Concrete won the riding surface a long time ago and the inventory records it plainly. FHWA’s Highway Bridges by Deck Structure Type table for 2016 counted 364,255 bridges with cast-in-place concrete decks and 61,416 with precast concrete panels, against 37,994 in wood or timber, 10,868 in corrugated steel, 2,737 with open grating, 1,727 with steel plate including orthotropic decks and 137 in aluminum. A further 120,974 carried the code for not applicable, which the inventory uses where there is no distinct deck element to classify. Steel remains the material of choice for long spans and competes with prestressed concrete for the framing beneath those decks. What it has largely lost is the surface.
Each material commits its owner to a different set of obligations that plays out over the following century, a fact that matters more than any ranking of which one is better. How long that period is assumed to run is itself a specification decision rather than an observed outcome, as how long bridges are designed to last sets out. On the question of what those obligations cost, the federal record is unusually specific about one material in particular.
Weathering steel is a maintenance contract wearing a material’s clothes
Uncoated weathering steel forms a stable oxide layer that protects the metal beneath it, which removes the recurring cost of painting. FHWA has published guidance on its proper use since Technical Advisory T 5140.22, issued 3 October 1989 and, according to NTSB’s account, updated in June 2017. The advisory is precise about where the mechanism does not work.
It names three environments to avoid. Marine coastal areas, where salt-laden air creates corrosion risk. Areas of frequent high rainfall, high humidity or persistent fog, described in the advisory as condensing conditions. Industrial areas where concentrated chemical fumes may drift directly onto the structure. It then names two locations warranting caution, and the second is quantified: grade separations in tunnel-like conditions that concentrate salt spray, and low level water crossings, meaning ten feet or less over stagnant sheltered water or eight feet or less over moving water.
The detailing guidance is equally concrete. Eliminate bridge joints where possible, and where a joint is unavoidable, paint all superstructure steel within a distance of one and a half times the depth of girder from it. Seal overlapping surfaces exposed to water. Provide troughs beneath joints. And the maintenance section tells owners to control roadway drainage, regularly remove all dirt, debris and other deposits that trap moisture, and regularly remove all vegetation that can prevent the natural drying of wet steel surfaces.
Every one of those is an ongoing obligation. The material trades a painting cycle for a cleaning cycle, and the cleaning cycle is cheaper, less visible and, per the funding rules behind rehabilitation and replacement, outside federal-aid eligibility.
The scale of the exposure is large. NTSB’s 2023 investigation report states that more than 10,000 bridges in the United States have been built using uncoated weathering steel, a figure the Board attributes to a 2014 paper in Transportation Research Record. The bridge that prompted that report was one of them. The Forbes Avenue structure over Fern Hollow in Pittsburgh, an uncoated weathering steel rigid frame that NTSB records as 447 feet long and that FHWA’s 2023 memorandum records as 442 feet 8 inches, collapsed on 28 January 2022. A laser scan of the failed legs found web plate areas at roughly 25 percent of the nominal half-inch thickness, with holes. The Board found corrosion documented in every inspection report from 2005 onwards and clogged deck drains identified continually from 2011 to 2021.
FHWA’s response, a memorandum issued in July 2023, directed state departments of transportation to identify every bridge with uncoated weathering steel in a primary load path, sort them into two groups by condition rating, report the higher-priority group to FHWA by 31 October 2023, and by 31 December 2024 review maintenance records for that group, verify that recommended work had been completed, update load ratings to reflect documented deterioration, and conduct special inspections where needed. A material that was specified to reduce maintenance became the subject of a national maintenance audit.
Concrete’s adversary is chloride, and the code answers it with cover
For reinforced and prestressed concrete the governing deterioration mechanism is corrosion of the embedded steel, initiated when chloride reaches it. The Service Life Design for Bridges summary guide records that AASHTO LRFD does not quantify exposure to chloride ions from de-icing chemicals, sea or brackish water, or other sources. What it gives instead are three named exposure situations with a cover requirement attached to each. Direct exposure to salt water calls for 4.0 inches over the steel. Coastal exposure calls for 3.0. Exposure to de-icing salts calls for 2.5, and all three figures precede any adjustment for water-cement ratio.
Prestressing raises the stakes on the same mechanism rather than changing it. Federal inspection policy lists reinforcing and prestressing steel corrosion together as a deterioration mode that risk assessment panels must consider for concrete members, and the practical difference is that a prestressing strand is a highly stressed element in which section loss translates directly into lost capacity. The alternative route the SHRP2 guide describes is avoidance rather than resistance: use stainless steel of a grade appropriate to the exposure for reinforcement or structural steel components, so that corrosion is not permitted to initiate at all, and use non-reactive aggregate to avoid alkali-aggregate reaction. Avoidance costs more at the outset and removes a maintenance dependency, which is the same trade weathering steel appeared to offer and did not deliver. Where the binder chemistry itself is the variable, the argument runs through the future of low-carbon concrete.
UHPC is precisely specified and modestly deployed
Ultra-high performance concrete has a definition, and quoting it correctly matters because the term gets applied loosely. FHWA’s TechNote FHWA-HRT-11-038, published March 2011, defines UHPC as a cementitious composite material composed of an optimised gradation of granular constituents, a water-to-cementitious materials ratio less than 0.25, and a high percentage of discontinuous internal fiber reinforcement. Its mechanical properties include compressive strength greater than 21.7 ksi, equivalent to 150 MPa, and sustained postcracking tensile strength greater than 0.72 ksi, or 5 MPa. The sustained tensile capacity is the property that distinguishes it. Ordinary concrete is designed on the assumption that it carries no tension.
Deployment scale is where care is needed, because the published figures are dated. That same 2011 TechNote states that in the United States UHPC had been used in three prestressed concrete girder simple-span bridges, two in Iowa and one in Virginia. FHWA’s current UHPC for Bridges page describes the material as improving the strength, simplicity and durability of prefabricated bridge element connections on new bridges, and as a means of strengthening, repairing and preserving existing bridges through deck overlays, girder end repairs and link slabs. It records that UHPC has been an initiative in FHWA’s Every Day Counts program since 2015, and it lists case studies from 2008 to 2020 across states including Iowa, New York, Oregon, Utah, Ohio, Florida, Delaware, South Carolina, Minnesota, Pennsylvania, New Mexico, Georgia and Nebraska, with material from several more states in other formats.
What that page does not publish is a national total. FHWA’s 2019 TechNote on the same subject, FHWA-HRT-19-011, describes field-cast UHPC connections as having captured the attention of bridge owners, specifiers and contractors across the country, without a count. So the honest statement is that UHPC is in service in most states in connection and overlay applications rather than as a primary structural material, that the only firm federal count of full girder applications is three and dates from 2011, and that anyone needing a current figure should go to FHWA’s own deployment map rather than to a secondary summary. Field-cast connections between precast elements, not girders, are where the material has actually gone to scale.
Fiber composites are excellent per pound and insufficient per pound
The case for fiber reinforced polymer is corrosion immunity and weight. The 2019 FHWA study of bridge strengthening techniques tabulates the comparison. Mild steel has a tensile modulus of 30,000 ksi, tensile strength of 65 to 120 ksi and a density of 0.282 pounds per cubic inch. Unidirectional glass and epoxy composite at 60 percent fiber volume has a modulus of 4,000 ksi, tensile strength of 114 ksi and a density of 0.061. Unidirectional carbon and epoxy at 62 percent comes in at 19,700 ksi, 278 ksi and 0.057.
The report draws the conclusion and then immediately qualifies it, in its own words: “In general, FRP composites are superior to metals with respect to specific strength and specific modulus. Nevertheless, glass fiber composites have a lower specific modulus than that of both steel and aluminum.” Glass, the cheap and common variety, is the one that loses on stiffness per unit weight, and stiffness rather than strength governs most bridge serviceability limits. That is why FRP appears first as external strengthening bonded to existing members, where the parent structure supplies the stiffness, rather than as a replacement for girders.
For decks the record is candid. FHWA’s Composite Bridge Decking final project report states that more than 100 fiber-reinforced polymer composite decks or structures are in service, that widespread use of these systems has not been attained for a variety of reasons, and that most have been proprietary in nature, which is generally not attractive to bridge owners. The proprietary problem sits in procurement, not materials science: an owner who cannot specify a replacement panel from more than one supplier has acquired a maintenance liability, whatever the panel is made of. Freeze-thaw exposure, which is what drives much of the deck deterioration these products are pitched against, is examined in freeze-thaw cycles and why roads break up in spring.
What actually decides it
Span length sets the shortlist. Environment eliminates candidates from it, and the weathering steel advisory is the clearest published example of environment doing exactly that. Cost then picks from what survives. That sequence, rather than any material’s headline properties, accounts for most of what bridge engineering coverage finds in the built record.
Federal inspection policy has since added a fourth term, and it is quiet but real. To qualify a bridge for a 48-month routine inspection interval under the simplified risk method, its superstructure materials must be limited to concrete and steel, and its span types to an enumerated set of arch, box, frame, girder, slab and culvert forms. Ordinary concrete and steel in ordinary configurations earn the cheapest inspection regime available. Anything more inventive does not, whatever its durability case. That is a small and permanent thumb on the scale, and it points where the inventory already went.