Industries / Construction and property / Heavy civil / Asphalt production
Recycled asphalt
Recycled asphalt is old pavement that has been milled off or dug up and crushed, called reclaimed asphalt pavement (RAP), along with ground roofing shingles (RAS), fed back into the plant so that their stone and hardened binder replace part of the new materials.
7
Milestones achieved, from 1976 to 2025.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1976
Federal recycling trials begin
Old asphalt pavement had been reused in some form since about 1915, but recycling as it is done today began in the mid-1970s, when inflation and OPEC's oil embargo drove up the price of binder. In June 1976 the US Federal Highway Administration (FHWA) started Demonstration Project 39, which helped pay for about 50 trial projects in hot, cold and surface recycling and led to the first mix design and construction guidelines. On the strength of that work, FHWA expected RAP to make up about 15% of hot mix production by the mid-1980s.
Credit
Federal Highway Administration
1996
Recycled mix matches new
From 1992 FHWA visited 17 state highway agencies to see how recycled hot mix had performed. Its 1996 report found that about 33% of the 45 million tonnes of RAP removed each year went back into hot mix, that many engineers still believed recycled mix was inferior, and that recycled mix designed and controlled during production performed comparably to new mix over as long as 17 years. Washington State, which built its first two recycled hot mix projects on Interstate 90 in 1977 and 1978, supplied much of the long-term record.
LessonEngineers' doubts, written into specification limits, held back recycling more than the material's performance did.
Credit
FHWA; Washington State Department of Transportation
2001
Superpave gets RAP rules
Superpave, the US mix design method that state agencies adopted in the late 1990s, came with no guidance for RAP and favored coarse mixes that left little room for the fine material in RAP stockpiles. Many states stopped allowing large amounts of RAP and switched to new materials to reduce variability. NCHRP Project 9-12 then produced guidelines, published in 2001 by R. McDaniel and R. M. Anderson, with three tiers: no change of binder grade below 15% RAP, a binder one grade softer from 15% to 25%, and blending charts above 25%.
LessonA new design method with no rules for recycled material can cut recycling until the rules are written.
Credit
National Cooperative Highway Research Program; FHWA's Superpave Mixtures Expert Task Group
2011
Thirty percent RAP holds up
The National Center for Asphalt Technology (NCAT) compared paired sections of new mix and mix with 30% RAP on 18 projects across North America, 6 to 17 years old, in the Long-Term Pavement Performance program. The sections did not differ statistically except for three kinds of cracking, where the new mix did slightly better overall, and NCAT concluded that in most cases 30% RAP gives the same overall performance. FHWA's 2011 review cites the result and notes that some states allowed more RAP once it was fractionated, that is, screened into coarse and fine sizes that can be fed separately.
Credit
NCAT; FHWA
2014
Shingle use peaks
Recycled asphalt shingles come from factory scrap or roof tear-offs, and their binder is much stiffer and more brittle than paving binder. US mix producers used about 1.96 million tons of RAS in 2014; by 2024 they used 641,000 tons, and the number of states where producers reported using it fell from 38 in 2013 to 20. In FHWA's accelerated loading tests at its Turner-Fairbank research center, mix with 6% RAS, which replaced 20% of the binder, lasted less than 15% as long in fatigue as a new-mix control and about half as long as mix with RAP, and the Texas Department of Transportation cut its use after reports of poor performance.
LessonOld binder can count toward a mix's binder content on paper and still leave the mix brittle.
Credit
FHWA; National Asphalt Pavement Association (NAPA)
2024
Europe's reuse rates diverge
The European Asphalt Pavement Association (EAPA) compiles national figures from its member associations. For 2024 it reports that Germany put 91% of its 12.9 million tonnes of available reclaimed asphalt back into new asphalt, Denmark 97% and France 86%, while Turkiye reused 1% and sent 99% to lower uses such as unbound road layers. Across the 16 countries with full data, 73% of the 40 million tonnes available went back into new mix.
Credit
EAPA and its national member associations
2025
Stockpiles tracked for 18 months
Cassie Castorena, Shane Underwood and colleagues at North Carolina State University sampled RAP and RAS stockpiles and the mixes made from them at seven plants of five contractors for about 18 months. RAP binder content ranged from 4.3% to 5.4%, and the plant that crushed its RAP every month with its own jaw crusher showed the most variable binder content, while plants that hired out impact crushing less often were more stable. Their analysis showed that lowering North Carolina's limit on binder replaced by recycled binder from 30% to 20% for PG 64-22 mixes raised the chance of meeting the intended binder grade from 75% to 95%.
Credit
North Carolina State University, for the North Carolina Department of Transportation (NCDOT)
Watch for
All-recycled mix in New York
Green Asphalt, which CONEXPO-CON/AGG News calls the first US plant making mix from 100% RAP, says it has placed more than 2 million tons of 100% RAP mix on New York City streets, and that in 2024 Heidelberg Materials joined it to pave 50,000 to 100,000 tons of municipal work in upstate New York.
Achieved whena road owner or independent laboratory publishes condition data over several years comparing streets paved with 100% RAP mix against conventional mix.
Claimed by
Green Asphalt, as reported by CONEXPO-CON/AGG News, July 2026
Open
01
Extraction-free grading, tested blind
A blind comparison of the ways to estimate how stiff the old binder in a RAP stockpile is without dissolving it out with solvents. Candidates include Auburn University's method, which reads the binder grade from cracking and strength tests on compacted RAP (the IDEAL-CT and a high-temperature indirect tension test), and infrared spectra (FTIR) of binder taken from the particles, which measure how far it has oxidized. The reference for every stockpile is the usual solvent extraction, recovery and binder grading, done by an accredited laboratory and kept sealed until the estimates are in.
The briefCredit · unclaimed02
Crusher type and consistency
A public analysis, from state agencies' quality records, of how much RAP binder content and gradation vary over time at plants that process RAP in different ways: jaw or impact crusher, in-house or hired crushing, fractionated or not.
The briefCredit · unclaimed
Open 01 · the brief
Extraction-free grading, tested blind
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A blind comparison of the ways to estimate how stiff the old binder in a RAP stockpile is without dissolving it out with solvents. Candidates include Auburn University's method, which reads the binder grade from cracking and strength tests on compacted RAP (the IDEAL-CT and a high-temperature indirect tension test), and infrared spectra (FTIR) of binder taken from the particles, which measure how far it has oxidized. The reference for every stockpile is the usual solvent extraction, recovery and binder grading, done by an accredited laboratory and kept sealed until the estimates are in.
Why it is still open
The 2025 North Carolina State study for NCDOT says solvent extraction and recovery "are not practical for routine use" and calls for practical ways to measure RAP and RAS binder properties. A 2025 Auburn master's thesis by Liz Valenca proposed an extraction-free framework, but it gave reliable estimates for two of three field RAP sources and has not been tested blind across many plants. NCAT, which ran the work with the Virginia Transportation Research Council as an NCHRP IDEA project, reported errors of 0 to 7 °C for two sources and 8 to 13 °C for the third, and plans further refinement and a draft AASHTO test method.
Built with what exists
IDEAL-CT (ASTM D8225) and indirect tension test frames, found in state and university asphalt laboratories; a dynamic shear rheometer and a bending beam rheometer for the reference grades; a handheld FTIR spectrometer such as the Agilent 4300; Python with scikit-learn.
First step
Ask one state materials laboratory or university asphalt laboratory for split samples of five stockpiles whose extracted grades it has already measured, and run IDEAL-CT tests on them.
Done looks like
Split samples from at least 20 RAP stockpiles at no fewer than five plants, each estimated by every method before the reference grades are opened, and a published table of each method's error in continuous high and low temperature grade, in °C, with the share of stockpiles each method places within 6 °C (one grade step) of the reference.
Open 02 · the brief
Crusher type and consistency
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public analysis, from state agencies' quality records, of how much RAP binder content and gradation vary over time at plants that process RAP in different ways: jaw or impact crusher, in-house or hired crushing, fractionated or not.
Why it is still open
In the North Carolina study, the plant that crushed in-house with a jaw crusher every month had the most variable RAP, but the study covered only seven plants and recommends investigating how crusher type affects consistency. US producers held about 134 million tons of RAP in stockpiles at the end of 2024, and fractionated RAP was about 19% of RAP use, by the NAPA and FHWA survey.
Built with what exists
Public-records requests for state agencies' RAP stockpile test results (binder content by ignition oven, AASHTO T 308, and sieve gradations); a short survey of producers for crushing method; an AI model that reads scanned test reports, such as Claude; Python with pandas and statsmodels.
First step
Request one state's RAP stockpile test records for the last three years and list which plants' records name a crushing method.
Done looks like
At least 50 plants in at least three states, each with at least ten stockpile tests and a recorded processing method, published with the standard deviation of RAP binder content by method and a test of whether the difference holds after allowing for state and RAP source.