Performance-Based Pavement Warranty Practices (2020)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

5 Introduction Transportation agencies often face pressure to reduce construction costs and traffic closures, while achieving high quality projects. European agencies have utilized pavement warranties to improve the quality of pavements and to protect highway investments from premature failure (Hancher 1994, D’Angelo et al. 2003, Cui et al. 2008, An et al. 2014). For example, Denmark, Germany, Sweden, and the United Kingdom utilize pavement performance warranties (with end result specifications); Spain and the United Kingdom utilize design-build-finance-operate (DBFO) (with end result specifications); and all utilize materials and workmanship–method specifications (D’Angelo et al. 2003). In the United States, one of the first documented pavement warranty projects, and the first concrete pavement constructed in the United States, was built in 1891 in Bellefontaine, Ohio, and included a $5,000 bond for a 5-year warranty (American Concrete Pavement Association 2016). Pavement warranties became a more common practice in the United States in the early 1900s and in the 1950s during the construction of the interstate highway system. However, they fell out of use since warranty-specified corrective maintenance activities were considered routine maintenance activities, and regulations prohibited expenditure of funds on pavement main- tenance using federal dollars (Qi et al. 2012). In addition, warranty clauses were prohibited in federally funded highway projects under regulation 23 CFR 625.413 (Hancher 1994). Pave- ment warranties started to regain popularity in the United States in the 1990s when revised fed- eral regulations returned to allowing SHAs to implement pavement warranties (Hughes 1997, Qi et al. 2012, FHWA 2017). Specifically, the Intermodal Surface Transportation Efficiency Act of 1991 allowed SHAs to use their own warranty policies and procedures for projects with federal funding. By 2003, the Federal Highway Administration (FHWA) indicated that 23 SHAs were using pavement warranties in some of their roadway projects (FHWA 2017). Warranty Types Warranties are defined as “a type of performance specification that guarantees the integrity of a product and assigns responsibility for the repair or replacement of defects to the con- tractor” (Transportation Research E-Circular E-C235 2018). The purpose of a warranty is to specify the expected performance of a roadway or product for a given period and to define who holds liability for the product or roadway. In practice, warranties are used for new and reconstructed asphalt and concrete pavement projects, as well as rehabilitation and preventive maintenance projects. In general, there are two types of warranties used on highway projects: • Materials and Workmanship. “Specifications that hold the contractor responsible for correcting defects in work elements within the contractor’s control during the warranty C H A P T E R 2 Literature Review

6 Performance-Based Pavement Warranty Practices period. . . . The contractor assumes no responsibility for pavement design or those dis- tresses that result from the design. Some responsibility is shifted from the agency to the contractor for materials selection and workmanship” (Transportation Research E-Circular E-C235 2018). • Performance. “Specifications that hold the contractor fully responsible for product perfor- mance during the warranty period. . . . For short-term warranties, contractor responsibility ranges from mix design to structural (thickness) pavement design. . . . For long-term warran- ties, pavement performance is the responsibility of the contractor” (Transportation Research E-Circular E-C235 2018). Highway construction projects are typically contracted using either design-bid-build or design-build project delivery methods. Design-bid-build is a traditional project delivery method where the agency is responsible for developing plans and specifications and separately for con- tracts, through competitive bidding or negotiation, a contractor to conduct the work (Design- Build Institute of America 2009). For design-build projects, the agency selects one entity to prepare plans and specifications and conduct construction under a single contract (Design- Build Institute of America 2009). Figure 1 further illustrates the differences in the design-bid- build and design-build methods. Public-private partnerships (P3) are an alternative project delivery process. In general, P3s are contractual agreements between the public and private sectors for building and equipping facilities (e.g., schools, hospitals, transportation). For transportation projects, the private sector typically takes on additional risks related to design, construction, finance, long-term operation, and traffic revenue (FHWA 2019a). Additional information on innovative financing can be found at the FHWA and the World Bank websites (FHWA n.d., World Bank 2019). Table 1 provides a summary of key factors by project delivery and warranty type. Warranty Periods The warranty period is the duration of time, starting from the acceptance of work, to the end of the evaluation period. Warranty periods typically range from 1 to 3 years for pavement preservation, 3 to 5 years for asphalt pavements, 5 to 10 years for concrete pavements, and 20 or more years for long-term warranties as part of a P3 project (Scott et al. 2011, Transportation Research E-Circular E-C235 2018). Usually, the warranty period is determined based on pave- ment performance data obtained from the agency’s pavement management system. The length of the warranty period should be long enough to balance the risk between the agency and the contractor in assuring the specified pavement performance, while minimizing project costs when the contractor is required to assume most of the risk (FHWA 2017). Table 2 provides a summary of SHA warranty periods. Project Selection Criteria According to Scott et al. (2011), few SHAs have developed project selection criteria, and for these few, the selection criteria varies by agency. Warrantied projects often range from low-risk projects to projects with unknown variables and risks. However, selected pavement warranty projects are predominantly low risk and exhibit sound base conditions. Table 3 summarizes common SHA project selection criteria.

a. Design-bid-build b. Design-build Figure 1. Project delivery methods (Design-Build Institute of America 2015).

8 Performance-Based Pavement Warranty Practices Consideration Materials and Workmanship Short-Term Performance Long-Term Performance Contract Type Design-bid-build Design-bid-build Design-build Design-build Provision Development Indicators, thresholds, and bond requirements + mix design criteria + structural design criteria + turn-back criteria Performance Criteria and Thresholds Characteristics indicative of early, catastrophic failure Characteristics predicative of long- term performance Characteristics predicative of long- term performance Characteristics predicative of long- term performance Specifications Method Primarily method w/contractor mix design Combination of method and performance Primarily performance Procurement Low bid Low bid w/expanded prequalification Low bid, one- or two- step best-value Two-step best-value Effort Required to Develop/ Evaluate Bid Nominal Nominal to minimal, depends on prequalification Minimal to extensive, depends on procurement Typically, extensive Contractor Design Expertise Not responsible for mix design Working knowledge of mix design Knowledge of mix design and allows contractor flexibility with structural design Expertise in pavement design and allows flexibility with planned maintenance Construction Inspection and Testing Agency or contractor quality control specification Contractor quality control for mix design Contractor- controlled Contractor- controlled Agency Staffing Least amount of effort Some effort for developing criteria, but significant resources can be realized during construction More effort for developing design and criteria, but significant savings during construction Most effort for initial development, savings during construction, but may result in long- term monitoring Contractor Innovation Does not promote Only in areas where contractor is given control Most opportunity Most opportunity Maintenance Cycle No impact Higher initial quality, potentially fewer call backs Higher initial quality, potentially avoid call backs, may include preventive maintenance Contractors benefit to maximize quality and minimize maintenance cycles Quality Can improve, based on specification Biggest improvement when applied on small scale, superior crews directed to project Biggest improvement when applied on small scale, superior crews directed to project Not enough projects to determine effect Cost Comparable to traditional Increase is proportional to level of risk shifted to contractor and length of warranty period Increase based on risk for design and construction, typically does not include preventive maintenance Maintenance costs include in initial bid price Table 1. Key factors by warranty type (Scott et al. 2011).

Literature Review 9 Monitoring and Evaluation Pavement condition surveys are typically carried out during the warranty period to monitor pavement performance. Pavement monitoring is usually the responsibility of the SHA, but some- times can be done by the contractor depending on the warranty type (Scott et al. 2011). There are different methods for conducting the pavement condition survey: • Visual inspection using pictures or video of the road segments. • Computer-aided devices such as inertial profilers and image-processing software. • Specific equipment such as the falling weight deflectometer. • Manual distress measurements. Pavement Type CO IL IN LA ME MI MN MS NM OH New Asphalt 5 5 ― 3 5(a) 5 ― 5 ≤ 20(b) 7 New Concrete 5 5 ― 3 5(a) 5 ― 10 ≤ 20(b) 7 Asphalt Overlays 5 5 5(c) ― ― ― ― ― ― ― Rehabilitation ― ― ― ― ― 5 2 to 5 ― ― ― Preservation ― ― ― ― ― 3 2 to 5 ― ― 2 to 3(d) (a) Contractor can propose additional 1-year increments up to 5 additional years. (b) In 5-year increments. (c) Asphalt overlays of cracked-and-seated or rubblized concrete pavements. (d) 2 years for chip seals; 2 years for microsurfacing; and 3 years for hot in-place recycling. Note: ― not used. Agency Selection Criteria Colorado • At least 75% of the project scope involves paving asphalt pavement, or the primary project scope involves paving concrete pavement • 20-year design • Asphalt pavement new construction or reconstruction • Concrete pavement constructed on subgrade or an unbonded overlay • Minimum concrete thickness of 9 in. (229 mm) for new construction • At least 3 mi (4.8 km), preferably at least 5 mi (8.0 km) for asphalt pavements • Project should be design-bid-build for asphalt pavements • Weigh-in-motion installed on or near the project Illinois • Projects are simple in scope • No unique design or materials Louisiana • New construction only Michigan • High-volume roads (initially, but no longer utilized) • All paving projects (may waive if repair is temporary fix) • Base conditions considered for rehabilitation projects Ohio • Warrantied work is within contractor’s control and measurable • Well and explicitly defined materials and workmanship performance attributes and/or failure thresholds • Aspects not under control of contractor have minimal effects on the warrantied pavement during the warranty period • Project may develop or incorporate innovative technologies • Existing project conditions are well defined • Performance requirements, monitoring methods, and acceptable thresholds must be defined • Construction quality parameters and acceptance criteria are clearly defined Wisconsin • Projects with excellent subbase Table 3. Warranty project selection criteria (Scott et al. 2011). Table 2. Warranty periods (years) by agency (Scott et al. 2011, Qi et al. 2012).

10 Performance-Based Pavement Warranty Practices Pavement evaluations may be conducted on segments within a specified section of roadway and extrapolated over the entire pavement section (e.g., sampling) or may include a 100% survey of the entire pavement section. Regardless, the length of the evaluation segment should be clearly defined by the agency. The segment length should be long enough to represent the overall pave- ment condition and allow for practical pavement evaluation and condition data collection. For sampling, typical pavement evaluation segment lengths range from 300 to 500 feet (91 to 152 m) (Scott et al. 2011). A summary of agency monitoring and evaluation practices is summarized in Table 4. Performance Indicators and Thresholds Performance indicators are measurable structural or functional performance characteristics for a given pavement type (Singh et al. 2005, Scott et al. 2011). Agencies are typically concerned with the structural capacity or integrity of the pavement, functional performance, and safety, Agency Monitoring Evaluation Segment Exclusions Colorado Pavement Management Data and Pavement Evaluation Team; conducted annually by agency Project divided into 1 mi (1.6 km) segments; evaluate each 0.10 mi (0.16 km) segments N/A Illinois Manual or automated condition surveys (cracking manual or image processing), lasers for roughness data; conducted by agency during and prior to the end of the warranty period 0.1 mi (0.16 km) Failures due to design, significant increase in traffic Indiana Annual pavement condition conducted annually by agency 328 ft (100 m) segment Factors beyond the control of contractor, projected Class 5(a) trucks > 50% Louisiana Distress survey conducted by the agency within 6 months prior to the end of the warranty period 500 ft (152 m) segments Naturally occurring transverse cracking in continuously reinforced concrete pavement Maine Agency selects and conducts evaluation on two segments 328 ft (100 m) segments Excess of cumulative equivalent single axle loads (ESALs) Michigan Conducted by agency; excludes distress beyond contractors’ control 492 ft (150 m) segments Design issues Minnesota Annual distress survey conducted by agency 1 mi (1.6 km); 500 ft (152 m) segments Accumulated ESALs 50% > projected ESALs Mississippi Annual distress surveys conducted by agency; rutting, roughness, and faulting collected 100% 500 ft (152 km) segment per mi (1.6 km) N/A New Mexico Condition surveys conducted by agency at least once a year Not specified ESALs > 4,000,000 Wisconsin Annual automated or semi-automated distress survey conducted by agency Two 0.10 mi (0.16 km) segments per mi (1.6 km); one segment 0.3 and 0.4 mi (0.5 to 0.6 km) from start of section Rutting waived if ESALs are more than 50% from projected Pavement thickness related issues (a) Two axle, six-tire, single-unit truck (FHWA 2019b). Table 4. Monitoring and evaluation practices (Qi et al. 2012).

Literature Review 11 while road users tend to focus on the roadway’s functional performance and safety. Typically, agencies will identify performance indicators related to structural capacity, functional perfor- mance, and safety and set thresholds to maintain a satisfactory level through the warranty period (Shober et al. 1996, Sinha et al. 2006, Scott et al. 2011). Pavement distresses, skid resistance, and ride quality are often used as performance indicators. Table 5 summarize SHA performance indicators for asphalt and concrete pavements. Performance indicator thresholds values are usually established based on expert opinion, performance models, and agency experience (Scott et al. 2011, Qi et al. 2012, An et al. 2014). Table 6 and Table 7 summarize SHA threshold values for asphalt and concrete pavements, respectively. Performance Indicator State Highway Agency Count CO IL IN LA ME MI MN MS NM OH TX WI Asphalt Pavements Rutting 12 Longitudinal cracking 11 Raveling 11 Transverse cracking 11 Alligator/fatigue cracking 10 Bleeding/flushing 10 Potholes 9 Ride quality or IRI 7 Shoving 7 Block cracking 4 Debonding/delamination 4 Edge cracking 4 Friction 3 Longitudinal joint distress 3 Patch/patch deterioration 2 Popouts 2 Slippage cracking 2 Reflection cracking 1 Settlement 1 Concrete Pavements Longitudinal cracking 6 Transverse cracking 6 Corner breaks 5 Scaling 5 Faulting 4 Joint seal damage 4 Spalling 4 Popouts or blowups 3 Map cracking 2 Diagonal cracking 1 Lane-shoulder separation 1 Patching/patch deterioration ― ü ― ― ― ― ― ― ― ― ― ― 1 Punchouts(a) ― ü ― ― ― ― ― ― ― ― ― ― 1 Ride quality or IRI ― ― ― ― ― ü ― ― ― ― ― ― 1 Shoulder or lane drop off ü ― ― ― ― ― ― ― ― ― ― ― 1 (a) Continuously reinforced concrete pavements only. Note: used and ― not used. Table 5. Asphalt pavement performance indicators (Qi et al. 2012).

12 Performance-Based Pavement Warranty Practices Agency Distress Type Threshold Value Period Colorado Bleeding, fatigue cracking, raveling < 50 ft2 (4.6 m2) per 0.1 mi (0.16 km) 5 years (new and overlay) Longitudinal cracking < 30 ft (9 m) per 0.1 mi (0.16 km) Transverse cracking < 5 count per 0.1 mi (3 per 0.1 km) Rutting < 0.50 in. (13 mm) per mi (km) Illinois Bleeding, raveling < 500 ft2 (46.5 m2) per 0.1 mi (0.16 km) moderate severity; no high severity 5 years (new and overlay) Block cracking < 100 ft2 (9.3 m2) per 0.1 mi (0.16 km) moderate severity; no high severity Fatigue cracking < 50 ft2 (15.2 m2) moderate severity; no high severity Longitudinal cracking < 10 ft (3 m) per 0.1 mi (0.16 km) moderate severity; no high severity in lane, at centerline, edge line Potholes None Ride IRI < 110 in/mi (1.74 m/km) Rutting < 0.30 in. (8 mm) Transverse cracking < 10 ft (3 m) per 0.1 mi (0.16 km) moderate severity; no high severity Indiana Block, fatigue, longitudinal, and transverse cracking < Severity 2 or greater 5 years (overlay of crack and seated or rubblized concrete) Friction > 25 Ride IRI < 133 in/mi (2.10 m/km) Rutting < 0.35 in. (9 mm) Louisiana Bleeding, fatigue cracking, raveling < 10 ft2 (0.9 m2) per 500 ft (152 m) 3 years (new) Edge cracking < 50 ft (15 m), crack width ≥ 0.25 in. (6 mm); < 100 ft (30 ft) per 500 ft (152 m) Longitudinal and transverse cracking < 50 ft (15 m), crack width ≥ 0.25 in. (6 mm); < 200 ft per 500 ft (50 m per 152 m) Potholes None Rutting < 0.35 in. (9 mm) per 50 ft (15 m) or < 0.50 in. (13 mm) per 500 ft (152 m) Shoving None Maine Fatigue cracking < 328 ft (100 m) per 328 ft (100 m), crack width < 0.28 in. (7 mm) 5 years (new, allows up to an Raveling and popouts < 0.4 in. (10 mm) over 16 ft2 (1.5 m2) per 328 ft (100 m) additional 5 years) Ride IRI < 79 in/mi (1.25 m/km) per 328 ft (100 m) Rutting < 0.3 in. (8 mm) per 328 ft (100 m) Depressions/shoving < 0.6 in. (15 mm) over 10 ft (3 m) per 328 ft (100 m) Transverse cracking spacing < 15 ft (4.5 m) and 0.3 in. (8 mm) crack width per 328 ft (100 m) Michigan Bleeding < 4% length per 525 ft (160 m) 5 years (reconstruct, rehabilitation); 3 years (preventive maintenance) Debonding < 5% segment length Longitudinal cracking < 10% length (or < 52.5 ft [16 m]) Raveling < 8% length per 525 ft (160 m) Ride Quality Index < 50 per 525 ft (160 m) Rutting < 0.40 in. (10 mm) per 525 ft (160 m) Transverse cracking < 2 counts per 525 ft (160 m) Table 6. Threshold values—asphalt pavement (Anderson et al. 2006, Scott et al. 2011, and Qi et al. 2012).

Literature Review 13 Agency Distress Type Threshold Value Period Mississippi Bleeding < 0.4 deduct points(a) 5 years Block, edge, trans. cracking < 3.0 deduct points Fatigue < 10.0 deduct points Longitudinal cracking < 4.0 deduct points Potholes and rutting < 5.0 deduct points Raveling < 0.2 deduct points New Mexico(b) Bleeding < 538 ft2 (50 m2) coefficient of friction < 0.05 ≤ 20 years (5-year increments) Fatigue cracking < 0.20 in. (5 mm) crack width Raveling < 54 ft2 (5 m2) Ride IRI < 80 in/mi (1.26 m/km) Potholes < 6 in. (152 mm) wide; 0.40 in. (10 mm) deep Rutting < 0.40 in. (10 mm) per 1.78 mi (3 km) Settlement and shoving < 0.6 in. (15 mm) per 10 ft (3 m) Transverse cracking spacing < 15 ft (4.6 m) and < 0.2 in. (5 mm) crack width Ohio Bleeding < 125 ft2 (11.6 m2) per 0.1 mi (0.16 km) 2 years (seals, microsurfacing), 3 years (hot in- place recycling), 7 years (new) Block, edge, fatigue, long., trans. cracking < 500 ft (152 m) per 0.1 mi (0.16 km), crack width > 0.25 in. (6 mm) Rutting < 0.375 in. (10 mm) per 0.1 mi (0.16 km) Texas Fatigue cracking and raveling < 1 yd2 (0.8 m2) per 0.1 mi (0.16 km) 3 years Friction 20% reduction from initial skid measurement Longitudinal cracking < 20 ft (6 m) per 0.1 mi (0.16 km), crack width > 0.0625 in. (2 mm) Potholes < 1 ft2 (0.1 m2) and < 1 in. (25 mm) deep Ride 20% increase from initial IRI measurement Rutting < 0.50 in. (13 mm) Shoving < 1 occurrence > 1 in. (25 mm) Wisconsin Bleeding < 20% length per 0.1 mi (0.16 km) 5 years Block and fatigue cracking < 1% area per 0.1 mi (0.16 km) Longitudinal cracking < 1,000 ft (305 m), crack width > 0.5 in. (13 mm) or < 25% band crack per 0.1 mi (0.16 km) Potholes None Raveling < 1 yd2 (0.8 m2) Rutting < 0.25 in. (6 mm) per 0.1 mi (0.16 km) Transverse cracking (count) Granular base < 25, > 0.5 in. (13 mm) width Overlay concrete < 50, > 0.5 in. (13 mm) width (a) Based on Pavement Condition Rating procedure, scale ranges from 0 to 100 (no defects). (b) Length of evaluation segment not specified. Minnesota Block, fatigue, longitudinal cracking, potholes None 2 years (mill and overlay); 5 years (bituminous pavements) Raveling < 1% area per 500 ft (152 m) Rutting < 0.375 in. (10 mm) per 25 ft (8 m) Transverse cracking (count) Low and medium severity < 3; high severity < 2; minimum 6 ft (2 m) per 500 ft (152 m) Table 6. (Continued).

14 Performance-Based Pavement Warranty Practices Agency Distress Type Threshold Value Period Colorado Corner breaks and longitudinal and transverse cracking Spalling < 2 in. (51 mm), faulting < 0.25 in. (6 mm), crack width < 0.125 in. (3 mm), corner not broken > 2 pieces per 0.1 mi (0.16 km) 5 years (new) Faulting < 0.25 in. (6 mm) per 0.1 mi (0.16 km) Joint seal damage < 2 ft (0.6 m) per 0.1 mi (0.16 km) Popups or blowouts < 2 per 1 yd2 (0.8 m2) or < 2 in. (51 mm) deep Scaling < 2.0 ft2 (0.2 m2) per slab Shoulder/lane drop off < 0.5 in. (13 mm) separation Illinois Corner breaks No moderate or high severity 5 years (new) Longitudinal and transverse cracking, and spalling Moderate severity < 10 ft (3 m) per 0.1 mi (0.16 km); no high severity Patch deterioration Moderate severity < 100 ft2 (9.3 m2) per 0.1 mi (0.16 km); no high severity Punchouts No moderate or high severity Ride quality IRI < 150 in/mi (2.37 m/km) Louisiana Corner breaks, joint seal damage, and longitudinal and transverse cracking None 3 years (new) Faulting < 0.125 in. (3 mm) per 500 ft (152 m) Scaling Diameter < 1 to 4 in. (25 to 102 mm) and depth 0.5 to 2 in. (13 mm to 51 mm) per 500 ft (152 m) Spalling < 2 in. (51 mm) wide 500 ft (152 m) Michigan Corner breaks, trans. cracking < 1 occurrence per 525 ft (160 m) 5 years (new) Joint seal damage < 10% length < 2 slabs per 525 ft (160 m) Longitudinal cracking < 5% length per 525 ft (160 m) Map cracking < 10% area per 525 ft (160 m) Spalling 10% each slab ≤ 2 slabs Ride quality (RQI) < 50 per 525 ft (160 m) Scaling < 15% area per 525 ft (160 m) Mississippi Corner breaks < 4.30 deduct points(a) 10 years Faulting < 2.70 deduct points Joint seal damage < 1.66 deduct points Longitudinal cracking < 1.40 deduct points Map cracking < 1.77 deduct points Spalling < 1.15 deduct points Transverse cracking < 1.97 deduct points Ohio Faulting < 0.1875 in. (5 mm) per 0.1 mi (0.16 km) 7 years Longitudinal cracking None 15 in. (0.4 m) from tied longitudinal joint Transverse cracking Crack width < 0.0625 in. (2 mm) for nonreinforced concrete, crack width < 0.25 in. (6 mm) reinforced concrete per 0.1 mi (0.16 km) (a) Based on the pavement condition rating procedure, scale ranges from 100 (no defects) to 0 (Mississippi Department of Transportation n.d.). Table 7. Threshold values—concrete pavement (Qi et al. 2012).

Literature Review 15 Specification Content Pavement warranty projects may require different construction specifications than typically used for nonwarranty projects. Regardless of the type of warranty used, the main purpose of having warranty specifications are to clearly define the project objectives, assign responsibilities for all parties, and to determine, if needed, a remediation process (Gallivan 2011). Warranty specifications may differ agency to agency, but the core elements summarized in Table 8 are typically included in the construction specifications. As part of the warranty specifications, exclusions waiving the contractor’s responsibility also need to be clearly defined. In most cases, exclusions are related to factors outside the contractor’s control such as pavement design, under predicated traffic levels, or poor (or weak) subgrade. Sometimes weigh-in-motion stations can be beneficial, especially in long-term warranties, to help minimize the contractor’s risks associated with underestimated traffic volumes and loads (Scott et al. 2011). In the event of warranty disputes, a common practice is to invoke a conflict resolution team (CRT) to provide a final decision between the agency and the contractor. The CRT typically includes three or five members, and consists of representatives from the agency, the contractor, and an external member mutually agreed by the agency and the contractor. Guidelines for Project Selection and Provision Development Anderson and Russell (2001) prepared Guidelines for Warranty, Multi-Parameter, and Best Value Contracting. In 2011, Scott et al. updated the guidelines to reflect the current state of practice, including criteria for project selection, asphalt and concrete pavement specifications for both materials and workmanship and performance-based warranties, and a project selection Content Description Project Description • Description of project scope covered under the warranty • Define responsibilities • Define the warranty period and start date Warranty Bond • Define bond coverage, amount, effective date, and warranty periods • Consider alternatives to bonds such as guarantees Performance Evaluation • Clearly define performance indicators • Set performance thresholds • If performance indicators reach or exceed the thresholds, specify the remedial actions to be performed Warranty Work • Clearly identify the coverage in the scope of warranty • Contractor’s remedial work plan (e.g., permits, lane closures, and traffic control) • Include emergency work details • Define exclusions or statements that will not hold the contractor responsible for distresses caused by factors beyond their control Maintenance Responsibilities • Define agency maintenance activities • Define contractor maintenance activities according to the warranty type • Define responsibilities for emergency repairs Dispute Resolution Team • Usually composed of contractor representative, the agency’s representatives and a third external party agreed by both parts Acceptance • Set criteria for initial acceptance • Set criteria for final acceptance and close-out Measurement • Consider various measurement methods (unit measurement or lump sum value) Basis of Payment • Establish basis of payment on the method of measurement Other Considerations • Final warranty acceptance Table 8. Warranty specifications—core elements (Anderson et al. 2006, Scott et al. 2011).

16 Performance-Based Pavement Warranty Practices tool. In general, the process for selecting and developing pavement warranty provisions are sum- marized in Table 9. European Experience European agencies started implementing warranties for several reasons, including resource shortages, to maintain the road infrastructure, and to improve construction quality (D’Angelo et al. 2003). Many European contracts have been awarded on a best-value basis compared with the low-bid basis used in the United States. Process Discussion Understand and Communicate Rationale for Warranties • Improve quality (end-product performance, workmanship, and quality balancer (e.g., incentive/disincentive, multi-parameter bidding) • Promote innovation and new technology • Reduce agency administration and inspection costs (depends on cooperation between agency and industry, industry capability and experience, scope, and duration) • Reduce premature failures (specifications have been strictly adhered to and enforced) • Transfer accountability and risk (motivate contractors to be more quality-conscious and innovative, replace or supplement agency inspection resources) • Reduce life-cycle costs or extend service life (develop agency standards for assessing life- cycle cost, provides contractor more flexibility with alternate designs, construction methods, or process controls) Program Considerations • Warranty concept buy-in (e.g., public, legislative, management, project staff, industry) • Phased development (gradually shift from method to end-result specifications) • Measuring performance (e.g., identify measures [IRI, cracking, faulting, rutting], ensure compatibility with pavement management and acceptable measurement methods, define reasonable threshold criteria, and clearly define remedial action) • Warranty monitoring (define responsible personnel, timing, frequency) • Surety bonds and alternatives, for example: o < 2 years: extend performance bond to eliminate a separate bond requirement o 3 to 10 years: use 1- to 2-year renewable bonds to limit capacity restrictions, alternative security, or prequalification process to guarantee performance o > 10 years: use renewable bonds or spread contractor payments across the warranty life (equal disbursements or based on level of work) • Other considerations (e.g., contracting and procurement, available agency resources, industry expertise, balance warranty type with desired outcome) Project Considerations • Contracting strategy and warranty type o Design-bid build (method specifications, low bid, separation of services, owner retains majority of risk, materials and workmanship warranty) o Design-build (performance warranty) o P3 (performance specifications, best-value, integrated services, investor/contractor assumes majority of risk, performance warranty) • Scope of work (new construction, rehabilitation, and preservation), agency determines: o Define warranty application (surface course, treatment, full depth) o Determine areas to exclude due to pre-existing conditions and anticipated fix o Consider if subcontracting will affect the warranted work o Determine who will be responsible for construction phasing, work sequence, and maintenance of traffic o Require contractor quality control or management plan • Conditions outside the contractor’s control (e.g., traffic accidents, natural disasters, acts conducted by third parties, failure due to existing conditions outside the scope of work) Table 9. Process for selecting and developing pavement warranty provisions (Scott et al. 2011).

Literature Review 17 In 1990, 1992, and 2002, U.S. contingents representing the highway construction industry visited European countries to obtain information in relation to pavement roadway construction practices (AASHTO 1991, FHWA 1994, D’Angelo et al. 2003). The intent of the tours varied; all three focused on asphalt pavements, and only the 2002 tour focused on pavement warranties; however, warranty information was also obtained from the previous two tours. The countries visited by tour year is shown in Table 10. Many of the countries have been using materials and workmanship warranties for at least 30 to 40 years, with an evolution in the early 2000s to performance-based warranties for design-build, DBFO, and pavement performance contracts (D’Angelo et al. 2003). Agen- cies also indicated that all pavement construction projects include warranties (D’Angelo et al. 2003). The majority of visited countries indicated a pavement warranty period of 2 to 5 years. Based on the 2002 tour, Denmark and Sweden use performance-based warranties, while Germany, Spain, and the United Kingdom use materials and workmanship warranties. For pavement structural design, three agencies (Denmark, Norway, and United Kingdom) conduct the design, three agencies (Austria, France, and Germany) use contractor designs, and one agency (Sweden) includes a joint effort by the agency and the contractor. In relation to quality control and acceptance, Austria, Denmark, France, Norway, and Sweden specify contractor quality control and acceptance; Germany requires contractor quality control and agency acceptance; and in the United Kingdom, the agency conducts both quality con- trol and acceptance. A summary of the European agency warranty practices is provided in Table 11. North American Experience As noted previously, the number of U.S. pavement warranty projects has grown since the 1990s. By 2011, U.S. agencies had completed more than 2,000 pavement warranty projects. Table 12 provides a summary of treatment type, warranty type, warranty period, and number of warranty projects constructed by agency. As shown in Table 12, most warranty projects have been applied to asphalt pavements and include new construction, rehabilitation, and preserva- tion projects. As summarized by Scott et al. (2011), several agencies evaluated and documented their experience with and effectiveness of pavement warranties. Table 13 summarizes agency pave- ment warranty experiences. Agencies 1990 1992 2002 Austria Denmark France Germany Italy Spain Sweden United Kingdom Note: visited and ― not visited. ― ― ― ― ― ― ― ― ― Table 10. Summary of countries visited by U.S. tours (AASHTO 1991, FHWA 1994, D’Angelo et al. 2003).

18 Performance-Based Pavement Warranty Practices Criteria Austria Denmark France Germany Spain Norway Sweden United Kingdom Warranty period (years) 2 to 5 5 (min) 10 4 to 5 1 (30)(a) 3 3 to 5 2(c) Warranty type (b) Perf(d) (b) M&W(d) M&W (b) Perf M&W Structural design Cont(e,f) Agency Cont Cont(g) (b) Agency Joint Agency Quality control and acceptance Cont Cont Cont Cont/ Agency(h) (b) Cont Cont Agency Performance indicators Cracking(i) (b) (b) (b) Potholes Raveling, joints Friction IRI Evenness Transverse profile Rutting Instability Crossfall Texture Measurement length, ft (m) (b) 328 (100) (b) 328 (100) 3,280 (1,000) (b) 65 (20) 328 (100) Measurement frequency(j) (b) Year 1 and 5 (b) Weekly(k) Bi- annual(l) (b) (b) Annual (a) 1-year for traditional projects and 30 years for DBFO. (b) Not specified. (c) For design-bid projects, 5-year warranty period, short-term performance, and end result specification. (d) Perf: performance-based; M&W: materials and workmanship. (e) Contractor. (f) Requires agency approval. (g) Within agency established limits. (h) Contractor conducts quality control with agency acceptance. (i) Includes longitudinal, transverse, and alligator cracking. (j) Countries also conduct performance measurements at the end of the warranty period. (k) Federal Ministry of Transport, local visual inspection for normal roads, thrice weekly for autobahns. (l) Includes annual bearing capacity evaluation. Note: used and not used. Table 11. Summary of European warranty practice (AASHTO 1991, FHWA 1994, D’Angelo et al. 2003, Goldbaum 2017).

Literature Review 19 Florida All pavement (94% asphalt) projects Materials and workmanship 3 > 700 Illinois Asphalt overlays (three projects), asphalt pavement (eight projects), and concrete pavement (16 projects) Materials and workmanship 5 27 Indiana Asphalt pavement preservation and asphalt overlay of rubblized or cracked and seated concrete Performance 5 10 (2)(a) Louisiana Asphalt pavement (two projects) and concrete pavement (one project) Materials and workmanship 3 3 Michigan Cold milling and single asphalt overlay, asphalt pavement crack treatment, microsurfacing, ultrathin overlay, paver- placed surface seal, and single and double chip seals Performance 3 > 1,000 Asphalt overlays, new asphalt and concrete Materials and workmanship 5 Minnesota Mill and asphalt overlay Performance 2 20 (1)(a) Asphalt and concrete pavement 5 Mississippi Asphalt pavement (13 projects) and concrete pavement (one project) Performance 5 (asphalt) 10 (concrete) 14 Ohio Preventive asphalt treatments Performance 3 > 200 Asphalt pavement Performance 7 Concrete pavement Materials and workmanship 7 Wisconsin Asphalt and concrete pavement Performance 5 > 80 (a) Number of pavement (pavement preservation) projects. Agency Treatment Type Warranty Type Warranty Period (years) No. of Projects California Chip seals, microsurfacing, asphalt overlays, and bonded wearing course Materials and workmanship 1 > 10 (12)(a) Colorado Asphalt pavement preservation and asphalt overlays Materials and workmanship 3 and 5 15 (1)(a) Table 12. U.S. SHA warranty projects (Anderson et al. 2006, Scott et al. 2011). The Ohio DOT noted an increase in bid unit price due to warranty application. Table 14 summarizes bid price changes over multiple years for warranty project. As summarized in Table 15, successful application of pavement warranties can provide improvements in the quality of materials and workmanship, reduce agency inspection and con- struction testing, promote contractor innovation, shift performance risk to the contractor, and improve pavement performance, thereby reducing life-cycle costs (Scott et al. 2011). Pavement warranties may also encourage the contractor to pay attention to details and clearly defines the roles and responsibilities of the contractor and the agency. The Ontario Ministry of Transportation (MOT) has included a 7-year asphalt pavement war- ranty on a total of 14 projects since 2006 (Lane 2018). The 14 projects have included full-depth reclamation (11 projects), cold in-place recycling (two projects), and mill and asphalt overlay (one project). The warranty specifications include a contractor-developed pavement structural design for rehabilitation or reconstruction (submit to department for review), while the depart- ment is responsible for specifying the design life (18 years); geometric, material, and construc- tion operational constraints; and performance requirements. Pavement performance criteria during and at the end of the warranty period is summarized in Table 16.

20 Performance-Based Pavement Warranty Practices SHA Experience California • Bids were competitive • Lower costs per mile (6% to 16% lower) on four of the five warranted projects • Contractors apply extra effort and pay more attention to aggregate and binder temperatures and spread rates Colorado • Short-term warranties: o No difference in performance compared to control o Slight improvement in performance, but not cost-effective • Long-term warranties: o Cost-benefit analysis showed no difference o Promotes innovative design and construction solutions Florida • No apparent cost difference due to prescriptive nature of warranty provision • Value-added warranty pavements are meeting expectations and have improved performance Illinois • Contractors may have distributed the warranty cost to other line items, making it difficult to quantify warranty cost • Did not see major difference in how contractor conducted work • Bid review and anecdotal observations indicate warranty projects cost slightly more for the same level of performance Indiana • Initial cost about 10% higher • Warrantied pavement has better quality and longer life • Warranty projects had less rutting and lower IRI • Analysis showed performance life extended 9 years • Life-cycle cost savings of 27% Louisiana • Warranty projects cost more and have the same quality as nonwarrantied projects • Determined to not be cost-effective Michigan • Contractors pay more attention to quality issues on warranty projects • Cost increase on warranty projects is due to the cost of the warranty bond • 13% of projects have required corrective action Minnesota • No significant changes to construction practices • No apparent cost increase for warranty projects, but contractors appear to increase mobilization item to address warranty costs Mississippi • Asphalt contractors indicated challenges with bonding for warranty projects • Comparative data and anecdotal evidence suggest warranty projects have improved performance • Anecdotal evidence suggests contractors using third-party mix designs, higher performance- grade binders, and better crews and equipment on warranty projects Ohio • Project selection includes simple projects with well-defined conditions and consistency • Average project bid increase was not significant; higher costs attributed to warranty bond • Warranty enforcement can be difficult Wisconsin • Initial cost about 7% higher, but life-cycle costs was less by about 10% • Analysis showed that warrantied projects were more cost-effective • Performance improved with warranty projects Table 13. U.S. experience and practice summary with warranty projects (Scott et al. 2011). Item Description Warranty Period 2000 Price Change 2005 Price Change Asphalt pavement (full depth) 5 and 7 +9% +1.2% Asphalt pavement (overlay) 3 +8% -1.8% Concrete pavement (11 in. [280 mm]) 7 +7% -7.8%(a) Concrete pavement (12 and 13 in. [305 and 330 mm]) 7 +15% Chip seal 2 N/A +12.5% (a) Average of all thicknesses. Table 14. Summary of unit bid price change (Ohio Department of Transportation 2007).

Literature Review 21 Requirement Acceptance Criteria Remedial Action Completion of Construction Surface condition No coarse aggregate loss or flushing, free of segregation, fat spot, bleeding, surface damage and contamination Repair or replace surface course Surface tolerance ≤ 0.12 in. (3 mm), measured with 10 ft (3 m) straightedge Diamond grinding or remove and replace IRI (average) < 79 in/mi (1.25 m/km) per 0.1 ln-mi (100 ln-m) Diamond grinding or remove and replace During Warranty Period Coarse aggregate loss No severe or very severe Repair or replace surface course Flushing No severe or very severe Repair or replace surface course Wheel track rutting No wheel paths 100 ft (30 m) in length with average rut depth > 0.6 in. (16 mm) Remove and replace rutted layer(s) Roughness (IRI) No segments > 139 in/mi (2.2 m/km) Partial depth remove and replace Final Year of Warranty Period(a) Coarse aggregate loss < 20% segment length, moderate severity or worse Repair or replace surface course Flushing < 10% segment length, moderate severity or worse Repair or replace surface course Alligator cracking < 5% segment area, slight severity or worse Reconstruction Single and multiple cracking < 20% segment length, cracks 0 to 1 in. (25 mm) wide Crack sealing < 160% segment length, cracks 0 to 1 in. (25 mm) wide Remove and replace asphalt full depth(b) No crack width > 1 in. (25 mm) wide Full-depth crack repair Wheel track rutting (average) < 0.3 in. (7 mm) per segment Remove and replace surface course IRI (average) < 95 in/mi (1.5 m/km) per segment, freeway < 101 in/mi (1.6 m/km) per segment, arterial < 108 in/mi (1.7 m/km) per segment, collector Partial depth remove and replace(b) Joint cracking < 60% segment length, 0.25 in. (6 mm) to 1 in. (25 mm) wide Crack sealing None > 1 in. (25 mm) wide Full-depth crack repair (a) 1,640 ft (500 m) segment length. (b) Contractor has option of financial payment in lieu of repair. Table 16. Ontario MOT asphalt pavement warranty requirements (Lee et al. 2016, Lane 2018). Potential Benefits Comment Quality of Work • Improved quality may include end-product performance and workmanship • End-product performance will vary depending on how performance is defined • Quality of workmanship dependent on contractor’s attention to detail Transfer Accountability or Risk • Motivate contractors to be more conscious about quality • Release some of the agency’s burden Reduce Life-Cycle Cost • Agencies may identify cost benefits by performing a net present value life-cycle cost analysis that accounts for initial costs and discounted future costs Promote Innovation • Usually this is achieved in performance-based warranty projects where most of responsibility is shifted to the contractor allowing the contractor more flexibility for innovative construction and use of alternative contract types Reduce Premature Failure • Materials and workmanship warranty projects can prevent early or premature failures due to poor contractor performance and quality control Table 15. Potential warranty benefits (Scott et al. 2011, Sadeghi et al. 2016).

22 Performance-Based Pavement Warranty Practices Summary The primary aspects of pavement warranties include: • Identify warranty type. Warranty types include materials and workmanship or performance, depending on the agencies’ expected outcomes. Materials and workmanship warranties require corrective action for pavement defects caused by contractor materials and application issues. Performance warranties include short-term warranties that typically transfer the mix design responsibility from the agency to the contractor, as well as long-term warranties that include high-level performance criteria and threshold limits to be met over the warranty period. • Select warranty period. Warranty periods can range from 2 to 20 years, and typically include 1 to 3 years for pavement preservation, 3 to 5 years for asphalt pavements, 5 to 10 years for concrete pavements, and 20 or more years for long-term warranties. • Define performance indicators and threshold limits. Pavement performance indicators generally relate to structural capacity, functional performance, and safety. The most common performance indicators for asphalt pavements include rutting, transverse cracking, fatigue cracking, raveling, longitudinal cracking, and bleeding. For concrete pavements, the most common performance indicators include corner breaks, longitudinal cracking, faulting, joint seal damage, spalling, and transverse cracking. Functional performance indicators typically include IRI and friction number. • Identify project selection criteria. Although agencies’ project selection criteria vary, two of the primary characteristics for pavement warranty projects include new construction and those with well-defined scope (e.g., existing condition, construction quality requirements and criteria, performance indicators and thresholds). • Specification elements. In addition to the above, other specification elements include warranty bond requirements, maintenance responsibilities, dispute resolution processes, acceptance criteria, methods of measurement, and payment.