National Academies Press: OpenBook
« Previous: Chapter 1 - Background
Page 5
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 5
Page 6
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 6
Page 7
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 7
Page 8
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 8
Page 9
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 9
Page 10
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 10
Page 11
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 11
Page 12
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 12
Page 13
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 13
Page 14
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 14
Page 15
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 15
Page 16
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 16
Page 17
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 17
Page 18
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 18
Page 19
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 19
Page 20
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 20
Page 21
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 21
Page 22
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 22
Page 23
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 23
Page 24
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 24
Page 25
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 25
Page 26
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 26
Page 27
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 27
Page 28
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 28
Page 29
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 29
Page 30
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 30
Page 31
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 31
Page 32
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 32
Page 33
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 33
Page 34
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 34
Page 35
Suggested Citation:"Chapter 2 - Research Findings." National Academies of Sciences, Engineering, and Medicine. 2011. Guidelines for the Use of Pavement Warranties on Highway Construction Projects. Washington, DC: The National Academies Press. doi: 10.17226/14554.
×
Page 35

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.

5The first phase of this project involved the collection and examination of the existing literature, critical discussions with practitioners in a workshop forum, and conducting selected DOT interviews to capture the current status of warranty use. The collected literature included national research studies, state-level evaluations, and warranty specifications. These state- level evaluations and selected pavement warranty specifications can be found on the FHWA’s National Highway Specifications website at www.specs.fhwa.dot.gov. The findings from these sources formed the basis for the final guidelines and model specifications developed under this research effort and are documented in the following sections of this chapter. General Literature (Comparison of National and International Experience) The FHWA-sponsored International Technology Scans generally concluded that the use of pavement warranties is entrenched in the international contracting model. European agencies began implementing warranties for a variety of reasons, including roadway maintenance needs, resource shortages, and the desire to improve quality and efficiency. This model uses a much higher degree of control by industry, and the warranty clause acts as a risk mitigation tool or an assurance to the agency that the pavement will perform. Furthermore, many European contracts are awarded on a best-value basis, using qualifications and price, in contrast to the U.S. practice of low-bid award. While the scans found that quality improve- ments can be linked to the application of a warranty, in reality, it is also probable that the sum of all European business practices produces a higher-quality product (AASHTO et al., 1991; Hancher, 1994; Bower et al., 2003; FHWA, 2005). Warranty practice continues to evolve in Europe. The 2002 scan reported that many of the host countries were experi- menting with alternative delivery methods that included operation and maintenance of facilities as an extension of the construction contract, in essence requiring longer-term guarantees. For example, the United Kingdom stated that its use of performance warranties had grown because design–build had become a contracting method of choice in the last decade and the addition of operation and maintenance was a natural extension. Both the United Kingdom and Spain have turned to the private sector to perform pavement maintenance or design–build–operate–maintain contracts with finance options. Warranty periods on European transportation projects range from 1 year for materials and workmanship warranties to 30 years for performance warranties under design–build– operate–warrant contracts, as shown in Figure 1. The 2002 scan report recommended the continued imple- mentation of warranties in the United States. The recom- mendations included promulgating legislation for best-value and prequalification procurement methods that incorporate quality and other technical factors in contract award and foster collaboration among federal agencies, DOTs, and industry. Warranty use in U.S. highway construction has grown, albeit at a slower rate more recently. The progression in the use of warranty contracting over the past 12 years is drawn from synthesis and research reports published in 1994, 1998, and 2002. Research published in 1998 indicated that several DOTs implemented warranty projects in 1996 and 1997, following FHWA’s publication of the final rule on warranty contracting. Figure 2 illustrates the spike in the number of warranty projects after publication of the final rule. These reports and the national survey performed under NCHRP Project 20-07/Task 201, “Use of Warranties in Highway Construction,” connected with this research project showed that the total number of pavement warranty projects completed by 2006 was more than 2,150. Table 1 shows that the majority of pavement projects are concentrated in certain states. Of the 24 DOTs that have implemented pavement war- ranties, several of those noted in Table 1 gained significant warranty experience where pilot programs evolved into C H A P T E R 2 Research Findings

6Source: Bower et al., 2003 Figure 1. Warranty periods reported in the 2002 European asphalt pavement warranties scan. 0 20 40 60 80 100 120 140 N o. o f P ro jec ts 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 Year Source: Anderson and Russell, 1998 Figure 2. Number of warranty projects constructed per year. standard practice, while many others have not carried their warranty programs beyond pilot implementation. Although significant warranty experience is concentrated in a relatively small number of DOTs, several others now using warranties on a limited basis reportedly plan to expand their programs. Joint DOT/Industry Pavement Warranty Workshop The research team conducted a two-day workshop on March 2–3, 2006, to collect information and insights related to current warranty practices within and outside of the United States. The participants included industry representatives from paving associations, domestic and international contractors, materials suppliers, the Associated General Contractors of America (AGC), owner representatives from six DOTs with warranty experience, and FHWA pavement warranty special- ists. The discussion addressed several topic areas, including the definitions of basic warranty types, the transition from materials and workmanship to performance warranties, what performance parameters and distresses are or should be under the contractor’s control, the benefits of warranty use, factors to consider in project selection and guidelines, and imple- mentation strategies. The list of workshop attendees and its meeting minutes can be found in Appendix C1. Warranty Types The workshop participants first considered and discussed warranty types/classifications and definitions. In general, warranty provisions have been implemented in conjunction with both method specifications under traditional design– bid–build delivery and performance specifications under design–build project delivery. Practitioners classified warranty provisions under one of three general categories: materials and workmanship, short-

term performance, and long-term performance. The FHWA used these general categories as part of the development of their selection procedures for pavement warranties. However, since these terms are not specifically codified in federal regulatory language, the researchers found significant latitude in how DOTs interpret these warranty types. A provision characterized as a materials and workmanship warranty by one DOT may be defined as a performance warranty in another, making clear distinctions between these three types of warranties difficult. There was considerable discussion among the participants concerning whether to exclude materials and workmanship warranties from the classifications because they were part of standard practices for many DOTs. The consensus among the group was that materials and workmanship should be included in the warranty classifications and guidelines because they differ from standard boilerplate code and commercial code requirements and provide additional protection under the bond or guarantee clauses. When the DOT survey evaluations were conducted as part of this research effort, classifications of warranty projects characterized by DOTs in terms of materials and workmanship and short-term or long-term performance were retained; however, the following classifications for these three types of warranties are offered based on the preponder- ance of the warranty specifications reviewed and the general consensus of the workshop participants. Type 1 Warranties Type 1 (materials and workmanship) warranties are imple- mented in conjunction with standard method specifications. Type 1 warranties require the contractor to correct early defects in the pavement caused by elements within the con- tractor’s control, namely the materials and workmanship of construction. The DOT uses a traditional, low-bid contract where the contractor assumes minimal performance risk. As shown in Figure 3, the contractor’s involvement is typically limited to construction and a small portion of the maintenance 7 DOT Warranted Pavement Components Years of Experience Total Number of Projects Michigan Pavement Pavement preservation Pavement markings 10+ 10+ 10+ 1,000+ (Standard practice; breakdown by type not available) Florida Pavement Pavement marking 3 2 700+ (Standard practice) 10 Ohio Pavement Pavement preservation Pavement marking 6 6 3 156+ 33+ 44+ +08 +51 tnemevaP nisnocsiW 72 5 tnemevaP sionillI California Pavement Pavement preservation 5 4 10+ 12+ Minnesota Pavement Pavement preservation Pavement markings 4 2 3 3 3 3 20 1 4 4 4 2 Colorado Pavement Pavement preservation 7 9 15 1 Mississippi Pavement Bridge deck overlays 4 3 11+ 3 Indiana Pavement Pavement preservation 10+ 3 10 2 Source: Anderson et al., 2006 Table 1. DOTs with the most significant pavement warranty experience. Figure 3. Type 1 (materials and workmanship) warranty.

period, and the warranty durations are relatively short-term, typically three years or less. Type 2 Warranties Type 2 (short-term performance) warranties shift more responsibility to the contractor for certain aspects of pavement performance during the warranty period. They have been implemented under traditional low-bid or alternative design–build contracts. Type 2 warranties are the broadest category of warranties since the amount of respon- sibility shifted to the contractor can range from design of the mix to design of structural aspects of the pavement, particularly when combined with a nontraditional design–build contract. As shown in Figure 4, the contractor may have some degree of responsibility for design and construction and a greater involvement in the maintenance period. The provisions of the warranty include elements of both method and perfor- mance specifications but can vary between being predomi- nately method-based or performance-based. Type 2 warranty durations generally fall within the range of 5 to 10 years. Type 3 Warranties Type 3 (long-term performance) warranties shift the responsibility for the long-term performance of the pavement to the contractor. Type 3 warranties typically use higher-level performance criteria, establishing pavement performance standards or thresholds that the contractor must maintain for the service life of the pavement or beyond, and include planned and unplanned maintenance. As shown in Figure 5, the contractor involvement extends from design through construction and includes planned maintenance. They are implemented under alternative design–build–warrant, performance-based main- tenance, or public–private partnership (PPP) or concessionaire agreements and typically are 20 years or longer in duration. Transitioning from Materials and Workmanship to Performance Warranties The workshop roundtable considered the range of options for a DOT to consider when transitioning from materials and workmanship to performance warranty types in the guidelines. The options considered were for mix quality management, mix constituents and design, structural design, drainage, equipment usage, manufacturers’ suggested practices, and traffic considerations. For example, when transitioning to a performance warranty, the possible options for quality management range from a conventional contractor quality control (QC) and owner quality assurance (QA) plan, to a reduction of owner QA where the owner would rely substantially on warranty provisions and records/data provided by the contractor, to the elimina- tion of owner QA requirements for construction where the contractor would be responsible for quality management and the DOT would rely on long-term warranty provisions to ensure quality. In terms of mix constituents and mix design options, gen- erally the mix design and methodology can be listed as start- ing points with progressively more allowance to deviate from them, depending on the length of the performance warranty. Substitutions for approved material or additives or admixtures are allowed with the understanding that the contractor can deviate from the list with positive performance proof offered. However, aggregate properties that may affect long-term per- formance and will not be detected during the warranty period (for example alkali silica reactivity or moisture sensitivity) may still need to be prescribed by the owner. In terms of selection of mixes and structural design/lift thickness, the workshop participants concurred that an owner needs to provide or allow progressively more access for the contractor to determine the pre-existing condition of the foundation when transitioning to a performance warranty. Obviously, there are other factors for thickness that the owner must have control of—shoulders, clearances, and guide rails are other design features. In terms of equipment requirements, the owner will often prescribe detailed equipment specifications under a short- term materials and workmanship warranty. As one moves to a performance warranty, the owner may relax equipment requirements and still rely on QA testing to ensure that the contractor is meeting minimum quality standards, or give the 8 Figure 4. Type 2 (short-term performance) warranty. Figure 5. Type 3 (long-term performance) warranty.

contractor the flexibility to select and operate equipment and reduce QC/QA testing, relying instead on a longer-term warranty to ensure performance. Lastly, in terms of traffic control, the contractor must have progressively more freedom in determining the traffic phasing when transitioning from a materials and workmanship to a performance warranty in conjunction with alternative delivery or lane rental and cost-plus-time bidding criteria. Table 2 summarizes the perceived levels of flexibility when transitioning from a materials and workmanship warranty to a performance warranty in terms of quality management, responsibility for mix and structural design, equipment, and traffic control. The transition from materials and workmanship to perfor- mance specifications gives greater flexibility to the contractor to manage quality and be responsible for pavement design and construction in return for a more comprehensive longer- term warranty obligation. These transitions in flexibility are considered in the guidelines and specifications for this research based on the type of warranty selected for a pavement con- struction project. Contractor Control: What Performance and Related Distresses Can Generally Be Regarded as in the Contractor’s Control? The following tables summarize a roundtable discussion during the workshop on performance parameters that could be included in a performance warranty. Again, control depends on the flexibility and extent of design responsibility given to the contractor as noted previously. In assessing the likelihood of including a particular distress or performance parameter in a performance warranty, the workshop participants considered whether the distress was attributable to actions within the contractor’s control and measurable and quantifiable through tests or analysis. Tables 3 and 4 summarize the consensus of the participants regarding the likelihood that distresses will be measurable and quantifiable and under the contractor’s control for asphalt and concrete pavements, respectively. These discussions and remarks served as one of the sources of information for the development of guidelines for imple- menting asphalt and concrete pavement warranties presented in Appendix A1. Benefits of Warranties— What Is the Motivation? During the meeting, various potential benefits or drivers for the use of warranties were offered by the attendees. For example, DOT representatives expressed that Type I warranties may prevent “lemons,” or early, premature failure based on poor contractor performance or quality control. They also may force the contractor to pay more attention to details, thereby improving quality. In the same vein, warranties may actually result in getting the product that is specified and show that it is performing the way the owner wanted it to perform. Other commonly cited benefits were that warranties may reduce DOT inspection levels, allow for innovation in materials and processes, or reduce contractor claims. One DOT partici- pant expressed that using warranties in conjunction with time- based incentives (A+B bidding) may be the quality (or Q factor) that balances time incentives (A+B+Q bidding). DOT workshop participants stated that warranties may provide clarity in roles and responsibilities between owner and contractor, improve public relations with the traveling public and legislators, or allow states to earmark or secure funding for maintenance over the long-term. A caution was that tying up money for this long period could also be perceived as a negative The research team tested these assertions through inter- viewing key DOTs with pavement warranty experience. One of the interview topics was feedback from practitioners on what the perceived benefits were and whether they actually accrued and could be confirmed. Once these potential bene- fits were vetted through the interview process, the research team planned to use them as an initial step in the warranty decision tool. Process Offered for Owners to Define and Manage Performance The workshop participants offered thoughts on what important steps to consider in developing a process for owners to define and manage performance under warranty provisions. These steps include 1. Determine when distress and/or performance levels should be identified. The consensus was that depending on the warranty type, distresses and performance levels or thresh- olds should be identified at three points in the life of the pavement: the end of construction, the end of the warranty period, and the end of the design life. In some cases, it is useful to develop performance curves based on historic data that represent expected performance during the service life of the pavement. 2. Identify which distress and/or performance levels are sub- stantially under contractor control. These would be the basis for the distresses used in the warranty provisions. 3. Quantify the end-of-construction and beginning of war- ranty period distress and/or performance level. This includes the distress measurement technique, the frequency and repeatability of measurement, the tolerances, the tester qualifications, and the test equipment calibration. The 9

Quality Management Mix Design Mix Selection/Structural Design Equipment Traffic -Conventional contractor quality control with conventional owner quality assurance. These would be standards normally used in non-warranty owner work. -Meets broad parameters established by owner with mix design work done in accordance with established owner-specified mix design protocols—Marshall, Superpave® or ACI, for example. -Owner-prescribed mixes and thicknesses of each layer. -Owner describes minimum plant and equipment specifications. -Owner specifies traffic control plan. -Conventional contractor quality control with no owner quality assurance. The owner would rely substantially on the warranty provisions. -Meets broad parameters established by owner with mix design work done using protocols from other sources. -Contractor may have option to adjust mix selection and thickness for equivalent performance, assumed to be based on established owner design procedures. -Owner allows variation to these specifications, with some evidence that the variation has been validated. -Contractor proposes an alternative traffic control plan subject to owner approval under a design– build, lane rental, or A+B bidding system. -Nonconventional contractor quality plan with no owner quality assurance. Records are available to the owner. sedivorp rotcartnoC- independent certification that the equipment and plan meet various standards—that weights and measures are legal. -No owner requirements. Warranty provisions are sole measure of quality. -Meets broad parameters established by owner with mix work done with proprietary procedures. Based on the European model, this would require some evidence of performance, with in-service data that validates the mix performance. -Contractor responsible for full structural design using mechanistic design criteria to meet performance requirements. Long-term issues such as fatigue may require some input from the owner. -Contractor has full freedom to select and operate all equipment. -Contractor has complete responsibility for developing a performance-based traffic management plan. Table 2. Flexibility to allow contractor variation from traditional DOT requirements.

owner should be assured that all performance distresses are dealt with prior to the start of the warranty period. 4. Identify the in-service, in-warranty measurement of dis- tress and/or performance level. This includes the distress measurement technique, the frequency and repeatability of measurement, the tolerances, the tester qualifications, and the test equipment calibration. Determine a rational way to measure traffic if it is included as a limiting factor in performance. 5. Determine the in-service, in-warranty reporting periods and the approach to remedial action, including the approval of specific fixes. While remedial action was not discussed in depth, it is important to recognize that any remedial fix should return the pavement to the performance curve. Clearly fixes that return the pavement to new or like-new condition may actually extend the service life beyond the original design life. 6. Determine the requirements for the completion of the pavement warranty and final warranty acceptance. These may include a hand-back plan with requirements for inspections and testing to meet defined performance requirements for hand-back. 11 Performance Parameter/Distress Within Contractor’s Control Measurable & Quantifiable Remarks eht yfitnedi ot ysae yllareneg si tI Y *Y gnittuR severity and the extent of rutting of the surface course, but the cause may extend to lower layers of pavement not under the contractor’s control. Smoothness Y Y Owner can establish standards based on whether it is rehabilitation or new pavement versus the thickness of lifts. Transverse and longitudinal cracking Y* Y Cracking in the wheel path is usually related to loading but can extend to the base or foundations. Non–wheel- path cracking can be related to paving operations or other causes. dna ,elbaifitnauq dna elbarusaeM Y Y gnivohS generally under contractor control. fo ytilauq ro erutxet ecafrus ot setaleR Y Y noitcirF aggregate. Can be used as a performance measure if contractor can select aggregates according to state standards and the DOT gives the contractor the skid numbers and some historical data. Liability or indemnification may be an issue and may restrict use. Raveling Y N A common distress not easy to measure. Potholes Y* Y Other factors outside the contractor’s control may cause potholes. Delamination (slipping) Y* N Difficult to pinpoint but measurable with photos. Bleeding (spot sections) Y N Bleeding will show up in conjunction with poor friction or rutting values. Could have some visual clues that are hard to measure. Segregation Y* N New tools are coming online to identify and minimize segregation using heat sensors and thermo imaging. The effects of segregation would show up as another distress. Reflective cracking N N Difficult to determine whether reflective cracking is attributable to the contractor. Fatigue and moisture sensitivity N N Almost never used in a performance warranty since the distress would appear well beyond a 5 to 7 year performance standard. *Forensic study may be necessary to determine whether distress is under contractor control. Table 3. Assessment of HMA pavement performance parameters.

Considerations in the Development of the Selection Guidelines The workshop participants identified several key charac- teristics that should be considered when selecting candidate projects for a warranty provision. The first consideration is using projects with sections/designs that have a performance history that can be used as a basis for selecting reasonable distress thresholds. Both the industry and the DOT need to buy into or believe in the reliability of this performance history. The second consideration is project size or complexity. Based on the current Colorado DOT procedures, the length of the project should be at least 3 miles. This would require a dedicated plant. Another consideration is the scope of work. If the primary scope of the project is paving, the prime con- tractor is directly responsible for the warranty. If the scope of paving is subcontracted, a pass-through warranty can occur. The prime contractor’s responsibility for a pass-through warranty and how the bond is structured must be carefully considered by the DOT. In terms of a procurement method, the design–bid–build approach will work but limits the ability of the contractor to control the design or performance outcome. It may be better to use prequalification and best-value under a design–bid–build system or move to design–build to give the contractor greater control for longer-term performance warranties. In terms of risks, they should be considered in the selec- tion. If they are estimated as high and will impact the bid appreciably, then the DOT should think twice about using a warranty. For example, foundation condition is one of the most significant risk issues and needs to be addressed. If pos- sible, the DOT should consider using the design–build concept to address contractor control of the foundation design. If failures caused by the foundation are not the contractor’s responsibility, the warranty would not be invoked. When considering warranties for concrete pavements, to reduce risk the design life should be 20 years or more and the minimum PCC pavement thickness should be over 9 in. Implementation Strategies Finally, the workshop practitioners commented on strategies that proved to be successful when implementing warranties. The first recommendation was to progress from simpler to more complex warranty projects. This may entail a transition plan from simpler to more complex longer-term warranties as DOTs gain more experience. Partnership with industry in the development and implementation of warranties is a key to success. As part of this partnership, pre-bid and constructability meetings should be held with industry to get input on the project. Based on this input, some projects could be weeded out. Also, to avoid issues related to excessive 12 Performance Parameter/Distress Within Contractor’s Control Measurable & Quantifiable Remarks Smoothness Y Y Can establish standards based on rehabilitation versus new pavements versus thickness of lifts. Transverse and longitudinal cracking Y* Y Are measurable and quantifiable. Transverse cracks may be temperature induced. Joints, corner distress Y* Y Can be measured and may have multiple causes. fi lortnoc s’rotcartnoc rednU Y Y noitcirF contractor can select aggregates according to state standards and the DOT gives the contractor the skid numbers and some historical data. Liability or indemnification may be an issue. Joint sealant Y* Y Can be used if there is sufficient data on product history to set tolerances. Complications may arise related to manufacturer installation requirements. Durability (D) cracking N N Doesn’t show up early. A long-term durability issue. Alkali silica reactivity N N Almost never used in a performance warranty since the distress would appear well beyond a 5 to 10 year performance period. *Forensic study may be necessary to determine whether distress is under contractor control. Table 4. Assessment of PCC pavement performance parameters.

equivalent single axle loads (ESALs), a weigh-in-motion station should be included in or near the project unless a current station exists in the vicinity, and the cost benefit of weigh-in- motion should be carefully evaluated. DOT Pavement Warranty Experience As noted in the objectives and research approach section, the next major activity involved surveying a representative cross-section of DOTs with varying levels of pavement warranty experience to identify the key factors in the warranty decision process, warranty outcomes, issues, and keys to successful implementation. The selected DOTs included a geographical cross-section of DOTs with significant pavement warranty experience at the time of the survey, as well as two control DOTs with no warranty experience and with no plans to implement them. A survey form was used to guide the telephone inter- views with selected DOT officials. A list of the DOTs, inter- view participants, and the form used during these interviews is included in Appendix C2. The research team included eight topic areas on the interview form, as summarized in Table 5. The results of interviews are captured in Appendix C3. In addition to the interviews, additional information on DOT practices was drawn from warranty evaluation reports or national research reports from these DOTs, as well as selected warranty specifications and guidelines. California DOT (Caltrans) Between 1993 and 1999, Caltrans implemented approxi- mately six rubberized asphalt concrete and chip seal warranty projects of varying length (1 to 5 years), most of which were emergency projects as a result of severe storm seasons that the state experienced between 1994 and 1998. In 2000, a formal pilot warranty program was initiated by Caltrans’ divisions of construction and maintenance to evaluate 1-year materials and workmanship warranties on a variety of pavement seal and ultrathin overlay projects, including asphalt rubber chip seals, polymer modified emulsion chip seals, microsurface seals, conventional asphalt concrete overlays, rubberized asphalt concrete overlays, and bonded wearing courses. Caltrans’ warranty specifications are structured to allow for excluded areas, meaning certain sections of the pavement project may be exempt from the warranty based on the results of a condition survey of the existing pavement. Existing conditions that may qualify a pavement section for exclusion include rutting greater than 9 mm, patches of cold mix asphalt concrete placed within the last 12 months, existing cracks greater than 6 mm, and existing cracks filled with emulsified filler within the last 12 months or hot applied filler within the last 4 months. If the total excluded areas amount to 15% or more of the total project, then Caltrans typically does not consider the project a good candidate for a warranty. Caltrans does not require a separate bond for the warranty period. The bonding industry agreed to extend performance warranties for one year beyond the project completion date in exchange for a warranty bond. Therefore, bonding issues do not negatively influence Caltrans’ warranty program. Based on an evaluation of the nearly 30 warranty projects constructed between 2000 and 2006, Caltrans concluded that (Cotey and Jones, 2005 and 2006): • Cost comparisons of five warranted asphalt pavement projects with the same non-warranted asphalt pavement types/applications resulted in lower costs per mile (6% to 16% lower) for four out of the five pavement types evaluated. The lower costs applied to rubberized asphalt concrete open graded high binder, polymer modified chip seals, dense graded asphalt concrete, and bonded wearing course. 13 Distress indicators and thresholds Identify typical and potential performance values that could be established for warranties and identify primary and secondary causes of distress evitcejbO tnioP noissucsiD General information Understand the background of pavement warranties in the DOT slairetam neewteb sehsiugnitsid TOD eht woh dnatsrednU ytnarraw fo epyT and workmanship and performance Guidelines and implementation Understand how guideline and implementation procedures are determined and conveyed to others etadidnac gninimreted rof ygolodohtem eht dnatsrednU noitceles tcejorP warranty projects Effects on contracting process Understand the effect of warranty on procurement, competition, and construction Warranty management Understand the responsibility of each party during the warranty period Benefits and future considerations Determine the perceived and/or documented benefits of implementing warranty projects Table 5. Summary of DOT pavement warranty interview forms.

• Contractors appear to apply extra effort during construction and pay more attention to binder temperature, binder spread rate, rock spread rate, and rock temperature. • One contractor took extra effort with materials by truck- ing in rock from a distance to ensure that the project would perform for the warranty period. • Bids were competitive. • Materials supplied by contractors meet specifications. • The warranty period provides added time for field mainte- nance personnel to observe performance of the pavement application. Caltrans reported that overall the highway surface treat- ments on the pilot projects were performing well. It plans for continued use with certain recommendations, several of which were aimed at specification revisions (Cotey and Jones, 2005 and 2006). One key recommendation was to revise the warranty specifications to allow contractors to determine the spread rate for the chip seal and slurry seal warranties. Another key recommendation was to allow contractors to propose alternative repair methods that would result in equal or better performance. These recommendations suggest that the DOT is moving toward more performance-based requirements as it expands the warranty program. Colorado DOT (CDOT) CDOT has experience with materials and workmanship and performance warranties, having implemented 3-, 5-, and 10-year HMA pavement warranty projects and 5- and 10-year PCC pavement projects. CDOT commenced its warranty program in 1997 as a result of Senate Bill 97-128, a legislative mandate requiring the implementation of three warranty pilot projects by the end of 1998. The first three warranties imple- mented were 3-year HMA pavement projects. After completing the pilot projects, CDOT revised the warranty specification and constructed three more HMA warranty projects before publishing its first evaluation report in December 2001. This report evaluated 3-year warranty projects and concluded that the cost–benefit evidence did not support continuing or stopping the 3-year warranty pilot program. Rather, the report proposed some changes to the warranty specification based on lessons learned and suggested further research on the cost–benefit of 3-year warranties and new research on the benefits of the 5-year and 10-year warranties. CDOT’s initial 3- and 5-year materials and workmanship asphalt warranty specification was less prescriptive than the standard specification, with the DOT specifying the pavement thickness and structural design and shifting the responsibility for mix design and placement to the contractor. Subsequent versions added prescriptive requirements for longitudinal joints, smoothness, and paving limitations for weather, as well as additional instructions for quality control. Under the 10-year pavement warranties, contractors are responsible for more aspects of the pavement design, including the mix, structural thickness, and vertical alignment. All of CDOT’s warranty provisions included a weigh-in- motion (WIM) station to track ESALs during the warranty period. Additionally, CDOT’s initial provisions required that warranty monitoring be performed by a pavement evaluation team (PET) consisting of three individuals not directly involved in the project, including one CDOT representative, one private consultant, and one industry representative. Almost all of the initial cost increases in the first warranty projects could be attributed to either the WIM or the PET. Subsequent versions of the specification allocated responsibility for monitoring to the CDOT representative on the PET rather than the three-member PET. All of CDOT’s warranty provisions included lane rental provisions for work performed during the warranty period; however, the 10-year provisions allow the contractors a specified number of days per year to perform work without lane rental fees. This provides an incentive for contractors to perform preventative maintenance without penalty. Interviews with representatives of CDOT revealed the following considerations regarding the warranty program: • The decision to apply a warranty to a particular project was made at the district level by district engineers; • Industry was very involved with the development and appli- cation of the warranty (e.g., pre-advertisement meetings, pre-bid meetings); • In addition to historical pavement management data, CDOT used model projects to help establish thresholds used in provisions and invited industry representatives to visit the model projects to give them a feel for the level of service required under the warranty provisions; and • CDOT established separate guidelines for applying a 3-year hot bituminous pavement (HBP) warranty, 5-year HBP warranty, 5-year portland cement concrete pavement (PCCP) warranty, and 10-year HBP or PCCP warranty. A cost–benefit analysis of short-term (3- and 5-year) warranties published in 2007 concluded that the use of 3- and 5-year HMA warranties was not a cost-effective tool for CDOT (Goldbaum and Aschenbrener, 2007). This conclusion was based on the evaluation of 10 warranty and control project pairs constructed between 1998 and 2003. CDOT determined that only one project resulted in improved performance at an overall cost savings. Table 6 summarizes the results of the CDOT cost–benefit evaluation. CDOT concluded that the average initial construction cost of the warranty project was $5,318 per lane mile more than the control projects. CDOT indicated that this amount could be 14

reduced by approximately $2,573 if CDOT eliminated the PET and the WIM station. It should be noted, however, that CDOT did revise the warranty provisions over the course of implementing these 10 projects, and a learning curve should be anticipated with each revision. Additionally, while every effort was made to match similar control projects to warranty projects for comparative purposes, in most cases the reha- bilitation strategies did not match exactly, which may have contributed to the differences in the performance data. Table 7 shows the rehabilitation strategies on the warranty and the control projects. Florida DOT (FDOT) FDOT began applying a 3-year materials and workman- ship warranty as standard on all pavement projects in 2004. FDOT refers to these pavements as value-added pavements. The state’s pavement projects are largely HMA pavements (94% asphalt). FDOT noted that HMA pavement is preferred over PCC pavement because there have been some significant failures in concrete pavements. In the specification development stage, FDOT organized a team consisting of FDOT and industry representatives in order to build consensus during the specification development stage. FDOT’s Type 1 warranty specifications are very prescriptive. In some cases, contractors may be given some flexibility in the mix ratios; however, FDOT noted that there is little dif- ference in the way the warranty project is constructed compared to projects under the Contractor Quality Control/Quality Assurance Program, which was implemented as a precursor to warranties. FDOT reported that there was no apparent difference in the price of the warranty project due to the prescriptive nature of the warranty provision and the elimination of the warranty bond. In lieu of a bond, FDOT uses prequalification to guar- antee the warranty. If a contractor fails to perform a warranty repair, the contractor is precluded from bidding on other work for a period of 6 months, or until the repair issue is resolved, whichever is longer. FDOT established a statewide Disputes Review Board (DRB) dedicated to resolving pavement warranty disputes that cannot be resolved at the district level. This DRB consists of representatives of academia, industry, and the FDOT. As of this reporting, there have been four failures, two of which were repaired by the contractors before taking it to the DRB. One dispute is currently being evaluated by the DRB. 15 Warranty Project Years of Performance Data Overall Cost Savings Overall Improved Performance oN oN 8 niatnuoF ta 52-I seY oN 8 *eF atnaS ta 074-C oN oN 8 roirepuS ta 63-SU I-25 North of Pueblo 6 No No seY seY 6 elgaE ta 07-I US-50 and Kannah Creek 4 No No oN oN 4 67-I ta 36-HS oN seY 4 noxiN yaR ta 52-I oN denimreted toN 3 sreyB ta 63-SU US-287 at Ted’s Place 3 Not determined Yes *In the sixth year after the warranty project was constructed, it needed minor rehabilitation and as a result, its performance was better than the control project after 8 years. Source: Goldbaum and Aschenbrener, 2007 Table 6. CDOT warranty cost–benefit evaluation summary. Warranty Project Warranty Rehab Strategy Control Rehab Strategy I-25 at Fountain 1”(NB) & 2.5”(SB) milling 2” milling C-470 at Santa Fe 0.5” milling 2” milling US-36 at Superior 1” milling 0.75” leveling course I-25 North of Pueblo 0.75” milling 2” milling I-70 at Eagle 1” leveling course Milling/recondition base/overlay US-50 and Kannah Creek Reconstruction/widening New construction/widening SH-63 at I-76 Full depth reclamation Full depth reclamation I-25 at Ray Nixon 2” milling 4” milling US-36 at Byers 4” cold recycle 4” cold recycle US-287 at Ted’s Place 0.75” leveling course 2” to 4” milling Source: Goldbaum and Aschenbrener, 2007 Table 7. CDOT rehabilitation strategies.

FDOT concluded that while it was still too early to conduct formal evaluations, the small number of disputes and callbacks indicated that the value-added warranty pavements are meet- ing expectations and protecting against early failures. FDOT noted that they are considering extending the term of the warranty by allowing contractors to bid the warranty in a best-value procurement, awarding higher scores for longer warranties; however, FDOT noted that they must develop performance-based requirements before significantly extend- ing the warranty. Illinois DOT (IDOT) IDOT began implementing warranties as a result of a 1999 legislative mandate requiring a 5-year warranty on at least 20 highway construction projects let between 2000 and 2004. The mandate required that at least 10 of the projects be designed for a 30-year design life. IDOT implemented 27 projects in total under this pilot program, including 3 bituminous concrete overlay projects, 8 bituminous concrete projects, and 16 con- crete pavement projects. All projects were Type 1 warranties, with IDOT retaining design, material selection, and construc- tion methods. IDOT noted that candidate project selection was largely influenced by the design life of the pavement, but other desirable candidates included projects that were simple in scope, had no unique designs, and did not incorporate any controversial design methods or materials. The warranty specifications were designed to work as an add-on without much change to the standard specifications since the goal of the program was to protect the department against premature failures due to inferior materials or work- manship. A committee consisting of representatives from several branches of IDOT and a representative of FHWA developed the specification. This committee examined pro- grams at Indiana, Wisconsin, and Michigan and also held a joint work group to allow representatives of the contracting and surety industry to review and comment on the draft specifi- cations. Historical data from the Illinois Pavement Feedback System (IPFS) and Illinois Roadway Information System (IRIS) were primarily used to select performance parameters, but the decision not to warrant reflective cracking on the overlay projects was made as a result of resistance from the industry work groups and acknowledgement that this type of distress may not be within the control of the contractor in an overlay situation. Thresholds and corrective action were determined based on statistical analysis and on the extent and severity definitions of distresses found in the Distress Identification Manual for the Long-Term Pavement Performance Program (Miller and Bellinger, 2003). IDOT structured the warranty as a separate line item in the bid documents. A preliminary bid comparison revealed that the cost of the warranty ranged from 0% to 2.38% of the total contract price. IDOT noted that even with a separate line item for the warranty, it was difficult to quantify the cost of the warranty because the large cost variation in the warranty line item for the 27 projects implemented in Illinois suggests that contractors may have distributed the true cost of the warranty to other line items in their bids (Wienrank, 2004). Based on a survey of resident engineers involved in overseeing warranty projects, only one responded that the contractor paid closer attention to construction details as a result of the warranty. Resident engineers noted that on most projects, construction was completed no differently than on standard projects. Preliminary remarks based on a bid review and anecdotal observations indicated that the warranty projects cost slightly more for about the same level of performance, and that longer periods and performance requirements were needed to realize the full benefit of the warranties. Warranties are monitored at the district level based primarily on surveys conducted as part of the pavement management data collection process. District personnel review digital images and data collected by the data collection vehicles and determine whether additional surveys and inspections are necessary. IDOT reported that this is an acceptable workload for district engineers since each district has only a few warranty projects, but as that number grows, dedicated personnel would be necessary to manage all the warranties. IDOT published an evaluation report in 2004; however, since none of the projects had reached the end of their warranty period and several were still under construction at that time, it was recommended that a decision on whether to continue the use of warranties be deferred until more performance data became available (Wienrank, 2004). Indiana DOT (INDOT) INDOT’s experience includes 5-year HMA and PPC pavement warranties. The first warranty project was let in 1996, and approximately 15 have been constructed to date. INDOT’s warranty projects incorporate the concepts of contractor QC/QA, cost-plus-time bidding, and lane rental. INDOT applied this combination of time- and quality-based contracting methods to major projects in high-traffic areas where the goal of the project is to produce high quality in the shortest construction time frame possible. INDOT classifies these warranties as performance warranties since the contractor is responsible for the job mix design while INDOT specifies the structural design and minimum aggregate and binder mixture requirements. INDOT published a report in 2003 evaluating the perfor- mance of approximately 7 HMA projects (involving HMA overlays on rubblized PCC or crack and seat PCC) constructed between 1996 and 2001 (Flora, Gallivan, and Huber, 2003). A subset of non-warranty HMA pavement projects completed 16

in the last 4 to 6 years was selected to be used in a compara- tive analysis of average values for rutting and International Roughness Index (IRI) on warranty and non-warranty projects. The results of the analysis were that the warranty projects had less rutting and a lower IRI than the non-warranty projects. The analysis concluded that the warranty would extend the performance of the pavement 9 years at an average initial cost increase of 10%, resulting in a 25-year life-cycle cost savings of 27%. Interviews with INDOT revealed that it worked closely with industry representatives during the development of the warranty specifications. Representatives of INDOT and the local industry collaborated to develop a white paper on warranty use. A list of recommended materials with a history of performing well was provided to aid contractors in devel- oping job mix formulas. Once warranty thresholds for IRI were established, INDOT met with industry representatives for a demonstration on these thresholds. The demonstration included a tour of pavements of various IRI values to aid the representatives in connecting the IRI threshold to a visible example of level of service. In some cases, INDOT modified the warranty provision based on the scope of the project. For example, on crack and seat projects, transverse cracking was removed as a performance indicator since transverse cracking would be expected based on the existing conditions. INDOT’s warranty monitoring system is tied to its exist- ing pavement management system, which is performed by an outside consultant, to mitigate the administrative burden of warranty monitoring on the agency. An INDOT repre- sentative reviews pictures taken by the consultant, and if visual defects are present, additional inspections may be scheduled. A separate verification inspection is performed by INDOT near the end of the warranty period for close-out purposes. INDOT indicated that internally there are mixed opinions about extending the warranty beyond 5 years, and further evaluation is needed to compare the initial costs and assess the potential for lowering life-cycle costs through an extended warranty. Iowa DOT Iowa DOT has no experience using warranties, and in fact, has expressed opposition to the use of warranties as a solution for improving quality on highway projects. It should be noted that PCC is the predominant pavement type in Iowa. The concerns raised by representative of Iowa DOT regarding warranties are summarized as follows (Grove, 2005): • Iowa DOT is better suited than individual contractors to assume the risk associated with long-term performance of pavements. • Warranties have the potential to reduce competition by precluding contractors that do not have extra bonding capacity available to maintain a warranty bond for the length of the warranty period. • Warranties only ensure quality through the warranty period, not the length of the design period. • Shifting responsibility for quality of the pavement to con- tractors will not automatically make pavement quality the number one priority of contractors. • Warranties do not promote innovation because they require contractors to minimize risk, which is accomplished by relying on what has been proven to work. Iowa DOT is better suited to shoulder the risk associated with innovation. • The myriad of factors beyond the paving contractor’s con- trol, such as grade conditions, traffic volumes, changes in legal truck load limits after the project is completed, and agency-controlled maintenance (or lack thereof) during the warranty period makes warranties highly susceptible to disputes. • Assessments of agency personnel savings on warranty projects must account for the significant startup cost to develop an effective warranty specification as well as the effort required to monitor warranties. • A better understanding of the tools necessary to monitor concrete pavement construction and predict performance is necessary before that risk can be transferred to contactors without exposing Iowa DOT to liability for that risk. Kentucky Transportation Cabinet (KYTC) KYTC applied warranties under a multi-parameter bid process that factored price, construction time, and warranty length into the contract award. The specification was developed for use with either HMA or PCC pavement. This warranty provision gave contractors the option to bid a warranty length ranging from 5 to 10 years, and bid values were adjusted for evaluation based on a pre-specified credit for the length of warranty proposed. KYTC specified minimum required structural thicknesses for both designs (HMA and PCC) and shifted responsibility for job mix formulas to the contractor under this warranty provision. The provision also specified separate threshold requirements for HMA and PCC. Louisiana Department of Transportation and Development (LaDOTD) LaDOTD’s warranty experience includes two 3-year HMA projects and one 3-year PCC project, implemented as a result of a legislative mandate. All three projects involved new construction and included a Type 1 warranty. LaDOTD did not publish a formal evaluation report on these projects; however, an interview with LaDOTD 17

resulted in the following conclusions concerning its warranty program: • The warranty period was largely determined by the indus- try’s willingness to accept a 3-year instead of a 5- or 7-year duration; • Standard inspection and testing responsibilities were carried out by the DOT on all warranty projects; • Performance indicators were chosen based on using engi- neering judgment to determine what defects would show within the first 3 years; • Thresholds for performance indicators were based on a statistical analysis of the pavement management system data; and • Initial impressions based on anecdotal observations are that the warranty projects cost more and are of equal quality to the non-warranted projects and were therefore not cost-effective. Michigan DOT (MDOT) MDOT is the most experienced agency implementing warranty contracting, having constructed over 1,000 warranty projects since 1997. MDOT implemented its warranty program as a result of a legislative mandate. MDOT has not yet con- ducted a formal evaluation of the warranty program, noting that quality data is not easy to correlate to the warranty given all the variables that can exist within a contract, and MDOT does not yet have enough data points for a fair, accurate per- formance review of the program. However, the consensus is that contractors pay more attention to quality issues as a result of the implementation of warranties (MDOT, 2005). Regarding project selection, the first criteria for candidate projects were high-volume roads. Warranties in Michigan are now applied as the standard practice for all paving projects. Exceptions are made in certain circumstances where the existing conditions justify that the warranty requirement should be waived. MDOT recognizes that it may not be able to apply the ideal repair in every rehabilitation situation due to limitations in funding. Therefore, the warranty requirement may be waived in situations where the project is scoped as a quick fix to buy more time until funding is available for the necessary larger rehabilitation. MDOT has developed a project selection tool that addresses when not to use a warranty based on where existing conditions cannot be addressed or the administrative costs would be high relative to the size and scope of the project (MDOT, 2002). Projects are classified into two categories: capital preventative maintenance (CPM) and rehabilitation and reconstruction (R&R). The majority of projects let each year are CPM projects, and MDOT has more HMA than PCC pavements. Projects included under the CPM program are • Cold milling and one-course hot-mix asphalt overlays, • HMA crack treatments, • Microsurfacing, • HMA ultrathin overlays, • Paver-placed surface seal, and • Single and double chip seals. Projects included under the R&R program are • New or reconstructed HMA, • Multiple course hot-mix asphalt overlays, • HMA placed on rubberized concrete, • HMA placed on crush and shaped base, and • New or reconstructed joint plain concrete pavement. Warranty periods are 3 years for CPM projects and 5 years for R&R projects; however, specifications for CPM projects are characterized as performance-based, while specifications for R&R are method-based, materials and workmanship warranties. While this does not appear consistent with typical warranty periods for materials and workmanship and per- formance warranties, the logic is consistent when considered in terms of design life. CPM projects have a much shorter design life than R&R projects. Therefore, a 3-year warranty on a CPM project covers a larger percentage of the design life than a 5-year warranty on an R&R project. For example, a 3-year micro-surface warranty covers approximately 60% of the design life, whereas a 5-year HMA pavement reconstruction warranty only covers approximately 20% of the design life. Therefore, MDOT determined that CPM project warranties should be performance-based, while R&R project warranties should be method-based. A warranty task force consisting of MDOT and industry representatives was established early in the warranty program. While MDOT had final decision authority for all aspects of the warranty provisions, the task force was consulted on durations, performance measurements, and thresholds. The warranty task force continues to meet on a semi-regular basis to discuss options for improving various aspects of the warranty program. MDOT has scaled back DOT-performed inspection and testing since the implementation of warranties, especially on performance warranty projects. Materials and workmanship warranty projects may have one inspector assigned to five projects at a given time, while performance warranties may not be inspected until final acceptance. Approximately 3% of all projects require remedial action, and there have been rel- atively few documented cases of disputes. Due to the size of the warranty program, the warranty management system is maintained separately from the pavement management system. MDOT noted that the cost increase on warranted projects is primarily attributable to the cost of the warranty bond. 18

FHWA has granted MDOT approval to use pass-through bonding because for 18% to 20% of MDOT projects, the prime contractor is not the paving contractor. In those cases, the prime contractor typically requires that the paving con- tractor carry additional bonding, and the MDOT may pay twice for the warranty bond. However, MDOT is also very interested in alternatives that would eliminate the warranty bond. MDOT noted that small- and medium-sized contractors have stopped bidding on warranty projects. It has also seen consolidation of asphalt companies, partly due to the use of warranties. With fewer companies able to participate, the bond- ing requirements may be reducing competition. Therefore, MDOT is looking into alternatives such as the guarantee model used in Florida and pooled risk insurance for contrac- tors, where contractors with better quality records would pay lower rates. MDOT does not typically include incentives for pavements that perform exceptionally well, but it is currently piloting a very small number of experimental projects where financial incentives are provided for superior pavement performance. MDOT outlined the keys to the successful implementation of warranties. They include • Good pavement management data, • Proper project scoping, including any necessary preliminary engineering (soil information, traffic information, etc.), • Pavement design method based on a widely accepted method (i.e., AASHTO) and providing for drainage of the pavement structure, • Performance measures linked to warranty length, • Performance thresholds based on real-life pavements, • Proper warranty administration, and • Contractor involvement in specification development. Minnesota DOT (MnDOT) MnDOT began incorporating warranties in the mid 1990s on both design–build and design–bid–build projects. MnDOT’s experience under design–bid–build warranties is primarily 2-year performance warranties on bituminous mill and overlay projects, but it has also developed a 5-year performance pro- vision for bituminous pavements. Contractors are responsible for developing the job mix formula under both the 2-year and the 5-year provisions. Under design–build projects, MnDOT has experience with both rigid and flexible pavement war- ranties ranging from 3 to 5 years. MnDOT’s innovative contracting summary stated that field personnel on projects with 2-year warranties identified no significant changes to the bituminous construction practices. It also stated that additional time was needed to assess the effectiveness of 5-year warranties on design–build projects (Johnson, 2004). MnDOT performed a cost analysis on its bituminous warranty projects to determine the cost increase on warranty projects. The analysis examined projects that had both war- ranty and non-warranty items with the same mix design and non-warranty projects with similar bituminous bid items that were let within three months of the warranty projects in the same districts. The analysis found no apparent trends to suggest that the unit prices for warranty items increased com- pared to non-warranty items; however, based on information from the MnDOT estimating unit, MnDOT concluded that contractors are including costs to address the warranty in the mobilization item (Johnson, 2004). MnDOT also considered implementing a 20-year warranty; however, it ultimately decided not to include the warranty due to several key issues that arose during the negotiation process, including the fact that the project was being constructed on excellent graded material, making subgrade failures less likely. Additionally, the maximum liability was capped at 1.5 times the warranty cost, and the warranty price included inflation of 3% to protect the warrantor against spikes in the bituminous cost. MnDOT decided that the implementation of the 20-year warranty would not be a cost effective tool. Mississippi DOT Mississippi DOT has experience with both HMA and PCC pavement warranties, having constructed approximately 13 HMA performance warranties and one PCC performance warranty since 2000. Mississippi DOT reported that initial bid increases on warranty projects were approximately 30%, but bid increases have leveled off at approximately 11% to 12%. Mississippi DOT developed a warranty program after receiving no bidders on a highway rehabilitation project. The project was split in half and re-bid as two projects, one as a traditional project and one as a warranty project. A case study on this project is included as Appendix C4. Mississippi DOT’s warranty projects are characterized as short-term performance warranties. Contractors are respon- sible for the mix design, while the DOT specifies the structural thickness. During construction, Mississippi DOT does not perform density checks in the field; however, it does verify smoothness and thickness through computerized profiling and coring. Price adjustments may be made based on the achieved smoothness and thickness. Warranty periods are 5 years for HMA and 10 years for PCC; however, Mississippi DOT noted that it plans to increase HMA warranties to 7 years, keeping the same thresholds, and it is also considering a reduced 5-year option for PCC. Mississippi DOT developed a software program (DEDUCT) that translates different distress thresholds for pavement performance measures into a common point system for pavement distress. The software program asks the user to 19

select a severity level for the distress based on the severity levels defined in the Distress Identification Manual for the Long-Term Pavement Performance Program (Miller and Bellinger, 2003). The program also asks for the extent of the distress in terms of percentage, length, surface area, or number relating the distress to the historical performance curve for that distress. The program then applies a point value for the distress based on the comparison of the distress level to the historical curve (Mississippi DOT, 2001). Mississippi DOT is traditionally an asphalt state, and only about half of the asphalt contractors are willing to bid on warranty projects. Most contractors attributed inability to obtain bonding as the reason for not participating on warranty projects. As a result, Mississippi DOT has seen some reduced competition in the asphalt industry as a result of warranties; however, the concrete industry has been supportive of war- ranties since they see it as an opportunity to increase the number of contracts let for concrete pavements. Mississippi DOT noted that the likelihood of obtaining competitive bids often plays a role in warranty project selection. Mississippi DOT conducts separate surveys for its warranty program and its pavement management program. Under the pavement management program, a survey of all roads in the DOT network is conducted every 2 years by an outside consultant. Under the warranty monitoring program, a survey is conducted every year on warranted pavements by a team from the research division. However, collection methods for the biannual survey and warranty survey are the same except for the collection of data on joint faulting. The biannual survey collects data using a laser profiler while the warranty survey collects this data using a Georgia fault meter. Mississippi DOT has not had any major disputes on warranty projects, and only one case in which the contractor was not liable for defects that occurred. In this case, a threshold for flushing was exceeded, but it was determined that the residue was actually being tracked by trucks coming from a nearby construction site. Mississippi has not published a formal evaluation report; however, comparative data and anecdotal evidence suggest warranties have resulted in improved performance. Anecdotal evidence of improvements include contractors consulting third parties on the mix design, using higher PG grading, and using better crews and equipment on warranted projects. Ohio DOT (ODOT) ODOT has the second highest level of experience using warranties, having implemented more than 200 warranties since 2000. ODOT began implementing warranties as a result of a legislative mandate requiring that a minimum of 20% of all projects include a warranty. The legislation was revised the following year, deleting the minimum requirement and substituting a maximum requirement for 20% of all projects to include a warranty. ODOT warranty periods are primarily 3-year for preventative asphalt treatments, 7-year for asphalt pavement, and 7-year for concrete pavements, though the ini- tial asphalt pavement projects were 5-year warranties. ODOT implements both materials and workmanship and performance warranties. Asphalt warranties are more performance-oriented, while concrete warranties are materials and workmanship. Ohio’s contractor association served on committees with DOT personnel to discuss and set performance indicators and thresholds; however, the legislation imposed a 6-month deadline on the timeline for the development of the warranty provision. Therefore, there was limited high-level review involved due to the time constraints. The same thresholds used to flag a road for maintenance or repair work were used as the thresholds to trigger the warranty. The committees also established the bond price, which was based on the estimation of the typical repairs costs. ODOT stated that project selection criteria for warranty projects included simple projects with well-defined pre-existing conditions and overall consistency compared to a typical urban main-street project in a downtown area where there are many crossings, signals, and transitions. The decision to apply a warranty is made at the district level, and district construction teams are responsible for monitoring pavement warranties on an annual basis. ODOT stated that there were some inconsistencies between the districts regarding inspec- tion processes and the level of inspection presence on-site. Unlike ODOT’s contractor quality assurance specification, the warranty provisions do not require contractors to submit quality records or reports. As a result, some districts choose to maintain these records. ODOT performed a comparison of bid prices for its pave- ment warranty projects. The analysis found that while there were significant increases in the initial projects bid in 2000, by 2005 the average bid increases had become insignificant. Table 8 summarizes the bid evaluation of pavement warranties. Cost increases were attributed primarily to the warranty bond, which was paid for under a bond line item that included the total cost for the performance, payment, and warranty bond. It was difficult to isolate the cost of the warranty bond, but ODOT estimated that the bond was approximately 1% of the total cost of the work. ODOT’s quality assessments are based largely on the perceptions of the district engineers overseeing the warranty work. According to ODOT’s latest report, the current percep- tion of warranty projects is that contractors are being more conscientious about their work, but they are not necessarily producing significantly better products. In some cases, con- tractors were found to be more proactive in improving poor soil conditions prior to placing the pavement, but in other cases, ODOT believed that contractors improved the surface course at the expense of the underlying non-warranted base 20

course. Concern has been expressed that in these instances, defects due to a substandard base course will not arise until after the warranty period, and ODOT is considering changes in the specification to address this issue. While the use of warranties continues, the level of use has declined in recent years. ODOT cited other concerns regarding warranty enforce- ment. Without comprehensive inspection records to document existing conditions and placement methods, determining the cause and enforcing the warranty may prove difficult in case of a failure. Another issue was the handling and addressing of superloads—large, one-time ESALs of 18,000 pounds or more in the warranty specification. While experience shows that the department designs are capable of handling such loads, many contractors disagree and believe such loads should void the warranty. ODOT has received letters from contractors during construction indicating problems with the base conditions and requesting that ODOT either authorize an extra work order for repairs to the existing base or void the warranty. Finally, ODOT noted difficulty in determining the start of the warranty period, especially on large, multi-phased projects. If the warranty is started upon opening to traffic, it may make monitoring an even more arduous task since there may be several different warranted segments. ODOT indicated that all of the above factors make enforcement of the warranty more difficult. Texas DOT (TxDOT) TxDOT has developed warranty specifications for hot-mix asphalt concrete, surface treatments, and microsurfacing, but has only implemented the microsurfacing specification. TxDOT reported issues in developing reliable thresholds for pavement warranties due to inconsistencies in its historical performance data, both in the type of pavement characteris- tic measured and how these characteristics were measured. This difficulty in determining thresholds has led to the slow implementation of warranties in Texas (Anderson et al., 2006). Currently, TxDOT has no plans to implement warranty specifications. Wisconsin DOT (WisDOT) WisDOT was the first agency to experiment with both HMA and PCC pavement warranties, each for 5-year periods. WisDOT began implementing warranties in 1995, following its implementation of a comprehensive quality control and quality assurance program. WisDOT noted in its 3-year progress report that the initial thresholds were easily achiev- able, which minimized the risk to contractors on the initial warranty projects. WisDOT’s long-term intention was to either tighten the threshold values or extend the warranty period once WisDOT and contractors gained some warranty experience (WisDOT, 1998). WisDOT stated that the intention of the warranty pro- gram was to give contractors as much freedom as possible while ensuring a quality product. WisDOT does not perform inspections or testing on warranty projects, and it identifies the ability to place these resources elsewhere as the most sig- nificant advantage to using warranties. Wisconsin contractors use their best crews on warranty projects; however, the majority of the state work is performed by a few large paving contractors. While these larger contractors have the bonding capacity necessary to support warranties, there are some small asphalt companies that have been pre- cluded from participating in warranty contracting due to limited bonding capacity. To reduce the impact of the bonding requirement and stimulate competition, WisDOT requires a warranty performance bond for the first year of the warranty, and then requires that the warranty bond be reissued in 2-year increments for the duration of the warranty. WisDOT published a 5-year progress report on its asphaltic pavement warranties in June 2001. WisDOT acknowledged that the limited amount of performance data available made assessing long-term trends difficult, but it offered a glimpse of comparative performance data by attempting to capture com- parative cost data between warranty and non-warranty con- tracts over a 5-year period (Brokaw et al., 2001). Costs evaluated under standard contracts over 5 years included mix bid prices, asphalt bid prices, tack coat bid prices, quality management bid prices, state delivery costs, and state maintenance costs. Costs evaluated under 5-year warranty contracts included warranted asphalt pavement bid prices and state delivery costs. Conflict resolution costs were found to be negligible on both project types. Additionally, distress surveys and further testing in cases of disputes were found to be negligible on warranty projects. 21 Item Warranty Period (Years) Avg. Bid Price Change (2000) Avg. Bid Price Change (2005) Asphalt (full depth) 5 and 7 +9% +1.19% %38.1- %8+ 3 )yalrevo( tlahpsA Concrete pavement (11”) 7 +7% -7.83% (all thicknesses) Concrete pavement (12” & 13”) 7 +15% Source: ODOT, 2007 Table 8. ODOT bid evaluation summary.

The results of the cost comparison were divided into two categories based on the year the project was let. Projects let in 2000 were broken out because of the addition of ancillary pavements to the warranty provision and the large increase in asphalt price that occurred that year. The evaluation showed warranty projects averaged $24.34 per ton compared to $27.72 per ton for standard projects from 1995 to 1999, and warranty projects averaged $29.34 per ton compared to $31.25 per ton for standard projects let in 2000. In both cases, the warranted projects appeared to cost less overall compared to non-warranted projects. This cost comparison concluded that even where an initial cost was up to 7% greater, warranty pave- ments were still more cost effective than standard pavements. The report also examined the comparative performance data on the warranty and non-warranty projects over the 5-year period. As shown in Table 9, the average IRI values of the warranted pavements over 5 years were improved compared to the average state IRI values. The pavement distress index (PDI) values were also significantly better than average state PDI values for non-warranted pavements. Based on the results of the 5-year progress report, WisDOT plans to continue using warranties and to make the warranty decision earlier in the design process. Other program recom- mendations include • Moving to a 7-year warranty while keeping the 5-year threshold values to tighten the requirements; • Considering the addition of an incentive that would reduce the warranty period for above-average performance, giving contractors bonding relief and providing a resource savings for the agency by eliminating unnecessary annual monitoring; • Investigating the use of optional warranty bidding; and • Considering implementing warranties as a standard. Type 3 Long-Term Performance Warranties Only a small number of U.S. highway projects have imple- mented Type 3 long-term performance warranties. As noted in Table 10, these have been implemented in Virginia, Missouri, and New Mexico, using hybrid construction manager at risk or design–build–warrant contracts, but the concept of long-term performance or maintenance contracts is being developed or explored in several states using PPPs or conces- sionaires, including in Indiana, Texas, and Virginia. Virginia is moving forward with PPP development and long-term operation and maintenance agreements under its Public– Private Transportation Act (PPTA) program (Interstate 81, Coalfields Expressway, and U.S. Route 460). These mainte- nance and operation agreements are typically concessions or long-term lease agreements (50 years or more) on toll roads. For example, the Trans-Texas Corridor (SH-130) and the Indiana Toll Road include performance indicators and distress thresholds (performance standards) for HMA pavements similar to those found in long-term warranty agreements. The agreements also may require the developer to maintain the pavement at a defined level of service for periods beyond the typical service life for the pavement and include pave- ment turnover standards at the end of the contract period (Seiders, 2006). 22 egA tnemevaPPerformance Indicators New 1 Year 2 Years 3 Years 4 Years 5 Years State average IRI – non- warranty 1.11 1.17 1.29 1.33 1.37 1.45 Average IRI – warranty 0.81 0.87 0.89 0.89 0.94 0.94 State average PDI – non- warranty 0 5 11 16 21 26 Average PDI – warranty 0 1 2 6 12 9 Source: Brokaw et al., 2001 noitaruD ytnarraW tcejorP TOD sraey 52 36 yawhgiH iruossiM sraey 02 44 etuoR etatS ocixeM weN sraey 02 882 etuoR ainigriV Table 9. WisDOT pavement performance data. Table 10. Type 3 U.S. pavement warranties.

Factors in Project Selection and Implementation Project Selection Few DOTs have established formal guidelines for warranty selection criteria, and authority for project selection is often delegated to the district level. Therefore, project selection criteria vary from state to state, and in some cases, within states. Table 11 summarizes how several DOTs delegated project selection authority. Caltrans, CDOT, and ODOT each established written guidelines or white papers on project selection criteria. These guidelines are included in Appendix C5. Despite the lack of formal guidelines and variations among districts, DOTs offered general comments regarding the types of projects typi- cally warranted. Comparison of the state experiences shows that warranty application ranged from safe projects with little risk of failure to highly urbanized projects that con- tained a myriad of unknown variables and risks. Table 12 illustrates how various DOTs compared with respect to project selection. DOTs interviewed acknowledged that applying a warranty to safe projects may result in little or no benefit over standard contracting practices; however, it became apparent during the interview process that while project-level criteria are an important part of the warranty decision process, programmatic- level criteria must also be considered in the project selection process. This was particularly true in states where warranty programs were initiated by legislative mandates. For example, initial legislative requirements in Ohio required warranties on a minimum of 20% of all projects let. ODOT reported that in some instances, warranties were applied simply to comply with legislation without much consideration to the project- level criteria. Legislation in Illinois required that warranties be applied to 10 projects with a 30-year design life as opposed to the standard 20-year design life. As a result, project selection was based largely on the feasibility of increasing the standard pavement thickness to comply with the legislation. Another programmatic-level criterion that appeared to play a part in the project selection criteria was level of industry cooperation. Mississippi DOT reported that the level of com- petition on warranty projects varied among districts, which often plays a role in the decision to apply a warranty. Contracting procedures can also affect project selection criteria. INDOT applies warranties in conjunction with A+B bidding with the goal of optimizing construction time and 23 ytirohtuA noitceleS tcejorP TOD California District with headquarters Colorado District engineers Illinois Pavement warranty committee Indiana Central office Louisiana State chief engineer Mississippi District engineers sreenigne tcirtsiD oihO Wisconsin District engineers DOT Project Selection Criteria Warranty Description pihsnamkrow dna slairetam raey-3 ylno noitcurtsnoc weN AL CA Projects with solid foundations; guidelines for cracking and exclusion areas 1-year materials and workmanship IL Simple in scope with no unique design, materials, or other innovations 5-year materials and workmanship pihsnamkrow dna slairetam raey-3 stcejorp llA LF WI Projects with excellent subbase 5-year performance CO Primarily paving scope with solid foundation, predictable traffic, and significant paving scope 3- and 5-year materials and workmanship OH Simple, consistent scope, not highly urbanized 7-year materials and workmanship MS Free of severe underlying defects 5-year performance MI All, unless the proper repairs cannot be addressed in the scope or project is short 5-year materials and workmanship MN Discretion of the district engineer 2- and 5-year materials and workmanship ecnamrofrep raey-01 setuor ciffart rojaM YK IN Major project in high traffic areas 5-year performance Table 11. Project selection authority. Table 12. Criteria for candidate warranty project.

quality. As a result, warranties are applied to highly visible, time-sensitive projects. In Louisiana, while it appeared that LaDOTD used appro- priate project-level selection criteria, programmatic issues such as limitations in the use of performance-based requirements resulted in dissatisfaction with warranty outcomes. LaDOTD reported that warranties resulted in an increased cost for the same performance. Based on the information drawn during the interviews regarding project selection, it was apparent that the selection tool developed for this research project should encompass more than just project-level selection criteria. As a result, the researchers focused on the development of a multidimensional warranty decision tool that addresses both programmatic and project-level section criteria. Furthermore, the existing project selection processes, particularly those developed by Caltrans and Michigan, led to the conclusion that the scoping of the project will be a key factor in the decision to use a warranty and will determine what warranty type can be implemented based on the risk allocation. For example, Michigan’s selection process for pavement rehabilitation considers base conditions, and if repairs are needed, the scope must include repairs to the base as a condition for using a warranty. Similarly, Caltrans defines warranty exclusion areas if crack repairs are not part of the project scope. In principle, DOTs can apply a warranty on any project incorporating a scope that reduces the risk to the DOT and to the contractor. Thus, warranties are applicable to rehabilitation projects where the scope of the rehabilitation can incorporate an appropriate fix consistent with the severity level of the existing pavement. It was also determined that in addition to addressing programmatic and project-level selection criteria, the tool must address the differences among the three types of warranties when deciding how to implement the warranty. Warranty requirements, such as the need for performance-based specifi- cations and contractor design expertise, vary depending on the type of warranty implemented. A comparison of programmatic factors related to warranty contracting for the three warranty types is included in Appendix C6. These factors, along with the project-level selection criteria identified through interviews, formed the basis of the warranty decision tool and guidelines discussed in Chapter 3 and presented in Appendix A. Despite the variation in reported outcomes of various types of warranties, the researchers decided to include all warranty types in the warranty decision tool. While several DOTs, including CDOT, MNDOT, and LaDOTD, have essentially concluded that materials and workmanship warranties have not proven to be beneficial over standard contracting practices, based on the collective state experience, it appears that these materials and workmanship warranties may serve as an important stepping-stone to larger warranty programs. These types of warranties give both contractors and DOTs the opportunity to gain experience with warranty contracting without placing excessive risks on either party. There is a lim- ited risk to the contractor under a Type 1 warranty since the contractor is only guaranteeing that it has constructed the job in strict adherence to the standard specification. Consequently, the only additional cost under a Type 1 warranty should be the cost of the warranty bond. Without evidence that Type 1 warranties result in improved performance compared to standard projects, many DOTs view these warranties as an unnecessary expense. However, DOTs that forgo a warranty bond may find that Type 1 warranties can improve consistency of the overall network and act as an additional guarantee against lemon projects at no additional cost to the DOT, as reported by FDOT and Caltrans. Several DOTs reported quality improvements as a result of Type 2 performance warranties, including Mississippi DOT, INDOT, and WisDOT. However, these DOTs admit that accurate, quantitative comparisons to support the effectiveness of warranties are difficult to achieve due to the many variables affecting project performance. DOTs with experience using Type 3 warranties contend that it is too early to determine the cost effectiveness or utility of these long-term warranties. However, with the increased interest and use of alternative design–build–warrant contract- ing and PPPs, Type 3 warranties will evolve into different forms within these long-term contracts that include similar performance criteria but significantly different administrative requirements. Implementation The implementation of construction warranties has varied based on DOT objectives, warranty type, contracting approach, industry feedback, and other factors. DOTs have shared infor- mation and experience on the use of warranties, which to some extent has promoted consistency in how the provisions are structured. Most practitioners would agree that key technical and managerial elements must be considered in guidelines as part of implementation regardless of objectives for use or type of warranty. In addition to project selection criteria, other key considerations include selecting performance indicators, setting distress thresholds, warranty durations, bonding, and risk allocation and contracting considerations. These are addressed in the following sections. Selecting Performance Indicators Warranty performance indicators are distresses, or func- tional characteristics of the warranted component that can be measured and linked to the performance of the warranted component. Performance indicators for a pavement may 24

include distresses or characteristics such as cracking, rutting, and ride quality. Based on the DOT responses, a variety of sources were used to identify and select performance indicators. The majority of DOTs indicated that historical information and experience were used to identify typical criteria that are easily, accurately, and routinely measured to assess performance of the product or component over time. ODOT further reinforced the impor- tance of historical data in a 2003 report (Hastak, Minkarah, and Cui, 2003), stating that a key criterion for evaluating warranty provisions is tracking how warranty projects are performing compared with historical data for non-warranted projects. Several DOTs (Mississippi DOT, for example) use standard- ized reference manuals, such as the SHRP Distress Identification Manual for the Long-Term Pavement Performance Program, to identify and define performance indicators and distresses (Miller and Bellinger, 2003). CDOT and ODOT stated that the indicator selection process was a joint effort including representatives of the DOT and industry. A representative of the Materials Division for Maine DOT stated that performance indicators used by other DOTs were researched and considered during the DOT’s indicator selection process. Often, DOTs use a combination of resources. For example, Illinois DOT described reviewing sample specifications from other DOTs, national reference manuals, historical performance data, and statistical analyses and using a joint DOT–industry work group to develop its performance-based warranty specifications. Table 13 shows representative samples of performance indicators used to gauge performance on warranty contracts. An issue to consider when selecting performance indicators is whether the monitoring process for selected indicators is consistent with established state practices for asset management. For example, distress characteristics routinely monitored through a pavement management program are relatively easy to administer under the existing program. If a performance indicator is not something that is routinely measured, the DOT must decide whether the benefit of the warranty will outweigh the burden of a separate monitoring process. A second issue is whether causes of premature failures are easily identifiable through inspection or forensic study. For example, Table 14 shows that all the sample HMA pavement specifications collected through the literature included rutting as a distress characteristic. Rutting is an easily identifiable and measurable pavement distress that is typically one of the first distresses to appear. Typically, it is easier to identify the cause of rutting than other distress characteristics, making it a good characteristic to monitor under a warranty. Indiana DOT selected rutting, IRI, longitudinal cracking, transverse cracking, and friction to evaluate its pavement con- ditions, aligning the warranty criteria with properties routinely collected under its pavement management system (PMS). INDOT did not include properties such as segregation, block cracking, flushing, and potholes in the primary evaluation because it decided that these distresses were more subjective and harder to measure or could be addressed as part of remedial action if any thresholds for the primary performance indicators were exceeded. Setting Functional or Distress Thresholds Warranty provisions specify threshold values for perfor- mance indicators. Threshold values are essentially measurable tolerances for the performance indicators. Warranty provisions may define maximum allowable tolerances for thresholds, which if exceeded trigger remedial action. They may also define zero-tolerance thresholds, meaning that the existence of any sign of distress requires remedial action. Thresholds are typi- cally tracked by visual inspection, laser profiling, or individual measurements. Survey respondents were asked to comment on the process for determining threshold values. DOTs develop benchmarks for thresholds using various resources, including statistical analysis of historical data, such as the state’s pavement manage- ment system, analysis of completed model projects perceived to be performing well, collaboration with industry representatives, information from other DOTs, and the state’s expectations for quality. 25 srotacidnI ecnamrofreP tnenopmoC TOD CA Pavement microsurfacing Rutting Raveling Flushing Streaking Defective areas IN HMA pavement Rut depth Transverse cracking Longitudinal cracking IRI Friction number MI Jointed plain concrete pavement Transverse cracking Longitudinal cracking Map cracking Spalling Scaling Corner cracking Joint sealant failure Table 13. Performance indicator examples.

Factors that affect the development of thresholds include the warranty classification (materials and workmanship versus performance), the availability of reliable historical data, and the level of industry cooperation. For example, DOTs aiming to establish consistent quality of the pavement network by strict adherence to the specification may establish thresholds that are consistent with historical, statewide statistical averages, while DOTs aiming to improve quality on a particular high- profile project may establish thresholds more stringent than historical averages. Most practitioners reported that threshold establishment began with a review of historical information. Solid, reliable historical performance data are a key to establishing thresholds consistent with the quality expectations of the DOT as well as industry standards. DOTs that struggled to establish thresholds reported that gaps or inconsistencies existed in the historical performance data. For example, when examining the feasibility of developing a pavement warranty specification in Texas, researchers reviewed historical pavement management data to determine whether accurate thresholds could be developed (Anderson et al., 2006). When examining rutting, researchers found that the method used to evaluate rutting in Texas DOT’s pavement management system was to measure rutting as a percentage of the section’s total wheel-path area. However, the SHRP Distress Identification Manual for the Long-Term Pavement Performance Program defines rutting severity as rut depth in inches (Miller and Bellinger, 2003). Furthermore, a survey of warranty provisions used in other states showed that rutting threshold values were always expressed as rut depth. Because the Texas DOT pavement man- agement system did not record rut depth, researchers were not able to establish a threshold consistent with industry-established terms and measurements of distress. Such inconsistencies in pavement data have hindered the implementation of pavement warranties in Texas and can act as a gap to implementation in other states (Anderson et al., 2006). DOTs that collaborated with industry throughout the developmental processes established reasonable thresholds for both the DOT and the industry. Some DOTs worked with industry representatives to familiarize the industry with dif- ferent levels of performance. For example, the Indiana DOT invited a group of industry representatives to participate in a demonstration of IRI values. The demonstration involved taking rides with these representatives on roads of varying IRI values to give them a better understanding of what the IRI is and how it relates to the warranty requirements. Other DOTs, including those of Wisconsin and Mississippi, started out with relatively easy-to-obtain threshold values to gain support and build experience within the industry. Wisconsin DOT is now considering modifications to either the threshold values or the warranty length as parties become more knowledgeable and comfortable with the application of warranties. Mississippi DOT reported that it initially chose a 5-year threshold for HMA and a 10-year threshold for PCC, but after letting two HMA projects, it decided to increase the HMA term to 7 years, with the same thresholds, on subsequent projects. Threshold values are structured several different ways. Some performance indicators use a single, minimum threshold value, while others may specify levels of thresholds with dif- ferent remedial procedures that correspond to the severity of the distress. Other thresholds are expressed in terms of 26 DOT R u tt in g R av el in g/ Se gr eg at io n C ra ck in g Lo ng itu di na l C ra ck in g Tr an sv er se C ra ck in g Bl oc k or Ed ge C ra ck in g D el am in at io n D ep re ss io n/ Sh ov in g Bl ee di ng /F lu sh in g Po th ol es A lli ga to r C ra ck in g Fa tig ue /R ef le ct io n C ra ck ing D eb on di ng D isi nt eg ra te d A re a Sm oo th ne ss R id e Q ua lit y o r IR I Pa tc hi n g Fr ic tio n N um be r CA x x x x x x CO x x x x x x x x FL x x x x x x IN x x x x x IL x x x x x x x x x x LA x x x x x x x x x ME x x x x x x x x MI x x x x x x MN x x x x x x x x x MS x x x x x x x x x x NM x x x x x x x x x x OH x x x x x TN x x x x x x x WI x x x x x x x x x Table 14. Performance indicators for HMA pavement warranties.

percentages of the overall segment or a predetermined length or surface area of the warranted component. Mississippi DOT developed the DEDUCT software pro- gram, which translates different distress thresholds for pave- ment performance measures into a common point system for pavement distress. DEDUCT asks the user to select a severity level for the distress based on the severity levels defined in the Distress Identification Manual for the Long-Term Pavement Performance Program (Miller and Bellinger, 2003). The program also asks for the extent of the distress in percentage, length, surface area, or number to relate it to the historical perfor- mance curve for that distress. The program then applies a point value for the distress based on the comparison of the extent of the distress to the historical curve. Table 15 illustrates how thresholds for the same HMA pavement distress are defined differently by various DOTs. Maine DOT indicated that the threshold values should account for both the number of defects and the dimensions of the defects. The comment was in reference to an issue arising on an HMA warranty project in Houlton. The warranty pro- vision called for remedial action if two or more transverse cracks were found in a single pavement segment. A segment was defined as a 100-meter (328-ft) portion of the pavement surface. During monitoring, the DOT found that many seg- ments of the project had one transverse crack, one of which was more than 50 mm (1.9 in.) wide, but these segments did not require remedial work because there was only one crack in the segment. Another potential issue with thresholds is determining whether to specify a distinct minimum threshold or a gradu- ated scale or performance curve during the warranty period. In other words, if the warranty period is 5 years, the DOT must determine whether the threshold should be set at the anticipated 5-year value or be based on a performance curve with differing values for years 1 through 5. A report prepared for the Montana DOT concluded that performance thresholds should be the minimum thresholds, as opposed to graduated thresholds over the warranty period, citing the disadvantage that the performance curve can be modified once remedial action is performed. The manner in which performance data is tracked and evaluated should be considered when setting threshold values. Selecting performance indicators similar to distresses tracked under the standard pavement management system is useful for developing comparative performance data; however, consideration should be given to the length of the segment over which data points are measured when setting thresholds. If the evaluation segments are too long, a local- ized area of poor performance can be diluted. Distresses on warranty projects are typically measured over shorter segment lengths than the segment lengths measured during typical pavement condition surveys to ensure that localized areas are not diluted. The effect of differences in segment length should be evaluated when considering thresholds or com- paring data points. To illustrate how a baseline performance threshold might be developed, the following steps provide an approach for establishing baseline IRI thresholds using PMS project data based on age and functional classification. 1. Review existing PMS data. For this example, PMS data is taken from an Indiana DOT 10-year HMA pavement using high-speed data collection methods (inertial profiler). Segments are typically 1.0 mile in length, as shown in Figure 6. These PMS segments are typically too long to accurately evaluate warranted pavement condition for shorter warranty segments. 2. To develop IRI thresholds for shorter warranty segment lengths, typically 0.1 mile, reprocess the existing PMS data for shorter sections by eliminating the data from areas with expected localized extremes caused by bridge approaches or other transitions in the pavement. 3. For an evaluation length of 0.1 mile (520 ft), use the reprocessed PMS data to compile the IRI and rutting data 27 sdlohserhT dna sisaB tnemerusaeM srotacidnI Mississippi Deduct Points Wisconsin Segment = 0.1 mi Minnesota Segment = 500 ft stniop 0.5> gnittuR >7.0 points ≥0.25 in. <0.50 in. ≥0.50 in. ≥0.375 in. (25 ft of the segment length) Transverse cracking >3.0 points >5.0 points >25 cracks that average 0.5 in. wide per segment (granular base) 3 cracks per segment with minimum length of 6 ft (med. severity) Longitudinal cracking >4.0 points >6.0 points >1,000 linear ft of cracks that average 0.5 in. wide None allowed Table 15. Threshold values.

for 520-ft (0.1-mile) sections. Figure 7 shows a sample IRI and rut data strip for a 501-ft section. 4. Determine the statistical distribution of the data and calculate the standard deviation (σ) of IRI, as shown in Figure 8. As shown in Figure 8, σ is a measure of dispersion. Assum- ing a normal distribution, approximately 68% of the data would fall within 1 σ, and 95% of the population would fall within 2 σ. As a starting point, the DOT may set the thresh- old at 2 σ (where only 5% of measured sections would exceed the threshold) to reduce the risk to the contractor for a 10-year warranted pavement. With additional experience or improved consistency, the DOT may decide to tighten the threshold (to between 1 σ and 2 σ) or extend the warranty. If one considers the variability or dispersion of distress values on the warranty versus the non-warranty projects, the consistency of the warranty projects is likely to be consider- 28 Figure 6. PMS data for IRI and rutting for INDOT 10-year HMA using a high-speed inertial profiler and rut bar based on 1.0 mile segment lengths. Figure 7. INDOT IRI and rut data for a 0.1 mile segment length. Figure 8. Distribution and standard deviation () for IRI over a 0.1 mile segment.

ably better than that of the non-warranty projects. Therefore, the consistency, and thus quality, of the warranty projects is improved, even if a comparison of the averages shows perfor- mance to be about the same. These factors should be consid- ered when setting thresholds or comparing data. Setting the Warranty Period Warranty periods vary by state but are typically much shorter than the service life of the component. Based on the literature and interview responses, warranty durations for projects that do not include planned maintenance typically cover 10% to 30% of the overall design life of the component being warranted. Table 16 summarizes the range of durations that have been implemented on pavement warranties in the United States according to the literature and interview responses. Long-term performance warranties or maintenance agree- ments (greater than 10 years) that include planned maintenance or major rehabilitation for pavements can cover from 60% to 100% or more of the overall design life. These are the exception rather than the rule in the United States, but the number of these projects is growing as DOTs turn to the private sector for long-term maintenance and operation of transportation assets. Survey respondents stated that warranty periods for shorter-term warranties were developed based on factors such as percentage of design life, historical performance data, experiences of other states, cost of bonding, and input from industry and sureties. In Illinois, legislation mandated specific warranty periods (5 and 10 years) for pavement. If a DOT is interested in setting warranty durations or supporting a proposed duration using a more analytical approach, it could analyze performance data for a particular pavement type and roadway classification and develop dete- rioration curves for key distresses over time (e.g., IRI and rutting from the PMS). An example of deterioration curves using regression analysis for data from different aged HMA pavement from the Indiana DOT’s Interstate highway system is shown in Figure 9. A DOT could then evaluate these curves to determine the expected IRI and rutting thresholds (from the PMS) that would trigger major maintenance or rehabilitation of the pavement. If, for example, the thresholds triggering maintenance were reached at approximately 12 years and DOT was considering a Type 2 warranty, it would not want the contractor to assume the risk (cost) of planned maintenance and might use a factor of safety setting the warranty duration at somewhat less than the age that would trigger major planned maintenance, say 8 years or 66% of the expected threshold. Alternatively, a DOT could analyze the dispersion of the data at the 15-year threshold and set the warranty duration based on standard deviation, as noted in Figure 8. In practice, DOTs apply a combination of analysis and practical knowledge gained through experience to determine the warranty length. Caltrans reported that, based on its experience, if a pavement preservation project did not show distress within the first year after construction, it had a higher potential for maintaining performance during the expected life of the surface treatment. Therefore, the benefit of a warranty period longer than 1 year did not justify the added cost. Minnesota DOT responded that economics played a factor in determining the length of its design–build warranties. Design–build warranties were cut from 5 to 3 years because of a significant increase in bonding cost for the longer war- ranty period. A common concern raised about warranties is that durations of 10% to 30% of the design life are often not enough to ensure improved quality and adequate performance over the design life of the component (if that is the DOT’s primary objective). For example, 5-year warranties on pavements and 10-year warranties on bridges provide some measure of quality assur- ance but not enough to guarantee performance over the design life of these warranted components. The findings indicate that warranties are often limited by bonding and other economic considerations, reducing their perceived benefit. An additional concern raised by practitioners in Minnesota and Mississippi was setting the effective start date for the warranty. Projects that involve multiple phases or lane shifts have the potential for disputes if the start date of the warranty is not clearly defined in the contract. The contract should define whether the warranty will start upon substantial completion or upon the opening of each of the warranted sections to traffic. As a disadvantage to pavement warranties, Minnesota DOT listed difficulty defining warranty start dates when the projects had multistage construction. Mississippi DOT commented that pavement warranties should start once the pavement has been accepted rather than waiting for the final maintenance release, which is often delayed by unrelated issues such as grass growth. 29 Component No. States Durations (years) Range Typical 5–3 52–1 32 tnemevap AMH 01–5 02–3 51 tnemevap CCP Pavement preservation 9 1–3 2 Table 16. Warranty periods.

Bonding Requirements In the United States, the cost of the warranty is typically factored into the unit price of the component being warranted; thus, the contractor collects payment for the warranted item upon completion of the construction. U.S. DOTs typically require a bond to cover contractor warranty obligations during the warranty period. A warranty bond is secured through a surety, which guarantees contractor performance throughout the warranty term. Should the contractor fail to perform, the surety is responsible for the cost of remedial work to the limits of the bond. DOTs vary significantly on the value of the bond, depending on the component being warranted. Bond values are typically determined by one of the following methods: • Total dollar value of the warranted item (construction value), • Percentage of the total dollar value of the warranted item, • Lower value between a percentage of the contract value and a set dollar amount (i.e., 5% or $1 million), or • Estimated cost to perform a full repair or preservation technique, as noted in the Indiana example in Figure 10. Because carrying a warranty bond reduces the contractor’s overall bonding capacity, many contractors have expressed concern that warranty projects will reduce their capacity 30 Source: Flora, 2003 Figure 9. Average IRI and rutting versus age of HMA pavement sections. The warranty bond is $2,000,000.00 for the warranted HMA/SMA pavement. The bond is intended to insure completion of required warranty work, including payments for all labor, equipment, materials, and closure periods used to remediate any warranted pavement distresses. Source: Indiana DOT HMA Warranty Specification Figure 10. Specification excerpt: warranty bond.

to take on future work, and sureties have set limits on bond durations based on their assessment of warranty risk. These bonding concerns have in some cases precluded contractors from bidding and contributed to lower numbers of bidders on warranty projects. Sureties are often reluctant to take on the risk of a longer- term bond, particularly for smaller or first-time contractors. Kansas reported that contractors could not find a bonding company that would warrant a project element beyond 5 years and discontinued PCC pavement warranties after deciding 5 years was not a significant enough percentage of the design life to make the warranty worthwhile. To avoid such obsta- cles, the warranty bond is often obtained for a shorter period (1 to 2 years) and then renewed for the life of the warranty. Many DOTs with warranty experience have invited represen- tatives of industry and sureties to discuss options during the specification development process that would balance the risk for all parties involved, but bonding remains a gap to imple- mentation for many DOTs examining warranties. Because of the issues related to the use of bonds to ensure performance during the warranty, several states have explored alternatives to single-term warranty bonds, including the following: • Extended-performance bond; • Letter of credit, certificate of deposit, U.S. currency, or other form of security approved by the department; • Warranty performance tied to the prequalification process for future work (guarantee); and • Pay-for-performance or retainage. Some of these alternatives have roots in European practice. For example, not all European agencies use warranty bonds. Instead of requiring a bond, the British Highways Agency uses prequalification to ensure that its contractors will correct defects in their work. Denmark uses a graduated bond for its 5-year warranties, reducing the bond to 2% of construction costs in the final 4 years to balance risk and reduce cost (Bower et al., 2003). Florida has moved from a warranty bond to prequalification for future work. If the contractor fails to perform the required remedial work, the contractor is precluded from bidding on future state work for 6 months or until the remedial work is completed, whichever is longer. Florida has coined this alter- native a “guarantee.” While Florida has stated that its guarantee process is an effective alternative to bonding, certain factors in Florida may not transfer well to other states. For example, the risk of being precluded from bidding on future work motivates contractors in Florida because most do not typically perform work outside of Florida. This concept would not be ideal in areas where it is feasible for contractors to bid on work in multiple states. For example, Michigan DOT is considering implementing a similar guarantee process, but is concerned about its effectiveness in areas where work is readily available should a contractor be precluded from bidding on future work in Michigan. Legislative changes would also be neces- sary for Michigan to implement a guarantee in place of a warranty bond. Minnesota has also expressed interest in the guarantee model. Minnesota has previously used a pay-for-performance specification. The pay-for-performance concept is similar to retainage, in which the contractor is paid a portion of the costs at the time the item is placed and then is paid on a graduated scale over time if the item performs to expectation. Minnesota implemented this alternative for warranties on its I-494 design–build project. For warranties of 1 year or less, DOTs have also extended the performance bond to cover the warranty period, making a separate warranty bond unnecessary. This is the case in California and North Carolina, which require standard 1-year warranties on most projects. Risk Allocation and Contracting Considerations Material and construction requirements vary depending on whether DOTs are implementing a materials and work- manship warranty or a performance warranty. When transi- tioning from materials and workmanship to performance warranties, progressively more responsibility is shifted to the contractor and they are typically given more freedom to control aspects of design or construction. Most materials and workmanship warranties require the contractor to conform to the standard method specifications. The contractor may have some choice over mix design and material selection, but the contractor is typically restricted to choosing materials from a state-approved list. Under per- formance warranties, the contractor is typically given greater control over material selection and mix design. The contrac- tor may or may not be restricted to a list of state-approved materials. Some performance warranty provisions also give contractors control over the methods used to construct the work. Based on the interview responses and an exami- nation of specifications, it was difficult to classify some of the pavement specifications as material and method versus performance because of subtle differences in responsibility for mix design. Contractors are typically given greater design responsibility under a design–build–warrant type of contract. Lack of clear definitions of rights and responsibilities between materials and workmanship warranties and performance warranties can be a risk in the successful implementation of warranties. 31

Quality Control, Inspection, Testing, and Acceptance During Construction Implementation of warranties often involves changes in traditional roles and responsibilities for quality control, inspection, and testing during construction. Warranty provi- sions differ on responsibility for these duties. For example, Indiana DOT implemented a hands-off approach, shifting responsibility for inspection, quality control, and testing to the contractor. Louisiana DOT, on the other hand, maintained traditional roles and responsibilities for these duties on its warranty projects. Typically, the contractor takes on greater responsibility for these tasks, with the DOT maintaining a quality assurance or verification role. DOTs in Indiana, Mississippi, Michigan, and Wisconsin reported resource savings benefits on warranty projects. Florida reported that the benefit of its resource saving was comparable to the benefit achieved through the use of a contractor QC specification. Based on the comments provided in the survey and evalu- ations performed at the state level on warranty contracting, DOTs differ on the role the DOT should maintain in inspection and testing on warranty projects. Colorado DOT responded that better quality control and more oversight by state forces are needed. Indiana DOT responded that the owner should not oversee contractor operations to avoid disputes over respon- sibility because the state witnessed the operation. Indiana DOT also responded that if inspection is performed, it should consist of a spot-checking process, not a full-time inspector. These different perspectives among warranty programs are driven in part by the comfort level DOTs have with shifting the responsibility for quality management to the contractor. Again, the warranty provisions should clearly define the roles and responsibilities of each party for inspection and testing in the warranty provisions. Alternative Contracting Design–Build–Warranty Recent international scan studies have reported that European agencies routinely use warranties in conjunction with design–build contracting and rely more on the private sector to maintain and operate highways. The European agencies also have increased the use of maintenance contracts and pavement performance contracts (FHWA, 2005). These trends toward greater private-sector responsibility are evident in the U.S. highway industry but are not as common. Based on a 2005 FHWA-sponsored design–build effectiveness study, about 30 DOTs have used design–build contracting on public works projects (FHWA, 2006). Among these, a smaller percentage of DOTs have implemented design–build–warranty contracts or entered into agreements with the private sector for long-term maintenance or operation of highways and bridges under a PPP agreement or performance-based maintenance contract. Because performance warranties shift progressively more responsibility for quality and performance to the private sector, contractors have expressed concerns that they cannot take on this performance risk without greater control of the design. Contractors from Ohio DOT’s 2003 warranty study expressed these concerns (Hastak, 2003), and they were also expressed in Michigan Local Technical Assistance Program workshops in 2005 (MDOT, 2005). When asked which factors would hinder contractors from bidding on warranty projects, the Ohio contractors cited the absence of design–build contracting, followed by the duration of the warranty and availability of bonds. In this vein, Maryland and Alabama solicited bids for warranted HMA pavement on design–bid–build contracts, but they failed to receive any bids and have no plans to use warranties on similar contracts. While the majority of DOTs use warranties with traditional design–bid–build contracts, some DOTs have combined design–build contracts with performance warranties to ensure quality in a design–build environment of reduced owner inspection and accelerated construction. Minnesota has com- bined short-term performance warranties with design–build projects to achieve these goals (MNDOT, 2005). Missouri and Virginia have also combined design–build contracts with the long-term performance warranties noted in Table 10. Given that design–build contracting shifts more control to the contractor for design and project performance, in the event of a failure the contractor is less likely to be excused from the warranty obligations because of design- or performance-related issues. Disputes documented in the literature and interview responses typically occurred on design–bid–build projects. Minnesota DOT did report that one of its 5-year design– build–warranty pavement projects was experiencing excessive transverse cracking and rutting, requiring corrective action by the contractor to rout and seal cracks. The contractor did not contest responsibility for the corrective action, but has contested the start of the warranty period, which affected the cost of the repairs (MNDOT, 2005). Virginia DOT also reported one case in which a contractor negotiated out of the warranty for a design–build job during construction. Public–Private Partnerships and Concessionaires In parallel with a long-standing and growing use in Europe, PPPs or concessions have been applied to a handful of high- profile projects in the United States, but their use has recently gained new momentum as transportation owners struggle to find resources to fund and deliver critical transportation projects. Some of the earliest examples of privately funded PPP projects in the United States were the Route 91 express lanes in California and the Dulles Greenway toll road in Virginia, both completed in 1995. More recent examples include the Virginia DOT Public–Private Transportation Act program, 32

the Florida DOT Port of Miami Tunnel project, the Indiana Toll Road, and the Trans-Texas Corridor (Seiders, 2006). Texas DOT established operational and maintenance per- formance standards for the Trans-Texas Corridor that the concessionaire must meet for the operation and maintenance (O&M) of the facility. The pavement performance standards define the minimum standards (thresholds) the concession- aire will be required to meet during the O&M period for the facility. Corrective action will be made if these thresholds are exceeded. The performance standards include the following (Seiders, 2006): • Pavement condition score. Measurements and inspec- tions necessary to derive a pavement condition score in ac- cordance with Texas DOT procedures. • Ruts—mainlanes, shoulders, and ramps. Depth measured using an automated device in compliance with Texas DOT standards. Straightedge used to measure rut depth for localized areas. • Ride quality. Measurement of IRI according to Texas DOT standard Tex-1001-S, operation of inertial profilers, and evaluation of pavement profiles. • Failures. Instances of failures exceeding the failure criteria set forth in the Texas DOT Pavement Management Information System Rater’s Manual, including potholes, base failures, punchouts, and jointed concrete pavement failures. • Edge drop-offs. Physical measurement of edge drop-off level compared with adjacent surface. • Skid resistance. ASTM E 274 Standard Test Method for Skid Resistance Testing of Paved Surfaces at 50 miles per hour (80.4 kilometers per hour) using a full-scale smooth tire meeting the requirements of ASTM E 524. These PPP performance standards and thresholds are very similar to performance characteristics and distress thresholds specified for warranty contracts, but they extend the perform- ance period in some cases well beyond the service life of the pavement, which would entail major rehabilitation during the O&M period. They also do not include the typical exclusions that may void the agreement. To achieve these standards, PPP specifications are performance-oriented. In other words, Texas DOT in theory will not specify pavement design and type and will limit its review and approval functions under these types of contracts. Performance-Based Maintenance An alternative to using warranties to guarantee post- construction performance is the use of performance-based maintenance contracts, such as in Virginia and the District of Columbia where a private entity enters into a long-term agreement with the DOT. The DOT typically pays the contrac- tor a set amount each year to maintain a specified performance level. These contracts primarily cover routine maintenance, but often include limited preventive maintenance duties for pavements such as the repair of potholes and joints. Multiparameter Bidding Several DOTs have combined warranties with cost-plus-time or A+B bidding and incentive–disincentive provisions to motivate contractors to balance time and quality goals. For example, Indiana uses pavement warranties in conjunction with A+B bidding and has reported improvements in quality and time with this approach. While the state has not substan- tiated it, Indiana believes that contractors receiving an incen- tive for accelerated project completion are motivated to apply greater resources and attention to quality than contractors not receiving an incentive. Kentucky piloted an A+B+C formula, asking contractors to bid, in addition to cost and time, a C duration for the warranty worth a $500,000 credit for each year offered. The results were advantageous to the state because it received 5 additional warranty years in conjunction with a shorter schedule. Exclusions A review of the warranty provisions collected for this report shows that most DOTs define specific exclusions limiting contractor liability under the warranty. For example, warranties on traffic signal and lighting posts typically include exclusions for conditions outside the contractor’s control, such as damage from lightning or vehicular accidents. Common exclusions include damage to warranted products resulting from the following: • State-controlled operations, such as routine maintenance or destructive testing, • Vandalism, • Vehicular accidents and hazardous material spills, • Military action, and • Acts of God and natural disasters. Some warranty provisions define design-related exclusions. For example, a materials and workmanship warranty for HMA pavement may exclude defects caused by existing base con- ditions or drainage design errors. As shown in Figure 11, pavement warranties also typically specify a maximum ESAL value or maximum number of heavy trucks that, if exceeded by a certain percentage, voids the warranty. Some pavement warranties also define one-time, heavy- load ESAL values. DOTs that have used ESAL exclusions for pavement warranties, including Florida, Ohio and Michigan, do not uniformly specify weigh-in-motion devices to monitor ESALs because of the significant up-front cost and instead rely on average daily traffic counts to estimate ESALs. 33

Monitoring and Remedial Action Warranty clauses always provide for condition surveys of warranted items, either at periodic intervals or, for short-term items such as pavement preservation, at the end of the warranty. For pavements, this survey may be conducted in concert with the DOT’s annual pavement condition survey or through an independent warranty program. This survey may be conducted with the contractor or unilaterally. If the survey shows that the thresholds established in the warranty provision are exceeded at any time during the warranty period and the cause does not fall under a defined exclusion, then the contractor is notified and called back to perform remedial action. Based on a review of the sample specifications gathered for this report, DOTs have developed different approaches for establishing the required remedial action. Several require or suggest a remedial procedure when a given threshold is exceeded. Some specify degrees of severity for the threshold distresses and require or suggest different remedial procedures based on the severity of the distress. For example, Minnesota DOT’s warranty for bituminous pavement suggests that con- tractors rout and seal transverse cracking of medium severity and mill and resurface transverse cracking of high severity. The severity of the cracking is determined by the number of cracks in a pavement segment. Most if not all of these warranty provisions require that the contractor submit for approval a remedial work plan detailing the proposed fix. If the contractor does not agree with the survey results or the scope of the remedial work or believes that the cause of the distress was beyond its control, it can dispute the results and refer the matter to a dispute resolution team or board to render a final or independent decision. Dispute Resolution Procedures Most warranty provisions establish measures for settling disagreements for potential disputes over remedial action. Responsibility for settling disputes is typically delegated to a conflict resolution team (CRT) or a DRB. CRTs are typically set up on a project-by-project basis. They consist of an equal number of representatives from the con- tractor and the DOT who may or may not be directly involved in the project, plus one outside representative mutually agreed on by the contractor and the DOT. Costs associated with the outside representative are shared by the contractor and the DOT. CRTs have also been called dispute resolution teams and pavement evaluation teams. DRBs are formal committees set up to resolve disagree- ments before they can delay or disrupt construction projects. DRBs typically consist of one to three members and can be set up for specific projects or on a district or statewide basis. DRBs can be responsible for settling all project-related disputes or only disputes related to warranties. Florida has established a three-member, statewide DRB dedicated to settling pavement warranty disputes that cannot be resolved at the project level. Florida’s DRB panel is drawn from the state, industry, and academia. Chapter Summary As evidenced by the experience of the workshop attendees, the DOTs surveyed, and the available literature, there is a signi- ficant range of opinion regarding the effectiveness of warranties, but the more-experienced DOT practitioners report that war- ranty projects will result in improved performance (cost and quality) if implemented for appropriate projects. There is also 34 Exclusions Remedial work will not apply if any one of the following factors is found to be beyond the scope of the contract: a. Determination that the pavement thickness design is deficient. The department will make available a copy of the original pavement thickness design package and design traffic report to the responsible party upon request. b. Determination that the accumulated ESALs (number of 18 kip equivalent single axle loads in the design lane) have increased by 25% or more over the accumulated ESALs used by the department for design purposes for the warranty period. In calculating ESALs, the average annual daily traffic (AADT) will be obtained from the department’s traffic count data and the T24 (percentage of heavy trucks during a 24-hour period) will be obtained from the department’s traffic classification survey data. c. Determination that the deficiency was due to the failure of the existing underlying layers that were not part of the contract work. d. Determination that the deficiency was the responsibility of a third party or its actions, unless the third party was performing work included in the contract. Source: Florida DOT Section 338, Value-Added Asphalt Pavement Specifications Figure 11. Specification excerpt: exclusion language.

significant variation in the way DOTs have defined and imple- mented warranties, contributing to a grey area or overlap in defining the types of warranties. Figure 12 uses a Venn diagram to illustrate how several existing warranty provisions align with respect to the definitions used in this report. The research team has adopted definitions of the warranty types generally based on the specifications that fall outside of the overlapping areas within Figure 12. While several DOTs have established that there is no tangible benefit to short-term Type 1 materials and workmanship warranties (3 years or less), based on collective state experi- ence, it appears that these Type 1 warranties achieve a specific objective, namely to prevent early failures, and may also serve as an important stepping-stone to implementing longer-term warranties. This approach gives both contractors and DOTs the opportunity to become accustomed to warranty contract- ing while limiting the risk to either party. Theoretically, there is a very limited risk to contractors under a Type 1 warranty since they are only guaranteeing that the project has been constructed in strict adherence to the standard specifications. Consequently, the additional cost to the project for a Type 1 warranty should be limited to the cost of the bond. Given this perspective, the research team decided to address all warranty types in the guidelines for project selection, project imple- mentation, and model specifications. While DOT selection criteria were limited in scope, the few examples of systematic selection tools used by Caltrans and MDOT have led to the conclusion that the project scoping is a key factor in the decision to use a warranty, and the decision process must align the project scope and project objectives with the selection of an appropriate warranty type to achieve a successful result. This chapter addressed the other key considerations for warranty implementation discussed in the literature, workshop, and surveys. Warranty implementation varies significantly, depending on whether DOTs are using short-term materials and workmanship or longer-term performance warranties, or in the context of traditional-versus-alternative contracting. This chapter also discussed how DOTs with experience have addressed these key implementation elements based on their internal goals and lessons learned. Finally, this chapter identified a number of issues raised by practitioners and discussed potential strategies to address these issues. These strategies include developing systematic project selection criteria and guidance for applying materials and workmanship versus performance warranties; setting performance thresholds based on historical data or quality goals; exploring alternatives to warranty bonds; appropriately allocating risk related to warranty types, contracting method, and exclusions; and establishing responsibility for pass-through warranties. Guidelines and sample specifications have previously been developed and used for implementing pavement war- ranties, particularly for asphalt pavements. These guide- lines and specifications have been updated in this report to reflect the current state of practice for warranties, includ- ing project selection criteria for warranty types, risk alloca- tion based on contract type, and model specifications for both HMA and PCC pavements covering both materials and workmanship and performance warranties. The next chapter discusses the approach to developing these guide- lines. Appendix A includes standalone warranty guidelines, a selection tool, and model specifications for HMA and PCC pavements. 35 Figure 12. DOT warranty provisions—comparison by type.

Next: Chapter 3 - Development of a Pavement Warranty Decision Tool, Best Practice Guidelines, and Model Technical Provisions for HMA and PCC Pavements »
Guidelines for the Use of Pavement Warranties on Highway Construction Projects Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 699: Guidelines for the Use of Pavement Warranties on Highway Construction Projects is designed to help guide state departments of transportation (DOTs) in establishing pavement warranty programs.

The guide identifies programmatic and project-level decision criteria that DOTs should consider when implementing and sustaining a program. The guide presents strategies to mitigate project-specific risks and also includes model warranty specification provisions.

The guide also includes a decision tool to help identify program-level issues and project-specific risks. The tool is included on a CD-ROM that is packaged with the printed version of the report.

The CD-ROM is also available for download from TRB’s website as an ISO image. Links to the ISO image and instructions for burning a CD-ROM from an ISO image are provided below.

Help on Burning an .ISO CD-ROM Image

Download the .ISO CD-ROM Image

(Warning: This is a large file and may take some time to download using a high-speed connection.)

CD-ROM Disclaimer - This software is offered as is, without warranty or promise of support of any kind either expressed or implied. Under no circumstance will the National Academy of Sciences or the Transportation Research Board (collectively “TRB’) be liable for any loss or damage caused by the installation or operation of this product. TRB makes no representation or warranty of any kind, expressed or implied, in fact or in law, including without limitation, the warranty of merchantability or the warranty of fitness for a particular purpose, and shall not in any case be liable for any consequential or special damages.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!