Volume 1: Practical Guide, Final Report and Appendix A

Under this research project, the prototype PRS (developed under previous FHWA research) was revised and expanded to make it more practical.^(1,2,3)   Different implementation levels of the PRS were defined (Levels 1, 2, and 3), with the simplified PRS (Level 1) designed to be immediately implementable.  This report (four volumes) was offered as a practical guide to using the revised prototype PRS (included as appendix A in this volume) for the acceptance of JPCP lots.  It discussed all aspects of the research, including the step-by-step procedures for developing and using the PRS, results of demonstrations of the PRS methodology (see volume II), explanations of supplementary laboratory and field studies (see volume III), and the user guide for the revised PaveSpec 2.0 PRS demonstration software (see volume IV).   Based on the research conducted under this project, the following conclusions and recommendations have been compiled.

 Conclusions

• The PRS that have been developed for JPCP highways are based on sound economic, engineering, and statistical concepts.  The PRS have been field tested in several projects as shadow specifications, and reasonable results were obtained from these projects, as documented in appendix B (volume II) of this report.  The PRS are now ready for the next step of implementation in an actual field situation where contractor pay is under the PRS.

• Implementation of the PRS for JPCP will require: (1) the tailoring of the PRS to the project and obtaining of many inputs and decision criteria and (2) adequate training of both SHA and contractor staff in PRS concepts and applications.

• The PRS for JPCP are specifications on key material and construction AQC’s that have been shown to correlate significantly with the performance of the pavement.   Through the use of PRS, a methodology is provided by which the quality of pavement construction can be related to the performance of the pavement and then to LCC’s.   Thus, rational contractor pay adjustments can be made for each constructed lot.

• PRS have the capability to identify an optimum level of construction quality that would minimize overall LCC’s, yet still meet performance requirements.  Selection of target quality levels of AQC’s (means and standard deviations) and other factors in the PRS are required by SHA’s.

• The PRS are driven by key distress indicators that control performance and, hence, the LCC of the pavement.  For JPCP, these include transverse cracking of the slab, joint faulting, joint spalling of the slab, and pavement smoothness.

• Based on the key JPCP distress indicators, the following AQC’s were selected for use in the PRS: concrete strength, air content, consolidation around the dowel bars, slab thickness, and initial smoothness.  The PRS can include any or all of these AQC’s.

• Both the mean value and the within-lot variability of each of the AQC’s are considered in computing the contractor pay factor.

• Three different specification levels (simply titled Level 1, Level 2, and Level 3) were developed to aid in the implementation process.  Level 1 is fully implementable and utilizes current SHA test procedures for AQC’s and simplified contractor pay factor computation procedures.  Pay factors computed using Level 1 are designed to closely estimate those obtained from the more accurate simulation conducted under Level 2.  Level 3 is a futuristic concept where all testing is nondestructive and rapid.

• Future LCC’s of the pavement are predicted, discounted, and summed as a present worth. LCC’s include both M & R costs to the agency, as well as certain highway user costs based on pavement roughness.  If these as-constructed pavement lot results indicate an increase or decrease in performance over that of the target as-designed pavement lot, appropriate pay adjustments are made.

• The simulated field trials made it clear that practical limitations must be placed on contractor pay factors.  These limits are placed both on pay factors for individual AQC’s and on the overall lot pay factor.

• A comprehensive laboratory/field investigation of concrete material- and construction-related variables was conducted and results were documented. This included the following:

1. Concrete strength testing (relationships between flexural and compressive strength and other tests, maturity methods, and the prediction of 28-day strengths from early age strengths at 3 to 5 days).

2. An investigation of testing and construction variability.

3. Correlation of concrete consolidation at dowels and measured load transfer.

• The Level 1 PRS were demonstrated by developing pay factor charts representing typical JPCP designs in a chosen SHA (Iowa).  The three typical designs were chosen based on medium, heavy, and very heavy traffic levels.  Three general trends observed from the exercise included:

1. Pay factors increased as the quality of the measured AQC mean improved.

2. At a given AQC mean, pay factors increased as the measured AQC standard deviation decreased.

3. Pay factor curves generally became flatter as traffic level increased.

A complete summary of these results is included in appendix B (volume II).

• The PaveSpec 2.0 PRS demonstration software is an integral part of the revised PRS approach, and is used in combination with the developed practical guide (volume I) to develop PRS.  The PaveSpec 2.0 software provides user-friendly capabilities for simulating a lot.  These include the simulation of the construction of a lot, sampling and testing, performance prediction, LCC summation, and computation of pay factors.  The PaveSpec 2.0 software is an invaluable tool for demonstrating and clarifying the revised PRS concepts.

 Recommendations

• The PRS are now ready for field implementation where contractor pay depends on the PRS.   Successful implementation by a SHA will require considerable preparation of the PRS, along with significant training of the agency and contractor personnel.

• The PRS are expected to lead to higher quality construction, improved performance, and lower LCC’s.  The main reason for this is the inclusion of several key AQC’s (concrete strength, slab thickness, entrained air content, initial smoothness, and percent consolidation around dowel bars) with incentive/ disincentive pay factors.   The rational computation of these contractor pay factors for each AQC and the overall constructed lot should give agencies support for implementing a PRS.

• Each SHA wishing to implement PRS for JPCP must tailor the specifications and pay computation procedure to local project conditions.  This includes the following: model calibration; AQC’s; target means and variability; pay factor limits; AQC limits; sample size; selection of key distress indicators; selection of tests for each AQC; field management of sampling and testing, and computation of pay factors; M & R plan; and procedures to choose trigger values for distresses.

• The PRS can be utilized to optimize design by helping to select AQC values that lower LCC’s while still meeting performance requirements.  Very little work has been done in this regard to date, and much more is needed to fully demonstrate the value of PRS.

• Many improvements to the existing PRS for JPCP are still needed.  The implementation of PRS will be a major effort, and limitations will certainly develop.   These must be overcome with further research and development.  Following are some specific areas of further development and improvements that are needed for PRS for concrete pavements:

1. Improved user cost models are clearly needed in PRS.  The existing user cost models need updating and do not include traffic delay costs from lane closures for M & R.

2. Improved distress and roughness prediction models are needed.  Those in PaveSpec 2.0 are the best available at this time, but they certainly could be improved upon given the extent of LTPP data now available.

3. Additional AQC’s could be included to make PRS more comprehensive.   These could include joint sawing depth, surface texture, concrete mixture components (cement, aggregates, etc.), and improved ways to measure early strength.   In addition, inclusion of base course quality, subgrade quality, and shoulders would be valuable to PRS.

4. Additional pavement types: continuously reinforced concrete pavement, jointed reinforced concrete pavement, and unbonded concrete overlays.

5. Development of more rapid, nondestructive testing for concrete strength measurement.

• Additional work is needed in developing and improving the PRS for Levels 2 and 3.   Level 2 requires the running of simulations to compute a lot pay factor.   While this provides a correct pay factor, it is a "black box" to the agency and contractor.  Level 3 needs more rapid nondestructive tests on the in situ pavement lot.
• Use of pavement management system (PMS) data support PRS implementation.  Such aspects as calibration of distress models, validation of distress models, and validation and improvement of M & R life-cycle costing are important.
• After the conduct of initial implementation projects, there will be a need to better investigate the benefits of PRS as compared to conventional QA/QC specifications.
• Finally, considerable training in PRS concepts and practice will be needed for SHA, contractor, and industry personnel.