Industries / Construction and property / Heavy civil / Asphalt production
Mixing plants
A mixing plant dries and heats crushed stone and sand, then coats them with hot liquid binder in measured amounts, turning out a few hundred tons of mix an hour for trucks to haul away.
5
Milestones achieved, from 1892 to 2021.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1892
A mixing plant on wheels
Amos H. Perkins of Chicago patented an "asphalt-paving apparatus": drums that dried sand, mounted on a car body, with conveyors, a measuring hopper and a mixer below it driven from a shaft on the car. His patent explains that sand must be "perfectly dry" for asphalt to stick to it, and the finished mix dropped from the raised mixer into a wagon parked beneath. The same steps, drying, weighing and mixing, still define a batch plant in the US Environmental Protection Agency's (EPA's) description of one.
Credit
Amos H. Perkins, for The Western Paving and Supply Company
1970s
The drum mixer returns
In a drum mix plant one long rotating drum dries the stone and then coats it with binder in a continuous stream, with no hot screens or weigh hoppers. Roads & Bridges magazine reports that drum mixing was first tried around 1910 and had mostly disappeared by the 1930s, when batch plants could make up to 1,000 tons in an eight-hour day, and that it came back from the 1960s. By the account of the equipment consultant Fred Prill, the continuous plants of the early 1970s made 300 to 500 tons an hour, against 150 to 300 for batch plants.
Credit
Boeing, which Prill names as one of the first makers of high-output drum plants
1986
Binder kept out of flames
The first high-output drum plants were parallel-flow designs: stone and burner gases moved the same way down the drum, so liquid asphalt and recycled asphalt met the hot gases and gave off "blue smoke", a haze of hydrocarbons that environmental laws limit. In a counterflow drum the stone moves against the gas, and binder and recycled asphalt are added behind the burner, in a zone shielded from its heat. Standard Havens filed for a counterflow drum mixer in June 1986, and Robert Farnham and Fred Prill filed for one for Barber-Greene in April 1987. EPA wrote that counterflow plants will "likely" give off fewer organic compounds, but that its stack data could not show the difference, so it gives both designs the same emission factors.
Lessona design adopted to cut smoke still lacks the stack data to show how much it cut, by EPA's account.
Credit
Standard Havens; Robert E. Farnham and Fredric W. Prill, Barber-Greene
1996
A census of US plants
EPA estimated that about 3,600 asphalt plants were active in the United States in 1996 and made about 500 million tons of hot mix. Of these plants 2,300 were batch plants, 1,000 parallel-flow drum plants and 300 counterflow drum plants; batch plants made about 240 million tons and drum plants about 260 million. EPA added that about 85% of the plants then being built were counterflow drums, that nearly all new plants could take recycled asphalt, and that 70 to 90% of the mix was made with natural gas as burner fuel.
Credit
US Environmental Protection Agency, AP-42 emission factors, Section 11.1
2021
Heated tanks keep exploding
Plants and terminals keep liquid asphalt hot in vented tanks. Reporters from Inside Climate News, working with Berkeley Journalism's Investigative Reporting Program, counted at least 17 explosions in a decade of heated tanks holding asphalt or No. 6 fuel oil, a heavy fuel oil, at US asphalt plants or terminals, from news reports and filings with OSHA, the federal workplace safety agency. One, at U.S. Polyco in Ennis, Texas, on 23 April 2018, threw a foreman using a cutting torch nearby about 30 feet and left him badly burned. The reporters found that neither OSHA nor the US Chemical Safety Board had studied these explosions as a group.
Lessonhot binder tanks can hold flammable vapour even when the liquid is well below its flash point, as a 1997 study by Robert Zalosh quoted in the report found.
Credit
Sabrina Shankman and Julia Kane
Watch for
AI tuning of plants
Tarmac, a UK producer, says that in 2025 it sped up the rollout of "AI optimisation boxes" that read real-time production data and recommend adjustments to plant staff, along with digital burner controls and on-demand heating of binder tanks. It reports a 10% cut in energy costs at its Elstow plant, a saving of one litre of liquefied petroleum gas (LPG) per tonne at Snodland after an analogue burner control was replaced, and a projected saving of 9,000 t of carbon dioxide equivalent over two years into 2026.
Achieved whenan independent auditor or a published study compares energy per tonne before and after at the same plants, with mix type, output and aggregate moisture taken into account.
Claimed by
Tarmac, in a case study filed in June 2026
Open
01
A public map of plants
An open map of the world's asphalt mixing plants, each with its location, owner where public, type (batch or drum, fixed or portable) and any air permit number, added to OpenStreetMap or published as an open dataset.
The briefCredit · unclaimed02
Moisture and fuel, measured openly
A public dataset, from at least three working plants over a paving season, that pairs hourly aggregate moisture with burner fuel per ton, production rate and mix temperature, and a model fitted to it that predicts fuel use from moisture.
The briefCredit · unclaimed
Open 01 · the brief
A public map of plants
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open map of the world's asphalt mixing plants, each with its location, owner where public, type (batch or drum, fixed or portable) and any air permit number, added to OpenStreetMap or published as an open dataset.
Why it is still open
On 18 November 2026 OpenStreetMap held 281 objects tagged as asphalt plants worldwide, while EPA alone estimated about 3,600 active plants in the United States in 1996. EPA's split of US plants by type dates from that 1996 estimate.
Built with what exists
EPA's Facility Registry Service filtered to NAICS industry code 324121 (asphalt paving mixture and block manufacturing); state air permit lists; OpenStreetMap's Overpass API; free aerial and satellite images (USDA's NAIP photos in the US, ESA's Sentinel-2 elsewhere); an object detector such as Ultralytics YOLO, trained to spot a drum, silos and a baghouse; an AI model such as Claude to pull plant type and capacity out of permit documents.
First step
Download the Facility Registry Service records for NAICS 324121 in one state and check each one against aerial photos.
Done looks like
At least 5,000 plants mapped in at least 10 countries, a random sample of 300 checked by hand against imagery with at least 95% confirmed as working mixing plants, and the data published under an open licence.
Open 02 · the brief
Moisture and fuel, measured openly
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public dataset, from at least three working plants over a paving season, that pairs hourly aggregate moisture with burner fuel per ton, production rate and mix temperature, and a model fitted to it that predicts fuel use from moisture.
Why it is still open
The National Asphalt Pavement Association's (NAPA's) 2024 carbon reference says plants are designed for aggregate at 5% moisture, that driving off that water takes about half the plant's heat, and that cutting moisture from 5% to 4% saves about 10% of burner fuel, a figure it takes from a 2007 source (Young). EPA's emission-factor data show carbon dioxide from drum plant dryers ranging from 2.6 to 96 lb per ton of mix (1.3 to 48 kg per tonne) across 180 plants. Linus Dep and colleagues tested a belt moisture gauge against more than 1,000 oven-dried samples at one Wisconsin plant in 2015 and 2016, but report the gauge's accuracy, not fuel use. Ivica Androjić and Zlata Dolaček-Alduk published a neural network model to forecast energy use in hot mix production (2018), and Androjić and colleagues reported that lowering aggregate moisture cut energy use by 36.3% in winter and 38.5% in summer (2020); this research found no public dataset from their work or from any other plant.
Built with what exists
Oven-drying of daily stockpile and cold-feed samples (ASTM C566, the standard moisture test for aggregate); the plant's own fuel meter, belt scales and control-system logs; local weather records; Python with pandas and scikit-learn.
First step
Ask one plant, through a state asphalt pavement association, for a month of burner fuel and production logs, and oven-dry one sample from each stockpile each morning of that month.
Done looks like
At least 1,500 hourly records from three plants, published with a model that predicts fuel per ton within 5% on weeks it was not trained on, and a measured fuel saving per percentage point of moisture set beside the 10% rule of thumb.