The federal highway agency keeps a written procedure for the case where its own analysis cannot pick a side. Technical Advisory T 5040.39, issued on December 20, 2012, explains how a state or local agency may put two pavement designs out to bid against each other and let the contract market decide which one gets built. The condition FHWA attaches is worth reading closely: alternate pavement type bidding is treated as a suitable approach for determining pavement type when engineering and economic analysis does not indicate a clear choice between different pavement designs.
An agency does not write a procedure for the indeterminate case unless that case is common. The interesting work is therefore not deciding which material wins. It is identifying the project conditions under which the answer stops being a coin toss.
The federal requirement runs to a single sentence
Title 23 of the Code of Federal Regulations, Part 626, asks of a pavement design only that it accommodate current and predicted traffic needs safely, durably and at reasonable cost. FHWA’s commentary on its own rule adds the part that matters: the regulations do not specify the procedures to be followed to meet this requirement. A state highway agency may work from the AASHTO Guide for Design of Pavement Structures, or from any other procedure that, on the basis of past performance or research, is expected to produce satisfactory pavement designs. FHWA reviews state design processes periodically rather than approving individual projects.
Nothing there expresses a federal preference between flexible and rigid pavement. The choice is delegated to fifty agencies with different climates, quarry and refinery geography, contractor bases and maintenance budgets, which is why national arguments about the two materials produce heat without resolving anything. A defensible answer exists only at the scale of a corridor.
What the money argument can settle, and where it goes quiet
The economic comparison runs on lifecycle cost analysis, and FHWA’s 1998 interim technical bulletin still sets the shape of the exercise. Its minimum analysis window, taken from the agency’s September 1996 policy statement, is 35 years, applied to new construction and rehabilitation alike, with anything from 30 to 40 years treated as defensible. That floor is not arbitrary: the window has to be long enough to contain at least one rehabilitation activity, and generally longer than the design period itself. The method behind the resulting present values is treated in highway lifecycle cost analysis.
That bulletin’s findings cut against arguments both industries like to make, starting with salvage value. Residual value is the net value recovered by recycling the pavement, and the bulletin states plainly that the differential residual value between design strategies is generally not very large and, discounted over 35 years, tends to have little effect on results. Serviceable life, the remaining life left in an alternative when the analysis window closes, is the more significant component by a wide margin. Recyclability is not where the money is. Remaining service life is.
Performance periods cut the same way, identified as having a significant effect because longer performance periods require fewer rehabilitation projects, and with them fewer agency costs and fewer work zone user costs. The alternate bidding advisory carries the same logic into procurement, directing that projects be carefully evaluated where anticipated work zone user delay costs differ by more than 20 percent between alternatives, since at that spread the alternatives are no longer equivalent in what they cost the public.
Climate selects the failure mode rather than the material
FHWA’s March 2023 state of the practice document on pavement resilience tabulates climate stressors against pavement vulnerabilities and adaptations, and the table is unexpectedly symmetrical. Higher average temperatures raise maximum pavement temperature and accelerate age hardening in asphalt binder, and they also increase temperature-related curling in concrete along with moisture-related warping where relative humidity falls. Higher extreme maximum temperatures add a distinctly rigid failure mode, concrete blow-ups caused by excessive slab expansion. More freeze-thaw events load the asphalt side toward thermal cracking resistance and the concrete side toward resistance to freeze-thaw cycling and chemical deicers. Lower summer humidity increases asphalt binder aging and long-term slab warping at the same time.
Every one of those vulnerabilities has a documented adaptation on both sides of the ledger: raising the high-temperature binder grade or adding polymer to the surface course, specifying concrete with lower drying shrinkage and a reduced coefficient of thermal expansion, shortening joint spacing, using smaller slabs with enhanced load transfer. What the table does not contain is a climate that rules a material out. It contains climates that make particular specification clauses non-negotiable and particular quality control failures expensive. An agency in a severe freeze-thaw environment can build either pavement well, and can also build either one badly in a way that surfaces in year eight.
Traffic class is where the divergence becomes real
The clearest published dividing line is a traffic threshold, not a climate one. FHWA’s TechBrief on long-life concrete pavements records that Caltrans requires a 40-year service life for pavement design along a corridor where the 20-year projected average annual daily traffic reaches 150,000 vehicles or the average annual daily truck traffic reaches 15,000 trucks. Above that volume, the calculation that matters is no longer the cost of the pavement. It is the cost of closing it, repeatedly, in an urban corridor where lane closures are acute.
State practice tracks the same logic. Minnesota adopted long-life concrete pavement as its standard design for high-volume urban highways in 2000, and 38 percent of Washington State concrete pavements were more than 35 years old as of 2006, having outlasted a 20-year design life with little or no maintenance. That record belongs in the argument.
So does the caveat attached to it in the same document. Minnesota’s own reservations, recorded alongside its standards, are that current materials specifications rely on accelerated laboratory testing whose results may differ significantly from long-term field performance, and that it remains an open question whether the permeability and air content systems specified for long-life pavement will deliver durability across the entire design period under the state’s extreme climate. Contractors unfamiliar with the supplementary cementitious material mixtures those specifications require reported difficulty handling them. A 60-year design life is a claim about materials nobody has yet observed for 60 years. The reasoning behind such designs, on both the rigid and the flexible side, is examined in long-life pavement design.
Recycling is asymmetric, and it is not the tiebreaker
On recycling the two materials are genuinely not alike, and the difference is what the recovered material becomes. The asphalt industry’s 2023 survey, conducted under a federal agreement, puts national use of reclaimed asphalt pavement in new mixtures at 96.1 million tons for the season, with more than 89 percent of what is milled up returning to new asphalt pavements and the balance going to unbound aggregate bases and similar applications. The material comes back to the same product at the same point in the structure, which is unusual in construction. Concrete’s path is a step down, and the U.S. Geological Survey states it without softening: cement is not recycled, though significant quantities of concrete are recycled for use as a construction aggregate. Crushed concrete is a genuine and widely used aggregate, and the clinker that made it is gone. The specification limits governing how much reclaimed material a mixture may carry are covered in recycled asphalt practice.
The temptation is to score that asymmetry as decisive, and the lifecycle bulletin has already explained why it is not. Discounted residual value barely moves a 35-year present value. The recycling difference is an argument about material supply and embodied emissions, which is where it belongs, and embodied carbon accounting is where it can be settled with numbers.
The maintenance profile is the part an agency lives with
Each material presents its owner with a different bill, in different increments, requiring different equipment. The threshold values FHWA lists for concrete distress describe what the eventual work looks like: cracked slabs at 10 to 15 percent of the total, joint faulting around 6 to 7 millimeters, roughness at 150 to 180 inches per mile, punchouts at 16 to 20 per mile on continuously reinforced pavement. Long-life practice assumes intervention will be minimal and specific, limited to ride and texture work, joint resealing and minor repairs. Illinois requires diamond grinding where the measured profile index exceeds 48 inches per mile on major highways, and Washington State designs the top inch of its slab as sacrificial so it can be ground away later.
Asphalt maintenance arrives in a different rhythm. FHWA’s resilience document lists asphalt preservation techniques that slow binder aging in the surface course, fog seals and chip seals among them, as an operations-level adaptation to a drying climate. These are cheap, frequent, and need a contractor already working in the region. The alternate bidding advisory sets the standard both approaches must hold across the performance period: good condition below 95 inches per mile of roughness, fair condition below 170.
Those are not interchangeable institutional capabilities. Grinding and joint work need specialist equipment and a schedule. Overlays and seals need a paving contractor and a lane closure. An agency with a thin concrete paving market and a deep asphalt one faces a real constraint that no lifecycle model contains, and the reverse holds equally. What an agency can reliably let and inspect each year is a legitimate input to pavement type selection, and it recurs across the materials and pavements section.
What the alternate bid is actually buying
The advisory’s central requirement is that the alternates be equivalent to the maximum extent possible, providing a similar level of service over the same performance period at similar lifecycle costs. Meet that honestly and the two designs become indistinguishable in every respect the travelling public can perceive: the same ride, the same years, the same number of closures, the same present value. What differs is the bid price, a function of local material and haul economics in the month of the letting.
So alternate bidding is not an evasion of the engineering question. It is a claim that on many projects the question has already been answered twice, correctly, in two materials, and that arguing further spends money on analysis instead of pavement. The projects where it stays live sit at the edges: the corridor carrying 15,000 trucks a day, the freeze-thaw environment with a marginal aggregate source, the agency that cannot staff a grinding program. Those deserve the full analysis. The rest deserve two good designs and an open letting.