Performance Specifications for Rapid Highway Renewal (2014)

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.

82 A p p e n d i x d Bridges Badie, S. S., K. T. Maher, and A. F. Girgis. NCHRP Report 584: Full- Depth, Precast-Concrete Bridge Deck Panel Systems (NCHRP Proj- ect 12-65). Transportation Research Board of the National Academies, Washington, D.C., 2006. This report provides recommended guidelines and specifications for the design, fabrication, and construction of full-depth, precast-concrete bridge deck panel systems. The general outcome criteria imbedded in these prescriptive specifications can serve as a basis for the development of performance specifications. Bickley, J., R. D. Hooton, and K. C. Hover. Preparation of a Performance-Based Specification for Cast-in-Place Concrete. Ready Mixed Concrete (RMC) Research Foundation, Silver Spring, Md., 2006. http://www.nrmca.org/P2P/. Accessed Dec. 3, 2007. This report was commissioned by the National Ready Mixed Concrete Association to review the state-of-the-art for performance-based speci- fications for concrete. It provides a historical and international review of performance specifications for concrete, including discussion of the other concrete materials specifications included in this bibliography. The report lists all the prevailing quality tests for concrete and discusses their suitability for administering a true performance specification of con- crete materials. FHWA. Prefabricated Bridge Elements and Systems in Japan and Europe. International Technology Exchange Program, Federal High- way Administration, U.S. Department of Transportation, 2005. http://international.fhwa.dot.gov/prefab_bridges/index.htm. Accessed Dec. 3, 2007. This resource describes a scanning tour of five countries in April 2004, which identified nine technologies including two superstructure systems, four deck systems, one substructure system, and two bridge movement systems. The tour found that bridge replacement using self-propelled mobile technology (SPMT) to move prefabricated bridges into position has been done overseas, as well as in Florida, Rhode Island, and New Jersey. SPMT for bridge or bridge component replacement appears especially relevant to rapid renewal projects. Use of SPMT is being documented in draft versions of guidelines being developed by FHWA. The guidelines are expected to produce performance criteria for rapid-replacement systems. Hughes, C. S., and C. Ozyildirim. End-Result Specification for Hydrau- lic Cement Concrete (VTRC 05-R29). Virginia Transportation Research Council, Charlottesville, 2005. This report discusses an end-result specification for accepting and pay- ing for hydraulic cement concrete (for use in bridges and pavements). It addresses the selection of quality characteristics, sampling and testing criteria, specification limits, and pay adjustments. The selected quality characteristics represent some of the most contemporary surrogates for long-term performance. Missouri Department of Transportation. Safe and Sound Bridge Improvement Program. Missouri DOT, Jefferson City. http://www .modot.org/safeandsound/. Accessed Nov. 29, 2007. If successful, the Missouri DOT Safe and Sound Bridge Improvement Program will outsource the rehabilitation and maintenance of approxi- mately 800 highway bridges for a period of 25 years. The program estab- lishes objectives for the original rehabilitation/reconstruction phase of the work, as well as performance expectations for the 25-year period of maintenance. Most of these detailed performance requirements are built around the Structure Inventory and Appraisal standards (SI&A) of the National Bridge Inspection Standards (NBIS). The program pro- poses to make the performing contractor responsible for delivering performance (through rehabilitation and/or maintenance) but to retain the DOT’s responsibility for determining whether performance expec- tations are met (via the routine bridge inspection program). Ramirez, J. A., J. Olek, A. Lu, and R. J. Frosch. Performance Related Specifications for Concrete Bridge Superstructures: A Four Volume Report. FHWA/IN/JTRP-2001/8. Division of Research, Indiana Department of Transportation, West Lafayette, 2002. This research is among the very few to address performance specifica- tions for entire concrete superstructures (the deck and major structural concrete elements). This work moves beyond materials issues to discuss performance in a larger structural context. Specific limit conditions (or performance criteria) in this research address corrosion-related deterio- ration, excessive crack width, flexural failure, and excessive deflections. Sprinkel, M. M. Performance Specification for High Performance Concrete Overlays on Bridges. VTRC 05-R2. Virginia Transportation Research Council, Charlottesville, 2004. This report describes the Virginia Department of Transportation’s first experience with a performance specification for concrete overlays of bridges. The specification accepts and pays for the overlay on the basis of air content, compressive strength, and permeability of the overlay material (concrete), as well as the strength of the bond to the original deck material. The research demonstrates how the new specification delivered an improved product. Annotated Bibliography

83 Tadros, M. K., and M. C. Baishya. NCHRP Report 407: Rapid Replace- ment of Bridge Decks. TRB, National Research Council, Washington, D.C., 1998. NCHRP Report 407 provides a number of techniques that facilitate rapid deck replacement and includes proposed special provisions for the removal of the deck, which is one of the most time consuming aspects of the replacement process. The report recommends the use of performance specifications. Also, the report includes details of a precast, partial-depth deck panel system that can be used to accelerate the replacement of the deck. Information from this project can serve as a basis for the develop- ment of performance specifications for rapid deck replacement. Geotechnical infrastructure Systems Bennert, T., and A. Maher. The Development of a Performance Specifi- cation for Granular Base and Subbase Materials. FHWA-NJ-2005-003. New Jersey Department of Transportation, Trenton, 2005. Performance testing, selected to simulate or evaluate aggregate materials from a pavement system, was conducted to evaluate the influence of aggregate gradation. The results provided guidance for the potential modification of the New Jersey DOT specifications for base and subbase aggregates used in the construction of pavements. Brown, D., R. Thompson, and S. Nichols. Performance Specifications for Drilled Piles. Proc., 10th International Conference on Deep Founda- tions, Amsterdam, Netherlands, 2006. http://danbrownandassociates .com/wp-content/uploads/2006/06/Performance%20Specs%20 for%20CFA.pdf. Accessed Nov. 12, 2007. This paper discusses the use of a performance specification in which the contractor can choose the most appropriate foundation system and also has greater responsibility for the performance of the completed founda- tion. Deep foundation contracting for transportation projects in the United States has typically followed the design-bid-build model in which contractors bid on a specific work product according to relatively inflex- ible prescriptive specifications. With new proprietary drilled displace- ment auger piling, there may be advantages to contracting according to the performance-based specification model, including opportunities to encourage innovation. There are also challenges for implementation in public works projects, notably the selection of qualified contractors and the verification of performance requirements on production founda- tions. This paper describes a number of features characteristic of perfor- mance specifications; in particular, it emphasizes the need for quality assurance (QA) and quality control (QC) using automated monitoring systems and the need for an appropriate testing program to assure the owners that the performance criteria are met. Edwards, P., N. Thom, P. Fleming, and J. Williams. Testing of Unbound Materials in the Nottingham Asphalt Tester Springbox. In Transporta- tion Research Record: Journal of the Transportation Research Board, No. 1913, Transportation Research Board of the National Academies, Washington, D.C., 2005, pp. 32–40. The current trend in mechanistic (analytical) pavement design is to use the mechanistic properties of pavement materials to optimize design. This is compatible with the move toward performance-based specifications and away from traditional, empirically based design methods and rec- ipe specifications. The performance parameters assessed, over a range of moisture and soaking conditions, are resistance to permanent defor- mation and resilient stiffness. The apparatus used during the unbound mixture assessments is the newly developed Springbox, which uses the standard Nottingham Asphalt Tester loading frame and software. Ellis, S., M. Nunn, and D. Weston. Development and Perceived Benefits of Performance Specifications in the UK. World Road Association, PIARC, Paris, July 2002, pp. 68–78. This article discusses the features of the United Kingdom performance specifications for surfacing and base layers of pavements. The benefits determined from UK experiences and conditions are discussed. A performance-related specification for asphalt materials was implemented in the UK in 1996. The clause was primarily implemented to ensure materials reach the standard assumed in the pavement design, to allow more scope for contractors to produce the most economical mix design, and to ease the introduction of alternative materials. Fleming, P., M. Frost, and J. Lambert. Geotechnical Specifications for Sustainable Transport Infrastructure. In Transportation Research Record: Journal of the Transportation Research Board, No. 1975, Trans- portation Research Board of the National Academies, Washington, D.C., 2006, pp. 73–80. The report pertains to the specification of materials and methods used in earthworks and foundations for highways, railways, and airfield run- ways that make best possible use of material properties. The functional requirements of a performance-based specification for UK highway foundations are described with the objective of a more sustainable use of resources. Reportedly, performance-based specifications better incor- porate the principles of sustainable construction and require a fuller understanding of material behavior for their development and imple- mentation. However, contractual issues and implications for construc- tion need to be considered carefully to allow a full performance-based approach to be adopted successfully. Fleming, P., J. Lambert, and M. Frost. In-Situ Assessment of Stiffness Modulus for Highway Foundations During Construction. Ninth Inter- national Conference on Asphalt Pavements, Copenhagen, Denmark, Aug. 2002. This paper reviews in detail several portable field devices that measure stiffness modulus, including the German dynamic plate test (also known as the lightweight drop tester), the TRL foundation test, the prima, and the Humboldt soil stiffness gauge. Laboratory and field data are presented to explain the many important influences on the mea- sured data and to demonstrate comparative performance with respect to the falling weight deflectometer. These field data show significant scatter and site-specific correlation. Fleming, P., C. Rogers, N. Thom, and M. Frost. A Performance Speci- fication for Pavement Foundations. Transportation Geotechnics Pro- ceedings, Thomas Telford Publishing, London, 2003, pp. 161–176. This paper describes how the taxation of virgin aggregates and the dis- posal of wastes in landfills are increasing the pressure to use novel, mar- ginal, and recycled materials in earthworks and engineered fills that form the foundations of transport infrastructure. The paper details the per- formance required of pavement foundations and materials, discusses the philosophy of the performance specification that has been developed, and addresses the issues associated with its implementation. It concludes that a two-stage implementation of the specification should be adopted and that the specification could be extended to any application that requires construction of a platform/foundation to support a pavement, such as industrial paving, railway track-bed, or general fill. Frost, M., J. Edwards, P. Fleming, and S. Arnold. Simplified Laboratory Assessment of Subgrade Performance Parameters for Mechanistic Design of Pavement Foundations. In Transportation Research Record: Journal of the Transportation Research Board, No. 1913, Transporta- tion Research Board of the National Academies, Washington, D.C., 2005, pp. 77–85.

84 With the increasing agenda for sustainability, the United Kingdom is attempting to move away from the empirical design of pavement foun- dations to develop a performance-specification approach that facilitates analytical design. The measurement of the subgrade performance parameters of resilient modulus and resistance to permanent deforma- tion is required for analytical design. Hefer, A., and T. Scullion. Materials, Specifications, and Construction Techniques for Heavy-Duty Flexible Bases: Literature Review and Sta- tus Report on Experimental Sections. Technical Report. Texas Depart- ment of Transportation, Austin, July 2005. For the majority of Texas highways, the granular base layer is the main structural component of the pavement system. Base specifications and construction practices from eight U.S. departments of transportation (DOTs) and two overseas countries were compared with Texas DOT’s cur- rent and proposed specifications. Currently, Texas DOT is the only agency that does not control the amount of fines (minus 200 fraction) in its bases. Research studies have indicated that high levels of minus 200 can severely affect both moisture susceptibility and cold weather performance. The newly proposed Texas DOT specifications, with limits on the fines content, are in line with the practices of other agencies in similar climates. Iowa Department of Transportation. Special Provision for Intermediate Foundation Improvements: Polk County, IM-NHS-235-2(498)11—03-77. Iowa DOT, Ames, Aug. 2004. This specification describes criteria for controlling settlement, bearing capacity, and stability for mechanically stabilized earth (MSE) walls by installation of an intermediate foundation system. Neither the methods of installation nor the product is specified; instead design parameters are given, including allowable bearing capacity, total settlement, and differ- ential settlement. The basis for acceptance is ultimately visual inspection by the engineer, who will consider the results of all verification tests (load tests, soil borings, and in-ground instrumentation), as well as consistent use of procedures, methods, and construction performance rates. Livneh, M., and Y. Goldberg. Quality Assessment During Road For- mation and Foundation Construction: Use of Falling-Weight Deflec- tometer and Light Drop Weight. In Transportation Research Record: Journal of the Transportation Research Board, No. 1755, Transporta- tion Research Board of the National Academies, Washington, D.C., 2001, pp. 69–77. Mechanistic, empirical-based specifications, which focus on the mechan- ical properties of materials, facilitate quantitative evaluations of alterna- tive construction practices and materials such as reclaimed materials. To represent these mechanical properties, quality-control and quality- assurance testing are expected to include stiffness along with density mea- surements. In two case studies, a falling weight deflectometer (FWD) was used during the construction of two major interchanges in Israel. The possible use of the German light drop weight (LDW) device for measur- ing the mechanical properties of the formation of flexible pavements was also examined. Given its cost per test, rate of testing, and quality of the data, the German LDW is a cost-effective testing device for quality control and assurance during subgrade and capping-layer compaction. Marr, A. Why Monitor Geotechnical Performance? Ohio River Valley Geotechnical Seminar, Feb. 2001. The reasons for monitoring geotechnical performance are discussed to help engineers develop justifications for geotechnical instrumentation programs on their projects. A simplified method is presented for esti- mating the potential benefits of a geotechnical instrumentation pro- gram. The techniques taught in decision theory can help estimate the potential monetary benefits of a geotechnical instrumentation pro- gram. Application of these techniques can help estimate how much money can be justified for spending on a project to reduce potential risk costs from undesirable consequences. These techniques may show where to concentrate the focus of an instrumentation effort to have the most benefit. Maryland State Highway Administration, PS 304: Geotechnical Perfor- mance Specification, Maryland SHA, Baltimore, 2007. This document is a design requirement specification that stipulates cri- teria for the geotechnical subsurface exploration, geotechnical design, construction, and submittals. Performance criteria are provided for bear- ing capacity for shallow foundations (factor of safety = 3.0), settlements, external stability, (FS = 1.5 global, and FS = 2.0 sliding), axial and lateral load capacity for deep foundations, retaining walls, embankments, soil improvement, cuts, subgrades (minimum k and/or Mr to be verified by FWD). The actual design values are references to ASSHTO 17th Edition. Verification of performance is based on a variety of techniques including FWD (subgrade), wave equation analysis of piles (WEAP) + pile driver analyzer (PDA) (deep foundations), cross-hole sonic logging (CSL) (drilled shaft foundations), settlement plates (embankment settlement), and in-ground instrumentation. McGuire, M., and G. Filz. Specifications for Embankment and Subgrade Compaction, Technical Report. Virginia Transportation Research Council, Virginia Department of Transportation, Richmond, May 2005. Six approaches were developed for specifying embankment and subgrade compaction and/or verifying compaction quality on Virginia Depart- ment of Transportation (VDOT) construction projects. This study con- sidered the use of a pay factor for embankment construction to motivate the contractor to deliver high-quality compaction. The pay factor devel- oped for this project links the contractor’s payment to the results of the field density tests. A shortcoming of this approach is that it increases the potential for disputes between the contractor and VDOT because every density test has the potential to influence the contractor’s pay. Minnesota Department of Transportation. Specification 2111: Aggre- gate Base. Minnesota DOT, St. Paul. http://www.mnroad.dot.state .mn.us/pavement/GradingandBase/G&BSpecs/2211_Sept_2003.pdf. Accessed Oct. 28, 2007. A key element of this specification is the requirement that the aggregate meet minimum penetration index values based on dynamic cone pen- etration (DCP) testing, which provides a performance measurement of strength. The specification requires the full thickness of each layer to be compacted to achieve a penetration index value less than or equal to 10 mm (0.4 in.) per blow, as determined by a Minnesota DOT (MnDOT) standard DCP device. For test purposes, a layer is consid- ered to be 75 mm (3 in.) in compacted thickness, but a testing layer can be increased in thickness to a maximum of 150 mm (6 in.) if compacted in one lift by a vibratory roller. At least two passing DCP tests should be conducted at selected sites within each 800 m3 (1000 cubic yard) of constructed base course. If either of the tests fails to meet the specified requirements, the material represented by the test should be recom- pacted and retested for DCP index compliance. Minnesota Department of Transportation. Special Provision 5305-55, (2106) Excavation and Embankment—Quality Compaction by IC, LWD, & Test Rolling. Minnesota DOT, St. Paul, 2007. The pilot specification developed and implemented on this project was written to require use of intelligent compaction (IC) technology as the primary QC tool. In brief, the contractor was required to develop a QC procedure that incorporated IC measurement values gathered from con- trol (or calibration) strips. After constructing the control strip, the con- tractor was required to detail how its QC procedure would be implemented on the remainder of the project (e.g., anticipated number, pattern and

85 speed of roller passes, potential corrective actions for noncompliant areas). The specification was written to ensure that grading materials were uniform and to confirm acceptable moisture contents. Following successful control strip construction and development of the QC pro- cedures, proof layers (predetermined layers that required QC mea- surements by the contractor and QA by the engineer) were constructed. For proof layers, the engineer (1) observed the final IC recording pass; (2) reviewed and approved the QC data, documenting that acceptable compaction results were obtained; (3) performed companion and veri- fication moisture content testing; and (4) observed test rolling results to ensure compliance (less than 50 mm rut under wheel of 650 kPa (95 lb/in.2) tire pressure). IC target values (IC TV) for all proof layers were obtained on the 1.2-m (4.0-ft) layer of each control strip—unless the layer thickness was less than 0.75 m (2.5 ft). In that case, the IC TV was obtained on a 0.6-m (2.0-ft) layer of the strip. All segments were to be compacted so at least 90% of the IC measurement values were at least 90% of the IC TV before placing the next lift. If localized areas had IC measurement values of less than 80% of the IC TV, the areas were to be recompacted. If a significant portion of the grade was more than 30% in excess of the selected IC TV, the engineer reevaluated the IC TV. Ohrn, G., and C. Schexnayder. Effect of Performance-Related Specifi- cation on Highway Construction. Practice Periodical on Structural Design and Construction, Vol. 2, No. 4, November 1997, pp. 172–176. On the basis of interviews with interested parties, this paper documents advantages and disadvantages of performance-related specifications related to highway construction. The interviews were used to identify major roadblocks to implementation, including (1) performance specifi- cations can be complex and (2) resistance to change comes from both transportation agencies and industry. Benefits to implementation include improved quality and not wasting effort on measuring parameters that are not meaningful to performance. Projects considered suitable for implementation include large highway projects but not small mainte- nance projects. Additional survey questions elicited discussion of the effect on change orders, safety, final quality, and construction claims. Rec- ommendations include defining what quality characteristics should be measured, developing a consensus on pay factors (including positive pay factors if negative pay factors are used), and providing test methods for assessing quality characteristics during the construction process. Ohrn, G., and C. Schexnayder. Performance-Related Specifications for Highway Construction. Journal of Construction Engineering and Management, Vol. 124, No. 1, January/February 1998, pp. 25–30. This paper builds on the preceding paper and expands on it by differ- entiating between traditional and performance-based specifications, explaining an approach for developing a specification, summarizing pre- vious research in this area, and presenting an example from New Jersey in which a pay factor was established for rigid pavement. Petersen, D., M. Erickson, R. Roberson, and J. Siekmeier. Intelligent Soil Compaction: Geostatistical Data Analysis and Construction Specifications. Presented at 86th Annual Meeting of the Transporta- tion Research Board, Washington, D.C., Jan. 2007. The authors describe two projects in which intelligent compaction specifications adapted from standard earthwork and embankment specifications were implemented. Used together, intelligent compaction rollers, specifications, and geostatistics offer the promise of reducing maintenance costs and increasing pavement life by helping project per- sonnel find and fix subgrade compaction problems before pavement placement. Petersen, J., S. Romanoschi, M. Onyango, and M. Hossain. Evaluation of Prima Light Falling-Weight Deflectometer as Quality Control Tool for Compaction of Fine Grained Soils. Presented at 86th Annual Meeting of the Transportation Research Board, Washington, D.C., Jan. 2007. Compaction of embankment soils is a key factor influencing premature pavement distresses. The authors investigated the use of the Prima 100 light falling weight deflectometer (L-FWD) to measure in situ soil stiff- ness and the feasibility of developing a stiffness-based specification for embankment soil compaction quality control that uses the L-FWD measured stiffness. They found that the predicted in situ resilient mod- uli did not correlate with the in situ measured moduli. Therefore, they could not develop a quality control scheme for embankment soil stiff- ness that employs stiffness requirements established on the basis of the results of laboratory resilient modulus tests. In addition, the testing proved that the in situ modulus of fine-grained soils has a high degree of spatial variability. This prevented the development of a quality con- trol procedure based on stiffness measured on a control test strip. Pinto, S. Proposal of Performance-Based Specifications for Selection and Placing of Natural Materials for Road Embankment Construction. Proc., 22nd PIARC World Road Congress, Durban, South Africa, 2004. The paper proposes to abandon soil classification and to select soils for road construction according to a CBR-based procedure which chooses compaction energy and compaction water content range to achieve good mechanic and volume stability. This procedure should allow the use of almost every soil for embankment, avoiding the need to consider materials not compliant with specifications. Rogers, C., P. Fleming, and M. Frost. A Philosophy for Performance Specification for Road Foundations. Proceedings of the Institute of Civil Engineers, Transport, Vol. 157, No. 3, Aug. 2004, pp. 143–151. This paper focuses on development of performance specifications for pavement foundations. A flowchart summarizes the performance specifi- cation philosophy. To develop a specification, the types of in situ tests and test equipment need to be identified. Performance parameters needed for pavement design include resilient elastic modulus, shear strength, and resistance to permanent deformation. During the design stage, laboratory material properties are determined. Next, to assess the materials and con- struction equipment, a field trial can be conducted. Finally, an assessment is carried out during field construction to verify achievement of the speci- fied parameters. Schaefer, V., and D. White. Quality Control and Performance Criteria for Ground Modification Technologies. Proc., Geotechnical Engineering for Transportation Projects, Geotechnical Special Publication No. 126, ASCE, Los Angeles, Calif., 2006, pp. 1935–1942. The paper describes a conceptual framework for linking QC and QA test- ing with performance-based criteria for ground modification techniques. The proposed method is general and can be applied to virtually any ground modification technology. The method provides a way to compare the relative merits of various ground modification strategies through a direct link to performance of the proposed system. The method has been demonstrated through examples of deep dynamic compaction and stone columns. Comparisons of QC and QA testing with performance results for additional case histories will be needed to verify the generality of the method to other ground modification techniques. Utah Department of Transportation. Geotechnical Performance Specification, Utah I-15 NOW Project. Utah DOT, Salt Lake City, Dec. 9, 2006. This document describes geotechnical investigations, analyses, and design for all components of the Utah DOT I-15 NOW project. A number of references are cited in the specification concerning design methodologies to be used. Design parameters included settlement bearing capacity, stabil- ity, and seismic performance criteria. Performance criteria were stated as

86 “(A) Perform analyses, prepare design, and construct the Project to limit the longitudinal and transverse settlement of the roadway, structures, embankments, and other Project facilities as specified in Part 9 Warranty Provisions; and (B) Mitigate and otherwise be responsible for all distress to structures and properties adjacent to the corridor that is caused by the Project (both directly and indirectly) as specified in the Part 9 Warranty Provisions. The Department will evaluate the Design Builder’s compliance with these performance requirements based on the profilograph measure- ments, as required in the Pavement Performance Specification and the allowable settlement criteria specified in the Part 9 Warranty Provisions.” Walsh, K., W. Houston, and S. Houston. Field Implications of Current Compaction Specification Design Practices. Journal of Construction Engineering and Management, Vol. 123, No. 4, 1997, pp. 363–370. Variations in both the field density and the laboratory-determined ref- erence maximum dry density arise from numerous sources. A corre- sponding spatial variability of relative compaction should therefore be anticipated. This paper provides a comprehensive evaluation of poten- tial problems in compaction control and addresses the sources of field variability in relative compaction. White, D. J., K. L. Bergeson, and C. T. Jahren. Embankment Quality: Phase III, Final Report. Iowa DOT Project TR-401. Center for Trans- portation Research and Education, Ames, Iowa, June 2002. Contractor QC and Iowa DOT QA special provisions were developed and tested by constructing a full-scale embankment project. Surficial density testing was shown not to be adequate for indicating the unifor- mity and stability of the embankment soils. The DCP test was able to detect nonuniformity and development of “Oreo cookie” effects requir- ing corrective action. One of the primary questions Phase III asked was, “Was the quality improved?” The project involved a “quality conscious” contractor, well-qualified and experienced Iowa DOT field personnel, a good QC consultant technician, and some of the best soils in the state. If the answer for this project was yes, the answer would unquestionably be yes for other projects as well. In the authors’ opinion, the answer for this project was indeed yes: the quality was improved, as evidenced by the DCP test data and the amount of disking required to reduce the moisture content to within acceptable control limits. pavement Portland Cement Concrete (PCC) Pavement ARA, Inc. Design of New and Reconstructed Rigid Pavements. Chap- ter 4 in Guide for Mechanistic-Empirical Design of New and Rehabili- tated Pavement Structures. Final Report, Part 3: Design Analysis, NCHRP Project 1-37A. Transportation Research Board of the National Academies, Washington, D.C., 2004. This document is part of the final report for the new AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). Chapter 4 covers the design of rigid (PCC) pavements. Although the MEPDG is a guide for design of new and reconstructed pavement, the models contained within the guide predict the performance of rigid pave- ments on the basis of project-specific inputs for design. Rather than relying solely on empirical performance models (as the PaveSpec soft- ware does), the MEPDG software includes mechanistic modeling and finite element analysis to predict pavement performance on the basis of site-specific inputs. The design process is described as follows: The process requires an iterative hands-on approach by the designer. The designer must select a trial design and then analyze the design in detail to determine if it meets the established performance criteria. The performance measures considered in this guide include joint faulting and transverse cracking for [jointed plain concrete pavement] JPCP, punchouts [and crack width and LTE] for CRCP, and International Roughness Index (IRI) for both pavement types. . . . The designs that meet the applicable performance criteria at the selected reliability level are then con sidered feasible from a structural and functional standpoint and can be further considered for other evaluations, such as life-cycle cost analysis and envi- ronmental impacts. Although this guide does not explicitly calculate pay factors for use in performance specifications, it provides an analysis tool with which the sensitivity of different construction-related inputs (or ACQs) can be evaluated such that pay factors can be determined for inclusion in a performance specification. One advantage of this guide is that it consid- ers both JPCP as well as CRCP for new construction. The guide also considers site-specific climatic conditions through the Enhanced Inte- grated Climatic Model (EICM) and site-specific traffic loading through axle load spectrum. ARA, Inc. PCC Rehabilitation Design of Existing Pavements. Chapter 7 in Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures. Final Report, Part 3: Design Analysis, NCHRP Project 1-37A. Transportation Research Board of the National Acade- mies, Washington, D.C., 2004. This document is part of the final report for the new AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). Chapter 7 covers the design of rigid (PCC) pavements for rehabilitation of exist- ing pavements. As described by this chapter, Several different rehabilitation strategies using PCC can be applied to existing pavements to extend their useful service life. These range from the combination of repair and preventative treatments such as full-depth repair and diamond grinding of existing jointed plain concrete pavement (JPCP) to the placement of unbonded JPCP or CRCP overlays over exist- ing flexible, composite, or rigid pavements, to the placement of bonded PCC overlays over existing JPCP or CRCP, to the reconstruction (including adding additional lanes) of existing pavements with JPCP or CRCP. As with the use of this guide for new pavements, the MEPDG can be used to predict pavement performance for PCC rehabilitation as well, such that proper performance parameters and pay factors can be devel- oped for true performance specifications. Considered in a rapid renewal environment, rehabilitation—rather than reconstruction—may be the best alternative, and the use of this guide will assist in developing per- formance specifications for such a purpose. California Department of Transportation. Replace Concrete Pave- ment (Rapid Strength Concrete). Standard Special Provision 40-020. Caltrans, Sacramento, 2006. This document is a special provision used by Caltrans for concrete pave- ment reconstruction using rapid strength concrete. The specification details the material requirements and construction practices to use for replacement of concrete pavement, but permits the contractor to select the mixture design for the rapid strength concrete. Specific mixture durability requirements are provided for the mixture supplied by the contractor. In addition to mixture durability requirements, pay factor adjust- ments (and rejection criteria) are specified on the basis of the required modulus of rupture at opening to traffic and at 7 days. In a rapid renewal environment, this specification provides guidance for the use of rapid setting concrete for rapid pavement reconstruction. One of the key fea- tures of this specification is that it permits the contractor to select the mixture design for the paving mixture.

87 Evans, L., M. I. Darter, and B. K. Egan. Development and Implementa- tion of a Performance-Related Specification: I-65 Tennessee. Research Report for Contract No. DTFH61-03-C-00109. Federal Highway Administration, U.S. Department of Transportation, March 2005. This report provides a summary of the development of a PCC pave- ment PRS and its implementation on a project constructed on Inter- state 65 near Nashville, Tennessee. Appendix A of the report provides the final PRS used for the technical special provision for the project on I-65 in 2005. The report discusses the selection of the AQCs and appro- priate target values for the mean and standard deviation, RQLs, and MQLs for each. For this PRS, 28-day compressive strength, slab thick- ness, and initial smoothness (Profilograph Index) were selected as the AQCs. The report compares the Tennessee DOT (TDOT) standard specification pay adjustments for these AQCs with the PRS pay adjust- ments. The report also presents the PaveSpec 3.0 inputs and analysis that were used to develop the pay factor charts for the selected AQCs. While the intent was to use the specification as a shadow specification on the northbound lanes and formally apply it for the southbound lanes, project time constraints only permitted its use as a shadow speci- fication for the whole project. After the project was completed, the PRS pay factors were computed on the basis of test results under the stan- dard specifications. Overall, TDOT, the contractor, and the QC repre- sentative all gave very positive feedback on the PRS process. The final project PRS provided in the appendix of the report does not appear to refer to any TDOT standard specifications. Evans, L., K. L. Smith, N. G. Gharaibeh, and M. I. Darter. Develop- ment and Implementation of a Performance-Related Specification in Florida: State Road 9A (I-295 Leg), Jacksonville. Research Report FHWA-HIF-07-. Federal Highway Administration, U.S. Department of Transportation, Dec. 2006. This report summarizes the development of the PCC pavement performance-related specification (PRS) and its implementation on a project constructed in Florida in 2004–2005. The report discusses the selection of the acceptance quality characteristics (AQC) and appropri- ate target values, rejectable quality levels (RQL), and maximum quality levels (MQL) for each AQC, on the basis of previous concrete pavement projects in Florida. This work included an analysis of maintenance and rehabilitation activities and costs for PCC pavements based on histori- cal data. The report presents the PaveSpec 3.0 inputs and analysis that were used to develop the pay factor charts for the selected AQCs. A postmortem assessment of the project and the PRS process produced favorable responses from both the Florida DOT (FDOT) and the contrac- tor. Contractors paid more attention to quality control and were pleased with the level of control they were given with developing the mix design and over construction operations. FDOT felt the project was successful, but because of the limited size of the project, a more thorough assessment of the PRS process was difficult. Overall, FDOT decided that moving toward PRS is beneficial and will lead to better PCC pavements in the state. FHWA. Guide to Developing Performance-Related Specifications for PCC Pavements, Volume 1: Practical Guide, Final Report, and Appen- dix A. Research Report FHWA-RD-98-155. Federal Highway Admin- istration, U.S. Department of Transportation, Feb. 1999. http://www .tfhrc.gov/pavement/pccp/pavespec/index.htm. This report is a precursor to the report by Hoerner, Darter, Khazanovich, Titus-Glover, and Smith (2000); it provides a more thorough history of the development of PRS for PCC pavements. This report also doc- uments the process for the development of the PaveSpec Version 2.0 software. The report provides additional insight into the goals of a true PRS for PCC pavement, notably, tying AQCs measured at construction with future life-cycle costs (LCC) of the pavement, such that rational pay adjustments can be employed during construction for the various AQCs. LCCs are computed using future maintenance and rehabilitation activi- ties which are determined on the basis of prediction models for dis- tresses, such as slab cracking, joint spalling, joint faulting, and roughness. The following figure taken from the report, denoted here as Figure D.1, graphically depicts the LCC-based PRS method. As Figure D.1 describes, as-designed mean and standard deviation values are established for the various AQCs, which are then used to compute as-designed life-cycle costs using distress indicator models and anticipated maintenance and rehabilitation costs. The as-designed LCCs are compared with as-constructed LCCs, which are computed using distress indicator models based on the actual measured AQCs during or shortly after construction. This comparison of as-designed to as-constructed LCCs permits the agency to establish rational pay adjustments on the basis of the measured AQCs. Of note in this report are the three different levels of PRS that can be anticipated given the amount of information available for each project and the AQC measurement capabilities. • Level 1 PRS (simplified PRS) uses acceptance testing similar to the procedures currently used by state highway agencies. Computation of pay factors is based on calculating independent pay factors for each AQC, assuming all other AQCs are equal to the target values. Pay factors are a function of the measured as-constructed mean and stan- dard deviation, target mean and standard deviation, and sample size. Final payment for each lot of pavement constructed is based on a composite pay factor equation. • Level 2 PRS (transitional PRS) compare simulated as-designed LCC with as-constructed LCC for each lot to determine the pay adjust- ment. LCCs are computed on the basis of AQCs, and pay adjustments are based on the premise of liquidated damages. Acceptance testing of AQCs is conducted with the best techniques available, preferably in situ and nondestructive testing. Using computer simulations of pavement performance based on all of the AQCs obviates the need to combine individual pay factors as with Level 1 PRS. • Level 3 PRS (ideal PRS) considers many more AQCs that are not cur- rently measurable. All AQCs that affect pavement performance are under the control of the contractor. Level 3 PRS uses LCC procedures similar to Level 2, except that all AQCs are determined using in situ, rapid, nondestructive methods. Figure D.1. PRS methodology for PCC pavements.

88 The appendix of this report provides a guidance performance-related specification for jointed plain concrete pavement which is based on the methodology described in the report (only for Level 1 and Level 2 PRS). FHWA. Guide to Developing Performance-Related Specifications for PCC Pavements, Volume 2: Appendix B—Field Demonstrations. Research Report FHWA-RD-98-156. Federal Highway Administra- tion, U.S. Department of Transportation, February 1999. http://www .tfhrc.gov/pavement/pccp/pavespec/index.htm. This publication is Appendix B of the report discussed above. It dis- cusses a prototype PRS that was developed under this effort and the specific PRS that were developed for shadow field trials on projects in Iowa, New Mexico, Missouri, and Kansas. FHWA. Guide to Developing Performance-Related Specifications for PCC Pavements, Volume 3: Appendices C–F. Research Report FHWA- RD-98-171. Federal Highway Administration, U.S. Department of Transportation, Feb. 1999. http://www.tfhrc.gov/pavement/pccp/ pavespec/index.htm. This publication contains additional appendices for the original report which include • Appendix C, Literature Search Summary; • Appendix D, Typical Acceptance Quality Characteristic Variability; • Appendix E, Distress Indicator Prediction Models; and • Appendix F, Annotated Bibliography. Florida Department of Transportation. Technical Special Provisions for Performance-Related Specification for Rigid Pavement. Project ID 209600-1-52-01, State Route 9A (I-295 Leg), Duval County. Florida DOT, Tallahassee, 2001. This PRS was developed for a PCC pavement project in Duval County, Florida, and is based on the PRS methodology developed under FHWA efforts using the PaveSpec 3.0 software. The AQCs used in this specifica- tion are 28-day compressive strength, slab thickness, and initial smooth- ness (Profile Index). RQLs and MQLs are provided for each AQC. A composite lot pay factor equation is provided which includes the pay factors for the three AQCs (strength, thickness, and smoothness), based on the mean and standard deviation for each AQC. This PRS is a trial specification used for a specific project, not a stan- dard special provision. This PRS does refer to FDOT standard specifica- tions for PCC pavement, with modifications noted in the specification. Hoerner, T. E., M. I. Darter, L. Khazanovich, L. Titus-Glover, and K. L. Smith. Improved Prediction Models for PCC Pavement Performance- Related Specifications, Volume 1: Final Report. Research Report FHWA-RD-00-130. Federal Highway Administration, U.S. Depart- ment of Transportation, Sept. 2000. This report documents the development of PRS for PCC pavement over the past 20 years. The work conducted under the effort described in this report was used to upgrade the FHWA PaveSpec PRS software to ver- sion 3.0. This upgrade to the PaveSpec 2.0 software provided (1) improved or validated distress indicator models, (2) the ability to calibrate or mod- ify the default distress indicator models, (3) sensitivity analysis capabili- ties, and (4) the ability to assess risks to both the contractor and agency through the development of project-specific expected pay charts. As stated in this report, A PRS is a construction specification that describes the desired level of key materials and construction acceptance quality characteristics (AQCs) that have been found to correlate with fundamental engineering properties that predict performance. These AQCs (e.g., smoothness, thickness, strength, air content, and percent consolidation around dowels) are amenable to acceptance testing at the time of construction. The basis for PRS for PCC pavement, described in detail in this report, is the use of distress indicator models to predict the performance of PCC pavement over time on the basis of specific AQCs that can be mea- sured during or shortly after construction. The AQCs which were con- sidered include • Slab thickness; • Concrete strength (flexural or compressive); • Air content; • Initial smoothness; and • Percent consolidation around dowels. Although included in the software, the final AQC listed (percent con- solidation around dowels) is generally not included in most PCC PRS because that AQC is difficult to measure accurately in the field at pre- sent. The report recommends consideration and inclusion of additional AQCs once distress indicator models can be developed for them. The additional AQCs include joint sawing depth, surface texture, concrete mixture components, base course quality, subgrade quality, air content characteristics, calorimetry, and coefficient of thermal expansion (CTE) of the paving concrete. On the basis of the AQCs presented above, distress indicator models are used to predict some of the most commonly observed PCC pave- ment distresses, which encompass both structural performance as well as functional performance. Improved distress indicator models were provided for • Transverse joint faulting; • Transverse fatigue cracking; • Transverse joint spalling; and • International Roughness Index (IRI) (roughness/roughness pro gression). A detailed investigation of available pavement performance data and distress indicator models was conducted to identify the most suitable models and performance data for inclusion. This investigation consid- ered pavement performance data collection efforts and databases from a number of sources including FHWA Rigid Pavement Performance and Rehabilitation (RPPR) study, Long-Term Pavement Performance (LTPP), NCHRP concrete pavement evaluation system (COPES), AASHO Road Test, and Mn/ROAD. This report provides a comprehensive summary of the distress indicator models that were used and the process of select- ing the models. Note that this report and the PaveSpec 3.0 software consider only new jointed plain concrete pavement and not continuously reinforced concrete pavement (CRCP), jointed reinforced concrete pavement (JRCP), or bonded/unbonded concrete overlays. Also note that the smooth- ness prediction models are based on IRI, even though most state high- way agencies currently use profilograph testing to determine initial smoothness. A significant effort was devoted to establishing a correla- tion between IRI and the Profilograph Index (PI) that could be used by the software. Indiana Department of Transportation. Special Provision, Section 509: Performance-Related Specification for Portland Cement Concrete Pavement. Indiana DOT, Indianapolis, 2003. This PRS for PCC pavement evolved from the work described in the FHWA and Hoerner et al. reports in this section. The Indiana PRS is per- haps one of the most mature as it was developed and refined on the basis of trial projects in Indianapolis (I-465 and I-70) and Clarksville (I-65), using PaveSpec 3.0 software and life-cycle cost PRS procedures. The spec- ification states, “Pay adjustment is based on the AQC quality-related

89 increase or decrease in future LCCs expected to be incurred by the Department over the analysis life of the project.” The AQCs selected by the Indiana DOT (INDOT) for this PRS are flexural strength (a 28-day equivalent of the measured 7-day strength), slab thickness, air content, and initial smoothness (PI). Acceptance test- ing of the AQCs is based on lots for strength, thickness and air, and sections for smoothness. RQLs and MQLs are provided for each AQC to ensure that no unacceptable levels of AQCs are provided and that the maximum pay adjustment is within reason. Additional acceptance cri- teria based on unit weight and water/cement ratio are also provided but are not included in the pay adjustment. A composite pay factor is pro- vided for strength, thickness, and air AQCs based on the mean and stan- dard deviation for each lot, and a separate pay factor is provided for smoothness based on the mean and standard deviation for each section. These are combined for an overall performance-related adjustment. The contractor is required to submit a quality control plan for the pav- ing operation at least 15 days before commencement of paving. While this is considered a performance-related specification, the speci- fication still prescribes material requirements and certain mix design limits, and it refers extensively to INDOT standard specifications for both material and construction requirements. INDOT has extensive experi- ence with QC and QA specifications and some experience with warran- ties for PCC pavement. The intent of moving to a PRS was to reduce the variability of the PCC paving operation that was experienced under QC/QA contracting. Minnesota Department of Transportation. Standard Specification Section 2301: Concrete Pavement. Minnesota DOT, St. Paul, 2005. Section 2301.2A5d provides an optional incentive for well-graded aggregate of up to $2.00 per cubic yard. The two provisions for grada- tion to achieve this incentive are based on the ACI 8%–18% retained chart (7%–18% for a lower incentive). Although this is not a true perfor- mance specification, MnDOT believes it will achieve better long-term performance from the pavement if a well-graded aggregate is used. This gradation incentive provides the contractor with some flexibility in achieving it. Section 2301.2A7b(5) provides an incentive/disincentive clause for water-cement ratio of the concrete mixture used for paving. Although this is not a true performance specification, it will help ensure that either a more durable, long-lasting concrete mixture is used or that the contractor does not receive full payment. This specification is based on well-understood principals that concrete mixtures with lower water- cement ratios generally result in more durable pavement. This type of specification will reward contractors who make an effort to carefully develop their mix design and monitor it carefully during the mixture production process. Section 2301.2A5c provides an incentive/disincentive clause for coarse aggregate quality based on absorption and percent carbonate for differ- ent classes of aggregate. Although this is not a true performance specifi- cation, it is based on payment for material quality, which directly affects pavement performance. If a better quality aggregate is used, better per- formance is expected from the pavement, and therefore the contractor receives an incentive. Section 2301 also contains a deduction/disincentive clause for slab thickness deficiency and an incentive/disincentive clause for surface smoothness, which are fairly standard clauses for PCC pavement specifications. Minnesota Department of Transportation. Special Provision S-111, Section 2301: Concrete Pavement. Minnesota DOT, St. Paul, 2006. This special provision provides a requirement for PCC pavement tex- ture in term of average texture depth, as measured according to ASTM E965 (sand patch test). Although this is not a true performance specifi- cation, it dictates what is required from the contractor in terms of tex- ture without prescribing the technique for achieving it. Missouri Department of Transportation. Standard Specifications Section 502: Portland Cement Concrete Base and Pavement. Missouri DOT, Jefferson City, 2004. Section 502.4.8.3 provides requirements for surface texture. Acceptance is based on the minimum texture depth provided as measured using ASTM E965 (sand patch test). The contractor can opt to diamond grind or tine the surface (in accordance with a prescribed technique) in lieu of this requirement. This specification represents a potential component of a performance specification. The contractor is required to achieve a minimum texture depth, but the specification prescribes the technique for achieving it. Although texture is not itself a functional performance measure, it has a direct impact on friction and tire-pavement noise. Morgan, P. (ed.). Guidance Manual for the Implementation of Low- Noise Road Surfaces. FEHRL Report 2006/02. Forum of European National Highway Research Laboratories, Brussels, Belgium, 2006. This report is one of the products from the Sustainable Road Surfaces for Traffic Noise Control (SILVIA) framework project in Europe. The SILVIA project was initiated to help develop solutions for addressing roadway noise issues (specifically, tire-pavement noise) at the pavement level by providing guidance for low-noise pavement surfaces. This report pro- vides background information on tire-pavement noise issues and an overview of low-noise solutions for pavement surfaces. The manual also summarizes the different measurement methods that are available for the evaluation of the acoustic performance of a road surface, presents a noise classification procedure that provides accurate and reproducible characterization of the acoustic performance of a specific pavement, and presents a conformity-of-production (COP) method for assessment. The manual also addresses some of the economic considerations of speci- fying low-noise surfaces. North Dakota Department of Transportation. Dowel Bar Warranty, Special Provision IM-2-094(064)275. North Dakota DOT, Bismarck, 2005. This special provision is a warranty specification for dowel bars. The specification is not explicitly for dowel-bar retrofits or new construc- tion and could likely be used for either. The specification is based on measurement of load transfer across the joint as quantified by deflec- tion testing, as well as visual inspection of the area around the dowel bars for visual distresses. Thresholds are provided for full payment, reduced payment, and rejection of joints. No warranty period is speci- fied, but two rounds of testing are conducted. The first test is conducted shortly after construction, the second test after approximately 1 year in service. Testing is conducted between September 1 and November 1, presumably when joints are neither locked up nor opened widest. Österreichische Forschungsgesellschaft (FSV). Austrian PCC Pave- ment Specification: Concrete Pavements—Pavement Construction, RVS 08.17.02. 2007. This specification is essentially a warranty specification for PCC pave- ments. However, it includes some unique components that could be considered for PCC pavement performance specifications. Notably, deductions are assessed for excessive rolling noise (functional perfor- mance) and also for poor performance in a freeze-thaw condition result- ing from poor air void parameters. The specification requires a guarantee (warranty) of 5 years by the contractor after the completion of con- struction, but requires a 2-year extension of that guarantee if certain performance parameters are not achieved. Section 9, Acceptance Factors, and Section 10, Deductions for Substan- dard Quality, describe the pavement characteristics which are measured

90 during and shortly after construction, and the potential deductions asso- ciated with each. While the deductions (penalties) for substandard qual- ity are more extensive than most used in the United States, the pay deductions are not clearly tied to life-cycle costs for the pavement. The acceptance parameters include the following: • Thickness. No deductions are specified for deficiencies in smooth- ness, but if the thickness measured from any given test core is more than 2 cm less than the required thickness, the guarantee period is extended for 2 years for the test lot. • Strength (splitting tensile). Deductions are specified for substandard strength. If the concrete strength is more than 15% lower than the prescribed strength at any test location, the contractor has the option to remove and replace the pavement, or accept an additional pay deduction, presumably reflecting the expected loss of pavement life. • Evenness (profilograph testing). Deductions are specified for substan- dard evenness. If the contractor elects not to take corrective action for substandard evenness, an additional deduction is applied. • Cracks. Deductions are specified on the basis of the area of cracked slabs. • Air content. Air content is measured at the point where the concrete is placed. If an impermissible air void parameter is established during the conformity tests on fresh concrete, cores are taken to determine the air void parameters. If the air void parameters are still substan- dard, the guarantee period is extended by 2 years. If damage occurs to the pavement in the presence of deicing chemicals during the guarantee period because of a lack of scaling resistance from substan- dard air voids, deductions are assessed. • Excessive rolling noise. For exposed aggregate surfaces, deductions are assessed for excessive rolling (tire-pavement) noise. • Substandard skid resistance. Deductions are specified for substandard skid resistance, which is evaluated at “handover” (up to 12 weeks after pavement has been open to traffic), during the guarantee period, and at 4 weeks to 16 weeks before the end of the guarantee period. The guarantee period is extended by 2 years if skid resistance is substan- dard at the end of the guarantee period. Additional deductions are assessed as a penalty for traffic restrictions when correcting skid resistance or for traffic restrictions because of low skid numbers. Sandberg, U. Low-Noise Road Surface Classification and Procure- ment System in Japan. Commentary. Swedish National Road and Transport Research Institute (VTI), Linköping, Sweden, 2005. A performance-based system for procurement of low-noise road surfaces has been used for several years in Japan. Performance is based on pre- scribed tire-pavement noise levels. Tire-pavement noise levels are mea- sured by the transportation authority using special vans (Road Acoustic Checkers) equipped with a special tire as a fifth wheel, and using a method resembling the close proximity (CPX) method. Currently, five such vans operate in Japan; they are calibrated once per year at a test track facility at the Public Works Research Institute in Tsukuba City. For construction acceptance, tire-pavement noise levels are mea- sured twice: soon after completion of construction and after 1 year of traffic exposure. The noise level should not exceed 89 dB soon after the completion of the road surface and should not exceed 90 dB 1 year later. An extra dB is allowed for projects in which an aggregate of poorer quality has been accepted. A special version of the procurement system allows for surfaces which are quieter than the normal requirement; the intent is to encourage development of quieter surfaces than those that merely meet the required limits. In this case, the bidder (the road con- tractor) specifies a lower noise level for his surface than the required limit; after construction, the measured noise level must not exceed that specified level. One year after completion, an increase of one dB is allowed. Beginning with only three procurement contracts under this system in 1999, the number of contracts increased to about 130 in 2003; and that number is still steadily increasing. Wisconsin Department of Transportation. Technical Special Provi- sions for Performance-Related Specification for Rigid Pavement. Item SPV.0055.01, Project ID 1011-01-88, IH-39/90/94. Lake Delton– Madison Rd., District 1, Dane County. Wisconsin DOT, Madison, 2005. This PRS was developed for a PCC pavement project in Dane County, Wisconsin, and is based on the PRS methodology developed under FHWA efforts using the PaveSpec 3.0 software. The AQCs in this speci- fication are 28-day compressive strength, thickness, air content, and ini- tial smoothness (Profile Index). RQLs and MQLs are provided for each AQC. A composite lot pay factor equation is provided which encom- passes pay factors for all four AQCs (smoothness, air, strength, and thickness), based on the mean and standard deviation for each AQC. Although this is considered a performance-related specification, it does refer to PCC pavement standard specifications for materials, mix- tures, and so on—much of which contain prescriptive requirements. Wisconsin Department of Transportation. Pavement Dowel Bars Retrofit Warranted. Item 416.0623.S. Wisconsin DOT, Madison. This document is a warranty specification for dowel bar retrofit (DBR) projects. Performance evaluation of the DBR is based on a visual dis- tress survey of retrofit. The warranty covers material and workmanship for a 3-year period. The specification provides materials and construc- tion procedures for DBR, including tolerances for dowel bar alignment. Thresholds and remedial actions are provided for the following: • Distressed joints within the DBR slot; • Cracking in the existing pavement between the slots or across slab to pavement edge; • Loss of surface and concrete patch material within the DBR slot; • Debonding of the patch concrete with existing concrete; and • Breakup or dislodgement of concrete patch material within the slot. Hot-Mix Asphalt (HMA) Pavement ARA, Inc. Design of New and Reconstructed Flexible Pavements. Chapter 3 in Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures. Final Report, Part 3: Design Analysis, NCHRP Project 1-37A. Transportation Research Board of the National Academies, Washington, D.C., 2004. This document is part of the final report for the new AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). Chapter 3 covers the design of flexible pavements and is most applicable for pave- ments with conventional dense-graded HMA mixtures; but it can also be used to a limited extent for stone matrix asphalt (SMA), polymer- modified asphalt (PMA), and recycled asphalt pavement (RAP) mix- tures. While the MEPDG is a guide for design of new and reconstructed pavement, the models contained within the guide predict the perfor- mance of flexible pavements on the basis of project-specific inputs for design. The models used in the software are not purely empirical, as with the original AASHO pavement design guide, but also use mecha- nistic modeling to compute stresses and strains to predict performance based on site-specific inputs. As with the rigid pavement section of the MEPDG, the flexible pavement design requires . . . an iterative hands-on approach by the designer. The designer must select a trial design and then analyze the design in detail to determine if it meets the established performance criteria. The flexible performance mea- sures considered in this guide include pavement deformation (rutting),

91 fatigue cracking (both bottom-up and top-down), thermal cracking, and smoothness (International Roughness Index). . . . The designs that meet the applicable performance criteria at the selected reliability level are then considered feasible from a structural and functional standpoint and can be further considered for other evaluations such as life-cycle cost analysis. Although this guide does not explicitly calculate pay factors for use in performance specifications, it provides an analysis tool with which the sensitivity of different construction-related inputs (or acceptance qual- ity characteristics) can be evaluated such that pay factors can be deter- mined for inclusion in a performance specification. One advantage of this guide is that it conducts a comprehensive analysis that considers site-specific climatic conditions through the EICM and site-specific traf- fic loading through axle load spectrum, among other variables. ARA, Inc. HMA Rehabilitation of Existing Pavements. Chapter 6 in Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures. Final Report, Part 3: Design Analysis, NCHRP Project 1-37A. Transportation Research Board of the National Acad- emies, Washington, D.C., 2004. This document is part of the final report for the new AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). Chapter 6 covers the use of HMA for rehabilitation of existing pavements. Reha- bilitation strategies considered by this guide include the following: • HMA overlay of existing HMA surfaced pavements, both flexible and semirigid; • HMA overlay of existing PCC pavement that has received fractured slab treatments, crack and seat, break and seat, and rubbilization; and • HMA overlay of existing intact PCC pavement (JPCP and CRCP), including composite pavements or second overlays of original PCC pavements (not including JRCP). As with the use of this guide for new pavements, the MEPDG can be used to predict pavement performance for HMA overlays, such that proper performance parameters and pay factors can be developed for true performance specifications. In a rapid renewal environment, rehabilitation—rather than reconstruction—may be the best alterna- tive, and the use of this guide will assist in developing performance specifications for such a purpose. Epps, J. A., et al. NCHRP Report 455: Recommended Performance- Related Specification for Hot-Mix Asphalt Construction: Results of the WesTrack Project. TRB, National Research Council, Washington, D.C., 2002. This is the final report from NCHRP Project 9-20, which sought to develop performance-related specifications for hot-mix asphalt pavement con- struction by examining how deviations in materials and construction properties, such as asphalt content and degree of compaction, affect long- term pavement performance. The experiment was also intended to pro- vide early field verification of the SHRP Superpave volumetric mixture design procedure. The testing program consisted of heavy truck loading on 34 different Superpave HMA pavement sections (26 original test sec- tions with eight reconstructed during the testing) constructed on a 1.75-mi closed loop test track (WesTrack pavement test facility) over a 2½-year period. On the basis of the work from this study, a prototype version of the HMA Spec software was developed. The software helps state highway agencies determine pay adjustments for the various HMA pavement quality characteristics such that they relate to the predicted life-cycle costs of the pavement. Part I of the report provides information on the overall WesTrack proj- ect, including construction of the test sections, trafficking operations, and materials characterization and performance models. Part II presents the development of the performance-related specification for HMA pavement and the development of the HMA Spec software. The intent of the PRS and HMA Spec software is to provide rational pay adjustment levels for as-constructed HMA pavement quality that are based on the difference between as-designed life-cycle costs and as-constructed predicted life- cycle costs. The report describes the two levels of performance models that were considered in the PRS. Level 1 models are based on direct regressions among specific performance measures (rut depth or fatigue cracking) and traffic (equivalent single axle loads, or ESALs) and mix characteristics. Level 2 models are based on mechanistic-empirical analyses which assume the pavement behaves as a multilayer elastic system. The performance criteria primarily considered in the PRS develop- ment process were permanent deformation (rutting) and fatigue crack- ing, though other ways to characterize performance exist (e.g., low temperature cracking, roughness, and friction loss). The AQCs consid- ered in the WesTrack experiment were HMA surface layer thickness, initial smoothness, asphalt content, air void content, and an aggregate grada- tion parameter (percent passing the no. 200 sieve). The initial smooth- ness performance criterion is not included in the HMA Spec software. Chapter 10 of the report provides a summary of the Guide Perfor- mance Specification for WesTrack that is contained within the HMA Spec software. The actual Guide Specification is provided in Appendix C of the report as AASHTO Provisional Specification PP 400. Huber, G. A., and J. S. Scherocman. Superpave and WesTrack: Did They Perform as Expected? Proc., 1999 Canadian Technical Asphalt Association Conference, Quebec, Canada, Canadian Technical Asphalt Association, Victoria, British Columbia, 1999. This paper presents a summary of the WesTrack study and a summary of a forensic investigation as to why some of the test sections did not perform as expected. The WesTrack study was intended to provide early field verification of Superpave HMA mixtures; so the poor performance of certain sections was of great concern to the HMA paving community as they were beginning to adopt the Superpave process and construct Superpave HMA pavements. Of particular note was the poor perfor- mance of coarse-graded test sections which were expected to perform better than fine-graded sections and did not. One of the key observations from this investigation was that the use of relatively thin HMA pavement layers—which were intentionally under- designed (thinner than the 20-year design thickness) to ensure that fatigue cracking would occur within the life of the project—may have had unintentional effects on the performance of the test sections. The thin- ner pavement sections resulted in significantly higher deflections under load than normally measured on interstate pavements. As a result, deflec- tions were influenced more than usual by the underlying base and sub- grade and influenced less than usual by differences in the asphalt mixture properties between test sections. This hypothesis and supporting evidence are discussed in greater detail in Appendix A of the FHWA report Per- formance of Coarse-Graded Mixes at WesTrack: Premature Rutting found at http://www.tfhrc.gov/pavement/pubs/westrack/westrack.htm#6. The underlying significance of this investigation is that because of the experimental design (specifically the structural design which would not likely be used for a typical interstate pavement), the pavement sections did not perform as expected. Therefore, the development of performance- related specifications based solely on the WesTrack experiment may not be appropriate. Monismith, C. L., J. A. Deacon, and J. T. Harvey. WesTrack: Perfor- mance Models for Permanent Deformation and Fatigue. Pavement Research Center, Institute of Transportation Studies, University of California, Berkeley, June 2000. This report presents performance models based on the results of the WesTrack experiment that can be used for PRS for HMA pavement. The

92 models developed are for permanent deformation (rutting) and fatigue cracking. The models presented are for both Level 1, based on direct regressions from measured performance, traffic loading, and mix char- acteristics; and Level 2, based on mechanistic-empirical analyses. Chapter 5 of the report discusses how the performance models can be used to develop pay factors that can be used for PRS for HMA pave- ments. The cost model presented considers the present worth of reha- bilitation costs resulting from the as-constructed versus as-designed quality of the pavement. The pay factors consider the quality character- istics of air-void content, asphalt content, HMA thickness, and aggregate gradation. NCHRP. Quality Characteristics for Use with Performance-Related Specifications for Hot Mix Asphalt. NCHRP Research Results Digest 291. Transportation Research Board of the National Acad- emies, Washington, D.C., Aug. 2004. This digest summarizes the key findings of NCHRP Project 9-15, Qual- ity Characteristics and Test Methods for Use in Performance-Related Specifications for Hot Mix Asphalt Pavements, which investigated sim- ple and rapid nondestructive testing procedures for evaluating proper- ties of as-constructed HMA pavements. The tests measure the quality characteristics included in HMA performance-related specifications. This project evaluated the following quality characteristics of as- constructed HMA pavement using the testing devices indicated: • Segregation, measured/evaluated using the Road Surface Analyzer (ROSAN), which detects and measures segregation, calculating esti- mated texture depth, as described in NCHRP Research Report 441; • Initial smoothness, evaluated using the Lightweight Inertial Profiler to determine the IRI; • In-place mat density, measured using the Pavement Quality Indica- tor (PQI) or nuclear density gauge (both appropriate only for dense- graded mixtures); • Longitudinal joint density, measured with the PQI (dense-graded mixtures only); and • In-place permeability, measured using the National Center for Asphalt Technology (NCAT) Field Permeameter, which provides a K-value for the surface. Initial specification criteria and threshold values for these five param- eters for a PRS are presented in the appendix of the final report from Project 9-15. Note, however, that with the exception of measuring smoothness and permeability, the digest recommends further evaluation and validation of these test methods and recommended values before full adoption in a PRS. New Jersey Department of Transportation. Ride Quality Specifica- tion for HMA Pavements. Section 406.13: Acceptance of Surface Course Rideability. New Jersey DOT, Trenton. This ride quality specification for HMA pavements in New Jersey is currently being implemented on pilot projects. The specification is based on the IRI, which best correlates to user perception of ride quality, as opposed to the Profilograph Index or Rolling Straightedge. The key feature of this specification is that the incentive and dis- incentive pay adjustments for the various levels of ride quality were set on the basis of expected pavement life (and associated cost of recon- struction) for those various levels of initial ride quality. For more detail, see the TRB paper, Conceptual Framework for Pavement Smoothness Specification (Weed and Tabrizi 2005; citation follows in this section of Appendix D). This specification is a true performance specification, as it pertains to ride quality, in that the pay adjustments for initial measure ride quality are based on expected performance of the pavement. Transit New Zealand. New Zealand Performance-Based Specifica- tion for Structural Design and Construction of Flexible Unbound Pavements. TNZ B/3 (provisional). 2000. This is a provisional specification (current status unknown) for the design, maintenance, and performance requirements for flexible unbound pave- ment layers for the construction of new pavements and reconstruction of existing pavements. The contractor is responsible for the pavement design, including selection of materials, layer thicknesses, drainage, and the binder type. The contractor is also responsible for maintaining the pavement and seal, including the shape and structural integrity of the pavement, for 12 months after construction. The notes to this specification state, This performance based specification has primarily been developed to allow the use of any material (lightly stabilized or otherwise) in the pavement. . . . There is some risk in allowing the use of alternative materials as other factors can not be assessed in the laboratory such as: constructability; seal adherence; and environmental performance. Therefore, it was considered appropriate that the Contractor is respon- sible for the pavements performance for the maintenance period of at least 12 months as per this specification. Compliance assessment requirements are provided for • Pavement design; • Pavement materials (quality plan required from contractors); • Pavement layer compaction; • Pavement stiffness (moduli) or strength; • Surface shape; • Rut depth; • Roughness; • Surface texture (minimum texture depth from sand patch test); • Chip retention; • Surface waterproofness; • Saturation before sealing (moisture content of pavement surface before sealing); and • Repairs. While this specification does provide certain performance parameters for the pavement, it stipulates a 12-month maintenance period, which is essentially a warranty. These performance parameters, however, should be analyzed and considered for potential performance specifications for HMA pavement. Transit New Zealand. New Zealand Performance Based Specification for Hotmix Asphalt Wearing Course Surfacing. TNZ P/23. 2005. This specification provides performance requirements for open graded porous asphalt (OGPA), textured high stress resistant type asphalts (e.g., SMA), and dense graded asphaltic concrete (DGAC). Material performance requirements are provided for the binder properties and aggregate properties tested using standard test procedures. Mixture and finished pavement performance criteria are provided as follows: • Open graded porous HMA: surface ride, permeability, and mix design properties; • Textured high stress resistant HMAs: surface ride, permeability, safety (texture depth), and mix design properties; • Dense graded asphaltic concrete: surface ride, permeability, mix design properties; and • Minimum voids in mineral aggregate (VMA) requirements for DGAC and SMA based on traffic conditions and mix nominal size. This specification is termed a performance-based specification, and it provides criteria for different types of mixtures in terms of material,

93 mixture, and finished surface properties. The key aspect of this specifica- tion is that it does not prescribe the mixture and construction require- ments; rather, it requires the contractor to develop the mixture (meeting certain performance requirements) and control construction operations such that the finished surface will meet certain functional performance requirements. Transit New Zealand. New Zealand Performance-Based Specification for Reseals. TNZ P/17. 2002. This specification provides performance requirements for reseal (chip seal) operations to restore HMA surface texture. The contractor is given control of design and construction of the reseal method. For compli- ance assessment, the surface texture and chip retention is tested between 10 months and 12 months after construction. The performance criteria, method of assessment, test methods, and threshold values specified for the reseal include the following: • Safety: skid resistance, light reflectance, chip retention, site safety, color uniformity, and roadmarking (striping) contrast; • Environmental: noise (measured using texture depth); • Waterproofness: impermeability (measured through chip size); • Economics: tyre [tire] wear [indicated by aggregate present system value (PSV) and texture depth] and rolling resistance (indicated by texture depth); and • Durability: aggregate (crushing value and weathering resistance), bitumen (durability and flux content), bitumen application rate ( texture depth). An additional performance requirement is specified: any areas repaired more than 9 months after construction will be subjected to an additional 12-month maintenance period; and if the area of repairs at the end of 12 months are greater than 10% of the section, the section will be subject to an additional 12-month maintenance period. This performance-based specification uses measurement of properties of the aggregates and finished surface. It describes the performance of the finished product but does not indicate that the performance parameters have an influence on the long-term performance of the reseal. Pay adjust- ments are provided for single-coat seals, however, based on the expected life of the reseal from the size of the chips used and the texture depth of the finished surface. Weed, R. M. Multicharacteristic Performance-Related Specification for Hot-Mix Asphalt Pavement: Complete Development Process. In Trans- portation Research Record: Journal of the Transportation Research Board, No. 1861, Transportation Research Board of the National Acad- emies, Washington, D.C., 2003, pp. 53–59. This paper proposes a simplified procedure for developing performance- related specifications for HMA pavements, which directly considers the effects of as-constructed quality characteristics on expected pavement life-cycle costs in the selection of pay adjustment factors for these quality characteristics. The procedure considers in-place air voids, thickness, and initial smoothness of HMA pavement as the primary as-constructed quality characteristics that affect pavement performance and expected pavement life. The paper presents a generic exponential model for com- puting expected pavement life on the basis of acceptable and rejectable levels of each quality characteristic. A separate model can then be used to convert expected pavement life to a pay adjustment and pay schedule or incentive/disincentive for the different quality characteristics. This procedure provides a rational approach to relating as-constructed quality to the life-cycle cost of an HMA pavement and, therefore, the justification for various levels of construction pay adjustments. How- ever, when using this procedure to develop the pay adjustment schedule, the owner agency must have a good understanding of the effects that deficiencies in each of the as-constructed quality characteristics have on pavement life, and the costs associated with rehabilitation resulting from those deficiencies. Weed, R. M., and K. Tabrizi. Conceptual Framework for Pavement Smoothness Specification. TRB Paper 05-0922. Presented at the 84th Annual Meeting of the Transportation Research Board, Washington, D.C., Jan. 2005. This paper presents the framework for the New Jersey IRI Ride Quality Specification for HMA pavements (New Jersey DOT). The paper high- lights the idea that pavements built smoother initially will last longer than those built rougher. The paper discusses the concept of relating expected pavement life to the initial percent defective pavement in terms of initial ride quality. Once the expected pavement life (based on ride quality) is known, specific dollar values can be assigned to the as-constructed ride quality based on the anticipated cost of future rehabilitation and recon- struction. The paper draws values of improvement in expected pavement life from FHWA studies which correlated pavement life to initial smooth- ness. As described in the paper (and required by the New Jersey DOT specification), the contractor pays very high penalties for constructing very rough pavement (PD > 90). This reflects the anticipated future cost to rehabilitate or reconstruct a pavement that is not anticipated to last as long as a pavement initially constructed with greater smoothness. Weed, R. M. Mathematical Modeling Procedures for Performance- Related Specifications. In Transportation Research Record: Journal of the Transportation Research Board, No. 1946, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 63–70. This report is a continuation of the PRS modeling procedure described in the previous paper. It presents a more general model which allows greater flexibility in developing multicharacteristic relationships. The refined model also allows “high” and “low” failures for two-sided require- ments such as high and low limits for air voids in HMA pavements since conditions either too high or too low can negatively affect pavement performance. The model provides a rational approach for tying expected pavement life to pay adjustments for as-constructed quality. Design-Build Documents with Pavement Requirements Maine Department of Transportation. I-295 Commercial Street Con- nector Design-Build Project. RFP, MDOT Project No. 7589.30. Maine DOT, Augusta, 2003. This document is the request for proposal (RFP) for a design-build project for the I-295 Commercial Street Connector in Portland, Maine. Section 6.7.1.3 specifies that pavement design shall be in accordance with the 1993 AASHTO design guide and with 20-year design life. It identifies the criteria (initial serviceability, terminal serviceability, reli- ability, and standard deviation) for use with the design guide. Section 11 describes the warranty for the asphaltic pavement struc- ture. The design-builder is to establish the job mix formula and select materials to be used, and is to provide a quality control plan. The war- ranty stipulated for the project is either (a) 5 years from final acceptance with the design-builder given the option to propose additional 1-year increments for a period of up to 5 years following the expiration of the initial 5-year term; or (b) “to the end of the calendar year in which the cumulative ESALs for a particular segment or project reach or exceed the amount determined in Attachment 2 of the Special Provision.” Attachment 1 provides the pavement performance criteria for the project. Thresholds and remedial action requirements are provided for

94 the initial 5-year warranty period and any additional warranty period for the following distresses: smoothness (IRI), rutting, cracking, ravel- ing and popout areas, potholes, depressions and shoving, and roadway settlement (near to and away from abutments). Maryland State Highway Administration. Maryland Intercounty Connector (ICC) Project. Contract A, Contract No. AT3765960, Part 3.0 Design Requirements, Appendix A: PS 307, Pavement Per- formance Specification. Maryland SHA, Baltimore, 2006. This document is part of an RFP for Contract A of a major design-build project for the Intercounty Connector (ICC) project. Contract A covers the westernmost portion of the ICC which is approximately 7-miles long. This portion of the RFP covers the pavement performance specifi- cation for the project. The intent of the pavement performance specifica- tion is to leave much of the pavement design and construction to the design-builder, providing only the essential performance parameters desired from the Maryland State Highway Administration (SHA) for the finished pavement. The performance specification provides standards and references from Maryland SHA, AASHTO, ASTM, and Montgomery County but permits the design-builder to select which standards to use. The con- tract does list the appropriate standards in order of priority such that the higher priority standard should be used if different standards conflict. The design-builder is responsible for all pavement engineering includ- ing, but not limited to, pavement investigation, pavement type selection, new pavement design, pavement rehabilitation design, and material selection. The requirement for a thorough pavement investigation by the design-builder is critical and helps ensure that contingencies are covered and no unexpected issues are encountered during the design and con- struction process. Pavement type selection is left to the design-builder. The primary stipulations for pavement design are as follows: • An initial structural design service life of not less than 25 years must be provided; • A consistent pavement type must be used throughout each roadway element; • The 1993 AASHTO and Maryland SHA pavement design guides are required for pavement design; • Flexible/rigid pavement combinations and CRCP are not permitted; and • Flexible pavements must use Superpave mix design criteria. Design criteria (for use with pavement design guides), including traf- fic and other inputs, are provided along with additional criteria, such as minimum thickness and the use of dowels for PCC pavement, for both flexible and rigid pavements. Design criteria are also provided for the sections of the project subject to rehabilitation. The finished pavement performance parameters stipulated for final acceptance of the pavement include the following: • Structural capacity, evaluated by monitoring thickness, strength, and quality of materials throughout design and construction. • Ride quality, evaluated using inertial profiler throughout and at the completion of construction. • Skid resistance, evaluated using ASTM E274 and E501. Average fric- tion number of 45 must be provided along with justification that it will remain at least 45 for 5 years after construction. • Visual appearance, evaluated according to overall appearance and by visual distress surveys. While this document leaves many of the pavement design and con- struction decisions to the contractor, the specification includes a number of prescriptive elements. There are no indications that pay adjustments are made for the various performance criteria. This document is a good example of a move toward true performance specifications. Minnesota Department of Transportation. Trunk Highway 100–Duluth St. Design Build Project. RFP, Project S.P. 2735-172, Part I: Scope of Work. Minnesota DOT, St. Paul, 2001. This document is part of an RFP for a design-build project for improve- ments to Trunk Highway 100 at Duluth Street in Golden Valley, Min- nesota, reconstructing Trunk Highway 100 from a four-lane divided highway to a six-lane divided highway. Section 1.1.3.1 of the RFP stipu- lates, “All design and construction must be performed in accordance with the Minnesota Department of Transportation Standard Specifica- tions for Construction, 2000 Edition.” Section 15.4 provides the material and workmanship warranty speci- fications. A 5-year warranty (after final construction acceptance or FCA) is stipulated for both flexible pavements. Definitions, threshold limits, and remedial actions are provided for the following pavement distresses: • HMA pavement: transverse, longitudinal, block, and fatigue crack- ing; debonding; raveling; flushing; and rutting. • PCC pavement: transverse, longitudinal, corner, map, and shrink- age cracking; joint spalling and sealant failure; and surface defects ( shattered slab, nonfunctioning joints, popouts, scaling). Utah Department of Transportation. Utah DOT I-15 NOW Project. RFP, Part 4: Pavement Performance Specification. Utah DOT, Salt Lake City, 2005. This document is part of the RFP for the Utah I-15 corridor reconstruc- tion design-build project (I-15 NOW). The document provides the requirements for pavement performance, both flexible and rigid. The contractor is given responsibility for pavement design and control of construction operations. Guidance documents are referenced in the RFP and given priority ratings in case of conflict in requirements between documents. The design requirements specified are a minimum 20-year design life for HMA pavement and 40-year design life for rigid pavement. Design ESALs for both flexible and rigid pavement design are provided in the RFP for various sections of the project. Performance requirements for the finished pavement include func- tional requirement for ride quality (profilograph measurement) and skid resistance (evaluated by the Utah DOT) thresholds both during construction and at the end of the project (final owner acceptance, or FOA). Various other requirements are also provided for drainage, mini- mum thickness, and selected material requirements for bases, paving materials, and asphalt grade. Utah Department of Transportation. Utah DOT I-15 NOW Project. RFP, Part 9: Warranty Provisions. Utah DOT, Salt Lake City, 2005. This document is part of the RFP for the Utah I-15 corridor reconstruc- tion design-build project (I-15 NOW). The document provides the warranty provisions for the project, including pavement warranties. For most elements, a 2-year warranty (after FOA) is specified. A 5-year war- ranty is specified for pavement settlement as well as rigid pavement cracking and joint deficiencies. Thresholds are provided for maximum permissible settlements for pavements, both flexible and rigid. For PCC pavements, threshold dis- tress levels, extent, and corrective actions (during the warranty period) are provided for the following distresses: cracking, joint deficiencies,

95 surface defects, and miscellaneous distresses. For flexible pavements, threshold distress levels, extent, and corrective actions (during the war- ranty period) are provided for the following distress types: cracking, patching and potholes, surface deformation, surface defects, and mis- cellaneous distresses. Washington State Department of Transportation, Guidebook for Design-Build Highway Project Development. Washington State DOT, Olympia, 2001. This document provides a framework for developing contract docu- ments for design-build projects in Washington State. The last half of the document provides a template for scope of work for design-build proj- ects. Some of the sections related to pavements in the scope of work include the following: • Section 416, Pavement Design, provides design criteria (e.g., design life) for pavement design, with specific criteria for asphalt concrete pavement (ACP) and Portland cement concrete pavement (PCCP) and for PCC pavement rehabilitation. • Section 1300, Product Warranty Provisions, specifies the perfor- mance parameters to be evaluated for all constructed pavements dur- ing the warranty period: ride quality, pavement friction, pavement surface condition, structural capacity, and material quality. • Section 1330, Asphalt Concrete Pavement, provides the require- ments for the elements of ride quality, pavement friction, pavement surface condition, structural capacity, and material quality for war- ranted ACP. • Section 1340, Portland Cement Concrete, provides the requirements for the elements of ride quality, pavement friction, pavement surface condition, structural capacity, and material quality for warranted PCCP. • Section 1350, Required Corrective Actions, provides tables which describe distress types, allowable levels of severity, allowable extent of severity, and corrective action for asphalt pavements, new concrete pavements, and dowel bar retrofits. Criteria specified for each pave- ment type include the following: 44 ACP: rutting and wear, alligator cracking, longitudinal cracking, transverse cracking; 44 PCCP: cracking, joint cracking/spalling, pumping and blowing, faulting, patching, scaling, wear, and joint seal damage; and 44 DBR: cracking within slot, wear within slot, bond failure within slot, faulting, spalling within slot. Performance-Based Maintenance Contracts with Pavement Requirements Engelke, T. Long Term Performance Based Road Maintenance Con- tracts in Western Australia. Proc., Bay Roads Exposed Conference, Rotorua, New Zealand, April 2003. This paper discusses Main Road’s Western Australia experience with performance-based road maintenance through six different contracts. This form of maintenance contracting provides performance levels for pavements, including intervention parameters and key performance indicators (KPI). Intervention parameters are quantified according to maximum intervention level or the severity of a defect, maximum response time, and maximum defective condition. Threshold levels are provided for each for different forms of maintenance. Intervention parameters for pavement surfaces include roughness, skid resistance, isolated pavement failure, and edge breaks. KPIs measure the contractor’s performance, and payment to the con- tractor is adjusted on the basis of KPIs. KPIs for maintenance performance for pavements maintenance (which falls under the category of asset man- agement) include pavement strength (measured using FWD), roughness, rutting, texture, skid resistance, and maximum defective condition. Robinson, M., E. Raynault, W. Frazer, M. Lakew, S. Rennie, and E. Sheldahl. DC Streets Performance-Based Asset Preservation Exper- iment: Current Quantitative Results and Suggestions for Future Contracts. Paper No. 06-2075. Presented at 85th Annual Meeting of the Transportation Research Board, Washington, D.C., 2006. This paper describes an experimental performance-based asset preser- vation project implemented in Washington, D.C., from 2000 to 2005. The goal was to reduce overall maintenance and rehabilitation costs for 75 miles of the National Highway System in the District of Columbia by encouraging innovative, cost-effective, and flexible preservation strate- gies. This type of contract is different from traditional maintenance con- tracts in that the owner specifies performance standards for the roadway, permitting the contractor to choose what materials and methods to use. In this case, 170 performance measures were included, covering all aspects of the roadways, including pavements. Each performance mea- sure had five levels of service; the minimum acceptable performance was “Good” or “Level 4.” For pavement surfaces, ride quality (IRI), pave- ment condition index (PCI), and friction number were the primary performance measures specified. Work Zone Management Curtis, D., and K. A. Funderburg. States Estimate Work Zone Traffic Delay Using QuickZone. ITE Journal, Vol. 73, June 2003, pp. 40–43. This article describes QuickZone, a software tool for estimating work zone delays and maximum queue lengths. The article touts the custom- izable nature of the product, citing specific states that have modified QuickZone to their needs. The implication is that transportation agen- cies could use QuickZone (or a similar product) to produce reasonable targets and tolerances for use in performance specifications on specific projects. Maryland Department of Transportation. Haul Routes and Access During Construction. Performance Specification 306: Maintenance of Traffic. Maryland DOT, Hanover, July 2006. This bid document outlines temporary traffic control restrictions for a highway project. It is largely prescriptive but includes a temporary lane and shoulder closure schedule and contractor disincentives related to that schedule. The document is useful in that it makes reference to the QuickZone program, requiring the contractor to use computer models to determine queuing impacts before construction. Also, the document includes a thorough temporary lane and shoulder closure schedule, clearly communicating acceptable days and times for closures. This gives the contractor significant flexibility with its maintenance-of-traffic plan, and penalties for deviating from the schedule are clearly defined. Michigan Department of Transportation. Contractor Proposal Using Best Value Practices. Special Provision for Highways for Life. Michigan DOT, Lansing, Aug. 2007. This bid document uses a true performance contracting approach with clear methods of measurement and contractor incentives/disincentives. Traffic measurements include an open-to-traffic date, number of work zone crashes, and motorist delay. While not truly a specification docu- ment, this is an excellent example of a structured performance approach to work zone traffic control. Clear parameters are defined for traffic- based criteria. For example, motorist delay includes explicit information dictating how the contractor will be paid based on the amount of delay measured during four random weekly on-site measurements.

96 Oregon Department of Transportation. I-5 Weaver Bundle 306. Temporary Traffic Control Performance Specification. Oregon DOT, Salem, Nov. 2006. As of September 2007, this document is the most up-to-date temporary traffic control performance specification for the Oregon DOT. It uses a specified measured-volume to projected-volume ratio target to assess contractor performance. Although it includes some prescriptive ele- ments, the overriding direction of this specification is performance- based. The document is specific regarding the design parameters to which the contractor is restricted but is not specific regarding contrac- tor penalties for deviating from those parameters. Pennsylvania Department of Transportation. Publication 352. Con- tract Management Division, Bureau of Design, Pennsylvania DOT, Harrisburg, June 2001, pp. 3.28–3.31. This is the Pennsylvania DOT’s sample specification (adapted from FHWA) for lane rental on both a daily and an hourly basis. Both versions include the clause, Failure to Complete Work on Time, which assesses liquidated damages. Since the specification is intended as an example, the values and prices included are for sample purposes only. The docu- ment concisely illustrates a straightforward lane rental agreement which is one type of work zone, traffic control performance specification. Pennsylvania Department of Transportation. Section XXX— Maintenance and Protection of Traffic During Construction. Penn- sylvania DOT, Harrisburg, Sept. 2006. This draft document is a straightforward, sample, lump sum, work zone traffic control specification. Some prescriptive elements are included, but the specification leaves much to the “sound engineering judgment” of the contractor. Duration of traffic stoppages and 85th percentile speed measurements serve as performance measures. This document is useful as an example of the means and metrics necessary to implement lump sum traffic control as a performance specification. Trauner Consulting Services. Work Product Stemming from FHWA Workshop Held in Seattle, Washington, (March 6–7, 2003). This document is a general work zone traffic control performance speci- fication, designed to be customized by the specifying agency. A large variety of possible performance measures are presented, although quan- titative thresholds are not included because of the intentional ambiguity of the specification. This document provides a valuable performance- based framework for a traffic control specification, despite the lack of sample values for restrictions and incentives. Utah Department of Transportation. I-15 NOW Project. RFP Part 4: Maintenance of Traffic Performance Specification. Utah DOT, Salt Lake City, Dec. 2005. This document is the maintenance of traffic performance specification for a major highway design-build project. Although the document is titled a performance specification and although it contains several per- formance elements, the document is wholly prescriptive in nature. This serves as an example of what several DOT’s are loosely terming perfor- mance specifications for work zone traffic control. Vassallo, J. M. Implementation of Quality Criteria in Tendering and Regulating Infrastructure Management Contracts. Journal of Con- struction Engineering and Management, Vol. 133, No. 8, Aug. 2007, pp. 553–561. This article describes contemporary contracting procedures designed to combine price and quality standards, specifically in the field of infra- structure management contracts. The author describes the increasingly common concept of lane rental, including a list of states in which the concept is fully operational. Washington State Department of Transportation. Lane Rental. Washington State DOT, Salem, Wash. http://www.wsdot.wa.gov/ Projects/delivery/alternative/LaneRental.htm. Accessed Nov. 9, 2007. This document, posted on the Washington State DOT website, includes an overview of lane rental, decision criteria for incorporating lane rental into a particular bid document, and associated sample special provi- sions. This brief document is most useful for the decision criteria sug- gested for determining appropriate uses for lane rental. Washington State Department of Transportation. Lump Sum Traffic Control. Washington State DOT, Salem. http://www.wsdot.wa.gov/ Projects/delivery/alternative/LumpSum.htm. Accessed Nov. 2007. This document, posted on the Washington State DOT website, includes an overview of lump sum traffic control, decision criteria for incorpo- rating lump sum traffic control into a particular bid document, sug- gested prebid procedures for use with lump sum traffic control, and associated sample special provisions. This brief document is most use- ful for the decision criteria suggested for determining appropriate uses for lump sum traffic control. public involvement Oregon Department of Transportation. I-5 Weaver Bundle 306. Pub- lic Information and Involvement Performance Specification. Oregon DOT, Salem, Nov. 2006. As of September 2007, this document is the most up-to-date public involvement performance specification for the Oregon DOT. While the document is titled a performance specification, it is wholly prescriptive in nature. The contractor is given latitude in the means of achieving the stated goal of “fully informed and meaningful participation by Stakeholders and the public for the duration of the Project,” but no performance targets or parameters are established. This document is useful in that it concisely articulates the goal of most public involve- ment processes. Also, the document is a good source of potential steps to be taken in a public involvement process even though the steps take the form of a prescriptive specification. Utah Department of Transportation. I-15 NOW Project. RFP, Part 4: Public Information and Performance Specification. Utah DOT, Salt Lake City, Dec. 2005. This document is the public involvement performance specification for a major highway design-build project. The document contains few per- formance elements despite its title. The document is indicative of the misconceptions that exist in several agencies regarding the use of the term performance specification, especially as it applies to public involve- ment. While the document does not prescribe extremely detailed steps for performing work, it is far too specific in its language to be termed a performance specification. Total Quality index Griffith, A. F., E. G. Gibson, Jr., M. R. Hamilton, A. L. Tortora, and C. T. Wilson. Project Success Index for Capital Facility Construction Projects. Journal of Performance of Constructed Facilities, Vol. 13, No. 1, Feb. 1999, pp. 39–45. This article documents research done in an effort to create a success index for facility projects. The study uses mail surveys and phone inter- views to collect historical data on facilities projects and bases the derived success index on four variables: budget achievement, schedule achieve- ment, design capacity, and plant utilization. The resulting index gives a snapshot of the cumulative success of the project, but the subjectivity of the inputs makes this a poor tool for use in evaluating contractors.

97 Lee, D.-E., and D. Arditi. Total Quality Performance of Design/Build Firms Using Quality Function Deployment. Journal of Construction Engineering and Management, Vol. 132, No. 1, Jan. 2006, pp. 49–57. This article describes an index that owners may use to rank design- build firms relative to their total quality performance; it also describes the development of that index. The article is useful in that it presents a logical matrix approach to correlating the owner/agency requirements with the technical characteristics of the contractor: quality function deployment. The model developed in the article has value as a product quality performance measurement tool, both for predicting future quality and assessing past value received. Michigan Department of Transportation. Contractor Proposal Using Best Value Practices. Special Provision for Highways for Life. Michigan DOT, Lansing, Aug. 2007. This bid document uses a true performance contracting approach with clear methods of measurement and contractor incentives/disincentives. A best value practice is used for contractor selection. Even though the indexing takes place before the contractor’s work, the best value practice illustrated in this document is useful in researching the concept of a total quality index. The contractor is scored on several aspects of its proposal. A higher score produces a multiplier that factors the contractor’s bid price downward (valuing dollars spent on a high-quality contractor higher than dollars spent on a low-quality contractor). The open-to-traffic date proposed by the contractor is one of the inputs that determine the mul- tiplier. Of note is that the contractor may be assessed liquidated damages based on the open-to-traffic date that it provides. North Carolina Department of Transportation. Design Build Package. Project No. 8.1674402. North Carolina DOT, Raleigh, Oct. 2001. This bid document describes the scoring system for evaluating technical proposals submitted by contractors for a design-build highway project. The document is useful as it provides an example of an adjusted bid price selection process in use at a state DOT. The process allows for a price adjustment of up to 15% based on the contractor’s quality score. Pongpeng, J., and J. Liston. Contractor Ability Criteria: A View from the Thai Construction Industry. Construction Management and Economics, Vol. 21, No. 3, Jan. 2003, pp. 267–282. The study described in this article aimed to develop a common set of contractor ability criteria for both government and the private sector. The study concludes that the ability criteria may be classified similarly to the hierarchal organization of a construction firm. Percentages of influence to be used to evaluate a contractor based on the project requirements— including time, cost, quality, and safety—are assigned to the various organizational units of a typical construction firm. The article is use- ful in that it analytically determines an indexing formula for the eval- uation of potential contractors. Each organizational unit (engineering/ construction, project managers, human resources, public relations, etc.) of a typical contracting firm has been assigned a weight of relative importance in determining a final total score. While not a true total quality index (especially in a project sense), the methods and findings of this study may be projected for use in determining a total quality index performance specification. Vassallo, J. M. Implementation of Quality Criteria in Tendering and Regulating Infrastructure Management Contracts. Journal of Con- struction Engineering and Management, Vol. 133, No. 8, Aug. 2007, pp. 553–561. This document describes contemporary contracting procedures designed to combine price and quality standards, specifically in the field of infra- structure management contracts. The article applies fundamental eco- nomic thinking to the problem of developing quality criteria and provides a sound framework for conceptualizing the quantification of quality. But the author does not develop numerical values for finalized equations. Risk Armistead, A. Performance Specifications and Contracts: The State of the Art. Proc., 21st Australian Road Research Board Conference, Cairns, Australia, 2003. Based on the work done by Opus International Consultants Limited for Austroads (association of Australian and New Zealand state highway authorities), this report narrows the focus of performance specifications and contracts to three contractual arrangements: design-construct- maintain contracts that typically extend to 5 years or 10 years; build- own-operate-transfer contracts that typically extend to periods over 20 years; and maintenance contracts. This report excluded performance- based specifications, performance-related specifications, and end prod- uct specifications because they rely on predictive methods to evaluate performance. The main conclusion is this: “The consultant has reviewed, to the extent possible, experience with performance contracts and specifications. . . . Due to the time frame of these contracts, and the fact that most of them are in relatively early stages, it is not possible to absolutely quantify the achievement of outcomes. There are still many issues surrounding these contract types that remain unanswered, and as many of these contracts move towards their conclusion, there will be an opportunity for further analysis and development of improved practices.” The benefits, risks, and constraints identified in this report are as follows: • Potential benefits: 44 Better customer focus; 44 Improved risk recognition, allocation, and management; 44 Reduced administration; 44 Budgetary certainty; 44 Greater ability to innovate; and 44 Improved use of industry skills. • Risks: 44 Performance measures that do not fully reflect the performance required; 44 Scale and tendering costs that will reduce competition; 44 Potentially greater effects of contract failure when compared with traditional contracts; 44 Reduced ability to deal with physical, political, or environmental issues; 44 Reduced road authority technical expertise leading to loss of “informed purchase” status; 44 Loss of road authority’s control, leading to reduction in standard of service; and 44 Restricted availability of innovation to the wider market. • Constraints: 44 Difficulty in defining the performance required; 44 Limitations with repeatability and reproducibility of condition data; 44 Inability to measure remaining life of pavements; 44 Prohibitive cost of tendering for contractors and the road authorities; 44 Absence of accepted methods of assessing ongoing benefits; 44 Time and resource capability limitations within the industry; 44 Absence of complete, accurate, and up-to-date asset data in some road authorities;

98 44 Politically unacceptable social consequences of large region-wide contracts; 44 Perceived threat of job losses in the road authority; and 44 Greater potential financial implications of contract failure. Shuler, S., T. Aschenbrener, and R. DeDios. Effect of Performance Warranties on Cost and Quality of Asphalt Pavements. In Trans- portation Research Record: Journal of the Transportation Research Board, No. 2040, Transportation Research Board of the National Academies, Washington, D.C., 2007, pp. 100–106. This paper presents the lessons from using warranties over a 3-year period in which the contractor was responsible for several forms of pavement distress. The contractor was allowed to choose materials and methods for constructing the pavements. Based on the cost-benefit relationship for the projects during the warranty period and beyond and comparison with the comparable nonwarranty pavements, the authors found no significant difference in competition or performance of the warranty projects compared with the control projects; the differ- ence in cost of the warranty projects compared with the nonwarranty projects was negligible. Cost comparisons included the initial hot-mix asphalt, maintenance, pavement evaluation team, weigh-in-motion sta- tion, and construction engineering. The warranty analyzed in this study was developed so the contractor had control of materials selection and construction methods. This meant that mixture design was under control of the contractor and not required to be approved by the Colorado Department of Transporta- tion (CDOT). The contractor was responsible for correcting defects in work elements within the contractor’s control, including distresses resulting from defective materials and workmanship, during the war- ranty period. CDOT remained responsible for the pavement design and those distresses that resulted from the design. Six warranted projects were evaluated in this analysis as follows: • I-25, Fountain (constructed, 1998); • C-470, Santa Fe to Wadsworth Boulevard (constructed, 1998); • US-36, Superior Interchange (constructed, 1998); • I-25, Pueblo (constructed, 2000); • I-70, Eagle to Avon (constructed, 2000); and • US-50, Kannah Creek (constructed, 2001). Control, or nonwarranty, projects were compared with the warranty projects to develop economic and performance comparisons. The con- trol projects were constructed using traditional CDOT specifications (nonwarranty) and were comparable to the warranty projects in terms of year of construction, rehabilitation strategy, traffic, environment, and original pavement condition. The performance of the experimental and control projects was measured by two methods to minimize errors: a pavement evaluation team made up of private and CDOT personnel, and the pavement management automated data collection van. The key recommendations follow: • Monitor competition on future warranty projects to determine if a reduction in the level of competition occurs as a consequence of using warranties; • Consider the role of agency specifications for QC/QA because con- tractors involved in this study had worked with QC/QA specifica- tions implemented by CDOT 10 years earlier; • Ensure that the distress thresholds for identifying warrantable and nonwarrantable distresses are carefully evaluated; and • Evaluate the requirement of a weigh-in-motion scale in conjunction with projected traffic flow increases over the warranty period. Cost-benefit evaluations should be based on the design life of the pavement and not just the warranty period. Blankenship, P., and D. R. Leach. Performance-Related Specifica- tions for Pavement Preservation Techniques. National Pavement Preservation Forum II: Investing in the Future (CD-ROM), Publica- tion No. FHWA-IF-03-019, Federal Highway Administration, Washington, D.C., November 2001. The authors gave this presentation at a conference in 2001. The authors’ premise is that prescriptive specifications do not always give the expected performance. The authors show a spectrum of different specifications that include material property/prescriptive specifications, performance-related specifications, performance-based specifications, performance-modeled specifications, and performance specifications. Without illustrating who is at risk, the authors claim that prescriptive specifications are high risk, whereas performance specifications, at the other end of the spectrum, are low risk. The presentation includes some slides about microsurfacing, macro- surfacing, and several types of pavement quality tests. The slides in them- selves are inadequate to discern what the authors might have explained in their talk, but one of the conclusions—that “performance related specifi- cations and innovative products lower overall costs”—is not supported anywhere in the presentation. Buttlar, W. G., L. Shanley, and S. Aref. End-Result Specification Devel- opment: The Illinois Demonstration Projects. Illinois Cooperative Highway Research Program, Project ICHRP, IHR-R23 (1999). In May of 1999 a committee consisting of industry, the Illinois Depart- ment of Transportation, and the University of Illinois at Urbana- Champaign was formed. The main charge of the committee was to develop a prototype end-result specification for asphalt pavement construction and to develop a series of demonstration projects to collect data for evaluation of the new specification in the year 2000. The key items addressed by the task force included the following: • Selecting the number and type of quality characteristics to be used as pay items; • Determining the required sample size to minimize pay factor uncer- tainty (risk); • Setting specification limits and pay factor equations; and • Evaluating the feasibility of using contractor test results in computation of pay factors, to minimize agency test burden to the extent possible. From the work of the task force, this paper specifically discusses the following: • Evaluation of a stratified-random technique for field density mea- surement in the new end-result specification; • Development and evaluation of a single lot system for establishing pay factors; and • Development of QA comparison limits in the new end-result specification. The key findings presented in the paper are as follows: • Random field sampling techniques, involving “dual-stratified” ran- dom sampling in both the longitudinal and transverse directions, appeared to lead to increased contractor awareness and response to densification levels throughout the pavement. • The use of contractor data for development of project pay factors appeared to be appropriate.

99 • Statistical analysis showed very little evidence of contractor bias in test results. (Only in the case of plant voids was a statistically different result obtained.) Evaluation of demonstration project pay factors showed a 0.7% increase in pay factor when comparing contractor and agency results. • It is feasible and in fact desirable to analyze each job as a single lot of material, which has the effect of a very large sample size and reduced payment risks. Simulation analyses were conducted to verify that this method would give appropriate pay factors even for bimodal data distributions. • A statistically based, tiered approach for developing quality assurance comparison limits was presented. The method minimizes agency test burden, while providing additional measurements when marginal comparisons are detected. A key benefit of this paper is the reduction in agency test burden. Although the analysis and the findings are credible, the repeatability on other projects is suspect. Gallagher, P. J., and D. Mangan. Risk Issues in Performance-Specified Flexible Paving Contracts. Proc., 19th ARRB Transport Research Conference, Sydney, Australia, Dec. 1998. The synopsis of this paper starts with this paragraph: “It is widely accepted that true performance specifications are a ‘pipe dream’ while there are no adequate performance models of pavement behavior. In the United States, the SHRP Superpave performance prediction model was found to have ‘significant technical problems’ which have prompted a review that will not be completed until the year 2005.” The background section concludes with the following: “There is little doubt that the new contractual forms (Design-Construct-Maintain, Build-Own-Operate-Transfer) have been driven by management and economic reform, rather than by technology. However, tenderers for con- tracts such as Design-Construct-Maintain need to be aware of the exist- ing boundaries of knowledge and technology if they are to exercise due diligence in relation to the risks which they are assuming.” More innovation, which is a premise behind the use of performance specifications, is weakened somewhat by an observation in this paper: “The final outcomes of the tenderer assessment process are regarded as commercially confidential, and unsuccessful proponents remain uncer- tain of the standards that they ought to pursue in further proposals.” An appendix to the paper lists the risks to be addressed in contractor’s detailed proposal (other than normal contractual risk of construction quality, climate, productivity etc.): • Adequacy of data acquisition on subgrade and traffic and of available climatic records; • Pavement design to be “fit for purpose” (i.e., to meet proposed level of service at all times); • Whole-of-life analysis for full design period (20, 30, or 40 years?); • Residual life analysis at end of warranty period; • Quality system for assurance of product quality and conformity with design intent, including independent verification as stipulated by the client; • Refinement of the client’s proposed criteria for pavement perfor- mance, if necessary; • Methods and frequency of measurement for each criterion; • Process/program for checking compliance for each criterion at prac- tical completion and during the warranty period; • Maintenance plan, if required, to detail 44 The intervention level in relation to the level of service for each performance parameter; 44 Inspection frequencies and procedures; 44 Maintenance procedures and work plans; and 44 Response plans to manage deterioration and emergency incidents. • Any particular local factors relating to raw material supply/quality, environmental, or other issues. A second appendix lists the specific risks affecting pavement perfor- mance but outside control of paving contractor: • Traffic: 44 Future volume and classification (Warranty term may be based on monitored ESALs or specified term as agreed.); 44 Allowable load limits (e.g., national proposal for increase); and 44 Individual overloads or damaging loads (e.g., army tanks). • Subgrade or subbase by others: 44 Uniformity in level and quality/uniformity of subgrade support; 44 Latent conditions for reasonable departures from client’s information; 44 Susceptibility to saturation/moisture damage; and 44 Pavement/subsoil drainage maintenance. • Network operations: 44 Nonfeasance versus malfeasance 4▪ Does private sector inherit indemnity for nonfeasance available to the agency? 4▪ What is the impact of the contract’s incident response plan on liability? 44 Accident damage (physical damage, spillages, fires); 44 Natural disasters (floods, bushfire, earthquake), severity; and 44 Utility services in the pavement 4▪ Leaks/defects; and 4▪ Quality of repairs/restoration. General issues for negotiation and inclusion in contract include • Independent audit of quality compliance during construction and of performance compliance during warranty period; • Agreed process for determining causes of failure and identifying con- tributing factors; • An alternative dispute resolution process in the event of differences arising on the nature and cause of failure; and • Construction program constraints on site access and on required completion date which may necessitate construction in adverse climate. The paper concludes with the following: “The future of [design- construct-maintain] and similar performance specified contracts seems assured by the current perception that they offer real economic benefits to tax payers and road users. If this perception is to survive and if this contract strategy is to be applied to a wider range of, and to smaller, projects, clients and industry must cooperate to collect more credible pavement performance data and to develop realistic perfor- mance criteria. Then, all parties can be confident that performance specified contracts will give us the better pavements that we are all looking for.” Gruneberg, S., W. Hughes, and D. Ancell. Risk Under Performance- Based Contracting in the UK Construction Sector. Construction Management and Economics, Vol. 25, No. 7, 2007, pp. 691–699. According to the authors, “the essence of a performance-based approach is that the focus is on what a building does, rather than its inputs.” The authors also state that “little empirical research work has been carried out on the management of risk under [performance-based contracting]

100 PBC” and “only those producers who are confident in calculating the subjective risks and reward structures would be willing to accept a PBC project.” Interestingly, J. T. van der Zwan (2003) argues that, because the results of empirical models apply only within the limits of the empiri- cism, producers cannot do much other than “control the recipe.” On the basis of a telephone survey of 22 construction companies in the UK, the authors of this paper identified 27 risks associated with PBC. The top 10 are Risk Frequency Fitness for purpose (FFP) Lack of insurance for FFP Lifetime costs (including responsibility for lifetime maintenance) Price (capital cost) Inadequate client specification Difficulty of changing mindset of clients, consultants, designers, and contractors Process of setting key indicators Measurements of key indicators over time “Wouldn’t touch it!” The FFP changing after completion 7 6 4 3 3 2 2 2 2 2 The authors conclude that although PBC gives contractors the free- dom, responsibility, and authority to perform their work as they see fit, they also face many risks. Longer-term risk obligations will require con- tractors to have new organizational structures. The difficulties of insur- ing and dealing with FFP are major obstacles to the adoption of PBC by contractors. Contractors prefer suppliers to assume responsibility for their products under PBC. In the end the question becomes one of bal- ancing the increased risks against the possibility of increased rewards. Given the nature of the construction market, it may not always be pos- sible for contractors to pass on expected increased costs to their clients. Kuzyk, P., R. C. G. Haas, and R. W. Cockfield. Performance-Based Specifications for Pavements, Canadian Journal of Civil Engineering, Vol. 18, No. 6, 1991, pp. 1054–1061. This article provides an interesting snapshot of the status of perfor- mance specifications as of 1991. The authors view a pavement manage- ment system (PMS) as being central to any efforts to improve pavement performance. Specifications and contract documents are one part of a comprehensive PMS. The authors illustrate the idea and the associated risk factors through an excellent graphic (Figure 1 in the article) and discuss the problems associated with the idea of transferring risks from the agency to the contractor. The authors’ study used an Ontario Pavement Analysis of Costs (OPAC) model that had the ability to separate pavement deterioration due to environmental exposure from that due to traffic loads. This dis- tinction is critical when using performance specifications because both the specifications and the contractor’s design have to be based on assumed levels of traffic supplied by the agency. Over the course of the contract period the actual traffic levels may have to be taken into consideration. To facilitate the implementation of performance specifications, the authors acknowledge the need to modify the system of contracting as follows: • Use sliding scale bonds. The sliding scale bond should include a por- tion based on the payments to date, putting the contractor’s prior profits at risk in case of default. • Develop a measure for traffic interruptions resulting from mainte- nance during peak and nonpeak hours over the performance period, with a predefined level beyond which penalties can be levied against the contractor. The contractor will have to submit a schedule ahead of time for performing maintenance. In conclusion the authors cite the reliability of models for estimating pavement performance as a major risk factor. Manik, A., and W. G. Buttlar. Monte Carlo Based Simulation for Managing Risk in End-Result Construction Specifications. (Paper supplied by Prof. Buttlar; publication details not known.) This paper begins with a sweeping claim: “Over the years many high- way agencies in North America have made a valued commitment to End Result Specifications (ERS). As a direct result, it is believed that the quality of our roadways has improved” (Smith, 1998; Benson, 1999). The authors also point out that the highway community contin- ues to struggle with the problems of comparing the owner’s QC tests to the contractor’s. The quality characteristics (defined as that charac- teristic of a unit or product that is actually measured to determine conformance with a given requirement) that are being used to deter- mine the quality of the pavement are generally in situ density of the constructed pavement, voids and asphalt content of the plant mix, aggregate gradation, and so on. These quality characteristics are believed to be related to performance, but the exact relationships are not yet firmly established. The risk addressed in this study is the payment risk, expressed as pay- ment made to the contractor (e.g., baseline, or “correct,” pay). Overpay- ment is referred to as agency risk, while underpayment is often termed contractor risk. A key contribution of this study is a strategy for using fewer samples/ cores than the current Illinois DOT end-result specifications, from a constructed pavement. The data from the samples feed into a simula- tion model that estimates the quality of the population from which the samples are taken. A simulated risk analysis (SRA) model developed by the authors is presented. The SRA computes the agency risk and the contractor risk as a function of many factors, including sample size, production and mea- surement variability, bias, pay formula and pay caps, and specification limits; and it considers the quality assurance and third-party testing schemes used. The SRA replaces all earlier models such as ILLISIM. The key contributors to payment risk in SRA are as follows: • Contractor data versus agency data; • Frequency of testing and/or number of samples; • Variability and/or bias of test device and/or test procedure; • Specification parameters, including 44 Specification limits (percent within limits); 44 Pay factor equation; 44 Pay caps; 44 Acceptance test frequency and acceptance tolerance; and 44 Third-party testing provisions. The authors conclude with the claim that SRA can be used to develop a better understanding of how changes in individual ERS specification parameters can affect the payment risk for the contractor and agency. This knowledge can be used to explore the possibility of developing desirable changes in an existing ERS—such as reducing sample size, reducing risk, optimizing tolerance limits, changing pay factor equations—and the pros and cons of pay factor equations with pay- ment caps.

101 (Smith, G. 1998. NCHRP Synthesis of Highway Practice 263, TRB, National Research Council, Washington, D.C.; Benson, P. 1999. Per- formance Review of a Quality Control/Quality Assurance Specifica- tion for Asphalt Concrete. Transportation Research Record: Journal of the Transportation Research Board, No. 1654, Transportation Research Board of the National Academies, Washington, D.C.) Piewerbesky, B., D. Alabaster, and J. Fulton. New Zealand’s Performance-Based Pavement Design and Construction Specifica- tions: Case Studies. Proc., 21st Australian Road Research Board Conference, Cairns, Australia, 2003. This paper agrees with J. T. van der Zwan’s argument (2003) that new materials are one key driver behind the move toward performance spec- ifications. That was Transit New Zealand’s rationale when the agency did not have the resources to develop new pavement and material speci- fications for every possible alternative material. New Zealand’s B/3 performance-based specification for structural design and construction of flexible unbound pavements, including chip seal surfacing, was introduced in 2000 to foster the use of marginal and nonconforming materials that give similar performance to standard basecourse and subbase materials. This paper provides an overview of B/3 and its accompanying document for materials (M/22: Notes for the Evaluation of Base and Subbase Aggregates) as they are and proposed improvements. Three pilot projects using B/3 were completed. Details of two of the projects are presented as case studies, and the relevant outcomes of each are discussed. Both projects involved the same con- tractor. Under the B/3 contract, the contractor was responsible for the design, construction, and maintenance of the pavement and seal. The contractor had to demonstrate that design, materials, and construction techniques were appropriate (through its quality assurance systems) and that the pavement performance at the end of the defects liability period (1 year to 3 years) was acceptable. The following measures were checked to ensure that the performance criteria were met: • Surface shape and rut depth; • Roughness; • Texture depth; • Skid resistance; • Surfacing aggregate retention; and • Surface waterproofness. The paper presents the following conclusions: • Contractors undertaking work involving performance-based specifi- cations require highly skilled and experienced pavement designers (either in-house or out-sourced); the road authority must also possess or have access to experienced, knowledgeable pavement engineers to adequately assess submitted proposals. • Statistical analysis techniques are a valuable tool for both the con- tractor and road authority in quantifying the condition of the pave- ment and verifying compliance with acceptance criteria, and should be an integral facet of performance-based contracts. The contractor in these case studies completed an extensive suite of tests on the subgrade and pavement layers during and after construction, and analyzed the data to determine statistically valid testing regimens for future projects. • Road authorities and industry must collaborate, including sharing knowledge and expertise, for performance-based specifications to be successfully introduced and implemented. These parties must also work together to ensure that pavement research is relevant to their needs; better, more accurate, and robust (but not complex) tech- niques must be developed for predicting pavement performance, so that the risks for the contractor and the road authority can be more readily quantified. The last conclusion begs the question: should performance specifica- tions be used in conjunction with reimbursable (not unit price) con- tracts for construction, similar to many private-sector owners, separating the maintenance work during the warranty period? Queener, J., T. Hill, and M. Horn. A Contractor’s Experience with the Caltrans Maintenance Warranty Pilot Program. National Pavement Preservation Forum II: Investing in the Future (CD-ROM), Publica- tion No. FHWA-IF-03-019, Federal Highway Administration, Wash- ington, D.C., November 2001. This paper includes a glossary of terms related to pavement failure and maintenance procedures. The paper itself describes a case study for an I-5 asphalt pavement overlay project near Fresno using Type O-HB asphalt on approximately 240 lane kilometers. The specification allowed the contractor to identify areas of the project that had specific defects before the placement, and those areas would be exempt from the war- ranty requirements. Granite Construction Company was awarded the contract, with a 1-year warranty. According to the contractor, what was unique about this project was the lack of history for this material in thin-lift overlays in Califor- nia. The specification required that the contractor provide warranty for rutting, raveling, flushing, and cracking within the limits specified. The specification excluded certain areas that were determined to have existing defects that could affect the performance of the asphalt- rubber overlay. The paper concludes that this was a positive experience overall for Caltrans. But the short duration of the warranty and the prescriptive exclusions make it a case study with limited relevance to other scenarios. Russell, J. et al. Asphalt Pavement Warranties—Technology and Practice in Europe. FHWA-PL-04-002. American Trade Initiatives, Alexandria, Va.; Office of International Programs, Federal Highway Administration, Washington, D.C. 2003. Warranty contract is defined as a type of performance-based contract that guarantees the integrity of a product and assigns responsibility for the repair or replacement of defects to the contractor. This report provides an excellent overview of warranty or similar contracting in Europe (Denmark, Germany, Sweden, Spain, and the United Kingdom) and gets into some detail of the pavement perfor- mance criteria used in Europe. Interestingly, the report notes that the use of pavement performance contracts in these countries at the time was quite limited, typically only a handful of contracts with the most in Sweden (about 10%). However, one of the conclusions is that “PPCs [pavement performance contracts] in particular may hold great benefit for counties and municipalities throughout the United States, and could gain acceptance relatively quickly.” The chapter on warranty eval- uation concludes, “Transparent warranty evaluation processes are a key to any warranty program’s success.” But the ARRB paper (Gallagher and Mangan 1998) rightly points out that pavement evaluation processes offered by successful proposers are likely to be treated as commercially confidential. Page 53 of this report states that in Europe contractor responsibility for pavement maintenance is a part of all warranty contracts if pave- ment performance criteria are not achieved or maintained. It also says, “The relationships and cooperation between owner agencies and

102 warranty contractors is significantly different than in the United States.” However, the report does not make clear what those differences are and how they manifest themselves in written contracts. Subramanian, R., and F. T. Najafi. Current Status of the Develop- ment and Use of Portland Cement Concrete Performance Related Specifications. Presented at the Annual Conference of the Canadian Society of Civil Engineering, Montreal, June 2002. This paper provides a statistical approach for contractors to maximize profit while improving quality in contracts involving performance specifications for PCC pavements. The project involved in this study was the Florida Department of Transportation (FDOT) I-295 project in which FDOT decided to base the specifications on concrete strength, slab thickness, and initial smoothness. The test proposed, and the tar- get mean and standard deviation for each of the three AQCs were established in the specifications. Other AQCs (air content and percent consolidation around dowels) were not included in the performance specifications. The paper also discusses the role of the seller’s risk and the buyer’s risk. The seller’s risk is the probability of rejecting or assigning a pay- ment reduction (disincentive) to a pavement lot that has a true mean and standard deviation that meets the targets. The buyer’s risk is the probability of accepting or assigning a payment increase (incentive) to a pavement lot that has a true mean and standard deviation that do not meet the targets. On the basis of tests involving five different concrete mixes ranging from 3500 psi to 5500 psi and five different slab thicknesses, the authors demonstrate how a contractor’s choice of the mix or slab thickness could depend on whether it is a risk-taking or a risk-averse contractor. The paper includes a graphic that ties life-cycle costs for projects using performance specifications to contractor pay adjustments. In conclusion, the authors note that recent FHWA studies validated the distress indicator prediction models for transverse joint faulting, trans- verse joint spalling, transverse slab cracking, and smoothness, and updated PaveSpec from version 2.0 to version 3.0. However, they sug- gest that additional AQCs be included in the modes, such asphalt air content and consolidation around dowel bars. van der Zwan, J. T. Functional Specifications for Road Pavements: A Question of Risk Assignment. Proc., 22nd PIARC World Road Congress, Durban, South Africa, 2004. This paper stands out because of its effort to understand and explain the theory behind performance specifications. It presents a “pyramid of demands,” with road user demands on top and demands on raw materials at the bottom; this is followed by a discussion of how speci- fications relate to the different levels on the pyramid. The pyramid serves as a theoretical basis for discussing the risks associated with per- formance specifications. The demands can be from national priorities, financing mechanisms, maintenance criteria, economic consider- ations, climatic circumstances, quality of natural raw materials, and safety considerations. The author identifies two main drivers of the move toward performance specifications: changing materials and new materials with which owners have little experience; and the continuing momentum (right or wrong) to transfer tasks traditionally done by owners, to contractors. Perhaps most important, the author questions the use of empirical models to justify the use of performance specifications because the results of empirical models apply only within the limits of the empiri- cism. As for the potential for innovation, this author believes that inno- vative products and techniques are out of the bounds of empirical knowledge; therefore, innovation as a premise for using performance specifications is risky, and the allocation of that risk is easier said than done. With a specific example, the author discusses how a contractor can- not make adjustments to production on the basis of functional prop- erties because that requires knowledge of how different parameters affect functional properties, for which adequate models do not exist. Also, for reasons already mentioned, empirical modeling does not help. Therefore, the producer cannot do anything other than “control the recipe.” The notion of the pyramid in this paper is important because it puts the issues in perspective. As the author notes, the changing methods of construction are not prompting the move to use performance specifica- tions; rather, the desire to transfer responsibilities from owner to con- tractor and change the contractual relationships provides the impetus. The author ends with an acknowledgement that owners and contrac- tors will be forced to rethink the traditional approach, thus creating knowledge that will improve quality and perhaps lead to innovation. However, he adds a cautionary note that greater risks for the contractor will result in higher prices, and that we don’t know what proportion of contractor costs at present reflects their risks. Villiers, C., Y. Mehta, G. Lopp, M. Tia, and R. Roque. Evaluation of Percent-Within-Limits Construction Specification Parameters. International Journal of Pavement Engineering, Vol. 4, No. 4, Dec. 2003, pp. 221–228. This paper presents the findings from a study that evaluated 10 Super- pave coarse mixtures used in the construction of four interstate high- ways in Florida. Two parameters were used to evaluate the mixtures: asphalt content and percent passing through a 2.36-mm sieve. The sensitivity of sampling frequency to the acceptable quality level (AQL) and rejectable quality level (RQL) is analyzed, to help in the develop- ment of more realistic percent-within-limits construction specifica- tions. At the time of this study, the state of Florida used an AQL of 90 and RQL of 50. Operation characteristic curves (probability of acceptance as a func- tion of percent-within-limits) are created for various sampling frequen- cies. A conclusion was that 10 samples were required to attain the AASHTO recommendation of seller’s risk of 1% and buyer’s risk of 5%. The seller’s risk is the probability that good quality will be rejected as unacceptable on the basis of test results. The minimum level of actual quality for the construction to be acceptable is AQL. The buyer’s risk is the probability that what is accepted is of unacceptable quality based on test results. The maximum level of actual quality for the construction to be unacceptable is RQL. The applicability of this study is contingent on development of oper- ation characteristic curves based on historical data for the critical pay parameters. Zhang, Z., and I. Damnjanovic. Quantification of Risk Cost Associ- ated with Short-Term Warranty-Based Specifications for Pavements. In Transportation Research Record: Journal of the Transportation Research Board, No. 1946, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 3–11. This paper begins with an acknowledgement of the concerns that state highway agencies have when deciding on the use of warranty specifications. Foremost among the concerns is the inability to quan- tify the risk cost or the warranty servicing cost. Other concerns include possible increase in the number of litigations and the poten- tial elimination of small contractors from the bidding process because of surety requirements.

103 The scope of the analysis in this study is limited to short-term perfor- mance warranties in which a contractor performs only preventive main- tenance, not rehabilitation, for the most vulnerable—burn-in—phase of pavement life cycle. The authors’ basis for the modeling presented in this paper is that, from the state agency’s point of view, the objective is to find an upper bound on the risk costs that contractors would be allowed to include in a bid in addition to their nonwarranty bid. In other words, the agency will be willing to pay the risk costs reflected in the increased bid price as long as those costs are less than the savings from using the short-term warranty. From the contractor’s point of view the objec- tive is to design or build pavements for the minimum total cost, including maintenance cost during the warranty period. Through an example the paper supports the argument that with the application of performance warranties, contractors will be motivated to implement strict quality control measures during construction. The modeling in this study is limited to “first-failure” after construc- tion and is therefore based on reliability performance models. The sequence of failures that occur after the first-failure are also dependent on rehabilitation, are more appropriate for consideration in long-term warranties, and are therefore outside the scope of this study. The paper comprises a notable effort to statistically model warranty risk costs for short-term warranties with a specific, empirical example. However, the extent to which the results can be applied to other sce- narios remains questionable. In conclusion the authors also state the need to address issues related to differences in the effects of different pavement maintenance actions and the need for a warranty policy that addresses hidden defects early in the pavement life cycle.