of the
mechanical agitations exerted on the asphalt by the router. The weak asphalt breaks due to the cyclic
wheel load from the traffic and creates spalling. Cracks with rout-and-seal were also found to be
experiencing a large amount of adhesion failures. This failure was seen more commonly with increased
crack spacing. The overall performance on rural roads with closely spaced cracks was nearly identical for
both clean-and-seal and rout-and-seal, except Site O. Cohesion failure has not been observed in any of
the seals during the first 18 months of service.

65
Table 12: Summary of Performance Evaluations of Newly-installed Crack Sealing Project Sites

Pavement Surface Layer Crack Sealing Avg. PI during

Traffic 2nd Winter
S. No. Site Route Year Thickness Year Avg. PI after
Description Sealant type (AADT)
Paved (Inch) Treated 1st Winter

Clean-and-seal (C&S)

1 A US-53 2008 Bit. Overlay (2360) 1.5 2017 C&S (3723) 10,300 - -

2 C US-53 2008 Bit. Overlay (2360) 1.5 2017 C&S (3723) 16,100 100.0 -

C&S (3725) and 99.0

3 I CR 5 - - - 2017 R&S (3725) - 99.5

4 M - - - - 2017 C&S and R&S* - 78.5 8.0

5 N - - - - 2017 C&S and R&S* - 75.0 58.0

6 O CR 10 - - - 2017 C&S and R&S* - 95.5 84.0

Average 90.0 80.5

Rout-and-seal (R&S)

1 D US-53 2009 Concrete Overlay 8 2017 R&S (3725) 7,800 88.0 63.5

2 E Mn-200 2012 Bit. Overlay (2360) 3.5 2017 R&S (3725) 1,200 84.0 71.0

3 F US-169 2010 Bit. Surf. (2360) Var. 2017 R&S (3725) 6,100 74.0 59.0

4 G Mn-1 2014 Bit. Overlay (2360) 3 2017 R&S (3725) 2,300 67.0 40.0

66
5 H TH 53 - - - 2017 R&S (3725) - 95.0 63.5

C&S (3725) and 96.0

6 I CR5 - - - 2017 R&S (3725) - 97.0

7 M - - - - 2017 C&S and R&S* - 95.8 8.0

8 N - - - - 2017 C&S and R&S* - 63.0 51.0

9 O CR 10 - - - 2017 C&S and R&S* - 100.0 98.0

Average 85.0 68.5

Note: ‘-’ indicate the data is not available; ‘*’ indicate sites (e.g., M) had both the R&S and C&S methods

67
4.4 PHASE 4: OLD CRACK SEAL SITES

This project also considered some old sealed cracks. Table 13 shows the list of crack seal sites
considered in Phase 4; one representative crack was considered in the site. Sealants were installed in
these sites between 2012 and 2017. These sites were located at the City of Duluth, Hutchinson, and
Andover and had relatively lower AADT compared to some of the newly sealed project sites discussed in
the previous section. All these sites were sealed with only the rout-and-seal method; suitable sites for
clean-and-seal method could be found. An effort was made to collect various details such as sealing
material, sealing procedure, costs, pavement, and traffic data. The sites in and around the City of Duluth
were monitored for three seasons: spring 2018, fall 2018 and spring 2019; whereas sites in the two other
cities were visited only once. The year of crack sealing for these sites were 2014, 2015, and 2016.

The methodology used for the performance evaluation of these sites was similar to what was
followed for the newly-installed crack seal project sites in Phase 3, except documenting a few additional
failures such as stone intrusion, pull-out failure, partial cohesion loss, and full cohesion loss. Examples of
these failure types are shown below in Figure 4.37. Note that these specific failures were much visible
during the fall season than spring season. The equation used for the performance index for the old crack
sealing project sites is given below:

𝑃𝐼 = 100 − [(𝑃𝐹𝐷𝐴) + (𝑃𝑃𝐷𝐴 ∗ 0.5) + (𝑃𝑆) + (𝑃𝐹𝐶𝐿 ∗ 0.5) + (𝑃𝑃𝐶𝐿 ∗ 0.25) + (𝑃𝑃𝐹) +
(𝑃𝑆𝐼 ∗ 0.25)] (4)

Where:

PFCL = Percent full cohesion loss by length

PPCL = Percent partial cohesion loss by length

PPF = Percent pull-out failure by length

PSI = Percent stone intrusion by length

Other variables were previously defined.

In this equation, percent full cohesion loss is multiplied by a factor of 0.5 since a portion of sealant remains
watertight. The percent partial cohesion is multiplied by a factor of 0.25 since these cohesive cracks
appear mostly on a surface level. The percent pull-out failure and the percent stone intrusion failure are
multiplied by a reasonable factor of 1 and 0.25, respectively. All failure types were noted as a
measurement of feet length.

68
Table 13: List of Old Crack Seal Project Sites Inspected during Field Visits

Year of
S.
Site City/County Project Location Crack
No.
Sealing
ID

1 F City of Duluth Northridge Drive 2014

2 H City of Duluth Hickory Street 2014

3 J City of Duluth E Palm Street 2014

4 I City of Duluth S Blackman Street 2014

5 A City of Duluth 3rd Street-21st AE to 1st AE 2014

6 B City of Duluth 2nd Street -24th AE to Mesaba Avenue 2014

7 D City of Duluth Sundby Road-Maple Grove to Haines Road 2015

8 E City of Duluth Swan Lk Road-Arrowhead to Basswood 2015

9 C City of Duluth 1st Street-21AE-8th AE 2015

10 L City of Duluth 1st Street-24 to 26AE 2016

11 K City of Duluth 2nd Street-24-26AE 2016

12 M City of Duluth 3rd Street-24-26AE 2016

13 Q City of Duluth 43AE Glenwood to Superior Street 2016

14 P City of Duluth 45AE Glenwood to Superior Street 2016

15 O City of Duluth 47AE Glenwood to Superior Street 2016

16 N City of Duluth 52AE Oakley to Superior Street 2016

G1 Arrowhead Road-Kenwood AV to Rice Lake

2016
17 City of Duluth Road (EB)

G2 Arrowhead Rd.-Kenwood AV to Rice Lake Road

2016
18 City of Duluth (WB)

69
 V1 City of
19 Hutchinson Segment 1: Hwy 22 to 1145 Michigan Street 2012

V2 City of Segment 2: 1145 Michigan Street to 5th Avenue

20 Hutchinson SE 2012

X City of
21 Hutchinson 1335 Jefferson Street SE 2017

22 Y City of Andover South coon creek Drive NW 2014

23 Z City of Andover Vale Street NW 2017

24 W City of Andover Wintergreen Street NW 2017

Note1: All these crack sealing treatments are rout-and-seal

(i) Stone Intrusion (ii) Pull-out Failure

(iii) Partial Cohesion Loss (iv) Full Cohesion Loss

Figure 4.37: Additional Seal Failure Types Observed in Old Crack Sealing Projects

70
4.4.1 Performance Analysis for Old Sites

Figure 4.38 shows photographs of conditions of a crack sealed in 2014 (Site J in Duluth) during the
three different seasons mentioned above. Such photographs were also taken for all other cracks during
each visit. PI was calculated using the documented information on the type and quantity of failure. Figure
4.39 shows the relation between PI and seal age for the sites sealed in 2014. It can be seen that the PI
slightly decreased with time in all the six sites.

PI–93.75 PI– 93.75 PI–91.25

Spring’18 Fall’18 Spring’19

Figure 4.38: Site J- E Palm St., Duluth; Crack Sealing Year - 2014

71
 100
Performance Index (PI) 90
80
F
70
60 H
50 J
40
I
30
20 A
10 B
0
0 10 20 30 40 50 60
Age (months)

Figure 4.39: Performance Evaluations of 2014-year Crack sealing Projects

Figure 4.40 presents photographs of a crack sealed in the year 2015; a total of three sites were
considered for the year 2015. Figure 4.41 shows that the PI of the two sites (Sites C and D) decreased by
approximately 20%, and the PI in the other site significantly dropped to 10%.

72
PI–87.50 PI– 87.50 PI–66.25

Spring’18 Fall’18 Spring’19

Figure 4.40: Site D- Sundby Rd., Duluth; Crack Sealing Year - 2015

100
Performance Index (PI)

80
60 D
E
40
C
20
0
0 10 20 30 40 50 60
Age (months)

Figure 4.41: Performance Evaluations of 2015-year Crack sealing Projects

73
 Figure 4.42 shows photographs of a crack sealed in the year 2016; a total of nine projects were
considered for this year. Figure 4.43 shows that two of the nine sites experienced a decrease of
approximately 60% of PI until the spring of 2019, while the other seven experienced variable but a lower
drop in the PI. The PI of the Sites P and Q significantly dropped in the winter of 2018 – 2019.

PI–90.00 PI-90.00 PI–52.50

Spring’18 Fall’18 Spring’19

Figure 4.42: Site Q- 43AE Glenwood to Superior St., Duluth; Crack Sealing Year – 2016

100
Performance Index (PI)

80

60

40
L K
M Q
20
P O
0
0 10 20 30 40 50 60
Age (months)

Figure 4.43: Performance Evaluations of 2016-year Crack sealing Projects

74
4.4.2 Performance Summary of Old Crack Seal Sites

Based on the sealant installation year and the last field visit year (Spring 2019), the ages of seals
in the old crack seal sites are approximately 1 (sealed in 2017), 2, 3, 4, or 6 (sealed in 2012) years. Figure
4.44 shows the average PI for these projects with respect to the year of sealant installation. Note that due
to the insufficient number of data points for the year 2012, the standard deviation was not determined.
The PI vs age of seals appeared to be following a good and declining trend, except for the seals those were
installed in 2014. The exact reason for this anomaly for the 2014 seals is unknown. However, the traffic
volume of the roads (mainly residential streets) on which they belong was very low compared to the other
roads.

120
Average Performance Index (PI)

100

80

60

40 3 yr. 4 yr.
1 yr. 2 yr.
20 6 yr.

0
2017 2016 2015 2014 2012

Year of Sealant Installation

Figure 4.44: Yearly Average Performance Index (PI): Rout-and-seal Projects

4.4.2.1 Influence of Pavement Shoulder Type on the Crack Sealant

In order to understand the influence of the shoulder type on the performance of the sealant, a
crack sealing site was visited in the city of Hutchinson where a road was sealed with rout-and-seal. The
road had two segments, mainly differing on the shoulder type: (i) Segment 1- gravel shoulder (rural), and
(ii) Segment 2- paved shoulder (urban) (Figure 4.45 and Figure 4.46). Even though both the segments
received crack sealing treatment in the same year 2012, the rout-and-seal on rural segment showed
complete failure; whereas, the urban segment was still in good condition when visited in 2018. Figure 4.47
shows the difference in performance index between these two segments. The City (John Olson, Personal
Communication) believed that the performance difference could be primarily due to the improper

75
drainage in Segment 1, where moisture could easily infiltrate through the gravel shoulder. Also, the 10-
inch thick single full-depth asphalt pavement in Segment 2, as opposed to the 9-inch thick three-layer
asphalt pavement structure (1.5-inch wear +1.5-inch binder + 6.0-inch base), could play a role too.

PI: 5 PI: 80

R&S (2012) R&S (2012)

Figure 4.45: Segment 1-Hwy 22 to 1145 Michigan Figure 4.46: Segment 2-1145 Michigan Street to 5th
Street, Hutchinson Ave SE, Hutchinson

100

80
80
Performance Index (PI)

60

40

20
5
0
Section 1 Section 2

Figure 4.47: Performance Index of the Rural and Urban Segments – Crack Sealed in the Same Year (2012)

76
CHAPTER 5: EFFECTIVENESS OF CRACK SEALING METHODS
5.1 PERFORMANCE-EFFECTIVENESS

5.1.1 Benefit

The performance-effectiveness of the crack seal was studied in terms of “benefit,” which was
derived from the area of the plot between PI and seal age. As discussed in the previous section, such plot
was drawn for each crack considered in the field study. Based on the performance trend of all the crack
seal project sites considered in this study and practitioners’ opinions on the seal performance, it was
decided that a minimum threshold PI of 50 would be reasonable to consider as the failure of the seal. This
threshold PI value indicates that 50% of the seal, mainly on the wheel paths, had failed either by one or a
combination of typical seal failures (cohesion, adhesion, etc.). The failed seals become ineffective in
resisting water infiltration. The pavement structure nearby the failed seal is likely to experience moisture
damages and result in secondary distresses. Thus the entire area of the PI vs crack seal age plot was not
used, the area up to the threshold PI value was only used for determining the performance-effectiveness
of the crack seal. See Figure 5.1. The following subsections discuss the procedures established for
determining the benefits for the rout-and-seal and clean-and-seals. As the seasons (especially winters)
between the years are usually not identical, for example, the 2018-2019 winter was harsher than 2017-
2018, the influence of the seasons on the crack seals differ from year to year. For this reason, benefits
were estimated separately, (i) once before the second winter, and (ii) then at the end of the second winter.

5.1.1.1 Rout-and-Seal

Figure 5.1 showed an example of PI vs. crack seal age relationship used for determining the
benefit; this particular plot was drawn for Site D (sealed with rout-and-seal) after the first winter. The PI
of all the eight cracks in Site D was averaged in this figure. The blue line in this figure represents the PI
values measured during the first year of service and the red line shows the forecasted PI beyond the first
year of service. As the seal performance data for the newly installed sites was not available until the
threshold PI (50), the PI for all these sites had to be forecasted for the period for which data was not
available. The trends observed in the relationship of the PI vs. age of crack seal, determined based on the
performance data of all the new and old seal projects as shown in Figure 5.2, were used for forecasting
the PI. The orange line represents the threshold PI. The area within these lines and the vertical axis marked
grey in Figure 5.1, represents the benefit of the crack seals. Such benefit plots were drawn for all the sites
after the first winter, and also after the second winter (discussed later). The PI values of all cracks for each
site were averaged together to make one representative plot for each site. Figure 5.3 shows the PI vs crack
seal age for all rout-and-seal sites; the first two plots ((a) and (b)) are for the data collected before the
second winter. The thick black line in Figure 5.3 (a) represents the average value of all sites combined. The
forecasted service lives (based on the data until the first winter) for different projects and the respective
benefit areas are listed in Table 14. The average benefit for all the rout-and-seal sites was 98. This table
also includes the calculated areas for the sites sealed with the clean-and-seal method. Note that Table 14
includes all the sites, including Site A and C, which were treated with micro-surfacing after crack sealant
application and not available to monitor the performance. Out of all these sites, only the Sites M, N and
O have both rout-and-seal and clean-and-seal methods.

77
 PI vs seal age trend line

Figure 5.1: PI vs Age of Crack Seal and Benefit Area Calculation (end of first winter)

78
 120

100
Performance Index (PI)
80 y = -11.5x + 98.9
R² = 0.95
60

40 Yr. 2015
Yr. 2016
20 Yr.2017 Yr. 2014
Yr. 2012
0
1 2 3 4 6
Age of Seal, Year(s)

Figure 5.2: Average PI vs. Age of Crack Seal

Table 14: Forecasting of Crack Seal Performance upon First Winter Evaluations

Site ID Seal Type Service Life (Age @ 50% Benefit Area

Threshold PI) (years)

A C-S under micro surface NA NA

C C-S under micro surface NA NA

D R-S 4.12 103

E R-S 3.98 90

F R-S 3.26 62

G R-S 2.79 49

79
 H R-S 4.92 135

I R-S 5.15 148

M R-S NA NA

M C-S NA NA

N R-S 3.69 82

N C-S 2.72 62

O R-S 5.35 135

O C-S 4.31 131

Average Rout-and-Seal Benefit Area 98

Average Clean-and-Seal Benefit Area 86

Note: C-S = Clean-and-seal; R-S = Rout-and-seal; NA = not available

Figure 5.3 (c) and (d) shows the PI vs. crack seal age for all the data collected for all rout-and-seal
sites until the second winter (including early spring). The figures indicate that the benefit and forecasted
service lives did not follow the same trend as anticipated before collecting the data in the second winter.
The cracks experienced larger failure than anticipated and expected to last shorter than forecasted; the
severely cold temperature during the 2018-2019 winter could have played a great role in this. The thick
black line in Figure 5.3 (c) represents the average value of all sites combined.

Table 15 shows the re-calculated service lives and benefits based on the data collected after the second
winter. The average benefit was found as 75.

80
 (a) All Rout-and-seal Sites (b) Avg. Benefit of all Rout-and-seal Sites (First
Winter)
(First Winter)

(c) All Sites (Rout-and-seal) (d) Avg. Benefit of all Rout-and-seal Sites (Second
Winter)
(Second Winter)

Figure 5.3: Performance Evaluations and Average Benefit Area of Rout-and-Seal Projects

81
Table 15: Forecasting of Crack Seal Performance after Second Winter Evaluations

Site ID Seal Type Service Life (Age @ 50% Benefit Area

Threshold PI) (years)

A C-S under micro surface NA NA

C C-S under micro surface NA NA

D R-S NA 79

E R-S 3.29 67

F R-S 2.30 46

G R-S 1.37 34

H R-S NA 67

I R-S 5.73 177

M R-S NA 8

M C-S NA 6

N R-S 1.62 55

N C-S 2.22 59

O R-S NA 167

O C-S NA 103

82
 Average Rout-and-Seal Benefit Area 78

Average Clean-and-Seal Benefit Area 72

Table 16 provides a list of the age of seal with respect to different PI values of the first winter
evaluations, determined by using the correlation shown in Figure 5.2. Based on a 50% threshold PI, the
service life of rout-and-seal is roughly four years. It is noted that the age corresponding to the 50%
threshold PI was referred to as the service life. Figure 5.4 shows the probability density function for the
service life for the rout-and-seal projects. This figure can estimate the service life of the projects at a
different level of reliabilities. Table 17 shows that the average service life of rout-and-seal is 4.2 years with
a reliability of 50% and it is 2.3 years with a reliability of 97.5%. Note that the benefit and the life cycle
cost analysis were performed for the service life with a 50% reliability.

Table 16: Estimated PI vs Service Life for Rout-and-Seal after First Winter

PI Age of Seal (year)

90 0.8

80 1.6

70 2.5

60 3.4

50 4.2

40 5.1

30 6.0

20 6.8

10 7.7

83
 Figure 5.4: Probability Distribution of Rout-and-Seal Service life

Table 17: Rout-and-Seal Service Life and Reliability after First Winter

Service life Reliability

2.3 97.5%

3.2 84%

4.2 50%

5.0 16%

6 2.5%

Upon documenting seal performance during the second winter, a new probability density function was
created. This new function shows that seal performance, as stated earlier, deteriorated more quickly on
these sites that had been anticipated based on the correlation of older crack seal sites. In order to create
this curve, sites I and O had to be excluded due to their much higher PI values. These high PI values are

84
associated with the tight crack spacing and low volumes of traffic seen at these sites, and they were
considered as the outlier as a result.

Figure 5.5: Probability Distribution of Rout-and-Seal Service Life after Second Winter Analysis

Table 18: Rout-and-Seal Service Life and Reliability after Second Winter

Service life Reliability

1.27 97.5%

2.19 84%

3.10 50%

4.02 16%

4.94 2.5%

85
5.1.1.2 Clean-and-Seal

Due to a lack of sufficient data for the clean-and-seal method, the PI vs. age of seal trend line for
this method was obtained indirectly by adjusting the trend line developed for the rout-and-seal method
(Figure 5.2). The main assumption of this adjustment was that the clean-and-seal method performs about
one year less than the rout-and-seal method as suggested by many previous studies (Table 21). Figure 5.6
shows a schematic indicating the adjustment made for the clean-and-seal method. The ‘corrected area,’
shaded in grey, refers the area for the clean-and-seal benefit area. The ‘uncorrected area,’ shaded in red,
refers to the difference in the benefits area between the rout-and-seal and clean-and-seal methods. For
this particular example, the clean-and-seal method provides 21.6% less benefit than what is provided by
the rout-and-seal method.

Using the previously mentioned procedure for the clean-and-seal method, the PI vs age of seal areas were
plotted for the sites that have PI data (8 sites) for the data until before the second winter, excluding the
Sites A, C and M. Figure 5.7 shows the PI vs age of seal for all these eight sites. The wider black line
represents the average value of all eight sites combined. Figure 5.8 shows the average benefit area for all
the eight sites combined, with a threshold PI of 50; the average service life for the clean-and-seal method
is 3.12 years with a benefit area equal to 86.

Figure 5.9 shows the probability density function for the service life for the clean-and-seal
projects. Table 19 shows that the average service life of clean-and-seal is 3.2 years with a reliability of 50%
and it is 1.3 years with a reliability of 97.5%.

86
 Difference = 21.6%

Figure 5.6: Clean-and-Seal Benefit Area Determination (end of the first winter)

87
Figure 5.7: Service Life of Clean-and-Seal method for Different Sites after First Winter

Figure 5.8: Average Benefit Area for Clean-and-Seal method (end of the first winter)

88
 Figure 5.9: Probability Density for Clean-and-Seal Service Life after First Winer Analysis

Table 19: Clean-and-Seal Service Life and Reliability after First Winter

Service Life Reliability

1.3 97.5%

2.4 84%

3.2 50%

4.1 16%

5 2.5%

By making the assumption that a clean-and-seal will last one year less than rout-and-seal to meet
its threshold performance index, Figure 5.10 is drawn to determine the benefit for the clean-and-seal
method including the data collected after the second winter. The expected service life probabilities for
clean-and-seal, based on the second winter data, is shown in Table 20 below. Based on the second winter

89
data, it appears that the service life of the clean-and-seal method could be approximately 2 years with
50% reliability.

Figure 5.10: Service Life of Clean-and-Seal method for Different Sites after the Second Winter

Table 20: Clean-and-Seal Service Life and Reliability after Second Winter

Service life Reliability

0.3 97.5%

1.2 84%

2.1 50%

3.1 16%

3.9 2.5%

90
5.1.1.3 Summary of Performance-effectiveness

Based on the second winter crack seal performance trend, it can be stated that the second winter
has caused a significant amount of failures of the crack seals irrespective of the sealing method. The
severely cold 2018-2019 winter has expanded many cracks beyond the typical winter average crack
widths, which stretched sealants beyond their allowable strains. Also, in many sites, new thermal cracks
have developed in between the previously recorded cracks (e.g., Site M). As this kind of aggressive winter
may not be typical, it is assumed that the service lives of crack seal methods determined based on the
first winter data from the newly installed crack seals (Phase 3) and old crack seal data (Phase 4, includes
data of the seals that experienced 1 to 6 winters) may be considered as more reasonable as opposed to
what was determined based on the second winter data alone (unusual winter). Therefore, the average
service life for the rout-and-seal and clean-and-seal are finalized as four years and three years,
respectively. The life-cycle cost analysis in this study was thereby performed using the above-mentioned
service lives.

The suggested average service lives for the two methods are also comparable with the service
lives obtained in other states of similar climate to Minnesota (Figure 5.11 and Table 21), including Al Quadi
et al. (2017) study that compared the performances of rout-and-seal and clean-and-seal methods. Note
that the 6-year period corresponding to the MnDOT construction log (Figure 5.11) is actually the service
period, not necessarily the service life. This only means that MnDOT usually applied a surface treatment
on the pavements on average 6-year after the crack sealant installation.

91
 10 Rout-and-seal Clean-and-seal
9
8
7
Service Life/Service Period, Years
7
6 6
6
5 5
5
4
4
3 3
3
2
1
0
Online Survey MnDOT Field Other States
Construction Investigations
Log

Data Source

Figure 5.11: Service Life/ Service Period of Crack Seals from Different Sources

Table 21: Service Life of Crack Seals Observed in Different USA States and Canadian Provinces

Method Crack Sealing/ Crack Filling/

State or Province Rout-and-Seal Clean-and-Seal

Minnesota (MnDOT, 2018a) 2 to 4 yrs 1 to 3 yrs

Illinois (IL DOT, 2010) 2 to 8 yrs 2 to 4 yrs

2 to 4 yrs 2 to 4 yrs

Colorado (Galehouse, 2004) (based on traffic levels) (based on traffic levels)

Virginia

(Al-Quadi et al., 2017) 2-3 yrs 1-2 yrs

Ontario, Ca (OHMPA, 2004) 3 to 5 yrs N/A

Alberta, Ca (AL Trans, 2003) 5 to 8 yrs 1 to 3 yrs

92
 Average Range 2-8 yrs 1-4 yrs

Average 5 yrs 3 yrs

5.2 COST-EFFECTIVENESS

Cost-effectiveness was established by determining the benefit to cost (B/C) ratio of each crack
sealing method. The B/C ratio analysis was initially performed for the crack sealing treatment cost alone,
then for all the costs incurred during the life cycle (analysis period) of the pavement. The average benefit
determined in the previous subsection was considered as the “benefit” in the B/C ratio analysis and it is
noted here that the “benefit” in this analysis is not a dollar amount. The following subsection discusses
different components of the B/C analysis performed in this study.

5.2.1 Unit Cost of Crack Sealing

The unit cost of rout-and-seal crack sealing was obtained from the Life-Cycle Cost Analysis (LCCA)
Spreadsheet available at the MnDOT website for the MnDOT District-1. Since the MnDOT’s LCCA
Spreadsheet does not provide unit cost for specific crack sealing treatment type, the given crack treatment
cost was assumed as the unit cost of rout-and-seal treatment. The cost of clean-and-seal treatment was
then determined using some representative bid letting abstracts from the MnDOT website. These bid
letting abstracts included both rout-and-seal and clean-and-seal methods. Using the unit prices available
in those bid letting abstracts, the year of the bid, MnDOT’s suggested discounted rate, and the present
year (2018) unit costs were calculated for the all the bids. The following equation was used to determine
the older year unit cost to the present year (2018) unit cost:

Present cost, Cp = Cu*(1 + 𝑖)𝑛 (5)

Where,

Cu=Unit Cost in the year of treatment

i= discount rate (1.22%, collected from MnDOT District-1’s LCCA spreadsheet)

n = the time in years until 2018

Table 22 shows the details of the present year unit cost along with the other information used. The present
year’s unit costs for all the rout-and-seal and clean-and-seal projects were then averaged, separately. The
average cost for the rout-and-seal and clean-and-seal methods were $101.3 and $55.4, per road station
(RDST). The ratio of the average costs of rout-and-seal to the clean-and-seal unit was found to be 1.8. This
ratio was used to determine the final unit cost for the clean-and-seal method. Table 23 provides the cost
of the crack sealing treatment per lane mile for both rout-and-seal and clean-and-seal methods.

93
Table 22: Unit Cost of Crack Sealing Treatments Obtained from Previous Bid Letting Abstracts

Project SP # Unit Unit Price n Discount Rate, i Present

Year Cost ($)
($) (up to
2018)

Rout-and-seal

2006 4508-26 RDST 51 12 0.0122 59.0

2009 8825-336 RDST 159 9 0.0122 177.3

2006 8827-68 RDST 65 12 0.0122 75.2

2007 3404-54 RDST 82.16 11 0.0122 93.9

Average Unit Cost per Road Station (RDST), $ 101.3

Clean-and-seal

2007 2001-33 RDST 35 11 0.0122 40.0

2008 2801-86 RDST 62.8 10 0.0122 70.9

Average Unit Cost per Road Station (RDST), $ 55.4

Cost ratio: Rout-and-seal to Clean-and-seal (=101.3/55.4) 1.8

94
Table 23: Crack Sealing Unit Cost per Lane Mile

Length Width Quantity per Unit Price Total Price

(Mile) (ft) lane mile Unit ($) ($)

Rout-and-seal 1 12 7040 SY 0.46 3,238.4

Clean-and-seal 1 12 7040 SY 0.26 1799.1

5.2.2 Pavement Life-Cycle Cost (LCC)

5.2.2.1 Initial Construction Cost

In order to determine the initial construction cost, a typical and stout pavement cross-section was
selected from the MnDOT’s pavement design guidelines (MnDOT, 2018b). Figure 5.12 and Figure 5.13
show the cross-section and layer thickness details of that typical section. Three pavement configurations,
such as one New HMA and two HMA Overlays with 13 and 20 years of design lives, were considered for
this analysis. Different milling and wearing course depths were used for HMA Overlay with varying design
lives. Table 24 to Table 26 show the details of the different items and their respective initial construction
costs for the three configurations. The quantities of each of these items were determined using MnDOT
District-1’s LCCA spreadsheet.

Figure 5.12: A Typical MnDOT Pavement Cross Section (after MnDOT, 2018b)

95
Figure 5.13: Typical MnDOT Pavement Structure Thickness Selection Guideline (after MnDOT, 2018b)

96
Table 24: Initial Construction Cost Per Lane-Mile, New HMA, Design Life = 20 Years

Depth (in) Width (ft) RDST Quantity per Unit Unit Price Total Price
mile ($)
($)

Subgrade Preparation 53 53 RDST 102.35 5,424

(RDST)

Select Granular 14 12 2737.8 CY 21.15 57,905

Embankment MOD 7%
(CV)

Aggregate Base (CV) 10 12 1955.6 CY 31.41 61,425

Class 5

Non-Wearing Course 4 12 1591 TON-B 62.75 99,835

Mixture (4, B)

Wearing Course Mixture 4 12 1591 TON-B 82.33 130,994

(4,C)

Grand Total 355,584

Table 25: Initial Construction Cost Per Lane-Mile, HMA Overlay, Design Life = 13 Years

Depth Width Quantity Unit Unit Total Price

(in) (ft) per mile Price ($)
($)

Mill Bituminous Surface 1.5 12 7040 SY 0.81 5,702

97
 Wearing Course Mixture 3 12 1392.2 TON- 77.58 92,571
B
(4, B)

Grand Total 98,274

Table 26: Initial Construction Cost Per Lane-Mile, HMA Overlay, Design Life = 20 Years

Depth Width Quantity per Unit Unit Price Total Price

(in) (ft) mile ($)
($)

Mill Bituminous Surface 2 12 7040 SY 1.22 8,589

Wearing Course Mixture (4,B) 4.5 12 1789.9 TON-B 72.85 130,394

Grand Total 139,983

5.2.2.2 Maintenance and Rehabilitation Activities and Costs

The maintenance and rehabilitation activity schedules were initially collected from the MnDOT
Pavement design manual (MnDOT, 2018b) for four different cases as follows: (i) Case 1: New HMA with
20-year design life and 35-year analysis period; (ii) Case 2: New HMA with 20- year design life and 50-year
analysis period; (iii) Case 3: HMA Overlay with 13-year design life and 35-year analysis period; and (iv)
Case 4: HMA Overlay with 20-year design life and 35-year analysis period. These schedules consider a
four-year service period between crack sealant installation and the follow-up surface treatment, which is
similar to the service life of the rout-and-seal treatment, determined in the previous section of this report.
The service life of clean-and-seal was determined as three years. As the clean-and-seal performs one year
less than four years, a re-sealing activity was considered for the remaining one-year service life; it was
assumed that 30% of the failed clean-and-seal seals would be re-sealed after three years of the service
period.

The relevant unit costs of maintenance and rehabilitation activities (other than sealing costs) were
obtained from the MnDOT District-1’s LCCA spreadsheet. The Net Present Values (NPV) of different items
were calculated using the following equation:
𝐶
Net Present Value (NPV) = 𝐶𝑖𝑛 + ∑𝑁 𝑎𝑡
𝑡=1 (1+𝑖)𝑡 (6)

98
Where,

Cin = initial cost of construction

t = the time in future in years

N = the number of years in the analysis period

Cat = the cost of activity t years in the future

i = the annual rate of interest in decimals (1.22%, collected from the MnDOT District-1’s LCCA
spreadsheet)

Table 27 to Table 30 show the details of maintenance and rehabilitation schedules and the
corresponding initial and net present value cost information for different cases considered. It can be seen
that the life cycle cost for the rout-and-seal method is slightly higher than the clean-and-seal method for
all the cases.

Table 27: Example LCCA per lane mile for a Typical New HMA with 20 years of design life and 35 years Analysis
Period

Rout-and-seal Clean-and-seal

Initial Cost ($) Age, NPV ($) Initial Cost ($) Age NPV ($)
Year

Initial Construction 355,584 0 355,584 355,584 0 355,584

Crack Treatment 3,238 8 2,939 1,799 8 1,633

Crack Re-sealing - None - 600 11 525

Surface Treatment (UTBWC) 50,688 12 43,824 50,688 12 43,824

Mill 1.5" and HMA Overlay 123,739 20 97,091 123,739 20 97,091

3.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 23 2,450 1,799 23 1,361

99
 Crack Re-sealing - - - 600 26 438

Surface Treatment (UTBWC) 50,688 27 36,535 50,688 27 36,535

End of Analysis period (2/17 14,557 35 9,523 14,557 35 9,523

Remaining Life)

Life Cycle Cost $ 528,900 $ 527,467

Table 28: Example LCCA per lane mile for a Typical New HMA with 20 years of design life and 50 years Analysis
Period

Rout-and-seal Clean-and-seal

Initial Cost Age, NPV ($) Initial Cost Age, NPV ($)
($) Year ($) Year

Initial Construction 355,584 0 355,584 355,584 0 355,584

Crack Treatment 3,238 8 2,939 1,799 8 1,633

Crack Re-sealing - - - 600 11 525

Surface Treatment (UTBWC) 50,688 12 43,824 50,688 12 43,824

Mill 1.5" and HMA Overlay 123,739 20 97,091 123,739 20 97,091

3.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 23 2,450 1,799 23 1,361

Crack Re-sealing - - - 600 26 438

100
Surface Treatment (UTBWC) 50,688 27 36,535 50,688 27 36,535

Mill 1.5" and HMA Overlay 123,739 37 79,004 123,739 37 79,004

3.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 40 1,994 1,799 40 1,108

Crack Re-sealing - - - 600 43 356

Surface Treatment (Chip Seal) 15,136 44 8,878 15,136 44 8,878

End of Analysis period (4/17 29,115 50 15,878 29,115 50 15,878

Remaining Life)

Life Cycle Cost $ 612,421 $ 610,458

101
Table 29: Example LCCA per lane mile for a Typical HMA Overlay with 13 years design life and 35 years Analysis
Period

Rout-and-seal Clean-and-seal

Initial Cost Age, NPV ($) Initial Cost Age, NPV ($)
($) Year ($) Year

Mill 1.5" and HMA Overlay 98,274 0 98,274 98,274 0 98,274

3.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 3 3,123 1,799 3 1,735

Crack Re-sealing - - - 600 6 558

Surface Treatment (Chip 15,136 7 13,904 15,136 7 13,904

Seal)

Mill 1.5" and HMA Overlay 98,274 13 83,941 98,274 13 83,941

3.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 16 2,667 1,799 16 1,482

Crack Re-sealing - - - 600 19 476

Surface Treatment (Chip 15,136 20 11,876 15,136 20 11,876

Seal)

Mill 1.5" and HMA Overlay 98,274 25 72,574 98,274 25 72,574

3.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 28 2,306 1,799 28 1,281

102
 Crack Re-sealing - - - 600 31 412

Surface Treatment (Chip 15,136 32 10,268 15,136 32 10,268

Seal)

End of Analysis period (3/17 17,342 35 11,345 17,342 35 11,345

Remaining Life)

Life Cycle Cost $ 287,588 $ 285,436

Table 30: Example LCCA per lane mile for a Typical HMA Overlay with 20 years design life and 35 years Analysis
Period

Rout-and-seal Clean-and-seal

Initial Cost Age, NPV ($) Initial Cost Age, NPV ($)
($) Year ($) Year

Mill 2" and HMA Overlay 138,983 0 138,983 138,983 0 138,983

4.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 3 3,123 1,799 3 1,735

Crack Re-sealing - - - 600 6 558

103
 Surface Treatment (UTBWC) 50,688 7 46,563 50,688 7 46,563

Mill 2" and HMA Overlay 138,983 20 109,052 138,983 20 109,052

4.5"
(9.5 Wearing - 4,B)

Crack Treatment 3,238 23 2,450 1,799 23 1,361

Crack Re-sealing - - - 600 26 438

Surface Treatment (UTBWC) 50,688 27 36,535 50,688 27 36,535

End of Analysis period (5/17 40,877 35 26,740 40,877 35 26,740

Remaining Life)

Life Cycle Cost $ 309,967 $ 308,485

5.3 THE BENEFIT TO COST (B/C) RATIO

The performance-effectiveness and cost-effectiveness were determined using the benefit to cost
ratio. First, the ‘benefit to treatment-cost ratio’ of the two sealing methods was compared. It can be seen
in

Table 31 and Figure 5.14 that the clean-and-seal method provides greater ‘benefit to treatment-
cost ratio.’ This indicates that if the only short-term benefit is considered, then the clean-and-seal may be
more cost-effective than its counterparts. Second, the ‘benefit to life-cycle-cost ratio’ of the two sealing
methods was compared. As shown in

Table 32 and Figure 5.15, it can be seen that when ‘benefit to life-cycle-cost ratio’ was compared
between the two sealing methods, the rout-and-seal method provides slightly more cost-effectiveness for
all the four different cases considered in this study. However, as the difference is very small, it would be
worth to look at other decision factors for recommending the crack seal method for a specific job.

104
Table 31: Benefit/Treatment Cost Ratio of Rout-and-Seal and Clean-and-Seal Methods

Benefit Cost per lane-mile B/C ratio

(1,000 $)

Rout-and-seal 98 $ 3.238 30.26

Clean-and-seal 86 $ 1.799 47.80

B / C RAT IO : T RE AT ME NT CO S T

47.80
30.26

R &S C &S

Figure 5.14: Benefit/Treatment Cost Ratio of Rout-and-Seal and Clean-and-Seal (Note- ‘benefit’ here is not in
terms of the dollar, so B/C ratio shall not be compared with unity (one))

Table 32: Benefit/ Life-Cycle-Cost Ratio of Rout-and-Seal and Clean-and-Seal Methods

LCC Benefit B/C(1000$)

Case R&S C&S R&S C&S R&S C&S

Case 1 $528,900 $527,467 98 86 0.185 0.1630

105
Case 2 $612,421 $610,458 98 86 0.160 0.1409

Case 3 $287,588 $285,436 98 86 0.341 0.3013

Case 4 $309,967 $308,485 98 86 0.316 0.2788

B / C RAT IO:L IFE CYCL E COS T (L CC)

R&S C&S

0.341

0.3013

0.316

0.2788
0.185

0.1630

0.160

0.1409

C AS E 1 C AS E 2 C AS E 3 C AS E 4

Figure 5.15: Benefit/Cost Ratio for Rout-and-Seal and Clean-and-Seal for Different Cases

106
CHAPTER 6: DRAFT RECOMMENDATION FOR THE SELECTION OF
APPROPRIATE CRACK SEALING METHOD
In this section, a draft recommendation was developed for selecting the appropriate crack sealing
method for asphalt pavement cracks. Easy-to-use decision trees have been developed to aid in deciding
whether to seal with clean-and-seal or rout-and-seal method based on many relevant variables, such as
crack severity, pavement type and age, traffic level, and subgrade soil type.

6.1 VARIABLES

6.1.1 Crack Severity and Crack Width

The crack severity largely influences the selection of the crack sealing method. The severity of crack
is defined with respect to the width of the crack (CW). While either of the rout-and-seal or clean-and-seal
method may be implemented on the low and moderate severity cracks depending on the other variables,
for the high severity cracks (wider than ¾ -inch), the crack width may be too wide for the router to reach
both sides of the crack during the routing operation. Therefore, the clean-and-seal method is more
appropriate for high severity cracks.

6.1.2 Pavement Type and Age

The type of the pavement, whether it is new construction or an overlay, play a role in the crack treatment
schedule and the selection of the crack sealing method. The age of the pavement, depending on the
sealing is intended for the initial years of the service or for towards the end, can influence the selection
of the crack sealing method. Table 33 shows a couple of examples of MnDOT’s crack treatment schedules
for new constructions and overlays with varying design life and analysis period (MnDOT, 2018a). In Table
33, it can be seen that the new pavement with a 20-year design life, with a 35-year analysis period will
receive crack treatments twice - at the 8-year and 23-year mark. Then pavement with a 50-year analysis
period, will receive crack treatments thrice - at 8-, 23-, and 40-year mark. A pavement overlay with a 13-
year design life and a 35-year analysis period receive crack treatments thrice - at 3-, 16-, and 28-year mark.
A pavement overlay with an 18-year design life and a 35-year analysis period receive crack treatments
twice - at 3- and 21-year mark. These different pavement ages or crack treatment numbers (e.g., first,
second and third treatments), and the pavement type (new vs. overlay) influence the choice of the crack
selection method.

Table 33. Crack Treatment Schedule (MnDOT 2018a)

(a) New HMA, Design life 20 years

35 Year Analysis Period 50 Year Analysis Period

Pavement Age Treatment Treatment

0 Initial Construction Initial Construction

107
 8 Crack Treatment Crack Treatment

12 Surface Treatment Surface Treatment

20 Mill & Overlay (1st Overlay) Mill & Overlay (1st Overlay)

23 Crack Treatment Crack Treatment

27 Surface Treatment Surface Treatment

35 End of Analysis Period

37 Mill & Overlay (2nd Overlay)

40 Crack Treatment

44 Surface Treatment

50 End of Analysis Period

(b) Overlay, Design life 13-17 years

Pavement Age Treatment

0 Initial Construction (1st Overlay)

3 Crack Treatment

7 Chip Seal

DL Mill & Overlay (2nd Overlay)

DL+3 Crack Treatment

DL+7 Chip Seal

2*DL-1 Mill & Overlay (3rd Overlay)

2*DL+2 Crack Treatment

2*DL+6 Chip Seal

35 End of Analysis Period

108
 (c) Overlay, Design life >17 years

Pavement Age Treatment

0 Initial Construction (1st Overlay)

3 Crack Treatment

7 Chip Seal

DL Mill & Overlay (2nd Overlay)

DL+3 Crack Treatment

DL+7 Chip Seal

35 End of Analysis Period

6.1.3 Traffic Level

Pavements are designed for different levels of traffic, from under a million to 30 million equivalent
standard axle loads (ESALs). According to MnDOT standard specifications for construction (2018), the road
with the traffic levels 2 (< 1 million ESALs) and 3 (1 to 3 million ESALs) are considered as the low volume
roads; whereas, roads with traffic levels 4 (3 to 10 million ESALs) and 5 (10 to 30 million ESALs) are
considered as the high volume roads. Even though a clear and convincing relationship between the service
period of crack seals and traffic volumes could not be established in this project (because of the limitation
of the data), it is believed that the traffic volume affects the life and performance of the crack seals. A
high volume road needs a superior performing sealing method, such as rout-and-seal but also depends on
other variables. It may be assumed that the higher the traffic, the greater the damage to crack seals.
Hence, for the intermediate crack treatment, it is reasonable to consider clean-and-seal for traffic levels
2 and 3 and rout-and-seal for traffic levels 4 and 5.

6.1.4 Soil /Subgrade Type

The moisture-holding behavior and frictional resistance offered by the subgrade soil depends on its
type (gravel, sand, silt, clay, etc.). The type of the soil, therefore, influences the seasonal movement of
crack width as well. In this project, it was found that some practitioners expressed better performance of
rout-and-seal for clayey and silty soils and clean-and-seal for sandy soils. Based on that observation, it is
reasonable to choose rout-and-seal for clayey and silty soil conditions, whereas the clean-and-seal for
sandy subgrade soils.

6.1.5 Cost factors

The initial crack treatment cost of the rout-and-seal method was found to be 1.8 times higher than
the clean-and-seal method due to the specialized equipment and additional time required for routing. In

109
today’s dwindling budget scenario, the crack treatment cost can play a significant role as well. Hence, it is
reasonable to consider the clean-and-seal method irrespective of other variables when a sufficient fund
is not available.

6.1.6 Practitioners’ Preference

The online survey conducted in this project revealed that many practitioners prefer the rout-and-
seal for sealing cracks at initial years of the service period, and the clean-and-seal for later years; this
practice is somewhat working as well. Keeping in view of this, it is reasonable to consider the rout-and-
seal for the first crack treatment and clean-and-seal for the last crack treatment in the analysis period.

6.2 CRACK SEALING SELECTION CRITERIA

As discussed above, the selection of the crack sealing method could be influenced by several
variables. Besides, these variables may not necessarily have an identical weightage. A decision worksheet
was developed to account for the non-uniformity in the weightages of these variables.

Table 34 shows the list of decision variables and their respective weightages. The weightage (%) of
the crack treatment cost, benefit/cost ratio, expected service life of the seal, ease of repair, practitioners’
preference, and traffic level were decided as 10, 15, 15, 10, 30 and 20 percentages, respectively. Although
the percent weightages of these decision variables are subjective, they were logically assumed based on
the experience gained in this project and reviewing different literature related to crack sealing practices.
As there was no significant difference found in the benefit/cost analysis, a relatively high weightage was
assigned to the practitioners’ opinion. Separate scores (%) were given to the two crack sealing methods.
The scores for some of the variables mentioned above were determined based on the findings from the
cost analysis performed in Section 5. For example, the crack treatment cost for rout-and-seal was 1.8
times more than the clean-and-seal; therefore, a score of 64 is assigned to the clean-and-seal, and a score
of 36 was assigned to the rout-and-seal (64:36 = 1.8:1). A similar procedure was followed for the scores
for the benefit/cost analysis as well. The typical service life of the rout-and-seal is generally four years
versus three years for the clean-and-seal method; the scores for the expected service lives for these
methods were thereby assigned based on this assumption. As there is no quantitative assessment
available for the remaining three variables listed in

Table 34, qualitative scores were assigned to them. The scores for the practitioners’ opinion and
traffic level were varied for the crack treatment numbers and traffic levels.

The weighted scores for each decision variable were calculated by multiplying the score with the
corresponding weight. The last two columns of

110
Table 34 provides the total scores for all the decision variables together and ranks for two crack sealing
methods for various traffic levels and crack treatment number.

Table 34. Decision worksheet for crack sealing selection criteria

Decision Variables Crack B/C Expected Ease of Performance Traffic Total Rank
Treatment Life Operation Opinion by Level Score
Cost Ratio* Practitioners

Weight 10 15 15 10 30 20

Alternatives

1st Treatment, All Traffic Levels

Rout- Score 36 53 57 33.3 60 50 1

and-
Seal Weighted 51.43
3.6 7.95 8.55 3.33 18 10
Score

Clean- Score 64 47 43 66.7 40 50 2

and-
Seal Weighted 48.57
6.4 7.05 6.45 6.67 12 10
Score

Intermediate, Traffic Levels 2 and 3

Score 36 53 57 33.3 50 40 46.43 2

111
 Rout- Weighted
and- Score 3.6 7.95 8.55 3.33 15 8
Seal

Clean- Score 64 47 43 66.7 50 60 1

and-
Seal Weighted 53.57
6.4 7.05 6.45 6.67 15 12
Score

Intermediate, Traffic Levels 3 and 4

Rout- Score 36 53 57 33.3 50 75 1

and-
Seal Weighted 53.43
3.6 7.95 8.55 3.33 15 15
Score

Clean- Score 64 47 43 66.7 50 25 2

and-
Seal Weighted 46.57
6.4 7.05 6.45 6.67 15 5
Score

Last Treatment, All Traffic Levels

Rout- Score 36 53 57 33.3 40 50 2

and-
Seal Weighted 45.43
3.6 7.95 8.55 3.33 12 10
Score

Clean- Score 64 47 43 67.7 60 50 1

and-
Seal Weighted 54.67
6.4 7.05 6.45 6.77 18 10
Score

*B/C Ratio: (PI, %) / (LCC, 1000$ per lane-mile)

112
6.2.1 Decision Trees

The rankings determined in the previous section was helpful in developing two decision trees that
can be used for the selection of the appropriate crack sealing method. The first decision tree is shown in
Figure 6.1. Using this decision tree, users can decide on the most appropriate crack selection method in
terms of the variables discussed in the previous sections. For example, for a low severity crack in a newly
constructed pavement with a 20-year design life and a 35-year analysis period, the crack sealing method
for first-time crack treatment would be the rout-and-seal. A clean-and-seal method would be appropriate
for a pavement with a sandy soil subgrade, and a rout-and-seal would be appropriate for the pavement
with clayey and silty soil subgrades, irrespective of other variables.

The second decision tree which is a simplified version of the first one is shown in Figure 6.2. This decision
tree can be useful when data on the decision variables are limited. As shown in Figure 6.2, a crack sealing
method can be decided based on the crack severity, traffic levels and crack treatment number (first,
intermediate or last). If the crack severity was high, a clean-and-seal method could be chosen. For low or
moderate crack severity, the choice depends on the number of crack treatment in the analysis period and
traffic level. If it is a first crack treatment for any traffic levels or the second crack treatment with traffic
levels 4 or 5, a rout-and-seal method is suggested. If it is the last crack treatment for any traffic levels or
the second crack treatment with traffic levels 2 or 3, a clean-and-seal method is suggested.

113
Figure 6.1. Decision tree for selection of crack sealing method

114
Figure 6.2. Simplified decision tree for selecting a crack sealing method

115
CHAPTER 7: SUMMARY AND CONCLUSIONS
Crack sealing is an important preventive treatment in a pavement preservation program. In order
to achieve a cost-effective crack seal, it is important to select a proper crack sealing method. While the
state of Minnesota usually seals cracks of asphalt pavements, there is no clear consensus on the most
appropriate crack sealing method for a specific job. The main goal of this study was to develop a guideline
so that an effective crack sealing method can be chosen based on the factors that influence the
performance of crack seals, such as, traffic level, pavement age, crack severity, crack sealing occurrence
number (e.g. sealing the pavement for the first time, 2nd time or 3rd time), etc. The following major tasks
were performed in this project: (i) literature review on the crack sealant installation practices and sealant
performances, (ii) collection and analysis of field performance data (iii) performance- and cost-
effectiveness analysis, and (iv) development of a guideline.

Various literature on the crack sealant practices and performances, including research reports,
synthesis, journal articles, and other relevant publications were reviewed. The crack sealing performance
data was collected through an online survey and reviewing the history of crack sealing projects
documented in several construction data logs of MnDOT. Performance data were also collected through
periodical evaluations of crack seals at 35 different sites located throughout Minnesota. The performance
of the crack seals was studied by quantifying the performance index of the crack seals. The effectiveness
of the crack seals was studied with respect to the benefit-cost ratio analysis. Finally, two decision trees
were developed to help to select the most appropriate crack sealing method, one of which can be used in
the pavement management systems and the other can be used by the preventive maintenance crews.
The major specific conclusions drawn from different tasks of this study are listed below:

Online survey (Phase 1)

 It was found that the rout-and-seal (of transverse thermal cracks) is the most commonly practiced
crack sealing method in Minnesota. Out of the 47 responses received in the online survey, 68%
revealed that they use rout-and-seal method and the other 32% use clean-and-seal method.
 A mixed response was received regarding the service life (time of sealant installation to failure) of
rout-and-seal and clean-and-seal treatments. The opinions ranged from 2 to 10 years for clean-
and-seal and 2 to over 15 years for rout-and-seal.
 The most commonly reported criterion for selecting a sealing method was crack/pavement
conditions (41%) followed by pre-determined schedules (24%), no criteria (17%), pavement age
(13%), subgrade material (2%), and budget (2%).

MnDOT Construction log data (Phase 2)

 The average service periods (time of sealant installation to next maintenance or rehabilitation
work) of the rout-and-seal and clean-and-seal methods were found to be around 6.4 and 6.0
years, respectively. The difference in average service periods between the clean-and-seal and
rout-and-seal was found to be statistically insignificant (at a 95% confidence level). However, it
shall be mentioned here that the amount of clean-and-seal data was limited compared to the
rout-and-seal data.
 Average service periods of 6.9 and 6.4 years were found for the roads with lower traffic (<10,000
AADT) and 5.4 and 5.3 years of average service periods were found for the roads with traffic higher

116
 (>10,000 AADT) for rout-and-seal and clean-and-seal treatments, respectively. Though a higher
service period was found with lower traffic, the difference between the average service periods
of these two traffic categories was found to be statistically insignificant (at a 95% confidence), for
both methods.

New crack seal project sites (Phase 3)

 After the first-year evaluation (one summer and winter), the average performance indices for the
rout-and-seal and clean-and-seal projects were found to be 85 and 90, respectively. For the
second winter evaluations, the average performance indices for both seal methods were found
to have dropped significantly, seals deteriorated quicker than it was anticipated. The reason for
this unanticipated drop of the performance index is probably the severely cold 2018-2019 winter,
which expanded many cracks beyond the typical average winter crack widths and stretched
sealants beyond their allowable strains.
 A large amount of spalling failure was observed in rout-and-seal sections. Although not formally
addressed as a seal failure for calculating performance in other studies, spalling was found to be
contributing to a large portion of performance loss in the cracks documented for this study.
 Sites documented with rout-and-seal also experienced a large number of adhesion failures. This
failure was seen more commonly with increased crack spacing.
 The short-term performance for closely spaced cracks on a rural road is identical for both clean-
and-seal and rout-and-seal.
 Cohesion failure has not yet been observed in this study.

Old crack seal project sites (Phase 4)

 Only rout-and-seal project data was available for the old crack seal sites. Although the PI drop for
these sites during the second winter was not as large as the new crack project sites, the drop was
noticeable.

Effectiveness of the crack sealing methods

 Data from the new and old crack seal sites were used to develop the relationship between the
average performance index and age of the seals; a reasonable correlation was found with an R2 =
0.95.
 The performance data of crack seals were used to study the effectiveness of the crack sealing
method. A threshold PI value of 50 was assumed to determine the service lives of the crack seals.
It was found that the rout-and-seal and clean-and-seal methods have approximately 4 and 3-year
service lives, respectively. The decision on these service lives was made based on the data
collected in this project as well as relevant crack seal performance data found in several literature.
 The life-cycle cost analysis (LCCA) and benefit-cost ratio analysis showed that the rout-and-seal
was more effective than clean-and-seal due to its longer performance period. However, if only a
short-term benefit is considered, then the clean-and-seal could be more cost-effective than its
counterpart.
 As the benefit/cost analysis did not yield a clear distinction between the effectiveness of the two
crack sealing methods, several other decision factors were then considered to determine the

117
 effectiveness of each crack seal method. Various factors such as treatment cost, B/C ratio,
expected life, ease of operation, practitioners’ opinion, and traffic level were considered.
 Two decision trees were developed for choosing the most appropriate crack sealing method. The
first one, which can be used for a pavement management system, needs more information such
as crack severity, pavement type (new vs. overlay), pavement analysis period and design life,
traffic level, and crack seal occurrence number. The second decision tree, which is a simplified
version of the first one and can be used by the preventive maintenance crews, needs less
information such as, crack severity, traffic level, and crack sealing occurrence number. In general,
the clean-and-seal method was found to be appropriate for high crack severity conditions. The
choice between rout-and-seal and clean-and-seal for low and moderate crack severity was found
to be varied based on pavement type and age, and traffic levels. The clean-and-seal method was
found to be appropriate for sandy soil subgrade and the low initial budget scenario, whereas the
rout-and-seal method was preferred for clayey and silty subgrades irrespective of other variables.

118
REFERENCES
Al-Quadi, I., Ozer, H., Youselfi, S., Sawalha, M., and McGhee, K. (2017). Validation of Hot-Poured Crack
Sealant Performance-Based Guidelines. Report No. FHWA/VTRC 16-R. Illinois Center for Transportation,
Rantoul, IL. Retrieved from <https://apps.ict.illinois.edu/projects/getfile.asp?id=5165>; Accessed
October 2017.

AL Trans. (2003). Crack Maintenance Guidelines. Version 2. Maintenance Guidelines and Procedures.
Alberta Transportation, Alberta, Canada. Retrieved from
<http://www.transportation.alberta.ca/PlanningTools/GMS/ProductList/BMP/CrTrtmntGdln.pdf&gt>;
Accessed July 2018.

ASTM. (2015). D6690-15 Standard Specification for Joint and Crack Sealants, Hot Applied, for Concrete
and Asphalt Pavements. Retrieved from <https://compass.astm.org/EDIT/html_annot.cgi?D6690+15>;
Accessed October 2017.

Caltrans. (2008). Maintenance Technical Advisory Guide Volume I – Flexible Pavement Preservation.
Second Edition. URL &lt: <http://www.dot.ca.gov/hq/maint/MTA_GuideVolume1Flexible.html>;
Accessed October 2017.

Galehouse, L. (2004). Development of a Pavement Preventive Maintenance Program for the Colorado
Department of Transportation. Report No. CDOT-DTD-R-2004-17. Colorado Department of
Transportation, Denver, CO.

Hajj, E. Y., Loria, L., & Sebaaly, P. E. (2009). Performance Evaluation of NDOT Pavement Maintenance
Activities. Draft Final Report. Research Division, Nevada Department of Transportation, Carson City, NV.

IL DOT. (2010). Pavement Preservation Manual. Illinois Department of Transportation, Rantoul, IL.
Retrieved from <https://idot.illinois.gov/Assets/uploads/files/Doing-Business/Manuals-Split/Design-
And-Environment/BDE-Manual/Chapter%2052%20Pavement%20Preservation.pdf&gt>; Accessed July
2018.

Johnson, A. M. (2000). Best Practices Handbook on Asphalt Pavement Maintenance. St. Paul, MN. No.
MN/RC-2000-04. Retrieved from <http://www.cee.mtu.edu/~balkire/CE5403/AsphaltPaveMaint.pdf>;
Accessed October 2017.

Schulte, K. (2018). Reading: Mechanical Weathering. Module 4: Weathering and Soil Formation. Lumen
Learning. Retrieved from < https://courses.lumenlearning.com/geo/chapter/reading-mechanical-
weathering/>; Accessed October 2018.

MnDOT. (2016). Standard Specification for Construction. Minnesota Department of Transportation, St.
Paul, MN. Retrieved from <http://www.dot.state.mn.us/pre-letting/spec/>; Accessed October 2017.

119
MnDOT (2018a). MnDOT Network Level Bituminous Decision Tree. Minnesota Department of
Transportation, St. Paul, MN. Retrieved from
<https://www.dot.state.mn.us/materials/pvmtmgmtdocs/Updated_Bituminous_DT_11x17.pdf>;
Accessed October 2018.

MnDOT (2018b). Roadway Data: Construction data log. Retrieved from

<https://www.dot.state.mn.us/roadway/data/const-projlog-bydistrict.html>; Accessed Feb 2018.

MnDOT (2018c). Traffic Forecasting & Analysis: County Maps. URL &lt:
<http://www.dot.state.mn.us/traffic/data/maps-county-alpha.html>; Accessed Feb 2018.

MnDOT (2018d). Construction and Maintenance Projects Abstract Page. Retrieved from
<http://transport.dot.state.mn.us/PostLetting/Abstract.aspx>; Accesses Feb 2018.

MnDOT. (2018e). Pavement Management. Retrieved from

<https://www.dot.state.mn.us/materials/pvmtmgmt.html>; Accessed Feb 2018.

MnDOT. (2018f). MnDOT Pavement Preservation Manual. Minnesota Department of Transportation,

Minnesota. Retrieved from
<http://www.dot.state.mn.us/materials/pavementpreservation/manualsandguides/documents/MnDOT
_Pavement_Preservation_Manual.pdf&gt>; Accessed on July 2018.

MnDOT. (2018g). MnDOT Pavement Design Manual. Minnesota Department of Transportation, St. Paul,
MN. Retrieved from
<http://www.dot.state.mn.us/materials/pvmtdesign/docs/newmanual/Pavement_Design_Manual_10-
31-14.pdf&gt>; Accessed July 2018.

MnDOT. (2018h). MnDOT Standard Specification for Construction. Minnesota Department of

Transportation, St. Paul, MN. Retrieved from <http://www.dot.state.mn.us/pre-letting/spec/>; Accessed
July 2018.

OHMPA. (2004). The ABCs of Pavement Preservation. Issue 1.0. The Ontario Hot Mix Procedures
Association, Mississauga, Ontario. Retrieved from <http://co-asphalt.com/wp-
content/uploads/2015/03/OHMPA_ABCs_of_PPM.pdf&gt>; Accessed July 2018

Rajagopal, A. (2011). Effectiveness of Crack Sealing on Pavement Serviceability and Life. Report No.
FHWA/OH-2010/14 Ohio Department of Transportation, Columbus, OH and Department of
Transportation Federal Highway Administration, Washington, DC. Retrieved from
<https://ntl.bts.gov/lib/43000/43000/43075/134364_FR.pdf>; Accessed October 2017.

Smith, K., & Romine, A. (1999). Materials and Procedures for Sealing and Filling Cracks in Asphalt-
Surfaced Pavements Manual of Practice. Report No. FHWA-RD-99-147 Federal Highway Administration,
Washington DC. Retrieved from

120
<https://www.fhwa.dot.gov/publications/research/infrastructure/pavements/ltpp/99147/99147a.pdf>;
Accessed October 2017.

Stoikes, J. (2017). How to Select Crack Routing Equipment. Retrieved from

<http://www.forconstructionpros.com/pavement-maintenance/preservation-
maintenance/cracksealing-equipment-materials/article/12203479/how-to-choose-a-crack-router>;
Accessed October 2017.

121