The largest solid waste stream in the United States is not household garbage and it is not office demolition. By weight it is mostly pavement and the structures that carry it.

In the federal analysis that underpins the numbers, construction and demolition debris generation in 2015 came to 547.8 million tons, of which concrete was 69.7 percent and asphalt concrete another 15.3 percent. Wood products accounted for 7.1 percent and everything else, drywall through steel, for the remainder. Split by where it came from rather than what it was made of, roads and bridges produced 242.3 million tons against 169.2 million from buildings and 135.0 million from other structures. EPA states the comparison plainly: roads and bridges contributed significantly more to generation in 2015 than buildings and other structures did.

Two materials fill the entire roads and bridges column, concrete at 158.4 million tons and asphalt concrete at 83.9 million. The column is short for a reason worth knowing before quoting the total.

What the 242 million tons leaves out

The estimate covers debris from constructing, renovating and demolishing buildings, roads and bridges and other structures. It excludes debris generated by land-clearing activity and by natural disasters, and the footnote explaining the first exclusion is candid about method rather than importance: the top-down approach depends on tabulated material consumption and typical material lifespans, which cannot see land clearing, and no separate methodology was developed because the management options for that material are decentralized and untracked, including management at the point of generation.

Anyone who has watched a rural alignment get built knows what falls into that gap. Clearing and grubbing produces stumps, root mats, brush and stripped topsoil, and where any of it goes is set by project specifications and state permits rather than by anything the national accounting describes. None of it appears in these figures at all.

The second omission is more awkward, because it sits inside the numbers rather than outside them. Steel consumption in buildings, a footnote records, also includes steel consumed for the construction of roads and bridges: data were not available to allocate steel across sources, so buildings were assumed to take the largest share. Every guardrail, plate girder, reinforcing bar and beam that left a highway project that year is counted in another column. The roads and bridges figure is a floor.

On a road job the waste arrives before the work does

Demolition, not construction, dominates the stream. Demolition debris comprised over 90 percent of total generation for every material except drywall and plasters, the one category where waste from new construction accounts for more than a tenth of the total.

Applied to a highway contract, that statistic has a scheduling consequence that gets missed. Millings, broken slabs, removed base and demolished structures show up at the front of the job, before the new pavement has consumed anything at all. The decision about whether that material becomes base course on this project or a truck ticket to a landfill has to be made during design, because by the time the milling machine is running the stockpile either has somewhere to go or it does not. Where the material ends up also shifts the project’s own carbon arithmetic, which is where waste accounting and embodied carbon accounting stop being separate exercises.

Diversion is high and the destination is the bottom rung

EPA’s public tables follow the material to its next use. In 2018, out of 600 million tons generated, 313.07 million tons went to aggregate, 131.59 million into manufactured products, 7.54 million to fuel, 2.46 million to compost and mulch, 1.89 million to soil amendment, and 143.78 million to landfill.

Work the same tables across the two reported years and the encouraging headline flattens. Landfilled debris was 132.19 million tons of 547.0 million generated in 2015, and 143.78 million tons of 600.3 million in 2018. The 2015 total is EPA’s slightly revised public-tables figure; the 2018 generation memo had put it at 547.8 million. The landfill share moved from roughly 24.2 percent to roughly 24.0 percent. Generation rose about 10 percent over three years. The proportion diverted did not meaningfully change.

Composition matters as much as the share. Aggregate accounts for about 69 percent of everything not landfilled, and aggregate is the lowest rung crushed concrete and pavement can land on. Material placed as unbound fill or base has avoided a landfill and displaced some quarried stone, which is a real gain, but its binder, cement paste and engineered gradation are no longer doing structural work. The distance between avoiding disposal and returning material to the layer it came from is the whole argument in how recycled asphalt is used in road projects, where the same tonnage supports two very different recycling rates depending on the question asked.

Concrete back into concrete is the unsolved half

Federal review of the concrete side found the pattern in the field as well as in the tables. After in-depth reviews of the programs in Minnesota, Utah, Virginia, Texas and Michigan, the national review reported that concrete is routinely being recycled into the highway environment in the United States, that its principal application has been as a base material, and that its use as aggregate in hot mix asphalt and in concrete is not as widely accepted.

The review’s stated purpose is the giveaway. It set out to capture the most advanced uses of recycled concrete aggregate for technical deployment and transfer that knowledge to all state transportation agencies. That is the language of an agency circulating existing practice, not of one waiting on a laboratory result. Recycling concrete is solved. Recycling concrete into concrete is a question of who is willing to specify it, which is also the live question in asphalt against concrete.

The barrier federal policy names is a sentence in a specification book

FHWA’s recycled materials policy has been in place for over a decade, and it reads as instructions to whoever writes specifications. Recycling and reuse can offer engineering, economic and environmental benefits. Recycled materials should get first consideration in materials selection. The determination of use should include an initial review of engineering and environmental suitability, with an assessment of economic benefits following in the selection process. Then the clause that names the obstacle: restrictions that prohibit the use of recycled materials without technical basis should be removed from specifications.

Read that in reverse and it is an admission. The barrier the federal agency chose to write policy against was not an unproven material or a missing technology. It was language already sitting in state books, prohibiting materials for reasons nobody could document. Policy statements do not delete specification clauses, and the reuse decisions that follow are made by the same field organizations covered across the construction section.

What is weighed and what is modeled

None of the tonnages above were weighed. EPA derived them from a materials flow analysis, tabulating historical consumption data and assuming typical lifespans to estimate when each material reaches end of life. Concrete debris in particular is estimated from cement consumption, split across buildings, roads and bridges and other structures using a Portland Cement Association estimate that in 2002, 47 percent of portland cement went to buildings, 33 percent to roads and bridges and 20 percent to other structures, then adjusted year by year with Census data on the value of construction put in place.

EPA names its own soft spots. Two inputs need updating in future estimates, the roofing shingle sales data from 2006 and the cement end-use estimate from 2002, and further research is needed to determine how steel debris divides across the three sources.

There is one exception in the same report, and it points at the fix. Asphalt concrete is no longer modeled from lifespans at all. It is estimated from the quantity accepted by asphalt mix producers plus the amount handled at state-permitted solid waste management facilities, figures that come from an industry survey and from state reporting, which is why no assumption about how long a pavement lasts is required for that line. The material with the best national numbers turns out to be the one that gets weighed on its way through a permitted gate. Everything else in the road waste stream is an inference from what was bought thirty years ago.