Reporting Live

By: Sameh Zaghloul, Ph.D., P.E., P.Eng., and T. Joseph Holland, Ph.D., P.E.

Concern for the environment and diminishing supplies of aggregates and petroleum products have made asphalt recycling an attractive option for highway agencies. In addition to conserving material resources and using less energy, asphalt recycling can provide economic benefits in the form of lower pavement life-cycle costs and increased business opportunities.

Although recycled asphalt pavements (RAP) have been in service for many years, well-documented field performance studies are rare. Most previous studies have instead approached the subject from a laboratory and materials point of view. Although laboratory testing is an important tool in the prediction of field performance, highway agencies and the public are, in reality, more concerned with actual in situ performance. Detailed field performance data also is essential for accurate life-cycle cost analysis, without which the true cost effectiveness of asphalt recycling cannot be assessed.

As part of a study being performed for the California Department of Transportation (Caltrans), Stantec Consulting has investigated RAP performance with a firm focus on in situ conditions. The investigation considered, firstly, how RAP performs across different environmental zones and, secondly, how it compares with other materials subjected to similar environmental and traffic conditions.

Field testing, including deflection, roughness and distress surveys, was conducted at numerous test sections located in three of California’s environmental zones—North Coast, Desert and Mountain. Laboratory testing also was performed on cores extracted from the sections.

15% of 60

Caltrans allows up to 15% RAP to be used as a substitute for virgin aggregate mix in hot asphalt concrete mix. Sixty RAP sections were evaluated in total, covering a wide range of layer thicknesses: total asphalt concrete (AC) thickness from 5.76 in. to 10.8 in.; total aggregate thickness from 4 in. to 15.6 in.; and total pavement thickness above the subgrade from 13 in. to 24 in.

Stantec employed a comprehensive evaluation program that covered all aspects of pavement performance. This program included four performance indexes. Structural performance was evaluated using the structural adequacy index (SAI), a function of the layer moduli backcalculated from deflection measurements. Functional performance was evaluated using the roughness index (RI), a function of the international roughness index. Distresses were evaluated using the distress index (DI), a function of observed distress types, severities and extents. In addition, the degree of uniformity achieved during construction, a common concern in the use of RAP, was evaluated using the construction consistency index (CCI). CCI is a function of in situ thickness, consistency of asphalt content and aggregate gradation in the AC mix and consistency of in situ structural capacity.

SAI, DI and RI are dynamic indexes that reflect a pavement’s deterioration over time. All employ a 0.0 to 1.0 scale, where 1.0 represents a perfect pavement and 0.5 represents the trigger level. CCI uses the same scale, but in this case 1.0 represents the highest level of construction consistency and 0.0 the lowest.

Since pavements deteriorate over time, a section that has been in service for 10 years, for example, cannot be directly compared with one that has been in service for only five. The 60 RAP sections considered in this study ranged in age from five to nine years. To overcome this issue, Stantec used normalized models to estimate each section’s SAI, RI and DI values at age five, giving a valid base for performance comparisons. Since construction consistency is not expected to change with time, no age adjustment was applied to the CCI data.

Different region, different language

Stantec evaluated the performance of the RAP sections in a two-stage process. Firstly, each section’s SAI, RI, DI and CCI values were compared with those of a typical asphalt pavement section and classified as excellent, good, fair or poor, depending on how well they matched this standard. Secondly, the expected service lives were calculated for each section based on SAI, RI and DI as the age at which the index would reach the trigger level of 0.5. The overall service life of the pavement section was assumed to be the minimum of the SAI, RI and DI service lives.

In the first round of analysis, Stantec evaluated how RAP performed in different environmental zones. A summary of the results is provided below:

The RAP sections located in the North Coast zone showed the best overall performance. The zone’s average SAI, DI and RI values were all considered excellent, with at least 80% of the individual sections showing excellent performance for each index;
The North Coast RAP sections showed the highest degree of variability in construction—at poor or fair levels for all sections;
The expected service lives for the RAP sections in the North Coast zone are 18, 21 and 17 years based on SAI, DI and RI, respectively. Therefore, it is anticipated that the North Coast RAP sections will be triggered for rehabilitation due to ride quality after 17 years;
The average SAI and RI values for the Desert zone sections indicated that RAP was performing at a good standard in terms of structural capacity and ride quality—at least 95% of the sections were ranked as good or excellent in both cases. However, in terms of DI, the performance of 90% of the sections was considered poor;
The RAP sections in the Desert zone showed the greatest construction consistency, with more than half of the sections considered to be good or excellent in terms of CCI;
The expected structural, distress and roughness service lives in the Desert zone were 15, nine and 15 years, respectively. Therefore, it is expected that the RAP sections would first be triggered for rehabilitation due to distresses after nine years;
In the Mountain zone, the RAP sections showed good structural and ride-quality performance overall. Distress performance was somewhat lower—more than half of the sections were considered fair, and the remainder poor, in terms of DI;
Construction consistency in the Mountain zone sections was considered fair, on average; and
The expected service lives for the RAP sections in the Mountain zone based on structural, distress and roughness performance are 11, 13 and 15 years, respectively. Therefore, it is expected that these sections will be triggered for rehabilitation due to structural condition after 11 years.

Going for seconds

In the second round of analysis, RAP performance was compared with that of other materials subject to the same environmental and traffic conditions. To make this comparison possible, Stantec selected test sections for the other materials that were within reasonable distance of the RAP sections, on the same route, with no major intersections in between. The following are the key findings of this comparison:

The North Coast RAP sections were compared with two groups of sections with similar AC overlay (referenced as AC1 and AC2). The distress performance was excellent for all materials. RAP and AC2 also showed excellent performance in terms of SAI and RI, while AC1 performance was considered good. All materials showed poor levels of construction consistency;
In the North Coast zone, the RAP sections would be expected to last the longest before needing rehabilitation—17 years compared with 15 years for
In the Desert zone, RAP was compared with RAC overlay (RAC) and with mill, AC overlay and RAC (MAR). RAP showed the best performance in terms of RI, but the worst in terms of DI. However, DI values were fairly low for all materials. SAI values for RAP were considered good, putting it above MAR, but below RAC in terms of structural performance. RAP showed the greatest variability in construction;
RAP had the shortest overall service life of the materials considered in the Desert zone—nine years compared with 11 for RAC and MAR;
In the Mountain zone, three materials were considered in addition to RAP: AC overlay (AC), RAC overlay (RAC) and mill and AC overlay (MA). Structural and ride-quality performance were considered good for all materials. RAP showed the worst performance in terms of DI, which was classified as poor. However, only MA showed good distress performance, while that of the other two materials was considered only fair. CCI values for all materials fell within the poor or fair categories; and
RAP sections in the Mountain zone were expected to be triggered for rehabilitation at 11 years, as were the RAC sections. For AC and MA sections, the trigger would be at 13 years.

Actual results

Most previous studies on RAP performance have approached the subject from a laboratory and materials point of view. With its focus on in situ field conditions, this study represents a step toward increased knowledge of actual, rather than laboratory-predicted, RAP performance. Based on the sections considered in this study, the long-term performance of RAP sections in California’s North Coast zone is expected to be better than those in the Desert or Mountain zones. However, within each of these three zones, RAP seems to be performing in a comparable way to other materials subject to the same environmental and traffic conditions.

It is hoped that other field-based studies will follow to create a more comprehensive understanding of the long-term performance, and therefore cost effectiveness, of recycled asphalt.

About The Author: Zaghloul is formerly of Stantec Consulting. He now works for H.W. Lochner Inc. and can be reached at [email protected]. Holland is with Caltrans and can be reached at [email protected].