Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback


<< Previous Contents Next >>

Development and Implementation of a Performance-Related Specification for I-65 Tennessee: Final Report

Chapter 5. Evaluation of the Performance-Related Specification

Both a quantitative assessment (from results of the AQCs and pay factors for each lot) and qualitative assessment (from results from surveys of the contractor, TDOT staff, and QC representative) are provided in this chapter. Expected pay charts for the PRS are provided in appendix C.

Quantitative Assessment

Quantitative assessments of the results achieved through use of the PRS specification can be made by comparing the final PRS pay factors and payments to those of the standard TDOT specification. Table 13 summarizes the PRS and standard TDOT quality requirements for strength, thickness, and smoothness. By design, they are similar. However, the pay factor curves for PRS are based on the expected change in LCC associated with actual variation in performance from the as-design target properties. The TDOT pay factors are based on the judgment of engineers regarding the incentives or disincentives for the contractor.

The differences in the PRS and TDOT pay factor curves for strength, thickness, and smoothness are shown in figures 9 through 11. TDOT specifications provide disincentives for below-standard quality levels. The PRS specification includes both incentives and disincentives, based on the expected level of the as-constructed quality values. In both specifications, concrete that does not develop compressive strength of 3,000 lbf/in2 (20.68 MPa) in 28 days must either be removed and replaced or accepted at reduced pay. The TDOT standard pay factors for thickness decline significantly more than the PRS pay factors for thicknesses between 12.0 and 12.8 in. (305 and 324 mm). For thinner pavement designs (e.g., 9 to 11 in. [229 to 279 mm]), these curves might be more similar, as thickness greatly affects performance. However, as previously discussed, because of the very conservative thickness design used (as determined by AASHTO at a high level of reliability), the PRS pay factors indicate that the pavement LCC is reduced only by about 10 percent when the thickness is reduced to 12.0 in. (300 mm). The target Rainhart PI (0.1-in. [2.5 mm] blanking band) that was used for PRS specification development was 7.0 in./mi (111 mm/km). Target values used by TDOT during the project included 10 in./mi (158 mm/km), 3.5 in./mi (55 mm/km), and 7.0 in./mi (111 mm/km).

Table 13. Quality Requirements for the Performance-Related Specification (PRS) and Tennessee Department of Transportation (TDOT) Standard Method Specifications
Factor Detail PRS TDOT
Strength Test methods AASHTO T23, T22 AASHTO T23, T22
  Lot AQC mean (std. deviation), lbf/in2 4,500 (500) 4,000
  Lot RQL, lbf/in2 * 3,000 3,000
  Lot MQL, lbf/in2 ** 5,500 -
Thickness Test methods AASHTO T 148 AASHTO T 148
  Lot AQC mean (std. deviation), in. 13.0 (0.5) 13.0
  Lot RQL, in. 12.0 12.0
  Lot MQL, in. 14.0 13.25
Smoothness Test methods Rainhart 0.1-in. blanking band Rainhart 0.1-in. blanking band
  AQC mean (std. deviation), in./mi 7.0 (1.0) 10.0
  Lot RQL, in./mi 9.0 10.0
  Lot MQL, in./mi 0.0 0.0
AQC = acceptable quality characteristic; 1 lbf/in2 = 6.89 kPa; 1 in. = 25.4 mm; 1 in./mi = 16 mm/km.
* RQL (rejectable quality limit)—Agency-chosen minimum limit for acceptable AQC specimen sample quality.
** MQL (maximum quality limit)—Agency-chosen maximum limit for acceptable AQC specimen sample quality.

Figure 9. Comparison of Performance-Related Specification and Tennessee Department of Transportation (TDOT) strength specifications.

Click on the image link for a description of the image.

Figure 10. Comparison of Performance-Related Specification and Tennessee Department of Transportation (TDOT) thickness specifications.

Click on the image link for a description of the image.

Figure 11. Comparison of Performance-Related Specification and Tennessee Department of Transportation (TDOT) smoothness specifications.

Click on the image link for a description of the image.

PRS pay factors for the as-constructed northbound and southbound lane lots indicate that the pavement in both directions was constructed to a quality above the design level. Lot quality levels and pay factors for strength, thickness, and smoothness in the northbound lane are shown in table 14. Table 15 includes the quality levels and pay factors for the southbound lanes.

Table 14. PRS Lot Quality and Pay Factors (PF) for the Northbound Lanes
Item Target Northbound Lot Number
1 2 3 4 5 6
Sample Units   3 4 4 4 3 5
Thick (mean), in 13.0 13.4 13.3 13.6 13.3 13.1 13.1
Thick (st. dev) 0.50 0.17 0.05 0.00 0.09 0.59 0.06
Thick PF (%) 100.00 101.63 101.44 101.96 101.37 100.31 100.89
Strength (mean), lbf/in2 4500 5177 5366 5321 4885 5292 5485
Strength (st. dev), lbf/in2 500 300 205 74 153 274 240
Strength PF (%) 100.00 103.25 104.1 104.2 102.07 103.76 104.57
Profile (mean), in/mi 7.0 5.5 3.5 5.0 8.9 5.8 5.2
Profile (st. dev), in./mi 1.0 0.48 1.14 0.94 1.09 0.53 0.34
Profile PF (%) 100.00 101.71 103.87 102.38 97.62 101.46 102.04
Lot PF (%) 100.00 106.73 109.7 108.58 101.00 105.60 107.65
Table 15. Performance-Related Specification Lot Quality and Pay Factors (PF)
for the Southbound Lanes
Item Target Southbound Lot Number
7 8 9 10 11 12 13 14
Sample Units   3 5 5 4 4 4 4 3
Thick (mean), in. 13.0 13.0 12.9 13.1 13.0 13.1 12.9 12.9 12.9
Thick (st. dev) 0.50 0.065 0.20 0.048 0.063 0.063 0.311 0.140 0.363
Thick PF (%) 100.00 100.52 99.87 100.70 100.18 100.58 99.50 99.60 99.39
Strength (mean), lbf/in2 4500 4676.7 4411.0 4766.5 5101.9 4761.3 4573.1 4631.3 5100.0
Strength (st. dev), lbf/in2 500 531.81 83.53 166.02 297.00 72.30 415.64 89.36 267.3
Strength PF (%) 100.00 100.79 99.89 101.52 102.93 101.57 100.41 100.95 102.94
Profile (mean), in./mi 7.0 3.7 4.9 4.5 2.1 3.2 3.5 3.1 3.8
Profile (st. dev), in./mi 1.0 0.34 0.63 0.65 0.96 1.07 0.36 0.25 0.17
Profile PF (%) 100.00 103.75 102.38 102.81 105.19 104.13 103.92 104.36 103.59
Lot PF (%) 100.00 105.11 102.14 105.11 108.46 106.39 103.83 104.93 105.98

A closer look at the values and pay factors provides additional insight. There appears to have been a dialing down by the contractor of the strength and thickness mean values as the project progressed from lot 1 to lot 14. Figures 12 and 13 both indicate this trend of the contractor beginning the project with conservative properties. Pay factors for both of these items also decreased as the project progressed.

Another interesting aspect is that the contractor reported that grade control for the northbound lots was accomplished using the adjacent pavement. In the southbound lanes, stringlines and better subbase grade control were used. The results are evident in figure 14, which shows reduction in both the average PI values and a trend toward reduced variability within the lots in the southbound lanes.

Figure 12. Comparison of Performance-Related Specification
strength specifications and results.

Click on the image link for a description of the image.

Figure 13. Comparison of Performance-Related Specification
thickness specifications and results.

Click on the image link for a description of the image.

Figure 14. Comparison of PRS smoothness specifications and results.

Click on the image link for a description of the image.

Figure 15 shows a summary of the PRS pay factors for each of the 14 lots used in the analysis. It also includes an overall pay factor, which averages 106.5 percent for the northbound lots and 105.2 percent for the southbound lots. High strength and thickness levels are the primary cause of increased pay factors in the northbound lanes. Although these values were reduced in the southbound lanes, the smoothness values increased, offsetting most of the pay factor reduction from strength and thickness.

Comparison with Existing TDOT Pay Factor Curves

The PRS pay factor curves provide for incentives and disincentives for strength, thickness, and profile. The PRS curves are based on economic analysis of LCC, indicating that there will be changes in pavement performance depending on the level of quality achieved during construction of these three acceptance quality characteristics. It is believed that the PRS pay factor curves will provide the contractor with more opportunity to achieve incentive pay and to avoid disincentives, thereby providing a pavement with a longer life and lower LCC.

A comparison of TDOT with PRS pay factors using data from each lot separately shows that the TDOT specification would provide 100 percent of the bid price. The PRS would provide for a positive incentive on the order of 106 percent.

Figure 15. summary of performance-related specification pay factor results.

Figure 15. Graph. summary of performance-related specification pay factor results. Pay factors (percent) are plotted for smoothness, strength, thickness, and lot total for the six northbound I-65 lots (1–6) and eight southbound lots (7–14). Average total pay factor northbound is 106.5 percent; southbound is 105.2 percent; all lots surpassed 100 percent, with lot 4 having the lowest total score at 101 percent. Four lots earned less than 100 percent: lot 4 on smoothness and lots 12, 13, and 14 on thickness.

Impact of Quality on Pavement Life

The result of the PRS was that the contractor would receive an average of 106 percent incentive pay for higher quality construction of all 14 lots. As was shown, this was due to PCC strength, PCC slab thickness, and initial PI being of general higher quality than the specified target values. The mean values were as shown in table 16 for comparison. The standard deviations of most of the AQCs were also of higher quality than the target values.

Table 16. Target and As-built Acceptance Quality Character Values
Acceptance Quality Characteristic Target As-Built
PCC compressive strength, lbf/in2 4,500 4,967
PCC slab thickness, in. 13.0 13.1
Profile Index (Rainhart 0.1-in. blanking band), in./mi 7.0 4.5
1 lbf/in2 = 6.89 kPa; 1 in. = 25.4 mm; 1/in./mi = 16 mm/km

In assessing benefits of the PRS, the following could be asked: will a 6 percent increase in construction cost due to higher quality result in a similar or greater increase in pavement life? This question was addressed using an independent method to predict pavement life for both the target and the as-built lot mean AQCs. The NCHRP 1-37A mechanistic-empirical pavement design and analysis software was used to predict the performance of the target (or as-designed) and the as-built JPCP.(10) The distress and smoothness models were nationally calibrated under NCHRP 1-37A and should be reasonably applicable to the Nashville, Tennessee, area. All inputs were estimated and the program run to provide an estimate of the target and as-built pavements expected life. The program predicts three main performance characteristics: slab cracking, joint faulting, and IRI and terminal levels of each were selected at which rehabilitation would be needed.

Results showed that the expected life of the target pavement turned out to be in excess of 50 years, which was due to the conservatism in the design. The expected life of the as-constructed lots was even longer, approximately 14 percent longer. Over the long predicted pavement life the IRI was the controlling factor (very low amounts of cracking and joint faulting were predicted) but eventually the pavements roughness increased. Therefore, for an increase in initial cost of 6 percent (from the positive quality incentives), an even greater percentage (over twice) increase in pavement life was achieved.

Qualitative Assessment

A meeting was held at the end of construction to obtain responses by the contractor, the QC consultant, and TDOT staff regarding the PRS implementation project. During the meeting, the results from the project were presented and explained, and questions were addressed. Then survey forms were provided to the contractors, the construction QC representatives, and TDOT personnel who participated in the PRS implementation. Included in the survey were questions assessing the functionality of the PRS, any related problems encountered in the process, and changes that were made in response to the PRS. Results of general questions are summarized in table 17, which indicates that the PRS documents were adequate, the PRS concept is desirable, and PRS implementation was not difficult. Additional detailed questions were asked of the contractors, QC managers, and TDOT personnel. Their responses are provided in the following sections.

Table 17. General Survey Responses
No. Question Contractors Quality Control Representatives TDOT
Yes Maybe No Yes Maybe No Yes Maybe No


Do you think the responsibilities and roles of the contractors and TDOT are well defined in the PRS document?








Do you think PRS (including the incentives) would improve the quality of concrete pavements in Tennessee?






Do you think that the PRS testing and sampling plan can lead to more accurate measurement of the quality of TDOT PCC pavements?






Did you think that the PRS process was complicated and difficult?






Would you like to see PRS implemented on more Tennessee PCC pavement projects?






PRS = performance-related specification; TDOT = Tennessee Department of Transportation

Contractor Assessment

Surveys were completed with representatives of the prime contractor (LoJac, Inc.) and the paving contractor (APAC, Inc.). Their responses are shown in the following tables.

  1. What average cumulative pay factor did you expect to receive for the PRS sections prior to construction?

    Pay Factor, %

    Reason for this estimate

    108 to 112

    Money saved for future construction, jobs are different.

    105 to 108

    Reasonable pay for additional control.


  2. Was the pay factor you received worth the effort you spent achieving it?




    Comments and suggestions

    Could include more pay factors like permeable base.

    Incentive promotes quality from the contractor.
  3. What problems did you see or encounter in preparing for and constructing the I-65 PRS sections?
    Problem encountered in: Description and suggestions
    Discussing the PRS specification with TDOT OK.
    Understanding the PRS specification OK.
    Adjusting processes to meet the PRS specification OK.
    Preparing subgrade and base Involved extra work to control grades.
    Setting grade stakes and string lines Need strandline for permeable base.
    Placing and finishing the concrete surface OK.
    Sampling and testing for strength, thickness, and smoothness Need faster turnaround on results.
    Understanding the PRS pay factors OK.
    Resolving any conflicts related to PRS Everyone willing to discuss.
    Other related activities None.
  4. What changes did you make or in the design and construction process to avoid penalties or receive bonuses under the PRS?
    Activities affected: Description of any changes
    Mix design  
    Subgrade and base preparation (1) Yes. (2) Better control.
    Grade stakes and stringlines (1) Yes. (2) Used more often.
    PCC batch mixing  
    PCC hauling to paver  
    PCC transfer to paver  
    Paving machine type and setup  
    PCC placement methods  
    Pavement surface finishing  
    Pavement curing  
    Surface grinding  
  5. What changes might you make in the design and construction process under similar PRS projects?
Possible changes: Description of any changes
1. Mix design Add over design factor in initial mix design.
2. Subgrade and base (1) Monitor and maintain a smooth, consistent subgrade.(2) Better control of base.
3. Grade stakes and stringlines Use wire for stringline. Base and paving contractor use the same stringline.
4. PCC batch mixing Monitor mix for consistency.
4. PCC placement methods Use a spreader in the paving train.

Other comments that were received included the following:

  • PRS “rewards contractor for exceeding product requested.” “Incentive promotes quality control.”
  • More accurate quality measurements can be achieved because PRS “relates actual product back to anticipated [design] product.”
  • PRS “promotes quality end product. Promotes payment for actual product received.”
  • The pay factor would have been worth the effort spent achieving it. “Incentive promotes quality from contractors.”
  • PRS provides “better pay for better work.”
  • “Need faster answers on test results.”
  • “Need more tests per sublot.”
  • “Incentives good—procurement process.”
  • PRS “eliminates low quality contractors.”
  • “Immediate feedback – yes!”
  • Contractor started strength high in the first 2 weeks of northbound paving and adjusted down to optimal strength in the southbound paving.
  • Contractors need 5 to 10 percent incentive to provide enough incentive for the necessary changes in construction processes.
  • If the incentive is greater than 10 percent, the target should be reset.
  • A smoothness specification for the permeable base would be helpful.
  • “There was more grade variability in the northbound sections. This affects smoothness and thickness.”
  • Northbound paving matched the existing PCC grade line. Southbound paving was tied to a stringline.

Construction QC Contractor Assessment

Surveys were received from the QC contractor, Florence & Hutcheson, Inc. of Nashville, Tennessee. Their comments are included in the following tables.

  1. What problems did you see or encounter in developing or implementing the I-65 performance related specification?
    Problem encountered in: Descriptions and suggestions
    Collecting data for PRS input (1) None.
    Selecting pay factor limits (1) Some judgment and subjectivity involved.
    Introducing PRS to contractors (1) I think they are receptive.
    Completing the PRS sampling (1) Based on square yards rather than lineal feet. Smoothness is fixed.
    Completing the PRS testing (1) None.
    Determining the PRS pay factor values (1) None.
    Informing contractors of bonus or penalty values (1) None, they know what is required up front.
    Resolving conflicts over payments (1) None, they know what is required up front.
    Other PRS activities (1) (2) Subgrade and permeable base grades.
  2. What other possible problems do you foresee in future performance related specification use?
    Potential problems Descriptions and suggestions
    1. Grades TDOT needs to tighten specifications for subgrade and permeable base (elevations).

These engineers also provided several additional comments:

  • No real problems were encountered in CE1 services.
  • PRS “would most likely reduce variability, thus increasing quality.”
  • PRS can provide more accurate quality measurements because “with reduced variability, actual test results are more realistic of actual pavement.”
  • “From testing and inspection viewpoint, don’t think any more complicated than current specifications.”
  • “Testing frequency is affected too much by pavement width; 24 ft versus 12 ft doubles testing.”
  • “Incentive versus disincentive should result in a better product.”
  • “Contractors can ‘go for’ an incentive rather [than] just focusing on not getting hit with a penalty, resulting in a higher quality of pavement.”
  • “The same data is obtained in the previous specification. More compilation of the data is required, but once the TDOT office is familiar with the software, it should not be complicated or lengthy.”

TDOT Assessment

TDOT engineers who had participated in the design, implementation, and management of the PRS project responded to the survey and followup interviews with generally positive responses. Following are their responses:

  1. What problems did you see or encounter in developing or implementing the I-65 performance-related specification (PRS)?
    Problem encountered in: Descriptions and suggestions
    Collecting data for PRS input (1) Smoothness data (for instance) was difficult to obtain in short sections. Road profiler was used.(3) It would have been very beneficial to use the impact echo device (for slab thickness).
    Selecting pay factor limits (1) Wanted to be fair with contractor and State so pay limits should reward appropriately.(2) How low do you go, or how high?
    Introducing PRS to contractors (1) The contractors seemed interested and helpful because of the potential plus incentives.(2) I think it is important to show that incentives can be earned by using ‘everyday’ practices.
    Completing the PRS sampling (1) This was a little difficult because it required additional samples to be taken. (2) I think it was good to define the sublot stationing up front, so that sampling could be easily tracked.(3) It’s hard to tell the contractor that we want to take 3 cores in their new pavement every 1,500 ft (e.g., 1 per 500 ft).
    Completing the PRS testing (2) Somewhat problematic with relating smoothness lengths to cubic yards.
    Determining the PRS pay factor values (2) Spreadsheet very good (for summarizing data and computing pay factors).
    Informing contractors of bonus or penalty values  
    Resolving conflicts over payments  
    Other PRS activities (2) Increased sampling and testing leads to increase in field personnel responsibilities. This could be an issue in an understaffed office.
  2. What other possible problems do you foresee in future performance related specification use?
Potential problems Descriptions and suggestions
1. Payment deduction issues If a penalty is levied by the data, concrete contractor will possibly look to place blame on permeable base and/or grading contractors.
2. Sublot/lot size If two lanes paving 500-ft sublot—good. If single lane, then two times as much testing.

Additional comments provided by TDOT engineers included the following:

  • “I think it [PRS] leads to elimination of less quality-oriented contractors.”
  • “The number of tests could be questionable in being accurate.”
  • “Compared to the discussion of all other specifications we deal with, the process was more straightforward and not as much time spent in discussion.”
  • “I like the direction it [PRS] takes us.”
  • “I see the contractor giving us a more concentrated effort to increase the quality of the product he produces.”
  • PRS “allows greater pay for better materials and quality of construction.”
  • “Since being involved with the process, it was not complicated. However, those who were not involved during the process may have seen it as complicated.”
  • “This trial shows that using everyday practices can lead to incentives, provided a quality, consistent approach is used.”
  • “Ultimately it [quality] is up to the contractor and how well they build the road. I think it [PRS] gives the contractor a reason to work harder and do better.”

Suggestions to Improve PRS for Concrete Pavement

Several suggestions for improvement of the specification and methodology can be gleaned from these comments:

  • Consider aiming for 10 percent maximum incentive. If this is exceeded, consider changing the specified requirements (e.g., modifying the AQC target values).
  • Consider tightening subgrade and subbase grade requirements, encouraging contractors to better control these elevations, or adding incentives.
  • Provide a mechanism for contractor to have PRS pay factor results quickly. More rapid testing would be one solution, such as a reduction in coring by use of alternative method to determine slab thickness (possibly the Wisconsin method of probing the plastic concrete).
  • Reconsider the required increased testing when paving width is one 12-ft lane rather than the normal two or three.
  • Adjust smoothness-sampling lengths or modify smoothness data analysis method to easily report PI for short lengths.
  • Consider methods for increasing the sampling rate and reducing the amount of destructive testing (this comment likely refers to coring to determine slab thickness).
<< Previous ContentsNext >>
Updated: 02/20/2015

United States Department of Transportation - Federal Highway Administration